commit ec2b45d1bf4c31c5420b48c213bf9ee1656adf9d Author: Ilia Mirkin Date: Mon Jan 28 23:25:58 2019 -0500 Bump version to 1.0.16 Signed-off-by: Ilia Mirkin commit 2376d1ebf2d9a96bc2ebf21d53a9f9841ce5c15b Author: Ilia Mirkin Date: Sat Jan 26 23:00:19 2019 -0500 update known chipsets list Signed-off-by: Ilia Mirkin Reviewed-by: Rhys Kidd commit ca13913aaf7e3995c3b49d3d80aad7e6cb33291a Author: Ilia Mirkin Date: Wed Jan 23 00:29:51 2019 -0500 Revert "wfb: Fix missing init function decls behind FB_ACCESS_WRAPPER" This reverts commit d7ba24fb6e4fce43adc3c73369c5d1035ae080a9. When built and run against Xorg 1.20.3, part way through init, I got an error about wfbPictureInit not existing despite not having wfb_enabled set. Signed-off-by: Ilia Mirkin commit fb2d4d27837ebf3ae2fdcf091495f52d2ad8c822 Author: Ilia Mirkin Date: Wed Jan 23 00:28:05 2019 -0500 present: rotated crtc's work fine The "pick best" logic takes rotation into account. However flipping a rotated CRTC can't work, so we disable that. Signed-off-by: Ilia Mirkin commit 92db605f0f5ac14957e38e59782ef3ed4808de05 Author: Rhys Kidd Date: Mon Jan 21 22:59:54 2019 -0500 Adapt to PixmapDirtyUpdateRec::src being a DrawablePtr Signed-off-by: Ilia Mirkin commit 8a90309401963aa43f89f91e204ec9cac5ade704 Author: Rhys Kidd Date: Mon Jan 21 00:52:23 2019 -0500 config: Suggest upstream to find macros in case they are missing Added upstream suggestion to error message in case xorg macros or xserver macros are missing. No actual package manager references in suggestion since these are 'user' and/or 'distribution' preferences. (Cherry picked from xf86-video-intel 90538ecfff9c324d61419f451d41d26ab026f232) Signed-off-by: Rhys Kidd commit fcddd2c8e6e4a738bf28028e6d4ad5be34c8b2ae Author: Rhys Kidd Date: Mon Jan 21 00:52:22 2019 -0500 config: Complete bug report URL for Xorg DDX nouveau driver Signed-off-by: Rhys Kidd commit a566c883b3663b685e467788b8b5de6d99ef9ae2 Author: Rhys Kidd Date: Mon Jan 21 00:52:21 2019 -0500 config: Set automake options consistently in one location Avoid use of two manners of setting automake configuration options, and consolidate them into AM_INIT_AUTOMAKE macro. Signed-off-by: Rhys Kidd commit 56e67e09925b0b5544884b279b9cc5552964fd25 Author: Rhys Kidd Date: Mon Jan 21 00:52:20 2019 -0500 config: remove unused sdkdir=$(pkg-config...) statement The sdkdir variable is not used, so remove the statement. (Copied from xf86-video-ati 36bd69affc996c92c40b7360a7fbaa1a3a46abfd) Signed-off-by: Rhys Kidd commit db51fcf52067e4412c96bb4c84d97ceac6c857a6 Author: Rhys Kidd Date: Mon Jan 21 00:52:19 2019 -0500 config: Remove unrequired AC_SUBST for LIBDRM_NOUVEAU and LIBUDEV These macros are called by the PKG_CHECK_MODULES macro, for all versions of pkg-config >= 0.24.0 (released May 23, 2010). Signed-off-by: Rhys Kidd commit f5250b1d46b6442550140b149341e06f7f27ae8d Author: Rhys Kidd Date: Mon Jan 21 00:52:18 2019 -0500 config: Add comments for main statements Signed-off-by: Rhys Kidd commit ed653f76b17ba4b252fe405e7bd39b36142dae05 Author: Rhys Kidd Date: Mon Jan 21 00:52:17 2019 -0500 config: Use LT_INIT option instead of AC_DISABLE_STATIC Since libtool 2.2 (released March 2008) the preferred initiation of libtool is with the LT_INIT() macro, instead of AC_PROG_LIBTOOL. As we are using LT_INIT anyway we might as well pass the disable-static option directly to it instead of using the AC_DISABLE_STATIC macro. See: - https://www.gnu.org/software/libtool/manual/html_node/LT_005fINIT.html - https://www.x.org/wiki/NewModuleGuidelines/ Signed-off-by: Rhys Kidd commit fe3fc188f17f000afae4c307ad5e54f1596fc6ff Author: Rhys Kidd Date: Mon Jan 21 00:52:16 2019 -0500 config: Utilize util-macros 1.8 for additional man page support Use MAN_SUBST now supplied in XORG_MANPAGE_SECTIONS The value of MAN_SUBST is the same for all X.Org packages. Signed-off-by: Rhys Kidd commit caaac38112fa9066089099655866e0e9f5e3a647 Author: Rhys Kidd Date: Mon Jan 21 00:52:15 2019 -0500 config: remove unrequired AC_HEADER_STDC Autoconf says: "This macro is obsolescent, as current systems have conforming header files. New programs need not use this macro". Was previously required to ensure the system has C header files conforming to ANSI C89 (ISO C90). Specifically, this macro checks for stdlib.h, stdarg.h, string.h, and float.h. This autoconf option was used to provide conditional fallback support via defined STDC_HEADERS. xf86-video-nouveau does not do so, so this macro is obsolete and unused, so let's drop it. Signed-off-by: Rhys Kidd commit 4cf6d02ecbd5d2201d66324ba33f9fcbed67092f Author: Rhys Kidd Date: Mon Jan 21 00:52:14 2019 -0500 config: Use XORG_DEFAULT_OPTIONS which replaces six XORG_* macros This macro expands into several macros and provide several build functions [0]. Refer to the module generated aclocal.m4 file as this is subject to change. At present, this includes: - XORG_CWARNFLAGS: platform adjusted compiler warning flags - XORG_STRICT_OPTION: additional warning flags - XORG_RELEASE_VERSION: defines variables for major, minor and patch level - XORG_CHANGELOG: contains a makefile rule which creates the ChangeLog file from Git - XORG_INSTALL: contains a makefile rule which provides the INSTALL file in the module root directory - XORG_MANPAGE_SECTIONS: platform adjusted man page section number The above macros invoke the following Autoconf macros: - AC_PROG_INSTALL - AC_PROG_CC_C99 - AC_PROG_SED - AC_CANONICAL_HOST [0] https://www.x.org/wiki/NewModuleGuidelines/ Signed-off-by: Rhys Kidd commit 945f95775a21cbcc721b579f7dc220f1474d1d25 Author: Rhys Kidd Date: Mon Jan 21 00:52:13 2019 -0500 config: libtoolize: put macros in AC_CONFIG_MACRO_DIR, 'm4' As suggested by one tool called by autoreconf. Necessitates change to: - configure.ac: Call AC_CONFIG_MACRO_DIR() - Makefile.am: Set ACLOCAL_AMFLAGS = -I m4 - Create placeholder m4 folders - Update gitignore files >From the autoconf documentation: "Starting from version 2.58, autoconf provide the macro AC_CONFIG_MACRO_DIR to declare where additional macro files are to be put and found. The argument passed to this macro is commonly m4." https://autotools.io/autoconf/macros.html Signed-off-by: Rhys Kidd commit 9b60c3bc74c69d607d44c864f60f5fda0b11cd48 Author: Rhys Kidd Date: Mon Jan 21 00:52:12 2019 -0500 gitignore: Update to mirror modern X.org drivers Avoids git showing autoconf/automake generated files (e.g. tarballs) that are local build artifacts and should not be committed. Cherry picked from xf86-video-intel as at: https://gitlab.freedesktop.org/xorg/driver/xf86-video-intel/blob/7c9dbc980b760e0053d83ca2d7cb147613285680/.gitignore Signed-off-by: Rhys Kidd commit 1761769d43ab3ab31b5e5d92e6014fdea21830bd Author: Rhys Kidd Date: Mon Jan 21 00:52:11 2019 -0500 config: remove AC_PROG_CC as it overrides AC_PROG_C_C99 XORG_STRICT_OPTION from XORG_DEFAULT_OPTIONS calls AC_PROG_C_C99. This sets gcc with -std=gnu99. If AC_PROG_CC macro is called afterwards, it resets CC to gcc. Signed-off-by: Rhys Kidd commit e27c0bb6c70b1a68ddf2aeeb0e4346e0a41b8d9c Author: Rhys Kidd Date: Mon Jan 21 00:52:10 2019 -0500 config: Rely upon list.h as at least xorg-server 1.8 required Fixes: edd1608 ("xorg_list: define the xorg_list* symbols") Signed-off-by: Rhys Kidd commit 6700548071763daf30ab74dbcfb6543ea0db101d Author: Rhys Kidd Date: Mon Jan 21 00:52:09 2019 -0500 config: Only check for presence of libdrm package once After a version-conditioned check for libdrm was added, no longer necessary to check a second time, especially without a version number check. Fixes: 779b8d4 ("Link against libdrm") Signed-off-by: Rhys Kidd commit 9b2b008eb76c6f4c31d5f71ca01a06254a03f7c9 Author: Rhys Kidd Date: Mon Jan 21 00:52:08 2019 -0500 config, nv: Use xorg-macros PACKAGE_VERSION_{MAJOR, MINOR, PATCHLEVEL} Use xorg-macros provided PACKAGE_VERSION_{MAJOR,MINOR,PATCHLEVEL} rather than rolling our own in xf86-video-nouveau. Macro has been available since xorg-macros 1.1.5 (we rely on >= 1.8) provided that XORG_DEFAULT_OPTIONS is called. Fixes: b82aacd ("configure: require xorg-macros 1.8") Signed-off-by: Rhys Kidd commit b098bdae321182d2f18d791e7f1c665dec21f571 Author: Rhys Kidd Date: Mon Jan 21 00:52:07 2019 -0500 config: Fix typo in comment Signed-off-by: Rhys Kidd commit 764b42babd38b545ccb3b43cd81c989bddf71c77 Author: Rhys Kidd Date: Sun Jan 20 22:21:34 2019 -0500 xv/nv30, nv40: Mark local NV30GetSurfaceFormat and NV40GetSurfaceFormat as static Avoids warnings with gcc 8.2: nv30_xv_tex.c:138:1: warning: no previous prototype for ‘NV30GetSurfaceFormat’ [-Wmissing-prototypes] NV30GetSurfaceFormat(PixmapPtr ppix, int *fmt_ret) ^~~~~~~~~~~~~~~~~~~~ nv40_xv_tex.c:138:1: warning: no previous prototype for ‘NV40GetSurfaceFormat’ [-Wmissing-prototypes] NV40GetSurfaceFormat(PixmapPtr ppix, int *fmt_ret) ^~~~~~~~~~~~~~~~~~~~ Signed-off-by: Rhys Kidd commit 0fa6ba423593cb5c8e43b8f6569f90f9fcc2b36c Author: Rhys Kidd Date: Sun Jan 20 22:21:33 2019 -0500 xv: Mark local NVSetupTexturedVideo function as static Avoids warning with gcc 8.2: nouveau_xv.c:2136:1: warning: no previous prototype for ‘NVSetupTexturedVideo’ [-Wmissing-prototypes] NVSetupTexturedVideo (ScreenPtr pScreen, XF86VideoAdaptorPtr *textureAdaptor) ^~~~~~~~~~~~~~~~~~~~ Signed-off-by: Rhys Kidd commit 2905ea76571faed91e7d17b5dc6e44758e71ff0b Author: Rhys Kidd Date: Sun Jan 20 22:21:32 2019 -0500 dri2: Mark local create/destroy buffer and copy region functions as static Avoids warnings with gcc 8.2: nouveau_dri2.c:38:1: warning: no previous prototype for ‘nouveau_dri2_create_buffer2’ [-Wmissing-prototypes] nouveau_dri2_create_buffer2(ScreenPtr pScreen, DrawablePtr pDraw, unsigned int attachment, ^~~~~~~~~~~~~~~~~~~~~~~~~~~ nouveau_dri2.c:109:1: warning: no previous prototype for ‘nouveau_dri2_create_buffer’ [-Wmissing-prototypes] nouveau_dri2_create_buffer(DrawablePtr pDraw, unsigned int attachment, ^~~~~~~~~~~~~~~~~~~~~~~~~~ nouveau_dri2.c:117:1: warning: no previous prototype for ‘nouveau_dri2_destroy_buffer2’ [-Wmissing-prototypes] nouveau_dri2_destroy_buffer2(ScreenPtr pScreen, DrawablePtr pDraw, DRI2BufferPtr buf) ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ nouveau_dri2.c:131:1: warning: no previous prototype for ‘nouveau_dri2_destroy_buffer’ [-Wmissing-prototypes] nouveau_dri2_destroy_buffer(DrawablePtr pDraw, DRI2BufferPtr buf) ^~~~~~~~~~~~~~~~~~~~~~~~~~~ nouveau_dri2.c:137:1: warning: no previous prototype for ‘nouveau_dri2_copy_region2’ [-Wmissing-prototypes] nouveau_dri2_copy_region2(ScreenPtr pScreen, DrawablePtr pDraw, RegionPtr pRegion, ^~~~~~~~~~~~~~~~~~~~~~~~~ nouveau_dri2.c:215:1: warning: no previous prototype for ‘nouveau_dri2_copy_region’ [-Wmissing-prototypes] nouveau_dri2_copy_region(DrawablePtr pDraw, RegionPtr pRegion, ^~~~~~~~~~~~~~~~~~~~~~~~ Signed-off-by: Rhys Kidd commit eb93278fadaab508ba8c2db65d2d3af2bd421efa Author: Rhys Kidd Date: Sun Jan 20 22:21:31 2019 -0500 wfb: Remove declaration for undefined function nouveau_wfb_init() Fixes: 794a277 ("nv50: use libwfb for pixmap access when driver pixmaps enabled") Signed-off-by: Rhys Kidd commit 5233f3c84a436e01dabac531c2615e0e9af5f4ac Author: Rhys Kidd Date: Sun Jan 20 21:31:19 2019 -0500 xv: Avoid shadowed declaration of 'int ret' in NVPutImage Warning reported by gcc 8.2: nouveau_xv.c: In function ‘NVPutImage’: nouveau_xv.c:1369:7: warning: declaration of ‘ret’ shadows a previous local [-Wshadow] int ret = BadImplementation; ^~~ nouveau_xv.c:1073:6: note: shadowed declaration is here int ret, i; ^~~ Fixes: a50bba6 ("xv: misc cleanups") Cc: Ben Skeggs Signed-off-by: Rhys Kidd commit 22dd49f60be54c5649a9e80c87c9bfc99e2974b8 Author: Rhys Kidd Date: Sun Jan 20 21:31:18 2019 -0500 xv: Avoid shadowed declaration of 'int i' in NVPutImage int i is accessed outside immediate scope so leave declaration at the highest common scope level: 1073: int ret, i; ... // Highest common scope ... 1193: if (newTTSize <= destination_buffer->size) { ... // Used in this scope ... 1248: } else { ... // Used in this scope ... 1316: } // Not used in this scope Warning reported by gcc 8.2: nouveau_xv.c: In function ‘NVPutImage’: nouveau_xv.c:1195:7: warning: declaration of ‘i’ shadows a previous local [-Wshadow] int i = 0; ^ nouveau_xv.c:1073:11: note: shadowed declaration is here int ret, i; ^ Signed-off-by: Rhys Kidd commit d7ba24fb6e4fce43adc3c73369c5d1035ae080a9 Author: Rhys Kidd Date: Sun Jan 20 19:20:08 2019 -0500 wfb: Fix missing init function decls behind FB_ACCESS_WRAPPER Fixes warning with xserver 1.20 and gcc 8.2: nv_driver.c:1443:9: warning: implicit declaration of function ‘wfbScreenInit’; did you mean ‘fbScreenInit’? [-Wimplicit-function-declaration] ret = wfbScreenInit(pScreen, FBStart, pScrn->virtualX, ^~~~~~~~~~~~~ fbScreenInit (See xserver 706e6d9cd074da606016ed4ecff51e9c2a822087) Signed-off-by: Rhys Kidd commit dbba3246c989a88f8ba5b0c906af58779bef0cdd Author: Rhys Kidd Date: Sun Jan 20 14:50:35 2019 -0500 xv: Silence build warning regarding const qualifier Silences warning with gcc 8.2: nouveau_xv.c: In function ‘NVInitVideo’: nouveau_xv.c:2247:68: warning: passing argument 2 of ‘vlCreateAdaptorXvMC’ discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers] adaptorsXvMC[0] = vlCreateAdaptorXvMC(pScreen, textureAdaptor[0]->name); ~~~~~~~~~~~~~~~~~^~~~~~ In file included from nouveau_xv.c:42: vl_hwmc.h:6:63: note: expected ‘char *’ but argument is of type ‘const char *’ XF86MCAdaptorPtr vlCreateAdaptorXvMC(ScreenPtr pScreen, char *xv_adaptor_name); ~~~~~~^~~~~~~~~~~~~~~ Signed-off-by: Rhys Kidd Reviewed-by: Ilia Mirkin commit cc284803fd75e0e340a5481342e0d0f7a5d7fa15 Author: Rhys Kidd Date: Sat Jan 19 15:45:36 2019 -0500 Check for xf86CursorResetCursor() If it's available, xorg-server calls it on each mode configuration change. It does what xf86_reload_cursors does (and more), so we don't need to call the latter anymore. Avoids gcc 8.2 warning: drmmode_display.c: In function ‘drmmode_set_mode_major’: drmmode_display.c:525:2: warning: ‘xf86_reload_cursors’ is deprecated [-Wdeprecated-declarations] xf86_reload_cursors(crtc->scrn->pScreen); ^~~~~~~~~~~~~~~~~~~ In file included from nv_type.h:10, from nv_include.h:69, from drmmode_display.c:36: /usr/include/xorg/xf86Crtc.h:1068:37: note: declared here static _X_INLINE _X_DEPRECATED void xf86_reload_cursors(ScreenPtr screen) {} ^~~~~~~~~~~~~~~~~~~ (Ported from radeon commit d670c5c9851b4eff21c845d26c7d7e4eb5ee0fa9) Signed-off-by: Rhys Kidd commit a4283f7fee5b0376f62ba305a2a47ee3dc5a822f Author: Ilia Mirkin Date: Fri Nov 10 23:28:08 2017 -0500 drmmode: update logic for dynamic connectors, paths, and tiles This largely copies the code from modesetting with minor adjustments. Signed-off-by: Ilia Mirkin commit 8dd1118369460e03ab4963987901dc9e13a2d53d Author: Mario Kleiner Date: Tue Jun 12 17:58:06 2018 +0200 Fix colormap handling at screen depth 30. (v2) The various clut handling functions like a setup consistent with the x-screen color depth. Otherwise we observe improper sampling in the gamma tables at depth 30. Tested at depths 16, 24 and 30 and tested at depths 24 and 30 that xgamma and gamma table animations work, and with measurement equipment to make sure identity gamma ramps actually are identity mappings at the output. v2: Also deal with X-Server 1.19 and earlier, which as of v1.19.6 lack a fix to color palette handling and can not deal with depths/bpc > 24/8 bpc. On < 1.20 we skip xf86HandleColormaps() setup at > 8 bpc. This disables color palette handling on such servers at > 8 bpc, but still keeps RandR gamma table handling intact. Tested on 1.19.6 and 1.20.0 to do the right thing. Signed-off-by: Mario Kleiner Signed-off-by: Ilia Mirkin commit c9ce4b8e86525b6e0ecc814b2c5ebb242fb1f107 Author: Ilia Mirkin Date: Sun Dec 9 13:40:55 2018 -0500 drmmode: provide better error when failing to set gamma Signed-off-by: Ilia Mirkin commit ac8f7b949e50d96f281b09f7fe96c92ea4d178ec Author: Ilia Mirkin Date: Fri Feb 9 23:00:27 2018 -0500 dri3: don't check permissions on render node The permission check fails if udev sets the render node to 0666 but leaves the card at 0660, as is done in at least udev-236. Signed-off-by: Ilia Mirkin commit 6e2b4bed7b226dce262c7e14f72a29b003791be8 Author: Ilia Mirkin Date: Sun Feb 4 13:50:29 2018 -0500 nv50/xv: add support for depth 30 xv output Ostensibly it should probably be looking at the pixmap format. However it's always the window pixmap, so we can assume it's what we expect. Signed-off-by: Ilia Mirkin commit ea83b9bd8c88dad67e6494031f1f65eed823b6f1 Author: Ilia Mirkin Date: Sun Dec 31 02:22:27 2017 -0500 dri3: remove bogus condition for creating pixmap Not clear what the depth % 8 was trying to protect against, but it was breaking 30bpp visuals with DRI3. Add it in to ensure that bitsPerPixel % 8 is 0, since there is plenty of bpp/8 math in the driver. Signed-off-by: Ilia Mirkin commit 048baf53872f2f93628f2c8637ec832fb4006057 Author: Ilia Mirkin Date: Sun Dec 10 23:40:51 2017 -0500 man: remove reference to glamor under DRI option GLAMOR backend was removed a while ago, but the help text stayed on. Reported-by: Jason Zaman Signed-off-by: Ilia Mirkin commit 69aecdd305defdd014bc92b82acc6000988511cf Author: Adam Jackson Date: Tue Jun 13 09:23:42 2017 -0400 modesetting: Validate the atom for enum properties The client could have said anything here, and if what they said doesn't actually name an atom NameForAtom() will return NULL, and strcmp() will be unhappy about that. [copied from xserver d4995a3936ae283b9080fdaa0905daa669ebacfc] Signed-off-by: Adam Jackson commit a7c190e0c230ed2d5e047ed6aba12d71a4866950 Author: Lyude Date: Fri Apr 21 14:41:17 2017 -0400 Bump version to 1.0.15 Signed-off-by: Lyude commit 3047e99a641d2b8720371aae292750206f5e91fc Author: Ben Skeggs Date: Thu Apr 20 15:57:33 2017 +1000 fix null pointer deref when building against >=libdrm 2.4.78 A new pointer got added to drmEventContext. As a result of us both: - Setting drmEventContext.version to "latest" AND - Not zeroing the struct We end up thwarting libdrm's compatibility checks, resulting in libdrm choosing to call through an invalid pointer. Fix this by zeroing the struct. Signed-off-by: Ben Skeggs commit e9418e434311336e905b70553a5ed740838d90ad Author: Mariusz Bialonczyk Date: Wed Mar 29 22:07:50 2017 +0200 Do not register hotplug without RandR When using Xinerama, RandR is automatically disabled, and calling RR routines will trigger an assert() because the RR keys/resources are not set, leading to an Xserver abort. Hotplug makes little sense without RandR, so no need to install a udev monitor if RandR is not available. Ported from xf86-video-intel commit 1a489142c8e6a4828348cc9afbd0f430d3b1e2d8, original work by: Chris Wilson Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98383 Signed-off-by: Mariusz Bialonczyk Acked-by: Ilia Mirkin commit e472b47d15634a864c8c981ed588d882aceaf26b Author: Ilia Mirkin Date: Tue Mar 21 21:25:33 2017 -0400 Add Pascal family support, identical to Maxwell Signed-off-by: Ilia Mirkin Tested-by: Lyude commit b71de83b7fae0abeb311251e6144294d319062cf Author: Lyude Date: Tue Mar 7 18:44:43 2017 -0500 Bump version to 1.0.14 Signed-off-by: Lyude commit 924083938c8f209d8f6ff472caf8692a644f7e78 Author: Lyude Date: Fri Mar 3 18:27:42 2017 -0500 Consider CRTCs disabled when DPMS is off It turns out there's a difference in X between a CRTC being "disabled" and simply having it's DPMS turned off. This is problematic though, because if DPMS is turned off you can't really use the CRTC as a normal CRTC anyway since page flipping and vblanks will be non-functional. As a result, we've been considering DPMS-on CRTCs as enabled and attempt to perform pageflips, vblank waits, etc. on them which inevitably fails. and usually breaks the display the first time any of the CRTCs have their DPMS turned on. This was a problem that didn't really show itself until kernel 4.10 when atomic modesetting was added which caused nouveau to stop trying to fulfill pageflips and vblank waits on disabled CRTCs. I'm not sure how pageflipping disabled CRTCs ever worked in the first place, but since not doing so is the proper behavior anyway I haven't investigated any further. So, copy the ms_crtc_on() function from the modesetting driver and add it here as drmmode_crtc_on(), then use that in all of the places where we should be checking for both DPMS off and disabled CRTCs. This fixes issues with the X ceasing to function (usually) after the first time a CRTC has it's DPMS turned on. Reproduction recipe: - Load up gnome-shell on a machine - Wait for the display to timeout from inactivity and turn itself off - Shake the cursor or press something on the keyboard. Chances are the monitor will come back on, but the display remains black until the next time the X server is restarted. Signed-off-by: Lyude Reviewed-by: Adam Jackson commit 1516d35b06c9cda399bef01d992805d6e63dcbae Author: Ilia Mirkin Date: Sun Oct 16 15:10:10 2016 -0400 recognize and accelerate GM20x Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit b00b73c3aa0da1d4cee5c9f580ca65a7bd344e0f Author: Ilia Mirkin Date: Mon Oct 17 23:23:41 2016 -0400 copy: add maxwell/pascal copy engine classes Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit 5a3ada3b0ff3d89441faf35d331bb8b6cf5e51c0 Author: Ilia Mirkin Date: Sun Oct 16 15:03:35 2016 -0400 nvc0: refactor TIC uploads to allow different specifics per generation This flips GM10x to using the updated format, which is what I tested with. However GM20x and GP10x also use this TIC format. Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit 514441f32c5b948ce87668234ab9f01e8ec28dac Author: Ilia Mirkin Date: Sun Oct 16 12:42:18 2016 -0400 nvc0: rename BEGIN_IMC0 to IMMED_NVC0 For consistency with mesa. It wasn't used anywhere previously. Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit 4459c04a6bc208924fac8082920f6786538bc5cd Author: Ilia Mirkin Date: Sun Oct 16 12:38:38 2016 -0400 nvc0: make use of the new hwdefs for TEX_CB_INDEX Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit ea6fd12ea059c6af832bd2180b26d75e00d4ea71 Author: Ilia Mirkin Date: Sun Oct 16 12:38:02 2016 -0400 hwdefs: update nvc0_3d, add gm107_texture for new TIC format These are copied directly from the mesa repository. Signed-off-by: Ilia Mirkin Acked-by: Samuel Pitoiset commit a24ded627e42d453546c553d609edc073f59cd57 Author: Ilia Mirkin Date: Thu Sep 25 20:15:51 2014 -0400 exa: add GM10x acceleration support rendercheck -f a8r8g8b8 passes as much as on a GK208, and xv appears to work. Very lightly tested. Instead of sticking coordinates into pushbufs, the vertex shader is modified to read them from a constbuf, indexed by vertex id. This approach could be used for all nvc0 generations, but I didn't want to rock the boat. Signed-off-by: Ilia Mirkin Reviewed-by: Samuel Pitoiset commit e6479845ec0db20dc733c621b7967b751840a552 Author: Ilia Mirkin Date: Tue Sep 20 00:31:06 2016 -0400 Bump version to 1.0.13 - Fixes for reverse prime offload - ABI 23 support for Xorg 1.19 Signed-off-by: Ilia Mirkin Date: Tue Jul 19 08:29:40 2016 -0700 Use NotifyFd for drm and udev fds NotifyFd is available after API 22, and must be used after API 23. Signed-off-by: Keith Packard commit a964931e8dc4e06c27749edb4993a656dc261d75 Author: Adam Jackson Date: Tue Jul 19 10:03:56 2016 -0400 Adapt Block/WakeupHandler signature for ABI 23 Signed-off-by: Adam Jackson commit 12f77348e7a9579b167b41228dec9e6f97b74de8 Author: Hans de Goede Date: Fri Jun 3 14:46:10 2016 +0200 Properly cleanup fb for reverse-prime-offload drmmode_set_scanout_pixmap(pix) adds drmmod->fb_id through a call to drmmode_xf86crtc_resize(), but on a subsequent drmmode_set_scanout_pixmap(NULL) it would not remove the fb. This keeps the crtc marked as busy, which causes the dgpu to not being able to runtime suspend, after an output attached to the dgpu has been used once. Which causes burning through an additional 10W of power and the laptop to run quite hot. This commit adds the missing remove fb call, allowing the dgpu to runtime suspend after an external monitor has been plugged into the laptop. Signed-off-by: Hans de Goede Signed-off-by: Ben Skeggs commit 1da8a937be19e41c51a3d516bd98cee988bca44b Author: Ben Skeggs Date: Thu Jun 2 11:23:53 2016 +1000 exa/nv50-: fix some potential incomplete pushes I don't think these should actually be able to happen, given that we already reject the unknown formats in the Check() functions. But, just in case... Signed-off-by: Ben Skeggs commit bb57f9af9d179af0962456dca369953521ea8c2c Author: Ben Skeggs Date: Thu Jun 2 10:17:58 2016 +1000 fix use of out-of-scope data Reported by Coverity. Signed-off-by: Ben Skeggs commit b824d36c28124955eda4aced5e637aa75eea4d6c Author: Ilia Mirkin Date: Wed Jan 13 11:55:06 2016 -0500 HAS_DIRTYTRACKING_ROTATION also supports multiple CRTCs A port of commit ff9a6b6f from xf86-video-ati. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93670 Reported-by: Thomas D. Signed-off-by: Ilia Mirkin commit b18bc036bf9997ea65f5cbd824fd057931e604fe Author: Ben Skeggs Date: Tue Dec 8 15:52:25 2015 +1000 Bump version to 1.0.12 Various random fixes have been pending for a long while now... Signed-off-by: Ben Skeggs commit 6e6d8ac1c7b4ee047a7b40b95dea1e65a7c3211a Author: Mario Kleiner Date: Sun Jun 28 02:33:49 2015 +0200 Take shift in crtc positions for ZaphodHeads configs into account. In multi-x-screen ZaphodHeads configurations, there isn't a one-to-one mapping of kernel provided drmmode crtc index to the index of the corresponding xf86Crtc inside the xf86CrtcConfig crtc array anymore, ie. for kernel provided drmmode->mode_res->crtcs[i], the i'th crtc won't correspond to the xf86Crtc in the i'th slot of the x-screens xf86CrtcConfig anymore, once ZaphodHeads has only selected a subset of all crtcs of a graphics card for a given x-screen, instead of all crtcs. This breaks the mapping of bit positions in the bit masks returned in kencoder->possible_crtcs and kencoder->possible_clones. A 1 bit in position i of those masks allows use of the kernels i'th crtc for the given kencoder. The X-Servers dix code checks those bit masks for valid xf86Output -> xf86Crtc assignments, assuming that the i'th slot xf86CrtcConfigPtr config->crtc[i] corresponds to bit i in the xf86Output->possibe_crtcs bitmask, and bails if the bitmask doesn't allow the specified assignment of crtc to output. If ZaphodHeads breaks the assumption of bit i <-> crtc slot i this ends in failure. Take this shift of crtc index positions wrt. encoder bitmask bit positions into account by bit-shifting positions accordingly when assigning encoder->possible_crtcs to output->possible_crtcs, so the proper indices match up again for validation by the dix. This problem wasn't apparent last year when testing the ZaphodHeads support on some Kepler cards, as apparently the encoder->possible_crtcs bitmasks returned for those cards by the kernel just had all 4 lsb bits set for all tested encoders/output, so each of the cards 4 crtcs could go with each output and things worked by chance. The current code breaks, e.g., on 2010 MacBookPro with nv50, where one crtc is hardwired to the internal lvds panel, and one crtc is hardwired to the external DP connector, resulting in a failure where dual-display on single-x-screen works fine, but assigning each output to a separate x-screen via ZaphodHeads fails due to the mismatched encoder->possible_crtcs bitmasks. This patch fixes the problem. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 1ff13a922535924681b91452235b017e43a4c6f6 Author: Ilia Mirkin Date: Sun Sep 13 15:29:15 2015 -0400 fix build after glamor removal Signed-off-by: Ilia Mirkin commit 3bb943b33bbe8e593afc00d13fe867dd7a708c3c Author: Ilia Mirkin Date: Thu Aug 21 20:36:41 2014 -0400 xv: use correct max width/height settings Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=70931 Signed-off-by: Ilia Mirkin commit 3e2e0faa2ee1cce9c1bb5c7ad80d0592460f3edc Author: Ilia Mirkin Date: Sat Jul 11 17:16:27 2015 -0400 remove maxwell GM10x support for now There is no EXA acceleration, and now no GLAMOR either. User is better off with modesetting until EXA makes an appearance. Signed-off-by: Ilia Mirkin commit 1f8b2b348526d94e9bde4a285f331a4934c11078 Author: Ilia Mirkin Date: Sat Jul 11 17:08:52 2015 -0400 glamor: remove If you want glamor, just use modesetting instead. It's hooked up incorrectly in nouveau, and there doesn't seem to be any point in fixing it when the modesetting driver does it all properly. Signed-off-by: Ilia Mirkin commit 6296145654b78518f3299bb5887f224f0d3810fd Author: Mario Kleiner Date: Wed Jul 29 14:39:02 2015 +0200 Add Option "DRI" to allow selection of maximum DRI level. (v2) Allow user to select the maximum level of DRI implementation to use, DRI2 or DRI3. exa accel supports both DRI2 and, if the kernel supports rendernodes, also DRI3. However, DRI3 still seems to have some bugs on current implementations, and additionally it doesn't work well at all for X-Servers older than 1.16.3 due to X-Server bugs. Therefore we default to DRI2 on exa, but allow the user to enable DRI3 with this new option. nouveau's glamor accel backend currently doesn't work under DRI2 at all, so we continue to use DRI3 whenever it is supported and ignore this new option for now. Also add a bit more output about status of Present and DRI3 to aid debugging. Note: This was originally meant to be a boolean parameter, to just select between DRI3 on and off, but changed here to a DRI level to make it consistent with the same option in the released Intel-ddx. v2: Use fixed up Bool return type of nouveau_present_init(). Signed-off-by: Mario Kleiner Cc: Ilia Mirkin Cc: Emil Velikov Cc: Martin Peres Cc: Ben Skeggs Reviewed-by: Ilia Mirkin commit 0deb1fcc6d1099db698ff8c512a4884708a458ec Author: Mario Kleiner Date: Wed Jul 29 14:39:01 2015 +0200 present: Fixup return type of nouveau_present_init() Make it a Bool consistently, as declared in header. Reported-by: Ilia Mirkin Signed-off-by: Mario Kleiner Reviewed-by: Ilia Mirkin commit 5212dbd7e5eba35fdd6578b30f5b0c8c8543081b Author: Ilia Mirkin Date: Tue Jul 14 19:23:01 2015 -0400 Avoid build fail without COMPOSITE Reviewed-by: Dave Airlie commit fdc528306e4018b314f2efcb2be4e37672bddd75 Author: Vadim Rutkovsky Date: Thu Jul 9 23:23:37 2015 +0200 Adapt to new dirty tracking api Mirrors commit b6d871bf2 (radeon: adopt for new X server dirty tracking APIs.) in xf86-video-ati. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91284 commit dfd827cefb940f46b18dc77c129f0a603c7c95fa Author: Samuel Pitoiset Date: Wed May 20 17:08:58 2015 +0200 fix a wrong use of a logical operator in drmmode_output_dpms() This is probably a typo error which has been introduced in 2009... This fixes the following warning detected by Clang : drmmode_display.c:907:30: warning: use of logical '&&' with constant operand [-Wconstant-logical-operand] if (props && (props->flags && DRM_MODE_PROP_ENUM)) { Signed-off-by: Samuel Pitoiset Reviewed-by: Dave Airlie commit 27234dbe00caf6dfd4efb24fc7bbc026fa0134fc Author: Ilia Mirkin Date: Wed May 27 20:26:42 2015 -0400 nv50/exa: add comments showing the code that makes up the programs As decoded by envydis. Perhaps these should move into separate files similar to nvc0, but this seems like a simple intermediate solution. Signed-off-by: Ilia Mirkin commit 08ef36dc6766080f173ae2b7e660b3182cab9f2d Author: Ilia Mirkin Date: Wed May 20 18:36:17 2015 -0400 older hardware prefers copy to clear, so use that Doing a solid copy of 0, or a clear, has the same effect. This is a partial revert of commit 5439a9905eab8c. Pre-nv50 hardware succeeds with copies more often. Signed-off-by: Ilia Mirkin commit 278ad73475bd137eac8a49ec7a22406bfc2867e7 Author: Ilia Mirkin Date: Tue May 19 16:26:16 2015 -0400 nv04-nv40: don't attempt to do 32-bit shifts A 32-bit shift is a no-op, which will also make the new planemask get or'd with ~0, thus negating the usefulness of the subsequent planemask != ~0 check. Only do this if it's a less-than-32-bit per pixel format, in which case it will have the desired effect of setting the high bits. Signed-off-by: Ilia Mirkin Reviewed-by: Martin Peres commit 7583471e1fe181dc7c26b7d0a0d16f70eb10dd21 Author: Ilia Mirkin Date: Tue May 19 16:11:17 2015 -0400 Check before trying a solid fill Pre-nv50 has all sorts of funny requirements for non-copy alu operations, and will bail out of solid fills left and right. Account for that case and fall back to the memset. Reported-by: Andrew Randrianasulu Signed-off-by: Ilia Mirkin Reviewed-by: Martin Peres commit 212fc5354ed60ba282d5eabf468c0e2dd0d0afe6 Author: Ilia Mirkin Date: Mon Mar 23 20:05:23 2015 -0400 nv50: explicitly disable ZETA buffer Signed-off-by: Ilia Mirkin commit 734b8404c0a1fa9ce8ee62a0d4cc5baae7c8fcdf Author: Ilia Mirkin Date: Sat Mar 21 11:46:35 2015 -0400 use defined method names where available Signed-off-by: Ilia Mirkin commit 5439a9905eab8c8d0f86e9963df6c893633bdbfd Author: Maarten Lankhorst Date: Mon Mar 23 19:41:38 2015 +0100 Actually do a clear instead of doing a bogus copy. Oops! commit f0fdc156039eaa79fb1ad82fcc6437f2432306c6 Author: Maarten Lankhorst Date: Sat Mar 21 22:30:10 2015 +0100 Try using EXA ops harder in drmmode_fbcon_copy fallback The memcpy in the fallback path fails with SIGBUS on GK20A because this memory is not coherent, try to use a solid fill instead. commit 7319b54291d59e1cd3cbe4f39ea8cc6dc8bf45bc Author: Maarten Lankhorst Date: Fri Mar 20 23:24:50 2015 +0100 Simplify nouveau_allocate_surface's pitch calculation. Judging from previous commits all linear buffers should have a 256 byte alignment. commit 91e51a7571e6e8cb9338fc168f817f56438b1901 Author: Maarten Lankhorst Date: Fri Mar 20 23:17:13 2015 +0100 Use drmGetRenderDeviceNameFromFd to obtain the dri3 node. commit d29d8baad4d845ec9d3d1d1bc8d1c1e1f6f6f086 Author: Maarten Lankhorst Date: Sat Mar 14 18:00:00 2015 +0100 Use the correct class and shader ops for GK20A. Signed-off-by: Maarten Lankhorst commit 6829d4c672fc05f4d3d993ac1bfde21d4b6c1200 Author: Maarten Lankhorst Date: Sat Mar 14 11:58:12 2015 +0100 Make the ddx load with platform devices Signed-off-by: Maarten Lankhorst commit 3fb97d78202213c1221a1d3ad8a5cebda78fdf44 Author: Maarten Lankhorst Date: Thu Oct 23 09:29:14 2014 +0200 fix null pointer deref commit 90ba0bb6a371a5beebdf6d363bcd73659b7c9f1f Author: Maarten Lankhorst Date: Thu Oct 23 08:25:01 2014 +0200 check memory allocation in nouveau_glamor_set_shared_pixmap_backing As reported by Roy Spliet. commit 4dfa13c77a1f6cee1affea05b5f3a0b26827d366 Author: Maarten Lankhorst Date: Thu Oct 23 08:21:55 2014 +0200 rework initialization to prevent some crashes commit 485902e61b18aff864226db62ba10399aab40128 Author: Maarten Lankhorst Date: Wed Oct 22 15:52:22 2014 +0200 fix !DRI3 case commit 79d438ec4d23280a242d36c2d81f2f43f6e90ada Author: Maarten Lankhorst Date: Wed Oct 22 15:42:55 2014 +0200 fix crash in nouveau_glamor_set_shared_pixmap_backing commit 217f6577afe44b10f7f8e066eaf55419fb0f5ee1 Author: Maarten Lankhorst Date: Wed Oct 22 13:16:33 2014 +0200 move down call to nouveau_present_init until after fb init. commit 241e7289f25a342a457952b9b0e539c2f0b81d99 Author: Maarten Lankhorst Date: Wed Oct 22 11:04:16 2014 +0200 enable dri3 support without glamor commit 762b22f5a2b9deed1c6b18d7098f8fddf2f4b759 Author: Ilia Mirkin Date: Tue Sep 23 21:44:12 2014 -0400 kepler: remove unnecessary texbars, add live-only tex for nvf0 Signed-off-by: Ilia Mirkin commit ce9d6e165cd810f06d158e3cfbb4b9f14b060ded Author: Ilia Mirkin Date: Tue Sep 23 21:37:46 2014 -0400 allow path to envyas binary to be specified Signed-off-by: Ilia Mirkin commit 02c1aee91ae22b58e777716ffd38397f9df0a087 Author: Ilia Mirkin Date: Thu Sep 4 15:27:53 2014 -0400 copy: don't bother trying to create copy on pre-nva3 Only the NVA3/5/8/F tesla chips have a copy engine. Don't bother trying to create one on earlier tesla chips (and thus printing an error on failure). Signed-off-by: Ilia Mirkin commit 3cd4c8494c296c7583dfa7f0823a272c9e932e03 Author: Mario Kleiner Date: Thu Aug 28 03:57:48 2014 +0200 Bump version to 1.0.11 for release. Highlights: - Support for server managed fd's. - Glamor support. - Maxwell support. - DRI3 and initial Present support. - vsync'ed kms pageflip performance fixes when running on Linux 3.13+ - Multi-display vsync, vblank, swap scheduling, timestamping fixes. - Multi x-screen support fixes. - ZaphodHead support on for multiple outputs per x-screen. - EXA nv-10 fixes. - Enable sync of swaps to vblank by default (Option GLXVblank "on"). - Disable pseudo-triplebuffering by default (Option SwapLimit "1"). Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 6d49f3f7aaaec872b6f9841f744a0bf09cc98492 Author: Mario Kleiner Date: Thu Aug 28 03:49:55 2014 +0200 dri2: Set default SwapLimit to 1 for double-buffering. This used to be 2 for pseudo-triplebuffering on XOrg 1.12+. However, it caused problems like unthrottled swapping at up to twice the video refresh rate for redirected windows under desktop composition, which can't be easily fixed under dri2. Default to double-buffering. Users can override this, but dri3 provides better solutions for this. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 10729a442be3cf018c5806c95ec4218977917ef3 Author: Mario Kleiner Date: Mon Aug 18 09:53:27 2014 +0200 xv/dri2/dri3: Switch to optimized crtc selection for drawables. Make crtc selection consistent with ati and intel ddx. Pick the crtc to use for vblank events, swap scheduling and kms-pageflip completion events as the one with maximum pixel area intersection between its viewport and the drawable. If multiple crtc's viewports display the same amount of area of a drawable, select the crtc whose output is the designated "primary output" for the x-screen. The latter allows, e.g., in a clone display setup, to select the important output on which tearing should be minimized. The former makes sure to minimize or avoid tearing on the display which shows the largest part of the drawable. This method is transplanted from the ati ddx. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit be96a5a62df5e1f286690311a6bc4df4f2a99e57 Author: Mario Kleiner Date: Sun Aug 17 02:30:36 2014 +0200 Add file "compile" to .gitignore Some auto-generated file by new make systems to deal with some old compilers. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 16c885ce9c7bebcfa1252a5d500956dd5f551043 Author: Mario Kleiner Date: Tue Aug 12 11:08:24 2014 +0200 Fix event handling on multi-x-screen configs. Only register the wakeup handler and event socket once per card fd and server generation, as the fd and device file is shared between all x-screens for a given card during a given server generation. Without this fix, vblank and kms-pageflip completion event processing don't work properly, as the server doesn't kick the wakeup handler for gpu events, and therefore the desktop will freeze, unless the user manually kicks the wakeup handler by moving the mouse or hitting the keyboard. Add proper reference counting and checks to make it so. This fix is derived from a similar and proven fix in the ati ddx for the same problem. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 7eab6806927ea8938b9bd143d1702e87fde2f2cc Author: Mario Kleiner Date: Wed Aug 13 06:48:10 2014 +0200 Make new ZaphodHeads work across server regenerations. Reset the mask of already assigned crtc's at start of each new server generation. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 6ad053788e804f3a8bd36ee3653ec95e7df62623 Author: Mario Kleiner Date: Wed Aug 6 07:37:32 2014 +0200 dri2 and xv: Fix NVxxSyncToVBlank() for multi-x-screen's. Use helper function drmmode_head() to choose proper hw-crtc to sync to. Fixes xv blits, and dri2 Copy-Swaps. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 17de663a71bcad6f69414eb6bc6bf9b76a6bf5f4 Author: Ben Skeggs Date: Mon Aug 11 09:48:42 2014 +1000 fix glamor header check Signed-off-by: Ben Skeggs commit 64ca88570b71289a019a2d130b1a30dc474ef52d Author: Ben Skeggs Date: Mon Aug 11 08:59:30 2014 +1000 present: use drmmode_head() where it's appropriate Signed-off-by: Ben Skeggs commit 6e71bdf5a9da15c60694fa8543243e37a32c5faf Author: Mario Kleiner Date: Mon Jun 23 11:25:41 2014 +0200 Allow/Fix use of multiple ZaphodHead outputs per x-screen. Defining multiple ZaphodHead outputs per x-screen in a multiple x-screen's per gpu configuration caused all outputs except one per x-screen to go dark, because there was a fixed mapping x-screen number -> crtc number, limiting the number of crtc's per x-screen to one. On a ZaphodHead's setup, be more clever and assign as many crtc's to a given x-screen as there are ZaphodHeads defined for that screen, assuming there are enough unused crtc's available. Tested on a triple display setup with different combos of one, two or three ZaphodHeads per one, two or three x-screens. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit fcf91a275e96f14e49a93dc3f8dd6798cf7d98f3 Author: Mario Kleiner Date: Mon Jun 23 11:13:57 2014 +0200 dri2: Fix drmWaitVblank on multi-x-screen configs. Old mapping of xf86Crtc -> display head index was too simple and only worked on a single x-screen config. Introduce a new helper function drmmode_hw_head_index() which returns proper mapping also for multi x-screens per gpu and ZaphodHeads setups and other non-standard multi- display configurations. Tested on triple-head setup with various ZaphodHead and server layout combos. This helper will be also needed for dri3/present code. Signed-off-by: Mario Kleiner commit 29572e44c24613be9dec0c98ff2731e46ce6203b Author: Ilia Mirkin Date: Sun Aug 10 02:25:15 2014 -0400 exa/nv10: src ops should use the SF macro Doesn't actually end up mattering. Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit 35e17905e1c54558059f6b72b8594a3377f1a57c Author: Ilia Mirkin Date: Sun Aug 10 02:25:14 2014 -0400 exa/nv10: fix blend function setup to check the correct mask Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit dfac18c67a17782a0263daf8cb4ee487374abd2c Author: Ilia Mirkin Date: Sun Aug 10 02:25:13 2014 -0400 exa/nv10: use color 0 when the pict format has no RGB component Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit ff7010e06b6da0fe27143b47b8a4ffd2f12894b7 Author: Ilia Mirkin Date: Sun Aug 10 02:25:12 2014 -0400 exa/nv10: use same clip settings as mesa driver The higher 0x800 was getting overwritten by the 0x7ff anyways, so it wasn't doing any good. The mesa driver just uses 0x800 for the low portion and doesn't set the 8 bit in the higher portion, so do the same thing here. Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit edd1608c397ebcf487b84948031f7807178e5e4d Author: Emil Velikov Date: Tue Jul 1 22:31:47 2014 +0100 xorg_list: define the xorg_list* symbols ...when building against pre 1.12 x. The build will fail as earlier versions of X were missing the xorg_ prefix of the symbols - struct and util functions. Reported-by: Pali Rohár Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit f6c1c755d8230494c8115e2c7899a85663f5c176 Author: Ben Skeggs Date: Tue Jul 1 12:56:49 2014 +1000 fix a harmless typo fdo#80738 Reported-by: Paulo Sérgio Travaglia Signed-off-by: Ben Skeggs commit 096fb1ebc10b14ba1b747832efe378369bcf44e6 Author: Mario Kleiner Date: Mon May 26 05:39:50 2014 +0200 dri2: Add support for handling more than 2 crtc's Need to use the DRM_VBLANK_HIGH_CRTC bits to allow selecting crtc's with id > 1 on latest gpu's with up to four display engines. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 4771ff5c668047a2dc56e4c6131175eed837e757 Author: Marcin Slusarz Date: Sun Jun 22 01:32:42 2014 +0200 present: build only when glamor is enabled nouveau_present_flip_exec references glamor_fd_from_pixmap, which do not exist when glamor is disabled Signed-off-by: Marcin Slusarz Signed-off-by: Ben Skeggs commit 41c1c4ee867753422d6486f13bf20b45905a5dd5 Author: Marcin Slusarz Date: Sun Jun 22 01:31:46 2014 +0200 glamor: fix build without glamor.h xorg-server can be built without glamor, which leads to: CC nouveau_xv.lo In file included from nouveau_xv.c:41:0: nouveau_glamor.h:12:20: fatal error: glamor.h: No such file or directory compilation terminated. Signed-off-by: Marcin Slusarz Signed-off-by: Ben Skeggs commit f0fa8313714c2a5b16e784b257b5ff79da3a443b Author: Mario Kleiner Date: Thu Jun 19 17:11:34 2014 +0200 dri2: Enable sync of bufferswaps to Vblank by default. Make this consistent with other drivers default behaviour. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 02d7e647ce3a4a5d48c4ee94c94bebcac2d4827d Author: Mario Kleiner Date: Sat Jun 21 01:26:16 2014 +0200 dri2: Fix kms pageflip completion timestamping. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 4d92001ebe6bec8c5eaab8156421e62a7e4dedc8 Author: Mario Kleiner Date: Fri Mar 14 18:27:53 2014 +0100 Add kernel async_flip cap detection. Part II of double-sync fix. Query if kernel supports the async_flip cap, thereby needs the new sync behavior. Linux 3.13+ nouveau-kms have this cap and need this fix. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 54dfbb2fed393b65d846e6ed9672acb47e0f19de Author: Mario Kleiner Date: Fri Mar 14 18:23:04 2014 +0100 dri2: Fix double-sync of pageflips on Linux 3.13+ - Part I Linux 3.13 and later sync kms pageflips to vblank in the kms driver, so we must not emit a sync to vblank pushbuf in the ddx on such kernels, or maximum framerate will be cut into half! A sync-to-vblank-pushbuf is emitted for copyswaps as in the past, also for older kernels which don't support async_pageflip's and don't sync by themselves. This adds the implementation, but not the detection logic for async_pageflip support in the kernel. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 303402e00d55b5296311738184cd61f4aadab74d Author: Mario Kleiner Date: Sat Jun 21 00:09:39 2014 +0200 Fix nouveau_copy_init() Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 65a70c4edd1f0df808a3a4dea6e8c9846cc7c920 Author: Ben Skeggs Date: Fri Jun 20 16:57:18 2014 +1000 another build fix... commit 882ebb4000803f6025e1bb3213a31b19f6d1a2c6 Author: Ben Skeggs Date: Mon Nov 18 14:42:02 2013 +1000 present: initial support Until glamor grows its own implementation. Signed-off-by: Ben Skeggs commit 0d94f32fce2759c4b0f4d22b99f3ced09c2aa5a0 Author: Ben Skeggs Date: Fri Jun 20 09:47:39 2014 +1000 kms: implement a more generic event mechanism Present wants to use this too. Signed-off-by: Ben Skeggs commit 97feb3b2104287eb385ffba8329f304e2808d808 Author: Ben Skeggs Date: Fri Jun 20 09:29:08 2014 +1000 dri2: move page flip handling out of drmmode Nothing to see here... This is to (hopefully) avoid busting DRI2 while implementing Present support. Signed-off-by: Ben Skeggs commit 60b876eab3708f1c277027addb5514974452644c Author: Ben Skeggs Date: Fri Jun 20 09:20:37 2014 +1000 dri2: move away from directly touching drmmode on page flips Signed-off-by: Ben Skeggs commit 36dda1341008dd87a6b5459f0616776097d11d9f Author: Ben Skeggs Date: Fri Jun 20 09:04:30 2014 +1000 dri2: shuffle some functions around Nothing to see here... Signed-off-by: Ben Skeggs commit 8bdec8ebd68ba5eaa696d8a666e18774bff3b2ad Author: Ben Skeggs Date: Fri Jun 20 12:37:33 2014 +1000 glamor: fix randr resize Signed-off-by: Ben Skeggs commit 3c45b590a31157abca140096b141aa83fecb671f Author: Ben Skeggs Date: Fri Jun 20 08:28:30 2014 +1000 sync: fix build against DRI3-less server Signed-off-by: Ben Skeggs commit 88ae58d786b0c41148e692dbfe8e2646e7533d9e Author: Ilia Mirkin Date: Thu Jun 19 02:37:45 2014 -0400 update man page with new chips, AccelMethod option Signed-off-by: Ben Skeggs commit 5cb36ba57493bf7edf560bf370c87eff79f1d1a6 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 default to glamor on maxwell We have no RENDER/Xv acceleration in EXA for Maxwell. Signed-off-by: Ben Skeggs commit 81148bb1dbc7007c021c59411d56cb31cfc74ef2 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 glamor: provide dri3 support when enabled Signed-off-by: Ben Skeggs commit 86024ceef015ffe31a204cc5bc6c326a19363ff1 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 glamor: initial support (no dri) Signed-off-by: Ben Skeggs commit fd0ce8839f307693d86c7602dd926ce79e6b777d Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 add support for maxwell, minus Xv/renderaccel More extensive acceleration will be via glamor, at least to begin with. It seems to make sense to jump on the bandwagon now, and deal with any issues that arise before it it becomes the standard for XWayland. Signed-off-by: Ben Skeggs commit 8ad0465e15bfc502abc259e886d6dcb448e0141a Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 50/c0/e0 -> tesla/fermi/kepler Signed-off-by: Ben Skeggs commit 4b138ab18f58c6d459e21dc319615f536c8e69c8 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 merge nv_dma.c into nv_accel_common.c Signed-off-by: Ben Skeggs commit 5c44e6b86195c22e092db5dae06e70208e8b66a0 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 pull copy engine handling into one place, a place that isn't also exa As a side-effect, we now allow Kepler to use async copies too. Signed-off-by: Ben Skeggs commit c375bacbdaeabd0288855c13b4717d0694be9e56 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 exa: move module load into nouveau_exa_init Signed-off-by: Ben Skeggs commit b24cae0bf5db6ece21439d4c6ff3668aed4c78d6 Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 dri2: move "is supported" checks out of nv_driver.c Signed-off-by: Ben Skeggs commit 07c26a0670dd2558acf8e1159f2f68989a7f753d Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 replace NoAccel with AccelMethod in preparation for glamor support Signed-off-by: Ben Skeggs commit e7cf48d9529160f7add2fef72d2e42be9370594a Author: Ben Skeggs Date: Thu Jun 19 14:54:20 2014 +1000 silence a warning Signed-off-by: Ben Skeggs commit 4a18dd74c0cc6e75937eb78347e8ea6c81d1373a Author: Dave Airlie Date: Mon May 26 17:30:24 2014 +1000 autogen: add support for gnome autobuilder Signed-off-by: Dave Airlie commit 8604a741baa5360164a882bd9f3c8474840744db Author: Hans de Goede Date: Fri Apr 11 14:20:43 2014 +0200 Fix building on older servers without xf86platformBus.h Signed-off-by: Hans de Goede commit 0c13843e300d1a9db2bfed18dc1b5df38c5adfdf Author: Hans de Goede Date: Mon Mar 17 17:37:31 2014 +0100 Add support for server managed fds v2 (Martin Peres): - Do not define "NVEntPtr pNVEnt" if not needed to avoid some warnings Signed-off-by: Hans de Goede Signed-off-by: Martin Peres commit 47f08aa6d713da4e6ae9c35ce691e790ffaed09d Author: Hans de Goede Date: Mon Mar 17 17:37:30 2014 +0100 Add a NVOpenNouveauDevice helper function This is a preparation patch for adding support for server managed fds Signed-off-by: Hans de Goede Signed-off-by: Martin Peres commit 480f0998ffed6d9a5c6656dba75182f00fd88a1b Author: Ben Skeggs Date: Thu Nov 7 14:56:48 2013 +1000 bump to 1.0.10 for release Signed-off-by: Ben Skeggs commit 59395c738ebdc21087109075ada67ea8d7783141 Author: Ben Skeggs Date: Thu Nov 7 14:56:06 2013 +1000 recognise GK208 as a Kepler board Signed-off-by: Ben Skeggs commit 03642323a954d0adb11662a98e92dfa20bee9728 Author: Dave Airlie Date: Wed Jul 31 11:19:50 2013 +1000 nouveau: fix build against older servers. Older servers had no GPU support. Signed-off-by: Dave Airlie commit 27323ce68d58fcb82a648ab14abbc99acdc36f44 Author: Dave Airlie Date: Wed Jul 31 11:01:15 2013 +1000 nouveau: add gpu identifier to connector names for secondary gpus this avoids clashes with identifiers from other GPUs. Signed-off-by: Dave Airlie commit 1df177f35a05db505577cdc929e63fde906a704b Author: Dave Airlie Date: Wed Jul 31 10:51:03 2013 +1000 nouveau: fix build on ppc by wrapping immintrin include. Signed-off-by: Dave Airlie commit 300c5a32d79cf507b5fede96710335bcd316eccb Author: Dave Airlie Date: Tue Jul 30 15:28:06 2013 +1000 bump to 1.0.9 for release Signed-off-by: Dave Airlie commit be44e7804862b4c276ed4d4717b1212920f428e6 Author: Dave Airlie Date: Tue Jul 30 15:26:46 2013 +1000 nouveau: fix crash when xinerama is enabled. Signed-off-by: Dave Airlie commit 588252556d1c84bf6793bc22edd48e6650a0f1b9 Author: Dave Airlie Date: Tue Jul 30 15:14:22 2013 +1000 fix make distcheck. add all the nvf0 stuff. Signed-off-by: Dave Airlie commit 29aecda1b5bdb95700894083874492a6d5dfc08a Author: Dave Airlie Date: Tue Jul 30 14:10:51 2013 +1000 drmmode: add support for multi-screen reverse optimus Initial reverse optimus didn't consider multiple screens, so this overhauls the code to use the new X server interface, and allows for multiple outputs on the nvidia to be used with the intel doing the rendering. Signed-off-by: Dave Airlie commit d7ae886260b480da0ea4cad22eb6f7b7d12f07b0 Author: Ilia Mirkin Date: Mon Jul 29 21:03:54 2013 -0400 nouveau: use uppercase for chipset name Right now the log will output e.g. NVaa instead of NVAA, which is how we usually talk about it. Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit 4492cf90c2a7fe27badf6f3f0f296755a65e1e83 Author: Ilia Mirkin Date: Mon Jul 29 02:40:39 2013 -0400 xv: speed up YV12 -> NV12 conversion using SSE2 if available memcpy() goes from taking 45% to 66% of total function time, which translates to a 30% decrease in NVPutImage runtime. Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit 08c667b06d28756d0d32f82bd736b0bdf2b4448e Author: Ilia Mirkin Date: Mon Jul 29 02:40:38 2013 -0400 xv: fix last pixel for big-endian machines in YV12 -> NV12 conversion Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit 8c1c5d4f212958559132506d51dcfeeb17f60371 Author: Ben Skeggs Date: Wed May 15 21:34:13 2013 +1000 nvf0: implement EXA and XVideo acceleration Signed-off-by: Ben Skeggs commit 779b8d44b5ff84b87b52600a2f010ac19cd396ad Author: Sven Joachim Date: Thu Jun 13 20:41:53 2013 +0200 Link against libdrm The DDX uses symbols from libdrm but currently does not link against it, but rather relies on indirect linkage from libdrm_nouveau. This could lead to insufficient dependencies in distribution packages, seen in http://bugs.debian.org/633786. Signed-off-by: Sven Joachim commit e96ae2b129df2a50d1d3839084df4fb1868d914d Author: Maarten Lankhorst Date: Wed Jun 12 11:05:23 2013 +0200 fixup the dri2 check Signed-off-by: Maarten Lankhorst commit 7f20a6af25ae28452ec9c971bfa2227857e6ee93 Author: Maarten Lankhorst Date: Wed Jun 12 10:46:39 2013 +0200 bump to 1.0.8 commit 1fda669c0e5a25eac7f9f1e9a743120cae50ff51 Author: Emil Velikov Date: Sat Feb 16 20:48:47 2013 +0000 configure: printout the configuration info Let the bikeshedding begin Signed-off-by: Emil Velikov commit 45dbcaac301efce04d0bb492337f4febf8129ed9 Author: Emil Velikov Date: Sat Feb 16 20:48:45 2013 +0000 nouveau: mandate dri2 build Building nouveau without dri2 is just silly Signed-off-by: Emil Velikov commit ea8d225fe100ecabb72e3cc0a92372f16afb10b3 Author: Emil Velikov Date: Sat Feb 16 20:48:44 2013 +0000 dri1: purge the final references Signed-off-by: Emil Velikov commit b5d1a9bd269c9d2758fee37492cdcccc976fcc3b Author: Emil Velikov Date: Sat Feb 16 20:48:43 2013 +0000 nouveau: Do not load dri {sub,}module Nouveau does not use dri1, thus loading XFree86-DRI is not needed On the other hand, we do use dri2 As a side effect, purge the 'set-but-unused' variable pLibDRMVersion Signed-off-by: Emil Velikov commit 7c3f6cd8d6b14aea561fb877f6006d339d156f35 Author: Emil Velikov Date: Sat Feb 16 20:48:42 2013 +0000 nouveau: stop using dri1 function DRICreatePCIBusID Step 1 to completely rip out dri1 out of nouveau Signed-off-by: Emil Velikov commit 3b1a45325d5fd46f0e7fad17168742b4c9a2465a Author: Maarten Lankhorst Date: Wed Jun 12 10:30:32 2013 +0200 update makefiles for make distcheck commit b82aacd3e72b52240181569c8b0abb55c1f716d5 Author: Emil Velikov Date: Sat Feb 16 12:48:46 2013 -0800 configure: require xorg-macros 1.8 Explicitly check for xorg-macros 1.8, as a subtle hint to update your build deps if building your own ddx Signed-off-by: Emil Velikov commit 4f667b40f41e311fbe302237efe173bcc24fb8bb Author: Maarten Lankhorst Date: Mon Jun 10 14:29:59 2013 +0200 fix a xfree from previous patch Signed-off-by: Maarten Lankhorst commit 2e34d11541e1f987c7ca0556592cb16f4d232299 Author: Maarten Lankhorst Date: Mon Jun 10 13:06:35 2013 +0200 nouveau: enable support for normal optimus configuration nvd7 has no display engine, only enable sink output capability if there is a crtc. drmmode doesn't work without crtc's, so prevent crashes by skipping the drmmode code in that case. Signed-off-by: Maarten Lankhorst commit bf72ae1f6574c540f0afc2d7845d41df43507a8f Author: Ben Skeggs Date: Wed May 15 22:05:01 2013 +1000 nve0: envyas isn't overly clever.. avoid short-form instructions Signed-off-by: Ben Skeggs commit 41cc2f24c43d2656ac722f594aaebad93e37dfec Author: Ben Skeggs Date: Wed May 15 21:13:11 2013 +1000 nvc0-nve0: decompile shaders into their source, add basic build system Far, far easier to maintain this way... Signed-off-by: Ben Skeggs commit 1edb5d7ae450a32fe6607d536653cf733444e7e2 Author: Ben Skeggs Date: Wed May 15 20:47:44 2013 +1000 nvc0-nve0: split shader code from their headers Signed-off-by: Ben Skeggs commit c18439f759bec267a3b67d375dcbfeb8a6fd7b6a Author: Dave Airlie Date: Mon May 13 13:55:51 2013 +1000 nouveau: clean up nv entity index declarations. Signed-off-by: Dave Airlie commit d3b52efe959f255784f5ead16d7276ca0fb4cdb1 Author: Dave Airlie Date: Mon May 13 13:35:12 2013 +1000 nouveau: attempt to fix zaphod since dri1 code removal j_v on #nouveau bisected b1a630b48210d6a3c44994fce1b73273000ace5c has breaking zaphod, on review it was trying to open the drm fd a second time which was unnecessary. Avoid the problem by storing the nv fd in an entity and have share it between the two scrn info recs. Signed-off-by: Dave Airlie commit 50426ac83f9f077791e4df8c04fc0ecefc06e8ab Author: Marek Vasut Date: Fri May 3 15:48:55 2013 +0200 nouveau: Allow 3D accelerators with output ports This patch let's cards with PCI class 0x30200 (3D controller) be operated by the nouveau driver as well. The nv Quadro NVS 450 is one such card, where the first GPU has PCI class 0x30000 and the other GPU has PCI class 0x30200. By ignoring the 1 << 9 bit in the PCI class mask, displays attached to the other GPU can also be used now. Signed-off-by: Marek Vasut Cc: Dave Airlie Cc: Ben Skeggs Signed-off-by: Ben Skeggs commit 2fa3397e348161a3394e2b456f065921272a056a Author: Ilia Mirkin Date: Fri May 3 09:59:39 2013 -0400 nouveau_xv: Avoid reading off the end of the source image on NV50+ The 'w' argument to NVCopyNV12ColorPlanes is used to index into the source image. line_len is rounded up to 8 on NV50+, so if the source image (+ left offset) is not rounded to 8, NVCopyNV12ColorPlanes could read past the end of the array and crash X. This change can cause the last few horizontal pixels of dst to not be initialized, but they should be truncated by the renderer anyways. See https://bugs.freedesktop.org/show_bug.cgi?id=63263 Signed-off-by: Ilia Mirkin Signed-off-by: Ben Skeggs commit 81f9ab38f0a5de7217227dd17a4d0e50dffd0d28 Author: Ben Skeggs Date: Thu May 2 11:38:07 2013 +1000 nvf0: recognise as supported chipset Signed-off-by: Ben Skeggs commit 653c95d60af0f7c19a1d2f3b76dd17bb169d9c9b Author: Ben Skeggs Date: Mon Apr 29 10:07:59 2013 +1000 nve0: explicitly ask kernel for copy2 object Signed-off-by: Ben Skeggs commit 6771424d79e541d2fa7253a582db3dc9108fd97d Author: Maarten Lankhorst Date: Wed Mar 27 09:50:03 2013 +0100 bump to 1.0.7 Signed-off-by: Maarten Lankhorst commit 84998320162a74a0861b7be1fcc230e50f08424a Author: Maarten Lankhorst Date: Mon Mar 25 13:55:37 2013 +0100 Clean up some errors on closing. If forced close happens, all ioctl's will fail. Some of the handlers also need to be unregistered before the module is unloaded entirely. commit a80785f79268ed3701fc69cbd9194ad12822c82c Author: Dave Airlie Date: Mon Feb 11 09:56:25 2013 +1000 nouveau: fix build against old servers (part 2) Should fix next bit of https://bugs.freedesktop.org/show_bug.cgi?id=60369 Signed-off-by: Dave Airlie commit 05dc9b57896689a3dbd9e6aeb16d57c49515b02e Author: Marcin Slusarz Date: Sun Feb 3 21:12:46 2013 +0100 nv30/exa: fix shaders on big-endian machines Direct port of commit d1bc38b6673c54af61196056c489383fba8dced8 "nv40/exa: fix shaders on big-endian machines" to nv30. Fixes https://bugs.freedesktop.org/show_bug.cgi?id=60050 Signed-off-by: Ben Skeggs commit 61a277488f7195914cbdda10355db0d758fc801c Author: Dave Airlie Date: Thu Feb 7 10:47:20 2013 +1000 nouveau: fix build on older X servers should fix https://bugs.freedesktop.org/show_bug.cgi?id=60369 Reported-by: fabio.ped@libero.it Signed-off-by: Dave Airlie commit e8f222fd10e21f39488d3d8c697e4948cd787cd0 Author: Dave Airlie Date: Wed Feb 6 16:21:50 2013 +1000 nouveau: make prime regression fix more robust. This makes this fix more robust and less crashy. Signed-off-by: Dave Airlie commit d762631c9306b6580b34db1e7eb57bbcac901390 Author: Dave Airlie Date: Wed Feb 6 15:52:33 2013 +1000 nouveau: align shared buffers to 256 We were getting 0xa5 error code on 2D transfers from tiled VRAM to untiled GART when using USB offload devices, this fixes this by upping the alignment on the 2D engine for shared buffers. Signed-off-by: Dave Airlie commit c6980b2add5752a7c285a846b2c9353c9d18afd4 Author: Dave Airlie Date: Wed Feb 6 15:24:09 2013 +1000 nouveau: fix regression in PRIME since 58284cf32778d54cde139423450bc33360784503 Signed-off-by: Dave Airlie commit 59fb3f0252e9b769c5877cfe83d8dbc51cdff232 Author: Dave Airlie Date: Wed Feb 6 14:58:08 2013 +1000 nouveau: create shared pixmaps in GART this creates the shared pixmaps in GART, and makes sure they are untiled. Signed-off-by: Dave Airlie commit 1fdd7db94b55c65ea62cc9eaefff620b20e9e4ea Author: Dave Airlie Date: Mon Jan 7 15:28:53 2013 +1000 nouveau: add reverse prime support This allows the nvidia card to scanout Intel cards rendering. Signed-off-by: Dave Airlie commit 58284cf32778d54cde139423450bc33360784503 Author: Christoph Bumiller Date: Tue Jan 29 16:38:21 2013 +1000 dri2: fix missing frame throttle on copy_region buffer swaps Signed-off-by: Ben Skeggs commit 912d418fdfd2e99eef1e5c631c76dda1d82cf451 Author: Marcin Slusarz Date: Mon Jan 21 00:15:38 2013 +0100 nouveau: put cursor BO in GART Keeping it in VRAM wastes CPU time, because cursor_set ioctl reads handed BO back to RAM, just to write it to actual cursor BO. Here (nv92/core i7), this patch decreases overall cpu usage of drmmode_load_cursor_argb from 4.6ms to ~90us and lets nv50_crtc_cursor_set disappear from perf traces. Signed-off-by: Ben Skeggs commit 8f934fad5d4934936b3265f272ca81e73c60d7dc Author: Dave Airlie Date: Mon Jan 7 14:13:10 2013 +1000 bump to 1.0.6 Signed-off-by: Dave Airlie commit 00834273e22f7234a3755269100ea3f10b521e35 Author: Dave Airlie Date: Mon Jan 7 14:04:40 2013 +1000 nouveau: fix dirty redisplay damage handling This processes the damage in the correct order. Signed-off-by: Dave Airlie commit 389b8a8dd91a48dee524eac96e60813c507f1a82 Author: Ben Skeggs Date: Mon Jan 7 14:05:06 2013 +1000 bump to 1.0.5 Signed-off-by: Ben Skeggs commit efe9c57af6ff4c12936e49ac3e188c54b8b2d5bd Author: Ben Skeggs Date: Mon Jan 7 11:35:56 2013 +1000 nvc0-/exa: fix typo which caused breakage in rendercheck tmcoords test Signed-off-by: Ben Skeggs commit 5cd86445272818ec2f598f4685a5d7f24e403d07 Author: Marcin Slusarz Date: Sun Nov 25 22:59:49 2012 +0100 add some missing newlines in error messages Signed-off-by: Ben Skeggs commit 3b33a0b426b8915f302247dcd40c64fb1a6dcdd2 Author: Emil Velikov Date: Sun Nov 4 18:06:18 2012 +0000 kms: Let the server know if we got more than 1 EDID block Verbatim copy from xf86-video-radeon commit 809cacb Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32341 Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit 018ab2d4634f3c8923038337ce9600973e2f0002 Author: Ben Skeggs Date: Fri Nov 9 12:16:00 2012 +1000 bump version to 1.0.4 Signed-off-by: Ben Skeggs commit 0d32d964dbba3864af0946aa8894cef858eb1319 Author: Dave Airlie Date: Wed Nov 7 14:49:03 2012 +1000 nouveau: drop unused nv private members. these are no longer needed. Signed-off-by: Dave Airlie commit b1a630b48210d6a3c44994fce1b73273000ace5c Author: Dave Airlie Date: Wed Nov 7 14:45:14 2012 +1000 nouveau: drop DRI1 device open interface. this is a DRI1 left over, the driver is meant to open the device itself now. this old interface did crazy stuff like drmAddMap. Reviewed-by: Ben Skeggs Signed-off-by: Dave Airlie commit 27a1a0616304e9b9f0ae842899b7d614f1026578 Author: Emil Velikov Date: Fri Nov 2 03:57:42 2012 +0000 nouveau: NVPlatformProbe, handle shareable entities During NVPciProbe, xf86ConfigPciEntity() implicitly calls checks and sets the entity as shared Here we are only allocating the screen, thus need to explicitly call the functions ZaphodHead and Xinerama should finally work Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=56347 Reported-By: Vladimir Chebotarev Reported-By: Damian Nowak Tested-By: Damian Nowak Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit e34cfbd5bd23f7f15372af52d8a39a5715ce7310 Author: Emil Velikov Date: Fri Nov 2 03:57:41 2012 +0000 nouveau: Factor out common code to NVHasKMS() As the name suggests checks if it has kernel mode setting, prints out the interface version and checkes if the chipset is supported Function is used in NVPciProbe and NVPlatformProbe Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit 78f24dc3798d635fca2e20856a09012c28bbbb8b Author: Emil Velikov Date: Fri Nov 2 03:57:40 2012 +0000 nouveau: introduce NVInitScrn() Common/helper function for NV{Pci,Platform}Probe() This introduces some the following functions to the NVPlatformProbe codepath * xf86SetEntitySharable() * xf86SetEntityInstanceForScreen() Necessary for dualhead, xinerama and/or zaphodhead v2: Rename nouveau_init_scrn() to NVInitScrn() Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit 47457fbd5006070ea9a06238a5a69ca8167dac77 Author: Emil Velikov Date: Fri Nov 2 03:57:39 2012 +0000 nouveau: Remove unneeded xf86GetEntityInfo() Retrieving the whole entity information is not needed as we only use the index - stored already in entity_num Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit 20995bb5920021668b8b607f886201c643ee0e9a Author: Dave Airlie Date: Fri Nov 2 09:04:43 2012 +1000 nouveau: fix regression in virtualbox drawing hopefully. Signed-off-by: Dave Airlie commit 8c3e1623b0be15f8cc590d893bfd19be87bd079a Author: Maarten Lankhorst Date: Wed Oct 24 11:19:04 2012 +0200 bump version to 1.0.3 commit 6644a3a724a11150eba0c0082c8d4fe167b1755e Author: Maarten Lankhorst Date: Tue Oct 23 20:17:29 2012 +0200 exa: remove exaMarkSync from download/upload screen They're noops on nouveau since sync markers are unused and download/upload are always synchronous anyway from exa's POV. Signed-off-by: Maarten Lankhorst commit 2a757b7dffa85c30879653a1e52e33b8f4fa8806 Author: Maarten Lankhorst Date: Tue Oct 23 17:04:37 2012 +0200 exa: fix comparison in download_from_screen Presumably a copy/paste error from upload_to_screen. Signed-off-by: Maarten Lankhorst commit 72230cd5a7846f7a26f929262a745dc0bdc34a3d Author: Maarten Lankhorst Date: Tue Oct 23 10:37:27 2012 +0200 exa: fix up download_from/upload_to screen Add a missing return TRUE to download_from_screen. Add a missing exaMarkSync to memcpy in upload_to_screen. Add a really loud warning to both if memcpy to/from tiled is done. Signed-off-by: Maarten Lankhorst commit 1d7890ecb8b3188177ab85608ea232812672d923 Author: Maarten Lankhorst Date: Tue Oct 23 09:53:39 2012 +0200 remove unused members from nv_type.h commit 0b2a548316b0ce88ba19d5e05fdca205fc139962 Author: Maarten Lankhorst Date: Mon Oct 22 13:12:06 2012 +0200 nvc0/exa: make solid fill actually work Some investigation showed that the code for accelerating solid fill didn't actually work because NVC0EXACheckTexture would fail. Making the check the same as in NV50EXACheckTexture fixes this, and seems to get acceleration for solid fills working right. Signed-off-by: Maarten Lankhorst commit b4231dd715a8a7f86b04519b5f4e8a8d93c2f561 Author: Mario Kleiner Date: Tue Oct 9 09:06:59 2012 +0200 dri2: Fix potential race and crash for swap at next vblank. This fixes a potential race + crash that wasn't properly handled by commit 248de8cdbd6d0bc062633b49896fa4791148cd3b and happened at least on one users machine. That commit wrongly assumed no special action would be needed for swaps at next vblank while triple-buffering is enabled on XOrg server 1.12 or later. Closer inspection of the x-server main dispatch loop shows it is possible that the client manages to get the server to dispatch a new DRI2GetBuffersWithFormat() call before the server calls the vblank event handler and executes the nouveau_dri2_finish_swap() routine. Such a race would cause a crash, as described in above commit. This commit handles the "swap at next vblank" case by calling nouveau_dri2_finish_swap() immediately without the roundtrip (queue vblank_event -> kernel -> deliver event -> x-server processes event -> nouveau vblank event handler), before control gets returned to the client. This avoids the race while retaining triple-buffering. As a bonus, time-critical swaps at next vblank get processed without roundtrip delay, increasing the chance of not skipping a frame due to vblank miss while sync to vblank is on. Thanks to Anssi for reporting this problem on the nouveau mailing list at 12th July 2012 and for testing this patch. Reported-by: Anssi Hannula Tested-by: Anssi Hannula Signed-off-by: Mario Kleiner commit 01c9a9cff7d6232ca6eafa59dd60833a0d9cdc5b Author: Ben Skeggs Date: Thu Oct 18 18:38:21 2012 +1000 shadowfb: fix segfault due to reading outside of shadow buffer Probably caused by the new libdrm port, the new libdrm sets the bo size field to the *allocated* size and not the *requested* size, making the max_height calculation here invalid. Switched to using virtualX/virtualY as the bounds, which should hopefully do the right thing.. Signed-off-by: Ben Skeggs commit 297fd0d0755bda698be1d0b30cc60a41d7673c0b Author: Dave Airlie Date: Tue Oct 16 16:15:16 2012 +1000 nouveau/dri2: fix pixmap/window offset calcs. This should fix prime rendering under kwin, and not break it under the others. Signed-off-by: Dave Airlie commit 5c9379b14cdabc81fd5d4c916dcd93dc77c4f683 Author: Sebastian Keller Date: Tue Sep 25 11:35:35 2012 +0200 exa: use CLAMP_TO_EDGE for RepeatPad This fixes border rendering in some gtk3 themes. https://bugs.freedesktop.org/show_bug.cgi?id=55310 v2 (Ben Skeggs): implement same fix for fermi/kepler too Signed-off-by: Sebastian Keller Signed-off-by: Ben Skeggs commit 302b5da84b4890c067f5f7e8b0f4e689909bbb51 Author: Adam Jackson Date: Tue Sep 25 10:44:32 2012 -0400 Remove mibstore.h Signed-off-by: Adam Jackson commit 20f3830ddf868c0d1a46c18fa381585c076bfb8d Author: Maarten Lankhorst Date: Wed Sep 12 13:42:43 2012 +0200 bump version to 1.0.2 commit 239f39977ebb64c9d545c6f3c905de3c0ad840d0 Author: Dave Airlie Date: Thu Jul 19 13:46:44 2012 +1000 nouveau: enable provider/capabilities this adds the provider object for nouveau and enables the capabilities on it. Signed-off-by: Dave Airlie commit 28abba276e36fd0176f416996d44297ddf13f5cd Author: Dave Airlie Date: Thu Jul 19 13:46:12 2012 +1000 nouveau/dri2: add support for DRI2 offload This adds the nouveau interfaces for DRI2 offload, this allows nouveau to act as an offload slave. Signed-off-by: Dave Airlie commit b0be0ac86bae871c9f1724c40298d50d7f2b08e4 Author: Dave Airlie Date: Thu Jul 19 13:45:32 2012 +1000 nouveau: add pixmap tracking/updating support. This should allow nouveau to act as an output master for driving output slaves. Signed-off-by: Dave Airlie commit baf485b1d304d741073f058d524ca698af3c15cd Author: Dave Airlie Date: Thu Jul 19 13:42:13 2012 +1000 nouveau/exa: add support for shared pixmaps. This adds the pixmap export/import hooks, along with code for nv50/nvc0 to validate shared buffers in GTT space. Signed-off-by: Dave Airlie commit 2f48b8f462a03cb92db9e9a7ae1957eb27473965 Author: Dave Airlie Date: Thu Jul 19 13:40:23 2012 +1000 nouveau: add platform bus support This adds the ability to load nouveau via the new platform apis. Signed-off-by: Dave Airlie commit 3aa9a296f508089a247f92b829036ca6a55b7b42 Author: Adam Jackson Date: Wed Aug 15 13:26:51 2012 -0400 Remove vbe and int10 includes This driver has never used int10 or vbe, presumably these were just copypasta from nv. Signed-off-by: Adam Jackson commit 9b9be11e99930136a591059db1bf4ceaca8806e7 Author: Adam Jackson Date: Thu Jul 19 18:22:44 2012 -0400 Implement ->driverFunc Copied from fbdev, makes it so we can run without iopl. Signed-off-by: Adam Jackson commit 6629066b87142cdd1d9e881b61ef55c97aedf949 Author: Ben Skeggs Date: Fri Aug 3 17:05:33 2012 +1000 nvc0/nve0: support sync-to-vblank if kernel is new enough Signed-off-by: Ben Skeggs commit d1bc38b6673c54af61196056c489383fba8dced8 Author: Ben Skeggs Date: Fri Jul 6 16:23:50 2012 +1000 nv40/exa: fix shaders on big-endian machines Signed-off-by: Ben Skeggs commit 0d2b414aff634c996e0468bb0df1894d6c28c09e Author: Ben Skeggs Date: Tue Jun 19 10:12:37 2012 +1000 bump version to 1.0.1 commit f84c28c9e6a0079e527655994689cb9c2ba51999 Author: Ben Skeggs Date: Tue Jun 19 09:49:02 2012 +1000 exa: fix remaining signed datatype issues Wasn't expecting this situation, apparently it happens.. Signed-off-by: Ben Skeggs commit 794141f22179a09ba6b2a094ba72316c298fee8b Author: Maarten Maathuis Date: Tue Jun 19 00:28:59 2012 +0200 exa: fix various issues in transfer bo handling Thanks to "Christoph Bumiller " for hinting at the cause of the memory leak. Signed-off-by: Maarten Maathuis Signed-off-by: Ben Skeggs commit 661d0cf42219a063f8faa0518f3490ddd38209ff Author: Christoph Bumiller Date: Mon Jun 18 17:13:55 2012 +0200 nve0: fix and improve NV12 shader Only the 1st set of coordinates is valid. NVC0 was fixed in 9a3e579f637267b32efc46cfce5d1a36a41323b1. commit a907efe44177298cb1c42a236aab087c14b71657 Author: Christoph Bumiller Date: Mon Jun 18 13:50:27 2012 +0200 nvc0/accel: oops fix method size in previous patch commit 8805e20bb9c1cfb7477f3eb7def64c5711fc91ea Author: Christoph Bumiller Date: Mon Jun 18 13:41:43 2012 +0200 nvc0/accel: use correct (signed) vertex data type in PUSH_VTX1/2s Could have kept s16 but decided against mask and shift. commit 6dbf4ea12600275775123f9f564469454415da55 Author: Ben Skeggs Date: Mon Jun 18 13:28:07 2012 +1000 bump version to 1.0.0 Signed-off-by: Ben Skeggs commit 5e74cee90eeb4e8e638b371a5f1a2c6737be70ff Author: Ben Skeggs Date: Mon Jun 18 15:36:47 2012 +1000 distcheck fixes commit 383aef5a8882a65efdf34e8d2681cd5ccec51b22 Author: Ben Skeggs Date: Mon Jun 18 13:31:37 2012 +1000 exa: drop massive gart allocation in favour of smaller ones, as needed Signed-off-by: Ben Skeggs commit 52836c15b1c0e4f28d92ca9e6c76872481917fd4 Author: Ben Skeggs Date: Mon Jun 18 13:20:18 2012 +1000 nv50/exa: somehow fix firefox tabs not being drawn properly since vp xfrm Signed-off-by: Ben Skeggs commit 36d3f8c28c55c05ccf4efa985831d2d249bad2fe Author: Ben Skeggs Date: Wed Jun 13 22:41:58 2012 +1000 dri2: kill debug spam Signed-off-by: Ben Skeggs commit 4dbc132f22721e3da30eb2e7fc97dea5b8458df6 Author: Viktor Novotný Date: Sat May 26 22:15:20 2012 +0200 nv04/exa: Reset destination surface offset in the same call of NV04EXACopy Fixes FDO bug #48954. Signed-off-by: Viktor Novotný Reviewed-by: Marcin Slusarz Signed-off-by: Ben Skeggs commit aab58ee9a8025422416458d637c203af5ba008fd Author: Dave Airlie Date: Tue Jun 5 12:34:06 2012 +0100 nouveau: i is used inside the function fixes build, reported by tallica on irc. Signed-off-by: Dave Airlie commit 619e99731f772ff8d9d93cd1d6d83de5d450574b Author: Dave Airlie Date: Tue Jun 5 11:02:54 2012 +0100 nouveau: port to compat server API. This ports to the new server API. Signed-off-by: Dave Airlie commit d2e16c62b5054f34a439e504bfa261fb71f5d7de Author: Dave Airlie Date: Tue Jun 5 10:57:30 2012 +0100 nouveau: drop flags arg to adjust frame Signed-off-by: Dave Airlie commit fadf83d7b373282ccbf0fa0c01928a35ff717a5e Author: Ben Skeggs Date: Thu May 31 15:46:57 2012 +1000 nvc0/exa: add support for async UTS/DFS copies Signed-off-by: Ben Skeggs commit 1600f7f202721c4b29c29e94f7fb69dc313b99dc Author: Ben Skeggs Date: Thu May 31 15:40:45 2012 +1000 nva3/exa: add support for async UTS/DFS copies Signed-off-by: Ben Skeggs commit ace77b6b1304826f4004bde23809b55d476b0615 Author: Ben Skeggs Date: Tue May 29 21:21:57 2012 +1000 disable fermi accel on 0.0.16 interface Kepler accel support broke some assumption made by the older kernel interface, and Fermi shares the same code. It can't work (without some annoying hacks anyway) with the 0.0.16 kernel anymore. commit 7041e30ab8beb627bbf569367961a658e79c2bdc Author: Dave Airlie Date: Wed May 23 14:18:24 2012 +0100 vl_hwmc: add missing compat include. Reported-by: tallica on irc. Signed-off-by: Dave Airlie commit 2abf8467cfb7a7648ce73ba5bcbbc62219d65d6d Author: Dave Airlie Date: Wed May 23 11:29:05 2012 +0100 nouveau: add compat-api.h to makefile. Signed-off-by: Dave Airlie commit 1d861ad716861c57b2b81531d21840d7c8de024b Author: Dave Airlie Date: Wed May 23 11:15:06 2012 +0100 nouveau: convert two more xf86Screens access to macros for some reason the script missed these two, just fix them manually. Signed-off-by: Dave Airlie commit 5625fb84efc699e65da0062ae101915a49f2969b Author: Dave Airlie Date: Wed May 23 11:13:30 2012 +0100 nouveau: convert scrn/screen to using new interfaces This commit was generated with the util/modular/x-driver-screen-scrn-conv.sh Signed-off-by: Dave Airlie commit 9f037f7be516ebf8fa4a67a5c9ff258594ab353e Author: Dave Airlie Date: Wed May 23 11:12:59 2012 +0100 nouveau: add initial compat-api header This just provides the scrn conversion and picture stuff so far. Signed-off-by: Dave Airlie commit 581564466c47ab0435ecca3cf4d82c03e1b30259 Author: Francisco Jerez Date: Tue May 8 14:02:44 2012 +0200 dri2: Don't try to page-flip pixmaps (fdo bug 49351). Reported-by: Marcin Slusarz Signed-off-by: Francisco Jerez commit b1efc807392d372e6a232aad9d16ae6e1b1c1916 Author: Marcin Slusarz Date: Sun Apr 15 13:46:42 2012 +0200 nv50: add missing pushbuf space check in NV50SyncToVBlank Regression from "WIP: port to new libdrm". commit 174f17072005971fee6e4eb76a273c3cc1aa10d3 Author: Ben Skeggs Date: Tue Apr 24 21:55:03 2012 +1000 nv40/exa: return false from gradient picture setup Not implemented yet. Shouldn't hit this anyway, fallback in check_composite, but, best be safe :) Reported-by: Maarten Maathuis Signed-off-by: Ben Skeggs commit 60edf2a87b928f413385443335493cb27da30a48 Author: Ben Skeggs Date: Tue Apr 24 10:54:51 2012 +1000 nv40/exa: support for solid pictures Signed-off-by: Ben Skeggs commit 067b8416ad1ab3e71ebbb8db8bfb82ff6041939b Author: Ben Skeggs Date: Mon Apr 23 17:49:57 2012 +1000 nv30/exa: implement support for solid pictures using register combiners Signed-off-by: Ben Skeggs commit b09a61780ea6710b679a6a099e35987e7a5f559f Author: Ben Skeggs Date: Mon Apr 23 14:49:22 2012 +1000 nv10/exa: implement support for solid pictures Signed-off-by: Ben Skeggs commit 9583753a205c585ec844dfb7168513a2c2a157ba Author: Ben Skeggs Date: Fri Apr 20 20:52:01 2012 +1000 nv10/exa: use fixed-function texture matrix for transforms Signed-off-by: Ben Skeggs commit c21c0ae57717bfb0fceab82505009a5382d1428d Author: Ben Skeggs Date: Fri Apr 20 20:30:44 2012 +1000 nv30/exa: use fixed-function texture matrix for src/mask transforms Signed-off-by: Ben Skeggs commit 24a5ca723539a940bc2dabcc50aa631f0d9d4219 Author: Ben Skeggs Date: Fri Apr 20 15:24:45 2012 +1000 nv40/exa: do transforms in vertex program Signed-off-by: Ben Skeggs commit 9a3e579f637267b32efc46cfce5d1a36a41323b1 Author: Ben Skeggs Date: Thu Apr 19 16:58:52 2012 +1000 nvc0/exa: port recent nv50 changes (vp transform, solid pictures, etc) Signed-off-by: Ben Skeggs commit 8ea09db1a1e8dff7c341dc3da17edefda7e56e6d Author: Ben Skeggs Date: Thu Apr 19 13:46:03 2012 +1000 nv50/exa: support solid pictures Signed-off-by: Ben Skeggs commit b48bcc094beecf521899dd63c8fdbccfd534e5cd Author: Ben Skeggs Date: Thu Apr 19 13:21:27 2012 +1000 nv50/exa: perform texcoord transformations in vertex program Signed-off-by: Ben Skeggs commit 5ac2ca8c56ec8b055878c8ac4cbc8ca74379abda Author: Ben Skeggs Date: Thu Apr 19 09:52:24 2012 +1000 nv50: implement and use PUSH_DATAu Signed-off-by: Ben Skeggs commit 6c41b3d7a8799daa9d0b34dcfbfc5c891c24fffa Author: Ben Skeggs Date: Tue Apr 17 11:07:05 2012 +1000 nv30/nv40: upload shaders directly from rankine/curie init Also removes A8 surface shader hack, and implements proper shaders for the specific purpose. Signed-off-by: Ben Skeggs commit ec019b3a50475b7ff1b5e0a5dbe14e5ef677bd9e Author: Ben Skeggs Date: Tue Apr 17 09:30:49 2012 +1000 Keep a single buffer for random accel data, rather than 3 different ones Signed-off-by: Ben Skeggs commit fb3a36b1e5af0f81bb266da894d3442eed8e4e55 Author: Ben Skeggs Date: Wed Mar 28 16:38:30 2012 +1000 nve0: initial exa/xv acceleration for kepler chipsets Signed-off-by: Ben Skeggs commit 95be16d8cd88db1ace0ac4568b4c67b51fd30fd9 Author: Ben Skeggs Date: Thu Mar 29 19:55:27 2012 +1000 nvc0: create scratch buffer before initialising m2mf Signed-off-by: Ben Skeggs commit 4896c87484e1cf1e5421de377b6b83890948664d Author: Ben Skeggs Date: Thu Mar 29 19:47:21 2012 +1000 nvc0: switch to kepler's subchannel assignments (see note) This will bust for kernels with nouveau ABI version < 1.0.0... It should be possible to add compat for 0.0.16 however... Signed-off-by: Ben Skeggs commit 4225b486cb8ab98d27ba23874096ff0879d45575 Author: Ben Skeggs Date: Thu Mar 29 15:56:02 2012 +1000 nvc0: tidy and cull a heap of fermi init (stuff in default grctx) Signed-off-by: Ben Skeggs commit 007dd62a8c5244bc5e71fe8f32520fb35b67a237 Author: Ben Skeggs Date: Thu Mar 29 15:28:03 2012 +1000 nvc0: punt shaders out to a header file Signed-off-by: Ben Skeggs commit 1c7651532975cf738b3aac4ec3c3b46187e24eea Author: Ben Skeggs Date: Thu Mar 29 14:25:15 2012 +1000 nvc0: do tic/tsc binding only once, at startup Signed-off-by: Ben Skeggs commit 180d8778764913ed061e68520603c0925477faba Author: Ben Skeggs Date: Thu Mar 29 14:21:25 2012 +1000 nvc0: replace direct inline m2mf with inline helper to do the same Signed-off-by: Ben Skeggs commit e70d801ae9287eab5e82f4d467dc8cd4be1b31a8 Author: Ben Skeggs Date: Fri Dec 9 18:15:32 2011 +1000 WIP: port to new libdrm calim: nv50,nvc0/xv: fix src_w/h width/height confusion commit ab7291d368ddc66ad21c5ad5caa0faeee42ccaf1 Author: Ben Skeggs Date: Wed Mar 21 14:56:55 2012 +1000 xv: don't try to init textured video without 3d engine object Signed-off-by: Ben Skeggs commit fe7e0425a4a43b2a7d4743a0a90a86724c3b6775 Author: Ben Skeggs Date: Mon Mar 12 11:59:29 2012 +1000 recognise kepler Signed-off-by: Ben Skeggs commit f5d1cd2cb6808838ae1a188cef888eaa9582c76d Author: Ben Skeggs Date: Tue Mar 6 08:16:51 2012 +1000 remove use of xf86PciInfo.h, deprecated now apparently Signed-off-by: Ben Skeggs commit 248de8cdbd6d0bc062633b49896fa4791148cd3b Author: Mario Kleiner Date: Thu Feb 16 00:45:24 2012 +0100 dri2: Fix corner case crash for swaplimit > 1 If a swaplimit > 1 is set on a server which supports the swaplimit api (XOrg 1.12.0+), the following can happen: 1. Client calls glXSwapBuffersMscOML() with a swap target > 1 vblank in the future, or a client calls glXSwapbuffers() while the swap interval is set to > 1 (unusual but possible). 2. nouveau_dri2_finish_swap() is therefore called only at the target vblank, instead of immediately. 3. Because of the deferred execution of nouveu_dri2_finish_swap(), the OpenGL client can call x-servers DRI2GetBuffersWithFormat() before nouveau_dri2_finish_swap() executes and it deletes pixmaps that would be needed by nouveau_dri2_finish_swap() --> Segfault --> Crash. Prevent this: When a swap is scheduled into the future, we temporarily reduce the swaplimit to 1 until nouveau_dri2_finish_swap() is done, then restore it to its original value. This throttles the client inside the server in DRI2ThrottleClient() before it can call the evil DRI2GetbuffersWithFormat(). The client will still be released one video refresh interval before swap completion, so there is still some potential win. This doesn't affect the common case of swapping at the next vblank, where this throttling is not needed or done. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 247465ec5e21ce7151ef1e6a3e7644bc74d48a5b Author: Mario Kleiner Date: Thu Feb 16 00:45:23 2012 +0100 dri2: Reimplement hack for triple-buffering on old X-Servers. X-Servers before 1.12.0 don't have the DRI2SwapLimit() API. On these, default to a swaplimit of 1 - double-buffering. This patch implements support for swap limit of 2, triple-buffering, on old x-servers via Francisco Jerez previous hack: Return DRI2SwapComplete() before the swap has completed, so clients don't get blocked on the pending swap. This allows for a "triple-buffering look-alike" behaviour, but breaks the swap scheduling and timestamping defined in the OML_sync_control spec, so applications which rely on conformant behaviour will break with a swap limit of 2 on pre 1.12.0 x-servers. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit a1902ffaab1f6807fdaac0fce49f784e7208d1a5 Author: Mario Kleiner Date: Thu Feb 16 00:45:22 2012 +0100 dri2: Allow vblank controlled swaps for redirected windows. Part II This part implements proper throttling for clients. For vblank synchronized blits, it defers DRI2SwapComplete() until 1 vblank after the framebuffer blit is submitted to the gpu. Rationale: For unredirected windows, this is the earliest time the "blit swap" can complete, as blits are submitted one vblank before the target vblank and synchronized with vblank in the gpu. This makes swap completion timestamps at least reasonable. For redirected windows, the compositor will probably pick up the "blit swapped" frontbuffer pixmap of the window quickly, but defer its own recomposition to the next vblank, at least if sync to vblank for the compositor is on. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit c8572d329014afc0c79b5765468dcdfb65be0868 Author: Mario Kleiner Date: Thu Feb 16 00:45:21 2012 +0100 dri2: Allow vblank controlled swaps for redirected windows. Part I Make sure that swaps for redirected windows under a compositor are still scheduled via vblank events, to avoid effects like 2900 fps swaps under a compositor. See discussion with Francisco Jerez at: http://lists.freedesktop.org/archives/nouveau/2011-September/009278.html http://lists.freedesktop.org/archives/nouveau/2011-September/009292.html This is part I of the agreed upon band-aid, in a separate patch. It allows to use vblank related functions on redirected windows and thereby fixes functions from sgi_sync_control and oml_sync_control extension, e.g., glXWaitForMscOML(), glXGetSyncValuesOML(), glXWaitVideoSyncSGI, ... Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 919643888045b102f3e10dfdeb6210ab14b6ecab Author: Mario Kleiner Date: Thu Feb 16 00:45:20 2012 +0100 dri2: Fixes to swap scheduling. Fix some small off-by-one errors and a mismatch between 32 bit kernel interfaces for vblank count and 64 bit dri2 interfaces for target_msc et al. Return corrected target_msc to swap scheduling in x-server. A revised version of the patch discussed here: http://lists.freedesktop.org/archives/nouveau/2011-September/009143.html Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 5ede4538366a352aa9726560f8d42e85df4aec82 Author: Mario Kleiner Date: Thu Feb 16 00:45:19 2012 +0100 dri2: Update front buffer pixmap and name before exchanging buffers Buffer exchange assumes that the front buffer pixmap and name information is accurate. That may not be the case eg. if the window has been (un)redirected since the buffer was created. This is a translation to nouveau of a fix that was originally developed by Ville Syrjala for the ati/radeon ddx to fix the same bug there. See thread at: http://lists.x.org/archives/xorg-devel/2011-May/021908.html Fixes FDO bug #35452. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 6fdf60b7288d49f889c80705aabc1db3bc327ba3 Author: Mario Kleiner Date: Thu Feb 16 00:45:18 2012 +0100 dri2: Add support for DRI2SwapLimit() API. Uses the new DRI2SwapLimit() API of X-Server 1.12+ to allow to change the maximum number of pending swaps on a drawable before the OpenGL client is throttled by the server. The new optional xorg.conf parameter "SwapLimit" allows to select a new swap limit >= 1. The default swap limit is 2 for triple-buffering on XOrg 1.12+, 1 for double-buffering on older servers, as we can't change the swap limit there. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 820916e722db2997a53128f2f370d14a02d4401c Author: Mario Kleiner Date: Thu Feb 16 00:45:17 2012 +0100 dri2: Implement handling of pageflip completion events. Requests pageflip completion events from the kernel. Implements pageflip completion handler to finalize and timestamp swaps. Completion handler includes a consistency check, and disambiguation if multiple crtc's are involved in a pageflip (e.g., clone mode, extendend desktop). Only the timestamp of the crtc whose vblank event initially triggered the swap is used, but handler waits for flip completion on all involved crtc's before completing the swap and releasing the old framebuffer. This code is almost identical to the code used in the ati/radeon ddx and intel ddx. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit 4da68cedbefef4f72c00bb088e179071490882c8 Author: Mario Kleiner Date: Thu Mar 1 19:11:14 2012 +0100 dri2: Fix can_exchange() to allow page-flipping on more mesa versions. can_exchange() returns false and thereby prevents page flipping on some drawables where page flipping would work fine. This due to non-matching drawable depths values between front buffer pixmap and back buffer pixmap, because front buffer pixmaps inherit the depth of the screen, typically 24 bits, whereas the depth value of back buffer pixmaps for a given RGB8 or RGBA8 visual depends on the mesa version in use, either 24 bits or 32 bits. Use bitsPerPixel instead of depth to decide if drawable is flippable. This will still catch really incompatible formats like 32 bpp vs. 16 bpp buffers. Tested for screen DefaultDepth 24 and also 30 bits (for RGB10 framebuffers) on NV-50. The problem was fixed in the same way in the ati & intel ddx. Signed-off-by: Mario Kleiner Signed-off-by: Ben Skeggs commit b0d7f4ddd94154491d04dcb9698d63d61e9240dc Author: Christoph Bumiller Date: Mon Jan 9 15:17:27 2012 +0100 dri2: CreateBuffer format may be 0, use depth in that case commit 3d9f6b32e0ad72dc8c7db07d039d87e83d8fdb1d Author: Ben Skeggs Date: Fri Dec 9 17:45:31 2011 +1000 nv50/xv: bump max dims to 8k Untested, but I don't see why this wouldn't work... commit 57c62228e2e67370feeeb421c369fbf940d30ca4 Author: Ben Skeggs Date: Fri Dec 9 17:44:07 2011 +1000 xv: stop using custom m2mf code commit ad402333157c437ccd2c639f5b3f3bcb81e78d6c Author: Christoph Bumiller Date: Sat Dec 10 13:32:24 2011 +0100 dri2: fix allocation of Z16 depth attachments commit 3d2a752ca8aafee5e1e94dabfd7deec439890e95 Author: Ben Skeggs Date: Thu Dec 1 11:31:49 2011 +1000 move away from libdrm's BEGIN_RING (and variants) macros subchan auto-assignment is now gone. commit b6cfdc01aeb5c2ab401ee81fff66bd0560abdfcc Author: Ben Skeggs Date: Thu Dec 1 11:25:10 2011 +1000 Append _CLASS to object class id definitions I want to be able to use ring macros of the same name commit 1dadb5826363d0d8ed1a72209be18277550f8345 Author: Ben Skeggs Date: Thu Dec 1 09:06:24 2011 +1000 exa: create general m2mf rect copy for each chipset, use for UTS/DFS No point having duplicated UTS/DFS logic in each implementation. This also splits nv50 out into its own code, in preparation for it being made to not use relocs at all. Signed-off-by: Ben Skeggs commit b5534a15a5f5e1c3bd6c92e3ed8391ff628c09bb Author: Marcin Slusarz Date: Wed Nov 30 18:56:54 2011 +0100 fix typo in conversion to rnn headers commit 5975ce34a12d1943b2852d0f5fbe09358754c2ad Author: Ben Skeggs Date: Wed Nov 30 17:40:02 2011 +1000 convert entire ddx to rnn headers commit 9fa0c6c74d0b9f7d66ebe3c99184eaa13abeb1a0 Author: Marcin Slusarz Date: Sun Nov 6 20:32:10 2011 +0100 Default to 16bpp for low memory cards. commit 5aa30ae6554c957609929ef29d1b27c89e9a7f06 Author: Marcin Slusarz Date: Sun Nov 6 20:32:09 2011 +0100 nv50/exa: wait for enough space for at least one composite operation commit 3d54426c4833a7e058e50624452f89139921e64b Author: Marcin Slusarz Date: Sun Nov 6 20:32:08 2011 +0100 add missing return in nv50_accel_init failure path commit d575a28a5020ff083836d3d1cf50018a99f49c9f Author: Maarten Lankhorst Date: Sat Oct 8 10:33:20 2011 +0200 dri2: Add vdpau driver name entry libvdpau has a driver loading mechanism that looks for a dri2 driver first before falling back to nvidia, so lets use that. commit 656cfe3bdb8b487827b710a0cea35ae1bcd5044e Author: Marcin Slusarz Date: Sat Sep 10 23:14:53 2011 +0200 xv: support AI44/IA44 It's needed for osd in mplayer's xvmc output. commit 169512fbe91f0671a90dfee5e280357f0a4ef701 Author: Maxim Levitsky Date: Tue Aug 23 13:10:59 2011 +0200 dri2: Disable the "exchange" swapbuffers path for the moment (bug 35930). Allow page flipping only for scanout buffer for now as simple swapping between off-screen pixmaps confuses compiz because there is no syncronization method to tell it about that swap Signed-off-by: Francisco Jerez commit b806e3f97a73701f057c45b1f45233e69e19f113 Author: Bryan Cain Date: Wed Jul 20 13:48:36 2011 -0500 g3dvl: update for compatibility with latest pipe-video branch in Mesa Signed-off-by: Bryan Cain Signed-off-by: Ben Skeggs commit de9d1ba7efeba64f319efa00df183d3cb78f24af Author: Emil Velikov Date: Tue Jul 19 00:35:38 2011 +0100 drmmode_display: Resolve missing brackets Correct some missing/misplaced brackets in drmmode_pre_init() The issue was exposed when trying a 4 monitor desktop using two cards/gpus Resolves https://bugs.freedesktop.org/show_bug.cgi?id=39099 Reported-By: Damian Nowak Tested-By: Damian Nowak Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs commit ef4957492956df54fafa6c75d576a07ce678ea0a Author: Ben Skeggs Date: Mon Jul 11 12:48:34 2011 +1000 kms: fix multiple rotations in noaccel mode commit ab89aa022f0c78154e525b5f5787feee8024171f Author: Ben Skeggs Date: Fri Jun 17 10:10:09 2011 +1000 nvc0: accept 0xdx chipsets Signed-off-by: Ben Skeggs commit 7dd8987b8b69f56d04a199207122d2e51b9caa6e Author: Ben Skeggs Date: Fri Jun 17 10:07:36 2011 +1000 nvc0: assume 0x9097's presence on all chipsets This appears to be the case thus far, and is what the 3D driver does. Signed-off-by: Ben Skeggs commit 98ee2105dbd8685e59c1967a587fccb6135743ac Author: Ben Skeggs Date: Mon May 30 10:10:58 2011 +1000 kms: clear new fb to black on fb resize commit 8378443bd3b26b57ef2ae424a700e01ead813d33 Author: Ben Skeggs Date: Thu Mar 24 02:13:12 2011 +1000 nv50-nvc0/exa: fix bug causing surface state to not be reemitted after flush Signed-off-by: Ben Skeggs commit 92db2bc192a074d4b7fc3e9c16e9aa62a8deaeb6 Author: Christoph Bumiller Date: Thu Mar 3 22:17:23 2011 +0100 nvc0/accel: allow nvc1,nvc3,nvc4 in 3D engine init and use 9097 commit ace98a492353e6de712f4f717e6d3f562e3591f0 Author: Ben Skeggs Date: Tue Mar 1 13:54:38 2011 +1000 dri2: return an error rather than crashing if we can't allocate a buffer Signed-off-by: Ben Skeggs commit bc5dec2ca7ca7edc340a99bd73946e228117dfd8 Author: Ben Skeggs Date: Thu Feb 24 15:15:00 2011 +1000 dri2: disable page flipping if any crtc is rotated Signed-off-by: Ben Skeggs commit c123618470b5ff32c6377a4cae0f148d4a518a5d Author: Francesco Marella Date: Thu Feb 17 03:48:16 2011 +0100 Fix gcc 4.6.0 set but unused variable warnings Signed-off-by: Francesco Marella Signed-off-by: Ben Skeggs commit 3531b2aa13e9c78288d96fe5e1492f0b6e899b00 Author: Francesco Marella Date: Thu Feb 17 03:47:46 2011 +0100 Page flipping man page update Signed-off-by: Francesco Marella Signed-off-by: Ben Skeggs commit 85a706df6f7986158c964e3afd0aa8e352ff5cea Author: Francesco Marella Date: Thu Feb 17 03:47:08 2011 +0100 Update configure.ac with autoupdate Signed-off-by: Francesco Marella Signed-off-by: Ben Skeggs commit 42c16ff33dcca6cd9bae62fe0d71a40452e03581 Author: Ben Skeggs Date: Wed Feb 16 13:51:54 2011 +1000 nv50: attempt to allocate compressed Z buffers This should automagically fallback to the non-compressed version on kernels that don't support compression. Signed-off-by: Ben Skeggs commit 46acb7e095e9c0d55b90070b77c64679fcfb9d4e Author: Marcin Slusarz Date: Sun Feb 13 20:46:42 2011 +0100 fix drmSetMaster error reporting On error, drmSetMaster returns -1 and sets errno. Signed-off-by: Ben Skeggs commit 54764980411b4b8124c36edb8730612b04f292a9 Author: Maarten Maathuis Date: Thu Feb 10 20:12:07 2011 +0100 Also flush from the BlockHandler. - The fact that commit 02c50db307a4e4eeb87f7db5d401bc53cce3b21f fixed missing text for some people proves this is needed. That commit should have only improved latency a bit. - The FlushCallback is still needed for DRI2 texture_from_pixmap. - See bugs.fd.o #33977 for discussion. Signed-off-by: Maarten Maathuis Signed-off-by: Ben Skeggs commit 02c50db307a4e4eeb87f7db5d401bc53cce3b21f Author: Maarten Maathuis Date: Mon Feb 7 19:37:47 2011 +0100 exa: Flush (S)IFC to the frontbuffer immediately. This avoids high latency while typing with core fonts for example. Signed-off-by: Maarten Maathuis Signed-off-by: Ben Skeggs commit e4fb9fd2f790ed66e4be61ca00d55522525619f4 Author: Ben Skeggs Date: Wed Feb 9 13:46:32 2011 +1000 dri2: allow page flipping to be disabled in xorg.conf Signed-off-by: Ben Skeggs commit 38e8809bb415bae5c182fc79c8fc62992c5e4ed0 Author: Ben Skeggs Date: Mon Jan 17 09:14:31 2011 +1000 nvc0: merge initial acceleration support Squashed commit of the following: commit e8ce1ef351ee51bec51211dca6fda88fdbfbefee Author: Ben Skeggs Date: Mon Jan 17 08:55:04 2011 +1000 fix hardcoding of nvc0 wfb path commit b414b1f948da18d96309bad84c54948d877efaa6 Author: Ben Skeggs Date: Mon Jan 17 08:52:53 2011 +1000 general tidy-ups in preparation for merge to master commit 73bf71d5de0cb990178a1d2a31e0ea1717c214a4 Merge: fb499a4 b795ca6 Author: Ben Skeggs Date: Mon Jan 17 08:47:40 2011 +1000 Merge branch 'master' into nvc0 commit fb499a4e9d95650dc89f4c1820b94d01344733f6 Author: Ben Skeggs Date: Mon Dec 20 12:54:43 2010 +1000 nvc0: switch to "standard" RING macros commit 64f0e1d0e9e29c5437ad4e52d01ccfb0b03a2ae9 Merge: eba7e17 c2092a4 Author: Ben Skeggs Date: Mon Dec 20 11:58:42 2010 +1000 Merge branch 'master' into nvc0 Conflicts: src/nv_dma.c commit eba7e173a5006ff50b950d0e8b8154564471a9ce Author: Ben Skeggs Date: Fri Dec 17 11:04:40 2010 +1000 nvc0/xv: use mad in nv12 shader, rather than mul+add Signed-off-by: Ben Skeggs commit b025ceb40a5e94a3657f3c7f76044c49a1a05e8d Author: Christoph Bumiller Date: Thu Dec 16 23:34:29 2010 +0100 nvc0: fix offset of 17bc buffer I had messed it up in 5e4da3451c3c0f645f98fea7fb0ce2629b1bfd2f. commit 6ff5a9a7f3411ef218f17fa481c583596ffde2cc Author: Christoph Bumiller Date: Thu Dec 16 01:07:25 2010 +0100 nvc0/xv: bind textures for the FP, not the TEP commit 5e4da3451c3c0f645f98fea7fb0ce2629b1bfd2f Author: Christoph Bumiller Date: Thu Dec 16 01:02:58 2010 +0100 nvc0: switch to rnn headers commit 0839041e79dbb0e629326d0f240220b33921db1a Author: Ben Skeggs Date: Wed Dec 15 16:42:41 2010 +1000 nvc0/exa: fix m2mf dfs commit 88c1c43302331e1cb8b7035bfb4677a5e7da3090 Author: Ben Skeggs Date: Mon Dec 13 12:00:23 2010 +1000 nvc0: add Xv implementation commit ad5b4cfe388a102a0c93f7281cb4bba532b1796b Merge: 1942a81 b6cb21b Author: Ben Skeggs Date: Wed Dec 15 14:57:04 2010 +1000 Merge branch 'master' into nvc0 commit 1942a81c0eb592972d4a3e8c46b80192ae8d1a59 Author: Ben Skeggs Date: Thu Dec 9 12:48:27 2010 +1000 nvc0: call grobj_alloc for all used object classes Though the hardware has no concept of object handles any more, the nouveau DRM still uses this call to know what engines are in use. commit 0c670a6d406fe0a57373fa9e03005be6f28ecc3e Author: Ben Skeggs Date: Thu Dec 9 12:44:41 2010 +1000 nvc0: don't try and create DmaNotifier0 commit 5d3d20d908a50ad7742567d58969c7f12dd944f7 Author: Christoph Bumiller Date: Thu Dec 9 01:36:32 2010 +0100 nvc0: change tile_mode to contain 0xZYX instead of 0xZY Warning: the kernel will not like that (EVO) commit dd7fd8263e0e9f92412fb20086e4c8acf8cab099 Author: Christoph Bumiller Date: Thu Dec 9 00:29:58 2010 +0100 exa: don't fall back to nv04 m2mf if nvc0 m2mf fails commit b4cf5a8f2276392945c2530f0d6fe96b5ab4b5da Author: Christoph Bumiller Date: Thu Dec 9 00:07:59 2010 +0100 nvc0/exa: remove noisy debug messages commit 5419ef6d2cfabb7329982cb8dbdb5bbc8e973fe3 Author: Christoph Bumiller Date: Wed Dec 8 23:55:51 2010 +0100 nvc0/accel: remove unneeded scratch method 0x3420 init It was intended be read by a MACRO method to get the address for constants upload. commit 178356391ff6831599eba257a5912079894641a1 Merge: 8573280 8bb8231 Author: Christoph Bumiller Date: Thu Dec 9 01:31:40 2010 +0100 Merge remote branch 'origin/master' into nvc0 Conflicts: src/drmmode_display.c src/nouveau_exa.c src/nv_driver.c src/nv_type.h commit 8573280871401e29f27fa24d0b7c6ac5e35b2cc1 Author: Christoph Bumiller Date: Sat Jul 3 14:27:36 2010 +0200 nvc0: initial implementation EXA should work, Xv not yet done. commit b795ca6e97fae9735843748585401098dae1c3e1 Author: Ben Skeggs Date: Fri Jan 7 13:33:36 2011 +1000 nv50: fix pitch of linear scanout buffers Un-breaks NoAccel on these chipsets. Signed-off-by: Ben Skeggs commit dc89dac734167bcbc667b39ca6ee2043871a60bf Author: Xavier Chantry Date: Sun Dec 19 23:17:34 2010 +0100 nv04/exa: kill useless WAIT_RING Signed-off-by: Francisco Jerez commit b76b43928bf4dcf5ce3a4b58c7fefe780070c0d3 Author: Ben Skeggs Date: Tue Dec 21 16:25:23 2010 +1000 add back generic include of nouveau_pushbuf.h So places that FIRE_RING() still work. Signed-off-by: Ben Skeggs commit c2092a458401377f421d9bdfe1d97ce580b99508 Author: Ben Skeggs Date: Mon Dec 20 11:50:57 2010 +1000 include nv04_pushbuf.h for ring macros, rather than nouveau_pushbuf.h Signed-off-by: Ben Skeggs commit aa2821a42706ac7b69703d1869e2d00a4ced9f4b Author: Ben Skeggs Date: Mon Dec 20 11:53:53 2010 +1000 kill NVSync(), it's rather useless Signed-off-by: Ben Skeggs commit 1fc564fe3494cf0abcc848d0e90bf2232f8fd272 Author: Ben Skeggs Date: Thu Dec 16 09:49:09 2010 +1000 nv50/xv: use mad in nv12 shader, rather than mul+add Signed-off-by: Ben Skeggs commit b6cb21b91f7d0409794a14ddbdd606728d4d49e3 Author: Ben Skeggs Date: Mon Feb 8 12:51:47 2010 +1000 set canDoBGNoneRoot if building against new enough server commit 11510f9ce8c9eab15e7f6d5962b83fc0052c8619 Author: Ben Skeggs Date: Tue Dec 14 15:39:01 2010 +1000 nv50/xv: implement brightness/contrast/saturation/hue controls Adapted from the code in xf86-video-ati. Signed-off-by: Ben Skeggs commit 214ea264cdb386520a1308a29f204e441f349cb1 Author: Ben Skeggs Date: Tue Dec 14 14:42:24 2010 +1000 nv50/xv: kick csc parameters out into a constant buffer These were previously inlined into the shader, which is simple, but makes implementing brightness/contrast adjustments impossible. Signed-off-by: Ben Skeggs commit 8bb8231236a6b877895663aeaa9cef731d67fe68 Author: Ben Skeggs Date: Tue Nov 30 15:27:36 2010 +1000 dri2: fix certain clients blocking forever when sync-to-vblank unavailable Signed-off-by: Ben Skeggs commit f29afe65b67c9b89ceba421d9b3aee590f7ded20 Author: Ben Skeggs Date: Mon Nov 29 09:23:35 2010 +1000 dri2: don't return failure from get_msc if sync-to-vblank isn't available Signed-off-by: Ben Skeggs commit 4b1d26578116f4d1745dd92a7c99413972a2bedf Author: Francisco Jerez Date: Thu Nov 25 21:33:46 2010 +0100 dri2: Don't return failure from schedule_wait if sync-to-vblank is unavailable. Reported-by: Ben Skeggs Signed-off-by: Francisco Jerez commit 1143e7a97dce1b741376e178b73b8b766e03849a Author: Jesse Adkins Date: Tue Oct 26 16:05:32 2010 -0700 Purge cvs tags. Signed-off-by: Jesse Adkins Signed-off-by: Francisco Jerez commit 22020457f43eceaf26532ecb50e46d743f4a691f Author: Francisco Jerez Date: Tue Oct 26 00:24:06 2010 +0200 dri2: Fix throttling on occluded drawables. Signed-off-by: Francisco Jerez commit 9737a812aacbc544b93eeaa031cf1c2e8adbd661 Author: Ben Skeggs Date: Mon Oct 25 10:37:08 2010 +1000 kms: initial pass at supporting zaphod-mode This works ok from some simple testing, there's undoubtedly things that will need polishing still however. Signed-off-by: Ben Skeggs commit d40c16f5497739c0215f37566db1651a74cb20ba Author: Francisco Jerez Date: Sat Oct 23 02:20:30 2010 +0200 dri2: Fix build on X server 1.8. Signed-off-by: Francisco Jerez commit a4d580bf05d7aac1d486e5d1ba9a2201c026e5f5 Author: Francisco Jerez Date: Fri Oct 22 23:09:30 2010 +0200 Bump the required X server version to 1.8. Signed-off-by: Francisco Jerez commit d0f2e16bc35660d01d3cd20a770cf8e613ab5323 Author: Francisco Jerez Date: Fri Oct 22 23:06:28 2010 +0200 dri2: Report damage on page-flipped drawables. Signed-off-by: Francisco Jerez commit c61e8d03161adcfbba4162172153814accf9e86b Author: Francisco Jerez Date: Fri Oct 22 23:05:24 2010 +0200 Make the sync-to-vblank helpers take a box as argument. Signed-off-by: Francisco Jerez commit a685a5c3be12f3e95a6ec1140af115e7ef28fcaa Author: Francisco Jerez Date: Wed Oct 13 18:45:04 2010 +0200 dri2: Implement the ScheduleWaitMSC/GetMSC hooks. Signed-off-by: Francisco Jerez commit eb83c830c87bce345748edef3b50660246143db7 Author: Francisco Jerez Date: Thu Oct 21 22:57:08 2010 +0200 dri2: Add pageflip/exchange support. Signed-off-by: Francisco Jerez commit c88f13e25b0040c1dd0f93e0ac40f62a6005ce59 Author: Francisco Jerez Date: Thu Oct 14 22:25:50 2010 +0200 Allocate a tiled scanout on pre-nv50 too. Pageflip between tiled and non-tiled surfaces isn't such a good idea. Signed-off-by: Francisco Jerez commit 4642d71bdffc04ce0dd458225541940c65f5443d Author: Francisco Jerez Date: Thu Oct 14 22:17:58 2010 +0200 dri2: Give out buffers ready for scanout. Signed-off-by: Francisco Jerez commit 75daada678d03d2e162f8003a6f2edcbe0525f92 Author: Francisco Jerez Date: Thu Oct 14 22:32:54 2010 +0200 Restructure tiled pixmap allocation in a single place. Signed-off-by: Francisco Jerez commit 70f0d2c886ceaa965ca4864788f4dd8e8f757a92 Author: Francisco Jerez Date: Thu Oct 21 22:55:02 2010 +0200 dri2: Add sync-to-vblank support. Note that you need a recent enough kernel for it to work. Use the "GLXVBlank" option to enable it (it defaults to disabled for now). Signed-off-by: Francisco Jerez commit 72dd4899ff036b09e3680ed2e09b3dbae3a7ba71 Author: Francisco Jerez Date: Thu Oct 21 22:31:45 2010 +0200 nv04-nv40/exa: Make the copy hook blit downwards for non-overlapping regions. Signed-off-by: Francisco Jerez commit 21ed642d3f76f286f23cd64ca817038ec94a1bf6 Author: Francisco Jerez Date: Wed Oct 13 18:22:22 2010 +0200 Make the sync-to-vblank helpers more consistent. Signed-off-by: Francisco Jerez commit 258e483d47f0f63155be18981c1118261f7675a8 Author: Francisco Jerez Date: Thu Oct 21 22:41:24 2010 +0200 kms: Register a wakeup handler to poll vblank/pageflip/udev events. Signed-off-by: Francisco Jerez commit 4f42708c26fa57bac7854701a6e757958c8d0654 Author: Francisco Jerez Date: Sun Oct 10 04:39:27 2010 +0200 Update to the last libdrm API. Signed-off-by: Francisco Jerez commit 8c8f15c9d3a31c35bf2ef20faf8c4133295a678f Author: Francisco Jerez Date: Sun Oct 10 04:36:17 2010 +0200 nv10-nv40/exa: Fix tiled pixmap pitch alignment calculation. Signed-off-by: Francisco Jerez commit a47c274f329b3b8bfbdcf155d6d832599d823dc8 Author: Francisco Jerez Date: Sun Oct 10 04:43:25 2010 +0200 Fix for the last xserver ABI. Signed-off-by: Francisco Jerez commit 9a5142fae6e1a0a470352f511427916e6fa01b9c Author: Maarten Maathuis Date: Fri Aug 27 21:11:59 2010 +0200 nv50/exa: Stop using wfb by default, but add an option to enable it. - Some scenarios, such as those heavy in core font rendering might prefer wfb. - Other (modern) situations are better of with wfb disabled. Signed-off-by: Maarten Maathuis commit d5a6f103113985c50bbe749ed341de34af1b47a5 Author: Francisco Jerez Date: Tue Sep 28 21:49:48 2010 +0200 dri2: Throttle frontbuffer flushes. There was no need for throttling until the hardware interchannel sync kernel changes because the pushbuf IOCTL blocked to sync the GL and X channels anyway. Signed-off-by: Francisco Jerez commit f1ac413d1d3dec2ccf63d4a8c79b9bd9ea578dcf Author: Ben Skeggs Date: Mon Sep 20 08:05:00 2010 +1000 nv50: use 0x8697 on NVAF Signed-off-by: Ben Skeggs commit 14c93acfc3fde04415e1c6a2a056465d15c86315 Author: Ben Skeggs Date: Fri Sep 17 13:17:37 2010 +1000 noaccel: fix rotation, as well as an assert during server shutdown Signed-off-by: Ben Skeggs commit 065576d5df0df4da61050d789f548801656b62d5 Author: Ben Skeggs Date: Thu Aug 26 13:22:33 2010 +1000 dri2: remove explit flush from copy_region() Unnecessary now that we kickoff from within the flush callback chain. Signed-off-by: Ben Skeggs commit 34025f0dc9284873d4bf7e4f547ed69fc562a3c9 Author: Ben Skeggs Date: Thu Aug 26 13:19:39 2010 +1000 kickoff from the flush callback chain This is a port of intel commit: 69d65f9184006eac790efcff78a0e425160e95aa Submit batch buffers from flush callback chain There are a few cases where the server will flush client output buffers but our block handler only catches the most common (before going into select If the server flushes client buffers before we submit our batch buffer, the client may receive a damage event for rendering that hasn't happened yet Instead, we can hook into the flush callback chain, which the server will invoke just before flushing output. This lets us submit batch buffers before sending out events, preserving ordering. Fixes 28438: [bisected] incorrect character in gnome-terminal under compiz https://bugs.freedesktop.org/show_bug.cgi?id=28438 Signed-off-by: Kristian Høgsberg Signed-off-by: Ben Skeggs commit 390f1c86d9583f4ceabd407ee47410073c2a1b55 Author: Ben Skeggs Date: Fri Aug 20 09:48:32 2010 +1000 nv50/xv: support UYVY Signed-off-by: Ben Skeggs commit 00d390952c912d4e9fc2c962caaeb90bf563d5b1 Author: Maarten Maathuis Date: Fri Aug 13 22:24:26 2010 +0200 nv50/xv: Fix the uncommon codepath to use width/height variable. - The common codepath should be the correct one, because that is triggered often, while the other is rare. - This isn't neccesarily a problem, but let's be on the safe side. Signed-off-by: Maarten Maathuis commit f5b486dd237679aec1987985f9ecffc6a4fd3105 Author: Maarten Maathuis Date: Fri Aug 13 22:16:51 2010 +0200 nv50/xv: We have 18 relocs, not 16, adjust MARK_RING. Signed-off-by: Maarten Maathuis commit 1cb6c91a9afe88d73f8a1180b00f56a9150a16ed Author: Roy Spliet Date: Sat Aug 7 20:59:06 2010 +0200 nouveau_exa.c: move line_count logic out of UTS/DFS loop Signed-off-by: Roy Spliet Signed-off-by: Maarten Maathuis commit 9f50b62229d5984ee5c16bc54953c7f8e2d5e575 Author: Ben Skeggs Date: Fri Aug 6 08:28:34 2010 +1000 nvc0: recognise these chipsets, and allow NoAccel mode on them Signed-off-by: Ben Skeggs commit b96170a4e51c1d3167601210ca31e98283c02885 Author: Francisco Jerez Date: Thu Aug 5 00:40:40 2010 +0200 nv10/exa: Bump maxX/maxY to 4096/4096. This way we can do accelerated 2D on larger pixmaps, it will still fall back to software when we hit the 2048x2048 texturing limit in a composite operation. commit 321eb2df5c8b1da4d6469037dd1d8888b508fabd Author: Francisco Jerez Date: Wed Aug 4 13:11:24 2010 +0200 Revert "nv04-nv40/exa: Match the blob behavior more closely on PrepareCopy." This reverts commit 9de0d97bd2fc2ee8800d48b5340a3d495525ad3e. Unnecessary after kernel commit "drm/nouveau: Ack the context switch interrupt before switching contexts.". commit db98ad23633958e52f84501c5a2061d42d346b64 Author: Francisco Jerez Date: Tue Jun 15 15:32:36 2010 +0200 dri2: Fix for the last DRI2InfoRec ABI. Just pre-fill the whole structure with zeros to avoid future segfaults. commit 9a3d0ab43f6335ab6cd48be4aa2f4f728386b386 Author: Ben Skeggs Date: Fri Jun 11 21:58:38 2010 +1000 call NVEnterVT from CreateScreenResources We need to grab DRM master here for things to work, so we may as well call NVEnterVT as we used to in NVScreenInit up until the previous commit. Fixes server regeneration. commit 964eeac6dc2209da6aa8861aa8b0bc028700ab3d Author: Ben Skeggs Date: Thu Jun 10 09:38:19 2010 +1000 fix for server 1.9 commit c079408e287ad09e954fa3985f95912c02cf06bc Author: Christoph Bumiller Date: Thu Jun 10 01:07:49 2010 +0200 nv50/accel: fix perspective interps in YUV to RGB shader commit 65153490f7d74d45a92adbd9c5e2bbe11c884b00 Author: Ben Skeggs Date: Wed Jun 9 16:12:28 2010 +1000 use libc directly for [cm]alloc/free commit 2958cf464dca9761f0fa21c8dd2d8fa8c8a96791 Author: Ben Skeggs Date: Thu May 27 10:06:47 2010 +1000 nv50/exa: rip out compat defines, we depend on 1.7 anyway commit af40bf0af47ebcb7de2dc418429f0b56f8c04d15 Author: Ben Skeggs Date: Wed May 26 16:11:01 2010 +1000 nv50/exa: new nouveau_class.h has correct blend func defines commit 8950c5b2a0d318e2568bec38a39d744b968c9b85 Author: Ben Skeggs Date: Wed May 26 16:09:48 2010 +1000 update nouveau_class.h commit 4b8f1a082069a17555d4142003213cf83d4e5464 Author: Lubomir Rintel Date: Mon May 10 09:25:55 2010 +0200 Handle startup without monitor attached more gracefully Refuse to configure a screen which has no valid modes. Otherwise X server would crash attempting to dereference to modes list later on. Signed-off-by: Ben Skeggs commit ac8f8d14a82ae8ba5b41fa88903edcad67f339c2 Author: Maarten Maathuis Date: Fri May 7 20:51:43 2010 +0200 wfb: use 28 bits instead of 32 bits for multiply_factor optimisation - Avoiding division in a hot code path is noticable even on an otherwise idle computer, but it's subject to rounding errors. If anyone knows a corner case and has better ideas, don't hesitate to contact me. For the moment this should make the problem less likely to trigger. Signed-off-by: Maarten Maathuis commit 13c10430ba8f7b0edff3ad8aae4a97672eea4a8f Author: Ben Skeggs Date: Fri Apr 23 14:09:35 2010 +1000 Bump version number to 0.0.16 Forgotten when the ABI broke a while back.. commit 6ff8ade0c7cd835b4172257ea310a8d88f28e757 Author: Ben Skeggs Date: Fri Apr 16 16:09:51 2010 +1000 Add uevent support When a hotplug event is received from the kernel we should notify the client side to reconfigure the display. Ported to nouveau from radeon driver, based on work by ajax in intel driver. commit 4063616938f76af8028491276039d422c0782b1b Author: Ben Skeggs Date: Fri Apr 9 20:43:51 2010 +1000 Import latest nouveau_class.h from renouveau commit bfb95cc0ca600135474ff02a99889b09d41b3c80 Author: Rico Tzschichholz Date: Thu Apr 1 20:39:23 2010 +0200 Fix "make dist". - remove README.NV1 reference from Makefile.am - add nv_rop.h to src/Makefile.am Signed-off-by: Francisco Jerez commit 2462b417fc550b71f021ca9736808f8f2d1cb6f0 Author: Ben Skeggs Date: Mon May 4 17:04:34 2009 +1000 xv: only use bicubic filtering when scaling >=2x commit e2146d3b29a4bea3d584c145e3190c3313692ed9 Author: Francisco Jerez Date: Thu Mar 18 14:32:54 2010 +0100 Fix for the latest nouveau_class.h. Signed-off-by: Francisco Jerez commit 7858345b6b384e88902ccc516715156401edec04 Author: Ben Skeggs Date: Tue Mar 16 13:08:37 2010 +1000 nv50: fix texturing from >=4GiB mark commit 6b8b15731b7610c9d8fb3625861797a0c9fd9546 Author: Xavier Chantry Date: Fri Mar 5 00:53:47 2010 +0100 Remove leftover in man page and nv_const.h Signed-off-by: Xavier Chantry Signed-off-by: Francisco Jerez commit c642b9f7a13bdeecd0a83ddcbf6d6d4f2c287501 Author: Francisco Jerez Date: Fri Mar 5 14:27:08 2010 +0100 kms: Don't try to copy fbcon contents when there's no framebuffer bound. drmModeGetFB complains loudly when it's given an invalid framebuffer id, bail out earlier if that's the case. Signed-off-by: Francisco Jerez commit efd7a526ca1cbac10b33b9d5d4cf3e542bc9b73b Author: Ben Skeggs Date: Thu Feb 25 09:32:49 2010 +1000 fix typo commit 70d0a48b6c3f1a817bf850acd3bae48d063e56b9 Author: Ben Skeggs Date: Tue Feb 23 15:08:13 2010 +1000 Protect against potential crash for unknown connector types commit 83bbeea0d842094dd647e754608420a404eef6e3 Author: Ben Skeggs Date: Tue Feb 23 15:03:38 2010 +1000 Add connector name for eDP commit 29647021044463768cbfa3eead1416ef1cf27fa1 Author: Ben Skeggs Date: Thu Feb 18 14:17:57 2010 +1000 remove drm patchlevel check, libdrm checks this for us commit 9b4118d6d1fa488db86cd7d2875beea9cdefb096 Author: Francisco Jerez Date: Wed Feb 10 18:43:39 2010 +0100 Revert "kms: work around some bong hits with dpms" This reverts commit 98c9e4edb58374f18249e5c9c53b392fb8b4a1d1. AFAIK it's no longer needed. Signed-off-by: Francisco Jerez commit f7deabfec0dcec4f203f85cd8494a70fe8fa6ebb Author: Ben Skeggs Date: Thu Feb 11 08:39:49 2010 +1000 Pass pScreen to drmmode_fbcon_copy() commit 43524fcbb786c5b9e9ec25776560790aec208719 Author: Francisco Jerez Date: Wed Feb 10 17:58:24 2010 +0100 dri2: Fix crashes on server regeneration. Signed-off-by: Francisco Jerez commit f50577c117c82eea05e750fecd1ff5653105c648 Author: Francisco Jerez Date: Wed Feb 10 13:23:46 2010 +0100 kms: Clean up the set_mode_major hook. The X server already takes care of restoring the previous CRTC parameters after a failed mode set. As a bonus this patch fixes RandR13 projective transforms. Signed-off-by: Francisco Jerez commit b16792b553dee41fc81d74c6e453832807e36368 Author: Marcin Kościelnicki Date: Tue Feb 9 20:51:17 2010 +0000 Remove some UMS leftovers. Signed-off-by: Francisco Jerez commit c56bce93f120d0ff18895149d1984896372cdd60 Author: Francisco Jerez Date: Wed Feb 10 12:31:54 2010 +0100 kms: Don't pretend it works on xserver-1.5. Signed-off-by: Francisco Jerez commit 0917665d1f2f1e76b6a0e7a4c027512f9f45f41b Author: Francisco Jerez Date: Wed Feb 10 11:48:47 2010 +0100 kms: Kill nv_cursor_convert_cursor. Well, mostly, the remaining pitch conversion will be unnecessary soon. Signed-off-by: Francisco Jerez commit e75dd23bd28f636b4c1759633d8dbc775e799add Author: Marcin Kościelnicki Date: Fri Feb 5 18:09:55 2010 +0100 nv50/accel: Fix mangled A8+A8 shader. Turns out we used a misaligned long instruction near the end, and the shader was getting killed after writing R, A components. This has gone unnoticed since the remaining G, B outputs aren't actually used. commit 41fc1fdec47fa824c598e1f3d828f4c8c4a2cb71 Author: Marcin Kościelnicki Date: Sun Jan 31 19:29:32 2010 +0000 Update to new nouveau_class.h Signed-off-by: Francisco Jerez commit c94544c83ef84e96f8d132ac9ff9296c22c4468e Author: Francisco Jerez Date: Sun Jan 31 18:26:03 2010 +0100 nv10/exa: Fix up nv11 accel. Fixes an off-by-one error made when UMS was removed. Signed-off-by: Francisco Jerez commit a397b94e32e6b0109096aba09bbe7b5dcfdd8231 Author: Francisco Jerez Date: Sun Jan 31 18:11:06 2010 +0100 dri2: Fix for the new DRI2InfoRec ABI. Signed-off-by: Francisco Jerez commit 8b5c67145a9c29cd4c82f411f15e57723d76d0c0 Author: Francisco Jerez Date: Sun Jan 31 17:43:54 2010 +0100 nv10/exa: Fix build after the last nouveau_class.h changes. Signed-off-by: Francisco Jerez commit 2630a1553d560ce2c7391bc103d4a543f7d3d0da Author: Marcin Kościelnicki Date: Wed Jan 20 07:46:32 2010 +0000 Remove nv_dripriv.h, nv_local.h commit 2d341f7e0a0dc70a28c6091ef6919fc6c0d60561 Author: Ben Skeggs Date: Thu Jan 21 14:58:37 2010 +1000 xv: use nouveau_bo_new_tile() commit 2a76ff1106ee8c76bd2e16a4954b3c05714d1db3 Author: David Heidelberger Date: Mon Jan 18 11:50:19 2010 +0100 Remove deprecated NV1 readme commit fe484a9b635ad56afec7be086a56c692b00dd6d1 Author: Ben Skeggs Date: Wed Jan 13 16:14:41 2010 +1000 exa: remove nouveau_pixmap_offset() Completely useless now we don't have to try and support non-driver pixmaps. Signed-off-by: Ben Skeggs commit bbf79469a878a8ab5b4c07faadcfdc8d85ceac83 Author: Ben Skeggs Date: Wed Jan 13 16:07:13 2010 +1000 exa: remove nouveau_exa_pixmap_{map,unmap} Signed-off-by: Ben Skeggs commit 18e3a649f7df73bc1e6d199ecf7a62ddcfe37d6e Author: Ben Skeggs Date: Wed Jan 13 14:26:04 2010 +1000 Check for drm+kms availability in NVPciProbe() Signed-off-by: Ben Skeggs commit 6267b85460ef364aa51a7d33006940bb00a049d9 Author: Ben Skeggs Date: Wed Jan 13 12:03:14 2010 +1000 bump xorg-server requirement to 1.7 Signed-off-by: Ben Skeggs commit da5c0dc6040096fd12ddb6a04a22fada89115ccb Author: Ben Skeggs Date: Mon Jan 11 13:23:06 2010 +1000 remove more unused code Signed-off-by: Ben Skeggs commit 104b3ea842d4cebd993d3e7e9d55ebcadae0cf28 Author: Ben Skeggs Date: Mon Jan 11 12:59:12 2010 +1000 exa: Remove non-driver pixmaps path Signed-off-by: Ben Skeggs commit e8798d91eedc5d8337223b2257ca607da8a197de Author: Ben Skeggs Date: Mon Jan 11 12:48:00 2010 +1000 Remove pNv->NVArch Signed-off-by: Ben Skeggs commit 7a6f785de1732ec0af429a0b66d5d6ba45427dcb Author: Ben Skeggs Date: Mon Jan 11 12:00:19 2010 +1000 Remove pNv->Chipset use Signed-off-by: Ben Skeggs commit 68c1f908be5739692c9f0bde9c23f3dc2a7210ea Author: Ben Skeggs Date: Mon Jan 11 11:56:19 2010 +1000 remove pNv->VRAMPhysical Signed-off-by: Ben Skeggs commit 06499197d6eb07a272a1812e1c8bce886798a1d7 Author: Ben Skeggs Date: Mon Jan 11 11:51:14 2010 +1000 dri1: remove DRI1 support Signed-off-by: Ben Skeggs commit 6975651b0b74dca6edaee759fb1003f65189f1df Author: Ben Skeggs Date: Mon Jan 11 10:07:11 2010 +1000 xv: ifdef out another problematic piece of overlay support for the moment commit 17485c234ff191cee3dd19e3dd693a80b024e189 Author: Ben Skeggs Date: Wed Jan 6 14:58:15 2010 +1000 Initial pass at removal of non-KMS support. It's likely there's a lot more that can be cleaned up/removed as a result of this, we can handle that as we come across it. Signed-off-by: Ben Skeggs Signed-off-by: Francisco Jerez Acked-by: Maarten Maathuis commit 061cb284b93612848296599759b915a4d66b6d01 Author: Maarten Maathuis Date: Sun Jan 3 02:14:36 2010 +0100 exa: Some compat defines for new pixman formats. Signed-off-by: Maarten Maathuis commit 8ca5a7383aad0e0557b3c9a1e9edd5431cd509fc Author: Marcin Slusarz Date: Wed Dec 30 21:46:04 2009 +0100 start using BEGIN_RING_NI commit b259b7ce5e57e1d20111878318b4703a14ec27ec Author: Maarten Maathuis Date: Fri Jan 1 20:16:13 2010 +0100 nv50: always set semaphore handle and offset - This is in preperation of using semaphores for more than one thing. Signed-off-by: Maarten Maathuis commit 8deeed25014b36f8d60f25c7ea2509d88fcb52ae Author: Marcin Slusarz Date: Mon Dec 28 18:39:16 2009 +0100 nv50: upgrade nv50_texture.h - copy changes from mesa/src/gallium/drivers/nv50/nv50_texture.h - remove 2 redefinitions - fix R<->B rename (see df189c9efc0fbcdce816af483f0147ab635280d1 (nv50: TIC/TSC fixes and additions) in mesa repo) commit 1877ee268774443a58a7bb73244bd70c82d67879 Author: Marcin Slusarz Date: Thu Dec 24 21:29:10 2009 +0100 nv50/exa: add support for more color formats commit 8b3753fc34e3557b91c3bdebe7c8addfb148b3bd Author: Christoph Bumiller Date: Wed Dec 30 18:28:25 2009 +0100 nv50: check drawable.bitsPerPixel instead of depth to get format commit 920a346e14cdc40efe9e89c3583d8e573e7e2f87 Author: Marcin Kościelnicki Date: Mon Dec 28 16:11:51 2009 +0000 nv50: Dehexify and bring up to date with new method defines. Signed-off-by: Francisco Jerez commit 7c72ac1f47e19af97c997f57012e3af12f941894 Author: Francisco Jerez Date: Sat Dec 26 15:56:58 2009 +0100 nv10/exa: Fix for the latest nouveau_class.h changes. Signed-off-by: Francisco Jerez commit 6992d0e7a0cb3c32b16af0b724246e44f7a35d7e Author: Francisco Jerez Date: Thu Dec 17 19:13:08 2009 +0100 dri: Use classic mesa on the fixed pipe cards. Signed-off-by: Francisco Jerez Signed-off-by: Ben Skeggs commit 87b9916dd4b471a9b8b02f166f6553f04f219769 Author: Maarten Maathuis Date: Wed Dec 23 21:53:36 2009 +0100 kms: Remove the one-off offset of connector names compared to the kernel. - Seeing DVI-I-0 in xrandr and DVI-I-1 in the kernel logs can be a source of confusion. Signed-off-by: Maarten Maathuis commit bb1947831d9a4e080b8d1e9dba086af6527ef479 Author: Ben Skeggs Date: Thu Dec 17 13:08:34 2009 +1000 Revert "exa: map the pixmap bo after creation to ensure a useable driver pixmap" This reverts commit 740c9f3e79729e4e06c693e66826afbad9bd1ce5. To get the same effect as this commit, update to latest libdrm git. commit 03c7d675f6d12886b6bca8e0375888ecf1b393df Author: Francisco Jerez Date: Fri Dec 11 15:47:05 2009 +0100 dri2: Make the render targets tiled for pre-G80 cards. Signed-off-by: Francisco Jerez commit 963aaeff6b414f371eff3e8ce016f7707c60f7c0 Author: Francisco Jerez Date: Fri Dec 11 15:44:47 2009 +0100 exa: Pre-G80 tiling support. For now, pixmaps will only be tiled if driver pixmaps is being used and we're told to with the NOUVEAU_CREATE_PIXMAP_TILED usage hint. Signed-off-by: Francisco Jerez commit 41d8d511124fcbe83a3cc68aaf22c3ce0fb4eb8a Author: Francisco Jerez Date: Mon Dec 14 00:51:17 2009 +0100 kms: Allow interlaced/doublescan modes. After bcafdfbed6 (an X server commit) we have to explicitly state we can handle them. Signed-off-by: Francisco Jerez commit 740c9f3e79729e4e06c693e66826afbad9bd1ce5 Author: Maarten Maathuis Date: Sat Dec 12 22:27:30 2009 +0100 exa: map the pixmap bo after creation to ensure a useable driver pixmap - Large pixmaps (think of 8000x8000 pixmaps on NV50) will cause crashes otherwise because exa assumes driver pixmaps to be mappable. Handling this elsewhere is problematic. Signed-off-by: Maarten Maathuis commit 0b5120cf8ef2528a15d88054afdc6fc33eb521f5 Author: Ben Skeggs Date: Mon Dec 7 15:31:53 2009 +1000 Allow texturing from GART Unfortunately NV_CONTEXT_SURFACES_2D on most cards (probably only PCIE cards work actually, but that's not confirmed) rejects a GART DMA object, so the usefulness is a bit more limited than I'd hoped. commit 60a394222b410da8bb832935fc525132218b4d94 Author: Ben Skeggs Date: Thu Dec 3 21:14:00 2009 +1000 nv50: avoid touching 0x16b0 on 0x8597, it causes a DATA_ERROR from the GPU Signed-off-by: Ben Skeggs commit 95fc663d5fb653fc0a50aa4d589bbf3b9c79e928 Author: Ben Skeggs Date: Thu Nov 26 13:10:45 2009 +1000 shadowfb: fix segv on fb resize Signed-off-by: Ben Skeggs commit 1cf0cfa926dd78698ff8b8c2d776bcdaa6a867bd Author: Ben Skeggs Date: Thu Nov 26 10:04:43 2009 +1000 kms: zfill fb if we have no accel to do copy, better than random garbage commit 92578fb4db0feb549bea82d0874ba0e39a2b1ae2 Author: Francisco Jerez Date: Thu Nov 5 18:30:28 2009 +0100 nv10/exa: Rename VIEWPORT_SCALE to VIEWPORT_TRANSLATE. Signed-off-by: Francisco Jerez commit 9de0d97bd2fc2ee8800d48b5340a3d495525ad3e Author: Francisco Jerez Date: Thu Nov 5 13:29:09 2009 +0100 nv04-nv40/exa: Match the blob behavior more closely on PrepareCopy. Fix a somewhat indeterministic corruption problem on nv17 when there is stuff going on the other fifos (e.g. gallium but I've also reproduced it with an app just SIFM-ing memory around): in some cases it made the blits the X server had scheduled fail corrupting the nearby screen areas. Signed-off-by: Francisco Jerez Reviewed-by: Ben Skeggs commit 16ef9ff7e394a693fc74764d8bc2630e784947f3 Author: Francisco Jerez Date: Thu Nov 5 03:28:00 2009 +0100 nv10/exa: Spring-cleaning * Kill the A8+A8 hack. Recent enough X servers (>=1.7) fall back to ARGB glyphs for drivers not supporting A8 render targets. * Kill all the global state. It doesn't matter a lot yet but it might if we get multicard working at some point. * Other random clean-ups with no functional changes. Some numbers from x11perf -aa10text -aa24text -comppixwin10 -comppixwin500: * Before, with A glyphs: 12000000 trep @ 0.0025 msec (394000.0/sec): Char in 80-char aa line (Charter 10) 3200000 trep @ 0.0119 msec ( 84300.0/sec): Char in 30-char aa line (Charter 24) 8000000 trep @ 0.0036 msec (280000.0/sec): Composite 10x10 from pixmap to window 60000 trep @ 0.4813 msec ( 2080.0/sec): Composite 500x500 from pixmap to window * After, ARGB glyphs: 20000000 trep @ 0.0015 msec (649000.0/sec): Char in 80-char aa line (Charter 10) 4800000 trep @ 0.0057 msec (175000.0/sec): Char in 30-char aa line (Charter 24) 8000000 trep @ 0.0036 msec (280000.0/sec): Composite 10x10 from pixmap to window 60000 trep @ 0.4813 msec ( 2080.0/sec): Composite 500x500 from pixmap to window Signed-off-by: Francisco Jerez Tested-by: Pekka Paalanen commit e1c2efd7e7703f5860e5fe99e0413e3be10576d4 Author: Ben Skeggs Date: Thu Nov 5 13:04:17 2009 +1000 kms: don't update drmmode_output->mode_output when vtSema is FALSE Fixes fast user switching issues. Signed-off-by: Ben Skeggs commit aed811e71b544dccfe84671f962503342acb75ab Author: Ben Skeggs Date: Wed Nov 4 18:03:12 2009 +1000 exa/nv10: handle reloc failures commit 5315034b7d46980a3527c814526857aed3bc98f7 Author: Ben Skeggs Date: Wed Nov 4 16:52:02 2009 +1000 xv/nv04: handle reloc failures + fail on bad surface format commit 99b1bf3d3e6b8c8ef2aa4aae3e33f3908cd00dab Author: Ben Skeggs Date: Wed Nov 4 16:46:35 2009 +1000 xv/nv40: don't even try with unknown surface format commit c1447574a025416f92f8dec2c361859c3945db92 Author: Ben Skeggs Date: Wed Nov 4 16:46:18 2009 +1000 xv/nv40: handle reloc failures commit 6ca14ba77104542a2915af8dca56236646be139a Author: Ben Skeggs Date: Wed Nov 4 16:41:58 2009 +1000 xv/nv30: don't even try with an unknown surface format.. commit 53ef7be58a2c46b2b5ffe912dcd408d74705ccbf Author: Ben Skeggs Date: Wed Nov 4 16:41:44 2009 +1000 xv/nv30: handle reloc failures commit 6facc0f765c1130aeb19ddfb76cbaaabaecf3f4a Author: Ben Skeggs Date: Wed Nov 4 16:36:09 2009 +1000 exa/nv40: handle reloc failures commit a97501a4e64d690be4232b01f9a3663fc2338c8a Author: Ben Skeggs Date: Wed Nov 4 16:31:56 2009 +1000 exa/nv30: handle reloc failures commit 222d13a03b41eb610a05063f81f0c47f57e6098e Author: Ben Skeggs Date: Wed Nov 4 16:02:50 2009 +1000 exa/nv04: handle reloc failures commit bd8aa7515f69c412e8f379777c49d59364df32fe Author: Ben Skeggs Date: Wed Oct 28 12:59:59 2009 +1000 xv: handle reloc failures commit 0214cb8b25a6959573bfc183b259b435f0a728bd Author: Ben Skeggs Date: Wed Oct 28 12:55:25 2009 +1000 xv/nv50: handle reloc failures commit e84e98d7f9260006a8b16a5785e4d2926d2370d4 Author: Ben Skeggs Date: Wed Oct 28 12:47:49 2009 +1000 exa/nv50: handle reloc failures commit 62c5c875e4c0812815be48f74bdef25d9f046515 Author: Ben Skeggs Date: Wed Oct 28 12:30:24 2009 +1000 exa: handle reloc failures commit c839f6098451aa59c7cb1fd7819e17a966e8ace7 Author: Marcin Slusarz Date: Sun Nov 1 22:16:50 2009 +0100 nv30/exa: support more 16bpp formats Signed-off-by: Marcin Slusarz commit 1d93e207db6e5a58b702c1812feaeb6b7252bc70 Author: Marcin Slusarz Date: Sat Oct 31 20:24:52 2009 +0100 nv/exa: fix 15/16 bits solid fill after this change nouveau passes all fill and blend tests of rendercheck (before: fill - 108/120, blend - 3323868/3569150) tested on NV34 Signed-off-by: Marcin Slusarz commit df5f17e59a7f42781b388074798d26f7981b2296 Author: Arthur HUILLET Date: Mon Nov 2 15:56:56 2009 +0100 nv10exa: improve codestyle Signed-off-by: Arthur HUILLET commit 5587f40c1b8af7d178f9a68d0f1fecdfc0ca9749 Author: Ben Skeggs Date: Fri Oct 30 14:01:09 2009 +1000 exa: force the use of sysmem pixmaps on low-mem cards Very similar effect to forcing MigrationHeuristic "greedy" on classic EXA. Far better than the migration ping-pong that'd occur otherwise Signed-off-by: Ben Skeggs commit 718a41b9a2b5649ac37c18f456b94eaf99584598 Author: Ben Skeggs Date: Thu Oct 22 12:35:32 2009 +1000 nv50: NVA8's 3D object class is 0x8597 commit 7494b18bf4061fb960a2126c745721161baddd01 Author: Francisco Jerez Date: Sun Oct 18 19:45:02 2009 +0200 nv50/crtc: Don't perform the LUT expansion twice. Signed-off-by: Francisco Jerez commit b161a7f7fa81b6d499c91c17f988a38ed537ed35 Author: Francisco Jerez Date: Sun Oct 18 19:41:11 2009 +0200 nv04-nv40/crtc: Don't perform the LUT expansion twice. Fixes bug 24595. Signed-off-by: Francisco Jerez commit 9cbc139176ae991150bd9ddf66b61db471b391f2 Author: Francisco Jerez Date: Sun Oct 11 23:42:01 2009 +0200 Partially revert "kms: always set gamma on modeset" 6c8d3e39b72abe897b08d9e90cae85e121863d2c breaks gamma setting with servers older than 1.7. Keep the call to xf86HandleColormaps even with KMS enabled, as it's a no-op in the cases we don't need it. Signed-off-by: Francisco Jerez commit cef4b61b4d0cf096428d89eff65a78fd16f612d5 Author: Francisco Jerez Date: Sun Oct 11 20:23:45 2009 +0200 Fix NVLoadPalette for 15/16 bit depth. Signed-off-by: Francisco Jerez commit cc4f7a733e3504260437d06091490fcb8a4c8063 Author: Francisco Jerez Date: Sun Oct 11 20:31:11 2009 +0200 Remove an useless call to NVLeaveVT from NVCloseScreen. Signed-off-by: Francisco Jerez commit 5a4afade51fe5fd149f8726fc243fbb69a660bf3 Author: Francisco Jerez Date: Thu Oct 8 03:33:34 2009 +0200 nv10/exa: Replace "celcius" with "celsius". Signed-off-by: Francisco Jerez Tested-by: Pekka Paalanen commit 2e744325827d702b564c5d6becaf65f3ad164761 Author: Francisco Jerez Date: Sun Oct 11 17:40:56 2009 +0200 nv10/exa: Allow component-alpha masks. Signed-off-by: Francisco Jerez Tested-by: Pekka Paalanen commit c3bb746ab6b371e843e1754a919dd25fa7b96b19 Author: Francisco Jerez Date: Sun Oct 11 17:38:44 2009 +0200 nv10/exa: Fix PictOpAtopReverse for destination pixmaps with no alpha channel. Signed-off-by: Francisco Jerez Tested-by: Pekka Paalanen commit f0a3d9f59817ebf2b9606097785f0ad1ae45da12 Author: Francisco Jerez Date: Sun Oct 11 16:15:30 2009 +0200 nv10/exa: Refactor the reg combiner/alpha blending code. Signed-off-by: Francisco Jerez Tested-by: Pekka Paalanen commit 3f020b0bf2e8d2bc54eceda771829c4b2535688a Author: Ben Skeggs Date: Thu Oct 8 16:19:16 2009 +1000 kms: reload cursor on modeset, needed on rotation (rh#524632) commit 38d7520986078fd01c19957a75ebfd9d91665225 Author: Ben Skeggs Date: Thu Oct 8 11:34:20 2009 +1000 kms: switch to CP+MPH instead of GetScratchPixmapHeader commit 6c8d3e39b72abe897b08d9e90cae85e121863d2c Author: Ben Skeggs Date: Wed Oct 7 15:57:49 2009 +1000 kms: always set gamma on modeset fixes issues vt-switching between console and X of different depths. commit b13c0509be63f27a4a9c918112d0a0fc84ccce10 Author: Christoph Bumiller Date: Wed Sep 16 18:17:30 2009 +0200 nv50/exa: use A8 RT format for PICT_a8 Signed-off-by: Ben Skeggs commit 7a63e4c61cd5d082caf074702cde79e1a66354b0 Author: Christoph Bumiller Date: Tue Oct 6 11:58:39 2009 +0200 nv50/exa: cleanup shader settings Since we only pass through 2D coordinates in the VP, no need to occupy extra output registers for HPOS z and, which can be coded as 0, 1 into VP_RESULT_MAP. We don't have to fetch all 4 attribute components if we only read 2 in the VPs. Using 0x1, 0x1 as terminator instructions looks odd, just make the last ones that do something long and set the exit bit. Setup VP_RESULT_MAP and INTERPOLANT_CTRL correctly. Signed-off-by: Ben Skeggs commit 60818bdeb975a9f469d3ae890932d4de1cbb59f9 Author: Ben Skeggs Date: Wed Oct 7 10:49:00 2009 +1000 kms: don't copy fbcon if depths don't match commit dd8339fc4e18fa4fd409f2997db741be209e852d Author: Ben Skeggs Date: Tue Sep 29 10:52:53 2009 +1000 Pass NULL fb pointer to fbScreenInit when using EXA Fixes 1.7 xserver getting very confused... commit 6dbdc80016c2a81c390d77db613833cd4d15c617 Author: Ben Skeggs Date: Tue Sep 29 09:00:59 2009 +1000 exa: make tiled scanout a separate condition from wfb being used commit 96d025232e292a8b1aebd54eaad3a21dbf6e2f77 Author: Maarten Maathuis Date: Fri Sep 25 19:50:11 2009 +0200 nv40/exa: fallback on special case xrgb non-repeat textures commit ed05bd6dab4fdd1289947154df02cd21c0839574 Author: Maarten Maathuis Date: Fri Sep 25 19:49:14 2009 +0200 nv30/exa: fallback on special case xrgb non-repeat textures commit dd7b86cb5351a60f21eee9cf81ae0ef61cf45f13 Author: Maarten Maathuis Date: Fri Sep 25 19:39:24 2009 +0200 exa: fallback on src pictures without drawable commit de0b095a1deaa1e53e1f555a6eb537482c488638 Author: Maarten Maathuis Date: Thu Sep 24 19:21:59 2009 +0200 nv50/exa: avoid fallback on xrgb repeat none when dst has no alpha - This should fix resizing slowness in evince amongst other things. commit df94ebdbcd89c1678ac243217e7f5b20cbbe857c Author: Ben Skeggs Date: Mon Sep 21 08:37:27 2009 +1000 kms: use displayWidth for pitch calc in transition, not virtualX (rh#523281) commit 5a6c3c427796f7f3dac0f597a320b09feb25ffc8 Author: Ben Skeggs Date: Wed Sep 16 15:13:08 2009 +1000 dri2: call CreatePixmap rather than yet another custom-built one commit 95086de507df7c5676c7ee712139ac4c91c9f6d2 Author: Pekka Paalanen Date: Tue Sep 15 20:31:33 2009 +0300 Partially revert "kms: wait for fbcon to be copied before continuing" This partially reverts commit 6c045fc44783454180d7b3d00b5f25436bd5544e. The nv_driver.c hunk of the above patch breaks the build on at least git and 1.6.3.901 xorg-servers due to: nv_driver.c:745: error: 'struct _ScrnInfoRec' has no member named 'canDoBGNoneRoot' Signed-off-by: Pekka Paalanen commit 6c045fc44783454180d7b3d00b5f25436bd5544e Author: Ben Skeggs Date: Mon Sep 14 13:04:12 2009 +1000 kms: wait for fbcon to be copied before continuing commit 22971d5545a2cd03e5c0b5d5d16fb85d7fff2aff Author: Ben Skeggs Date: Thu Sep 10 16:05:15 2009 +1000 dri2: switch to bo_new_tile() commit 806eaf6b0b36cb05ca9d883ff4572629812a1d48 Author: Ben Skeggs Date: Thu Sep 10 12:00:44 2009 +1000 kms: rework fbcon transition This does 2 things. Most importantly, it avoids a situation where we get multiple GEM handles for a single object, making it possible to confuse the kernel if both appear in the same push buffer. Secondly, it moves the transition to only on X startup, and not on every modeset etc. commit f90edc168c4a18b20750043d33c193bdc1b89156 Author: Ben Skeggs Date: Wed Sep 9 09:23:58 2009 +1000 Revert "nv50: use 0x7000 everywhere for now, until we can fix mixed mode alignment" This reverts commit c80f125a97003c2d0db77f54e22d9de447a06f63. This was not supposed to sneak in.. commit ff6bdb14ac13168c1a4b7866784e4833a1162dbc Author: Ben Skeggs Date: Wed Sep 9 09:19:42 2009 +1000 Revert "kms: fetch active encoder during init if there is one" This reverts commit 4f66fc6e871969448eba25817bfb383f2fdd4b34. See reason in previous commit. commit 9708e50e820970d526b084e0ab736c1c7a4f50c0 Author: Ben Skeggs Date: Wed Sep 9 09:18:48 2009 +1000 Revert "kms: update encoder config during detect" This reverts commit 25610704d4b16293f5bc24d5ad18cc76e20896d2. Very bad thinko, drm_mode_get_connector.encoder_id is *not* the same thing as our DRM's nouveau_connector.detected_encoder. Need to find a better way of achieving this. commit c80f125a97003c2d0db77f54e22d9de447a06f63 Author: Ben Skeggs Date: Tue Sep 8 16:20:33 2009 +1000 nv50: use 0x7000 everywhere for now, until we can fix mixed mode alignment commit 25610704d4b16293f5bc24d5ad18cc76e20896d2 Author: Ben Skeggs Date: Tue Sep 8 11:33:51 2009 +1000 kms: update encoder config during detect commit 4f66fc6e871969448eba25817bfb383f2fdd4b34 Author: Ben Skeggs Date: Tue Sep 8 11:26:29 2009 +1000 kms: fetch active encoder during init if there is one commit 2b5ec6ad708164eb9b2df00f453ea84f4898153a Author: Ben Skeggs Date: Fri Sep 4 10:35:57 2009 +1000 kms: only call drmModeSetCursor when cursor is visible It seems we get load_cursor_argb() calls even when the cursor is hidden, and the drmModeSetCursor() call would then decide a cursor is available and show it on outputs where the cursor shouldn't have been actuve. commit 569a17a98d03d965f9500c37c0cb14e0fb462311 Author: Francisco Jerez Date: Thu Aug 20 18:44:38 2009 +0200 kms: No need to upload gamma on set_mode_major in older X servers. Fixes build for CRTC ABI < 3. Signed-off-by: Francisco Jerez commit 32fc8a9a3104b38fe89427ffa60d21dad1572c15 Author: Francisco Jerez Date: Thu Aug 20 18:43:58 2009 +0200 kms: Fix commit e4e6f863b626d77934935e13714c6cbfd014d916 Author: Maarten Maathuis Date: Thu Aug 20 12:04:26 2009 +0200 kms: actually call kernel when loading cursor image - In the past we depended on getting a cursor show to call the kernel and actually make it visible. commit 95f09df23efd0a36d29cb4810faa659f839194f0 Author: Ben Skeggs Date: Thu Aug 20 15:48:26 2009 +1000 kms: prevent X from hiding cursor before updating image This wasn't noticeable when everything was being done in userspace. However, with KMS there's more of a delay between the hide<->update<->show steps and this produces a noticeable flicker. If there's any good reason we shouldn't be updating the cursor image while it's being shown, we'll handle it transparently in the DRM as required. commit fe2b5e6c6f81f4c4a76f5e4c5816812459872af3 Author: Maarten Maathuis Date: Wed Aug 19 22:06:42 2009 +0200 kms: fix gamma - crtc->active is needed in order to set the gamma manually - the initial gamma must be set by us commit 9d270f65ee85dedc6bf9cd9cf24e3d8df062f670 Author: Maarten Maathuis Date: Tue Aug 18 21:59:10 2009 +0200 exa: update to latest CreatePixmap2 + do not align width to 64 - Only the pitch needs to be aligned to 64. commit b50110d00e32396e186168271adbb88ace8ba986 Author: Francisco Jerez Date: Mon Aug 17 16:54:48 2009 +0200 kms: Some TV-out fixes. commit 523feba8fdc6d285929ac72be0a6e037019f019e Author: Maarten Maathuis Date: Mon Aug 17 00:09:42 2009 +0200 wfb: more correct and more error tolerant - In a strange world it is possible that a high index is tiled, and a lower index empty. - This should avoid that loophole and also emit a warning when we run out of indices. commit 3db3e5350e2f6dea50cd2bce0c5fb0437fe08dcf Author: Francisco Jerez Date: Wed Aug 12 02:18:27 2009 +0200 kms: Implement output->get_property when RandR1.3 is available. This allows getting updated values when the kernel is modifying them behind our back. commit ba9817b88d31364a5d43d945359e357535aeb09a Author: Francisco Jerez Date: Wed Aug 12 02:18:26 2009 +0200 kms: Don't hardcode the output properties Replicate any properties the kernel exposes. Mostly taken from the intel DDX. commit 92f271020a1b9486bd5890a00649aed3678ead62 Author: Ben Skeggs Date: Thu Aug 13 15:40:52 2009 +1000 nv50/xv: provide SYNC_TO_VBLANK attribute commit ada1d36d7097c389c8e21707d36b3cdd6378db94 Author: Ben Skeggs Date: Thu Aug 13 15:38:14 2009 +1000 xv: disable overlay with kms for the moment commit 85b1c86983a39cb921bd68a8601785975c948e2d Author: Ben Skeggs Date: Mon Aug 10 11:42:15 2009 +1000 dri2: call exaMoveInPixmap() before accessing driver pixmap private commit 292f46a0a0e031fc8a4256aa1eed91e3b2167acd Author: Ben Skeggs Date: Mon Aug 10 11:36:56 2009 +1000 xv: call exaMoveInPixmap() for driver pixmaps commit 0210fbf01d61fb0541770d541b71396537bcb728 Author: Christoph Bumiller Date: Sun Aug 9 14:53:28 2009 +0200 nv50/exa: fix pointer increment in UploadSIFC We always output 'size' dwords to the ring buffer, so the pointer should be advanced accordingly, independent of cpp. commit 1ac7d83c1e46a164c4a3a3688de1bcde9e5a58e0 Author: Maarten Maathuis Date: Sat Aug 8 19:41:02 2009 +0200 wfb: don't default to linear when encountering a pixmap without bo - With exa "mixed" there are actually very valid cases where this happens. - This fixes certain types of corruptions, like the yellow and black in a gimp window. commit 64790e950e7f46e6f4f5b8486f8d0bd929cb5504 Author: Maarten Maathuis Date: Fri Aug 7 16:12:41 2009 +0200 Revert "nv50: enforce certain allignments on a tiled frontbuffer, as well as overallocate it a bit" This reverts commit dd72640c8133ee78304f875023134f214d94a630. Moved to the kernel. commit dd72640c8133ee78304f875023134f214d94a630 Author: Maarten Maathuis Date: Fri Aug 7 11:56:29 2009 +0200 nv50: enforce certain allignments on a tiled frontbuffer, as well as overallocate it a bit - There is reason to believe that with driver modifications (part of) this won't be needed. - Until then this should fix "tiles" appearing at the upper left or lower right corner. commit bfbe4d33679cc25939db0da729d89ba2a4295764 Author: Ben Skeggs Date: Fri Aug 7 16:02:27 2009 +1000 exa: default to enabling driver pixmaps if everything else is new enough commit 30ddd08bb2739514edf6d6e137f533b34a4369d2 Author: Ben Skeggs Date: Fri Aug 7 16:01:27 2009 +1000 exa: turn on new exa migration mode for driver pixmaps commit 712064e6188d65e2c7d85078b95a0292edcad0e7 Author: Ben Skeggs Date: Wed Aug 5 14:00:51 2009 +1000 nv04-nv40/exa: set cliprect to drawing area for IFC commit 7f63da83d8622176114ecd55048736fecd703322 Author: Ben Skeggs Date: Wed Aug 5 12:04:46 2009 +1000 dri: close before dropping master commit 0cc44b713ec07dfb56a4b4a6d7ae1c1a75c74971 Author: Ben Skeggs Date: Wed Aug 5 11:57:01 2009 +1000 Set/Drop master in Enter/LeaveVT commit 087e9f61b81ea90a1591d8b43e29bda091156405 Author: Ben Skeggs Date: Wed Aug 5 11:51:54 2009 +1000 dri: close before taking down accel Fixes "failed to destroy server context" errors on exit. commit 96ac8c5a9473d81b737afa116f4fd388e3152f50 Author: Ben Skeggs Date: Wed Aug 5 11:36:39 2009 +1000 nv04-nv40: deal with framebuffer resize in !kms_enable mode commit b8ea3b41b0540635585bc68334691414f7cff53d Author: Ben Skeggs Date: Wed Aug 5 11:30:51 2009 +1000 nv50: deal with framebuffer resize for !kms_enable commit d0dcc89825acb4b480f3849d8896b17a80df02f4 Author: Ben Skeggs Date: Wed Aug 5 11:28:45 2009 +1000 No wfb without KMS commit 2c7502680e1ab3a4ea4869a78d0317746084dd92 Author: Ben Skeggs Date: Wed Aug 5 08:55:09 2009 +1000 dri2: fix CopyRegion() commit 2a4c93a064353cf61c30dd1e96f93203895beb15 Author: Ben Skeggs Date: Wed Aug 5 08:36:26 2009 +1000 dri2: merge DRI2BufferRec and nouveau_dri2_buffer into a single struct commit 4f386f983fde197e9136ffe6fbf941a343996897 Author: Ben Skeggs Date: Wed Aug 5 08:20:46 2009 +1000 dri2: remove support for < version 3 commit 0f691257e6d2c3662f9c76adc3287f5320a8822d Author: Ben Skeggs Date: Mon Aug 3 11:30:46 2009 +1000 zfill scanout buffer, not offscreen commit d893600356b3480d3df536a04e9860bfada5a159 Author: Maarten Maathuis Date: Sun Aug 2 19:43:58 2009 +0200 kms: remove fb in NVCloseScreen otherwise server regeneration fails - The bo was deleted, but fb_id remained non-zero, so it never picked up the new one. commit ac2a644773e5b160b6794a252b85c0bf8acc5169 Author: Ben Skeggs Date: Sun Aug 2 20:10:35 2009 +1000 xv: really fix offscreen pixmap calculation sleep.. good.. commit 60635e5073e69c6f690dc3c876772b5478013783 Author: Ben Skeggs Date: Sun Aug 2 19:42:17 2009 +1000 xv: fix check for offscreen pixmaps commit 1ecd4358c851b3d7bebfb00f5f7ac3c7c151bc20 Author: Ben Skeggs Date: Fri Jul 31 15:30:08 2009 +1000 Recalculate displayWidth if we do a late NoAccel fallback Fixes G80, where we'd have calculated the pitch for tiled scanout. commit 08ef8aaf2bd7b9565cc36679412b721499f9a367 Author: Ben Skeggs Date: Fri Jul 31 15:23:49 2009 +1000 Don't allocate offscreen memory or GART when using NoAccel commit 64a1515fab9764d0ada304426e7f0d7ec995b84e Author: Ben Skeggs Date: Fri Jul 31 13:49:49 2009 +1000 Fix NoAccel/ShadowFB with fb resize commit e9034a2e8d8285cf45cd1410b9e8933991f82a2b Author: Ben Skeggs Date: Fri Jul 31 13:29:11 2009 +1000 NoAccel + wfb == bad idea commit 69a40131b700a4fbc4de9432110045870de0806e Author: Ben Skeggs Date: Fri Jul 31 13:28:02 2009 +1000 Fix up screen pixmap's devPrivate.ptr after resize commit 4ea65c594a1f10f586da2ca32e499de2ea0a2605 Author: Ben Skeggs Date: Fri Jul 31 13:25:45 2009 +1000 Point fb at the correct buffer.. commit c6b98608daec4162a7d23ce8c1d50a4a9c3ca3dc Author: Ben Skeggs Date: Fri Jul 31 13:09:56 2009 +1000 Switch on can_resize for non-kms commit 4bb007ebb8e60c87c815f9c4ce63b8d1d46e8b75 Author: Ben Skeggs Date: Fri Jul 31 13:03:45 2009 +1000 Fall back to NoAccel if channel/object creation fails commit 641fdacbd1aeb9882caa30da6d568ff9a0bd78f5 Author: Ben Skeggs Date: Fri Jul 31 11:46:47 2009 +1000 nv50/exa: use scissored triangle for composite too commit c6c0ef1b5a837838a0fc2e8d1659b1b4d76d26b0 Author: Ben Skeggs Date: Fri Jul 31 11:23:04 2009 +1000 nv50/xv: enable sync-to-vblank commit b3d6861fa482a92807aadd9222ca2c8c605beaeb Author: Maarten Maathuis Date: Thu Jul 30 17:23:25 2009 +0200 nv50_xv: switch to a scissored triangle to avoid strange artifacts when vsync is enabled commit 9b37bc96d86882b7f9c355e92cbeff517090a98d Author: Ben Skeggs Date: Thu Jul 30 19:24:19 2009 +1000 more cleanup commit d8c10eb14f8d6b5cad88a39a5da7104a403060cd Author: Ben Skeggs Date: Thu Jul 30 10:17:38 2009 +1000 non-kms fb resize commit 3b1a861aa56f823e8d26f33b7139e290bf44b79d Author: Ben Skeggs Date: Thu Jul 30 07:14:17 2009 +1000 free scanout buffer in NVUnmapMem() commit 4d20547df30d630b94b0cfc28dcf7ead564f7369 Author: Ben Skeggs Date: Tue Jul 28 18:14:03 2009 +1000 nv50: tile scanout buffer for "classic" exa if xserver new enough commit 03bc3a604e5c12829f81f2582374d37012da7f36 Author: Ben Skeggs Date: Tue Jul 28 11:45:51 2009 +1000 kms: only use rounded height for allocation Fixes minor cursor corruption in some circumstances with driver pixmaps. commit e4fb3e7064df14b21a6ea34b919ccf2bf9fa0ca2 Author: Ben Skeggs Date: Tue Jul 28 11:27:09 2009 +1000 More fixing of pitches + use 0x7000 tile_flags for tiled 16bpp on nv50 commit 97cca7630441876414da5dfabb9ad9f2b208e29d Author: Ben Skeggs Date: Tue Jul 28 10:21:11 2009 +1000 kms: implement rotation support without driver pixmaps commit 786ff2994aa4f614b7cc9d466455eef183697d42 Author: Ben Skeggs Date: Tue Jul 28 09:43:00 2009 +1000 kms: simplify drmmode_is_rotate_pixmaps args commit ff62da668a468796e4001523958e73eb2060b7f2 Author: Ben Skeggs Date: Mon Jul 27 22:07:38 2009 +1000 dri1: dodgyness to "survive" a framebuffer resize (or rather, move) This can never work well without a lot of work, which I'm not willing to put in for DRI1. This will at least let it keep sort-of working, any DRI1 apps running when a resize event occurs will continue rendering into the old framebuffer. commit b77d580ebcc9b371153774010d29038645effffb Author: Ben Skeggs Date: Mon Jul 27 21:43:25 2009 +1000 We already map offscreen and store the virtual, no need to keep doing it commit bdf2c9aee02a100b44c083347a2415c100bd2546 Author: Ben Skeggs Date: Mon Jul 27 21:35:19 2009 +1000 kms: implement framebuffer resize without driver pixmaps commit 595b1c74e52f16ef482be6f14e4a33b4649b8bfd Author: Ben Skeggs Date: Mon Jul 27 07:44:41 2009 +1000 Fix build with 0.0.15 commit 9b879b9b38809e7d206c9e78610836bdd9c2d09b Author: Ben Skeggs Date: Mon Jul 27 21:09:42 2009 +1000 Cleanup commit 0336eb15ebc351ab6900b48972e12f5cc6abf1bb Author: Ben Skeggs Date: Mon Jul 27 17:52:15 2009 +1000 Switch to using scanout buffer detached from EXA offscreen memory area commit 0403cf3698ecb9a290bb4b7e24d0deedc957ad2a Author: Ben Skeggs Date: Mon Jul 27 17:45:29 2009 +1000 Allocate scanout buffer independently of EXA offscreen area This is just wasted memory for the moment... commit f8ac9e35bd20391f923c5c5ccc509251ea23a4c5 Author: Ben Skeggs Date: Mon Jul 27 17:30:54 2009 +1000 Remove NoDRM mode commit e259925bbdcebe4c3d136f476fc54f426eea9261 Author: Ben Skeggs Date: Mon Jul 27 17:14:00 2009 +1000 xv: fix some badness in nv3x/nv4x textured video adaptors commit 3a10daca16752b52ec3187deb8b1c25d6c6fd641 Author: Ben Skeggs Date: Mon Jul 27 16:05:03 2009 +1000 rename pNv->FB to pNv->offscreen Not strictly true right now, but it's about to be. commit 91bfda81982a77a79c2baee39929711b38b1ae6a Author: Ben Skeggs Date: Mon Jul 27 11:46:51 2009 +1000 exa: allow acceleration to rotation bo commit d59aaa81590b52b6227b69028f08d8ab62505525 Author: Ben Skeggs Date: Mon Jul 27 11:02:44 2009 +1000 exa: use PixmapIsOffscreen hook for non-driver pixmaps path too commit c442376d4e9a8296760c1171fee5ca155534d011 Author: Ben Skeggs Date: Mon Jul 27 07:44:14 2009 +1000 Fix typo exposed by removal of NV_* card type defines from DRM header commit 15bfc676094d19e57711bbdff40edd5c8e8592d6 Author: Ben Skeggs Date: Mon Jul 27 07:40:20 2009 +1000 Warning fixes commit 6eb790b6113a999727895694d01910b83348647a Author: Ben Skeggs Date: Fri Jul 24 13:54:49 2009 +1000 kms: keep rotate buffer pitch in crtc struct, rather than recalculating commit 25f29426293fcac2e7a17b986673b2874e7a7264 Author: Ben Skeggs Date: Fri Jul 24 13:18:22 2009 +1000 Wrap CreateScreenResources and set screen pixmap's BO there Avoids reallocating pNv->FB yet again when we hit the resize hook. commit d1628dc9159343400f09f696b9c4ef9d0fe1acbf Author: Maarten Maathuis Date: Tue Jul 28 10:08:40 2009 +0200 nouveau: remove some more RAC bits - The assumption is that RAC is dead anyway on post-libpciaccess xservers. commit 75fba5dfd2a00f3b8212d0e8af544273c14ee0c5 Author: Dave Airlie Date: Tue Jul 28 13:53:06 2009 +1000 nouveau: rip out resources code whats left at least commit 094e4e075f8e48ace534a6f011f95b6ebf1abc3b Author: Ben Skeggs Date: Thu Jul 23 11:25:22 2009 +1000 kms: don't skip resize call if we don't have a backing bo for screen pixmap commit bff801941e23dfe3fb7449af5780ee477b10d092 Author: Ben Skeggs Date: Thu Jul 23 11:23:15 2009 +1000 Fix a DRM vs DDX confusion typo commit e9ebddae840789535acf56ac2281b3fde3cfee96 Author: Maarten Maathuis Date: Wed Jul 22 12:51:08 2009 +0200 Fix build without AM_SILENT_RULES. commit f70287cf7cacc03d8390376519165d50a325cec5 Author: Peter Henriksson Date: Mon Jul 20 23:56:13 2009 +0200 Add support for AM_SILENT_RULES Make use of the AM_SILENT_RULES functionality in automake 1.11 Defaults to quiet output. Verbose output can be enabled with "./configure --disable-silent-rules" or "make V=1" Signed-off-by: Ben Skeggs commit 69ba1095a9aef29ce8fdb71ecf48f15cb13202c2 Author: Ben Skeggs Date: Tue Jul 21 16:29:13 2009 +1000 nv50/xv: attempt sync-to-vblank if driver pixmaps are being used commit 994c37927289de7a0ab6e951892f0d66880a6210 Author: Ben Skeggs Date: Tue Jul 21 14:05:59 2009 +1000 nv50: fix a number of tiled buffer sizing issues in kms paths commit 9c1eefd221366b9f5efbd30e4a28fba244347d4f Author: Ben Skeggs Date: Tue Jul 21 13:43:14 2009 +1000 Don't set driver pixmaps enabled with NoAccel commit 52882e82708f79ef198dde51f9bbebee752154be Author: Ben Skeggs Date: Tue Jul 21 13:19:07 2009 +1000 exa: fix nouveau_exa_pixmap_is_onscreen() for driver pixmaps commit 9fb28d107ad9fcd6a4bbb0a35c1fc0f719d531ae Author: Ben Skeggs Date: Tue Jul 21 12:52:35 2009 +1000 exa: remove ModifyPixmapHeader hook, it's not required commit 574fa36c73a1b04b67289762a10b5bbda7be083f Author: Ben Skeggs Date: Tue Jul 21 12:51:26 2009 +1000 exa: only support driver pixmaps on servers with CreatePixmap2 commit 365ab69499f466aeebac1437578609a7790f5dc4 Author: Ben Skeggs Date: Fri Jul 17 13:41:07 2009 +1000 remove explicit pin of pNv->FB on creation commit a757db1aadfd7a7d206dfc7968da8647dd91e65f Author: Ben Skeggs Date: Fri Jul 17 13:38:52 2009 +1000 no need to pin GART bo commit 474ebdeea1266959fddeabf04b24e04aaab8449e Author: Ben Skeggs Date: Fri Jul 17 13:37:29 2009 +1000 remove pinned creation from nodrm path commit b9333d5d34386d88b58d2267b388bbcc89b73f47 Author: Ben Skeggs Date: Fri Jul 17 13:35:20 2009 +1000 nv04-nv40: explicitly pin scanout buffer when required commit 53b5b68cd168b15c377f83e5e068cee92361f51d Author: Ben Skeggs Date: Fri Jul 17 13:31:13 2009 +1000 nv50: no need to pin rotate scanout buffers on creation now commit a57a9bb825e04bbe4aa94d9ac1149e2fca6525a4 Author: Ben Skeggs Date: Fri Jul 17 13:28:36 2009 +1000 exa: only allow driver pixmaps when KMS enabled The pre-KMS code can't deal with driver pixmaps in some circumstances, and rather than leave it there for people to accidently trip on, remove support. The other option would have been to fix the UMS code, but, who really cares? commit 5d7a0b954c0da879f81a6fa1a9dc8070c58a0fe6 Author: Ben Skeggs Date: Fri Jul 17 13:25:31 2009 +1000 remove remains of prehistoric NOUVEAU_EXA_PIXMAPS stuff commit 784a0337c3165c8d8fb696e4fa30bbdc8cf01b98 Author: Ben Skeggs Date: Fri Jul 17 13:18:06 2009 +1000 nv50: explicitly pin scanout buffer in non-kms path commit 09db6180e4205ef55516c24ac5dfc3bd861d9fd6 Author: Ben Skeggs Date: Fri Jul 17 13:12:25 2009 +1000 nv30: don't pin shader memory commit 7da55d99e3c93cadda50dcad06b09c52daa0cc5b Author: Ben Skeggs Date: Fri Jul 17 13:10:16 2009 +1000 xv: don't pin buffer unless using the overlay commit e8b9f689b778909e710fc5bd5616e6497b922a09 Author: Ben Skeggs Date: Fri Jul 17 13:07:04 2009 +1000 kms: don't pin scanout buffer, no need, kernel will do it commit b1b233099853b56cfdcff9c4c31d3550b76b949e Author: Ben Skeggs Date: Fri Jul 17 08:22:02 2009 +1000 Fix build with xextproto 7.1 commit c089a71a94e7887c405af7141e188302a5e2d393 Author: Ben Skeggs Date: Thu Jul 16 09:32:59 2009 +1000 dri2: fix for xserver abi change commit ef3a386a6df4569343e63332eb798c1d722fe164 Author: Ben Skeggs Date: Mon Jul 6 20:33:49 2009 +1000 Update manpage + finish removing useless options commit 4b6e0b96188e32cb2a13f8d2bd5973f0287c827b Author: Ben Skeggs Date: Mon Jul 6 20:25:23 2009 +1000 No need for pci access wrappers now.. commit 1e22673f43acd0f02ba36e189b39e28b635cd50c Author: Ben Skeggs Date: Mon Jul 6 20:16:02 2009 +1000 Remove non-XSERVER_LIBPCIACCESS paths, we depend on a new xserver anyway commit aafb62f86adebb3418876649e8bd189c1523e961 Author: Ben Skeggs Date: Mon Jul 6 20:08:03 2009 +1000 Some more bits of NVRec to go... commit 0ba588cb75af67a644f51add35778925d4088cf9 Author: Ben Skeggs Date: Mon Jul 6 19:58:15 2009 +1000 Rip out vgaHW + more unused vars from NVRec commit 3e22547b6826ae741b168c4b3391b6ade56b4829 Author: Ben Skeggs Date: Mon Jul 6 19:46:26 2009 +1000 Nuke nv_hw.c, and a heap of unused vars from NVRec commit eea5259d951a5919fa7800877570347c14411a32 Author: Ben Skeggs Date: Mon Jul 6 19:13:44 2009 +1000 Remove nv_dac.c commit a38b91a7a5c3337d405aa217830857832c5399e0 Author: Ben Skeggs Date: Mon Jul 6 19:09:09 2009 +1000 Cull everything in !pNv->randr12_enabled blocks commit 4815d312911cb4cfad4b4a5b194fce41d82f4869 Author: Ben Skeggs Date: Mon Jul 6 11:44:26 2009 +1000 kms: fix CRTC offset when rotation enabled commit 6d143272f497a61c5578752f58261befbfaf911a Author: Ben Skeggs Date: Wed Jul 1 14:03:06 2009 +1000 nv50: initial support for depth 30 commit 9b0919c3f3269b88fb338c043bbad34f20c22f41 Author: Ben Skeggs Date: Tue Jun 30 13:20:15 2009 +1000 kms: don't try transition when we don't have src+dst buffer ids commit c0bf670ac6b58cff60a01ab6b174ece6b1b7d892 Author: Ben Skeggs Date: Thu Jun 25 07:26:54 2009 +1000 wfb: fix stupid thinko + more safety commit a12cb5c8c0a85a601188178c5fc0d84a13abd87e Author: Ben Skeggs Date: Thu Jun 18 14:28:25 2009 +1000 bios: oops commit e66867e4cf9f0bc8a1971664ccc3d5c56b08b2fb Author: Ben Skeggs Date: Tue Jun 16 10:28:00 2009 +1000 bios: use image from PRAMIN in preference to PROM on NV50 There's at least one known case (rh#492658) where the DCB table present in the VBIOS image from PROM is not suitable for use. It contained all 16 entries filled, each entry valid in itself, but contradicting other entries. The VBIOS image in PRAMIN however, still has all 16 entries filled, but the first few entries now match what is present on the hardware, and the rest are set as type 0xf, which we ignore. commit 42c5730a4f6c243b75fbd40f1dc26e8433191c7b Author: Ben Skeggs Date: Tue Jun 16 09:48:39 2009 +1000 wrap BlockHandler a little earlier commit 09e663aea82403f69a0e44e3b7cd89dfe96970d5 Author: Ben Skeggs Date: Mon Jun 15 16:09:27 2009 +1000 kms: safer fb resize func commit 90be5d5d6a696564bda23e9a2dad14a3a1808868 Author: Ben Skeggs Date: Mon Jun 15 13:43:21 2009 +1000 wfb: use straight memcpy hook if no tiled wraps present commit 86d905d34a5693e3d9b986660d4a2bcd6f30a06a Author: Ben Skeggs Date: Mon Jun 15 13:41:21 2009 +1000 wfb: be more cautious in a few places commit 1bcbc4b26f71f90345bd82b836fc634e741aced8 Author: Ben Skeggs Date: Mon Jun 15 10:19:53 2009 +1000 kms: small cleanup commit c2d3550aafeb3e7c4df3334f0ac3554a8aa15da3 Author: Ben Skeggs Date: Mon Jun 15 09:16:10 2009 +1000 kms: fix rotation buffer pitch commit 6fd982979e03ce3eace7dc7523c428aca2c30b4a Author: Ben Skeggs Date: Mon Jun 15 09:00:02 2009 +1000 kms: point rotation pixmap at correct buffer commit 67f8ebe77f047c12d76e508b4d7f6384baed39ac Author: Ben Skeggs Date: Mon Jun 15 08:52:36 2009 +1000 nv50: create non-linear scanout buffers for rotation commit b7e3306ddc3693699f6f9de7b22913ee22ed31ed Author: Stuart Bennett Date: Sun May 31 00:26:13 2009 +0100 Add DVI-A output info for Mac card missing bios tables (#21273) The reporter indicates the card also has DVI-D and ADC, but those can't be tested to deduce the fake correct data at present commit 01b19c9fc25b147596e1719008ed9d861246ced9 Author: Ben Skeggs Date: Wed Jun 10 18:00:05 2009 +1000 nv50: sigh commit 4d9e63baff079fad727430fc605cb1e43c4303a7 Author: Ben Skeggs Date: Wed Jun 10 11:38:43 2009 +1000 wfb: "mode0" tiles have the same pitch as the others This appears to be the case now (as in: after making the GPU do some of the dirty work) even though it didn't in the previous wfb patches. commit 02ed6b66e37e13ab4064974b89db2e73613108ef Author: Ben Skeggs Date: Wed Jun 10 11:23:37 2009 +1000 wfb: work-around wfb suckage wfb sucks far more than anticipated: - can only expose a single set of rd/wr hooks, not per-wrap as wfb doesn't treat the hooks as per-pixmap. - the hooks will get called on buffers that aren't even pixmaps commit 669c59ecb69a6f84ef8b8714cda6e87eacb3de9f Author: Ben Skeggs Date: Wed Jun 10 09:15:37 2009 +1000 wfb: need a 64-bit datatype for multiply_factor commit 86dedb748cb6d047b195e50e921e840c1c403f6f Author: Ben Skeggs Date: Tue Jun 9 22:25:01 2009 +1000 fix some pitch issues commit 794a27724fb9ab173dc7ea8062730797e1b4de04 Author: Ben Skeggs Date: Tue Jun 9 10:54:54 2009 +1000 nv50: use libwfb for pixmap access when driver pixmaps enabled commit 11d9690dd9e1190d1f6873bb9a5df8f98a39d4f5 Author: Ben Skeggs Date: Tue Jun 9 09:10:57 2009 +1000 exa: use the sane CreatePixmap hook, if available commit 6ee453381f00063522187939cb54ab9447b73ec7 Author: Ben Skeggs Date: Thu Jun 4 14:43:54 2009 +1000 xv: fix some issues with driver pixmaps commit 67c5287f31bf20ab3a38ffcdee410d9d49ce4d2b Author: Ben Skeggs Date: Thu Jun 4 14:19:14 2009 +1000 exa: wait_marker becomes a NOP with driver pixmaps mapping/unmapping buffers will sync as necessary commit f44e52852a199196f8428822bf9fb71711b9c4f9 Author: Ben Skeggs Date: Thu Jun 4 14:08:10 2009 +1000 nv50: make use of larger tile sizes commit 30c44ce3142ba44fc7a6c95ceaa9bf028cc112c1 Author: Ben Skeggs Date: Thu Jun 4 13:41:36 2009 +1000 exa: create unacceleratable pixmaps (ie. 1bpp) in system memory commit 1aa22d5894c4e4efe746592bf1c51cdaf0c31c24 Author: Ben Skeggs Date: Thu Jun 4 13:20:37 2009 +1000 nv50: use non-linear scanout buffer when driver pixmaps enabled commit 13d8d49f74fb10f631bcd38b169068e691b9ca99 Author: Ben Skeggs Date: Thu Jun 4 13:11:40 2009 +1000 nv50: use tile_mode from bo should have no effect currently, but needed later commit bd9f5f2cbe7687aba7bb3a7f397d9b1bf87a8ef3 Author: Ben Skeggs Date: Thu Jun 4 12:11:48 2009 +1000 exa: fix compile against latest libdrm_nouveau commit 5f97afe38095a274938b8c2e9c5ebd17b7733bb6 Author: Stuart Bennett Date: Wed Jun 3 12:16:23 2009 +0100 randr12: split from pre-randr12 structs into new header, de-typedef New nv_crtc->state shortcut and some renaming included for convenience commit 0c17b8790086a529ee9e4b5aca4b6f19f31900a9 Author: Stuart Bennett Date: Thu May 28 22:40:30 2009 +0100 Move head getting into nouveau_hw.c commit 12314fa6abb2dd6d325a370724c7d16c371c229c Author: Stuart Bennett Date: Thu May 28 21:45:36 2009 +0100 randr12: improve uniformity of props code commit b541c1c4400942c59be37f62b7a25719dbdaa1a1 Author: Stuart Bennett Date: Thu May 28 18:23:37 2009 +0100 Remove "TMDS table script pointers not stubbed" bios warning Scripts appear to be init scripts (+7 one for 0x6808b?, +9 for 0x6828b?), which when the pointer is non-zero (and the script isn't 'q') get called at POST (at least on my nv4b). The warning, from a time when the scripts were thought to perhaps be needed for setting up outputs, seems redundant. commit 8af50289136e562198b5acc851629606d9ce450f Author: Adam Jackson Date: Thu May 28 15:08:35 2009 -0400 Remove useless loader symbol lists. commit 317b58166bf9f3189f33e99493cf8a1b40e014b8 Author: Stuart Bennett Date: Wed May 13 19:05:25 2009 +0100 randr12: disable lvds (invalidate all modes) if bios lvds parsing fails commit e897191c7bcf61193576ecf52d23fcddf5b70634 Author: Stuart Bennett Date: Thu May 28 16:42:29 2009 +0100 randr12: better behaviour (avoid crash) when fp native mode can't be found commit 03167483b41f7205dd1efc5b571ec73ec4b3cdb5 Author: Stuart Bennett Date: Thu May 28 02:18:52 2009 +0100 randr12: line length improvements commit b60c16cdcbeff2d02ea0866dfbb742c3880b7ae4 Author: Stuart Bennett Date: Wed May 13 02:21:17 2009 +0100 Simplify tests for digital fp outputs commit 992d4b5d3aa04f88c22e52e25dc2e3e7d9bd776e Author: Stuart Bennett Date: Thu May 28 15:59:12 2009 +0100 randr12: a bunch of trivial improvements and tidyups commit 6f0a324a503431d757b2247e7232de2eb7c318a5 Author: Stuart Bennett Date: Wed May 27 16:17:38 2009 +0100 Fix oopsy from 81bbdd4e causing broken framebuffer commit c5799186b1f70df0484c6cef63c8485ddf924733 Author: Stuart Bennett Date: Wed May 27 16:01:42 2009 +0100 randr12: fix two colour cursor on second head Colour caching was preventing cursor from being set correctly on second head commit 81bbdd4ee2e942d91d056a10ef993a8f41651c64 Author: Ben Skeggs Date: Tue May 26 15:48:25 2009 +1000 explicitly mark buffers mappable commit 6c09ad5296820a845571d07cbe42d88bb19560df Author: Stuart Bennett Date: Wed May 27 00:53:51 2009 +0100 Tolerate missing fp bios table (rh#502371) Modern desktop IGPs frequently miss it out, and now it seems even some laptop parts (MXM style) do too commit db9ff958291416dd5b72d20d32d7412ec45c1e60 Author: Ben Skeggs Date: Mon May 25 08:31:07 2009 +1000 kms: clip sw transition dims to smallest common area commit 2688c97169d8ea579ba09b9db049e31704f96626 Author: Ben Skeggs Date: Mon May 25 08:29:47 2009 +1000 don't run vbios parser when kms enabled commit 9656762ba186e91dd2b3b7f3f9427ba520982f9d Author: Ben Skeggs Date: Thu May 14 10:26:44 2009 +1000 nv50: fix multiple-display hangs when encoders swap crtcs commit d70eed91b2909ba0b20415ab1d5a099fb6b39b08 Author: Stuart Bennett Date: Tue May 12 16:54:33 2009 +0100 randr12: fix digital dpms regression introduced in 3971dda57004894d5d4fc9420aa00da400815af9 setting two different "blank" DPMS modes (say Suspend, followed by Off) consecutively would lead to the saved fp control value getting overwritten with FP_TG_CONTROL_OFF flavoured state, so DPMS on would then never work also, an improvement to LVDS DPMS state caching commit 10721038b854d9945ae551c6aa382b90f5d743e1 Author: Ben Skeggs Date: Thu May 7 21:39:41 2009 +1000 exa: use exaDriverAlloc() to prevent issues across exa changes commit 8502a8024354b88a4f19c7ad355268bb20772356 Author: Stuart Bennett Date: Wed May 6 18:01:54 2009 +0100 Fix server regeneration again The DRM is now opened in PreInit, so it must not be closed in CloseScreen, as the DRM wouldn't then be open in the subsequent ScreenInit commit 7a796a94d0c0e18a77c5c50a0f9ca9c218d7201f Author: Stuart Bennett Date: Thu Apr 30 16:04:59 2009 +0100 randr12: off-chip lvds for nv28 (rh#487456) Quite how this possibly works with the off-chip DVI connector the laptop in question also claims to have is unclear: I suspect that (at least in nouveau's implementation) it doesn't. Sadly we've no dvi-in-use reg dump to see what should be done. commit 5959512ef5ce20aa52bbd41d5d987392f27b9e65 Author: Stuart Bennett Date: Wed May 6 14:44:40 2009 +0100 'drmCheckModesettingSupported' implicit function declaration fix commit f62719e2ac8b04d4496dc56d5fa181934bcea9f7 Author: Peter Hjalmarsson Date: Wed May 6 14:43:22 2009 +0100 Unused variable warning fixes commit c8260ef52b58234335bdbfab31364d3d25840c21 Author: Pierre Pronchery Date: Wed May 6 14:32:34 2009 +0100 Fix crash with Xinerama enabled From the mailing list, minor tweaks by Stuart Bennett commit 1a478edbd58bb2eb1a41f22dea893c6d85902318 Author: Stuart Bennett Date: Wed May 6 02:23:42 2009 +0100 randr12: de-magic ramdac general control values commit 6c209a87d2d90b39fddc446378b370bd7fb81f6b Author: Stuart Bennett Date: Tue Apr 7 14:37:05 2009 +0100 randr12: pre-nv17 load detection Sampling heuristic as close to nvidia's as mmiotrace-based inference admits Works on both nv05 and nv11 commit 3971dda57004894d5d4fc9420aa00da400815af9 Author: Stuart Bennett Date: Thu Apr 30 19:31:27 2009 +0100 randr12: fix fp_control again again again As prophesied, 9c991d3a84ad16f893b739282403eb01e02bb57f broke something: turning off a digital output, then turning on an analogue output on the same crtc as previously used by the digital output led to FP_TG_CONTROL being turned on again when the (unmodified by analogue mode_set) fp_control state was written out So, belt and braces attempt to get this right, one more time commit 0ef16031c7e27ffabb0299e9463def125bd5e72c Author: Stuart Bennett Date: Thu Apr 30 15:27:33 2009 +0100 randr12: solve off-chip encoder crtc exclusion in prepare, rather than dpms commit 0447ce06989d5599e268fdb8a4ba7bebf72f5c95 Author: Stuart Bennett Date: Sat Apr 4 20:08:32 2009 +0100 randr12: simplify nv11 digital encoder-crtc binding Always setting |0x10 on the appropriate crtc for nv11 digital is fine; it's what the proprietary driver does, even if the bios doesn't Fix some line lengths in nv_output_mode_set() too commit f69b34aa92050987e8f3da07d92d5698e9333b75 Author: Ben Skeggs Date: Fri May 1 09:07:55 2009 +1000 nv50: return immediately in GetDDCModes if no DDC on connector... commit 45e16187612bc09b234b598dff25bbcc8069fde1 Author: Ben Skeggs Date: Thu Apr 30 16:37:29 2009 +1000 bios: use NV_ARCH_50 to match G8x chips There's at least a couple of chips where we get 0x77 for chip_version, and fail to match the chip_version >= 0x80 test.. commit 3d61697c477dd7358e5fd4e5bc608b4fe763b5d3 Author: Ben Skeggs Date: Thu Apr 30 10:23:11 2009 +1000 nv50: remove reading back vbios-programmed lvds native mode commit bade2499db8e02dd2904bc811399c891bf126208 Author: Ben Skeggs Date: Thu Apr 30 09:33:21 2009 +1000 nv50: group encoders into connectors by i2c port commit 9ee2ac19d7adc067a5551e6d6853a19b89946d46 Author: Ben Skeggs Date: Wed Apr 29 16:26:41 2009 +1000 bios: remove dodgy mode-table search for g80 This appears to work in exactly the same way as on earlier chipsets, the strap value just moved slightly.. commit 6282574dbeeb7075a00ce03940c4f2a597a048b5 Author: Ben Skeggs Date: Wed Apr 29 16:25:43 2009 +1000 bios: modify get_fp_strap() for g80 commit 09b832e24b9bd887275ba882beabe9bba56f0207 Author: Ben Skeggs Date: Tue Apr 28 09:46:55 2009 +1000 nv50: merge NV50SorSetClockMode and NV50SorSetClockModeLVDS commit ed2c185caeb75c595dfb2e814748a856a23b2b35 Author: Ben Skeggs Date: Tue Apr 28 08:44:01 2009 +1000 bios/nv50: initial parsing of display script tables, not quite complete commit 3fccc9e83a194fd58c333c6c4793a27cb05ce026 Author: Ben Skeggs Date: Tue Apr 28 09:38:21 2009 +1000 bios: some nv50 init scripts have flags in reg values, deal with (1<<30) commit 2b4b3da73babbe9ebf4acc26238a4b2b8a6e6ee8 Author: Ben Skeggs Date: Wed Apr 29 11:00:17 2009 +1000 bios: dcb location is only 2 bits commit 7b5bfff21916461ce7f6a2adb27dc443ad0b8201 Author: Ben Skeggs Date: Tue Apr 28 08:12:05 2009 +1000 bios: parse BIT U table commit e4260ec4f0303b286f16eab4fe5a1bb66ff89e3d Author: Stuart Bennett Date: Wed Apr 1 02:56:05 2009 +0100 randr12: more use of logging abstraction in modesetting code commit 4d0fdba9abcd0af736821d54baca5f0c9a8d6ce1 Author: Stuart Bennett Date: Mon Apr 6 01:14:04 2009 +0100 randr12: fix for off-chip dual link digital (enable extra wide interface?) commit bb246d72a01285299599cbf8ff82c67f14682a4f Author: Stuart Bennett Date: Mon Apr 6 01:00:41 2009 +0100 Partial fix for nv28 lvds bios parsing (part of rh#487456) For rh#487456, though this is neither a complete fix for that bug, nor a complete fix for nv25/28 lvds (the EDID matching mentioned in the commit is omitted until someone appears with a laptop requiring it) commit 2915926702b7ae9ccb1e900c07ca14a2f32e4512 Author: Stuart Bennett Date: Mon Apr 7 23:20:12 2008 +0100 plls: max_log2p_bias is likely actually max_log2p As explained in nouveau_bios.h, a separate max_usable_log2p is needed (lack of change from 6 to 7 tested by setting a stable small mode, say 400x300, then manually tweaking the multiplier up and post divider down). The max_usable_log2p values hardcoded here are unchanged from those previously taken from the bios pll calculation code For reference, the blob will quite happily attempt to set clocks using the raw parsed max_log2p values (and go up to 7 therefore) when setting modes around 200x200 (3MHz pxclk), but at such a low clock my monitor won't agree to give a sensible picture anyway. Reg dumped data using "nvidia" below: nv40: log2P up to 7 (dual pll mode), at very low clocks driver gives up and sets 0x0006ff0d 0x80001f04 nv43: log2P up to 7 (dual pll mode), at very low clocks driver gives up and sets 0x0006ff0d 0x80001f04 nv11: single pll, therefore goes quite happily down to at least 0.25MHz nv31: log2P up to 7 (dual pll mode), at very low clocks card locks up nv34: log2P up to 5 (single pll), goes down to at least 2.25MHz nv4b: log2P up to 6 (single pll mode), at very low clocks driver gives up and sets 0x0006ff0d 0x80001f04 commit 47bb00fee8b9906adc03b372efa1ae813bc9ca9a Author: Ben Skeggs Date: Tue Apr 21 08:57:07 2009 +1000 nv50: missed a WAIT_RING commit 523ccecee26454fc43a65e9b33d91a7d0ff8487d Author: Ben Skeggs Date: Fri Apr 17 14:18:54 2009 +1000 nv50: use vbios-programmed lvds mode as native mode if no ddc Why was this even removed? Its replacement (read VBIOS fp mode table) has proven useless in the majority of cases I've seen. commit fa2f1115d5f7f3e3767d423a77c6d929a4ae8d0f Author: Ben Skeggs Date: Fri Apr 17 10:25:57 2009 +1000 Remove pNv->GART conditional for selecting CB_LOCATION in AGP/PCI Kinda useless, the DRM knows better whether GART is available. commit 7100c06be099bacc0f8bb8898bbf7eb34ff1cc6e Author: Ben Skeggs Date: Mon Apr 13 20:21:51 2009 +1000 kms: fix displayWidth in resize commit ef2de256dabfde2d619e4d451bd417973c919329 Author: Ben Skeggs Date: Mon Apr 13 19:09:32 2009 +1000 do e-edid for legacy modesetting path commit d8545e669aaa4362a236ba2e7231ca27ffdea1bd Author: Ben Skeggs Date: Wed Apr 8 16:15:02 2009 +1000 bios: logic typo from earlier commit commit 960a5c82a80272a1577dede6a05ab7dc5db37cc3 Author: Ben Skeggs Date: Wed Apr 8 14:44:06 2009 +1000 nv50: modify ddc<->nvreg assignments again.. commit 3a6c7091d9706778a2e3971b7f36601d79fe7a83 Author: Ben Skeggs Date: Wed Apr 8 14:34:20 2009 +1000 kms: fix setting the randr edid property commit a1194b3690ced5db552d63be118a605508a1b965 Author: Ben Skeggs Date: Tue Apr 7 19:24:25 2009 +1000 bios: fix typo commit 11451cabbd5357dd71ab65e6155d7ca7d9766db7 Author: Ben Skeggs Date: Tue Apr 7 18:34:57 2009 +1000 bios: some G8x don't have PBUS at 0x1800, us 0x88000 unconditionally there commit 62d69bdf807423a644d7c2c493f94948ea11e491 Author: Ben Skeggs Date: Sat Apr 4 01:41:47 2009 +1000 bios: add exception for chipset 0x73 for lack of fp table commit 620d519564fdc6b3a7d55184bd481fa69035371b Author: Ben Skeggs Date: Fri Apr 3 23:07:56 2009 +1000 consistent connector naming across Date: Tue Apr 7 00:40:31 2009 +0100 randr12: set 1 on cr59 for off-chip digital, 0 otherwise (#21023, rh#492399) The ubiqitous setting of 1 on cr59 for Date: Fri Apr 3 23:39:22 2009 +0100 randr12: unlock CR21 *after* turning off digital output otherwise the analogue output still fails to come up. That'll teach me to rearrange things just before committing commit 11be9a982073d66a68cd3db2bfc611eb58d3ea81 Author: Ben Skeggs Date: Fri Apr 3 12:22:04 2009 +1000 nv50: fix i2c port addresses commit e2aa03794420324ae5a682bbc415a79a4d830f7e Author: Stuart Bennett Date: Fri Apr 3 00:57:19 2009 +0100 randr12: fall back to standard timings when finding native mode (rh#492819) commit 6965663ddcc8dd3330842d3c4c4eb4cbf4e4fc39 Author: Stuart Bennett Date: Wed Apr 1 05:38:29 2009 +0100 randr12: fix hw cursor for fully transparent pixels ARGB8 pixels of 0x00000000 were getting bumped to 0x01000000, which was _just_ visible as a box around the cursor commit d12f70cedb0ba598234bbfc26dd6beed23e3dfe7 Author: Stuart Bennett Date: Sun Mar 29 00:51:05 2009 +0000 randr12: pre-nv17 digital fixes 1) set panel regs better (don't do min_front_porch adjustment) 2) scaler programming - nv11 and nv20 don't need / 2 for aspect scaling commit ea567dbc7bc9c8d2ee5d7d9e0dcd33b03c638edb Author: Stuart Bennett Date: Sat Mar 28 01:58:09 2009 +0000 randr12: fix restore for cards where CR21 is not left unlocked at POST More nv11 inspired fun. If CR21 is left without a 0xfa mask after boot, we'd restore the old CR21 value and the subsequent restoration of VGA CRTC values wouldn't get through. RE'd doc of CR21 behaviour included commit bd263d83139fcc93b967c71c99fd5e63e1ba6315 Author: Stuart Bennett Date: Fri Mar 27 18:27:18 2009 +0000 randr12: make nv11 dvi work for both crtcs commit 52c287dc64a97a750eb953ac5f725d794cbf0c64 Author: Stuart Bennett Date: Fri Mar 27 00:23:04 2009 +0000 Every card since nv5 can do panels We'll only try to do ones using scripts atm though Note that as we don't have a load detect function for pre-nv17 you'll likely get a 1024x768 mode set up on VGA-0 even if only the DVI output is plugged. Unclear if this is a bug or feature :) commit 3d371a78bc9522e9a1b2c967957a090bb527a30c Author: Stuart Bennett Date: Thu Mar 26 22:21:57 2009 +0000 randr12: enable tmds reg access on pre-nv17 I think this nv11 hates me commit 1bb85c8020802719e82c1335a18df9367215a8ff Author: Stuart Bennett Date: Wed Apr 1 04:09:09 2009 +0100 randr12: reorder ramdac reg access commit 4465fddb2467ad9d923120a2fd2c648222603771 Author: Stuart Bennett Date: Wed Apr 1 04:10:18 2009 +0100 randr12: remove nv30 special case 0x68?89? are just an alternative way to set scaling (set | 0x40 on 0x68.880, then horiz scaling in 0x68.898, vert in 0x68.89c) and afaict don't have any special significance on nv30 commit 72eced5eeeef438979dff06b6c1cbaa120738cfe Author: Stuart Bennett Date: Wed Apr 1 03:12:30 2009 +0100 randr12: fix/improve a load of version/feature tests Main improvement is introduction of flag for nv17+ display architecture Other version fixes include: * allow access of fp control regs on all cards (defined since nv4) * only read crtc_850 and gpio_ext on nv40+, since only written on nv40+ * set dither state on all versions (only written out on relevant cards) * PRAMDAC_DACCLK is only defined/usable on nv17+ * no point saving the lvds/tmds head for restore if there's only one head * fix a bunch of places the nforce (0x1a) case had been missed commit 37c6916d3eac54cc13812276fca99662162c4bbc Author: Ben Skeggs Date: Wed Apr 1 15:00:03 2009 +1000 bios: fix chips with pll limits table version 0 commit 76c5a05603b0eaa0831e031268e48ab6ffdee3a9 Author: Ben Skeggs Date: Wed Apr 1 10:36:34 2009 +1000 bios: xf86DrvMsg->NV_ERROR commit b1b9bcb111f1f5ec15e19d5a84b7ca576983c4a1 Author: Ben Skeggs Date: Tue Mar 31 11:43:26 2009 +1000 bios: support for pll limits table v3.0 commit 8c85b4b7451bd09b0607b1f4ef8ce4b833006328 Author: Ben Skeggs Date: Tue Mar 31 11:11:12 2009 +1000 kms: CONNECTORn -> CONNECTOR-n commit 4e1c3239648510a4024917289127ad1e982ed34d Author: Stuart Bennett Date: Tue Mar 31 20:59:26 2009 +0100 Restore call to NVDRIGetVersion, to ensure the dri module is loaded commit c973f756cfdf36db3a5c7095f6e2c57ff5841388 Author: Stuart Bennett Date: Tue Mar 31 17:07:26 2009 +0100 Abstract logging in nv_bios.c commit b17bebd811316b5ccca21146b58f8fbf4773f0ad Author: Stuart Bennett Date: Mon Mar 30 20:52:29 2009 +0100 nForce DIMM check is now in DRM Since it only at most suggests an optimization, there's no real need to wait for a DRM version bump commit 9c991d3a84ad16f893b739282403eb01e02bb57f Author: Stuart Bennett Date: Mon Mar 30 03:31:01 2009 +0100 randr12: FP_TG_CONTROL 2: dpms improvements For DPMS on digital outputs nvidia does not just OR the 0x68.848 value with 0x20000022, they mask off 0x10000011 too. To do the same thing here, don't change in-memory state of programmed reg value (hopefully nothing breaks) commit 5fb2888d79d39a1c74ab005d8d518686ad823224 Author: Stuart Bennett Date: Fri Mar 27 18:27:18 2009 +0000 randr12: FP_TG_CONTROL 1: turn off digital path when using VGA encoder On pre-nv17, having a CRTC's digital output path turned on when programming the CRTC for VGA confuses the chip and often results in the analogue output not working. Indeed, having it turned on even just before programming the CRTC seems to break it. Nvidia solves this by turning off the digital path early in modesetting and then delaying for 50ms, so (if the digital path was on to start with) that's reproduced here. Although this problem has not been seen on nv17+, nvidia uses the same approach, so we may as well too. Some tidying of the output_prepare function thrown in as well. commit 8023dc3e2dcfa8a4cc11a5a64750a7505e8aa704 Author: Stuart Bennett Date: Mon Mar 30 02:27:50 2009 +0100 Split DCB 1.5 parsing from 2.0+, get closer to 80 cols commit ea027b35b28c0305aca03edc8c1ff5cd84bbf3ef Author: Stuart Bennett Date: Fri Mar 27 00:03:38 2009 +0000 The dcb prior to v1.5 is pretty hopeless, just add a crt. Also, we don't use dcb 1.2 i2c for modesetting due to cards with broken tables (ala #14821), and I've got an nv11 with duff entries (v1.4), so don't raise hopes of init_dcb_i2c_entry working on these versions either commit 921fbccd53eae37ed86ecece6dc1fcae4b7bf3c4 Author: Stuart Bennett Date: Thu Mar 26 16:41:02 2009 +0000 randr12: avoid nv11 chip lockup when saving palette regs commit 9d46930f0eec7b787666671681e9a93fc33dce62 Author: Ben Skeggs Date: Mon Mar 30 20:11:03 2009 +1000 xv: oops commit 9213c39d3a8d9a7fc90a44f9427c1fb8218243e2 Author: Ben Skeggs Date: Mon Mar 30 11:14:20 2009 +1000 xv: post damage after we draw (rh#492239) commit d68a0527aa2a3e93c3408526d4bbb93db89e9b3c Author: Ben Skeggs Date: Mon Mar 30 10:45:59 2009 +1000 kms: small cleanup commit c9cb6a6f7f79233bdab737fa6b6d0916dab45971 Author: Stuart Bennett Date: Fri Mar 27 16:43:27 2009 +0000 rh#492511 has a CRTC_OWNER of 0x7 for some reason not that this probably affects the bug in any way commit 95bff61597ddf21d6415b40759258802a5f42150 Author: Stuart Bennett Date: Wed Mar 25 02:40:17 2009 +0000 randr12: fix LVDS legitimately disconnected case DPMS off still needs to work, but there's no native_mode for the pxclk. Fortunately LVDS_PANEL_OFF doesn't actually need a pxclk, so just use 0 commit f86e395231c97155c1a211010835fd02b8e4dc5f Author: Stuart Bennett Date: Wed Mar 25 02:07:15 2009 +0000 randr12: permit vga outputs to be force enabled in xorg.conf You need: Section "Monitor" Identifier "VGA-0" Option "Enable" "true" EndSection commit 8427b39a2f6cb32496bbbe370b793d5ba78f6b6c Author: Stuart Bennett Date: Tue Mar 24 03:32:24 2009 +0000 randr12: move a load of hw mode programming into nouveau_hw.c Pull in pll setting (common, no reason to bloat nv_bios with it), state load/unload from nv_crtc, and useful bits of nv_hw.c (remnants in nv_hw.c are only used on the non-randr12 path) commit 163bdce98c0f9ca5c1125e33a9a2bf2d2d9a1286 Author: Stuart Bennett Date: Tue Mar 24 03:19:38 2009 +0000 Turn the smaller mmio wrappers into static inline commit 21ee927f28d1ca8d3807acc767ef14783e43f057 Author: Stuart Bennett Date: Mon Mar 23 21:34:26 2009 +0000 Put arbitration and mnp calcs in separate file Make bios->chip_version public so the pll calcs can use it commit 83dc890f0b1fc88d2fbc7559f5d4decf4c84f340 Author: Maarten Maathuis Date: Sat Mar 28 01:25:32 2009 +0100 nv50: use E-EDID when available + apply edid quirks commit 2c130cf71587706a4f1dcd8f5db1de47a10522f5 Author: Pekka Paalanen Date: Fri Mar 27 23:16:59 2009 +0200 nv50: fix xf86GetDefaultModes() call for older servers Thanks to stillunknown for this build fix. commit f1907dcef8d06d7ee4ef10ba22bb7decef700110 Author: Ben Skeggs Date: Fri Mar 27 15:37:17 2009 +1000 nv50: add default modes to mode pool for LVDS panel At some point X stopped doing this if the display doesn't support continuous timing. For a lot of G80 LVDS panels, they only report their native mode via DDC. We'll add the default modes ourself, and use the GPU scalers. commit 7da6fdb8b477d2007c83b47e9cbcc2476ae40f22 Author: Ben Skeggs Date: Fri Mar 27 14:17:29 2009 +1000 kms: fix bug which prevented getting edid from the kernel commit f431e204d3a9be92df5ca606e86ed7c6d82103f0 Author: Ben Skeggs Date: Fri Mar 27 11:50:22 2009 +1000 randr: fix crash when rotation requested commit 36dedd04da1b3c89be61a95270414477d284b2ef Author: Ben Skeggs Date: Fri Mar 27 11:26:09 2009 +1000 kms: check for mm_enabled as an additional test for kms presence commit a923bc1e4840c0386301f8648add2ccbfbf79a88 Author: Ben Skeggs Date: Thu Mar 26 16:24:01 2009 +1000 nv50/xv: correct rendering to partically obscured windows Won't be exposed by current git, due to lack of accelerated front-buffer rendering. But should fix rh#492173, rh#492229. commit 01cee2906686897c4b795d308270b3b69a3e286e Author: Ben Skeggs Date: Thu Mar 26 11:43:09 2009 +1000 nv50: call NVSync() in CloseScreen() before restoring video mode Without this, if accel is still being done to the front buffer (usually at offset 0), we'll restore the video mode and then the engine will clobber the text mode fonts right afterwards leaving a messed up console. commit aa7c0375b51d813be54de5e19b96e958ac183f17 Author: Ben Skeggs Date: Thu Mar 26 06:36:17 2009 +1000 kms: implement AdjustFrame, should fix crash in fdo#24236 commit 61879b8079bee1caeedc0bd9c11061423ff27b4a Author: Ben Skeggs Date: Wed Mar 25 10:09:39 2009 +1000 kms: drm_mode_modeinfo struct changed names at some point commit 79306fc29f6b27fd61fe51855bad5489fc4fd986 Author: Ben Skeggs Date: Wed Mar 25 08:15:41 2009 +1000 dri: fail harder commit da1ba93f1dc2b02615e4f0366450872659176799 Author: Ben Skeggs Date: Wed Mar 25 08:12:26 2009 +1000 device close doesn't belong in DRI close, can have DRM without DRI commit 56b11c773ac9392069c6a589f4f87c5cc79e020e Author: Younes Manton Date: Tue Mar 24 14:27:40 2009 -0400 Fix mouse cursor disappearing when near top/left edge of screen. commit 4067ab466fe3aa817e0323959f70c7dd3494de0a Author: Ben Skeggs Date: Mon Mar 23 14:43:22 2009 +1000 another ppc fix commit 862dba8b6ca6354d915f2430826a5581f41d1002 Author: Ben Skeggs Date: Mon Mar 23 14:22:15 2009 +1000 more ppc.. commit 30634865617737171835f6ac9437df597e5a8951 Author: Ben Skeggs Date: Mon Mar 23 13:57:52 2009 +1000 fix ppc build commit d80fe7878379aefd3345746a149f2b46d3a84805 Author: Francisco Jerez Date: Mon Mar 23 03:27:16 2009 +0100 Fix nouveau_hw_decode_pll on NV30/35. commit 62aa81deca5c63d30407872c0e2b784f5084acb3 Author: Stuart Bennett Date: Sat Mar 21 01:43:14 2009 +0000 Some laptop mode finding fixes resulting from rh#487456 On closer inspection, it seems BMP laptops only depend on use_straps_for_mode to ban DDC. Therefore use pxclk alone to determine EDID case in parse_lvds_table, and invert sense of fp ddc flag to get a default of "false" commit aa6edfaf1c32432b07c298c42b68f2d93f3b337c Author: Stuart Bennett Date: Sun Mar 15 15:18:16 2009 +0000 randr12: name VGA CRTC fields Now with added consistent indentation Use fields in a couple of other cases while we're about it commit 7b7c44c828eda2edfad2fa9bea95f70889c59453 Author: Stuart Bennett Date: Sun Mar 15 03:45:54 2009 +0000 randr12: rename reg state members to follow reg names commit 4ae6c22957a929760e583d5704e350adb4404c3d Author: Stuart Bennett Date: Sun Mar 15 03:11:40 2009 +0000 randr12: name some FP flags, move some unrelated stuff out from nvreg.h commit b50b49f6f54148ee64ba7500e797f432b198a324 Author: Stuart Bennett Date: Sun Mar 15 04:21:40 2009 +0000 pre-nv50: use nvidia's names for PRAMDAC regs where known Mostly adding 'P' to the previous NV_RAMDAC regs -- either when known correct, or when totally unknown (eg NV_PRAMDAC_580); made-up names untouched. Lack of code changes checked with cpp. Unused defines removed too. commit a9df3049a7e38b3bf4478459e5f0736e10927247 Author: Stuart Bennett Date: Sun Mar 15 04:08:56 2009 +0000 pre-nv50: use nvidia's names for PCRTC regs where known Mostly adding 'P' to the previous NV_CTRC regs -- either when known correct, or when totally unknown (eg NV_PCRTC_850); made-up names untouched. Lack of code changes checked with cpp. Unused defines removed too. commit 2c0710d334e3e1dcb04cc4fc0b46ec4c8cdb9ba6 Author: Stuart Bennett Date: Tue Mar 17 18:50:18 2009 +0000 Separate maximum limits for encoder devices and i2c records, DCB_ prefixes From the comment with pNv->connector in nv_type.h I've assumed nv50 was using the previous define as an I2C limit, rather than encoder limit commit 0db27a6f1aee023782de14ae2f260fec4e60229d Author: Stuart Bennett Date: Tue Mar 17 13:43:13 2009 +0000 Add DCB 1.5 entry from a GeForce2 Go reported by Martin Ketzer on the ML commit 3e7fa97a50e42c0255876c79137dc1a40a6052a6 Author: Ben Skeggs Date: Wed Mar 18 09:36:33 2009 +1000 dri: pass object handle instead of offset if using real mm commit 88efe405213a3ea23edd9bed971da1d2492021eb Author: Ben Skeggs Date: Mon Mar 16 17:45:49 2009 +1000 dri2: initial implementation, only with driver pixmaps Also, no DRI1 when driver pixmaps enabled. commit 675126e0f0c1c077bde4b25ccb64f87d1d08547b Author: Ben Skeggs Date: Mon Mar 16 15:06:31 2009 +1000 kms: support framebuffer resize if driver pixmaps enabled commit 8b354b92a859cf277b19e51fc6cb109f4ea44b7a Author: Ben Skeggs Date: Mon Mar 16 13:13:07 2009 +1000 kms: copy old fb to new fb on modeset Requires driver pixmaps enabled, which suck on G8x currently. Paves the way for a nicer transition from console->X. There's still a flash where the screen goes black before X starts here, but this is on the way to what we want. commit 6658403cfd91556d06afecdde03feec7384c5398 Author: Ben Skeggs Date: Mon Mar 16 12:34:49 2009 +1000 Call NVEnterVT from NVScreenInit rather than duplicating it. commit f59fef15c94fc5c6f216d0f5dcdcb3d3188c0049 Author: Ben Skeggs Date: Mon Mar 16 09:17:29 2009 +1000 exa: fix multiple prepare/finish_access on a pixmap commit 75a1c4158a19bf0857de49b736fb87a40c095048 Author: Stuart Bennett Date: Sat Mar 14 15:48:39 2009 +0000 Wrap bios endian casting Also convert some functions to use individual bios pointers, rather than bios->data[offset + ... Plus some minor bmp structure documentation and tidying commit 3666d73333ecbdd3a4941a76af2528113db427bb Author: Stuart Bennett Date: Wed Feb 25 20:21:42 2009 +0000 Switch to using pll_vals struct for passing pll stuff around commit a3dbc06d4f9dc340f5c167b94d8e10ac8386c5b4 Author: Stuart Bennett Date: Fri Apr 11 16:51:30 2008 +0100 randr12: cull old unused pll code commit 71c821b8119cbac8dcd91b27e3b7556d00ceb546 Author: Stuart Bennett Date: Sat Mar 14 19:06:12 2009 +0000 Don't pointlessly overallocate cursor storage commit 7284c07a0ad6551ea99c28454f4e7308699835e6 Author: Stuart Bennett Date: Fri Mar 13 21:23:48 2009 +0000 randr12: nv10 hw cursor fixes/changes Remove access to PCRTC_GPIO, it's unused and on NV10 overwrites PCRTC_CURSOR_CONFIG leading to all sorts of cursor weirdness NV10 hw cursor is now set-up on the assumption it can do 64x64 (we've been using 64x32 up to now), but if it can't do cursors 64px high it can easily be fixed to behave like nv0x at 32x32 (64x32's pretty silly) It's possible that with the PCRTC_GPIO thing resolved the nv10 cursor can be more featureful: the code would also be simpler if it supported 32bpp cursors (test by seeing if setting NV_CRTC_CURSOR_CONFIG_32BPP has any effect). It may even do alpha -- change pNv->alphaCursor to be (NVArch >= 0x10) to find out. commit bcc3a14b15cdad3647cc7df0f540df1111843be5 Author: Stuart Bennett Date: Fri Mar 13 02:31:35 2009 +0000 randr12: use nv0x cursor colour expansion funcs on later cards, unify argb paths commit b21c807d2ced35ac1b06300f6e68d661710b336a Author: Stuart Bennett Date: Fri Mar 13 03:52:54 2009 +0000 randr12: store nv0x hw cursor image in VRAM nv_driver.c is already doing the alloc, nvidia uses the same method, and it saves code :) commit 3ea4b194e83735145d5e9d5c5357f92ae6c6a569 Author: Stuart Bennett Date: Thu Mar 12 15:37:59 2009 +0000 Kill pre-randr12 hwcursor code, and tidy randr12 hwcursor code If you're not using randr12 you should be. Misc fixes included: Ensure PNVM bit set on PCRTC_CURSOR_CONFIG for consistency Fix NV11 alpha cursors Add missing bo unref commit b988160dfa5a535fd042062d7318ab99fc7efa5a Author: Stuart Bennett Date: Thu Mar 12 16:31:24 2009 +0000 Remove calls to load ddc and i2c modules -- they're built-in since server 1.3 commit d5893ecf3dd915fc6937c61e6eae7c221edeb6e2 Author: Stuart Bennett Date: Thu Mar 12 21:46:02 2009 +0000 Hack around missing display table in rdivacky's GeForce 7050 PV/nForce 630a commit 369e1efd23c0180e210dbe898318b6b9efa1faad Author: Stuart Bennett Date: Thu Mar 12 16:09:27 2009 +0000 Fix xserver shutdown for randr12 off commit 8343d6b2c0ebddf1a1a7d146890ed90278559fea Author: Stuart Bennett Date: Wed Mar 11 01:11:02 2009 +0000 Remove bios lvds parsing pxclk test; ddc_permitted is a sufficient condition commit 8b56c699ad39c64326d76658c741801bb4118c5f Author: Stuart Bennett Date: Wed Mar 11 00:43:33 2009 +0000 Name some regs/values and reduce crtc reg struct size TVOUT_LATENCY taken from nv11 .scp file CRTC[]: even nvidia only use up to 9f, and they know what the regs do commit 79d23d87b489332af521fbda418a13b9ae032b7d Author: Ben Skeggs Date: Fri Mar 13 11:43:54 2009 +1000 improve drm/dri initialisation code The main motivation here was to fix the scenario where X is started without nouveau.ko loaded, checks if kms is enabled, is told no because nouveau.ko isn't loaded, and then later nouveau.ko gets loaded with kms enabled by default and starts fighting with the ddx for the display. One cool side-effect is that we're not depending on DRIScreenInit for our DRM fd now, which means that if DRIScreenInit fails for a non-fatal reason (like, NVIDIA's libglx being around instead of xorg's) we can still run with full 2D acceleration. commit bd1448216715681beebabe82b88c8b4d0e62eced Author: Ben Skeggs Date: Wed Mar 11 16:07:41 2009 +1000 kms: cleanup set_property commit 7a455929785e00beb2d1e65007a3407427ca693a Author: Ben Skeggs Date: Wed Mar 11 15:51:59 2009 +1000 kms: get current values for output properties on startup commit 7095e8942c74953fcf1615900e2d7864024ece5c Author: Ben Skeggs Date: Wed Mar 11 11:30:08 2009 +1000 kms: work around some xserver stupidity with DPMS The server calls the output DPMS functions with the intention of turning off outputs before reprogramming them. It probably shouldn't be doing this in the case of the randr driver using set_mode_major, and leave it up to the driver. The DRM modesetting will usually reenable the outputs after a modeset anyway, which is the behaviour that the xserver currently relies on. However, there's a shortcut used in the drm to allow far faster fb swaps without doing a modeset. When this path is hit, the DRM will rightly not touch any power states and the display will never get turned back on. commit 8f9a58070c2fc55c922e9d82d0796e32959a8431 Author: Stuart Bennett Date: Tue Mar 10 00:20:35 2009 +0000 Another DCB 1.5 entry (seen in rh#455194) Plus a few minor PLL tidyings commit 9a719900adcebf47442a30fe1989509f93e13e16 Author: Stuart Bennett Date: Mon Mar 9 15:57:06 2009 +0000 randr12: add output properties for digital vibrance and image sharpening commit 678ddc2e56aea3deb1e95b2309b47e1f2cce0635 Author: Stuart Bennett Date: Mon Mar 9 15:34:03 2009 +0000 randr12: functionate output properties creation notes: * setting pending at change time is necessary for props created as pending, so that PostPending doesn't overwrite the correct current state with the incorrect pending one * our call to PostPending can go; relic from when props were updated in update_output_fields commit ed9bd8848fcfe6a77894bfe520a5aa25a9b3f769 Author: Ben Skeggs Date: Wed Mar 4 15:10:24 2009 +1000 kms: remove early-exit from gamma_set, the kernel side works now commit 98c9e4edb58374f18249e5c9c53b392fb8b4a1d1 Author: Ben Skeggs Date: Wed Mar 4 13:40:49 2009 +1000 kms: work around some bong hits with dpms This'll need to get fixed elsewhere, but in the meantime, we have working DPMS *and* can actually see what's happening when X starts! commit 88ae2c92c0efc1919dbaa7e8db4c37b86987046c Author: Ben Skeggs Date: Wed Mar 4 13:36:36 2009 +1000 kms: implement dpms commit 5269cc8060e0ce5160dfca408b0807a4cf2ff17c Author: Stuart Bennett Date: Sun Mar 8 20:07:23 2009 +0000 randr12: make props code slightly less awful to behold commit 0b87c4964076101a2d8373fd3addd9355fb7765e Author: Stuart Bennett Date: Sat Mar 7 21:13:55 2009 +0000 More robust PLL upclocking for old cards commit cb237a34f31f8b8c841974cb64141a2f2e36c53b Author: Stuart Bennett Date: Sat Mar 7 04:38:12 2009 +0000 bios/randr12: mode validation fixes (includes fix for #20298) Remove arbitrary 12000kHz mode clock lower bound (#20298) Use DCB supplied max dac clock (and remove a memcpy) commit 9cfccd7281a715e45a4296226d9d52854230f346 Author: Stuart Bennett Date: Sat Mar 7 04:34:20 2009 +0000 Handle failure to calculate pll values better Ensure pll limit table get parsed on nv30 and nv35 too commit 1e994400913bc656b34440df67aa105aa2b211c4 Author: Stuart Bennett Date: Fri Jan 16 22:26:28 2009 +0000 xv: return BadAlloc if destination pixmap is not put into EXA offscreen (#15792) Lets the player get X errors, instead of the card choking trying to do system memory addresses We could use exaDrawableIsOffscreen() here, but that's only in new xservers. commit 77347da64f86822f128e0dae3e89a6b80282c4ba Author: Stuart Bennett Date: Fri Feb 27 01:22:13 2009 +0000 Remove duplicated DCB parsing Store stuff in DCB private instead commit 1b6cabeee4c0160335c174bd4e9111dbd9535aaa Author: Stuart Bennett Date: Tue Mar 3 01:03:45 2009 +0000 Remove VBIOS access in parse_dcb_entry commit 72fd0ae537e4db208bbfa142dd70144901d7f267 Author: Stuart Bennett Date: Tue Mar 3 21:51:46 2009 +0000 randr12: no need to re-detect edid for lvds, it's not going away Reuse of nv_connector->edid is safe as nv_lvds_output_get_modes doesn't call nv_output_get_edid_modes after a mode has been found once commit ae981a8075395af40cf530f2aa24ac68d2ca0e77 Author: Maarten Maathuis Date: Thu Mar 5 20:21:39 2009 +0100 nv50: some minor modesetting changes commit 42f99e68940a526f804e2046676c1d06d1bbbb42 Author: Ben Skeggs Date: Thu Mar 5 12:11:03 2009 +1100 nv30/xv: restore original viewport/clipping after putimage Fixes two cases of corruption I seen while using the texture adaptors with a composite manager. - desktop getting clipped to video size sometimes - missing fonts on some menu items after xv+compmgr commit 56bb8f8e81e9d152217df7863e59781009a9bb77 Author: Stuart Bennett Date: Wed Mar 4 19:48:29 2009 +0000 Allow parse_fp_mode_table for x86 cards that find themselves on ppc commit 1c2f33ed33791f391d059778b1490222e8cd1e1d Author: Stuart Bennett Date: Wed Mar 4 19:40:37 2009 +0000 Don't rely on the bios mobile feature bit for BMP biosen (#19986) Quadro bios builds also set the mobile bit (not on nv10gl, but for all observed subsequent Quadros before nv4x (intro of BIT bios structure). So for BMP cards, use bit 6 of cr4b on head A to detect mobile biosen (all sampled mobile bios images set this in their init scripts, and non-mobile don't) commit e436c7ee98c2eda9d5ba45f7d897d68497b52799 Author: Stuart Bennett Date: Wed Mar 4 20:59:42 2009 +0000 randr12: make treatment of CR4B more closely resemble blob's behaviour Possibly totally unnecessary anyway commit c61102b326455f6c63a55c8c73ed629994281588 Author: Stuart Bennett Date: Wed Feb 18 01:17:29 2009 +0000 Parse bios fp mode during modesetting, removing extra alloc Also, it seems that the mode parsing was wrong on big endian for years, so fix that (though Apple laptops have DDC panels anyway) commit 631e15a03b8a790d342aefe677f8c5eb33935de6 Author: Stuart Bennett Date: Tue Mar 3 22:03:00 2009 +0000 Get dual_link and dithering bios info through lvds table parsing function Now that lvds table is only parsed during modesetting, return the parsed data in out parameters to make the ordering requirements obvious Since calling this on nv50 is pointless, that code was commented out commit 2d4377193f5e2166af3df32ddcfb1235be82764e Author: Stuart Bennett Date: Tue Mar 3 21:52:04 2009 +0000 Call parse_lvds_manufacturer_table during modesetting only (#19986 partial fix) Also remove SetOwner call now that locking crtcs doesn't change the set head Moving parse_lvds_manufacturer_table to modesetting means it only gets called for lvds outputs, removing the dependency on checking for the mobile bios build. Another benefit is making the strap and EDID FP paths common. A side effect is that as the lvds dithering toggle depends on the results of parse_lvds_manufacturer_table, setting that moves to after calling the parsing function commit ce8f5d71dacbe0d56cfcb17dfe60f1c4c6cf1773 Author: Stuart Bennett Date: Tue Mar 3 17:24:09 2009 +0000 Variety of LVDS script changes Remove lvds invocation static No LVDS_INIT for BIT biosen Merge in one-time call to LVDS_INIT script to main script function Make sel_clk preservation stuff use regular reg writing functions commit 071d5373df744c0c9578134fa634bce1e202d4f3 Author: Stuart Bennett Date: Tue Feb 17 16:45:50 2009 +0000 Read embedded bios edid only on demand commit f04689cb326da409448d54694e2660cf4b4b7950 Author: Ben Skeggs Date: Tue Mar 3 10:14:26 2009 +1000 move memset of fb after NVSave(), hopefully fix lost VGA fonts commit 6961efb84dd6894aaaf010a9f620231dfcbad004 Author: Ben Skeggs Date: Mon Mar 2 11:35:05 2009 +1000 kms: fix check for xf86drmMode.h commit 76fa6569ffd1a7f61e295753383bd01b246b5348 Author: Stuart Bennett Date: Mon Mar 2 16:43:06 2009 +0000 Detypedef bios_t, rename bios io functions, group various helper functions commit 39c9d0231c0d24bf91dbe9ab262f54fcad1b3b2f Author: Stuart Bennett Date: Fri Feb 27 18:57:35 2009 +0000 Move parsed DCB to bios structs, split into public and private parts commit 275daaf7be46e54ca59067d3f2a6e6790e1ce7e0 Author: Stuart Bennett Date: Fri Feb 27 17:46:26 2009 +0000 Split bios struct into struct for nv_bios.c, and exported public info struct commit 8b3e966e3230add04aa05a9ded3f8a647afe0d91 Author: Stuart Bennett Date: Sun Mar 1 20:50:15 2009 +0000 Remove never updated bios opcodes 0x31, 0x5A and 0x73 Notices of removal left in case any bios using them ever appears, so they can be found in git history Also nuke a few unused reg access defines commit ce9eb05f516899f56e66cc290fc7c4f9d3085882 Author: Stuart Bennett Date: Fri Feb 27 23:14:24 2009 +0000 Trivial bios convenience variable changes commit cfeef1144943e72d62a2c0d9701ddf4a639deeca Author: Maarten Maathuis Date: Fri Feb 27 17:30:40 2009 +0100 exa: forgot to check if EXA_SUPPORTS_PREPARE_AUX was defined commit 37a85dcbd6f6134a43b5c9bb883553fdaf7c0bd9 Author: Maarten Maathuis Date: Fri Feb 27 17:23:42 2009 +0100 exa: we support PREPARE_AUX commit 1249fd25a529702a4c6fb4831ef92e801f0b6122 Author: Maarten Maathuis Date: Fri Feb 27 17:21:47 2009 +0100 Revert "exa: support major version 3" This reverts commit b7f99aec7282ca7c81d19ff3ad7d9ff22275d476. commit 1c4a284a80ebed9f9d1e01c47b929481801566b5 Author: Stuart Bennett Date: Fri Feb 20 20:02:38 2009 +0000 Rework PLL decoding in terms of pll_vals, save/restore pll_vals on vt switch If your nv04-nv4x vt after running nouveau is more broken than it was with the previous commit, file a bug commit 81b4732162915ab1ac0593e0ef21b56b089dafae Author: Stuart Bennett Date: Fri Feb 20 19:11:41 2009 +0000 New common struct for pll values Makes the compiler do our dirty accessing work for us commit adf2e350341b983d27123550041c3b1ddaac4526 Author: Stuart Bennett Date: Wed Feb 25 02:11:27 2009 +0000 Check whether NV_RAMDAC_580 has changed when setting PLLs commit b7f99aec7282ca7c81d19ff3ad7d9ff22275d476 Author: Maarten Maathuis Date: Tue Feb 24 19:13:23 2009 +0100 exa: support major version 3 commit d91fc78859b11d00c822e5cae70f0fc511b26701 Author: Ben Skeggs Date: Tue Feb 24 08:31:40 2009 +1000 kms: fix resize func Don't change virtualX/virtualY when we're not resizing the actual framebuffer, just change displayWidth to account for the adjusted scanout size. With driver pixmaps, we'll be able to do "proper" framebuffer resize. commit 535498bcfe79858e08ae1f01125025f56bcd504f Author: Ben Skeggs Date: Mon Feb 23 08:20:34 2009 +1000 clear the scanout buffer on statup to avoid seeing garbage on-screen commit 022767373474a0dacf50ee6fcd6401db15018aca Author: Ben Skeggs Date: Fri Feb 20 13:59:58 2009 +1000 nv50/exa: properly align offset for mem_tile ioctl I really really hate this ioctl().. commit e09f50e5f9126c682289b7ba790f16b93a80b876 Author: Stuart Bennett Date: Wed Feb 18 21:04:00 2009 +0000 bios/randr12: fix dual link TMDS (#20006) TMDS04 needs setting, just as for LVDS, and bit 31 of fp_control should also be set (although in this instance the scripts ensure it is set) Furthermore, stop dpms clobbering any changes made to fp_control by scripts Many thanks to bmaass for testing commit b968db77037790dd5be91d6fafa0552eb7e9520f Author: Stuart Bennett Date: Mon Feb 16 21:38:31 2009 +0000 Return of "bios: fix create_i2c_device for g80" commit 0a439ba05aa192e7433d430acd91b60719d20caf Author: Stuart Bennett Date: Wed Feb 11 20:52:43 2009 +0000 Parse DCB before running init commit 870b6f0d2df3e64af5ac921594b1742f82329c49 Author: Stuart Bennett Date: Tue Jan 20 05:29:38 2009 +0000 Split parsing and init into separate functions commit 0280a0c975830b5c4b2edf4fff8a79fb7d67b9df Author: Stuart Bennett Date: Wed Feb 11 18:53:56 2009 +0000 Attempt to remove pNv and bios_t use in parsing the DCB It's meant to parse a section of rom image into useful things, so use of pNv ought to be unusual. Access to things contained in pNv and bios_t makes it harder to prove parsing ordering requirements commit 9179c6667cf859248a525c8d01922b69ed45c17b Author: Stuart Bennett Date: Wed Feb 11 18:33:47 2009 +0000 Variety of DCB enhancements * ensure parsing of dcb_table.i2c_default_indices happens * only merge entries on dcb versions < 2.1, and note consequnces of merging * correct dcbent->index on post-merge compacted entries * drop pointless memcpy * return an error if no outputs found * define dcb structure in nouveau_bios.h (ought to rename pNv->dcb_table to pNv->dcb one day) commit 317ca477cdbb936f29360984bcee25a0ecda3aa3 Author: Stuart Bennett Date: Mon Feb 16 21:15:51 2009 +0000 Revert "bios: fix create_i2c_device for g80" This reverts commit ca72ca24784b9da09b6e3e0b39f3aa32b8f7b502. It'll be back in a minute. commit 7b25a3019beba6342505fa16327a5bcd43669f90 Author: Maarten Maathuis Date: Mon Feb 16 21:23:56 2009 +0100 nv04: unbreak copy commit 60c8bb1391a51188a4981b4936c190c149b79b7f Author: Ben Skeggs Date: Mon Feb 16 13:09:39 2009 +1000 default to autodetecting whether to enable the driver's kms paths commit a31b1d0118dc7b4d07a2c232fcd5aafbf94f69b9 Author: Ben Skeggs Date: Mon Feb 16 09:49:39 2009 +1000 exa: reimplement driver-controlled pixmaps commit d97e993072dc773ecfeb80b5e138b5e4598a1c3d Author: Ben Skeggs Date: Mon Feb 16 08:50:02 2009 +1000 exa: preparation for reintroducing driver-controlled pixmaps commit 1e00c7d1e6edb7f6b4e6d12ef227d27413948f8f Author: Maarten Maathuis Date: Sun Feb 15 18:52:59 2009 +0100 shadowfb: clip copied regions to frontbuffer size. - There have been reports of crashed when moving windows outside the visible screen. commit efe27964fa25f71dcbdc7a751210860369ea0e11 Author: Maarten Maathuis Date: Sun Feb 15 16:54:18 2009 +0100 nv50: correct a poorly written comment commit 2573c065423af47e14263479c66f2308259f29b8 Author: Ben Skeggs Date: Fri Feb 13 10:46:13 2009 +1000 exa: remove exaWaitSync() call in nouveau_exa_pixmap_map() unnecessary, nouveau_bo_map() will sync if required. commit 500053517cb5fe058468c8965f43e6c3209ba7b5 Author: Ben Skeggs Date: Thu Feb 12 09:04:43 2009 +1000 exa: don't need exaMarkSync on DFS, the operation will be done already commit a7e7c557dbf9b2fe54db401793608128de08bdf5 Author: Ben Skeggs Date: Thu Feb 12 09:01:49 2009 +1000 xv: remove notifier usage commit 8ecb8d5331bdb910beea416cdb0b662c624c5c96 Author: Ben Skeggs Date: Thu Feb 12 09:00:07 2009 +1000 exa: remove notifier waits in UTS/DFS nouveau_bo will handle CPU/GPU sync if necessary when bo_map() is called. commit cfb884b816eebbdaa2103dd233c6f975834e536d Author: Maarten Maathuis Date: Fri Feb 13 22:29:33 2009 +0100 bios: add some more valid regs commit ca72ca24784b9da09b6e3e0b39f3aa32b8f7b502 Author: Maarten Maathuis Date: Fri Feb 13 22:18:18 2009 +0100 bios: fix create_i2c_device for g80 commit bbd7041791eacd357afa775236d4123dd98ab9c9 Author: Maarten Maathuis Date: Fri Feb 13 20:30:24 2009 +0100 bios: implement opcode 0x76 (INIT_IO_CONDITION). commit 5dd2cbe454ee8560c814f3fd96d7c28e0fec1dc1 Author: Maarten Maathuis Date: Thu Feb 12 17:13:42 2009 +0100 exa: add nouveau_exa_pixmap_is_tiled commit 15ccf721791fa4f61784ad00449630c6d2496328 Author: Stuart Bennett Date: Sat Feb 7 16:40:46 2009 +0000 randr12: anti-crash measures for #19854 commit 17e9d4bb8c46d67e201b2a369eff7db805dff07b Author: Stuart Bennett Date: Fri Apr 4 19:01:20 2008 +0100 randr12: use bios PLL setting routines for modesetting Only used for setting modes currently, old code remains for restore commit e444e0e348b3aab59671ffdeb605f9a57c46bf70 Author: Stuart Bennett Date: Tue Jan 20 03:23:40 2009 +0000 Readability improvements for nv_hw.c and nvreg.h commit f5ea66c2ef2c726b6a0aff72843c379ab25dbaff Author: Stuart Bennett Date: Tue Feb 10 19:01:43 2009 +0000 Pull in extras from CRTC PLL code to bios routines * g7x magic bits * 0xc040 manipulation for vplls commit 416636509152cf12c801ac537cb8c1bd20c19b5e Author: Stuart Bennett Date: Tue Jan 20 02:11:08 2009 +0000 Pull out ramdac580 setting and powerctrl_1 shift code to separate functions And only twiddle powerctrl_1 for vplls on nv3x single stage pll chips commit 8b7041833c11f32feb7fa5edc21648cc9db1a333 Author: Stuart Bennett Date: Tue Feb 10 19:01:05 2009 +0000 Miscellaneous fixes to bios PLL code * bios variable was only used for chip_version. simplify. * set plls faster on single pll cards after nv15/16 * allow early return for single stage plls when already set * single stage pll mode is only for nv41+ * 80 cols fixes commit 01bb4c1576a45217953632c9f5612aea9173f95f Author: Stuart Bennett Date: Mon Feb 9 21:14:43 2009 +0000 Make PROM score badly, out of available bios images, if checksum bad Problem exposed on a Mac following 0387ac32 Having PROM first is good, if the image is fine. But if the read-only PROM has problems, not a good idea to use it if there are other options available Also avoid result contamination, if a read method has returned early and not overwritten previous image in *data commit 30848dd378d05917c646a24df404846e6f9b2352 Author: Stuart Bennett Date: Tue Jan 6 16:23:33 2009 +0000 randr12: bodge-in continuous timings when the scaler is in use (#19203) Similar to 3354e660b0744976871683ce226f17e873f26b50 in the intel driver, except I can't see much benefit in allowing bigger-than-native resolutions, so no sync range hacking A pity the xserver has no idea of scalers and so each driver ends up with a hack commit 0387ac32bef6c714f54917d5d36071ae1224458c Author: Stuart Bennett Date: Sun Feb 8 09:19:53 2009 +0000 Allow reading of PROM and PCI rom on PPC (for people using x86 cards) commit 6ad66c0b6165822dde167fd1700bd14727ce7087 Author: Stuart Bennett Date: Sun Jan 25 20:09:36 2009 +0000 bios/randr12: be strict about when to use modes from bios vs modes from ddc Again, the Inspiron 8100 in #14567 is an example of wackiness, having a non-0xf i2c_index, but using straps If the previous commit didn't break your DCB 2.0 EDID equipped laptop, this will commit b5391eb909f8b477c3c438f1136d788acfe6bfd8 Author: Stuart Bennett Date: Sun Jan 25 03:03:29 2009 +0000 Simplify FP strapping semantics and trust bios parsing of use_straps_for_mode In many ways a spiritual revert of 11de8d The compound fp.strapping member faithfully implemented what nvidia sticks in cr58 (for cr57 == 0xf), but this was never directly useful and had differing overloaded interpretations with bios version, causing confusion. Return to just using the result of get_fp_strap (g80 code of which unused and so dropped) As noted in the commit, there have been DCB 2.0 laptops (as in bug 14567) which don't set use_straps_for_mode when they should, so we set it for all DCB 1.4-2.0 cards. If EDID-ful DCB 1.4-2.0 laptops exist they will be broken, and will need fixing later (simplistic fix: use_straps_for_mode = (get_fp_strap() == 0xf) ) Will also break nv40 straps-based laptops whose tables used reserved values, but that's pretty broken anyway, and I've no evidence they exist commit 34c2b1539bb2d69dd3bd5911c4fac62e46385d20 Author: Stuart Bennett Date: Sat Jan 24 18:41:31 2009 +0000 bios/randr12: misc (mainly lvds related) clarifications and cleanups get_fp_strap() moved (& unused g80 bodge removed) in anticipation of next commit commit 98b8cada6c355d437925a92ef0413e96751ed567 Author: Maarten Maathuis Date: Sat Feb 7 21:02:31 2009 +0100 Always allocate 2 hw cursors. - This hopefully fixes the legacy codepath again. commit f1099810bb3cfe451877667a0224eb3d664e442f Author: Maarten Maathuis Date: Sat Feb 7 20:04:54 2009 +0100 Fix a few warnings. commit 691589d8c05b8b0354e359b4942711f6fffa9903 Author: Maarten Maathuis Date: Sat Feb 7 19:59:28 2009 +0100 Fix some cursor fallout, mostly in legacy code. commit 945f0cb644b6e4f34cc04927177fe53529aeb5cf Author: Ben Skeggs Date: Fri Feb 6 00:44:14 2009 +1000 kms: update drmmode_display commit ba1f897159d910a25560796d11c9d0357869e1c0 Author: Ben Skeggs Date: Thu Feb 5 23:15:24 2009 +1000 keep resources around after leavevt, far too problematic currently commit 4dfd0b1b21415c48c4b6e0e8d335f3fc70922d27 Author: Ben Skeggs Date: Wed Feb 4 15:35:53 2009 +1000 oops commit e02813aab0376271ee777a2b8dafc9381f96311a Author: Ben Skeggs Date: Wed Feb 4 14:50:35 2009 +1000 link against libdrm_nouveau (installed with libdrm) commit 2c06308a6c5878fca3ff4f2f9c969705d966607f Author: Ben Skeggs Date: Wed Feb 4 14:09:21 2009 +1000 require drm 0.0.12 commit 29aa1341fb28578968953f7e7dc3de0ec66f70f5 Author: Ben Skeggs Date: Mon Feb 2 11:13:55 2009 +1000 fix NoAccel from when map/unmap became stricter commit 62adc557560fa82c2086e32cd16bc51f67a28b48 Author: Ben Skeggs Date: Tue Feb 3 08:10:45 2009 +1000 Revert "Improve the NoAccel situation, but it's not working yet." This reverts commit 285027a180b676ac63f4c7e158e94ffbe22bf11d. Sorry, fixed this yesterday and forgot to push the commit. commit 285027a180b676ac63f4c7e158e94ffbe22bf11d Author: Maarten Maathuis Date: Mon Feb 2 11:15:02 2009 +0100 Improve the NoAccel situation, but it's not working yet. - nouveau_bo related functions will cause problems for fake bo's (since they don't have a bo_priv). - NoAccel still doesn't work for me (the scanout contains random crap). commit 5e63c7866ac93f39fde49af185387d2094110627 Author: Ben Skeggs Date: Mon Feb 2 13:46:01 2009 +1000 Revert "nv50: make entire offscreen area tiled, use extra blits to scanout buffer" This reverts commit 6c3f8da1bff3c94353748321b40a523add824327. EXA doesn't use UTS/DFS hooks when it's accessing the "front buffer" for software fallbacks, so sw rendering to it was garbled because it doesn't understand tiling. The Prepare/FinishAccess hooks could be used to point the pixmap at the linear scanout buffer instead for sw access, but on VT switch for some reason exa get called to try and do things, hit the PrepareAccess() hook and die horribly because in LeaveVT() the driver deallocates the buffers. The other alternative was to use libwfb to wrap the fb accesses and handle tiling, but exa isn't quite ready for that yet either - it's being worked on. Another point made about this patch was that we should figure out the display engine and get it to scanout the tiled buffer directly, not commit shoddy workarounds. Even for this "correct" solution, the two above-mentioned issues still apply. commit 6c3f8da1bff3c94353748321b40a523add824327 Author: Ben Skeggs Date: Mon Feb 2 08:52:18 2009 +1000 nv50: make entire offscreen area tiled, use extra blits to scanout buffer This fixes some pretty bad performance issues with window resize and desktop switching where ARGB windows are concerned. It also allows Xv to be used without a composite manager. The drawbacks are the use of a little more memory for the shadow scanout buffer, and a bit more GPU time is used for the extra blits. But it's worth it. If you want to get rid of this, figure out how to scanout tiled buffers :) commit 61791028ee9321748b635a85d4e53ed80d143df5 Author: Ben Skeggs Date: Sun Feb 1 18:58:27 2009 +1000 nv10: use the correct 3d object on NV1A commit dfd87ece014402339f82ecc26bc27b2e47e213df Author: Younes Manton Date: Sat Jan 31 16:52:46 2009 -0500 xv: Unmap filter table BO after init so it can be validated (nv30,40) commit b8de749bbaeacbbe677df1939f7fd39d9d3863a7 Author: Stuart Bennett Date: Sat Jan 24 05:40:02 2009 +0000 Avoid any risk of parsing oscillation on if_is_24bit Would depend on having a BMP laptop with EDID equipped panel and a bios that doesn't match current suspicions, but better safe commit 0a0bb720da63e4a323978e4a52a4d6e1e4f4ae27 Author: Stuart Bennett Date: Wed Jan 21 00:33:55 2009 +0000 Split up parsing the bit structure, complain when necessary tables are missing Added bios major version 0x60 (g80+) check since it now complains about missing the BIT "A" table on nv4x roms Also permits on-demand parsing of individual tables if needed in the future commit 2d86018d52c387d4f2370b7c1ff83798ad6e1cf7 Author: Stuart Bennett Date: Sun Jan 25 20:12:41 2009 +0000 Swap the order of calling parse_fp_mode_table and parse_lvds_manufacturer_table Last seen in this order pre 11de8d7a commit 8c6b3a800ae2df35b6475a46aebab32f0577261a Author: Stuart Bennett Date: Tue Jan 20 04:04:43 2009 +0000 Make calling of init bios functions common Bios structure parsing stage should no longer require hw access This involves serializing BIT hw-accessing functions after BIT structure parsing, but order is maintained so should cause no issues Also, set a digital_min_front_porch default on BIT PPC (G5s), as they have no BIT "display" table (might even fix some wonky display issues like #17571) commit 9b32034eb72e580d4cb5a5e2fa22a76eff0e7f26 Author: Stuart Bennett Date: Thu Jan 29 04:07:04 2009 +0000 Don't read obviously useless bios images, and drop 5x reading on PROM If some nv43 actually need the 5x bodge (no pramin image) we can add it back as a method of last resort PCIR check taken from mmiotrace of the blob's PROM behaviour commit 63aa210f7aee4674337fed3ebfebd198ff1e699b Author: Stuart Bennett Date: Wed Jan 28 16:08:42 2009 +0000 Implement bios I2C opcode 4E (untested) Courtesy an old trace by RAOF commit a81f1590d384bad214111139db9c988ff7676e83 Author: Stuart Bennett Date: Wed Jan 28 15:39:24 2009 +0000 Handle I2C device 0xff fallback for I2C bios opcodes Courtesy an old trace by RAOF Also, use i2c record entry length when available and make rundigitaloutscript function name legible commit 34e967e689a186cdde2f6c3ec222a6870b3584ea Author: Stuart Bennett Date: Wed Jan 28 15:48:27 2009 +0000 Fix pasta error in bios opcode 4D commit 786dd366474006b10c740c5012630ff5f1cd2a0f Author: Stuart Bennett Date: Thu Jan 29 17:56:40 2009 +0000 Fix nv04+ IFC upload and frag prog upload commit 3ab65cf6af0d3f892ade584c29f861f9f8a9d021 Author: Ben Skeggs Date: Thu Jan 29 16:51:53 2009 +1000 add vm_vram_base field to nouveau_device instead of hardcoding 512MiB commit d833c81fadf7dd63ab4ca94b953397fcbe0be581 Author: Ben Skeggs Date: Thu Jan 29 16:25:14 2009 +1000 nv50/xv: previous xv commit was bonged - fix that It actually still worked just fine, but it was wrong. It's possible we'll have more than 1 clipbox, in which case the check for enough space in the pushbuf for a single frame is insufficient. Check for enough space just before emitting quads, if there's not enough flush and reemit state - then go again. commit 0f741eb09c556f8d7893628964fedea862322771 Author: Ben Skeggs Date: Thu Jan 29 16:10:18 2009 +1000 exa: small cleanup commit c10c3c9bbabe8682fe1500098dc53f631690b182 Author: Ben Skeggs Date: Thu Jan 29 15:43:51 2009 +1000 exa: mass renaming and movement of nv04 solid/copy code to nv04_exa.c commit 81ace5dd1fa2c2760036fb1371cc61a383b2698e Author: Ben Skeggs Date: Thu Jan 29 15:29:37 2009 +1000 exa: remove driver pixmaps path that never worked here anyway Bits and pieces remain (disabled) in various modesetting bits for reference. Some of it may be useful when it reappears again. commit 4b1829e2c66f92a19a233afb749f3fb946930db5 Author: Ben Skeggs Date: Tue Oct 7 00:44:46 2008 +1100 exa: remove NVCopyROP table.. commit 410788dc896d99b658508e3a7ebcb4940a039f35 Author: Ben Skeggs Date: Thu Jan 29 14:32:13 2009 +1000 blow up horribly if GPU access to mapped buffer is attempted ...and fix the numerous cases where this was done. The current bufmgr couldn't care less, and synchronisation is done elsewhere. This won't be the case for much longer. commit 4c9346bd750797ee0f0485195f7ce8beb2114b04 Author: Ben Skeggs Date: Thu Jan 29 14:24:57 2009 +1000 bo_del->bo_ref, closer to ng api commit c533f86b512b9d84ec698163a8fbc54233648fbb Author: Ben Skeggs Date: Thu Jan 29 14:13:06 2009 +1000 exa: same for UTS state commit 02677c07f9af01fa3b1eb787d6d320aa43c9140b Author: Ben Skeggs Date: Thu Jan 29 14:10:51 2009 +1000 exa: ensure DFS state is updated across flushes commit 69d02bc73409c8010d19c9f87780b0c0e7621e43 Author: Ben Skeggs Date: Thu Jan 29 13:52:08 2009 +1000 nv50: remove some redundant init, this stuff is done as required now commit 60aea9057a78255a0ded1f64a9b9efd75017347f Author: Ben Skeggs Date: Thu Jan 29 13:50:28 2009 +1000 nv50: forgotten buffer access flags during init commit 5a382dddc948c49f3f5159c078875ee6cdf56db7 Author: Ben Skeggs Date: Thu Jan 29 13:04:04 2009 +1000 nv10/exa: resubmit state in composite as needed commit d4ff8db7b0fb34e0494042b128251ad78c19ad09 Author: Ben Skeggs Date: Thu Jan 29 12:58:30 2009 +1000 nv30/exa: resubmit state in composite as needed commit b71ebcb2c8a0c97837f3de09f034c0bbb35f562c Author: Ben Skeggs Date: Thu Jan 29 12:56:39 2009 +1000 nv40/exa: resubmit state in composite as necessary commit 6ddaddbb7b9442000f1e665bf61994b16d52cd25 Author: Ben Skeggs Date: Thu Jan 29 12:11:24 2009 +1000 exa: rework common pre-nv50 code in the same way nv50 code just was Still have card-specific composite hooks to go. commit f36fa31401b060b9785d1835845c8ce09f8fe6b1 Author: Ben Skeggs Date: Thu Jan 29 11:25:22 2009 +1000 nv50/xv: ensure entire frame will fit in pushbuf without causing a flush commit 1b48331f6bd077f39797bb2d25b62c3c9df5ccb3 Author: Ben Skeggs Date: Thu Jan 29 10:38:53 2009 +1000 nv50/exa: ensure buffer usage is tracked correctly commit 7e644265267b4e35a05328b1c4cb1f2ef5beea58 Author: Ben Skeggs Date: Thu Jan 29 10:13:57 2009 +1000 remove unconditional FIRE_RING() in DoneComposite() hooks Far, far too much flushing. Will completely destroy performance with kernel command submission. commit 34fc4461a6f612e4e00db483effe3a2b9760e70c Author: Ben Skeggs Date: Thu Jan 29 09:47:21 2009 +1000 hack a flush_notify() hook in to match ng preparation for rework of accel code so it'll be more suitable for ng-style command submission. commit d9da090cae1d2dfc539c9db4c923c0e665bce975 Author: Ben Skeggs Date: Thu Jan 29 09:38:39 2009 +1000 Drop GPU resources on LeaveVT, reaquire on EnterVT This allows GDM's switch-user functionality to work on some systems. commit 1eb8b8778037336e530e113b8bbd61d724b1cc52 Author: Ben Skeggs Date: Wed Jan 28 13:50:55 2009 +1000 Enable ShadowFB for DRM-less mode. commit 487463895390500fd4e899f4e3badbb44f2b5429 Author: Ben Skeggs Date: Wed Jan 28 13:41:44 2009 +1000 Allow NoAccel operation without the DRM module present. commit f78aeed720ef185ccdea2d62e02b2714f00ee49c Author: Ben Skeggs Date: Wed Jan 28 10:30:43 2009 +1000 if CARD_INIT ioctl fails, report it instead of continuing commit cf65b875ba3467060d8a3806914b75c710504b78 Author: Stuart Bennett Date: Wed Jan 21 05:22:34 2009 +0000 Fix stupidity introduced in 25de6b867f319099dac05ba84f170da0f0e6c803 Wow. (function executes before bios is parsed, hence version check never works) commit 55e9cdda6e555937e8f5f3998a2e603669b84df4 Author: Stuart Bennett Date: Tue Jan 20 05:09:22 2009 +0000 Split out a function for finding the panel power sequencer stuff and add a bios_t convenience member commit 1b3546b0687a1ae8080009f0c58b26074f08d3af Author: Stuart Bennett Date: Tue Jan 20 04:11:11 2009 +0000 Move the bios flat-panel pointers back into the bios struct Last there in 2007, removed in 82fdd3f76e20fc7dd10fcf00cf1c13520dc56364 :-) commit d29c4815b61635ed53ded94361ba52ee61ea5ac7 Author: Stuart Bennett Date: Tue Jan 20 03:46:05 2009 +0000 Remove a BITism from BMP parsing path, and move old style BMP init into common path commit 3e971f63b6b6859acc19a52a4003bb6041cc5374 Author: Stuart Bennett Date: Mon Jan 19 23:02:09 2009 +0000 randr12: lock independent crtc base registers separately (#19620) Thanks to Mjules for tireless testing, surprising nobody else noticed it earlier commit ffbca7272878d440a06a613df95ce5d7be46f288 Author: Stuart Bennett Date: Sun Jan 18 17:24:47 2009 +0000 randr12: code motion and allow ramdac580 setting on nv40 commit 75a037623adecbe460198066b5059c349de427aa Author: Stuart Bennett Date: Sun Jan 18 15:41:23 2009 +0000 randr12: better selection of nv4x single pll mode and some notes commit 767fcea420fc21c5b5ec6c9f2db56c1a667db306 Author: Stuart Bennett Date: Mon Jan 19 01:51:21 2009 +0000 randr12: no mmiotrace indication that the blob ever clears these bits some hardware drops them on readback, but I don't think that matters also make defined bits shifts comprehensible commit 985ac728276633d14fa5577eacbc2095582b2587 Author: Stuart Bennett Date: Sat Jan 17 23:21:19 2009 +0000 randr12: no mmiotrace evidence for nv47 using the special pll bits commit 3ef6b41a1c072d2f3545915180b522e43e0c8754 Author: Stuart Bennett Date: Sun Jan 18 16:13:57 2009 +0000 twoStagePLL is actually a description of cards using two PLL regs commit c7a423a24ad7da7b510007a0d5a779999bd8cf73 Author: Stuart Bennett Date: Sat Jan 17 04:34:48 2009 +0000 nv04 blitter: use correct number of dma params for real nv04 commit 25de6b867f319099dac05ba84f170da0f0e6c803 Author: Stuart Bennett Date: Fri Jan 16 20:22:16 2009 +0000 Complain more about weird bios images also use bios chip version to test for g80 commit 4d72981cc7d156708683aba51489d4691106438c Author: Stuart Bennett Date: Wed Jan 14 18:15:21 2009 +0000 randr12: avoid classic macro error Guess there's not many ppc nv11 users, as it's been here since Feb '08 commit 133c1a55134853161a25f358432fd05e736718b5 Author: Stuart Bennett Date: Tue Jan 6 17:47:07 2009 +0000 No need of separate allocation for bios image nor of manual allocations for short lived I2C devices commit 8a25049abd2b755187f47572f92b11e9072e522a Author: Stuart Bennett Date: Tue Jan 6 16:55:42 2009 +0000 Make turning on bios execution logging slightly easier and add a minor parsing check commit 54d43c4760b573b5d376859281268046b61fb2f4 Author: Maarten Maathuis Date: Fri Jan 9 20:02:44 2009 +0100 nv50: small fix - planemask is rarely used, so it's just a cleanup. commit 5c9f8d3ce8964ad974175f94b858f9efc11756bb Author: Maarten Maathuis Date: Fri Jan 9 17:48:07 2009 +0100 nv50: Add a fallback for non-repeating XRGB pixmaps. - Other card generations will most likely need this too. commit 0321e6741d9749f9f6b15ac2099139f24e88eef9 Author: Ben Skeggs Date: Thu Jan 8 14:31:39 2009 +1100 nv50: don't segfault if preinit fails before display setup done commit 7dc567d432fc6a3734bf64244e57bcc3612d5ec2 Author: Ben Skeggs Date: Mon Jan 5 10:51:03 2009 +1100 nv50: correct size/layout of uploaded data This should fix the minor corruption that is present while playing some videos. commit 159c18cea2ecb92da74f9a9fdff0a9f5ef03df53 Author: Maarten Maathuis Date: Sun Dec 28 17:56:55 2008 +0100 Make CHECK_TIMEOUT() a little more correct. - Decrease the maximum counter value by a factor 16. - Determine t_start at the start. commit 5d281a2439de1e8c1848b6b700f30476575966e0 Author: Stuart Bennett Date: Wed Dec 24 23:38:02 2008 +0000 nv04-nv4x: support RandR 1.3 panning commit 3c4705cb4472680b6cdc02fd5152d565d214226f Author: Stuart Bennett Date: Mon Dec 15 01:54:12 2008 +0000 Implement bios opcodes 4C and 4D commit 1fd60f202d41d6bb2028955d3bf9e6fc3921e3ff Author: Stuart Bennett Date: Mon Dec 15 01:19:26 2008 +0000 Group all DCB I2C info for a given bus in a structure Furthermore, use the port type field to determine how to drive a bus, rather than the chip(s) characteristically associated with that port type commit 31c9958779d95ae84993f4772607367927a0e313 Author: Maarten Maathuis Date: Fri Dec 19 19:59:53 2008 +0100 nv50: implement get_crtc - Won't help much right now (pre-KMS), but it doesn't hurt to prototype now. commit f7a757827c1d62ac8a7246fe3b64fea1891a7a16 Author: Maarten Maathuis Date: Thu Dec 18 19:57:50 2008 +0100 nv50: implement set_origin (mostly used for panning) commit 74b2e4e92822fb321a9304f0e077c18de57e585f Author: Maarten Maathuis Date: Wed Dec 17 22:42:20 2008 +0100 configure: Avoid compiling drm modesetting code by accident. commit d104f145260ff183b5f7e4c16cebc1680cf771b7 Author: Maarten Maathuis Date: Tue Dec 16 23:54:39 2008 +0100 nv50: Improve lut code. commit 20f93a94ae034ee7c744947f42a29a7c0b50ee9d Author: Stuart Bennett Date: Wed Nov 26 01:37:38 2008 +0000 randr12: pass cursor visibility changes through mode state structures commit 763e0f88c79105a726ea05219875a25fc0129e1f Author: Stuart Bennett Date: Wed Nov 12 20:32:47 2008 +0000 randr12: allow ddc while crtc is locked commit 9d8bd459485392ec21d90c405510f3f097a61921 Author: Stuart Bennett Date: Sat Nov 22 21:20:07 2008 +0000 Punt all bios structs and defines into a new header commit a34a2c07662b0175e03161cb48abaea9d814c1eb Author: Stuart Bennett Date: Thu Nov 20 17:35:59 2008 +0000 Do not attempt to parse FP mode tables for Mac laptops (#18636) The two varieties of these laptops I've had the misfortune to encounter use DDC, and the NV34M based one appears to not include a fp mode table at all commit 1aba513a0b5a00996f28462cd4118214dffd87e2 Author: Ben Skeggs Date: Fri Nov 21 13:16:38 2008 +1100 nv50: program physical addresses into CRTCs, not VM addresses. commit 65b956fb407818117ce17ced6bbe66c55b4acb7c Author: Stuart Bennett Date: Mon Nov 17 20:02:19 2008 +0000 randr12: fix stupidity from 539f4990 found by AVRS2 commit 1666e858dc027bc028083c15618da5cde5a1b6f1 Author: Maarten Maathuis Date: Sun Nov 9 23:10:15 2008 +0100 nv50: use bios provided load detect value commit a020a44a2c48ff20bca6188469cb9cbbb5bf6093 Author: Maarten Maathuis Date: Sun Nov 9 22:56:30 2008 +0100 bios: support g80 loadval table commit 220cbb6347cb0404bd24f457ac6cb5e39917cdbf Author: Maarten Maathuis Date: Sat Nov 8 18:01:18 2008 +0100 nv50_randr: plug a small memory leak commit cea05e14f94dcb842adbfcec515e17e9f5655089 Author: Stuart Bennett Date: Sun Oct 26 03:51:36 2008 +0000 randr12: make colour-map setting less complex commit 0e4b01a5e06d8b5323c010ce89723369f17b10b2 Author: Stuart Bennett Date: Sat Oct 25 02:11:37 2008 +0100 randr12: convenience functions for vga crtc state access commit fe9bfd9b1b5ce6fd0cd0360ec850ae48b4abac41 Author: Stuart Bennett Date: Fri Oct 24 21:02:00 2008 +0100 Tidy EnterVT, and don't call it from ScreenInit commit 3e1ccff8744d9cabef0cd3901e987deeeb38ec12 Author: Stuart Bennett Date: Fri Oct 24 02:49:19 2008 +0100 Condense crtc locking We unlock all crtcs whenever the driver is active, and relock them when returning to VT (which is what we were doing, but in a more complicated manner) commit 83b541e0a9c454059d0e55a932a2ecdbbf650d60 Author: Stuart Bennett Date: Mon Oct 27 19:20:22 2008 +0000 Common function for the bios condition table commit 36c6aa449063387b34c31015d618e53d4b6634e0 Author: Stuart Bennett Date: Mon Oct 27 18:55:28 2008 +0000 Allow more time in INIT_CONDITION_TIME (#18234) commit bf810b32bc959f3821bca18cf158f711dd314a26 Author: Stuart Bennett Date: Mon Oct 27 18:41:43 2008 +0000 More warnings for C51's magical unaligned regs commit d481b5e2e1dcb7ac5b5a7583bc6542b42567eac2 Author: Stuart Bennett Date: Sat Oct 25 17:48:35 2008 +0100 Rework bios parser error handling commit 387f30821ff7add22376de33a203744cdc719849 Author: Stuart Bennett Date: Sat Oct 25 00:38:29 2008 +0100 RivaTuner dumps call the `how much memory have I got?' reg NV_PFB_CSTATUS commit bb20175001748e5d2fab1b9cb2075bbe5d94e5ed Author: Stuart Bennett Date: Fri Oct 24 17:42:53 2008 +0100 pedantry fix for nv11 locking, and some minor cleanups commit 9aecc4226ca9067667004150f7d885b2247273b2 Author: Stuart Bennett Date: Thu Oct 23 23:17:15 2008 +0100 Add a function to lock/unlock all crtcs, use it in nv_bios code Also split initial head owner determination into separate function commit 48581da4d0e1663456ddcbf7a456029cf961d4cd Author: Stuart Bennett Date: Thu Oct 23 02:35:30 2008 +0100 Attempt to make the usage of cr44 rational instead of cargoculted and a couple of minor cleanups in nv_crtc commit 1c7bee95610a47175064b17afd6aa2a45e311b1a Author: Stuart Bennett Date: Thu Oct 23 02:46:03 2008 +0100 Robust nv11 head setting (from mmiotrace) commit cf5162b5ed99666d3dc8db8328115afd6fff6a8e Author: Stuart Bennett Date: Tue Oct 21 02:12:24 2008 +0100 randr12: update scrn field on server regen, and emit an error if rotated (updating the scrn is easier than removing and re-adding outputs) commit 13739cf36893e8ab6d3d4911c76a45f0b8bd5f4c Author: Stuart Bennett Date: Mon Oct 20 21:47:07 2008 +0100 randr12: don't change virtualX/virtualY all the things in the ddx that use virtual* (setting displayWidth, allocating the FB, EXA init etc) use the values present before this function is called these prior values are derived from any Virtual line in xorg.conf and are what we want to use setting virtual* to the crtc size here, rather than the desired Virtual size, breaks everything using them in ScreenInit on server regeneration as the rhd guys say: /* This is strange... if we set virtualX/virtualY like the intel driver * does, we limit ourself in the future to this maximum size. * The check for this is internally in RandR, no idea why the intel driver * actually works this way... * Even more curious: if we DON'T update virtual, everything seems to * work as expected... */ so I'll take that as precedent. commit 98a4dcb112221fa66d82b87ef9229f5c443daf90 Author: Stuart Bennett Date: Mon Oct 20 18:59:35 2008 +0100 Delete notifiers and grobjs in CloseScreen to avoid channel member becoming stale on server regeneration commit 12fce00b2ecf4c6be6cc86c8a7374878bb7dfc20 Author: Stuart Bennett Date: Wed Oct 22 02:47:45 2008 +0100 Remove all object with mmaps in CloseScreen, so that drmClose actually calls the drm release method (and so subsequently, we can be DRM master on the next drmOpen) commit 8e3f27a30078675f1a7974ba436a68a5b45e0924 Author: Stuart Bennett Date: Wed Oct 22 03:32:51 2008 +0100 No reason to call AccelCommonInit twice in ScreenInit, but calling it in EnterVT helps resume... commit f495fa9153ee57d3ee0dc735c3e9e836d84a2ccc Author: Stuart Bennett Date: Wed Oct 22 02:41:17 2008 +0100 nouveau_channel_free should remove mmaps made in nouveau_channel_alloc commit 1cedb8e49c832890efd12904e78dfd8ab5855691 Author: Stuart Bennett Date: Wed Oct 22 02:15:06 2008 +0100 Do frag prog allocation and shader upload in TCL init Similar in spirit to the nv50 inplementation. Removes many dependencies on state stored in static variables, which in turn makes reinitializing shaders on resume / server regeneration tractable. commit 9e2c08937b0d1bfed71997b2b16bded91830af71 Author: Stuart Bennett Date: Wed Oct 22 01:44:10 2008 +0100 Only hackup shaders once commit afc6668d2b4a1971c211c421b6a415fe74dbd2ea Author: Stuart Bennett Date: Sat Oct 18 02:23:18 2008 +0100 cr26 is just another view of 0x3c0 remove superfluous crtc unlock too commit 539f4990c8f1a32da57992dc863f277f8ede8194 Author: Stuart Bennett Date: Wed Oct 15 03:59:06 2008 +0100 Use symbolic define values where known commit 791666a76c68880136ae26d0477bf484cc667f72 Author: Stuart Bennett Date: Wed Oct 15 00:25:26 2008 +0100 Convert CIO, DIO and VIO use to use defines from nvreg for index and data reg offsets commit a1b7f8d9b20bff59956ea1dc2130e5590a60aa4c Author: Stuart Bennett Date: Wed Oct 15 00:12:11 2008 +0100 Rename relevant functions, sizes and offsets to PRM.IO from P.IO, in keeping with the nvidia scheme commit 267c0eef02f5e2a48217235d170a0530de6a94d3 Author: Stuart Bennett Date: Tue Oct 14 23:58:11 2008 +0100 Name crtc index regs according to rules.xml commit 9261c34e873356b02a0fda00e9911d9795d9f1f2 Author: Stuart Bennett Date: Tue Oct 21 17:46:37 2008 +0100 SaveGeneration isn't doing a lot add another close screen thing too commit c29c190036c721dac2e58e66d0c5a70a24377a1d Author: Stuart Bennett Date: Tue Oct 21 16:27:00 2008 +0100 Replace a few memsets with initializers commit 3b53f6ceb86458f9df9b7becc18a79e74b21c91f Author: Stuart Bennett Date: Sun Oct 12 15:52:43 2008 +0100 Eliminate separate NVRec ctor and dtor commit 1b18db28f124bd5021f61b72389e5c0ce003b212 Author: Stuart Bennett Date: Sat Oct 11 21:05:57 2008 +0100 randr12: remove pointless debug in nv_output and nv_crtc commit 240d51dc928cea38f6b9d842b59011ad7c5500c0 Author: Stuart Bennett Date: Sat Oct 11 18:40:35 2008 +0100 randr12: deBoolification and eliminate `override' arg to nv_crtc_load_state_ext commit b95c3e23f0eac7d12d5ee271af8ec5707ac4ef8e Author: Stuart Bennett Date: Sat Oct 11 18:29:40 2008 +0100 Make all CR 57/58 access use the proper functions and defines Use a define for usleep in bios too commit 9988ae594f6f03385b261b0af83b324981f11593 Author: Stuart Bennett Date: Sat Oct 11 17:56:23 2008 +0100 Redo bios logging commit 0e6a9cc61c88deaf0e0935d5a7d269621c3d6cd0 Author: Stuart Bennett Date: Sat Oct 11 18:04:58 2008 +0100 Improve some nv_bios messages, remove others commit b175bfcd6485d9dcddb66ab5f79104567d256336 Author: Stuart Bennett Date: Sat Oct 11 17:59:02 2008 +0100 Bios register list is easier to read when sorted numerically and remove NULL initialization for calloc'd struct commit e51b49c938483ccbd172ff07d0385a6fb474b57e Author: Stuart Bennett Date: Sat Oct 11 02:35:40 2008 +0100 randr12: panels with edid have only a constant mode prove me wrong. commit 765494ebf9b129bd613121cb714e928b09e19500 Author: Stuart Bennett Date: Sat Oct 11 02:16:50 2008 +0100 Remove unnecessary "Setting owner" message commit bf585ad38a80f7828b821936296316bcee4658f5 Author: Maarten Maathuis Date: Sat Oct 11 01:22:35 2008 +0200 The fifo channel should be removed along with closing of the fd, so only clear the client memory. commit 8b7e42461add708e75d3f7d675245f5bbf3652e8 Author: Maarten Maathuis Date: Fri Oct 10 23:37:55 2008 +0200 git-log is no longer valid for git 1.6, switch to git log. commit 933fd803f0aadc934f448937dbeb305256fe8893 Author: Maarten Maathuis Date: Fri Oct 10 23:14:57 2008 +0200 Some fixes that bring me closer to surviving to the 2nd X server generation. - Not quite done, i get a fifo dump when something tries to render. - Something is keeping an fd, which is preventing a full takedown, but this is not a valid reason for failing (kernel modesetting for example will never have full takedown). commit daee3c38a02f48f9094f712800174b88b065f0ff Author: Stuart Bennett Date: Thu Oct 9 17:09:58 2008 +0100 randr12: separate cursor pixmap location setting function commit caf2e920f18cad1c57e945a4e474e63a65ec6a43 Author: Stuart Bennett Date: Thu Oct 9 17:07:30 2008 +0100 Move cursor show/hide funcs to nv_hw, document nv40 bug, and set curctl2 before applying nv40 fix commit f76074c7041b1246efdf2c0a4b0c467e2fe0bef1 Author: Stuart Bennett Date: Thu Oct 9 17:24:23 2008 +0100 Use correct refclk for nv_get_clock, and use bool type as appropriate in arbitration commit bd5b8074b4a250610cf8c9e2f181c2946e902bf8 Author: Stuart Bennett Date: Thu Oct 9 15:25:16 2008 +0100 Tidy arbitration prototypes and split randr12 cursor pixmap setting from arbitration Also fix ages old warning in nv_hw commit 5dc41fdcf4ba76d7154d32f030d2e99756681e4a Author: Stuart Bennett Date: Fri Oct 10 01:11:46 2008 +0100 Merge nv4, nv10 and nForce UpdateArbitrationSettings functions commit 0ffd012c922d241aec22c8e44416171e39e6bf0c Author: Stuart Bennett Date: Fri Oct 10 00:57:45 2008 +0100 Add PCI slot reading defines, and use them for nForce code Also move nForce dimm checking code to nv_setup, so it is only tested once commit 520a76ef3df3cdc735691cfda329ceaae0ec3123 Author: Stuart Bennett Date: Thu Oct 9 20:22:34 2008 +0100 Some tidying of arbitration functions commit 3ceb2b60a09379ef51635f8309fdc82060681e8f Author: Stuart Bennett Date: Thu Oct 9 19:05:01 2008 +0100 Reindent arbitration stuff commit d7439fc50a924f6f1a6b5b59ba3ec076cf90c52e Author: Stuart Bennett Date: Wed Oct 8 23:45:13 2008 +0100 Wrap usleep when called in bios code commit 62cf9f371c420abe7a74ac4bb9fa2da081f1d29b Author: Stuart Bennett Date: Fri Oct 10 16:41:29 2008 +0100 randr12: don't use in use (by another vga output) crtc for load detect (avoids flicker) commit f1d0fd0595b40d3a7f555bada88f1f79b517d610 Author: Stuart Bennett Date: Tue Oct 7 21:39:10 2008 +0100 Better behaviour on allocator fail also fix randr12 code which doesn't know how to call calloc and be sure to use the number of actually inited crtcs commit 2b67015e5911cb7e321da72d7d3d5f0fefedd8e0 Author: Stuart Bennett Date: Tue Oct 7 02:35:44 2008 +0100 randr12: reorder nv_output functions roughly into order of use commit 9f4eef0165766be2e1aa5516eae409e94c09553b Author: Stuart Bennett Date: Thu Sep 18 18:32:36 2008 +0100 randr12: distinguish detected encoder from active encoder don't swap output props until a mode is set on new encoder either commit 549cda114ee378640dda89621b0ce9c937c473da Author: Stuart Bennett Date: Wed Oct 8 01:22:17 2008 +0100 Reduce includes, everything necessary is in nv_include.h commit 30eaa0f08250f72b0171a858f2694ca8a62c3ebd Author: Stuart Bennett Date: Sun Sep 21 00:52:56 2008 +0100 Oops. commit 2481e5646ab3ea5e18d3f5ee4042fa20a8dad3de Author: Stuart Bennett Date: Sun Sep 21 00:44:50 2008 +0100 Optimizing memory usage for unaccelerated case is not interesting commit 19f0ea63a505e6a840cf07bc554fd8f1787616b4 Author: Stuart Bennett Date: Sun Sep 21 00:43:57 2008 +0100 randr12: rename mon -> edid commit 0cb081758f52e73c3055758ed7aced12b93c2519 Author: Stuart Bennett Date: Sun Sep 21 00:31:40 2008 +0100 randr12: crtc destroy also alleviate world shortage of curly braces commit 37adca1dd68d5791b0371eed4b96f6613fc747c2 Author: Stuart Bennett Date: Sat Sep 20 23:57:06 2008 +0100 Need to increment dcb entry index when fabricating additional entries commit e6b6d8c9189f3b5ebbc77f55e318246be0fa1bd4 Author: Maarten Maathuis Date: Fri Sep 19 21:23:04 2008 +0200 Also remove man page entry. commit cb1e9c1ab288bcf380789f775533e4c8ff8705db Author: Maarten Maathuis Date: Fri Sep 19 21:14:09 2008 +0200 NV50: remove {Prepare,Finish}Access since i realise now it's not so useful - despite that the xserver code suggests that pixmaps can be offscreen on this call, that is not done as far as i can tell - not to mention that there is no way to know for sure if a pixmap is offscreen (exaPixmapIsOffscreen is not exported), making the offset useless - this will (obviously) become relevant as soon as driver managed pixmaps come into play, but that is for another day commit 3e397f575df6c3fee6eb9ee810fb3fc70d2c0cc8 Author: Maarten Maathuis Date: Fri Sep 19 20:41:59 2008 +0200 NV50: The TMDS dual link threshold is meaningless for LVDS commit 599d258948be0505f02947251ad6877e7168d9a8 Author: Stuart Bennett Date: Mon Aug 25 01:47:45 2008 +0100 randr12: multiple encoders per connector (DVI-I) that this doesn't break things in subtle and unexpected ways is too much to hope for. one not-subtle and expected way things will break is that xorg.confs with specific setups for DVI-A-n and DVI-D-n connections will need renaming (to DVI-I-n). as the last patch in the series, this also fixes an issue introduced with 144b04a2 (randr12: fix dpms, detect, destroy, ...), that for a screen connected to a dvi connector, both the analogue and digital variants were detected as connected commit 5e8ac848936a1b70babc417454599c0231343399 Author: Ben Skeggs Date: Wed Sep 17 22:12:07 2008 +1000 nv50: 0xa0 family uses class 0x8397 for 3D commit 9c36eefe991daf47d20badffb3009def99473bdd Author: Stuart Bennett Date: Sun Sep 14 23:01:49 2008 +0100 randr12: avoid weirdness when tv-out happens to share i2c_index with another output commit 144b04a258eaa8e3b63c48b5e574524b40a59ac7 Author: Stuart Bennett Date: Sun Aug 24 04:15:20 2008 +0100 randr12: fix dpms, detect, destroy, save and restore for multiple encoders per connector commit 2b30b8914a32049e87205cf896b2e89ab6649bca Author: Stuart Bennett Date: Sun Aug 24 04:26:19 2008 +0100 randr12: re-do detection of dvi-a vs vga commit 50bac3fcd24072b8926ba2dbebd571424fbb7d6d Author: Stuart Bennett Date: Sun Aug 24 03:56:16 2008 +0100 randr12: unify output funcs for analogue and tmds commit 69e30e6727895a8edeffc9a348b13a353cd0c6d0 Author: Stuart Bennett Date: Sun Aug 24 03:12:58 2008 +0100 randr12: unified entry path for dpms commit a794a0c864e5b651bea5e40a4e10978aee9da5b6 Author: Stuart Bennett Date: Sun Aug 24 04:45:48 2008 +0100 randr12: nouveau_output -> nouveau_connector commit 0a5119bd8a927da1f3f4ed6caa2e93acdae29c04 Author: Stuart Bennett Date: Sun Aug 24 02:45:41 2008 +0100 randr12: separate encoder struct commit 352b8fd98c519f85ade02aa679d342bc64faa508 Author: Stuart Bennett Date: Sun Sep 14 16:53:12 2008 +0100 Another DCB 1.4/1.5 TV (#17471) I think we've got enough for TVs to no longer be fatal commit e8ba4e0baa62f02ba85258847963512791d89385 Author: Stuart Bennett Date: Sun Sep 14 16:44:32 2008 +0100 Only use OF bios image on PPC commit 54b8a7ea43ec20c798dc93d881cdffa5cb1125a0 Author: Ben Skeggs Date: Thu Sep 11 03:29:46 2008 +1000 nv50/exa: work around corruption issues (see detailed commit msg) It's not possible for us to do this properly without using the exa pixmap hooks, they require a real memory manager, and is being worked on in the ng branches, which doesn't currently support nv5x at all. For a better solution until then: http://nouveau.freedesktop.org/wiki/NV50Support commit f8c0af3886142ea59936a77006ead0cd2c4ff795 Author: Stephane Marchesin Date: Fri Sep 5 23:08:14 2008 +0200 Partly revert ea152819f45b6cf92f6742ed3f9f639cdbbd53f0, alignment is needed. commit ea152819f45b6cf92f6742ed3f9f639cdbbd53f0 Author: Maarten Maathuis Date: Fri Sep 5 22:51:27 2008 +0200 exa: align offscreenBase to something sensible - Aligning virtualY makes no sense whatsoever. - displayWidth is the "pitch-width" of the framebuffer. commit 95c19b98b38310c543758fe4288a9e47b970fe67 Author: Stephane Marchesin Date: Thu Sep 4 07:51:03 2008 +0200 Add 8200 detection. commit 6dd8ad4ed3f89c6ed9e61708a587d417c8d984fe Author: Maarten Maathuis Date: Sat Aug 30 17:51:05 2008 +0200 nv50: solid fill shouldn't use a pattern rop + minor changes commit b2b726d59fd8af819a7e464824a49f6e6fc10b84 Author: Stuart Bennett Date: Sun Aug 24 00:18:40 2008 +0100 randr12: merge output detection functions commit a223e581b06ffb308aecc2365795951e2c021f29 Author: Stuart Bennett Date: Sun Aug 24 00:08:41 2008 +0100 randr12: remember the edid from detection for getting modes later also split native mode finding into separate function commit 16d9e89d9cfae1c21999d880c557ae283e14178e Author: Stuart Bennett Date: Sun Aug 24 19:56:24 2008 +0100 randr12: detypedef private structs and use a define to get to the privates commit 3b3f3cdbe0f69c57153e070f248ff001cd5a4a17 Author: Stuart Bennett Date: Sun Aug 24 00:31:35 2008 +0100 Remove unused code commit 1e2b87c459a3a0eacd630f1de8b567c36cc792bd Author: Stuart Bennett Date: Mon Aug 25 20:23:05 2008 +0100 randr12: make the nouveau experience less green commit f89af0e6bc5ac88d0755f9402b772ca823518298 Author: Maarten Maathuis Date: Sat Aug 23 03:44:28 2008 +0200 nv50: micro cleanup commit 94bf1060cc9d2f4c0a14bc036f8b76c4d9bfc7d7 Author: Maarten Maathuis Date: Sat Aug 23 03:39:39 2008 +0200 nv50: do ROPs properly this time - Table should work for pre-nv50 as well, i just don't have a card plugged in to be sure. commit 6bd14e4ffe1d5fd9420c4ebb46c9c0844d0547f9 Author: Maarten Maathuis Date: Fri Aug 22 17:56:31 2008 +0200 nv10: composite is an async operation by default, no need to call exaMarkSync(). commit 9f11d13e68bbf72be814b37fa3d03853eec981b9 Author: Maarten Maathuis Date: Fri Aug 22 17:53:11 2008 +0200 exa: FIRE_RING at the end of composite commit 137d099294617adead2cf3d5cc3372d18579eaef Author: Maarten Maathuis Date: Thu Aug 21 23:17:06 2008 +0200 nv50: move VERTEX_{START,END} to {Prepare,Done}Composite - If PrepareComposite returns TRUE, then everything upto DoneComposite is required. commit 34dc05fbc4835ec68de5a1f70cc094ff6efcf310 Author: Maarten Maathuis Date: Thu Aug 21 22:50:49 2008 +0200 nv50: minor tweaks to composite - Make the blend stuff more consistent with other composite implementations. - Verified with "rendercheck -f a8r8g8b8,x8r8g8b8,r5g6b5,a8 -t {blend,composite,cacomposite}" for all accelerated ops. commit eca997750b1cc2263f804afba242f16b6d960e02 Author: Maarten Maathuis Date: Thu Aug 21 20:01:49 2008 +0200 nv50: handle ROPs better - Handle ROP's more like pre-NV50 code. - Special case for 32bpp which seems to work. - This does NOT fix all rendering issues, just an improvement. commit e468df855abd8d8e4fecd2445c82bb8f824297fb Author: Maarten Maathuis Date: Wed Aug 20 19:14:54 2008 +0200 nv50: clip SIFC and add {Prepare,Finish}Access hooks. - when w*cpp is not aligned to a dword, then it's possible to end up with crap at the end of a line, so clip it. - {Prepare,Finish}Access hooks force tiled memory to be migrated upon cpu access. - Added an option to disable these hooks in case you hit a xserver bug that existed up until very recently. commit 0c0c1ab2ce786def653e8adb95c7c5a8ff9c7eca Author: Stuart Bennett Date: Tue Aug 19 00:02:00 2008 +0100 randr12: don't cache pll values (fixes newrestore removal regression) for 5ms, it's really not worth it commit 544fd3c502287754f2161c21514203ccaab8aed9 Author: Stuart Bennett Date: Mon Aug 18 21:50:37 2008 +0100 Remove some of the inaccuracies in the manpage commit e3d9cb926e56d28b7cadd0026ae76c076faced75 Author: Stuart Bennett Date: Mon Aug 18 21:40:36 2008 +0100 NV_ARCH_04 can have rotated output, it's just not accelerated commit 40e920fafb9f22da23ef11c941561bb14966dc47 Author: Stuart Bennett Date: Mon Aug 18 21:39:18 2008 +0100 Delete ShadowFB rotation And nothing of value was lost. commit 6352d7db24451b014b5f6b873881d103fc1eb316 Author: Stuart Bennett Date: Mon Aug 18 00:59:32 2008 +0100 Bye bye 8 bit depth commit 33d52e476b64cd91dc4b01c6826892f2da9499b0 Author: Stuart Bennett Date: Sun Aug 17 23:52:28 2008 +0100 CrtcNumber option is long gone commit a9393a87d36cc486114842275bb78e7e75de754e Author: Stuart Bennett Date: Sun Aug 17 23:17:01 2008 +0100 Micro-tidyups to nv_driver and nouveau_xv commit 8ed30d05afe4a5bb077c088c48c28741414c09d3 Author: Stuart Bennett Date: Sun Aug 17 16:57:17 2008 +0100 randr12: a few line length improvements commit 666ab70754035c569eaa2550ca77c92fe3eaaff3 Author: Stephane Marchesin Date: Sun Aug 17 22:44:11 2008 +0200 Kill some 8 bit code. commit bee8450931395b23040a4cb1ddcb09f8a476cb93 Author: Stephane Marchesin Date: Sun Aug 17 22:25:43 2008 +0200 Fix IFC for 16bpp. Also fix the color accuracy of 16bpp solid fills. commit 609b10ff00c9e8d51fd7c9a5fb62120516780aed Author: Stuart Bennett Date: Sun Aug 17 03:51:39 2008 +0100 0x48 exception does not apply on BIT cards commit a9e2d47e2cee7db447063d7c33f3e42a8d865c92 Author: Stuart Bennett Date: Sun Aug 17 02:48:33 2008 +0100 randr12: some code simplification, rearrangement and tidying We don't claim to support a bpp of 15 or 24, so some old calculations can be simplified commit 5715f95b37ad182377024f468982608b48eb3409 Author: Stuart Bennett Date: Sun Aug 17 02:27:42 2008 +0100 NewRestore: remove, due to lack of interest nobody wants to do text modes manually. for any future masochists, it's here in the history commit d2e924e847273c5b4d342d0ad611de7d106df429 Author: Stuart Bennett Date: Sun Aug 17 01:37:41 2008 +0100 randr12: remove another field member commit 55f490c8c29b103d7d8853e2126a8366304f4125 Author: Stuart Bennett Date: Sun Aug 17 01:26:20 2008 +0100 randr12: fpWidth and fpHeight no more commit 4cbbd44db614d250fdaaf648db2d70ec9bc2f6d9 Author: Stuart Bennett Date: Sun Aug 17 00:38:23 2008 +0100 randr12: improve aspect scaling code commit 8a364682fd67c10c153cb0778c914a4754e82ddd Author: Stuart Bennett Date: Sat Aug 16 23:15:29 2008 +0100 randr12: for native scaling, a clock check isn't enough 1280x1024 and 1280x960 can have the same clock commit 8793bb1d7b79632678d7ce9f1e94c79dfd55fa3c Author: Stuart Bennett Date: Fri Aug 15 23:02:48 2008 +0100 Apply minimum front porch only when necessary (#15949) Also, read minimum front porch from bios table commit 83bb26fd9a8bb283c406bc34d1d7ad0b53d7098a Author: Maarten Maathuis Date: Sat Aug 16 17:18:21 2008 +0200 NV50: disallow doublescan modes on TMDS/LVDS commit 7d5b345f69bf63d36207a7f6ac4d01001d62e79c Author: Maarten Maathuis Date: Sat Aug 16 15:33:13 2008 +0200 NV50: extra warning message commit 743c696c8b0e420a67ce1552ac0cec49c0fdc0f2 Author: Maarten Maathuis Date: Sat Aug 16 15:28:43 2008 +0200 NV50: a half decent attempt at doing something when LVDS has no DDC commit fcbaa10a945e04d77c6ea8f83b738fb6c9e47dd5 Author: Maarten Maathuis Date: Sat Aug 16 14:35:05 2008 +0200 Forgot something. commit 628acddedc7755a16471e9beee3d95d3b8f11150 Author: Maarten Maathuis Date: Sat Aug 16 14:24:18 2008 +0200 NV50: support extended i2c ports (4 and 5) commit c845f5ac0a54b17a9911155030234c229e80626c Author: Maarten Maathuis Date: Sat Aug 16 14:02:51 2008 +0200 nv50: fix some of the rendering bugs - It seems that 32bpp solid fill only works well for for the non-rop case. - The SRCCOPY_PREMULT comes from nv and seems to work well, in combination with the above fallback. - This does not affect the strange stripes with font rendering issue. commit 18288208bdb7f872f0a2e227d20b77c3397747c2 Author: Stuart Bennett Date: Fri Aug 15 01:40:01 2008 +0100 Redo parsing for panels with EDID, and fix nv3x in the process (#17138) breakage a possibility :) commit c9d4e86022bffd00d89076a921b734645145f7db Author: Stuart Bennett Date: Tue Aug 12 14:15:03 2008 +0100 Try harder to load a good vbios image commit ae594786ffc1cdca5bdd8c8638225819f851061b Author: Stuart Bennett Date: Wed Aug 6 00:31:39 2008 +0100 Deal with mobile cards that scribble over the fp strap at POST It's probably safe to assume a mobile card will have been POSTed at boot, so this should be fine. Hopefully there are no cards which have bit 2 set at 0x48 and now break... commit 776670613f79cce3891c65d2f688979767cd1e6d Author: Stephane Marchesin Date: Fri Aug 15 00:43:02 2008 +0200 Fix a 16bpp issue, also remove old code that mostly worked around core EXA issues in older xservers. commit 6854f819f6bb89d67bfb737fd24c3e8b16e67e46 Author: Maarten Maathuis Date: Tue Aug 5 13:08:22 2008 +0200 NV50: Disable messages from hide, show and load cursor, as they clutter the logs. commit 4b24be811b839f9f836b9f8574fc808d2182dd30 Author: Ben Skeggs Date: Mon Aug 4 16:09:28 2008 +1000 xv: fix thinko from when source was reformatted commit 30f54f2a4b81afde129170a26189cdb2bd3c7e9c Author: Younes Manton Date: Sun Aug 3 12:29:57 2008 -0400 g3dvl: Temporarily disable IDCT. commit 89d786435267af58bcecaff206a09b035e666df5 Author: Ben Skeggs Date: Tue Jul 29 23:38:45 2008 +1000 nv50: fix mystery typo commit c0b67f3f3490e826c86eedb32bbcf327592aaaf5 Author: Ben Skeggs Date: Tue Jul 29 01:26:20 2008 +1000 remove use of implicit variables from pushbuf macros This matches the gallium winsys now. At some point the ddx will get updates so the "common" code is once again in sync to make it easier to do mm changes to both the ddx and dri simultaneously. But, I want to get this part out of the way first as the changes might take a bit of work - and I don't want to have to deal with merge conflicts everywhere if someone changes the accel code before I'm done :) commit 80278c9f53f7c357e499bf2b3c5a1dff75fb9651 Author: Younes Manton Date: Sun Jul 27 16:48:58 2008 -0400 g3dvl: Basic XvMC initialization. commit 1f8e27fd1d92d1c07dd8387f1a5e751ab3fab54a Author: Maarten Maathuis Date: Sun Jul 27 00:09:02 2008 +0200 NV50: I was proven wrong quicker than i thought, reinstate a slightly prettier hack. commit 41d46f59eb055130d5d415f0573fbede957c66c3 Author: Maarten Maathuis Date: Sat Jul 26 21:53:26 2008 +0200 NV50: forgot something commit 1b905249c492af4485a9a2f4e7ae3ad5441c8eb5 Author: Maarten Maathuis Date: Sat Jul 26 21:50:34 2008 +0200 NV50: my best guess at the lvds bios table, time will tell if it's correct for more than a few cases commit f3ec6e0e2d84e172cb634fd1348a54aef7256e74 Author: Dave Airlie Date: Thu Jul 24 15:41:57 2008 +1000 nouveau: fix macro as pointed out my moondrake on irc commit 6652e9c85fe8e807ed3d8ff76ec18c250aef436f Author: Ben Skeggs Date: Thu Jul 24 10:13:05 2008 +1000 nv50: support YUY2 in textured video adaptor commit 3534f40b64496d9f6f95148ab81b9321057f7af1 Author: Ben Skeggs Date: Thu Jul 24 09:39:09 2008 +1000 nv50: remove 32bpp solid fill fallback It seems to work, but there's probably a good reason it's a fallback in nv, so may be reincarnated with a more specific fallback if needed. commit 87f7d1f7084032110c347cb73f63e9a929d8c00e Author: Ben Skeggs Date: Thu Jul 24 09:38:27 2008 +1000 Revert "NV50EXA: read the notes + cleanup + enabled 32bpp solid fill" This reverts commit 0b6249fed6909af30856e1afc1a0dd4cd98c0b9e. commit ba7c239e7a13c4726dab7c686ee1c7f34d03a593 Author: Maarten Maathuis Date: Thu Jul 24 01:30:16 2008 +0200 NV50EXA: serious brain fart of my part, sorry - prepare copy may set up a rop and will set operation appropriately. commit 0b6249fed6909af30856e1afc1a0dd4cd98c0b9e Author: Maarten Maathuis Date: Wed Jul 23 21:01:57 2008 +0200 NV50EXA: read the notes + cleanup + enabled 32bpp solid fill - I observe occasional corruption, some of this goes away when using a compositing manager. - The above statement contradicts my earlier statement in the previous commit message. - 32bpp solid fills seem to work fine, but i left the original fallback in case something pops up. - There is definately something wrong somewhere, i just don't know what. commit edbfbd09ac0a05c4125c3217f5d9c8f198d9345d Author: Maarten Maathuis Date: Wed Jul 23 13:34:28 2008 +0200 EXA: put NOUVEAU_FALLBACK in a do { } while (0), fixed opera corruption for me. - I don't understand why a single line define would need this, but i noticed that turning on the fallback info fixed rendering. commit 36d13082dd2c2ba5d490b3dbc6e82d9ccfed8669 Author: Maarten Maathuis Date: Sun Jul 20 15:44:47 2008 +0200 NV50_KMS: some basic scaling and dithering output property support commit 022a9ed5c2657466dd2471a1b29d5e8bc92ca57a Author: Ben Skeggs Date: Sat Jul 19 22:14:50 2008 +1000 nv50: no more sync in composite() :) 0x502d mthd 0x0110 some kind of 2D cache flush? commit 6d8096c72e1aa160cef6729443ed06305a8ae870 Author: Ben Skeggs Date: Fri Jul 18 13:57:13 2008 +1000 nv50: sync after composite for the moment This corrects some severe text rendering issues that get *very* bad with server 1.5. Will look into a proper fix soon. commit c68d8806b309ef4b48d7a98d6fc7aeccb8bb05ac Author: Ben Skeggs Date: Fri Jul 18 13:39:59 2008 +1000 nv50: punt vertex emission macro out to header commit 4ad74cb03070f5279e0b1de2ed326f1167400d60 Author: Ben Skeggs Date: Thu Jul 17 11:15:59 2008 +1000 exa: missed a reloc delta Can't actually hit this bug currently, but one day perhaps :) commit 98a9056224a86244f99ef7e844cda363e563a446 Author: Ben Skeggs Date: Thu Jul 17 11:15:08 2008 +1000 nv50: remove acquire/release surface stuff It's left over from code that never made it to git in the first place, long overdue removal. commit f0305a29c2b4010894362d17f21cdfec39cd7b92 Author: Ben Skeggs Date: Thu Jul 17 11:14:02 2008 +1000 exa: rename m2mf locals to prevent some confusion commit 140c36ed936f4f28842af88efdd8b824c01bb097 Author: Danny Date: Fri Jul 18 03:07:38 2008 +0100 randr12: fix dithering output property endian issue (#16624) tweaked and commented by Stuart Bennett commit b36802fe90165cf0fea61728b06605be5ac25d8d Author: Pawel Worach Date: Fri Jul 18 01:17:18 2008 +0200 Fix for #14858 - freebsd compile. Someone still needs to fix the kernel part though. commit 6f5e90ae739c64478ddf3e042e2bc9c4ff2b951c Author: Ben Skeggs Date: Sat Jul 12 02:24:54 2008 +1000 nv30: more header name changes... commit 4b8427af583a8f84a09e8bd8d6ad614575a4f9c7 Author: Ben Skeggs Date: Sat Jul 12 02:15:08 2008 +1000 nv30: more bustage from header update commit 1672a783c4af58c80ed549fb3dc63c07fc75a882 Author: Ben Skeggs Date: Sat Jul 12 02:03:58 2008 +1000 nv30: un-change something that somehow changed in an earlier commit. oops! commit dae59585abec4b514789a8e580d30757f10d0357 Author: Ben Skeggs Date: Tue Jul 8 12:28:18 2008 +1000 nv50: exa/xv share some stuff, punt it out to nv50_accel.[ch] commit 17fbd810748b61b298c1ea9729d94f7d1177253b Author: Ben Skeggs Date: Tue Jul 8 12:09:09 2008 +1000 nv50: Xv support The NV12 YCbCr->RGB shader is hand-written just like the composite shaders, hopefully correct! I'll probably put a disasm next to each opcode at some point so it doesn't look like just a bunch of magics. You'll need to use a composite manager if you want to watch videos, we can't render directly to the screen for the same reasons as mentioned in the composite commit. Though, a simple workaround is possible in the Xv case, so that'll probably come at some point. For now, PutImage() will fail with BadMatch. The Y and UV plane upload may not be 100% for what NV50 requires, I want to rework all that at some point anyway. For now, at least there's some form of accelerated video to work with. Enjoy :) commit ec452786d9aa3b6db41f1ab3c89aaa20a758f89e Author: Ben Skeggs Date: Sun Jul 6 20:38:44 2008 +1000 nv50: de-magic things before someone accuses me of being an NVIDIA employee commit ad56c5e27642674d8e435bbbe3d6030097a01ff9 Author: Ben Skeggs Date: Sun Jul 6 01:30:21 2008 +1000 nv50: exa composite commit 33fad279e935a5a134ee5cc1777b8ed7510101d0 Author: Ben Skeggs Date: Sun Jul 6 01:21:39 2008 +1000 nv50: tile offscreen pixmaps Until someone figures out how to make the display engine scan out from tiled surfaces things will be a bit painful. Future commits for exa composite and Xv will require the use of the 3D engine, which will *not* render to linear surfaces.. a composite manager will probably be a must for it to be usable. I really don't want the NOUVEAU_MEM_TILE ioctl in the end, but we'll have to deal with it until we can drop support for Date: Sun Jul 6 00:57:08 2008 +1000 exa: pass window coords to m2mf-based UTS/DFS commit b1f316977e2c79fdf5c83092e953bdfc988299e1 Author: Ben Skeggs Date: Mon Jul 7 01:03:02 2008 +1000 xv: remove some disturbing abuses of buffer related things Also brings to attention a point about the overlay when we use a real memory manager. We'll need to find a way for the overlay to signal it's finished with a buffer, or just resort to pinning pPriv->video_mem.. commit 0ce028b6aaa6c20a1578148add2a96b2d37d9108 Author: Maarten Maathuis Date: Sun Jul 6 13:45:32 2008 +0200 EXA: add a pixmapIsOffscreen hook. commit 86c70df254e63ebc5c405803d6faa9ae33df4995 Author: Maarten Maathuis Date: Sun Jul 6 12:52:16 2008 +0200 NV50_KMS: support gamma changes commit 6de89c8cb219cc82b46e0d0d723fa9d1cefe4ac8 Author: Maarten Maathuis Date: Sun Jul 6 12:29:13 2008 +0200 NV50: Unaccelerated rotation support. - This will need a PixmapIsOffscreen hook before it will be considered for acceleration. commit 788fefa2e78a4abf7bd5c396ab9639bc6ff78451 Author: Maarten Maathuis Date: Sun Jul 6 01:03:44 2008 +0200 NV50_KMS: Unaccelerated rotation support. - This will need a PixmapIsOffscreen hook for it to be ever considered for acceleration. commit 2bc1cb832e0ded6cd1411757bec2a4b03536b7ea Author: Maarten Maathuis Date: Sun Jul 6 01:02:45 2008 +0200 Revert "EXA: Add PixmapIsOffscreen hook for EXA versions that allow it." This reverts commit 5a5aee10f84fbea70eee69ece17ddf6e797497a6. This was never supposed to end up here as it's broken. commit 5a5aee10f84fbea70eee69ece17ddf6e797497a6 Author: Maarten Maathuis Date: Sat Jul 5 23:07:58 2008 +0200 EXA: Add PixmapIsOffscreen hook for EXA versions that allow it. - This should allow more flexibility for rotation shadow buffers. commit b9f2358ac35caedbba0339649ef13b03f209f2a7 Author: Ben Skeggs Date: Sun Jul 6 03:50:03 2008 +1000 xv: remove some defines that are in a header now commit 9f53b2a341a81e1bff87fee8bf5c2754cac17849 Author: Ben Skeggs Date: Sun Jul 6 02:28:26 2008 +1000 xv: remove drawable handling from hw-specific PutImage.. it's common... commit a50bba662ddfe49f1b0a7f8faa3668f5e810811f Author: Ben Skeggs Date: Sun Jul 6 01:57:17 2008 +1000 xv: misc cleanups commit 08e9489b79395b234282b0c03d94601ce57678b8 Author: Maarten Maathuis Date: Sat Jul 5 12:47:48 2008 +0200 NV50_KMS: minor change commit ea7e85b493ff0f684f9ccad92f84348cd835719b Author: Maarten Maathuis Date: Fri Jul 4 23:05:20 2008 +0200 Fix some compile warnings. commit 3298249d8267c27235653f17379f8c9f18e4f344 Author: Ben Skeggs Date: Thu Jul 3 23:54:13 2008 +1000 xv: reformat source so mere mortals can understand it commit a1e8ccdf85c5a0177df38a8a2d8c99d60890ec96 Author: Ben Skeggs Date: Thu Jul 3 17:07:45 2008 +1000 xv: nv_video.c to nouveau_xv.c commit 3e17781625c0ef0ec34ee23a6e999de7f1fe4524 Author: Maarten Maathuis Date: Thu Jul 3 08:30:43 2008 +0200 NV50_KMS: autodetect kernel modesetting when active commit 48c2116fb31872111239f5434215f30c8707091e Author: Maarten Maathuis Date: Wed Jul 2 19:04:58 2008 +0200 NV50_KMS: no int10 please commit 11102af2896b3fdf2364854b2d7939b2802c5b2a Author: Stuart Bennett Date: Fri Jun 27 16:21:19 2008 +0100 Backport console font save/restore from nouveau_ms commit ca2c55dfbf65ac268d43ba62db2399976eaf8f0a Author: Stuart Bennett Date: Thu Jun 26 14:44:45 2008 +0100 randr12: fix mode detection for when screen size changes also some mem leaks commit 9c6eb667aa48f87f6bbbdae10cc9f4d237bdb0c6 Author: Maarten Maathuis Date: Sat Jun 28 15:47:18 2008 +0200 NV04-NV4E: Save and restore FB_START commit 82fe9a21596ab80157df20b269ce8df41315de83 Author: Maarten Maathuis Date: Fri Jun 27 18:49:16 2008 +0200 NV50_KMS: switch to output dpms commit 13ce1f4c1d02e70333a192db2efefb3f6a6765e3 Author: Maarten Maathuis Date: Wed Jun 25 19:45:12 2008 +0200 Properly close drm when exiting + misc cleanup. commit 7cb0eed603ea0bd196a7471e43d60fa5e518c7fb Author: Maarten Maathuis Date: Wed Jun 25 19:07:58 2008 +0200 Fixup and add some license statements. - The changed statements are a response to nv commit 2fdcda8ea62eba1b7885f0a75249884715936247. commit 27c7c88431cf368c3d3a953daf03ce3f9f15bc3a Author: Ben Skeggs Date: Wed Jun 25 04:01:21 2008 +1000 change hack to work with latest drm changes.. commit 99a4b686d2b1f8e6768b8cd1cc3fcb6cea2bf6d3 Author: Maarten Maathuis Date: Tue Jun 24 00:02:25 2008 +0200 NV50: unbreak kms and implement a basic off mode/dpms commit 1a23dbd1878909709fc130e00f6171508dcd41bf Author: Ben Skeggs Date: Tue Jun 24 02:24:42 2008 +1000 nv50: stub out NV50SorSetClockMode for LVDS, it doesn't work there commit 2ccc80bd87a76a9f0498cfec531e8a46cbd2c384 Author: Ben Skeggs Date: Mon Jun 23 02:10:09 2008 +1000 nv50: bring in int10-constole-restore hack from nv driver Lack of console restore has been annoying me personally lately while working on 3D, I lack the time/modsetting skills to fix it properly at the moment, so this'll do. commit 73b4fcc20334ac6cd2076c658a4937f9a98ee06d Author: Ben Skeggs Date: Mon Jun 23 01:47:55 2008 +1000 fix non-kms build/run commit 75d8947d40e19c5fc039caf4289cf94fd1570eda Author: Maarten Maathuis Date: Sun Jun 22 16:34:22 2008 +0200 NV50: Some basic code to get kernel modesetting going. - This will get a cleanup later on. commit df52dc4664df11d57fcf8a6c9ae3f87a149e5392 Author: Stuart Bennett Date: Fri Jun 13 02:11:09 2008 +0100 No need for G80+ LVDS conf bits reports I've got a few already. commit 1813559a20f0786b4decdb645f7ba3f52bc30739 Author: Stuart Bennett Date: Fri Jun 13 01:58:41 2008 +0100 More DCB 1.5 TV variants (#16252 and #16342) commit 29c4a58bb2ba476f9558bda31f1082b4685555fc Author: Stuart Bennett Date: Fri Jun 13 01:40:31 2008 +0100 Be a little more discerning about the BIT bios signature Thank you ABIT commit e8c790721d99e171c57b8dd497f5b3e82126c78b Author: Stuart Bennett Date: Fri Jun 13 01:25:49 2008 +0100 LVDS_INIT should not always use head A While the script doesn't need the head variable set correctly, link_head_and_output still gets called and will therefore break panels expecting to be on head B Fixes Apple NV34 console restore commit fc137f23bed7405221b7f483d9151e296b99caf0 Author: Maarten Maathuis Date: Fri Jun 13 19:12:38 2008 +0200 NV50: lvds is probably still broken, but it's a step in the right direction - Just something i had lingering around. - Note that lvds still doesn't run by default, but you're more than welcome to fix it ;-) commit e034616ab2bbbc486f90c5f64e2bd2de194cf92e Author: Maarten Maathuis Date: Thu May 22 21:08:48 2008 +0200 NV50: A small extra piece of init that may prove useful to someone, somewhere, at some point for something. commit c21594cf452a1263682386d04718531d76a7a9b8 Author: Maarten Maathuis Date: Tue May 20 19:19:53 2008 +0200 NV50: Be consistent with rules-ng. commit 9c1d87fa9e6152bffc5c5cdf16b1931cc6e58399 Author: Maarten Maathuis Date: Sat May 10 21:58:00 2008 +0200 NV50: Some misc things. commit 5f08db9057f2970171186b20a25d346386dcb0f2 Author: Stuart Bennett Date: Thu May 8 21:36:05 2008 +0100 Cosmetic g80 laptop dcb parsing enhancement commit 6a460eae941f3e9a7a8a0297efbbd12dca71a9dd Author: Stuart Bennett Date: Thu May 8 14:17:38 2008 +0100 randr12: some reordering, tidying &c commit 7455bc89116bfc5e88fbd444d83c1415a6020a7f Author: Stuart Bennett Date: Tue Apr 29 19:42:54 2008 +0100 Enable RandR 1.2 code-path by default commit 98a751e773c098382e7675e41ee60ce3d9a67ce0 Author: Stuart Bennett Date: Wed May 7 21:25:03 2008 +0100 Don't call call_lvds_script with a null dcbent pointer Stumbled over by hiyuh :) commit 77b24bc93659471bb20254847da124fdaba13a54 Author: Stuart Bennett Date: Mon May 5 00:56:28 2008 +0100 randr12: tidy up some nv50 remnants in the pre nv50 code commit 8997f83cd2727b76240e9dc31ec7751f6d69cf6b Author: Maarten Maathuis Date: Tue May 6 19:46:52 2008 +0200 NV50: Some misc fixes. - Even though tv-out is mentioned, it by no means implies that is (nearly) functional. commit ae574ec7d930d8708e4dd9d1d47558666019afe5 Author: Maarten Maathuis Date: Mon May 5 11:02:07 2008 +0200 nv50_exa.c had no license statement. commit a7af057ff67cf5d7baf78d852dee5ac69b24e706 Author: Maarten Maathuis Date: Mon May 5 01:26:37 2008 +0200 NV50: Rework the modesetting code into a different model. - There is now interface independent code for most things. - This structure, although not exclusive, is only used for NV50 atm. - The change in approach is that the concept of a connector is added to the model (randr12 sucks in this aspect). - I'm hacking around randr trying to create a single output for each connector, the indices are the bus numbers. - Limitations: Hot swapping vga and dvi of same monitor will not make randr12 think it changed, so for the moment force a change. - Concept is largely inspired by the radeonhd driver. commit 5e1b5708d3a7e4298f80b1a1b8bb3fafae0c69bd Author: Stuart Bennett Date: Fri May 2 20:36:59 2008 +0100 xv: nv10 double-buffered overlay mode was getting set on nv04 commit 85536023ed7050632d121299ab99382e096b9984 Author: Stuart Bennett Date: Wed Apr 30 00:59:47 2008 +0100 Use the actual number of BIT table entries commit 78588c71ffd19e18602b95a55186bc53eebb51c3 Author: Stuart Bennett Date: Mon Apr 28 16:51:01 2008 +0100 randr12, xv: manual overlay clipping for randr12 (see #12825) There could be something to be said for doing this clipping when using the blitter and texture adapters too commit 6f7d00a0cb4ec85b058afb0150a122a124542b5f Author: Stuart Bennett Date: Tue Apr 29 22:53:40 2008 +0100 randr12: remove unneeded separate dcb_entry and "or" members Make off-chip outputs clearer too commit 52e58c7e799697989fcfbf95050ce10a4c3d1f8f Author: Stuart Bennett Date: Tue Apr 29 22:08:01 2008 +0100 Pass DCB entry structs in bios functions, rather than index commit 0c1da6907d8e29ed1c058f9655d436c228ea91b6 Author: Stuart Bennett Date: Tue Apr 29 21:53:22 2008 +0100 Pass output field, rather than dcb indices, for tmds access functions commit 2333bc907a4be3c7e36bdd02f27e22b63a7ee7c3 Author: Stuart Bennett Date: Tue Apr 29 21:58:17 2008 +0100 Add an index field to dcb entries, and carry pointer to dcb entry in output struct commit 21f062cad184f21d33d7f5b43f86937979291b58 Author: Stuart Bennett Date: Tue Apr 29 21:27:07 2008 +0100 randr12, xv: oops, transform_in_use doesn't exist on xserver 1.3 commit ee78dc0abef626e29f7617a4b324e847bab1abcc Author: Stuart Bennett Date: Mon Apr 28 02:12:17 2008 +0100 randr12: C51 I2C commit 9b70dfaf4f0317e622ca64fc0804f74a31158108 Author: Stuart Bennett Date: Tue Apr 29 18:46:24 2008 +0100 randr12: only set cr59 to 1 for tmds (fixes nv34 issue) commit c8cec9b2a0b18cd1159613d4dd218725695aef7c Author: Stuart Bennett Date: Mon Apr 28 17:00:27 2008 +0100 randr12, xv: no overlay when the crtc is transformed commit 3233ebba07e7adab347dc3c6f9d17f02e1a9d32b Author: Stuart Bennett Date: Sat Apr 26 02:08:24 2008 +0100 randr12: fix spread spectrum setting for lvds and turn it off during dpms off Assuming that the bios has set this up correctly for LVDS initially is probably safe commit 467f83aa43bffdf545b513fdb7b11b4fe2205b4f Author: Stuart Bennett Date: Sat Apr 26 19:59:58 2008 +0100 randr12: make nv50 code depend on Architecture commit 8b165722d9f9461353abf84b4136e33b816ac842 Author: Stuart Bennett Date: Sat Apr 26 01:21:50 2008 +0100 CR27 appears to contain the chip revision commit 59ff20738e88d6621f9beaea94810274029a21a3 Author: Maarten Maathuis Date: Sun Apr 27 19:52:41 2008 +0200 NV50: A little bit more information, no functional change. commit bb5a2a105f264502d0985ce70e76d5a3d3b6463c Author: Stuart Bennett Date: Fri Apr 25 22:09:46 2008 +0100 randr12: sel_clk binding fix commit 5a85626cccbba80ef3d6fb2fb90ae609b7abde15 Author: Stuart Bennett Date: Fri Apr 25 02:09:18 2008 +0100 randr12: unified mode validation doublescan/interlace limits applied too commit 38ad350118aa9592bdbcb530fe63b85983fb9f64 Author: Stuart Bennett Date: Fri Apr 25 00:03:41 2008 +0100 randr12: only set up one spread spectrum clock on pre nv40 avoids inadvertently getting wobbly crt pictures commit dfd8b42f91257f2cb0762b3d9b07612a7edac549 Author: Stuart Bennett Date: Thu Apr 24 22:23:50 2008 +0100 randr12: deny doublescan modes to digital outputs commit d150ce5c633c90a82411b4b8e44fa4b91647e6a8 Author: Stuart Bennett Date: Thu Apr 24 01:22:49 2008 +0100 randr12: off-chip dvi dpms and fp_control dpms fixed properly Also solves output exclusion problem: two crtcs claiming the same output through cr33 does not work commit dd152ce186fea953968805cf4a1b309a1fd3987b Author: Stuart Bennett Date: Thu Apr 24 00:24:35 2008 +0100 randr12: set CRTCX_LCD in nv_output This makes my nv28 external dvi work. commit b8d401b0a56d23db7199147e6a5dee1edccff4ff Author: Stuart Bennett Date: Thu Apr 24 00:01:50 2008 +0100 These made up defines are more misleading than helpful commit fa8c8437798cfc5494a639d649f0f91f0e7275ce Author: Stuart Bennett Date: Tue Apr 22 22:42:12 2008 +0100 randr12: remove FPP setting and other FSEL changes commit a083bb1a55d0983991c45d113eb7e9814369fff4 Author: Stuart Bennett Date: Tue Apr 22 23:25:40 2008 +0100 randr12: oops (fix to 9edf6d60) commit 16f420e339f0dd6d55d7b87763d97c2eb8e77e66 Author: Stuart Bennett Date: Tue Apr 22 21:34:52 2008 +0100 randr12: RAMDAC_OUTPUT is only for analogue outputs commit 52174047f3f817fa9ce3674440a13687b73ccea8 Author: Stuart Bennett Date: Tue Apr 22 21:29:48 2008 +0100 Revert "randr12: correct SEL_CLK register beheaviour for outputs with or=2" This reverts commit 77af887089be6fef7b793e3c3afdb98ac90a2e8c. External DVI (which is source of or == 2) do not use SEL_CLK commit 61e416153b24516951962aca407ceaa9b7885e22 Author: Maarten Maathuis Date: Sun Apr 13 02:40:55 2008 +0200 NV50: Another few regs. commit 5328370d2f7e10c61485572f53031deace69c515 Author: Maarten Maathuis Date: Sun Apr 13 02:16:27 2008 +0200 NV50: Some more regs. commit e5511a564564350bd8755b7f18a1448ea7c796de Author: Maarten Maathuis Date: Sun Apr 13 01:21:20 2008 +0200 NV50: I now know what UNK8A8 is, so update the value register as well. commit 2c9027679db1657e6fe02cf6674024ca97f4cd6e Author: Maarten Maathuis Date: Sat Apr 12 02:11:32 2008 +0200 NV50: Re-implement output status caching. commit d7681e8ad70c07ec963ac9425b868a2686d79769 Author: Stuart Bennett Date: Fri Apr 11 17:28:16 2008 +0100 randr12: separate pll and arbitration stages of calc_state_ext, and call from mode_set directly commit 0bf08d7699d7c3a1748060febded1c8b3950c9bd Author: Stuart Bennett Date: Fri Apr 11 16:20:21 2008 +0100 state->config is not used commit 08b8bd8a97b580a57d7bb75b3da7827d65c046f5 Author: Stuart Bennett Date: Fri Apr 4 18:59:29 2008 +0100 randr12: unite pre nv40 and nv40 pll restore commit 5b443084ec7b03230d0ec7490c5a3378a9f7187e Author: Stuart Bennett Date: Tue Apr 8 02:25:34 2008 +0100 Use the correct reference clock when reading (V)PLLs commit 0f533b8cc12f49f5d5ac017838f7330a4f8b47ea Author: Stuart Bennett Date: Tue Apr 8 02:25:22 2008 +0100 Various PLL code improvements commit 4e1caa6edfe9499e6c501e70a8fba14bae510462 Author: Maarten Maathuis Date: Wed Apr 9 18:55:06 2008 +0200 NV50: Some misc stuff. commit 099128175ad31956fd1c6622166b65f4ab1fe370 Author: Dave Airlie Date: Tue Apr 8 16:48:37 2008 +1000 nv50: fix distcheck since header removal commit 4af376fa6031bd41dc247e950319f3a151c49f22 Author: Stuart Bennett Date: Tue Apr 8 00:28:03 2008 +0100 MCP67 does not do the second stage gain tests commit dbca8b65d7b12a1dd2eb7aee3ac37ece0bd4ce65 Author: Stuart Bennett Date: Tue Apr 8 01:31:14 2008 +0100 PLL stuff commit 39b2e0b842f4a4e1c3de96c94e380aa9a0327c12 Author: Stuart Bennett Date: Sat Apr 5 05:20:54 2008 +0100 If no PLL coefficients compatible with the constraints exist, don't brick the hardware commit 0b54450e1db7c6411528bc4a3bac4d1a42510aca Author: Maarten Maathuis Date: Mon Apr 7 22:12:44 2008 +0200 NV50: Scaling works fine on VGA connectors, so support that as well. - The default is SCALE_PANEL instead of SCALE_ASPECT for obvious reasons. commit 891e72f4316c4e7e1b9d06f39e075b1a8a8a8c53 Author: Maarten Maathuis Date: Mon Apr 7 21:18:54 2008 +0200 NV50: Deobfuscate some more commands. commit 13c701f9d23a598c281ced726070a219e1b78607 Author: Maarten Maathuis Date: Mon Apr 7 19:29:55 2008 +0200 Make some CARD32 die. commit a3a0710604017f3ceb9c552552d71c3f5dcfab0f Author: Maarten Maathuis Date: Mon Apr 7 19:09:35 2008 +0200 NV50: Give a few registers an UNK label. commit be72a7c9104c1c1d6db960b1d80cc4d820d37440 Author: Maarten Maathuis Date: Mon Apr 7 18:35:18 2008 +0200 randr12: NV04/05/06 can't accelerate rotation, so it's better to disable it. commit 9cf76c91a661587d9c54845e6a06fb25cf9b55a2 Author: Maarten Maathuis Date: Sun Apr 6 23:36:15 2008 +0200 NV50: Some more stuff. commit c725e33078a2fed374fbb52175beba92b0b9fbe8 Author: Maarten Maathuis Date: Sun Apr 6 19:49:11 2008 +0200 NV50: minor rename to avoid confusion commit fcda539ce60ca02f2f4589157f8b8c522430571f Author: Maarten Maathuis Date: Sun Apr 6 15:16:40 2008 +0200 NV50: Deobfuscate another register (partially). commit fa93541b257a8055ea8b55c883475d2db717d971 Author: Maarten Maathuis Date: Sun Apr 6 15:05:33 2008 +0200 NV50: Fix another deadlock in the clock setting code. - Swapping my monitor from dvi to vga causes no mode (=clock) change, but still the supervisor expects a clock change. - This meant it got stuck in stage 2, with no clock that is supposed to change. commit c001fea3b0ae3921c063e7f44c298cafdfef51f6 Author: Maarten Maathuis Date: Sat Apr 5 23:33:22 2008 +0200 Forgot a few things. commit 17a8e1743478e5548a90caf8f84a387366020416 Author: Maarten Maathuis Date: Sat Apr 5 23:23:22 2008 +0200 NV50: rename randr-1.2 functions to be nv50_something_function commit a71ef4da76e33e84303a8b9181dbcfd655e78aa9 Author: Maarten Maathuis Date: Sat Apr 5 23:00:39 2008 +0200 NV50: Some more cleanup. commit aeab563cc89b2fc3cd6cb306aab880b6e28f2b64 Author: Maarten Maathuis Date: Sat Apr 5 22:52:58 2008 +0200 NV50: Some reordering. commit b899ab75781a3076898791381604be575c6d8a70 Author: Maarten Maathuis Date: Sat Apr 5 21:17:23 2008 +0200 Some minor corrections. commit b0ef158a7e281329825c7b6c10cab6097e627b54 Author: Maarten Maathuis Date: Sat Apr 5 21:06:03 2008 +0200 NV50: Document dvi hotplug detection. - This can only be used once modesetting is moved to the kernel. commit be54b9299b4a261fc707b74407cf4d847fcbb243 Author: Maarten Maathuis Date: Sat Apr 5 16:16:26 2008 +0200 NV50: Remove some questionable guesses. commit c06f202cfbee8568fb8358ea1caffec41c71e648 Author: Maarten Maathuis Date: Sat Apr 5 01:54:06 2008 +0200 nvbios: Drop a N2/M2 requirement on NV5x hardware, as it prevents getting a mode on 1024x768'ish and lower. - Limitations are very strict in nv5x bios tables and N2/M2 can drop below 2 (the blob does this too). commit 5ce26b9aec214dd99e5399fe0f864d8faad15d2c Author: Maarten Maathuis Date: Fri Apr 4 22:49:06 2008 +0200 nvbios: Add some valid register ranges for nv50. commit ce86484979d873325ec4c74ed84107be9ad9ba62 Author: Maarten Maathuis Date: Fri Apr 4 22:38:30 2008 +0200 NV50: Switch to common pll calculator function. commit 821ad3d643f9564d043c7f3909e0b8be28e94965 Author: Stuart Bennett Date: Wed Apr 2 16:58:45 2008 +0100 randr12: delete stuff commented out a week ago that no one's complained about commit e90e534728e9cbe65ee6c70faf596db449495665 Author: Maarten Maathuis Date: Fri Apr 4 17:13:25 2008 +0200 NV50: Minor comment updates. commit 08550984aba2bfa3eaeeddbbf6af25206819abe8 Author: Maarten Maathuis Date: Fri Apr 4 17:04:49 2008 +0200 NV50: Some more stuff. commit ef4339aacf72bf5422b22ab18b73416f2738b284 Author: Maarten Maathuis Date: Fri Apr 4 16:49:00 2008 +0200 NV50: Some more registers. commit 3183df1aa61f3c441a93e5a8b50db6628d804465 Author: Maarten Maathuis Date: Thu Apr 3 22:42:31 2008 +0200 NV50: Update my knowledge of interrupts. commit 04fd80d29913ed02524de177fa7c87b616147d15 Author: Stuart Bennett Date: Thu Apr 3 03:00:13 2008 +0100 Small declaration reordering commit 92934e4b7571a578386327ebdb59eb6d10992763 Author: Stuart Bennett Date: Thu Apr 3 01:03:38 2008 +0100 Move some variable declarations in the bios parser commit 927384005a01a606835452c9d1e26d7df98f81cb Author: Stuart Bennett Date: Thu Apr 3 00:11:31 2008 +0100 randr12: some fixes to dithering property commit 52c880c97acf29d5ed8dfeec5810fcc7d8bb7b35 Author: Stuart Bennett Date: Wed Apr 2 19:45:43 2008 +0100 randr12: better analogue clock limits (taken from pNv->{Min,Max}VClockFreqKHz) commit 29714e7b9ae952aecd036086c053ec030aad94c9 Author: Maarten Maathuis Date: Wed Apr 2 21:31:07 2008 +0200 NV50: Some minor changes to the display supervisor (mostly comment). commit 9ac32d3e2bf400f6a6220118b8008804cd11e435 Author: Stuart Bennett Date: Wed Apr 2 19:03:07 2008 +0100 Some bios parser PLL writing improvements commit 4a192ea6cf9df39d95368f46482c777524471118 Author: Maarten Maathuis Date: Wed Apr 2 18:21:01 2008 +0200 NV50: Add my name to some files. commit db6220b8611829a77d354de1e45f6fe29e118b2a Author: Stuart Bennett Date: Wed Apr 2 16:16:13 2008 +0100 randr12: since the crtc function records are now mutable, unify pre nv50 templates commit b9762ba272a3ee817278e3967794c55a4fe73961 Author: Stuart Bennett Date: Wed Apr 2 16:02:43 2008 +0100 randr12: don't provide rotation functions in NoAccel case commit a24ab3d1559d023270915f373527efb6eefacd8d Author: Stuart Bennett Date: Wed Apr 2 16:05:30 2008 +0100 Revert "randr12: Fail rotation in NoAccel case." This reverts commit b33317ed98957db42691f5f9c41c119c10c91be6. commit 33d4c2f2fe6c6c400ce945f4c7ecc09ee08123a5 Author: Stuart Bennett Date: Wed Apr 2 15:46:50 2008 +0100 Add quirks for Apple *book backlight commit 8a3f051dc9f007c3d801942d32899b8d6c09fd34 Author: Maarten Maathuis Date: Wed Apr 2 14:54:16 2008 +0200 NV50: Some more things. commit b33317ed98957db42691f5f9c41c119c10c91be6 Author: Maarten Maathuis Date: Wed Apr 2 09:28:45 2008 +0200 randr12: Fail rotation in NoAccel case. commit d017dfa93e7eff988339b8030af1a563347156bc Author: Stuart Bennett Date: Tue Apr 1 21:23:35 2008 +0100 NV30/35 setPLL bios support commit 32f550a534dc3e19f7cd3640549336846ad6052d Author: Maarten Maathuis Date: Tue Apr 1 22:33:15 2008 +0200 NV50: Fix a minor cursor issue. commit 27e819dbcd5ce36be71c7dd3be332f777a80caf4 Author: Maarten Maathuis Date: Tue Apr 1 22:20:04 2008 +0200 NV50: Add some debugging statements to important functions. commit 99857933e8edc3be2e4af0c8c7e3fe3b1e221de9 Author: Maarten Maathuis Date: Tue Apr 1 20:33:16 2008 +0200 NV50: Some i2c stuff. commit 9199307195c51b52979c483b3c0adf49f2301281 Author: Stuart Bennett Date: Tue Apr 1 05:16:29 2008 +0100 Use new PLL reading code Of course, having spent some time perfecting the reading of NVPLL and MPLL, it turns out to be totally unnecessary on >= nv30 :-/ commit a3f82d92ae6cbd779d341fe11687d12c5aac1762 Author: Stuart Bennett Date: Tue Apr 1 05:03:14 2008 +0100 Unified PLL reading commit 350fb202e113bc03a71fd7e451e2c26ec990a61a Author: Stuart Bennett Date: Tue Apr 1 05:29:05 2008 +0100 randr12: get_clock_from_crtc only necessary for TMDS commit f741bf7b724dd0698f71b28c461337526f20fe31 Author: Stuart Bennett Date: Tue Apr 1 00:23:06 2008 +0100 Better PLL limits for some nv3x commit 94ca1951feda10b9d6ad828816a459af750de286 Author: Stuart Bennett Date: Tue Apr 1 01:00:16 2008 +0100 randr12: deal with nv30 and nv35 PLLs individually commit 626d8d62eef205dc542e3a7cd82beb329a23cdaf Author: Maarten Maathuis Date: Mon Mar 31 19:27:28 2008 +0200 Init global gamma on all crtc's, even if they are currently not in use. commit f54cea5662fdfc17762711a88892a27aed62f74c Author: Maarten Maathuis Date: Mon Mar 31 19:17:50 2008 +0200 NV50: Some clock related cleanups + minor changes. commit 7393d8fc76f7bbb8aa156587d81ee1a01c17470e Author: Maarten Maathuis Date: Mon Mar 31 14:10:42 2008 +0200 NV50: Some misc cleanup. commit 9edf6d60b2b6698e88edc7798285bd843bfce0f4 Author: Stuart Bennett Date: Sun Mar 30 21:48:13 2008 +0100 randr12: different handling for sel_clk for pre and post nv40 This won't set the spread spectrum bits for pre nv40 from null initial state, but that's not a major concern right now commit 84ab4be08947ef35900da6b84702bc9acb4121bb Author: Stuart Bennett Date: Sat Mar 29 21:26:37 2008 +0000 Minor tidyup to output merging commit 1fe1d8161059afcc6f6ba152b9331e1576ef1f93 Author: Maarten Maathuis Date: Sun Mar 30 22:42:47 2008 +0200 Kill a useless register offset. commit 75680d723144270c12a6d500d8e2c2c421e4ebc1 Author: Maarten Maathuis Date: Sun Mar 30 15:18:09 2008 +0200 NV50: remove unneeded function commit 263ec94ec0f553f5ab6a22d3fd335bebff2c029b Author: Maarten Maathuis Date: Sun Mar 30 12:21:29 2008 +0200 NV50: Some minor cursor cleanups. commit 86722ad60fb939743f5ab73413f721cd26bf2c3c Author: Maarten Maathuis Date: Sun Mar 30 03:37:16 2008 +0200 NV50: Deobfuscate some of the modesetting, a few things were hidden away in innocent looking variables. - Some mode variables were overridden and didn't represent their name anymore. commit 633a8f4f0a8ff3bcb5e6afa9bbcf3f6ed44ef905 Author: Stuart Bennett Date: Sat Mar 29 19:01:30 2008 +0000 randr12: make sel_clk behaviour match comment (which seems to be correct) If nv40 breaks, this whole thing will need another look commit 7a8bbb7f0890ef99de539f0f7acccd469f33dce9 Author: Stuart Bennett Date: Sat Mar 29 00:03:27 2008 +0000 Fix Apple backlight code, broken in pci-rework conversion commit 4d0974c48eff8cd51af66f8fded4fad50044958a Author: Stuart Bennett Date: Fri Mar 28 01:10:31 2008 +0000 Consistency fix + constify commit acd007843f00cb99cf26c4a5f885e365c3a8065b Author: Maarten Maathuis Date: Sat Mar 29 21:47:04 2008 +0100 It seems wise to NVSync for all hardware on NVLeaveVT. commit ea27bc38938277599228ba403a082963bf89d702 Author: Maarten Maathuis Date: Fri Mar 28 19:56:36 2008 +0100 NV50: Better safe than sorry. commit d4ca83e2d1f1432dc7de290087f03a6ff3553d98 Author: Stuart Bennett Date: Thu Mar 27 23:55:30 2008 +0000 randr12: shrink pitch alignment stuff and move commit 43ff98cd89e3dc43c281e41f3a36a76bcd3e5a65 Author: Stuart Bennett Date: Wed Mar 26 18:35:14 2008 +0000 randr12: tidy-up 3/3 - TMDS04 twiddling Code commented for easy reinstatement should problems arise commit 9491ee3f3fbb8b50244444424ef6148217ace944 Author: Stuart Bennett Date: Wed Mar 26 18:34:28 2008 +0000 randr12: tidy-up 2/3 - DPMS CR57/58 twiddling Code commented for easy reinstatement should problems arise commit 2e001c140eacc5d5f02435b9c81b13e5d40db71f Author: Stuart Bennett Date: Wed Mar 26 18:33:59 2008 +0000 randr12: tidy-up 1/3 - FP_DEBUG_0 twiddling Code commented for easy reinstatement should problems arise commit ec9fa3c958d6763ff96d8b884ed36c80d911475a Author: Stuart Bennett Date: Thu Mar 27 19:24:04 2008 +0000 randr12: allow doublescan and low clock modes 320x240 ftw. Also improved mode rejection error messages commit 048f31014092c1dccb1944d2b1a84c7a301be48e Author: Stuart Bennett Date: Wed Mar 26 18:58:47 2008 +0000 Minor fix to lvds script caching commit commit 3feaa48b53f51c094718708e52f08b546b0b0e27 Author: Maarten Maathuis Date: Wed Mar 26 20:01:19 2008 +0100 NV50: Implement per CRTC gamma correction. - This remains untested, so if you're in the mood to test ;-) commit 77af887089be6fef7b793e3c3afdb98ac90a2e8c Author: Maarten Maathuis Date: Wed Mar 26 10:02:46 2008 +0100 randr12: correct SEL_CLK register beheaviour for outputs with or=2 commit cb5d4cdb3dc83f9dbd53eae9097c0df458313fff Author: Stuart Bennett Date: Wed Mar 26 02:27:33 2008 +0000 Cache running of LVDS scripts, rather than blocking LVDS_RESET Fixes a regression on some nv40 where LVDS_RESET is needed to fix some bad behaviour in the off/on scripts commit 6924d1ac900524d0978a02f8f8509faab1809b62 Author: Maarten Maathuis Date: Thu Mar 20 20:20:16 2008 +0100 NV50: Although not essential, it's nicer to save a register as uint32_t. commit 1042c2bed3aee8561411fba85438c3fe5ee2cdea Author: Maarten Maathuis Date: Thu Mar 20 20:10:13 2008 +0100 NV50: Some more stuff. commit b58e5e99aa06e1851ce255211d524a88f1fb297a Author: Maarten Maathuis Date: Thu Mar 20 19:59:46 2008 +0100 NV50: Educated guesses at some of the control registers. commit ef25cb9224d37c7937e67824a262937a747c91cc Author: Maarten Maathuis Date: Thu Mar 20 18:49:58 2008 +0100 NV50: Forgot some more stuff. commit 64b3d1ab1a4226fa688722dabc40fe05a0e30d52 Author: Maarten Maathuis Date: Thu Mar 20 18:35:20 2008 +0100 NV50: Forgot something. commit 81929c1810fe4825ff9bd5d4bc84996c35c8ceb7 Author: Maarten Maathuis Date: Thu Mar 20 17:43:30 2008 +0100 NV50: More dehexing. commit ee1ac0646c7c2f018abdfd4f58d49d51f5340784 Author: Maarten Maathuis Date: Thu Mar 20 17:19:37 2008 +0100 NV50: Some more dehexing. commit ca1b8bbebd4b786f0fe8531ba7145aa932e5abcc Author: Maarten Maathuis Date: Thu Mar 20 14:18:18 2008 +0100 NV50: Some more register guesses. commit aa99fb07cd438f436fdfa4e5dff6d87d27950358 Author: Stuart Bennett Date: Wed Mar 19 18:16:03 2008 +0000 randr12: do DPMS off better Half the time that DPMS off is done, there is no crtc bound, which meant outputs never got turned off. It seems xf86DisableUnusedFunctions is always run *before* setting up the new outputs (when using xrandr): if this is not the case, this commit may not be quite right. commit 063b393ac12bec286e0b28c2b18a98cedd3422c1 Author: Stuart Bennett Date: Wed Mar 19 02:56:31 2008 +0000 Optimizations on LVDS script code commit c5203647ddf262978e7d6a4912661a9cc448da66 Author: Stuart Bennett Date: Wed Mar 19 23:12:59 2008 +0000 randr12: unbreak LVDS and primary I2C for < NV50 NV50 check disallowed pre-NV50 cards using I2C on first head, and made garbage get written to CR0 on LVDS This should work for both pre-nv50 and nv50. commit ebc75dc9e8394ee1d13a853341ba9995e9d9d04f Author: Maarten Maathuis Date: Wed Mar 19 21:31:10 2008 +0100 NV50: fix i2c for real commit c6e3a6f57a7ae67c187aff0ad87153e988961784 Author: Maarten Maathuis Date: Wed Mar 19 08:57:33 2008 +0100 NV50: Fix a typo and hopefully fix i2c. commit a989b4caac077f65034fb328a80ab4ae758c1162 Author: Maarten Maathuis Date: Mon Mar 17 23:40:16 2008 +0100 NV50: Fix major thinko in "NV50: kill output_resource and use or directly." commit cd5aaeaa319e2b681c7c6ccc135d76b7c027bd0d Author: Maarten Maathuis Date: Mon Mar 17 22:05:12 2008 +0100 NV50: Handle some things that weren't fixed during rebase. commit f73c2361a7bba5aa9abb4fe65697bfcb2a407e99 Author: Maarten Maathuis Date: Sun Mar 16 16:09:11 2008 +0100 NV50: kill output_resource and use or directly. commit de0fc0ece55997f8b2ad86070cb940e2afd2d7e8 Author: Maarten Maathuis Date: Sun Mar 16 12:46:40 2008 +0100 NV50: Mostly resorting regs. - Also a dithering register define and rename cursor related reg. commit 849bab73da8da8ad8b2987f80fcf90bf86299171 Author: Maarten Maathuis Date: Sun Mar 16 12:33:19 2008 +0100 NV50: Some more educated guesses at register names. commit 0244dc1dd0698be161d7fd629efa10dcb3df106a Author: Maarten Maathuis Date: Sun Mar 16 01:09:48 2008 +0100 NV50: Another subtle difference. commit 5e2d220edf65357f4d7bd8ce97bd829086e0781c Author: Maarten Maathuis Date: Sun Mar 16 01:05:03 2008 +0100 NV50: Fix a potentional bug. commit cef158cee51ecc53c2f5a9910a88c9e802d49980 Author: Maarten Maathuis Date: Sun Mar 16 00:40:59 2008 +0100 NV50: Share output properties. commit 31e7a740a1cddc5edfdede05187e6bd9a6f1b993 Author: Maarten Maathuis Date: Sun Mar 16 00:27:41 2008 +0100 randr12: Implement dithering output property. commit 05f0464669eef0a3caff2466eeffd9843c18a7b7 Author: Maarten Maathuis Date: Sat Mar 15 19:02:55 2008 +0100 NV50: Merge output creation and i2c init. commit 285a13ab41e869544681260b1f2415143456bdc9 Author: Maarten Maathuis Date: Sat Mar 15 18:20:34 2008 +0100 NV50: Merge crtc creation. commit c5dac45a199b7f4fa3692727aa6af1cd9a6b8d35 Author: Stuart Bennett Date: Mon Mar 17 23:20:39 2008 +0000 randr12 offbyone debug commitnoise semantic version control messages ftw? commit 4cd5e430d164cb5a834e4946d5e5dc79061cadfb Author: Stuart Bennett Date: Mon Mar 17 17:21:28 2008 +0000 randr12: remove NVGetOutputFromCRTC This function had a major problem for the long term: assuming a maximum of 1 output per CRTC, which is simply not valid commit 27c21c0574bac47bbb311094c32f17e7ac321ba5 Author: Stuart Bennett Date: Mon Mar 17 17:09:23 2008 +0000 randr12: split out fp-only registers commit c09d190ffe2bab1e1df52d8b83009c10fafbc2bc Author: Stuart Bennett Date: Mon Mar 17 03:03:14 2008 +0000 randr12: better handling for output types in nv_crtc commit 78ded96a4e8f88dba038e5ef0c452b7c7bf9a2dc Author: Stuart Bennett Date: Mon Mar 17 17:16:27 2008 +0000 randr12: remove duplicated regs commit f7013760f9e9ec91bc3bd3739b6cce9018e34bcb Author: Stuart Bennett Date: Mon Mar 17 00:59:31 2008 +0000 randr12: fix noscale scaling mode Also, a bit of consistency missed in the earlier commit commit a9c01eff4bcf3af52faf73269f63a5fd46b3b3b3 Author: Stuart Bennett Date: Sun Mar 16 14:27:14 2008 +0000 Misc startup info mostly duplicated the state storage stuff commit a02f82e9ce79ae24a4f2eba71a83e51f3b6d9e52 Author: Stuart Bennett Date: Sun Mar 16 00:51:09 2008 +0000 randr12: make output parsing consistent commit f04c44eff67adb9099eca7b298b10639dc0b1759 Author: Stuart Bennett Date: Sat Mar 15 01:36:13 2008 +0000 G80 I2C parsing Changed to on-demand parsing commit 718e42478c02c7497cd3917b8ea78ed692f30f3f Author: Stuart Bennett Date: Fri Mar 14 21:22:48 2008 +0000 randr12: make nv_crtc_calc_state_ext definition smaller commit 772bf76b3697d1d1725017787378802c2940093e Author: Stuart Bennett Date: Tue Mar 11 03:39:43 2008 +0000 randr12: merge common PLL paths commit 1326e0176582fd21e16359764e2792cd125548e6 Author: Stuart Bennett Date: Tue Mar 11 02:19:10 2008 +0000 randr12: make plls belong to their CRTCs commit 97b074889feef7b4a3b6401825a46d53321d1c7e Author: Stuart Bennett Date: Wed Jan 30 14:38:59 2008 +0000 Use PMC defines commit ee11c6abe875a758bdfa4469ae6326ac25732178 Author: Maarten Maathuis Date: Fri Mar 14 11:11:07 2008 +0100 randr12: Let an unneeded quirk die. commit 8851d9f3ef4a8cb9b35e1cb8ed80b3ff213eeee0 Author: Stuart Bennett Date: Wed Mar 12 16:29:17 2008 +0000 randr12: sel_clk values are really an FP thing, so move calculation to nv_output commit 63527f978d923317a97dec060be6df16c7baf17d Author: Stuart Bennett Date: Thu Mar 13 20:26:07 2008 +0000 randr12: tidy up output restore commit 769436b3840e10da00b51b94d72d0b2ef510fecc Author: Stuart Bennett Date: Thu Mar 13 19:52:52 2008 +0000 randr12: common tmds access functions commit c455ca0d3df3bb2322154ded40fbe6b86c6a8acd Author: Stuart Bennett Date: Wed Mar 12 03:23:21 2008 +0000 Improve bios' handling of ramdac_580 commit bd5ad7c4e1f824d57517da540658c97b5283d5d0 Author: Maarten Maathuis Date: Thu Mar 13 20:49:53 2008 +0100 NV50: Accept G9X cards. commit 0318b31e19da12ba0d4dc456ad08d6b059d08d1e Author: Maarten Maathuis Date: Thu Mar 13 16:57:24 2008 +0100 NV50: Add back output_resource. commit a4397563c6845668ad1f4b2352b90c872957f8bc Author: Maarten Maathuis Date: Wed Mar 12 23:16:53 2008 +0100 NV50: Unbreak "NV50: Kill the connection status caching (which was broken btw)." commit 63ffb8dce809267706eb621f03da3c3501935bbd Author: Maarten Maathuis Date: Tue Mar 11 23:15:11 2008 +0100 NV50: Merge the two output private structs. commit b4c3bee2f8cc824cd52aca61121634c55d5f5a60 Author: Maarten Maathuis Date: Tue Mar 11 23:09:19 2008 +0100 NV50: rename scale to scaling_mode commit b0d2f93d3adbb798f5d9bfa2d188b666e4c5fc16 Author: Maarten Maathuis Date: Tue Mar 11 23:06:16 2008 +0100 NV50: Rename nativeMode commit e5b8efad35df21d35a3893a6c8b10e1f02272544 Author: Maarten Maathuis Date: Tue Mar 11 23:00:19 2008 +0100 NV50: Rename or to output_resource. commit 19ca2db7683b8cdcfb74f6cc14c9cb6c3876d2c7 Author: Maarten Maathuis Date: Tue Mar 11 22:43:33 2008 +0100 NV50: kill set_pclk commit eef3a665c01c530238c3dc02324573d6bcf6c3f2 Author: Maarten Maathuis Date: Tue Mar 11 22:24:46 2008 +0100 NV50: Remove the remnants of cached_status. commit dc0bb590a40f3d8b18d2136ed906e8dac10ec598 Author: Maarten Maathuis Date: Tue Mar 11 22:22:12 2008 +0100 NV50: kill partner entry commit f66463713a71b0e3090f33c68186884594255cc0 Author: Maarten Maathuis Date: Tue Mar 11 22:12:57 2008 +0100 NV50: fix warning commit cb50cd3276825569a6da1f25a2d89bfd417b5034 Author: Maarten Maathuis Date: Tue Mar 11 21:33:29 2008 +0100 NV50: Kill the connection status caching (which was broken btw). commit 6e2af71c73263c3cf2d61f66c02363decd354da5 Author: Maarten Maathuis Date: Tue Mar 11 21:15:21 2008 +0100 NV50: change i2c to pDDCBus commit d07b3063953c0b17585249dfbc2938a433bea88f Author: Maarten Maathuis Date: Tue Mar 11 19:50:25 2008 +0100 NV50: Switch to other scaling defines. commit 4ddb92b8e936985b5b1429a0a21890879635bcde Author: Maarten Maathuis Date: Tue Mar 11 19:29:11 2008 +0100 NV50: Switch to the same output type as the rest of the code. commit e605fbfc659d3fe1bd71d478c21429b563fbfd80 Author: Stuart Bennett Date: Thu Mar 13 15:21:49 2008 +0000 randr12: nothing uses output_resource. kill it commit 53925cfc3540b89dedd317bbae4318cf53daada1 Author: Stuart Bennett Date: Thu Mar 13 15:31:00 2008 +0000 randr12: fix the setting of NV_RAMDAC_OUTPUT And turn on/off the DAC for DPMS commit 493c79405be23a543cd21c2583c466342399509d Author: Stuart Bennett Date: Thu Mar 13 03:53:18 2008 +0000 Simplify TMDS writes a little commit 4cf1bd9ef016a2c9fe7784fdc83f559bec464f9e Author: Stuart Bennett Date: Tue Mar 11 17:20:28 2008 +0000 randr12: don't allow scripts to break overlay commit 95f2b8439f433afb5d637a3580f933e2cd44488a Author: Danny Tholen Date: Tue Mar 11 17:14:54 2008 +0000 Fix video overlays to treat double-scan correctly when using randr12 commit 460cb260c30467799fa31aef11946d1fe2dd6523 Author: Stuart Bennett Date: Mon Mar 10 23:47:43 2008 +0000 Correct RAMIN offset and some bits missed in previous nv_setup tidying commit e8a5d4638e74b5957fc12b05865f49dd995d0435 Author: Stuart Bennett Date: Mon Mar 10 21:46:38 2008 +0000 randr12: changing tmds2 for dvi-d seemingly unnecessary commit 30ddbe9c2b82bc19d1480a1b38df6b5ca0dd057b Author: Stuart Bennett Date: Mon Mar 10 23:49:08 2008 +0000 randr12: unbreak sel_clk changes commit 03f38725304842b053b7976db5df97e61159d07d Author: Stuart Bennett Date: Mon Mar 10 19:50:55 2008 +0000 randr12: allow spread spectrum bits to be updated on pre nv40 commit 323b7ef9c1c26996b84fd1f67297b2958f68fcc9 Author: Stuart Bennett Date: Mon Mar 10 01:18:56 2008 +0000 randr12: quirk for 17" powerbook, and better sel_clk handling We set sel_clk *before* running the output script, so make sure the set pll->head binding is not altered commit 58f9079cda17bb6a76ba16c14ddc56bd4ac640fc Author: Stuart Bennett Date: Sun Mar 9 16:08:46 2008 +0000 nv_type.h mandates XF86DRI being declared commit 0dbffc868a7b60aa99ca550d33b8488ac9de8d00 Author: Stuart Bennett Date: Sun Mar 9 03:59:57 2008 +0000 Fix up a bunch of prototypes, remove unused wrappers commit e5ded02925e7d39c67e69d0389778ac795bfe743 Author: Stuart Bennett Date: Sun Mar 9 01:31:47 2008 +0000 Move setting of alphaCursor to nv_setup, and restyle useful portions of nv_setup commit 0b8629a9a29203553397eb764b9b9c87d38d4b9f Author: Stuart Bennett Date: Sat Mar 8 22:05:52 2008 +0000 Kill NVFBLayout Using pScrn->currentMode is wrong for randr12, but it's no worse than what it was commit 29cfb08c1b75e050aa58fa6019f7d9be063f8555 Author: Stuart Bennett Date: Sat Mar 8 21:02:35 2008 +0000 Warning fixes And remove a few unused members from nv_type.h commit b87059fcccfac693df0b2dc5ba85cb358ad1eba6 Author: Stuart Bennett Date: Sat Mar 8 20:53:20 2008 +0000 Drop the NV*VGA defines and adjust users commit a636b02eba03454b8b757f69e6a62aae2c3fd571 Author: Maarten Maathuis Date: Sat Mar 8 21:18:34 2008 +0100 randr12: Disable a quirk that has adverse sideeffects on NV36M. - NV31M owners, please test if you have: - LVDS on CRTC 1 and or = 3. - VGA on CRTC 0 and or = 1. - Testing requires enabling both lvds and vga. commit f4228861abec9df310efb863833c82022136e464 Author: Maarten Maathuis Date: Sat Mar 8 18:23:58 2008 +0100 Merge CRTC private structs. commit b3d0d0c01664e874f3b69b36383ad1a7845d7d91 Author: Maarten Maathuis Date: Sat Mar 8 18:10:53 2008 +0100 Death to some unnecesary stuff. commit 03a7a1d7065b55f16385fd16e7a2eff938d59cea Author: Stuart Bennett Date: Sat Mar 8 01:56:36 2008 +0000 Various minor tidyups commit 624148ca7067f748c142db3cd4d31314aa31ee0f Author: Maarten Maathuis Date: Fri Mar 7 20:59:30 2008 +0100 NV50: Convert back to absolute registers. - Someone with a known to work setup should test this, as i have no NV5x card. commit 10bdffb65e2d8f59399556fe3224cbf8e8a171a9 Author: Stuart Bennett Date: Fri Mar 7 19:08:50 2008 +0000 PLL limit table v0x11 commit 2e1af24826e7db0e3f53440a2c5912789ade6c97 Author: Stuart Bennett Date: Fri Mar 7 18:43:16 2008 +0000 Make bios init code work on 6600 by changing pll setting stuff a bit I suspect we're going to need some card specific checks in here. ugh. commit 5fd2dd7b4c6e15d83cb7e7daafb6ee801fa823e8 Author: Maarten Maathuis Date: Fri Mar 7 07:52:09 2008 +0100 It seems libpciaccess release became version 0.10 not version 0.10.0 commit 3e4db03a39770e5fa94ca094deb690cba8175196 Author: Maarten Maathuis Date: Fri Mar 7 07:36:16 2008 +0100 randr12: Forgot to clean up when switching output resources. commit 8a5c6e8b4d00871b17b7d564d765ea677596f79a Author: Jaime Velasco Juan Date: Sat Feb 16 15:01:23 2008 +0000 Do not access beyond source pixmap in NVAccelUploadIFC commit 7afb739418fbac5058d13d05188a69b142529fb4 Author: Stephane Marchesin Date: Thu Mar 6 23:40:08 2008 +0100 Oops. commit 0da8c84cceb178b04ab535edb4e3f0ced204d00a Author: Stuart Bennett Date: Thu Mar 6 22:35:30 2008 +0000 randr12: workaround for xserver randr bug, where the hwcursor always gets displayed after a mode change commit c424144e854563709a01ed2d30a0482edc57fbf3 Author: Stephane Marchesin Date: Thu Mar 6 23:15:31 2008 +0100 Fix my nv10 cursor. commit 633047c65708f1a39e371c2c5da5c111027af1af Author: Maarten Maathuis Date: Thu Mar 6 23:05:43 2008 +0100 randr12: A major rework of output resource conflict handling. - This mostly affects laptops, such as some geforce5 and the famous 7300go laptops. - Test this if you know to have a strange vga/dvi port of your laptop. commit c5230961d181f15cd965c4569813219d9292681a Author: Maarten Maathuis Date: Thu Mar 6 19:29:50 2008 +0100 randr12: Readd some useful information about how an output is routed. commit 97ac866a4e433c70d0de7a1e8fb66a3994194ffd Author: Maarten Maathuis Date: Thu Mar 6 19:16:36 2008 +0100 randr12: Kill almost all usage of ErrorF + some misc cleanup. commit ea8fba29bcfe8158c9bf17754a20684430027aee Author: Stuart Bennett Date: Wed Mar 5 18:53:54 2008 +0000 randr12: don't reset panel unnecessarily for mere backlight control commit 22ec35504fbef83ba5790c87d27cd13e936fccb0 Author: Stuart Bennett Date: Wed Mar 5 14:40:08 2008 +0000 randr12: remove switchable_crtc stuff commit 3b0a1a11f662bfa705806f4afc5898338234ca65 Author: Stuart Bennett Date: Wed Mar 5 14:35:41 2008 +0000 randr12: if no ddc, return XF86OutputStatusUnknown, for !twoHeads and nv11 commit 424500cf828903e3449d4bca22a083087258cc99 Author: Stuart Bennett Date: Wed Mar 5 02:15:28 2008 +0000 randr12: do not change CR43 from its initial value commit a764d3ca2741243f49f34e9a4930a59272927902 Author: Maarten Maathuis Date: Wed Mar 5 09:44:46 2008 +0100 Give some defines more consistent names. commit 234ce16558e59dff302f13ea3d8eb48f46eedf0d Author: Stuart Bennett Date: Wed Mar 5 01:25:20 2008 +0000 Use the BMP structure's I2C info, instead of DCB 1.2's, as marcheu's bios is broken commit bb5429f08b3ffd16d6786bd65d6a4c8db1210ade Author: Stuart Bennett Date: Wed Mar 5 00:38:15 2008 +0000 Add another DCB 1.4 entry commit 5998ba0a35bc1c897916f66b0bcfb9a3e0a7c54d Author: Stuart Bennett Date: Wed Mar 5 00:34:16 2008 +0000 Don't zero out i2c_read and i2c_write pNv is calloc'd, so zero inited, and explicit zeroing overwrites the stuff parsed in parse_bmp_structure. Duh. commit 2d74bcb4602bb0cdae18ed1ffdd17520a3b57a78 Author: Maarten Maathuis Date: Tue Mar 4 21:18:49 2008 +0100 Put braces around a bitmask. commit c7ab8e84bf8ad8ee6106b21a1c07b6d8470e7f11 Author: Stuart Bennett Date: Tue Mar 4 13:34:16 2008 +0000 Add an LVDS output setup quirk for Powerbooks Again, as a result of much testing by moondrake and sbriglie. Hopefully it even works this time :) commit fd9b37b31896e1731bdd56ec847c86c0502ba0f8 Author: Maarten Maathuis Date: Mon Mar 3 19:04:09 2008 +0100 randr12: Only change vpll value when needed. commit 89882b416de3528cf504c989cc147a86d3572053 Author: Maarten Maathuis Date: Sun Mar 2 20:25:45 2008 +0100 randr12: Keep the gamma values when setting a new mode, instead of always resetting to default. commit e8c363b7ca4a416be37a1ea1c97bbd99b21d5be4 Author: Stuart Bennett Date: Sat Mar 1 15:05:09 2008 +0000 Improve link_head_and_output Thanks to moondrake and sbriglie for their work on this commit 4a16414c77ad77e7612f62f4ca1859e0d6a24cc2 Author: Patrice Mandin Date: Fri Feb 29 22:17:30 2008 +0100 nv30: Remove fix that is useless now, and maybe broken in some cases commit 5cb164876b824e8bf52aa44ccf4a797bab310602 Author: Stuart Bennett Date: Thu Feb 28 00:43:17 2008 +0000 Less magical way of setting dual link on EDID equipped panels Eldritch coupling ftl commit 1127974ed8517f28ef9c451d81bf1c61a2b0e6ae Author: Maarten Maathuis Date: Wed Feb 27 23:45:34 2008 +0100 NV40EXA: Support extend/repeat type pad. - Note that all xservers, up to about a minute ago, did not try to accelerate repeat type reflect or pad. commit 6980d0e24a9dae52903e6971c5c8a18240a7dbc1 Author: Stuart Bennett Date: Wed Feb 27 12:55:54 2008 +0000 Add default dual link transition frequency For card bios versions before the relevant field was added Should fix Apple laptops. commit 27e2e55698ab05bf3469e0f68474d4c3f021ca31 Author: Stuart Bennett Date: Tue Feb 26 18:07:39 2008 +0000 Deal with pre-nv40 laptops with EDID commit 41992eb90828976105cbc8bba87fe7d079d7c06a Author: Patrice Mandin Date: Fri Feb 22 17:29:12 2008 +0100 Revert dummy commit commit 073d852f36df4a18563ea892e89eca269efd55ef Author: Patrice Mandin Date: Fri Feb 22 17:27:51 2008 +0100 Dummy commit commit 5db7920763aee3a1b124139273566f7add5e27df Author: Stuart Bennett Date: Sun Feb 10 23:14:52 2008 +0000 Use some of nv_hw's functions in nv_bios.c commit fa8891e78f37edc02537704172156ba77371a307 Author: Stuart Bennett Date: Wed Feb 20 00:50:48 2008 +0000 Unbreak console restore again, and use a less ambiguous function name for head locking commit 38d42a7857480a930e2025e4f63aa7f5defca5c5 Author: Stuart Bennett Date: Tue Feb 19 23:59:12 2008 +0000 randr12: rename NV*VGA functions for api consistency Also nuke NVCrtcSetOwner wrapper commit c77a54768c3fed5939755f23c5c067bb86f49805 Author: Ben Skeggs Date: Thu Feb 21 00:55:39 2008 +1100 xv: don't crash on init commit cf963a128374bf7c2070e8a380e236b5e083adc1 Author: Maarten Maathuis Date: Wed Feb 20 09:36:42 2008 +0100 randr12: Only NV11 needs byteswapping + NV11 uses premultiplied cursors i think. commit 87f1676a243cddcaece9eeb0487a85b7e74b0f13 Author: Maarten Maathuis Date: Wed Feb 20 00:14:12 2008 +0100 NV30/40TEX: Create two adapters, one with fancy filtering, the default without. - Also enable sync to vblank by default, strange that it was off. - Remove the GET_TEXTURE(D)_PRIVATE macros, as one of them pointed to the blitter and one is now ambigious. commit 7ee1bb47d86396cc1a0f7879a98d04588cb31eee Author: Stuart Bennett Date: Thu Feb 14 23:16:01 2008 +0000 randr12: make more hw funcs generic commit 11de8d7a6f9a2dfd5a4f93f031865e284268a0f3 Author: Stuart Bennett Date: Tue Feb 19 21:44:56 2008 +0000 Use fp strapping to determine whether straps are used for flat panel modes commit c1ac6b977a4756e2dbfdb75ab5b53c7c10e76cf4 Author: Stuart Bennett Date: Tue Feb 19 14:53:41 2008 +0000 Rehabilitate bios opcode 6d commit afb42307cb2810ca2d759df62240fbf1f0cc7f7c Author: Stuart Bennett Date: Tue Feb 19 14:23:06 2008 +0000 Spruce up NV04 bios parsing a little commit dfabaab48034bf36c19510fe5de0f1bdaa2157ef Author: Stuart Bennett Date: Tue Feb 19 00:37:27 2008 +0000 More DCB v1.5 entries Courtesy HenneNWH's NV11 commit 2a4373d4f3b092600651b7dc8e2f9f9b520c7e0d Author: Stuart Bennett Date: Mon Feb 18 17:32:43 2008 +0000 Treat 0x4 and 0x8 bits in LVDS DCB v2.0 configuration as meaning the same thing commit daa557c0dadb3ad5caf39a4e90d3ebd16b12fc96 Author: Danny Tholen Date: Mon Feb 18 05:00:03 2008 +0100 Fix PPC DFS/UTS by using PCIgart. commit f8f469286fb3acc1ee694c61fe21e42123a5456a Author: Ben Skeggs Date: Sat Feb 16 03:41:38 2008 +1100 dri: fix broken display in some cases commit a293f32b2172837b6e2bcc3f5771c5219fccfb14 Author: Stuart Bennett Date: Wed Feb 13 00:52:34 2008 +0000 Fix bios pll setting regression Oops. commit 7121fe9dfb1c9162c8be808e09ed6b4b3ce3999e Author: Stuart Bennett Date: Tue Feb 12 18:20:43 2008 +0000 Use a loop to read RAMIN bios image, rather than memcpy commit 3ad7c35740905ad267faf8e248075640d5ee9cea Author: Stuart Bennett Date: Tue Feb 12 17:23:38 2008 +0000 randr12: reorder some variable declarations in nv_crtc and nv_output commit ab40a46430f369a7bd9cef8db4e86414125c8437 Author: Stuart Bennett Date: Tue Feb 12 15:19:58 2008 +0000 randr12: use integer math for scaling No floats in modesetting path commit 4a44468c29c37b5ad4828439f37d023865a9d457 Author: Stuart Bennett Date: Tue Feb 12 15:03:46 2008 +0000 randr12: tidy up some mode retrieval stuff commit 50d77c4199a3a5d2f3ead8327341e91ac2ec5684 Author: Stuart Bennett Date: Mon Feb 11 16:27:27 2008 +0000 Trim useless nv_type members Also, crtc_active isn't used on randr12 path commit 0c1e48a2192555856894d7e2b3025cd342c2f448 Author: Stuart Bennett Date: Mon Feb 11 01:42:04 2008 +0000 Remove some DDX cruft commit 5a8e246fcab3e349f48d0ce0c4da812a411767f2 Author: Stuart Bennett Date: Mon Feb 11 00:34:27 2008 +0000 Use mmio access wrappers and reorder some variable declarations in nv_bios.c commit 2ffcd89bfb0383fb32e8ec41f1cd472764b9b99d Author: Stuart Bennett Date: Mon Feb 11 00:03:50 2008 +0000 randr12: Use absolute regs for 8 bit regs commit c9b1f020d1c5be069dae19f99b3c26bb126052ae Author: Stuart Bennett Date: Fri Feb 8 19:44:37 2008 +0000 Move non-crtc specific hw funcs from nv_crtc to nv_hw commit 6cb8fb6cc63bf55f18b4dde53c393336e817474e Author: Stuart Bennett Date: Thu Feb 7 02:04:20 2008 +0000 Ordering mechanism for BIT table parsing, and (hopefully) automagic laptop detection commit 4f690235d70ae249323439380a69d8309a8c9981 Author: Stephane Marchesin Date: Wed Feb 6 22:18:46 2008 +0100 Replace a const with its name. commit 75c5912fffde88bfe2ddd158a446588b76ef44d9 Author: Maarten Maathuis Date: Wed Feb 6 19:06:14 2008 +0100 Fix bad debugging in NVWritePVIO(). commit 50da8b8f5b5feb6787ed50f221529730d25b5467 Author: Stephane Marchesin Date: Wed Feb 6 17:44:52 2008 +0100 Fix UV swapping on PPC. commit 4ff11d13be6e8050e0b726dc7af8e4bc11bd9d0d Author: Stuart Bennett Date: Mon Feb 4 22:34:21 2008 +0000 Missed out "DEBUG" part of the name in the define commit 181ee3bcecfc778c9f671853a696ab3224ac4747 Author: Stuart Bennett Date: Mon Feb 4 22:31:02 2008 +0000 Load correct PLL limits for single stage PLL cards with blank PLL tables commit cbcb1070db6fbb4d13ddcc35739d2d3e49f3766d Author: Maarten Maathuis Date: Mon Feb 4 07:47:06 2008 +0100 Once again kill usage of an exa private function. - Must have creeped back in during all the texture adapter changes recently. commit f14d27d638e75faa5bbda903f65a4d5479222ead Author: Stephane Marchesin Date: Sun Feb 3 17:15:50 2008 +0100 Remove include guards & add copyright. commit 7522640ee8d342c5ebb62989b051f73e230f5ce7 Author: Stephane Marchesin Date: Sun Feb 3 17:01:50 2008 +0100 Last round of fixes for nv30/40 bicubic Xv. commit 78553ab9a0309c7d27632fe7184f3233cfde6610 Author: Stephane Marchesin Date: Sun Feb 3 06:07:35 2008 +0100 Fix tearing. commit 7240f861a98c5f8d9779a4f3a743b748af775337 Author: Stephane Marchesin Date: Sun Feb 3 05:15:41 2008 +0100 Understood and fixed some NV30 FP_CONTROL bits. commit bb2a7bd3461eec37f179748a9bdb9375b1c3da04 Author: Stephane Marchesin Date: Sat Feb 2 05:54:47 2008 +0100 Fixes to nv30/nv40 textured video. commit c10d30dfefe93a9911d89eaffe3d29d3cff66a96 Author: Stephane Marchesin Date: Thu Jan 31 06:15:02 2008 +0100 Misc textured video fixes. commit cfdfd5a99bb7ac959a1188994db4c75a97de2e34 Author: Stephane Marchesin Date: Thu Jan 31 06:00:59 2008 +0100 Of course, add the nv30 video texture file. commit 970a8e474c666d6e33351f7ca6a3a5273024baf4 Author: Stephane Marchesin Date: Thu Jan 31 05:59:08 2008 +0100 nv30 video texture. Doesn't work with bicubic yet, only bilinear. commit c1eba822f1579637f59933b04e31591e86b975fc Author: Stephane Marchesin Date: Wed Jan 30 02:38:14 2008 +0100 Remove (void)ing that's not necessary any more. commit 7356ad90b387d01c2bf25aa653c41ca9365c5c20 Author: Stephane Marchesin Date: Wed Jan 30 02:30:39 2008 +0100 Remove unneeded stuff. commit e06dfd2964f9bd7f0855a101f5e6971020dfa6be Author: Stephane Marchesin Date: Wed Jan 30 02:19:08 2008 +0100 Separate the shaders from the rest of the code. commit a69c50780cacf0d0f2948208cfa7e0558a8c3644 Author: Stephane Marchesin Date: Wed Jan 30 02:18:26 2008 +0100 Macros for the PME range. commit 47c8ed472f37d68f3cc241a58ee5b7b8643e537d Author: Stephane Marchesin Date: Wed Jan 30 02:17:22 2008 +0100 Add PME range. commit 63f47c9bf59ccb9f1b6adbefd55deaec1435faf7 Author: Stephane Marchesin Date: Tue Jan 29 21:25:28 2008 +0100 Small cleanups. commit c98e9fa5c572deeafd7f605bb29d6c83cd38de9d Author: Stephane Marchesin Date: Tue Jan 29 21:24:39 2008 +0100 nv40 video texture: first pass at bicubic video. commit 10967a4dac8c3eaf828359dcba7a6e1272561aa3 Author: Stephane Marchesin Date: Mon Jan 28 04:52:01 2008 +0100 Update to newer nouveau_class.h commit 82728b3df45e8a5c0a45462d2ce847a41fd31f07 Author: Stephane Marchesin Date: Sat Jan 26 19:02:50 2008 +0100 Update to latest nouveau_class.h commit 9d9a5acae1deae8c76b512362634066e0c32a817 Author: Jaime Velasco Juan Date: Sun Feb 3 15:45:32 2008 +0100 Don't FIRE_RING() when in NoAccel mode. commit 11c1ee1cb0ee9c8096ea371d95d188c538ac3fb2 Author: Maarten Maathuis Date: Sun Feb 3 15:43:35 2008 +0100 Some minor reindenting. commit 0db347294e09d3d17200ca130e719df2d22f69b6 Author: Ben Skeggs Date: Mon Feb 4 01:19:23 2008 +1100 exa: those limits were a bit conservative on some hw. AFAIK nv2x up should be able to handle 4096x4096 textures and render targets. commit eb72e67641215b710868da7bb66b466f6ccbbe25 Author: Ben Skeggs Date: Mon Feb 4 01:07:07 2008 +1100 exa: more accurate coordinate limits commit 8645d2d9554ee46397240fbd9904fd4f905aefb2 Author: Maarten Maathuis Date: Sun Feb 3 14:54:41 2008 +0100 Fix NVExaPixmapMap() for NOUVEAU_EXA_PIXMAPS case. commit dd81efd0cc64a070f7489decbac345cc1b4fa9ce Author: Maarten Maathuis Date: Sun Feb 3 12:03:43 2008 +0100 Fix undefined symbol. commit 2a1116cbf61e84b16366076366f4bbd5f550fb9f Author: Maarten Maathuis Date: Sat Feb 2 20:34:57 2008 +0100 Remove a few leftovers to actually make the last commits work. commit 714bea6449aab47a9f2211718183ee84a3c70129 Author: Maarten Maathuis Date: Sat Feb 2 19:53:32 2008 +0100 Kill NVShowHideCursor and use a wrapper. commit 741acded643945c5a2da7bef088ed1ff7f5fa85e Author: Maarten Maathuis Date: Sat Feb 2 19:26:13 2008 +0100 Remove nvRead/WriteVGA in favor of nvRead/WriteCurVGA wrappers. commit d429cabd9ae9865a45cc88f43ab6d9d61b3fcf36 Author: Maarten Maathuis Date: Sat Feb 2 19:17:32 2008 +0100 Kill NVLockUnlock and make it redirect to NVLockUnlockHead. commit dc2aee0ef69c85aaffdc3c1da5a228642d049a30 Author: Maarten Maathuis Date: Sat Feb 2 15:12:00 2008 +0100 More wrapper death. commit a57fd67a2a6ae66af218fedd704a04d855260a2f Author: Maarten Maathuis Date: Sat Feb 2 15:06:02 2008 +0100 Let some wrappers die. commit b9a4eb6f6ea4f8f0747baa50037e33a0584cfd29 Author: Maarten Maathuis Date: Sat Feb 2 12:47:32 2008 +0100 Undo last commit and move it to drm. commit 288b56545a8da1e7b5aa64b1093316ca2f28f868 Author: Maarten Maathuis Date: Sat Feb 2 12:28:44 2008 +0100 NV40: Fix longstanding issue with nouveau crashing after using blob. - Took me a while to track down, but for me this is definately it. - In my case i got the X cross and then it would freeze and i would get a fifo dump. - Don't be shy, let me know if this works for you too. commit 0ec1b10a82c411181ff7a8643bb2390cfe08e4e4 Author: Stephane Marchesin Date: Sat Feb 2 04:04:58 2008 +0100 Don't allocate the whole AGP gart as a scratch buffer, since we still need to allocate the fifo afterwards. commit 5b79b0a9e54fe99eaae5b411f46a601b6944b773 Author: Maarten Maathuis Date: Fri Feb 1 23:34:32 2008 +0100 Don't use functions that are not supposed to be used in randr12 mode. - Also add NVLockUnlockHead() and NVSetOwner() functions. commit cedb604ee79cd73a9ec3940c7b056109811ba670 Author: Maarten Maathuis Date: Fri Feb 1 22:41:09 2008 +0100 Remove usage of "illegal" wrappers in code that also used for randr12 mode. - I've got something prettier in mind to clean this up, which will likely follow. commit 1a5247856929877442e1ec00f190c6d497ca64fe Author: Maarten Maathuis Date: Fri Feb 1 19:14:38 2008 +0100 randr12: Remove a hack, confirmed on 7600GS that it is unneeded. - This was one of the cards that did need it in the past. commit b6972e7bec0455250851c06d142336cb53be1a74 Author: Maarten Maathuis Date: Fri Feb 1 14:47:32 2008 +0100 randr12: Some minor fixes. commit 988dce0f452f6b80b2d02e1a0f78f826674cc3c0 Author: Arthur Huillet Date: Fri Feb 1 10:04:32 2008 +0100 NV10 EXA: limit regcomb state re-emission, fix logic for A8 + A8 yielding a 3% perf improvement :p commit a84852c2e0b19cf848495dc7e4bde4426b7bd199 Author: Maarten Maathuis Date: Wed Jan 30 22:57:09 2008 +0100 (hopefully) Fix build on powerpc. commit e8a975e6a8afe5ae9c77d20f4d59d3d10221f7d9 Author: Arthur Huillet Date: Wed Jan 30 18:37:39 2008 +0100 Xv: NV30 overlay can not scale down by more than 0.5x commit adc26892f6e711f08a5ed189a643e3fe9646e15c Author: Stuart Bennett Date: Wed Jan 30 17:43:16 2008 +0000 Fail gracefully for too-big resolutions / colour depths commit 09c9d7c3c50b630e1509f787ff946cc0656f7cba Author: Maarten Maathuis Date: Wed Jan 30 18:50:37 2008 +0100 randr12: Misc change. commit 6a46bf59f15d68404b536d436bf70337cd88e33d Author: Maarten Maathuis Date: Wed Jan 30 18:35:18 2008 +0100 Check if the framebuffer can fit into the offscreen memory. commit 9607d7828b16b0cb4a428bc29e2ce7c9bfd95371 Author: Arthur Huillet Date: Wed Jan 30 18:24:44 2008 +0100 Xv: NV04 overlay can not scale down at all, as confirmed by DirectFB commit 7edfb473533ce41db9e2b3709d32bb8ca7da85a4 Author: Stuart Bennett Date: Wed Jan 30 14:44:14 2008 +0000 PBUS define additions commit 6e090591bd2dab051362d8528a3bc76f30418d94 Author: Stuart Bennett Date: Wed Jan 30 14:35:25 2008 +0000 PCRTC define additions commit e51799382e40c033b2d6d09dd65dd7f7c1e5e4d6 Author: Stuart Bennett Date: Wed Jan 30 15:03:06 2008 +0000 Missed these in the ->absolute conversion commit ee0d557a26d326af27445bd0b283a5c6ead7c1cd Author: Arthur Huillet Date: Tue Jan 29 18:46:16 2008 +0100 Xv: detection of composite for the overlay commit 3a5f92d48e32d251e41bab561e52082033d26d9d Author: Stuart Bennett Date: Wed Jan 30 01:31:43 2008 +0000 Make bios port io use PVIO as appropriate, and have harsh validity checks commit 92c0d9630843957212d8ce296fea6bca8d802069 Author: Stuart Bennett Date: Wed Jan 30 00:36:59 2008 +0000 Make PLL setting quieter commit f58d174bd50943c3dfb2d6c909d0a98dbbd35dae Author: Stuart Bennett Date: Tue Jan 29 23:10:27 2008 +0000 Move some regs from nv_bios.c to nvreg.h And use more symbolic names in nv_bios commit d0022c35f39fde74ee5f3978ea78e053988fb608 Author: Stuart Bennett Date: Tue Jan 29 22:41:40 2008 +0000 Tighten bios reg checks a little commit b1b1d76eb187f4ab3020bbd69c2b1cc9b39d0d83 Author: Stuart Bennett Date: Wed Jan 30 00:30:46 2008 +0000 randr12: make nv4x plls less special commit 8f05d66bad8e74129cffa2918f9ecf6824c9f0b9 Author: Stuart Bennett Date: Tue Jan 29 22:33:34 2008 +0000 Don't attempt to parse BIT D and L tables on non-mobile cards commit 8e834ece6592df13bd3eb7e9f28c01b4eb0c9791 Author: Stuart Bennett Date: Tue Jan 29 18:57:28 2008 +0000 Add some mobile pci-ids, and reorder numerically commit 0b586ae0a2998339a9d5d3590e91dfa62af9ab6a Author: Maarten Maathuis Date: Tue Jan 29 19:04:25 2008 +0100 randr12: Remove excessive writing of VTOwner. commit c522ebb293a1c37142e29bf02bb82524e99336b0 Author: Stuart Bennett Date: Tue Jan 29 17:30:32 2008 +0000 Make sure nv04 has some defaults (as the bios parsing is non-existant) commit 62b6c5fe490578990c249042657f20bb0cf49a01 Author: Stuart Bennett Date: Sun Jan 27 22:11:31 2008 +0000 Use absolute regs in the DDX Due to having to make a few defines for regs that were hardcoded, there could be breakage (nv04 overlay, this means you) Hopefully not. commit f5c9f9910a46cfa41d817ef7ca345bfea278b358 Author: Stuart Bennett Date: Tue Jan 29 14:14:26 2008 +0000 Reorganise pll limit retrieval commit 26ff67f03be38925905a572dab3ec61653cd8286 Author: Maarten Maathuis Date: Tue Jan 29 10:53:14 2008 +0100 NV30EXA: Always init texture origin, please report if your viewport setup needs to be different. commit 38df9a95740e934b9691dff1ed1b597cd2254cfa Author: Maarten Maathuis Date: Mon Jan 28 20:55:47 2008 +0100 Forgot to be paranoid. commit 2ea0314859569c3ae4356272a35f0b316abf4df7 Author: Maarten Maathuis Date: Mon Jan 28 20:51:20 2008 +0100 Small improvement to nouveau_dma_wait, at least makes it show up in profiling in a clear way. - This may also slightly reduce the cost of a very short wait, but not by much. commit dd483e226519bc4d2c088e97490039613014dfac Author: Stuart Bennett Date: Sat Jan 26 20:22:08 2008 +0000 randr12: 2nd attempt on nv11 console restore commit 5b0fa02101c82a83e11f7d279d366689d54bfacb Author: Stuart Bennett Date: Sat Jan 26 18:09:10 2008 +0000 Use NV_{WR,RD}08 to avoid double logging io And one less bios FIXME commit a099bbff8bb3766201a7deb9c4a718fff7edbe51 Author: Stuart Bennett Date: Mon Jan 21 23:54:51 2008 +0000 randr12: remove unused PLL code commit 0eb428e5a8b60e52d18b3bdffed46ae3f07b7bc0 Author: Stuart Bennett Date: Fri Jan 25 16:05:59 2008 +0000 randr12: attempt to fix brokenness of head setting / getting on nv11 commit eea2fd41ebc17f72b43c776a0391df0733b8d6e4 Author: Stuart Bennett Date: Sat Jan 26 14:20:01 2008 +0000 randr12: don't do load detection on nv11 commit a2c0facefbc878815e00c734464b3c87b5cce842 Author: Stuart Bennett Date: Sat Jan 26 01:46:59 2008 +0000 Assume 18 bit laptop panels by default, add a message that may trigger on 24 bit panels Revert "Invert 18/24 bit interface flag, so that the default value (0) does not imply dithering" This reverts commit 5b9dc7ad6eb2f16c4ca8824c2a19c85bbe1f126b. commit ec548d20239a49f8e9eb0399317d4c68df24f4a3 Author: Stuart Bennett Date: Sat Jan 26 01:36:41 2008 +0000 Detect mobile BMP biosen I'd really love to know how where this info is in BIT commit 9f932eb684814e2a04c83d5aee172b9e020d82a3 Author: Maarten Maathuis Date: Fri Jan 25 23:31:58 2008 +0100 NV30EXA: Fix viewport setup + switch to triangle based composite. commit afca5f2665a62ef2d025c15f07fe1cdc2b40d8d5 Author: Stuart Bennett Date: Fri Jan 25 01:18:50 2008 +0000 randr12: fix / annotate a couple of ddc thinkos catch a longstanding source of error too commit 3e6ea79c3292c3ed5bd676653c4c40532772ae5f Author: Stuart Bennett Date: Fri Jan 25 00:33:07 2008 +0000 Implement opcodes 0x66, 0x67, and 0x68 Courtesy jwstolk's trace Names taken from nv11 biosmod scps commit cdae2e26cc432ba572f92522f1b3454264d38f43 Author: Maarten Maathuis Date: Fri Jan 25 00:10:08 2008 +0100 randr12: Rework native modes for TMDS. - Choose a preferred ddc mode when available, otherwise the one with the highest refresh. - Fall back on cvt modes, only when needed. commit 46636052498d1e33985e3d98a6ce9462639bd4ff Author: Stuart Bennett Date: Wed Jan 23 23:02:46 2008 +0000 oops commit e067a0eee42324ec7c01e4d84a33554054817dd9 Author: Stephane Marchesin Date: Wed Jan 23 23:40:51 2008 +0100 Use proper version. commit 324c19acf6b506aae47e3699dd0fcefb72473590 Author: Stuart Bennett Date: Wed Jan 23 01:13:58 2008 +0000 Fix C51 refclk commit 178b8bbf0e09c190fef8b2f85acddd3172ace629 Author: Stuart Bennett Date: Wed Jan 23 20:59:50 2008 +0000 randr12: calculate (and set, for nv40) sel_clk before calculating plls commit 8067bd08d9a928c97b83f2b6ebff996b320c5813 Author: Maarten Maathuis Date: Wed Jan 23 19:43:42 2008 +0100 Fix big bug in commit "Mostly log message changes". - Please don't hide such important changes in a commit like that ;-) commit c5329a42c509efe9336573ff7e32978082ba9355 Author: Stuart Bennett Date: Wed Jan 23 16:37:31 2008 +0000 Redo sel_clk setting and update its documentation My apologies if this breaks stuff. commit f07ab92e7bb419df5473e1ed843a79ef4ff08f30 Author: Stuart Bennett Date: Wed Jan 23 02:09:56 2008 +0000 Fail on LVDS DCB entry parsing issues commit 12efa8cdfe68c312458368ab876030b02919a922 Author: Stuart Bennett Date: Wed Jan 23 00:09:11 2008 +0000 Mostly log message changes commit 0faa15d9b84b2cad23424dfc1679f119c5b45108 Author: Ben Skeggs Date: Wed Jan 23 23:45:37 2008 +1100 not an appropriate place for this type of thing.. commit 9e982df106028edd0f7ce0feedbe8e082f29f2ac Author: Maarten Maathuis Date: Wed Jan 23 11:42:55 2008 +0100 randr12: Increase cvt mode to 72 Hz. - I've had occasions were i did slightly notice something at 60 Hz, and no problems at a higher setting. - I must admit i was a little bit tired at the time, but i figure it won't hurt (much). commit 71435dde5b2fd1c197ef5dc31b22ba40abcbca7e Author: Maarten Maathuis Date: Wed Jan 23 11:20:47 2008 +0100 NV40EXA: Use the same approach as the texture adapter for rendering. - Verified to work for abitrary rotation. - This should eliminate diagional tearing altogether. commit d9149bddc758cc0644630b26fe10fc563ba38ce9 Author: Maarten Maathuis Date: Wed Jan 23 10:28:25 2008 +0100 NV40TEX: Switch to a big triangle and clip it to a quad. - On some systems (NV43, NV46 are known) an artifact shows when 2 triangles are combines and the size is larger than roughly 512x512. - The trick is draw a single triangle and use scissors to turn it into a quad. - Someone once asked me if i ever made drawings, well i did in this case and i've included it for future reference. commit e578d6e0a8c0fb8877d8b4b6ba1046dccf28ee6d Author: Stuart Bennett Date: Tue Jan 22 18:33:42 2008 +0000 Fix LVDS with DCB 2.0 Also use the X_ERROR facility for errors, not ErrorF commit fd3195f503420591b522249bb423c817db0709e7 Author: Maarten Maathuis Date: Tue Jan 22 19:00:49 2008 +0100 Xv: Remove illegal usage of private exa symbols. commit 3a995b06ddb507655279565e992604bd40036b19 Author: Maarten Maathuis Date: Tue Jan 22 15:53:38 2008 +0100 NV40TEX: Use width, not (x2, y2). commit abaa8e822be49b9d2129d31c4996dfcf8034ea2a Author: Arthur Huillet Date: Tue Jan 22 11:42:19 2008 +0100 Xv: NV12 overlay copy - removed wrong byteswapping on PPC, reported by moondrake commit 782237e4e942f2e8a9ed3dd111d60c665459963c Author: Stuart Bennett Date: Mon Jan 21 23:44:47 2008 +0000 randr12: some tidyup commit 91ca67056cb236e5770d4fb39dd7fbb9818ff093 Author: Stuart Bennett Date: Mon Jan 21 22:35:58 2008 +0000 randr12: more beautiful dithering commit 1d5dc47324ddd1939db166af60df7a3a0e055810 Author: Stephane Marchesin Date: Mon Jan 21 22:19:24 2008 +0100 oops :) commit 7080f93a83e0d680f53a7962df300d527edd121a Author: Stephane Marchesin Date: Mon Jan 21 21:43:11 2008 +0100 Quickly hacked support for NV6x. commit 2d399af61424cc4b4e0080ea016423f1b22c02f3 Author: Stuart Bennett Date: Mon Jan 21 20:16:56 2008 +0000 randr12: fix dual head restore commit 8be20d6fc21a3869786e8c19656bf8b6569fee6e Author: Stuart Bennett Date: Mon Jan 21 18:51:52 2008 +0000 randr12: save and restore palette commit d32f310afa05ad85d6675348400880506d98ac1e Author: Stuart Bennett Date: Mon Jan 21 15:03:59 2008 +0000 Make LVDS bios script logs easier to understand and fix a backwards comparison commit 7c7f0194678c7bd7c5ce791c07b38ac8eeeef608 Author: Stuart Bennett Date: Mon Jan 21 14:14:07 2008 +0000 randr12: avoid SIGFPE commit b9a3aab7a7f6c5c4a5c440e46bbca7d76fd8df68 Author: Stuart Bennett Date: Mon Jan 21 14:09:46 2008 +0000 debug_modes breaks build on old xservers commit d2e078b4ebf11b286e152e0ea10b17c41bf61b67 Author: Stuart Bennett Date: Mon Jan 21 14:06:42 2008 +0000 randr12: use correct clock for LVDS modesetting part 2 commit 52154e97b5a40b014fbee747fc542f765b4390bc Author: Stuart Bennett Date: Mon Jan 21 13:46:15 2008 +0000 randr12: use correct clock for LVDS modesetting commit d54a9f4b76f3a8c29c5fce6dfe7cc3fbc20f3952 Author: Stuart Bennett Date: Mon Jan 21 12:27:36 2008 +0000 Don't log strange characters commit 07e01061e9b90348ea38b8889ff9e8daf207b820 Author: Stuart Bennett Date: Sun Jan 20 21:53:30 2008 +0000 Use more imaginative log levels than X_INFO commit 18017632cbc96f36bbc72b454bd5b48070985a6f Author: Stuart Bennett Date: Sun Jan 20 20:17:28 2008 +0000 Use stdbool.h types in nv_bios.c It makes vim highlighting work better and reduces shift key wear. Coincidentally it's what the drm uses too. bool, true and false are defined in IEEE 1003.1 (2003) and POSIX since C99 (as is stdint.h, which we use already) commit e1714a4f4329728cf897f8953acc480b47a4e5d7 Author: root Date: Sun Jan 20 18:38:11 2008 +0000 randr12: fix early cards commit 271844d3a761af65c9625d1564e1c08a9f64d159 Author: Maarten Maathuis Date: Sun Jan 20 19:06:49 2008 +0100 NV40TEX: Don't sync when compositing, will have bad sideeffects if the compositing manager tries it too. commit c338a14ea688953b03f274d0e436b579d0411f18 Author: Maarten Maathuis Date: Sun Jan 20 17:28:13 2008 +0100 Some misc cleanup. commit 2cd6e157cebac663f1b3c88e3653a4c2c9971ff1 Author: Maarten Maathuis Date: Sun Jan 20 17:11:27 2008 +0100 randr12: Fix a major screwup i made during a cleanup. commit c3127a0c9cf4599c40b27e24a411a7fe4969152a Author: Maarten Maathuis Date: Sun Jan 20 17:00:58 2008 +0100 NV40TEX: Implement sync to vblank. - I don't know why this didn't work the first time i tried, maybe i missed it due to another bug (that also caused tearing). - This should also work reasonably well for composited situations, but ofcource we can't guarantee that. commit 512f4ba33528dafb505bc2db5f9bb4d1b92c3a7b Author: Stuart Bennett Date: Sun Jan 20 14:47:52 2008 +0000 randr12: init some more Xv state commit 3f6a681a38fccbd53dd10aa42cc8d1b9bfe0eb71 Author: Stuart Bennett Date: Sat Jan 19 14:02:43 2008 +0000 minor debug correction commit efd3cd086036cd5f8817aee771ee630066a92830 Author: Stuart Bennett Date: Sun Jan 20 01:56:00 2008 +0000 INIT_8E opcode again, from pq's trace commit f0bb436796a2d4dc693ec92b73c7abb27d428cda Author: Stuart Bennett Date: Sat Jan 19 01:37:34 2008 +0000 INIT_IO opcode Based on pq's g8x trace commit c1b69fdaa089898821338b14c1f72eda354c2caa Author: Maarten Maathuis Date: Sat Jan 19 19:24:04 2008 +0100 randr12: hide cursor before locking + misc cleanup. commit dfa8901e847e1fbcbcb6de0c0726f53531b15b34 Author: Maarten Maathuis Date: Sat Jan 19 16:39:26 2008 +0100 randr12: Remove unneeded vgahw stuff. commit 47452756724007318eb384912c40302b15523998 Author: Maarten Maathuis Date: Sat Jan 19 16:29:57 2008 +0100 randr12: Move away from some of the nv_hw wrappers. commit 9e3a2d7c8c36d4dc43396be6b742b6137dfef455 Author: Maarten Maathuis Date: Sat Jan 19 15:59:48 2008 +0100 randr12: Remove usage of nvRead/WriteVGA0/1. commit 936e9fe8adb2660cf75959265971c60f378cd7d6 Author: Maarten Maathuis Date: Sat Jan 19 11:28:41 2008 +0100 randr12: Some cleanup. - framebuffer consoles can use most of the normal stuff, so change it back. - some cursor deobfuscation commit a9bae6ba734882308ac29a9fe2403eaa027bd518 Author: Stuart Bennett Date: Fri Jan 18 20:49:58 2008 +0000 More detailed modesetting debug (both randr12 and oldstyle) commit a28216639c3cf01d8f7cc85030d69fd58be86ed9 Author: Stuart Bennett Date: Fri Jan 18 18:10:40 2008 +0000 Remove no longer needed panel quirk commit 59fec415ef32c291363408d5c8a24a2041450434 Author: Stuart Bennett Date: Fri Jan 18 17:02:32 2008 +0000 Use bios PLL functions on nv4x And a minor completeness correction commit 7efaf3d97a51d2447310954c9224fd7ec888caf9 Author: Maarten Maathuis Date: Fri Jan 18 23:45:51 2008 +0100 randr12: Refine CR4B, anyone know what this does? - There is more to this register. commit 41646042ab9d9d0c3bc8a76b25a4bff45f039922 Author: Stuart Bennett Date: Fri Jan 18 16:50:50 2008 +0000 Rerefactor getting pll limits for nv4x commit 174d8f7e015250af4c0b27e044f5088f5465d851 Author: Stuart Bennett Date: Fri Jan 18 14:24:48 2008 +0000 Use pll limits infrastructure on single stage plls And add default PLL limits on old cards commit b3ca42d8589bed843ce5e772a1807c4c95377b92 Author: Stuart Bennett Date: Fri Jan 18 01:47:05 2008 +0000 Don't speculate on the unknown limit field, and call the others log2p, not p commit d9db305e2d4bfb017a8542a7407ad4df27c86121 Author: Stuart Bennett Date: Fri Jan 18 01:18:42 2008 +0000 Better setPLL_single commit 2a94184112cf06225672ef34836e19723761140a Author: Stuart Bennett Date: Fri Jan 18 14:33:01 2008 +0000 Bios parsing safety checks commit 6251f2fa774b913f5622ff9d07f56b2a5b9d1888 Author: Maarten Maathuis Date: Fri Jan 18 12:17:50 2008 +0100 Some misc fixes. commit b21789f93c7e7df9db320dea093deab0f1056955 Author: Maarten Maathuis Date: Fri Jan 18 09:23:39 2008 +0100 randr12: Actually using SavedReg does tend to help ;-) commit f15ea75b88b68fadb64fdeef75a23329e2a3f903 Author: Maarten Maathuis Date: Fri Jan 18 09:13:24 2008 +0100 randr12: Fix a huge gaping hole in NewRestore. - Restore CR5758 in a better place and actually restore something proper (use SavedReg instead of ModeReg). - Minor other fix. commit cd427da27d69db000f07ec50d27ca56ded97994e Author: Stuart Bennett Date: Thu Jan 17 23:37:23 2008 +0000 Invent a DVI-A output for the DVI-D output, on DCB1.4/1.5 commit fbcb57827a1c4ac4b3c7ce66ca935f6b095f020d Author: Stuart Bennett Date: Thu Jan 17 22:49:37 2008 +0000 And another DCB 1.4 entry commit a355e5b0dcd5f5c78da20bd07f581b5841b3a2ba Author: Maarten Maathuis Date: Thu Jan 17 22:25:12 2008 +0100 randr12: Forgot to clear a few state variables on restore. commit ba0d0e8571b657b5196d0508347f7641cd7f04cb Author: Stuart Bennett Date: Thu Jan 17 20:48:16 2008 +0000 Add me commit 9cfc736ea4b089c74daaf1b0fac6393fec459bc0 Author: Stuart Bennett Date: Thu Jan 17 20:29:30 2008 +0000 That warning wasn't there a moment ago commit 7eded6883404ce0cff83051497518a99c1d8a832 Author: Stuart Bennett Date: Thu Jan 17 17:07:24 2008 +0000 Fix for situations with fixed second stage gain and hook code in to init scripts And make sure bad PLL values are easily seen commit 075a14ac2525b364edc583caac50e03d171f67a7 Author: Stuart Bennett Date: Thu Jan 17 19:59:55 2008 +0000 PLL limits reg off-by-four is dealt with elsewhere Also some more imaginative pll limit member naming and minor get_pll_limits debugging corrections commit a3ef2cb34ad9de4567d30625a3bf87720a5ed8da Author: Stuart Bennett Date: Thu Jan 17 17:30:19 2008 +0000 setPLL for nv40 lowreg PLLs commit c3bef786b5291a638791b8a5125b3ff60cf9d1b1 Author: Stuart Bennett Date: Thu Jan 17 17:30:36 2008 +0000 setPLL for nv40 highreg PLLs commit 996b93cdc73e3d6e581a766cbbf4a29bb34b8eb2 Author: Maarten Maathuis Date: Thu Jan 17 18:27:51 2008 +0100 randr12: Some misc fixes. commit aa561c487754210ca22a1fc1d17c54d8e199e7a1 Author: Maarten Maathuis Date: Wed Jan 16 22:42:18 2008 +0100 Some misc fixes, mostly related to setting the right crtc for overlay (randr12). commit 276d4e35673bb71410c882badf03550bcee48571 Author: Maarten Maathuis Date: Wed Jan 16 08:03:51 2008 +0100 randr12: Fix mode privates again, hopefully ;-) commit f89243e4c2ba33a822b15de9136736f44b4c1579 Author: Stuart Bennett Date: Wed Jan 16 00:25:08 2008 +0000 Minor reg adjustment commit df0f66750d1d893b2d91b0cfc91ecd4f0a1ba80e Author: Stuart Bennett Date: Tue Jan 15 16:30:00 2008 +0000 Refactor pll table parsing And correct table reg matching on version 0x21 tables, which is off-by-4 commit 3171b470f00a11d3528067c7d91e751257aa603c Author: Stuart Bennett Date: Tue Jan 15 14:59:49 2008 +0000 Use pll_lims in getMNP_double Also move quirks elsewhere, and optimize commit 7bc921485c63232fecb59e488ef203166c48c6e1 Author: Stuart Bennett Date: Tue Jan 15 15:14:46 2008 +0000 Store PLL limit frequencies in kHz, since that's what we use them in commit 51a739771b59a4725e75660b2c8edb54aa763ba5 Author: Stuart Bennett Date: Tue Jan 15 01:32:25 2008 +0000 Two stage PLL limits parsing corrections And some corrections to getMNP_double commit 7013116596ce8b065df265f16b3fdb9ac9e3df37 Author: Stuart Bennett Date: Tue Jan 15 09:53:58 2008 +0000 randr12: Oh no! More pNv->twoHeads commit c7081039bf27aad12e17063d1a9becf160e7ee49 Author: Maarten Maathuis Date: Tue Jan 15 21:05:08 2008 +0100 randr12: PrivFlags are not so private, switch to a real mode Private. - This hopefully fixes the obscure issues people have had these last few days. commit cd14cd8c520361a15f5f8d255171c85810d83033 Author: Stephane Marchesin Date: Tue Jan 15 20:44:48 2008 +0100 Order the reg ranges, add some. commit 17fa42166fb4e8102cbe880be6021f1145268d16 Author: Patrice Mandin Date: Tue Jan 15 18:45:30 2008 +0100 nv30: set VIEWPORT_TX_ORIGIN to 0 for all cards, except real nv30 commit 777a3df246feb364d440282db0864fb9b0908872 Author: Maarten Maathuis Date: Tue Jan 15 17:52:19 2008 +0100 NV30EXA: Chipset should be masked by 0xFFF0, switch to NVArch while i'm at it. commit b55608bac3f4f909f1ef8380a1ac69ca2f262ea2 Author: Maarten Maathuis Date: Tue Jan 15 08:10:46 2008 +0100 randr12: Reenable cvtmode. commit 0db297bb6199878c0f77935acc447c568a87c900 Author: Maarten Maathuis Date: Mon Jan 14 22:54:42 2008 +0100 randr12: Undo some changes of the past + always set vpll on modeset. commit 2879f4518e6830cf0b490cde5b4daca3e931383d Author: Patrice Mandin Date: Mon Jan 14 23:26:14 2008 +0100 nv30: Only change VIEWPOR_TX_ORIGIN on real nv30 commit 405fd62ee8b56e3b1d6cc78df7ecfa654cd54dec Author: Maarten Maathuis Date: Mon Jan 14 20:17:05 2008 +0100 randr12: "NewRestore" Always reset some state regs after restore. commit 8093b83c3d1393ba225f9def1ce055b04cba5a8f Author: Maarten Maathuis Date: Mon Jan 14 19:29:28 2008 +0100 randr12: "NewRestore" proper depth vs bpp awareness. commit 8733ea89d9acece8721a2ae2d62a68916746c374 Author: Stuart Bennett Date: Mon Jan 14 18:21:47 2008 +0000 randr12: yet more pre nv10 fixes commit c0d1117782673771b0490fe4d19abd83c6620033 Author: Stuart Bennett Date: Mon Jan 14 11:54:28 2008 +0000 Possible fix for C51's misaligned regs commit 6ac824446d65a28519f378b3c01b1a6683a14d1b Author: Stuart Bennett Date: Sun Jan 13 21:36:53 2008 +0000 randr12: better load detection (enabled) commit ef28e91b24ee34911b3238e681064a95be7afd3a Author: Maarten Maathuis Date: Mon Jan 14 00:48:25 2008 +0100 randr12: We didn't copy the Flags to the adjusted mode, very bad. commit d41d5ae34a7d7f4658a1d86635fb8c59b80a1355 Author: Stuart Bennett Date: Sun Jan 13 21:35:33 2008 +0000 Be more flexible on init table length commit 4744e5410c27d8bd3c2f92bee4554801639dba29 Author: Stuart Bennett Date: Sun Jan 13 19:37:28 2008 +0000 More DCB 1.4/1.5 support commit c043b3fa7cf6297e010dfa6fcff24f263b8a8fd8 Author: Maarten Maathuis Date: Sun Jan 13 17:09:04 2008 +0100 randr12: Minor fix. commit 3f7f0cab91416f6f4f4378ee6bdabeb9e75cf230 Author: Maarten Maathuis Date: Sun Jan 13 16:05:22 2008 +0100 randr12: Some "NewRestore" fixes for framebuffer modes. - This is mostly it for framebuffer modes. commit d8e7c5068835004741b74577344bd4defad4d139 Author: Maarten Maathuis Date: Sun Jan 13 15:05:15 2008 +0100 nvbios: Fix a bmp bug for malc0. commit 5c01f96caa2b3b7223383f37d6903fbcaf3643cd Author: Maarten Maathuis Date: Sun Jan 13 14:12:00 2008 +0100 randr12: Some "NewRestore" fixes. - Improvements to restoring framebuffer modes, still a few minor bugs to fix. - VGA modes work partially, i doubt this will make mainline though, since it's fragile. - This is a work in progress, only for testing. - Added some other functions to NOUVEAU_MODESET_TRACE. commit 364497a4234410adea25e4e51e0bdb17ccd3ba4a Author: Stuart Bennett Date: Sat Jan 12 00:39:59 2008 +0000 randr12: more fixes aimed at nv0x commit ac546e77fe71ffbbc76d9c48c5ab0b6faac832d7 Author: Stuart Bennett Date: Fri Jan 11 22:05:11 2008 +0000 Remove some nv_bios TODOs and shut up some warnings commit fc04d71de6ec36158040617fb973218a8bdd0903 Author: Maarten Maathuis Date: Fri Jan 11 23:23:43 2008 +0100 randr12: Add new experimental restore mode. - Option "NewRestore" - Only tested on my 16bpp 1280x1024 vesafb console. - It may work on text consoles. - This should eventually replace the somewhat frail restore system that we currently have. commit 361bbe074a0d674012615b05fd3da81e8e1ce34e Author: Maarten Maathuis Date: Fri Jan 11 22:30:14 2008 +0100 randr12: Intermediate commit. commit 3b6e551e1afefe86338ab830505dc08412b2e80b Author: Maarten Maathuis Date: Fri Jan 11 18:59:53 2008 +0100 randr12: Intermediate commit. commit 5a989b23322e6ba1abe6456246300067c3bf42cb Author: Stuart Bennett Date: Fri Jan 11 21:33:47 2008 +0000 Tidy up and correct DCB merging commit ec7889a5a0787929c8c8ae08aaf729929c483035 Author: Stuart Bennett Date: Thu Jan 10 23:44:59 2008 +0000 Fix bios opcode 0x56 and add reg alignment check commit 9b356d08f9169e35bdbff079c3351c475e53ec9b Author: Stuart Bennett Date: Fri Jan 11 16:34:28 2008 +0000 randr12: undo overzealous code disablement commit 44295b59f57be8956a5226415b5fdfbf49e175c1 Author: Stuart Bennett Date: Fri Jan 11 12:28:16 2008 +0000 randr12: twoHeads and other fixes for pre nv10 (nv_crtc.c) commit 80fa1e1aef1b093b0310c51817ee9685e4d9d251 Author: Stuart Bennett Date: Fri Jan 11 11:16:21 2008 +0000 randr12: some tidyup to output save/restore functions, and some twoHeads fixes commit 04f64725b5784ecfd7a5c3218d4710c2ad0f7f0f Author: Stuart Bennett Date: Fri Jan 11 11:20:45 2008 +0000 randr12: tidy up dpms functions also, don't set cr57/58 in output_prepare as the dpms funcs do it, and it shouldn't be done on early cards anyway commit cc6794abb637e5ac8c98d7c6385bd36aa43d1ebc Author: Stuart Bennett Date: Thu Jan 10 20:25:50 2008 +0000 randr12: use clocks in kHz throughout commit 8337b77b40da9c3bba106837b4553433fa4e8a76 Author: Stuart Bennett Date: Thu Jan 10 19:09:16 2008 +0000 The bios transition freq is in 10s kHz - so x10 commit d65f59144f8a79116bca2f09de41d5786a7a4f84 Author: Maarten Maathuis Date: Thu Jan 10 18:50:47 2008 +0100 randr12: Some minor things for supporting vga modes (in the future). commit 0e646b0332b52e23ef13301765f7f1e8f881e165 Author: Maarten Maathuis Date: Thu Jan 10 17:41:39 2008 +0100 randr12: Intermediate commit. commit a5122460cfa07d45cf91ae112db9170fecee6966 Author: Stuart Bennett Date: Thu Jan 10 16:21:49 2008 +0000 Remove weirdo special case handling on bios reg writes I don't think these ever get used, and the code only seems to be on gf6 anyway commit 31b1a894079ca3983d17ac7a81d6ecd420f7d377 Author: Stuart Bennett Date: Thu Jan 10 14:14:24 2008 +0000 randr12: dtrt wrt dual link on lvds when !straps commit e02d2cc8b77443fde4e267ce4d9691ed680eef01 Author: Stuart Bennett Date: Thu Jan 10 12:03:43 2008 +0000 randr12: improve setting of fp_control commit d8eb924a43576c0e360965a43f5363d0ba2aecd0 Author: Arthur Huillet Date: Thu Jan 10 12:03:30 2008 +0100 Xv: cleanup. commit 1f727083cfafa9097ed98266420a2fd16905ae10 Author: Stuart Bennett Date: Thu Jan 10 00:12:52 2008 +0000 randr12: enable LVDS and refactor output creation Heavily based on initial patch by seventhguardian commit c08c6de0f452c7f5cd1bac8bf974aaac872c8e2c Author: Stuart Bennett Date: Wed Jan 9 16:18:30 2008 +0000 deifdefication - we requre xserver 1.3, so randr 1.2 is guaranteed commit 1556144a1af3c89eaa14cbfe8bdc9dc966fef8be Author: Stuart Bennett Date: Wed Jan 9 15:47:50 2008 +0000 randr12: optimize away excessive LVDS_RESETing commit 55e6750038edd7010ae8304fe3c3b4ac9e7b963c Author: Maarten Maathuis Date: Wed Jan 9 22:08:49 2008 +0100 randr12: Fix minor bug, related to output resource conflicts. commit c1fe3766bd181d9ed273850e6c0d1939c0efc8aa Author: Maarten Maathuis Date: Wed Jan 9 20:20:03 2008 +0100 randr12: Reset crtc when changing output routes. commit 01b9a081c288c69fa0bbc0d3829a0026796e731a Author: Maarten Maathuis Date: Wed Jan 9 19:30:17 2008 +0100 randr12: Don't change vpll's twice on restore. commit 3643b2635dcaba3e6ec969a3cd6eda43623c8992 Author: Stuart Bennett Date: Wed Jan 9 15:12:30 2008 +0000 Detect an LVDS panel if there's a hardcoded EDID commit 64d8032c419ef8cf14cdf3f83228d3e0f5714909 Author: Stuart Bennett Date: Wed Jan 9 15:11:24 2008 +0000 If no PROM, and no PRAMIN, use the PCI rom image commit 99a64d31979a8c83ac9d52fabe3114c576c94bba Author: Stuart Bennett Date: Tue Jan 8 22:58:37 2008 +0000 fp bios edid oops commit commit b470cf257433dad174d62cdbd4331ada2809334e Author: Maarten Maathuis Date: Tue Jan 8 23:54:51 2008 +0100 randr12: More fixes for strange output routes. commit cfbdbd6ac846af79d9ddf2b4597736cba74d5916 Author: Stuart Bennett Date: Tue Jan 8 20:57:10 2008 +0000 randr12: try to use bios fp edid, if nothing else works commit d4e4892244c90b481bfd8c6d28d2390fe696e160 Author: Stuart Bennett Date: Tue Jan 8 20:07:10 2008 +0000 Read EDID from BIOS if appropriate commit 6fab536aa13b8639d9b419a37fccfb30a4313b77 Author: Stuart Bennett Date: Tue Jan 8 16:18:09 2008 +0000 Darktama's C51 is a mobile card commit 921f2e226ccbe7468cc8e51919872c3e8cf28c4e Author: Stuart Bennett Date: Tue Jan 8 15:37:51 2008 +0000 BIT cards use the init scripts to do the power sequencing commit 33ea2cdef8c46ee9837b6fc9809df82d4bdca339 Author: Maarten Maathuis Date: Tue Jan 8 22:36:12 2008 +0100 randr12: Small fix for NV3x, more will be needed. commit 175b91d48644bdbfc4e48f4c2fe6c9cf936802b5 Author: Maarten Maathuis Date: Tue Jan 8 20:23:38 2008 +0100 randr12: Fix reverse pll reading code for NV30. commit 4add8cfb68404e9e2764451c4e48ab4d0cf04482 Author: Maarten Maathuis Date: Tue Jan 8 19:07:50 2008 +0100 randr12: Some minor things for NV30. commit 74b03f7d4f5506222b6a95ccf8b5c9e0ff69e4dc Author: Stuart Bennett Date: Tue Jan 8 13:52:00 2008 +0000 Assorted small bios/randr12 bits and pieces Make bios execution a bit faster Better self documenting code for DCB dual link criterion Tidy up some ineffectual LVDS mode code commit 76f3c47736661df86490b0d133ac82c371cab73d Author: Maarten Maathuis Date: Mon Jan 7 23:38:08 2008 +0100 randr12: Handle NV30 vpll's correctly. - This only concerns the real NV30, not the other NV3x cards. commit 1a25def0a64ac6437c6ba5f51ddb8743d4a88d00 Author: Maarten Maathuis Date: Mon Jan 7 19:06:16 2008 +0100 randr12: Clear dpms cache when needed. commit 3f0008128e2afd2b1a1b595829c6324e46a4bf1d Author: Arthur Huillet Date: Mon Jan 7 19:03:53 2008 +0100 Xv: overlay BW workaround applied to NV40 as well commit d5dabd8c316371605d81be55a5c6c82e6351b591 Author: Arthur Huillet Date: Mon Jan 7 19:01:07 2008 +0100 Xv: removed useless #includes commit 510f3575df87733857dbb5bfe398ce04aebadfd4 Author: Arthur Huillet Date: Mon Jan 7 17:13:43 2008 +0100 Xv: added NV04 overlay brightness attribute. Original patch by Andrew Randrianasulu commit a71de657e1e2718800ef1c319dab5dafcc9ea0cd Author: Ben Skeggs Date: Tue Jan 8 02:18:59 2008 +1100 avoid symbol clash with 3d driver. commit 0463dc0682fb192ce82037601289f75913f65e3c Author: Ben Skeggs Date: Tue Jan 8 02:16:26 2008 +1100 nv50: support hostdata UTS commit 9fb186c57bdad07281f1a0bdebac98a4b18dc6e3 Author: Ben Skeggs Date: Mon Jan 7 18:33:03 2008 +1100 nv50: small fix commit b679d6a8539403f075a57d882bb4d0eb33b180d9 Author: Stephane Marchesin Date: Mon Jan 7 02:39:19 2008 +0100 nv40 video texture: sometimes less is more. commit b9a364a985bb23ee30dc094ba13e19ea2f4f6bdf Author: Maarten Maathuis Date: Sun Jan 6 20:44:56 2008 +0100 NV40EXA: Remove stair shaped tearing i get. commit 80ac301acaf8dba6a93f5b04b97f8dca2748b92c Author: Pierre Ossman Date: Sun Jan 6 20:04:10 2008 +0100 We do not get a valid CRTC configuration structure if we're not using RandR 1.2, so make sure we can handle the old mode with some form of grace. Signed-off-by: Pierre Ossman commit f1aaf3fe616f5a0633e685867cf33b37f2c3f07b Author: Stephane Marchesin Date: Sun Jan 6 19:14:03 2008 +0100 nv40 video texture: bilinear filtering for UV. commit 57fc0631428f29af0d24036c939826349e3018b5 Author: Arthur Huillet Date: Sun Jan 6 17:41:50 2008 +0100 Xv: overlay auto CRTC switch with randr12, blitter fallback Now switching the overlay to the right CRTC when displaying video with the overlay and dual head, with a nice fallback to the blitter when the window is on both CRTCs. Fixed a bug preventing the overlay from working on CRTC1. commit b6ba53ee1cbda5cb91e43ae6b84e1737ebc367f4 Author: Jeremy Kolb Date: Sun Jan 6 12:58:45 2008 -0500 port "G80: Reduce load detection false positives." from xf86-video-nv commit d483ecd851bb84014f1fa86f0ada4540b3ad801e commit b4fa9b060308090f4bab3f8716f7e2ecaf234be7 Author: Jeremy Kolb Date: Sun Jan 6 12:51:55 2008 -0500 Return NULL not FALSE. commit 47dde8283bcc409037682098cad8f0a0c3f9e3b2 Author: Pierre Ossman Date: Sun Jan 6 17:48:48 2008 +0100 Xv: don't wait for vblank on CRTC0 _and_ CRTC1 commit c8a9b4d0c2145d13c74a00cb24e1310e247a2718 Author: Pierre Ossman Date: Sun Jan 6 17:45:37 2008 +0100 wait for correct crtc vsync Make sure we wait for the CRTC that will actually display the video frame we're blitting. commit 25fac91dfc7072405243d2891f9ec1f0cef49a3d Author: Stephane Marchesin Date: Sun Jan 6 16:20:43 2008 +0100 Switch to latest nouveau_class.h commit bc82548514daab356fe0e0ee942ff75018fbc383 Author: Stephane Marchesin Date: Sun Jan 6 16:19:50 2008 +0100 Use clamp to edge to avoid the thin black lines. commit acaf2501d17030097761b3f8d369f14874261f6d Author: Maarten Maathuis Date: Sun Jan 6 13:17:49 2008 +0100 randr12: Some fixes for NV31M. commit 7b1521c263511016538286806f54221721c60608 Author: Maarten Maathuis Date: Sat Jan 5 19:43:08 2008 +0100 randr12: NV11 and NV20 stop at CR52. commit 9441ca0fd579cb648577f107777236d22a2b1ead Author: Maarten Maathuis Date: Sat Jan 5 19:32:32 2008 +0100 randr12: NV11 and NV20 don't have CR57 and CR58. commit 8f520754459c363313353605207f7598e259ea77 Author: Maarten Maathuis Date: Sat Jan 5 18:54:14 2008 +0100 Add a missing function to NOUVEAU_MODESET_TRACE. commit d5bf6a0149dc80d5d76ab8857415d20b42188656 Author: Stuart Bennett Date: Sat Jan 5 17:22:33 2008 +0000 randr12: fix PLL breaking typo commit 21f02f5f248bce5f49e0fcee5594c83155b532b5 Author: Stuart Bennett Date: Sat Jan 5 03:10:21 2008 +0000 Need to restore EXTDEV before fp table parsing and a nice hwsq message commit 4e4819f6722a120d3a22eb9829fe73780a0c1087 Author: Maarten Maathuis Date: Sat Jan 5 18:30:08 2008 +0100 randr12: Minor cleanup. commit 758e1cf14d61725d69ca75b7268f22e11a32a5ba Author: Maarten Maathuis Date: Sat Jan 5 18:27:10 2008 +0100 randr12: Fix a thinko, generate data before restoring, not at the beginning. commit f315ce85fa33894767892772c1e7a0966150ba6b Author: Maarten Maathuis Date: Sat Jan 5 14:05:55 2008 +0100 randr12: Some NV40 quirks. commit 08b1a2622708be9f714c27fd31372e28277d7c5b Author: Maarten Maathuis Date: Sat Jan 5 13:03:40 2008 +0100 randr12: Small quirk for NV11. commit 73101dbcfac68688914e2a3f46008ddd69d0dc62 Author: Stuart Bennett Date: Sat Jan 5 02:46:52 2008 +0000 Load NV17 power sequencing microcode commit 2ec6cf4a6e59b861c1672471872fc4ca75cfce27 Author: Maarten Maathuis Date: Sat Jan 5 01:22:42 2008 +0100 randr12: Cache dpms calls to avoid unnecesary blinking and such. commit 4f586ac169e7acd0e0bfa4139bd74e718d7598e2 Author: Arthur Huillet Date: Sat Jan 5 01:38:52 2008 +0200 Xv: linearly interpolate U and V samples when converting from YV12 to YUY2 Original patch by ahuillet, pq did one last fix. commit 5f1a721070c20a4287d17a8797ef1aabe19a579e Author: Stuart Bennett Date: Fri Jan 4 17:59:54 2008 +0000 More DCB 1.4/1.5 fixes commit 95485646eb61aaa84a8d3b6550552ab417be8e16 Author: Maarten Maathuis Date: Fri Jan 4 18:33:31 2008 +0100 randr12: Forgot to switch to vpll1_a, etc in one place. commit c43bb4990f44d50d91785f8c9860e27c3357c0a3 Author: Stuart Bennett Date: Fri Jan 4 01:14:40 2008 +0000 randr12: fix BIT bios LVDS dpms commit 5d9e42bbb5878f52cf7fd6d96532253d868b2a27 Author: Stuart Bennett Date: Fri Jan 4 13:26:38 2008 +0000 Better guesses at DCB 1.4/1.5 commit 73da4362cd7e2918bd066528a9feb157019edb5e Author: Maarten Maathuis Date: Fri Jan 4 11:54:27 2008 +0100 randr12: Remove unnecesary usleep. - I once did this thinking it may solve a bug, but it turned out a palette issue. commit c35f532fdee6141e871560f8a400040c2eb82891 Author: Maarten Maathuis Date: Fri Jan 4 11:43:16 2008 +0100 NV40: Improve VCO2 detection in nvGetClocks() commit 9369c03788d7410150044d2e3a83f0bcb876af42 Author: Maarten Maathuis Date: Fri Jan 4 11:39:37 2008 +0100 randr12: Improve speed of mode switches. - Moved pre-NV40 to vpll1_a, vpll1_b, etc variables. - Only change vpll when needed. commit ec0fe5294fdee7a79e9bfaf21ea42cfd3f52ec75 Author: Stuart Bennett Date: Fri Jan 4 02:31:21 2008 +0000 Infra for regs stored over suspend and a sort-of implementation of INIT_COMPUTE_MEM commit 786e133577edcf91b519dd8dede24250ec6fdcdf Author: Stuart Bennett Date: Thu Jan 3 02:21:35 2008 +0000 Better bios register read function prototype and a ReadEXTDEV purge commit 951f95ba553f301de2a540ef67cf04152cf72344 Author: Stuart Bennett Date: Thu Jan 3 00:52:12 2008 +0000 Refactor NVParseBios commit 8b82f24e8a526c0483120f9979a3b9370da3b297 Author: Stuart Bennett Date: Thu Jan 3 00:32:24 2008 +0000 Revert "randr12: LVDS screens use center mode." This reverts commit 1142b15c56a0b9dab8b929fc196a825815457374. They don't, necessarily commit 876d7b11c754dd66738bcc26a07e0d707b81a99f Author: Maarten Maathuis Date: Thu Jan 3 22:45:00 2008 +0100 NV40: Handle broken VCO2 values. commit 50eb355e44017a51636c1e7b8c140bc3b48813af Author: Maarten Maathuis Date: Thu Jan 3 22:36:04 2008 +0100 NV40: A better attempt at YUV->RGB conversion constants + remove vblank sync. - The vblank sync didn't work (made things worse actually), someone needs to reverse engineer the proper way for the 3d engine. commit 22cdfec9b0858c186ab37ff649248a2a825eb513 Author: Maarten Maathuis Date: Thu Jan 3 12:48:43 2008 +0100 NV40: Avoid a SIGFPE in nvGetClocks() when VCO2 is off. commit 5015e2cefcf2069ee0054c433c7ae1763d2f95eb Author: Maarten Maathuis Date: Thu Jan 3 12:37:10 2008 +0100 NV40: Actually add the attribute. commit 5b60344486c8ddeba28204dbea6f1e11704cf951 Author: Stephane Marchesin Date: Thu Jan 3 05:56:23 2008 +0100 Add a vsync attribute to nv40 textured video. It doesn't seem to work so well for me (6200), I suspect because my card is too slow anyway. Possible solutions include reducing the shader instruction # and forcing drawing from top to bottom. commit f7a356250a9a0cb214dbb1871696506c04fab946 Author: Maarten Maathuis Date: Wed Jan 2 12:15:21 2008 +0100 randr12: Hopefully these were all the function declerations that need updating. - Build didn't even show any errors, so that's why i missed them. commit aa608573918ed34968054fb5a93b11fc5c8aa1a8 Author: Maarten Maathuis Date: Wed Jan 2 00:19:12 2008 +0100 Dump some CARD8 and CARD32 in favour of uintX_t. commit b44af66323b622411d5ffa4a78a210abefdb1e2f Author: Maarten Maathuis Date: Tue Jan 1 22:38:34 2008 +0100 randr12: Small fix for NV4x cards with strange output routes. commit 6b204d5ef3ff61236ef86eadcc7e137a722aef72 Author: Maarten Maathuis Date: Tue Jan 1 21:51:31 2008 +0100 randr12: Improve output routing. commit 64546788e50ae3482698f9505869e06f727c0e50 Author: Maarten Maathuis Date: Tue Jan 1 21:07:36 2008 +0100 randr12: Really hopefully fix output resource conflict detection now. commit b71a52ae9689026e7d96d2c2752eec0c9854b0c3 Author: Maarten Maathuis Date: Tue Jan 1 20:38:24 2008 +0100 randr12: Hopefully fix output resource conflict detection. commit 9ac2765ff13f82746991c85649f45c13475e851d Author: Maarten Maathuis Date: Tue Jan 1 19:41:07 2008 +0100 Forgot the NOUVEAU_EXA_PIXMAPS case. commit 53565fc3e854abc0b5e7614da251024a60ddfd9e Author: Maarten Maathuis Date: Tue Jan 1 19:36:28 2008 +0100 Stricter alignment requirements for offscreenBase, to avoid framebuffer corruption. commit c9cedb7712719e3dee450245ebeb7db18cd40770 Author: Maarten Maathuis Date: Tue Jan 1 17:36:20 2008 +0100 randr12: Time to let some routing insanity die. - I hope nvidia is just crazy sometimes when setting CRTC1 on both output resources. - Or is it some override used on load detect? commit 9f1494ebf1ec1957f362366ab4abfc31c2227092 Author: Maarten Maathuis Date: Tue Jan 1 17:28:41 2008 +0100 randr12: Always set bit 26 in FP_DEBUG_0 if already present. commit a83c71244b8dceb2cac968195777636bd5c5f639 Author: Maarten Maathuis Date: Tue Jan 1 17:01:19 2008 +0100 randr12: Minor tweak to flatpanel regs. commit afe174cd57173c22011718d3290db5d939076ce6 Author: Maarten Maathuis Date: Tue Jan 1 14:25:54 2008 +0100 Remove AM_MAINTAINER_MODE from configure.ac commit ef1a613e0f1a1fffb54a50fd14c10069ed9eac4b Author: Maarten Maathuis Date: Mon Dec 31 20:24:41 2007 +0100 randr12: Fix a bug in an earlier commit. - This was local, but i wanted to keep some things seperate. commit 01313e3dca07615fa3b2a2b09f6d45489cc8d796 Author: Maarten Maathuis Date: Mon Dec 31 20:15:03 2007 +0100 randr12: Add a quirk for 1280x800 panel on 7300go. commit 78bf63088d5805762460801c59f0f93b43ec96a7 Author: Maarten Maathuis Date: Mon Dec 31 19:34:33 2007 +0100 randr12: Duplicate native modes, to avoid hitting an invalid pointer. commit 83aad1ad6e8a5e48ee91346b4c7037ddc8456a73 Author: Maarten Maathuis Date: Mon Dec 31 18:57:49 2007 +0100 randr12: Set the adjusted mode in output code. commit 23d9d05a68cf2eac51b500d48d31bae3b484bbe8 Author: Maarten Maathuis Date: Mon Dec 31 14:20:41 2007 +0100 randr12: Deal with output resources in a better way. commit a4d973214158efa90a1a23da879630d2b53ad313 Author: Maarten Maathuis Date: Mon Dec 31 13:18:16 2007 +0100 randr12: Disable crt load detect again, until i know how to use it properly. commit 59346e15688186b4c00ee5348413ae8312ac9a30 Author: Maarten Maathuis Date: Sun Dec 30 19:11:48 2007 +0100 Revert "randr12: Another quirk for the 7300GO cards." This reverts commit 2f2921e6c33a0482f1f30e4d69820dc82690310a. It was unneeded after all. commit 2f2921e6c33a0482f1f30e4d69820dc82690310a Author: Maarten Maathuis Date: Sun Dec 30 17:07:13 2007 +0100 randr12: Another quirk for the 7300GO cards. commit 642baacdab215f6ad7173757b8d7e00f716d6dbe Author: Maarten Maathuis Date: Sun Dec 30 16:44:23 2007 +0100 randr12: Fix quad view on some 7300GO cards. commit df36ea6d9cd2a22b51a0386bbfe4f6760ed98d97 Author: Maarten Maathuis Date: Sun Dec 30 15:01:47 2007 +0100 Add some headerfiles to Makefile.am commit 29851581b2420570efd4fca2d4f6a19113ea6947 Author: Maarten Maathuis Date: Sun Dec 30 13:35:47 2007 +0100 randr12: Only load palette's when appropriate. commit c3359789d8f288b403aedbcdfa8795391fc13deb Author: Maarten Maathuis Date: Sun Dec 30 13:26:24 2007 +0100 NV40: Some fixes and safeties. commit a7c922b58af72e3a48b92056c2fcd71c22477c2f Author: Maarten Maathuis Date: Sun Dec 30 11:36:10 2007 +0100 NV40: Fix some comments and a damage reporting bug. commit b6cd9964f9c20c40eda4283a93e447601beddb96 Author: Maarten Maathuis Date: Sun Dec 30 02:31:38 2007 +0100 NV40: Fix the texturing adapter when not rendering offscreen. - Now clip restrictions are taken into consideration. - No streching when clipping the border of a screen anymore. commit 4efdd1b560228aae7f6f5d6ed549461f8fdf0d00 Author: Maarten Maathuis Date: Sat Dec 29 22:34:32 2007 +0100 randr12: Minor cleanup. commit 5af6cd8dea2fa51b362201961191a17b0813fa30 Author: Maarten Maathuis Date: Sat Dec 29 21:46:18 2007 +0100 randr12: Micro cleanup. commit 4f91dd1ae08d348b2977b8ede0908ddd2add1cb4 Author: Maarten Maathuis Date: Sat Dec 29 20:15:25 2007 +0100 NVBIOS: Fix duallink detection. commit 559a640a542a89238afba02a78cf6399b8edc2fc Author: Maarten Maathuis Date: Sat Dec 29 19:34:12 2007 +0100 randr12: Do not remove bits in DITHER register. commit 4b980011742a6fa0bdeb312a7b367d6a56675fcb Author: Maarten Maathuis Date: Sat Dec 29 15:09:08 2007 +0100 NV40: Preliminary Xv Texture Adaptor. - Only supports YV12. - No attributes. - Only the Y component gets bilinear filtering. - You're welcome to improve it ;-) commit a4c045aa14b471fcfe29cdbec0cd846b2c5be2a6 Author: Peter Winters Date: Sat Dec 29 12:15:40 2007 +0100 Xv: nforce2 overlay can do yv12 commit 1142b15c56a0b9dab8b929fc196a825815457374 Author: Maarten Maathuis Date: Fri Dec 28 21:11:49 2007 +0100 randr12: LVDS screens use center mode. commit 8215e970445d99312c8fbbc31944466f70f3e186 Author: Andrew Randrianasulu Date: Fri Dec 28 20:59:17 2007 +0100 Fix IFC on nv4/5 by using the proper object. commit c6ae3fe9f077c38f616fd3ad07275dcf8cb65bbc Author: Stuart Bennett Date: Fri Dec 28 05:45:10 2007 +0000 Make sure the CRTC regs are unlocked when doing bios init Also make the single stage PLL setter less exploitable by the whims of the hardware commit bb467ad4baa30fc46a05bbce11d6aa4e3dac89a9 Author: Stuart Bennett Date: Fri Dec 28 08:23:17 2007 +0000 Fake a DCB entry for cards from the pre-DCB era Also, drop DCB parsing to after the init parsing again; having an INIT_TMDS in the init scripts would be silly (the original reason for bumping DCB parsing up) commit cdfe4fef9bd7d0a69d9ba19a79779152baef5489 Author: Stuart Bennett Date: Fri Dec 28 07:36:58 2007 +0000 Madness. Bios parsing for NV05 commit d4e12684617d07dc859e774f3eff9b944bc4c89e Author: Maarten Maathuis Date: Thu Dec 27 20:01:14 2007 +0100 randr12: Minor change. commit ab9dc943dd5036545e3396868c0332e9b727f31f Author: Maarten Maathuis Date: Thu Dec 27 19:50:27 2007 +0100 randr12: Small change. commit 66d343ad594f1019eb8f4d23308eb97dd1bbd94d Author: Maarten Maathuis Date: Thu Dec 27 19:25:00 2007 +0100 randr12: minor change commit cf5072e0c1d9dbee4f22ef190082aa1db78dbb0d Author: Maarten Maathuis Date: Thu Dec 27 18:37:30 2007 +0100 randr12: Undo unnecesary change. commit a7c643601073de5a63f718ddcdc9eb9f4873f2bf Author: Maarten Maathuis Date: Thu Dec 27 17:12:17 2007 +0100 randr12: Some fixes for LVDS and an odd 7300GO. commit b17f8ca13604f6b085b474219843d740b31918e8 Author: Stuart Bennett Date: Wed Dec 26 20:59:36 2007 +0000 Fix the setPLL optional code to actually compile commit 05b929ccd9c7cd50f60b84bf987fc5220ee83008 Author: Stuart Bennett Date: Wed Dec 26 21:31:16 2007 +0000 Fix up getting modes for both TMDS and LVDS commit ba92eb062aaf8b53c7f5c1858eb876d3e2f9af3e Author: Maarten Maathuis Date: Wed Dec 26 21:21:15 2007 +0100 randr12: Remove this, since it's not essential and can cause problems. commit 150d3362c5e0a300c35f1cb0c7da0c98cdd6b4de Author: Maarten Maathuis Date: Wed Dec 26 20:28:43 2007 +0100 randr12: Some minor fixed for dvi and lvds. commit 076f86dd67d6e69bd7d2a68f372cbbbb35785a3c Author: Maarten Maathuis Date: Wed Dec 26 14:51:49 2007 +0100 randr12: Some assorted fixes for dual link outputs. commit 15489f15e8130ffa74bee7c52086857554a69a6b Author: Stuart Bennett Date: Tue Dec 25 23:50:10 2007 +0000 On some chips we shouldn't look for a 27MHz crystal commit 8fa59a6231e293a8b669517bf11d35dcc15f34ff Author: Maarten Maathuis Date: Tue Dec 25 19:43:47 2007 +0100 randr12: Some minor changes. commit 80bdd0e9f4bd664e217722a57a8db79d56968194 Author: Maarten Maathuis Date: Tue Dec 25 11:58:09 2007 +0100 randr12: Restrict TMDS register backup to digital outputs. - The NV11 locks up when accessing these and it has no digital outputs, so ;-) commit 5574e96ae3e4608b95b5dd5f2dd5676386bb4ffe Author: Stuart Bennett Date: Mon Dec 24 22:33:21 2007 +0000 Better (not complete) support for DCB1.1/1.2 commit 2aa6a2347fa283dbd9b10258053602bd37880c26 Author: Stuart Bennett Date: Mon Dec 24 21:00:31 2007 +0000 "type1" bios scripts commit e0c5f65c0091b0f0ac7715fbf399a1f6fea60dc6 Author: Stuart Bennett Date: Mon Dec 24 18:13:47 2007 +0000 Rehabilitate BMP parsing commit 2789da6b34c9495a88d349b73cb838f4626fb152 Author: Stuart Bennett Date: Mon Dec 24 22:51:51 2007 +0000 Fix BMP length for version 5.16 commit 93448b40627d35fd89d8fd5629dae5f487082f43 Author: Stuart Bennett Date: Mon Dec 24 15:23:41 2007 +0000 randr12: use new PLL code on < NV40 commit 3d90e13c26e389f43253a5d2721ce93a628f156d Author: Stuart Bennett Date: Mon Dec 24 05:31:37 2007 +0000 setPLL routines for single and two stage plls plus some parsing fixes commit eabe1e02504467c7425e62673a68f4dc7dd83aab Author: Stuart Bennett Date: Mon Dec 24 02:29:54 2007 +0000 Two stage nv31 PLL mnp calculator and bios limits parsing commit c8b50bc7dc146e01ceb67ee3633b5c8958345fcd Author: Stuart Bennett Date: Sun Dec 23 03:09:26 2007 +0000 Single stage PLL MNP calculator commit 935f809f4bc9dab37e80c59007141d7b9ad673a8 Author: Maarten Maathuis Date: Sun Dec 23 22:05:10 2007 +0100 Reading into the -1 index of array is a bad idea, this should fix subchannel binding issues. commit f357ba26b7f560cdec03f70a2c3b79b65d70172b Author: Maarten Maathuis Date: Sun Dec 23 15:48:24 2007 +0100 randr12: Add modesetting debug option (showing most register writes). - Contribution of ahuillet. commit 95387b9779173cc3e51a7fffc60975010253b391 Author: Maarten Maathuis Date: Sun Dec 23 15:35:24 2007 +0100 NV40EXA: Suport as many of the repeat types as possible + random cleanup(randr12). commit 8e80f7504f872166a811b6ce20663aaf97fd7181 Author: Arthur Huillet Date: Sun Dec 23 13:49:24 2007 +0100 Revert a manifestation of my stupidity. commit 86a3a5fd40005a4860d11f0046444ba232db0d1e Author: Arthur Huillet Date: Sun Dec 23 13:47:08 2007 +0100 Xv: added an attribute to change the CRTC the overlay is on - useful with dualhead commit 129f04612cdbad756dbe213f77d2f12482b62c47 Author: Arthur Huillet Date: Sat Dec 22 23:32:36 2007 +0100 aunedsoau~ commit 26953280895698a94c1f886cf7c0d2f389167b1c Author: Maarten Maathuis Date: Sun Dec 23 11:46:17 2007 +0100 randr12: Extend dvi clock setting to nv3x also + cleanup. commit 063d1bd50b164e841e3a725b75ac243827aede50 Author: Maarten Maathuis Date: Sun Dec 23 10:46:09 2007 +0100 randr12: Actually load the default palette on modeset. commit 810901abb50a54b8f1eb73029f5348805a53531d Author: Maarten Maathuis Date: Sun Dec 23 10:32:18 2007 +0100 randr12: Simplify code a bit by using the more sane code for NV4x also. commit eb8441824ea2a7dc9a1d7a86a4305fae48a26a0c Author: Maarten Maathuis Date: Sun Dec 23 01:02:18 2007 +0100 Minor cleanup. commit 9eef67abaf8beb9ed93eb5b6ecdf7255f5b06e4c Author: Maarten Maathuis Date: Sun Dec 23 00:18:13 2007 +0100 randr12: Always enable MPLL and NVPLL on pre-NV41 cards. commit fa82eb23904bc1a70209778d2aafcbef6f27cfa1 Author: Maarten Maathuis Date: Sun Dec 23 00:07:01 2007 +0100 randr12: Put all NV2x (and lower) cards into restricted mode. commit 33d7313d7dca4501b7ac2e078a87f3a10ab1b501 Author: Maarten Maathuis Date: Sat Dec 22 23:55:19 2007 +0100 randr12: Correct CRTC selection for pre-NV40 cards. commit d6243fe33223d2934c1a819c58bb58b9ebe15c25 Author: Maarten Maathuis Date: Sat Dec 22 23:39:51 2007 +0100 randr12: Older cards do have switchable CRTC's, i was again wrong. commit a96553544ae76239e394ab34b33814a57252d31c Author: Maarten Maathuis Date: Sat Dec 22 23:23:32 2007 +0100 randr12: Minor cleanup to clock writing for pre-NV40 cards. commit 9b63f42bc7da9b45ecbbd32ef81cae55d2a57220 Author: Arthur Huillet Date: Sat Dec 22 18:44:36 2007 +0100 NV10EXA: disabled A8+A8 hack on PPC machines. commit 308fec0111d182fb17536ff2746e079c1af32d5a Author: Ben Skeggs Date: Sat Dec 22 23:18:57 2007 +1100 nv30: Don't touch 0x1a4 commit 62916c7c8c6b467436e4628c7d08a4e0de1f21f3 Author: Maarten Maathuis Date: Sat Dec 22 12:02:31 2007 +0100 randr12: Undo a previous mistake for NV1x/NV2x. commit 6f9eb17637e2c3b8d49a07332f06a0102d2fe016 Author: Stuart Bennett Date: Fri Dec 21 20:18:57 2007 +0000 Beginnings of PLL limits for BMP bioses And the correct way to use the debug_modes member (still disabled for now) commit 73bc1db51e9431c5d1861d8a753aa8dd9abe9882 Author: Stuart Bennett Date: Sat Dec 22 02:57:40 2007 +0000 On demand PLL limit parsing for NV40 Also some nv_bios message formatting consistancy pedantry commit 57b5dbfac4ff65258e94a17717815156a54cc273 Author: Maarten Maathuis Date: Sat Dec 22 01:29:38 2007 +0100 NV40EXA: Fix borders of images that recieved arbitrary rotation. - It was getting padded up to the edge of the smallest square that could fit the rotated image. - Please try this on older cards as well. commit 39ef0262be49feaed2ba1c6caeab5134181ef0b9 Author: Maarten Maathuis Date: Sat Dec 22 01:26:44 2007 +0100 randr12: Some minor changes. commit 371b6e52bef743f14003e3783ea3dbc624050594 Author: Maarten Maathuis Date: Fri Dec 21 18:02:33 2007 +0100 randr12: Restrict crtc's on nv1x and nv2x. - If anyone can show a mmio-trace showing how to switch them (on cards that have a dcb table claiming they can), then please do. commit c2ac36927827439cf92ff53bbf2414bfaf45a6c3 Author: Maarten Maathuis Date: Fri Dec 21 00:05:37 2007 +0100 randr12: Detect switchable crtc's. commit 33a2cf86520f42d549d03c46134b818e90d51a41 Author: Peter Winters Date: Thu Dec 20 23:47:11 2007 +0100 NV10 EXA: use the proper variable commit 71c3b8ff09adcadb10d800d3696762c3ef58be3d Author: Stuart Bennett Date: Wed Dec 19 23:44:23 2007 +0000 randr12: fix old and new style scaling for LVDS, and use an enum for the scaling modes commit 6bea51b905a461b4e703053082ab45d9e1074112 Author: Maarten Maathuis Date: Thu Dec 20 00:07:29 2007 +0100 randr12: Fix secondary (analog) outputs with ffs(or) != 3 on some NV4x cards. commit e0e04eb561f4b127087befc4e3768d8012a82bfe Author: Stuart Bennett Date: Wed Dec 19 04:29:34 2007 +0000 randr12: don't use invalid crtc pointer for lvds dpms commit a68ac97979a3178ea210470bbb3d64f3f9325e07 Author: Stuart Bennett Date: Wed Dec 19 04:00:14 2007 +0000 Turn LVDS panel on again, after resetting, when we're shutting down commit 1cfd659fc1a5f8dc2faba30155fca46bbe31f3a4 Author: Stuart Bennett Date: Tue Dec 18 16:58:57 2007 +0000 minor link_head_and_output cleanup commit 73a5d09b3e9fa13d04d6696ca38b6402795acc3d Author: Stuart Bennett Date: Wed Dec 19 03:43:59 2007 +0000 randr12: CRTC member wasn't big enough yay gcc. could fix console colour breakage commit 0a63198195f689b32388b99db61ff942e16fbc6d Author: Maarten Maathuis Date: Tue Dec 18 23:43:57 2007 +0100 NV10EXA: Fix rotation. commit f2a3807618cfeac01d7d28e9d76721ad61495f7f Author: Maarten Maathuis Date: Tue Dec 18 23:23:59 2007 +0100 randr12: Add new pll calculation for NV4x cards (from NVClock). commit ebf96c8b651d1a0dc5a19576cf57efcbb989d51d Author: Maarten Maathuis Date: Tue Dec 18 21:38:56 2007 +0100 randr12: Hopefully reset outputs better between mode changes. commit 5561f54f94e0b31394ecaf461e342470ca768281 Author: Maarten Maathuis Date: Tue Dec 18 07:49:05 2007 +0100 Some endianess fixes for the previously committed code. commit 11104ce9c499f30fb5abe29e65ba1918c563f4cd Author: Stephane Marchesin Date: Tue Dec 18 07:40:08 2007 +0100 Rename NV30TCL to NV34TCL in preparation of the merge. commit 3e6ea7af5b1613b9e0e4e29433910d998def4899 Author: Maarten Maathuis Date: Mon Dec 17 23:13:12 2007 +0100 Add pll limits bios detection code for NV4x cards. commit e56512d283f401a75c6165a0a02e8a7a5905fc4c Author: Maarten Maathuis Date: Mon Dec 17 19:45:42 2007 +0100 NV30EXA: Fix rotation. commit 9ff42b7a83da06a3b9b7076ca888ea447eaa16f3 Author: Maarten Maathuis Date: Mon Dec 17 17:50:11 2007 +0100 randr12: Fixes + cleanup. commit 83d99bc3e15d11d3476c54053f6bbe0ff9f094be Author: Maarten Maathuis Date: Mon Dec 17 00:40:24 2007 +0100 randr12: Some minor additions. commit 60c08fedec09dd036b4165881b3001874c4d349a Author: Stephane Marchesin Date: Sun Dec 16 23:40:26 2007 +0100 nv30 exa : handle gart case. commit efb10849e136048c9c49cce752d4073402219c8c Author: Maarten Maathuis Date: Sun Dec 16 23:26:49 2007 +0100 Undo a bad change. commit dd8beae22883822ea3555122b0352d4e3cf2c266 Author: Maarten Maathuis Date: Sun Dec 16 21:35:18 2007 +0100 A little more cleanup. commit 9ff169ba62dcb2eeefce8a020060de731b72cb6c Author: Maarten Maathuis Date: Sun Dec 16 21:32:32 2007 +0100 More cleanup. commit 2d03d4b999a0a37ff2f9fc87bd20959e789a152e Author: Maarten Maathuis Date: Sun Dec 16 21:19:13 2007 +0100 randr12: Select sensible maximum heights and widths for framebuffer. commit 0e47fa4ab35e529926f5fdccdb44f9d054e1886a Author: Maarten Maathuis Date: Sun Dec 16 21:17:04 2007 +0100 Some reindenting and minor cleanups. commit 2be4b73905dc5767e593bf4753607144fe67b7ab Author: Maarten Maathuis Date: Sun Dec 16 20:24:26 2007 +0100 Some cleanup. commit 6686adbea6638dd894eef697e044941352cc2674 Author: Maarten Maathuis Date: Sun Dec 16 20:10:39 2007 +0100 randr12: Restrict digital outputs with ffs(or) == 1 to primary crtc + some cleanup. - The blob does not reveal what the possible way is (so far), and guesswork didn't help either. commit 9c8da2f562ca571a56224d7fecfb9c3fb3a8d3d0 Author: Stephane Marchesin Date: Sun Dec 16 18:33:52 2007 +0100 And of course, that comment has to go. commit 02e9a01064930a7081569474be9544fa1f20a0f4 Author: Stephane Marchesin Date: Sun Dec 16 18:32:52 2007 +0100 Fix front buffer alignment issues (probably needed because of tiling). commit b134b8f1552066b6706521a9c85d9241d99b3ccc Author: Maarten Maathuis Date: Sun Dec 16 13:41:38 2007 +0100 randr12: Fix compile error. commit ec0c9cc9a69c00a960e98e796c00d35d996cf360 Author: Maarten Maathuis Date: Sun Dec 16 12:55:29 2007 +0100 randr12: A minor change to OUTPUT register. commit 69782a57285bb53df73c7f45d29882ae990f6054 Author: Patrice Mandin Date: Sun Dec 16 00:09:45 2007 +0100 nv30: revert viewport clip change, add myself, add unknown command 0x2b8 (viewport origin?) commit 8dd238f3ce2e7eda7151670ea50a75abdb5b6217 Author: Maarten Maathuis Date: Sat Dec 15 23:49:58 2007 +0100 randr12: Some minor changes. commit 08a50d878571f7c9579b15703a08c5c2e52924f2 Author: Maarten Maathuis Date: Sat Dec 15 19:54:11 2007 +0100 randr12: Some minor reordering of load/save functions + don't lock crtc when leaving X. commit 4209c06800b97e98261c10e0a625a68f7352ca6c Author: Maarten Maathuis Date: Sat Dec 15 14:48:04 2007 +0100 randr12: Also support accelerated rotation on older xservers (and default code). commit 933b40a2911486a78615984fb17e9ea6757fcfa8 Author: Ben Skeggs Date: Sat Dec 15 23:45:03 2007 +1100 nv40: just for stillunknown... Fix composite transforms. This is getting to a point where it may be worth doing this in the vertex program perhaps. For now this works, it'll do. commit 7212400070264e27a53564a7f5fc95dd9c99f82e Author: Ben Skeggs Date: Sat Dec 15 23:44:17 2007 +1100 Revert "NV40EXA: Fix composite rotation." This reverts commit 02d4538b207cf318737ace8c3d8ded76a54e5886. commit 1a622257687a499ce74b3a608618d957784dccf0 Author: Maarten Maathuis Date: Sat Dec 15 13:39:23 2007 +0100 randr12: Avoid a vram/memory leak in some situations. commit 02d4538b207cf318737ace8c3d8ded76a54e5886 Author: Maarten Maathuis Date: Sat Dec 15 13:35:45 2007 +0100 NV40EXA: Fix composite rotation. - Thanks to radeon (code) for inspiration and sample code. commit 77d47348dfe01ca3972a683d3c478f910301e910 Author: Maarten Maathuis Date: Sat Dec 15 12:02:36 2007 +0100 randr12: Add some ifdefs to make compiling better on older xserver's. commit dbb7efc5b81145ad61d8e6502b918a1e909f3cfb Author: Patrice Mandin Date: Sat Dec 15 11:44:41 2007 +0100 nv30: need to set default depth range commit 6ab8436df0865aab7f12cd8527600353b4728df7 Author: Patrice Mandin Date: Sat Dec 15 11:37:43 2007 +0100 nv30: only 4 texture units commit 370d97a4578528c26c3f1546a9522caf40aa2151 Author: Patrice Mandin Date: Sat Dec 15 10:34:26 2007 +0100 nv30: fix viewport clipping setup commit 1190d046dfef97dea04d72c993f90a203d81d86c Author: Maarten Maathuis Date: Sat Dec 15 02:00:37 2007 +0100 Fix ugly bug, that creeped in. commit 46f321985ae57457edcedf80f82ba23e53fd9cd2 Author: Maarten Maathuis Date: Sat Dec 15 01:44:59 2007 +0100 randr12: Preliminary support for rotation, please read notes. - It works, but it's slow, very slow. - Each crtc now has it's own cursor map (fixing the bugs you never saw). - Dualhead with one rotated screen may show minor corruption near the edges, this will disappear if a fullscreen app opens. - Acceleration is possible, BUT: - You need a git xserver. - Enable NOUVEAU_EXA_PIXMAPS in nouveau_local.h (*maybe* this will become default later). - Most of the composite hooks have problems with rotation, so left and right won't work, inverted does work for me. - So please fix the composite hooks ;-) - Note: No slowdowns are apparent when using proper acceleration. commit 0f07ec3f681b87762fbe7328f47b628711bdbaae Author: Stuart Bennett Date: Sat Dec 15 00:01:04 2007 +0000 Better restoration of TMDS regs on LVDS commit 34be8d005fbfff373cc7fe52ca2fa09f5cd5ef00 Author: Maarten Maathuis Date: Fri Dec 14 23:58:25 2007 +0100 randr12: It seems DVI on laptops wasn't as strange after all. commit d678932a1818c5c5f4a4f3097992626ce9620c0e Author: Stuart Bennett Date: Fri Dec 14 01:40:14 2007 +0000 INIT_COPY_NV_REG and INIT_ZM_REG_GROUP_ADDRESS_LATCHED opcodes Thanks to RAOF for the trace with these Also correct a doc oopsy, and a bunch of superfluous whitespace commit 5b9dc7ad6eb2f16c4ca8824c2a19c85bbe1f126b Author: Stuart Bennett Date: Thu Dec 13 02:30:06 2007 +0000 Invert 18/24 bit interface flag, so that the default value (0) does not imply dithering commit a878b48c7b5a7aafac9aaa877706f5c7570d3057 Author: Stuart Bennett Date: Thu Dec 13 01:10:12 2007 +0000 Name bios opcodes 4f and 50 Extrapolating from previous names, but marked as non canon commit 59b46e695b1badbf00f2c03bc78ce58f6887f739 Author: Stuart Bennett Date: Thu Dec 13 01:03:25 2007 +0000 Get VBIOS from pNv (cleanup) And a minor doc correction commit c45930d0819668612547b4bf1a8d963bfe40e38b Author: Stuart Bennett Date: Thu Dec 13 00:15:21 2007 +0000 randr12: some minor cleanups, no actual changes commit 0e778b2d6970f761e5c0b5eb3054e91589cd2c46 Author: Stuart Bennett Date: Wed Dec 12 15:54:53 2007 +0000 Setup NV40 LVDS commit 269d83cb46075971fd8dc5402589d4da3eb92f22 Author: Stuart Bennett Date: Wed Dec 12 22:20:43 2007 +0000 INIT_CONDITION_TIME opcode commit 9b3072d195aae383a993526b3ab4350c500c4277 Author: Stuart Bennett Date: Wed Dec 12 18:08:07 2007 +0000 Opcode 4F commit 8ee100ced7da347904e45ce590350d7b7c47ed36 Author: Stuart Bennett Date: Mon Dec 10 22:50:40 2007 +0000 FP BIOS mode finding for nv40 commit 9131985f5c359bdb077eb52649369a5c06831544 Author: Stuart Bennett Date: Tue Dec 11 23:23:27 2007 +0000 CR57 doc addition commit d595eb86790432f6a75dec3388080ffc9ffff8d8 Author: Maarten Maathuis Date: Wed Dec 12 22:12:33 2007 +0100 randr12: A good guess at the lower bits of SEL_CLK (i hope ;-)). commit c5bc9899c218abe801e024919c89896e8183dff2 Author: Maarten Maathuis Date: Wed Dec 12 21:03:34 2007 +0100 randr12: Add some extra safeties (to mode checking). commit 43c05d306485eb5d8c8b9b39f9bd4a3f470f20a2 Author: Maarten Maathuis Date: Tue Dec 11 21:43:02 2007 +0100 randr12: Enable dithering on lvds when needed. commit 8752f18d27a6095485e36f4fb20e9252f27a1e1b Author: Maarten Maathuis Date: Tue Dec 11 20:22:23 2007 +0100 randr12: SEL_CLK attempt 10, return to simplicity, which may actually work ;-) commit c1227d3a2b9194d50a4274a48fa0b3bcf2628e36 Author: Maarten Maathuis Date: Tue Dec 11 17:52:39 2007 +0100 randr12: (mostly) Some fixes for NV40. - The NV40 is a strange card, as it partially behaves like an NV3x. commit 34d9e971e17e763aee96b823bb34da54a281339c Author: Maarten Maathuis Date: Tue Dec 11 11:50:28 2007 +0100 randr12: Fix SEL_CLK assignment again (for non-mobile cards) + fix crosswiring. - I made a good guess for mobile cards, but i need more mmio-traces. - I accidentally broke crosswiring for my card (and other's like it). commit bf28c3e7d9976f8695198a938397f001de62836b Author: Maarten Maathuis Date: Tue Dec 11 00:04:38 2007 +0100 randr12: Fix an earlier mistake, make a decent guess about dvi on laptops. - I could use some more mmiotraces of LVDS+DVI on NV4x. commit 58531ebd4a72368b13d3c1564110ef85fb6d6644 Author: Maarten Maathuis Date: Mon Dec 10 20:43:26 2007 +0100 randr12: Improve crosswiring for digital outputs + minor changes and cleanup. commit a388cc5fe21f8692e8ee1dbd71caea88ea2ca5df Author: Maarten Maathuis Date: Mon Dec 10 16:31:47 2007 +0100 randr12: Properly set SEL_CLK based on bus, not based on output. - This hopefully fixes non-working dvi on some desktop cards and on mobile cards. commit f96a6c2d5d251af560de673ccd91b88b3f5cc535 Author: Maarten Maathuis Date: Mon Dec 10 00:03:49 2007 +0100 randr12: Add config option for scaling mode. commit fc73ddad60c551289629be5f2f87b39e26f42674 Author: Maarten Maathuis Date: Sun Dec 9 22:46:39 2007 +0100 randr12: Some cleanup. commit 2543e301402b04ae94f349e21a7760beb6f5bdc6 Author: Stuart Bennett Date: Sun Dec 9 19:58:58 2007 +0000 Bios register write function special case and some INIT_RESET improvements nv32_wr weirdness only seen on NV40+, but it shouldn't hurt for earlier cards commit 4408ce8d2f902a9ace550a045e6022df2936fbce Author: Stuart Bennett Date: Sun Dec 9 17:09:03 2007 +0000 Some renaming for bios indexed port io commit 9eb4294e3f5814f675b403078761bf002540868b Author: Maarten Maathuis Date: Sun Dec 9 18:36:26 2007 +0100 randr12: Revert dither value to that used by nv. commit cfbe4bdef7f4ba958462743dc28bf253b456c1a3 Author: Maarten Maathuis Date: Sun Dec 9 18:34:10 2007 +0100 randr12: Set FP_CONTROL correctly for NV2x cards. commit d5dd3afc50e491689f765a5e6a72892efa1768a3 Author: Maarten Maathuis Date: Sun Dec 9 18:10:29 2007 +0100 randr12: Improve output selection. - Now detect if there is a dual link capable output at all. - NV2x cards can't switch crtc on analog outputs. commit fda83a3c489469d2bb4cbd20c205016b19dfdec5 Author: Stuart Bennett Date: Sun Dec 9 15:51:02 2007 +0000 Additional LVDS bios script sanity check Avoids current incorrect situation when LVDS attempted on NV40 commit 9932e36d4cf2988d1f6f40d615cd247acc21b592 Author: Maarten Maathuis Date: Sun Dec 9 16:29:14 2007 +0100 randr12: Fix minor (potentional) bug. commit 3f83d7308f3b0aca4943262b0ba2715a32b61d58 Author: Maarten Maathuis Date: Sun Dec 9 12:47:55 2007 +0100 randr12: Add noscale mode and some misc fixes. commit d688008b38609814de9c5c36a2c8bc569455d14e Author: Maarten Maathuis Date: Sat Dec 8 18:34:04 2007 +0100 randr12: Fix typo, it's "our" not "or". commit 0f7c018af0c7f929c2bf8efb12c510d6e3170885 Author: Maarten Maathuis Date: Sat Dec 8 18:25:41 2007 +0100 randr12: Assuming an output always has a crtc is a bad idea. - Don't rely on current crtc to restore mode. commit 86f248cc2e52de652b1768e15a4ba62eefefd81c Author: Maarten Maathuis Date: Sat Dec 8 16:49:21 2007 +0100 randr12: Avoid that two outputs have the same clock selected. commit 6be67d6cf0f047ceba0aa62989dc44bb58b0db87 Author: Maarten Maathuis Date: Sat Dec 8 15:50:39 2007 +0100 randr12: Fix bug in sel_clk assignment. commit ddfec2f1c749af227dd57dbde6751320ba72adb6 Author: Stuart Bennett Date: Sat Dec 8 02:18:41 2007 +0000 Base error :) commit dff95ae101f701d737c8d2134181693f36178beb Author: Stuart Bennett Date: Fri Dec 7 16:56:26 2007 +0000 Fix failure to cope with varying table lengths commit 10731d2a8fe06fa8675c26d201615c2a019de737 Author: Stuart Bennett Date: Fri Dec 7 02:54:36 2007 +0000 Add a new valid register range, which appears in some gf7 biosen. NVClock has it down as appearing in gf8 as a new PBUS range, but the use on gf7 looks PBUS-like too commit b25049233613eff99d767a8ec0c97b5140088213 Author: Stuart Bennett Date: Fri Dec 7 03:13:52 2007 +0000 Some more bios script opcodes Thanks to swany for the trace for these commit 6da4923088e8ca1d2de464030205b41d893e65b3 Author: Stuart Bennett Date: Fri Dec 7 02:17:37 2007 +0000 Some M table parsing, and defer I table execution commit 7e5e477b67279e00d1e56736f9c953972bda4dec Author: Stuart Bennett Date: Fri Dec 7 00:18:25 2007 +0000 Improve TMDS04 setting for LVDS commit dd757910f546a97bb42dbefdc7eae058980b1ffb Author: Stuart Bennett Date: Thu Dec 6 23:37:01 2007 +0000 TMDS regs setting for pre NV40 commit 38f139d473eb3201daaac47e655516f65c7528a6 Author: Stuart Bennett Date: Fri Dec 7 00:30:14 2007 +0000 BIOS version parsing commit 20b6be6c95a71a41456c4c9b5cb1a1e0238f2e70 Author: Stuart Bennett Date: Thu Dec 6 17:01:49 2007 +0000 Refactor BIT specific details out of TMDS setting commit f665b4737f1fbbd5e4a6495b561e723f17bc6137 Author: Stuart Bennett Date: Thu Dec 6 18:48:04 2007 +0000 The 'T' table is clearly a TMDS table, call it as such commit 3759cc9d602b0cd933a0ee7e6f64626822383e19 Author: Stuart Bennett Date: Thu Dec 6 05:03:49 2007 +0000 The TMDS scripts being non-stubs does not seem to matter - they're still not executed Message left as log clutter in case there's some future case of broken TMDS when the scripts might be useful commit d6e896dac1018fc3190a5a5283a405c4b031d1ba Author: Stuart Bennett Date: Thu Dec 6 04:14:56 2007 +0000 Test for LVDS_INIT script commit c50bc602861fdcdb238e1a4db1146b4870ce83ba Author: Maarten Maathuis Date: Thu Dec 6 21:48:13 2007 +0100 randr12: Small fix for dual link dvi. commit 457d4217624cc73df22ca7f7db0134c54ebac5e5 Author: Maarten Maathuis Date: Thu Dec 6 19:54:08 2007 +0100 randr12: Improve compatibility with DCB 2.0 commit 5046cb7b3d1af89b72327660932ae06c06d0869b Author: Stuart Bennett Date: Thu Dec 6 02:28:18 2007 +0000 randr12: oops, lost the restore case commit 54c7bdf0f3e53c23c0853eac8243407d280d0732 Author: Stuart Bennett Date: Thu Dec 6 01:54:45 2007 +0000 randr12: a common function to fix TMDS04, and new use of LVDS_RESET Plus some corrections/additions to call_lvds_script, and a minor CR5758 doc update commit a7811521ed5d227a18af23e2e761ea39fb7eb86f Author: Maarten Maathuis Date: Thu Dec 6 00:28:49 2007 +0100 randr12: Fix comment. commit 4771b4bd4b30f404fb087dd1fd6eac0e12fdd041 Author: Maarten Maathuis Date: Thu Dec 6 00:25:48 2007 +0100 randr12: Some assorted fixes, which hopefully improve things for dual link dvi. commit 4d3a015631a0d95392daeec6fd8e99c65c83cb53 Author: Maarten Maathuis Date: Wed Dec 5 20:41:27 2007 +0100 randr12: replace preferred_ramdac with preferred_output. commit d3036012db63f9a386d4c1fd8e624b86bcd6c48b Author: Maarten Maathuis Date: Wed Dec 5 20:27:57 2007 +0100 randr12: Clean up the unneeded ramdac variable. commit 37f3a06be4f50c34ee8b82617e5e6eca3a4d66fe Author: Maarten Maathuis Date: Wed Dec 5 19:03:52 2007 +0100 randr12: Pick right register set, to correctly "guess" the bios/VT clock. - A wrong register set leads to improper detection of crosswiring, thus selecting the wrong clock. - I noticed a 0.675 MHz difference in the clock (which is vpll2), when instead i should have had vpll1. commit b843dbddbd496ab7d374018e6e05ac1a05ea00e2 Author: Maarten Maathuis Date: Wed Dec 5 16:51:04 2007 +0100 randr12: Small fix. commit 4b049994f32058e380543cee81b1d0b76658ea9e Author: Maarten Maathuis Date: Wed Dec 5 16:47:07 2007 +0100 randr12: Final commit, restore TMDS regs before everything else. - Use the bios to program the tmds registers. - Restore TMDS registers in nv_crtc_restore, to make sure it's early enough (see code note). - It works for me, let me know if it breaks stuff ;-) commit 6ff8c1cb2014c7984503a8481cf6ad4f9cccf79a Author: Maarten Maathuis Date: Wed Dec 5 07:57:30 2007 +0100 randr12: Intermediate bios based flat panel init commit, still a few minor issues. commit c14c7deb725c2509c804bae72faa584109f3021c Author: Stuart Bennett Date: Wed Dec 5 00:35:35 2007 +0000 Add LVDS script function, and use it for DPMS commit c587731d08bbfd58ff5c2725b7aed4e6477e9f7f Author: Stuart Bennett Date: Wed Dec 5 00:22:22 2007 +0000 Parse the LVDS Manufacturer table Use dual link flag to set bit 31 of FP_CONTROL properly commit affcf6988f1af2c3ea8c47a2f1fdf608cbafcfce Author: Stuart Bennett Date: Tue Dec 4 20:31:06 2007 +0000 Modify DCB dual link flag, and DCB config parsing for LVDS The DCB duallink flag indicates that this setup may be used for duallink; it obviously can't know whether a duallink display is plugged or not Use LVDS DCB conf flags for mode detection / selection commit c822e32d95a7afa3d5ef076f97179a1843835f12 Author: Maarten Maathuis Date: Tue Dec 4 09:06:10 2007 +0100 randr12: Add fullscreen gpu scaling (only as randr property for the moment). commit a1054c7bef3594f6876d611dde1cf3f0528fb8e1 Author: Stuart Bennett Date: Tue Dec 4 01:31:24 2007 +0000 Better T table stuff. Hopefully now usable by calling parse_t_table(...) at the appropriate place Also collapse some messages, and add an array bound check commit edb9a59ac8d3b16cccd52bde24606c43ae7d36d0 Author: Stuart Bennett Date: Tue Dec 4 01:40:37 2007 +0000 Allow BIOS execution at runtime, if wanted And comment the opcodes I haven't double checked yet commit 82fdd3f76e20fc7dd10fcf00cf1c13520dc56364 Author: Stuart Bennett Date: Tue Dec 4 01:31:24 2007 +0000 Make the bios a bit more accessible And make a few things use the new structure commit d15bfdfb7e09dce85c3817036ced9efa6ac4522c Author: Stuart Bennett Date: Tue Dec 4 00:00:40 2007 +0000 randr12: set CR58 appropriately for 0 and 2 cases commit 02a4f498d903bda55ac38a6e19b1beca0d128e8a Author: Stuart Bennett Date: Mon Dec 3 23:35:01 2007 +0000 randr12: refactor dcb usage heads is also a better name than head commit 1571f81210b48407c9d20264437406d79dfcc784 Author: Stuart Bennett Date: Mon Dec 3 22:26:33 2007 +0000 randr12: some code economy Only addition is adding call to output2->funcs->commit when stealing a ramdac, no logical changes otherwise AFAICT commit 7debf566c7a88258ac8ca925d7138187c1ed8755 Author: Stuart Bennett Date: Tue Dec 4 00:06:05 2007 +0000 randr12: Improve CR57 and CR58 infrastructure and saving/restoration commit dab828c4793fd7e48e4d8e1d7f8d0a2f8198484c Author: Maarten Maathuis Date: Mon Dec 3 20:55:26 2007 +0100 randr12: A few minor comment changes. commit d959d880b6808adc153c752ea01f3a228879fe3c Author: Maarten Maathuis Date: Mon Dec 3 20:51:16 2007 +0100 randr12: Select CRTC 1 on pre-nv40 (stupid bug) + initial support choosing scaling method on tmds. - The output property SCALING_MODE accepts "gpu" and "panel". - The first obeys aspect ratio, the second always goes fullscreen. - This is just a first attemp at exposing an output property. commit 1d2ea797704d55f63cfe8c00cc000a6b13f1093d Author: Maarten Maathuis Date: Sun Dec 2 22:10:56 2007 +0100 Show the id's of unknown bit structures in the bios. commit fc98b11eb70f1dd7dfb4e7c6eb8d888dfebdc890 Author: Maarten Maathuis Date: Sun Dec 2 12:02:08 2007 +0100 randr12: dual-dvi cards need extra consideration. commit 2beece0c634c2c60996da1f1d15b24c5f231a6be Author: Stuart Bennett Date: Sun Dec 2 02:33:57 2007 +0000 Fix endian bug on BIT biosen commit d3563cccb43355927d4ba355bc31a957ea01587f Author: Stuart Bennett Date: Sun Dec 2 02:14:32 2007 +0000 Don't use PROM if the checksum is bad commit 4097af9bad85031fa1da9c3d21b084f587412604 Author: Stuart Bennett Date: Sat Dec 1 09:55:31 2007 +0000 randr12: print "or" _pre_ ffs-ing commit a8faa53355f9ba3851f47e290c821bc1b9a03e85 Author: Stuart Bennett Date: Sat Dec 1 09:53:05 2007 +0000 DCB parsed flag for duallink outputs Light on proof, strong on suspicion with this one commit 63fc1ab4ce597a2f542c9bb7d3edae67551bda29 Author: Maarten Maathuis Date: Sun Dec 2 02:08:20 2007 +0100 randr12: Always use vpll2_true on nv4x cards. commit 97a99c42fb24320177bdae884c4c710c5a25ce4f Author: Maarten Maathuis Date: Sun Dec 2 01:48:01 2007 +0100 randr12: A safer bet for TMDS register 0x2e. commit 3358f4eec53845cf90b9f9ef1f2a70c231bac588 Author: Maarten Maathuis Date: Sun Dec 2 01:34:49 2007 +0100 randr12: Remove preferred_crtc, since it's probably incorrect. commit 5994655925173a4e8b1c0ac902f7dc6e1ffe64f1 Author: Maarten Maathuis Date: Sun Dec 2 01:08:41 2007 +0100 randr12: Improve output register setting. commit df8d6848a004e724e92c3dd63fc608198c7bd940 Author: Maarten Maathuis Date: Sat Dec 1 18:40:46 2007 +0100 randr12: Fix up some dvi related issues, thanks to MighMoS. commit 84cc907beffd9e8f91c4333a6db919ee2efa2740 Author: Maarten Maathuis Date: Sat Dec 1 11:23:50 2007 +0100 randr12: Fix NV4x dvi clock selection, mostly affects G70 cards. commit c97ad9850da89b1898f125df1796bb1d00d124bb Author: Stephane Marchesin Date: Sat Dec 1 10:33:53 2007 +0100 Make the debug variable global to all DMA calls. commit 19d75f5018fd442920378a5083dbe1316cc6f62b Author: Stuart Bennett Date: Sat Dec 1 06:29:09 2007 +0000 The BIT bios T table. T might stand for TMDS. Parsing once on start may not be the correct answer, could be a runtime thing commit 42a02513a1db7ada586dcaa2eaa5b68e97d67892 Author: Stuart Bennett Date: Sat Dec 1 06:37:17 2007 +0000 Bump DCB table init so that INIT_50 can use it commit ddc4b0946dfd567fc0ea136db0d8b695c28bc3a2 Author: Stuart Bennett Date: Sat Dec 1 05:35:40 2007 +0000 "or" can be 0xc Quite what this means, or why Nvidia couldn't give it a more descriptive variable name in the g80 driver is anyone's guess commit 384115a5abd756c5e3ee79ba50e15636d159f19c Author: Stuart Bennett Date: Sat Dec 1 05:17:47 2007 +0000 Use Nvidia's naming for CRTC_INDEX_COLOR commit 2344b5b5a5a0efa439ba64f92d5b08097178dd6c Author: Stuart Bennett Date: Sat Dec 1 05:16:20 2007 +0000 Bios script code 0x50 commit 78c6ff292f4c2aa0e06bce095108d0716555104a Author: Stuart Bennett Date: Sat Dec 1 00:45:32 2007 +0000 Add dcb "location" field. 0 for on-chip, !0 for external commit 312fb3d65b8db6e3c2d123c0b192da82b7de4f71 Author: Maarten Maathuis Date: Fri Nov 30 16:47:26 2007 +0100 randr12: Don't unset bit 8-15 in sel_clk. commit ea10d579fb6053e064a5cd14b47e1da55d4120ac Author: Maarten Maathuis Date: Fri Nov 30 16:43:33 2007 +0100 randr12: Big commit, read message. - Outputs (on nv4x) should now select vpll2 instead of vpll1 when neccesary (this took a long time to figure out). - Register 0x00680594 is now set correctly to allow analog secondary outputs to work (for me at least). - I found the bit that triggers non-VGA modesetting in nv10reg, so in the future we could copy the blob in that area. - Misc fixes to a lot of things. commit a7a8993f4074e942ec4b3ef58ede0162a2d3c415 Author: Maarten Maathuis Date: Mon Nov 26 22:11:03 2007 +0100 randr12: Make sure the vpll is properly set when we steal a ramdac from another output. commit c11d3f68720b698736d5a28a82d3e45b55379bc6 Author: Maarten Maathuis Date: Mon Nov 26 19:40:59 2007 +0100 randr12: Fixed a minor cursor bug. commit 59d98374e082c40b57737de56b17173f5b54910e Author: Maarten Maathuis Date: Mon Nov 26 19:24:03 2007 +0100 randr12: Set the value of sel_clk correctly (hopefully) for NV4x. commit 8672d7a33fe369645786c21e130f51560a656d3c Author: Maarten Maathuis Date: Sun Nov 25 22:08:23 2007 +0100 randr12: cleanup + double register assignment (vga CRTC) removed. commit 11809882f24e40b1b8c77f4caca27ce2ebd3ccbf Author: Maarten Maathuis Date: Sun Nov 25 18:15:27 2007 +0100 randr12: Convert premultiplied alpha cursors to normal alpha cursors. commit ece41daabeb0cfd727daec14d9734dac08159c7a Author: Maarten Maathuis Date: Sun Nov 25 16:49:01 2007 +0100 randr12: Fix alpha cursor corruption. commit 5abbf55fb5c957c4be425e9ebd2660ca14d44e27 Author: Maarten Maathuis Date: Sun Nov 25 14:15:30 2007 +0100 randr12: alpha cursors don't need background changes. commit 37c36a284508e63af152a9315aa19a729e5516cd Author: Maarten Maathuis Date: Sun Nov 25 13:06:10 2007 +0100 Reindent a few cursor functions. commit 1ad212271cc99b95ff52eb719d24ac58d392c0eb Author: Maarten Maathuis Date: Sun Nov 25 03:20:43 2007 +0100 randr12: Improve TMDS and LVDS registers (and enable for LVDS). commit 39d94139910c53f9092f28ea6d742133666ac18d Author: Maarten Maathuis Date: Sun Nov 25 01:31:39 2007 +0100 randr12: Optimize tmds_regs a little bit. commit c6dce92c47dacfc822315e570bff70a8e42f5dc2 Author: Stuart Bennett Date: Sat Nov 24 21:51:14 2007 +0000 randr12: use bus field to determine dvi pairs commit 328197b4613f5e25eacfb67888701bb0e298c765 Author: Stuart Bennett Date: Sat Nov 24 20:44:27 2007 +0000 Remove a load of unnecessary casts, and reorder functions by opcode Plus a micro opt to checksum thing commit 8cee41bc6112a96f43f242b5178a9c968d2b2d9e Author: Stuart Bennett Date: Sat Nov 24 20:25:13 2007 +0000 A bunch more cleaned up bios opcodes Thanks to stillunknown for the bios trace commit 5d6f214db4cd5f0cdcb6a4395ce2ce4acf322257 Author: Maarten Maathuis Date: Sat Nov 24 16:22:04 2007 +0100 randr12: Be more considerate of outputs that can only work on ramdac 1. commit 6e1c37e26f630caa03da9061a5cec200708babbd Author: Maarten Maathuis Date: Sat Nov 24 15:47:04 2007 +0100 randr12: The result of another refinement of dvi crosswiring. commit ce57eb9b34095f2776ef36b09249a4b202824a86 Author: Maarten Maathuis Date: Sat Nov 24 13:12:49 2007 +0100 randr12: Disable crt_load_detect for ramdac 1 and set some better valued for test_control. commit 525997f86c5fe0f79cf0fa584b4cb617b83a9386 Author: Maarten Maathuis Date: Fri Nov 23 23:12:53 2007 +0100 randr12: add bit30 to vpll register for nv4x cards - This seems the most common situation, any idea what the bits do? commit f249b3b33c025417d11a85a29848db2112450dce Author: Patrice Mandin Date: Fri Nov 23 22:25:47 2007 +0100 nv30: missing command commit b56ca2ef87fe15132dcbc43361344d8af446c8a7 Author: Patrice Mandin Date: Fri Nov 23 22:25:06 2007 +0100 nv30: another reference to an object commit a8af1725ef7ae6c1cd19659d0607e2a702dc148a Author: Patrice Mandin Date: Fri Nov 23 22:24:19 2007 +0100 nv30: rename viewport clip commit 14e631691a2e8704ee016828e94c48de7eb96409 Author: Patrice Mandin Date: Fri Nov 23 22:03:21 2007 +0100 nv30: was not using gart handle commit 9648fd259f9906f5db854e8c4b57fd2675f4d5c5 Author: Maarten Maathuis Date: Fri Nov 23 21:08:46 2007 +0100 randr12: Reenable a properly working crt load detect. commit 3010847c1ccf14f0b5356aeb05f2c8c0c519ce91 Author: Maarten Maathuis Date: Fri Nov 23 18:58:49 2007 +0100 randr12: Remove some unneeded stuff. commit 2b092ca5277405a166ea54de418a8114fb6a8cf6 Author: Maarten Maathuis Date: Fri Nov 23 17:32:12 2007 +0100 randr12: fix comment typo commit 5ca53ef57ece7ae547cd5b11e83eb0d15fe69033 Author: Maarten Maathuis Date: Fri Nov 23 16:40:02 2007 +0100 randr12: Allow arbitrary ramdac and crtc combinations. - Support for the misterious ramdac 2 still has to be reverse engineered. commit e16fccde3f817e9eab3a09398a4821566967a86d Author: Maarten Maathuis Date: Fri Nov 23 13:25:38 2007 +0100 randr12: Move a lot of registers from output to crtc. commit cd8293954e1d004ef837d339ae87c4689fd45e1f Author: Maarten Maathuis Date: Fri Nov 23 10:32:08 2007 +0100 randr12: There was more logic to sel_clk than previously thought. commit 0b55ba9f9fbd60044bad880f260e0c70e6f577f0 Author: Maarten Maathuis Date: Thu Nov 22 22:18:47 2007 +0100 randr12: revert to old vpll regs on nv4x cards, because the others can't be written - If anyone knows the bit to allow those registers to be written, then please tell. - This also fixes db1 vclk ratio issues, hopefully. commit 18117df17ec2ce355344953d70caf33a0272697f Author: Stuart Bennett Date: Thu Nov 22 00:30:29 2007 +0000 Bios CRTC head chosing fixed Also, a slightly tighter limit on values for or, as I don't currently think bit 4 being set makes any sense commit 425b7cd9dcddf1735d59e9ef60e81bf91db1c65c Author: Stuart Bennett Date: Wed Nov 21 20:02:17 2007 +0000 randr12: unbreak LVDS console restore by reverting tmds_regs change commit e3f77b1b4332f5c072be713b5ea6dd234c194bb0 Author: Stuart Bennett Date: Tue Nov 20 19:02:17 2007 +0000 Complete init_index_addr_latched opcode commit 05f40b2dd3f8c14da2464b19b21aa66defec17e3 Author: Stuart Bennett Date: Wed Nov 21 01:58:12 2007 +0000 More DCB entry output, and a default for DCB 1.4 The table seems to be the same for all DCB 1.4 biosen I've seen, and POST seems to make no effort to parse it. I wonder if it is in fact used at all commit d96176c2f07efde4224c35b7c03b8380a73e1831 Author: Maarten Maathuis Date: Tue Nov 20 20:41:38 2007 +0100 randr12: better guess for ramdac register 670 (nv44 and up) commit 817b8f1a1af70402c18c45edccec5c746063d4e8 Author: Maarten Maathuis Date: Tue Nov 20 19:55:15 2007 +0100 randr12: fix serious bug in vclk calculator for nv4x cards. commit 040e5479344b0f794e79e4ffc6c547c58835401a Author: Maarten Maathuis Date: Tue Nov 20 18:49:39 2007 +0100 randr12: Fix a small mistake that could cause blank screens for NV44 and up. commit 1cc05c9e08a7e4f0c27bb0e08d47265d31123bf9 Author: Maarten Maathuis Date: Tue Nov 20 12:56:28 2007 +0100 randr12: I hope i really fixed db1 vclk ratio this time (for nv4x). commit 016f7e84c4ef62840e300d61e5e3b2ddfadd5e32 Author: Maarten Maathuis Date: Tue Nov 20 12:44:25 2007 +0100 NV50: Make a small mistake. commit 2cbd24ad579d3e68eaef2f8500b22e101fecd936 Author: Maarten Maathuis Date: Tue Nov 20 12:25:21 2007 +0100 NV50: Some deobfuscation. commit ccfc6910f8cd8eb8d4b628cc14ef7db417f576ea Author: Maarten Maathuis Date: Tue Nov 20 09:52:47 2007 +0100 randr12: Also enable DB1 vclk ratio if PLL_SELECT had it enabled. commit 5b3d86562afd35c14d2c8c07618b08bbc98ad775 Author: Maarten Maathuis Date: Mon Nov 19 23:55:58 2007 +0100 randr12: Proper support for db1 vclk ratio, thanks to Thunderbird for some info. commit 22d8ba5ebf659f6f126a1380190339d08ee824c6 Author: Maarten Maathuis Date: Mon Nov 19 09:23:37 2007 +0100 randr12: Add a gamma_set hook. commit cbd70303a949f387d42497cb4ea52186ed938b9c Author: Maarten Maathuis Date: Sun Nov 18 22:03:03 2007 +0100 randr12: A first attempt at some lvds programming, not hooked up. - This is a start for someone with a laptop to test and refine. commit 58c3d6a886c5d5557def490ee8c3e75c9f8d2cd7 Author: Benjamin Herrenschmidt Date: Sun Nov 18 02:50:33 2007 +0100 Enable IFC uploads. commit 619ae1f7a028f24293f9a4e4f7777cc51ef8d16d Author: Maarten Maathuis Date: Sun Nov 18 00:15:34 2007 +0100 randr12: Final commit, switch to nv40 native vclk regs. - I've also seperated the save load pll functions, to improve clarity. commit 601adb4f431b638d4b711ba01aa2d4c3408a7bd3 Author: Maarten Maathuis Date: Sat Nov 17 23:47:12 2007 +0100 randr12: Intermediate commit (using the native nv40 vpll regs). commit a29eca9dec176388b1c4666ad801ae75360a1219 Author: Maarten Maathuis Date: Sat Nov 17 20:08:27 2007 +0100 randr12: Some comment fixes. commit 6f3d335f6e3b9866457730660538c5f073862dfe Author: Maarten Maathuis Date: Sat Nov 17 22:23:32 2007 +0100 randr12: Reinstate m-divider limitations. commit e2b7f57cbe7e1080736e6088d2997cbcf966fc3d Author: Maarten Maathuis Date: Sat Nov 17 18:36:24 2007 +0100 randr12: Some minor changes. commit 2c0c7612bd08938496abb47b34498065e668ec34 Author: Maarten Maathuis Date: Sat Nov 17 15:57:20 2007 +0100 randr12: Change vpll handling, note that it still is a work in progress. commit caec3d1ab69c523b6c0a7e0019d1a9f115cf5da3 Author: Ben Skeggs Date: Sat Nov 17 20:59:25 2007 +1100 EXA 1.4 pixmap hooks (disabled) Really, you don't want to turn this on yet... :) commit b8e7c64476948426d5801367bab5cdf15db5ce84 Author: Ben Skeggs Date: Sat Nov 17 17:17:34 2007 +1100 Add a user callback for when a channel hang is detected. commit 293b32d0f66e965b3a31d54859b91210b3e35b3d Author: Ben Skeggs Date: Sat Nov 17 17:02:23 2007 +1100 Use handles from grobj, instead of hardcoded vals from nv_dma.h commit 84bc37723c2cabe7652050b4cbaf9f5218fdb071 Author: Ben Skeggs Date: Sat Nov 17 16:43:05 2007 +1100 Turn vram/fb handles into "real" grobjs. commit f3cdd5086f410711103c6b3453629d85ef4c2788 Author: Stephane Marchesin Date: Sat Nov 17 01:22:31 2007 +0100 Put back memory barrier and write posting. commit 762f732c7aaa8aa3dd7d54a8734cbcc54cd10494 Author: Stephane Marchesin Date: Fri Nov 16 23:38:49 2007 +0100 Add verbose debugging code to help with fixing push buffer issues. commit c8f7a9fcdcc356b4dc5030112b1532d2fa3eaf0f Author: Stephane Marchesin Date: Fri Nov 16 23:37:10 2007 +0100 Oops. Correct either the header or the packet but not both. commit 5ce573875387490eb7feaa87a3354b888beb4410 Author: Stephane Marchesin Date: Fri Nov 16 23:32:07 2007 +0100 Unmatched OUT_RING(0). commit cc0c1af932a96cec6aabc72cbe9ca508fbc80a36 Author: Stephane Marchesin Date: Fri Nov 16 21:59:53 2007 +0100 fix typo. commit 4b7601c297fefc46adf3bc37a463575d3caafedb Author: Stephane Marchesin Date: Fri Nov 16 21:56:14 2007 +0100 fix 32/64 bit ism. commit bee8bc5ce321245362847253b7ccc7530899e514 Author: Maarten Maathuis Date: Fri Nov 16 21:29:20 2007 +0100 randr12: Minor change. commit 735fae4bb0b92f097665f675ba5127a9bb11d8fa Author: Maarten Maathuis Date: Fri Nov 16 19:19:23 2007 +0100 randr12: That wasn't supposed to end up there ;-) commit 2848c3e7a4d25e5ca311ea1a88458eb2a77976ed Author: Maarten Maathuis Date: Fri Nov 16 19:16:08 2007 +0100 randr12: Disable changing ramdac reg580, until i know why this causes issues. commit 75222fa199f1f5bc1abd66a4e80a446b5d597523 Author: Ben Skeggs Date: Sat Nov 17 15:31:49 2007 +1100 nouveau: fix valgrind complaint commit 1a98f1ae8298ca3b069b146a2866f96e934afb74 Author: Maarten Maathuis Date: Fri Nov 16 17:00:59 2007 +0100 randr12: Restore old timings + misc changes. commit 1b0cd6602cf9d96f41242054efe2f96cc70eaa96 Author: Ben Skeggs Date: Sat Nov 17 09:08:03 2007 +1100 oops, forgot these... commit 9c46ec6833ccfc29e4a67b0ce4226d8b38b772ab Author: Ben Skeggs Date: Sat Nov 17 07:23:41 2007 +1100 Initial hacky relocation stuff. Entire DDX ported across to OUT_RELOCx/OUT_PIXMAPx, workable in it's current state for the current BO backend.. definitely not for TTM yet :) commit ddda33decb42ba2f5f9586e1d5de4533d1d88674 Author: Ben Skeggs Date: Sat Nov 17 06:11:01 2007 +1100 Initial buffer object work. Implemented on top of the current MM interface, and currently used in the exact same way NVAllocate/FreeMemory was used. Can transition to something more ttm-friendly gradually. The "old mm" bo backend is extremely primitive. Buffers can only be created in mappable memory, forcibly pinned, and never moved once they're created. Some of these things could be improved, but probably not much point. commit 9b9d995a517b48cd074030c567f9c78bee19b129 Author: Ben Skeggs Date: Sat Nov 17 03:36:04 2007 +1100 Report engine hangs again, instead of spinning forever. commit 2af4a28ae1df46686e264bcda417a896b7cfa59a Author: Ben Skeggs Date: Fri Nov 16 09:18:33 2007 +1100 Hook up DMA funcs to channels/grobjs etc etc commit 6e3d05e8c60fd7bab371af86151e45eec0a65b3d Author: Ben Skeggs Date: Fri Nov 16 07:08:23 2007 +1100 notifiers.. Lacking timeouts at the moment so if X would've crashed before, it'll spin forever now. commit 020740d92b5d623bee65e598ac60020187011b04 Author: Ben Skeggs Date: Fri Nov 16 06:07:56 2007 +1100 And grobjs. commit 929e2b7d53435d097b07c23ed1d0caaa47924f8a Author: Ben Skeggs Date: Fri Nov 16 05:10:58 2007 +1100 And channels.. commit 22909d96b8ea720169b34326b93c7b29eb3db013 Author: Ben Skeggs Date: Fri Nov 16 04:36:08 2007 +1100 Implement nouveau_device struct. commit b96e813773b0705bbac57f2d64adff3aad845622 Author: Stuart Bennett Date: Fri Nov 16 03:26:56 2007 +0000 Begone, foul g5-bios-hack also, fix the DCB termination test; I'm amazed nothing broke commit 7eee9585fa8e8c45f4ab4712a3b990664203d72b Author: Maarten Maathuis Date: Thu Nov 15 19:33:13 2007 +0100 randr12: Only NV4x have two seperate PVIO ranges. commit 276892b069fd045b20f44d7888a7971371d73ad3 Author: Maarten Maathuis Date: Thu Nov 15 19:27:12 2007 +0100 randr12: Improved values for FP_CONTROL register. commit 4a16f14061c88ed49b05ef8779a2584f5053c3f1 Author: Maarten Maathuis Date: Thu Nov 15 18:44:52 2007 +0100 randr12: Some minor register changes. commit b436a6809c62e0fe0696d2c6bad066c3afcf9166 Author: Arthur Huillet Date: Thu Nov 15 15:01:47 2007 +0100 NV10 EXA: removed left out debug statement commit 98e223ebb01440810e01e902e904524c50b463cb Author: Arthur Huillet Date: Thu Nov 15 12:57:20 2007 +0100 NV1x EXA : refactored A8 + A8 code, enabled on NV2x, fix A8 txfmt for NV2x more compact and more readable A8 + A8 code along with some codestyle fixes now enabling this acceleration on NV2x cards, with the A8 format value hack commit 3cdc4b274f371f1f524d140fddd79dad2987b06c Author: Ben Skeggs Date: Fri Nov 16 03:28:45 2007 +1100 nouveau: fix a host of warnings commit 6eba6c7fab534efd15c8e96daeeb9a33972c2b6d Author: Stuart Bennett Date: Thu Nov 15 02:09:45 2007 +0000 Big endian bios fix commit 69a3a849e65a022609db3a5f33277b744463a6ae Author: Patrice Mandin Date: Wed Nov 14 23:03:06 2007 +0100 nv30exa: set link with image blit commit 2435836202efe7ea42a7bece71f125fb46aac3ba Author: Stuart Bennett Date: Wed Nov 14 19:02:28 2007 +0000 nv_bios indentation commit 7b0ca260f57e92e45d7f48f16642bfcd67b10eaf Author: Stuart Bennett Date: Wed Nov 14 18:47:40 2007 +0000 More code and logging beautification - wrap I/O better Plus some actual fixes commit d00420a1286e368fbc20551cae5260bf65ed1b15 Author: Stuart Bennett Date: Wed Nov 14 19:07:06 2007 +0000 IO flag conditions, beginnings of init_idx_addr_latched plus usual whitespace and pointless renaming commit 4e8af31a89b7d732b52706f87e1dc66e29ab717e Author: Stuart Bennett Date: Wed Nov 14 03:01:27 2007 +0000 More opcodes, and some monstrous reindenting all part of an evil plan to have the most lines in the Xorg log commit 89617cfcdfd1ee19b99888cfbcfec1ee0550d26c Author: Stuart Bennett Date: Wed Nov 14 00:36:02 2007 +0000 Updating some BIOS parser opcodes plus various debug / whitespace fixes commit cc5db05bed630350613c7cbfa418d814ed5a5093 Author: Stuart Bennett Date: Wed Nov 14 19:27:06 2007 +0000 randr12: More DCB abstraction to lessen parsing nastiness for old versions commit d46b61f88225d3e2b28a6c4a4ac4b8a1fe942a39 Author: Stuart Bennett Date: Tue Nov 13 18:24:46 2007 +0000 Gratuitous code renaming, moving + warning fixes for bios commit c87e3570a198b5d69046387d111a2709a571d5e3 Author: Maarten Maathuis Date: Wed Nov 14 22:30:11 2007 +0100 randr12: Reinstate the "magic factor" until a better solution comes along. commit 5b62e77ded9e64d439b70ce0b7dce27a7ae1df4f Author: Arthur Huillet Date: Wed Nov 14 20:54:24 2007 +0100 NV1x EXA : implemented A8 + A8 via a dirty hack it works, but probably needs to be optimized commit 3db89822ba94a94eeb1ab891488d7f203a0dfff8 Author: Maarten Maathuis Date: Wed Nov 14 19:48:53 2007 +0100 randr12: Disable ramdac register 0x900, until i know for certain what it does. commit 842b33e2dc82e70a7da992642de4503d4cb0a1fc Author: Maarten Maathuis Date: Wed Nov 14 16:02:56 2007 +0100 randr12: fix comment typo commit c7fc4b9981bb242830ddb830863fb26301429c74 Author: Maarten Maathuis Date: Wed Nov 14 08:34:44 2007 +0100 randr12: Enable all pll on pre-nv40 and improve usage of ramdac_670 register. commit 992d079d64369a4a8445012b438ec980968c11d8 Author: Maarten Maathuis Date: Tue Nov 13 21:46:23 2007 +0100 NV50: fix stupid type from commit d5dfb82f7f39304eb39dd4b1c34a2a2801d90a25 commit 7ad7b6ad453b6401ef0bf5128f7ac9790df95bb2 Author: Stephane Marchesin Date: Tue Nov 13 21:36:13 2007 +0100 Don't use GART on PPC. Workaround DMA bugs there. commit b451a21f77678b1fb1d31f3df7e3b1aea6984876 Author: Stephane Marchesin Date: Tue Nov 13 20:58:51 2007 +0100 Remove ENABLE_NV30EXA, which is not needed any more. commit 9dc1e1e1fc049c7a535da8e1bc8ce79fd2d620b5 Author: Stephane Marchesin Date: Tue Nov 13 20:54:53 2007 +0100 Enable nv30 exa on PPC. commit 4fb7b608cbdf53084da63526c066eb7f772c6c14 Author: Stephane Marchesin Date: Tue Nov 13 20:53:58 2007 +0100 Fix nv30 exa for big endian commit 875960dacda54268fc1d318995c145b2db09ca5b Author: Maarten Maathuis Date: Tue Nov 13 19:43:41 2007 +0100 randr12: Set a few new registers + misc fixes. commit cf9aaa2b074eca67cf0119172c3d15328418416d Author: Stuart Bennett Date: Mon Nov 12 19:24:34 2007 +0000 Some type / warning fixes for nv_bios.c commit f00f6d1df06d2671255fd988f514a6f6ab0f8b18 Author: Stuart Bennett Date: Sun Nov 11 17:50:55 2007 +0000 Misc bios parsing bits commit fdc7f4a353c4dda07fffd9800b477ae5bc43baac Author: Patrice Mandin Date: Mon Nov 12 23:54:22 2007 +0100 Remove disable switches, big cleanup, requires xorgs server 1.3 again commit e93a5d3abe6d637e00bd9c998e9735c9766eecfb Author: Maarten Maathuis Date: Mon Nov 12 21:05:00 2007 +0100 randr12: Improve vclk setting. - It now considers crosswired TMDS situations. commit d32420c82b1e63bbc84768c92c336391af624952 Author: Maarten Maathuis Date: Mon Nov 12 20:17:04 2007 +0100 randr12: Fix bad comment. commit 8a04babb88111ec8ccec9af56c3591d41afba9b3 Author: Maarten Maathuis Date: Mon Nov 12 20:14:10 2007 +0100 randr12: Some minor changes. commit 821d4b83ed780d9a0a8d950a2595a0e15ef3f103 Author: Maarten Maathuis Date: Mon Nov 12 16:21:14 2007 +0100 randr12: Set SEL_CLK appropriately based on wether a TMDS is crosswired. commit d017ce4879ca755fc6d56f232fe3162864f4b38e Author: Maarten Maathuis Date: Mon Nov 12 10:34:54 2007 +0100 A little code cleanup. commit 16fb5b92dba9a0380d2fdd1ece1b9b0691d79d3a Author: Maarten Maathuis Date: Mon Nov 12 00:17:52 2007 +0100 NV50: Port an LVDS detection fix from nv. commit b01a87f18d122d80fac6daf1dfb1c404d7f3a814 Author: Maarten Maathuis Date: Mon Nov 12 00:03:19 2007 +0100 NV50: Port one of the nv patches. commit f22691661f8b72cedb8aa4fc96b41fcff4171e12 Author: Maarten Maathuis Date: Sun Nov 11 21:53:10 2007 +0100 randr12: Some code cleanup. commit aceceb22775bb039cb0e207f39f7cbd93c2db194 Author: Maarten Maathuis Date: Sun Nov 11 21:32:31 2007 +0100 NV50: Clean up nv50_output.c a bit. - I did not make wrappers for the i2c stuff, since i do not know what that range does, besides i2c. commit a0835f7c64cf5f62626332ccba21cfe436584275 Author: Maarten Maathuis Date: Sun Nov 11 21:17:40 2007 +0100 NV50: cleanup of nv50_sor.c commit f463a7f6213bede18917191546c8cd61cf4612e8 Author: Maarten Maathuis Date: Sun Nov 11 20:51:01 2007 +0100 randr12: Make sure we don't set 2 "special" bits in sel_clk. commit 27cbf73bc572f435870decc2017cbf1c4f38264d Author: Maarten Maathuis Date: Sun Nov 11 19:32:26 2007 +0100 randr12: Overhaul vpll set code + misc fixes. - Hopefully the problematic G70's work again. commit 78e88d1f4e2f058c487978c01e19a828271f3f0b Author: Stuart Bennett Date: Fri Nov 9 03:17:12 2007 +0000 vbios fptable v2 parsing (for BIT biosen). Not that I know how to index it. Plus some fptable v1 fixes, bmp structure length checks, and support for a weird bios version that may be in some gf2go commit 15c771a6b3ecb9b36f9725c518141cf889aa953c Author: Stuart Bennett Date: Fri Nov 9 01:19:04 2007 +0000 G5 6600 DCB improvements, but I'm sure to be some bswaps short This will all become so much nicer when I can do leNN_to_cpu()... commit 9408fcbbbbab80608c8923ab2b7757dfa76b9767 Author: Maarten Maathuis Date: Sat Nov 10 17:02:38 2007 +0100 NV50: Last part of nv50_dac.c commit 85b5e6bfc17c347a9976155a755637f7c8d9304f Author: Maarten Maathuis Date: Sat Nov 10 16:59:49 2007 +0100 NV50: More cleanup (this time nv50_dac.c). commit 23d9f09a4600abf5426cb6d069dc5d9aa876a771 Author: Maarten Maathuis Date: Sat Nov 10 16:32:51 2007 +0100 NV50: That's the initial cleanup for nv50_display.c commit 25cef08b183d6406ce065b2b3cd224dd1840525b Author: Maarten Maathuis Date: Sat Nov 10 16:24:32 2007 +0100 NV50: Should have run make first ;-) commit d5dfb82f7f39304eb39dd4b1c34a2a2801d90a25 Author: Maarten Maathuis Date: Sat Nov 10 16:23:48 2007 +0100 NV50: Even more cleanup. commit 391c498e676752ac6bf03b668a1cb00094c1a049 Author: Maarten Maathuis Date: Sat Nov 10 16:19:32 2007 +0100 NV50: More cleanup, small commits to make bisecting easier. commit 59788f9d4e6c264a6b666918989ad6b778026bd9 Author: Maarten Maathuis Date: Sat Nov 10 16:05:42 2007 +0100 NV50: Some initial cleanup. commit 1a7d7ccb4c12f6a10aa5e8b41daee1aae52efa8a Author: Maarten Maathuis Date: Sat Nov 10 15:39:04 2007 +0100 NV50: Another wrapper. commit 3ca3170c8bbd758da81a1bc5b5a84cf3ca642c07 Author: Maarten Maathuis Date: Sat Nov 10 15:29:13 2007 +0100 NV50: A new wrapper + a few functions renamed. - I'm trying to clean up the code a bit, but lack of the hardware prevents me from testing it. commit 77dc99b2621523e20a956a8eaa01a100c992c458 Author: Maarten Maathuis Date: Sat Nov 10 00:35:04 2007 +0100 randr12: This may break stuff for some g70 cards, if it does, then please scream. - I hope i've got most situations covered. - If there are issues getting the second head to work on nv4x cards, then please tell. - nv4x is geforce 6 and 7 series for those who do not know. commit c169480b0bb177e8c5efb292d38c55554dcdfcdd Author: Maarten Maathuis Date: Fri Nov 9 20:41:48 2007 +0100 randr12: Disable load detect again, until we can identify it from a DVI monitor. commit 7237a7014b426acc4857d32967e0f10974877819 Author: Maarten Maathuis Date: Fri Nov 9 20:40:46 2007 +0100 randr12: Improve detection of DVI-D/DVI-A pairs (for naming). commit a8125c458fa53544625da2071a69e3fd27d7c084 Author: Stuart Bennett Date: Fri Nov 9 19:27:37 2007 +0000 randr12: Revert "randr12: lose ffs wrapper", it's wrong commit d83cb471cfcc18973eb4a5f319bc17a06e0a883f Author: Stuart Bennett Date: Fri Nov 9 19:02:00 2007 +0000 randr12: lose ffs wrapper commit dedf68e10f4b0cde6f9c49d60df737ed7b4148c0 Author: Stephane Marchesin Date: Wed Oct 31 02:36:41 2007 +0100 More nv30 exa cleaning up. commit d3679e8a5c0ce54ee3ef4de15be5a1fcd353df8e Author: Maarten Maathuis Date: Fri Nov 9 18:44:37 2007 +0100 randr12: Do not mess with ramdac register 580 until it's proven to be neccesary. - I cannot seem to find out what the conditions are to change this register, except that it's related to VCLK ratio DB1. commit a49d51c638c2004641e58dc0061652553e9a6751 Author: Maarten Maathuis Date: Fri Nov 9 18:31:02 2007 +0100 randr12: Fix vertical scaling. commit 2e6f5e18231a327cac7de4a927435585d1eadcee Author: Patrice Mandin Date: Fri Nov 9 18:13:43 2007 +0100 oops, trailing whitespace commit 596f4e6a78900f3cbb3beb4c99da7606dc630179 Author: Patrice Mandin Date: Fri Nov 9 18:11:25 2007 +0100 Also disable nv40exa if nv30exa disabled commit d9f80b9d5f4ae1c00e79b9e903c2fc0c23b37a13 Author: Patrice Mandin Date: Fri Nov 9 18:10:53 2007 +0100 Reading DFP tables need randr1.2 commit 50aaccb5e0d3b9a978701c55ebe6545a6f7b1456 Author: Stuart Bennett Date: Thu Nov 8 18:59:50 2007 +0000 randr12: invert fpScaler to regain original meaning (see man page) commit 3f6db3ac15b51a58353396fce2126a369b4cfd89 Author: Stuart Bennett Date: Thu Nov 8 20:30:28 2007 +0000 randr12: Do scaling / native fixup in the fixup function Nuke the LVDS specific fixup code Then use adjusted_mode accordingly commit 51a874a59d82678bbb790b708f9ba62985d99344 Author: Stuart Bennett Date: Wed Nov 7 18:37:57 2007 +0000 randr12: rename DIGITAL and PANEL to TMDS and LVDS Following on from earlier renaming, this is less confusing, at least to me. commit 61c92f10e1182403fc6fc8a666ed3d8b14795481 Author: Stuart Bennett Date: Thu Nov 8 17:00:44 2007 +0000 randr12: blank screen before we're ready to show it commit da64328dd7be7f0e7ee1dd0dd774174fc126948b Author: Maarten Maathuis Date: Thu Nov 8 20:11:25 2007 +0100 randr12: Some outputs are only valid on secondary output. commit fe3d49ccb24ff20fba79ee97cceba64850d9387f Author: Maarten Maathuis Date: Thu Nov 8 19:20:41 2007 +0100 randr12: Improve output naming scheme. commit 42542b36d771142d38f72b3365da8b53e9d25c55 Author: Maarten Maathuis Date: Thu Nov 8 19:13:03 2007 +0100 randr12: Add ramdac assignment code to the lvds dpms function. commit 2f1bf8138ba970ad18afe1ba9ee10228ea213926 Author: Maarten Maathuis Date: Thu Nov 8 18:42:35 2007 +0100 randr12: Misc fixes. commit 8b733ab61e0b9eb375783e986de776eef5fd49ff Author: Ben Skeggs Date: Fri Nov 9 03:29:21 2007 +1100 get rid of ALLOCATE_LOCAL etc. commit bfb78d2090ee5c189cf00e6af5107a756e864e98 Author: Maarten Maathuis Date: Thu Nov 8 00:20:53 2007 +0100 randr12: It seems the situation was a bit easier to handle, luckily. - I've tied ramdac and crtc number together. - Switching crtc's doesn't cause a problem anymore (even when crtc0 is not in use). commit d56e6e498ed9acc35bfa540881243acf03981571 Author: Maarten Maathuis Date: Wed Nov 7 22:25:40 2007 +0100 randr12: Intermediate commit. commit f13659a8c8e855dc0b871173ccfdd0b70236be71 Author: Arthur Huillet Date: Wed Nov 7 23:03:54 2007 +0100 NV10EXA: added support for R5G6B5 as source and destination (tested only with rendercheck) added A8 + A8 via ARGB + ARGB hack, disabled by default because it cannot work updated instrumentation to dump pixmap sizes and repeat flag commit 3d671bf3b06e64b49254db666980be3a3b3652cc Author: Stuart Bennett Date: Wed Nov 7 15:08:48 2007 +0000 randr12: remove some duplication and reduce mem leaks commit cdbec606e4ec398d6bef87d635499100559e3a19 Author: Stuart Bennett Date: Wed Nov 7 14:54:18 2007 +0000 randr12: refactor ddc stuff for convenience commit ed8473b228579024a4fdc9c2a9ecf4cb71135378 Author: Stuart Bennett Date: Wed Nov 7 14:33:50 2007 +0000 randr12: micro-optimization on use of "or" commit f6d0b446eda687bb55649599e1aedfa231feaaa9 Author: Stuart Bennett Date: Tue Nov 6 21:20:10 2007 +0000 randr12: LVDS warning commit 62f8fed79f1617ffd0289123045fda0a3c0e92b2 Author: Stuart Bennett Date: Tue Nov 6 01:29:33 2007 +0000 randr12: Use NV_RAMDAC_PLL_SELECT_PLL_SOURCE_ALL for LVDS Using only VPLL seems to be broken for LVDS. Traces show consistent use of ALL. commit 0b958dc66a5533a7db61b7b2f456df876e9905ff Author: Maarten Maathuis Date: Wed Nov 7 09:33:40 2007 +0100 randr12: Also save flatpanel regs on analog monitors. commit c8ac3f0c7cfd6eca85f332e11e280f715e80283e Author: Stuart Bennett Date: Tue Nov 6 21:09:51 2007 +0000 Various minor fixes Now that ramdac selection uses sensible numbers, it turns out writing to TMDS regs is a bad move on LVDS. Also, change FP_CONTROL handling a bit for LVDS (bit31 needs preserving) commit d7be5bdc27fee0758df33feb3a331c9c79dd94e0 Author: Stuart Bennett Date: Tue Nov 6 19:53:08 2007 +0000 Start using OUTPUT_PANEL where appropriate commit f1160ac702a251708f2385b78c1657f95c2916ab Author: Stuart Bennett Date: Thu Nov 1 00:24:13 2007 +0000 Beautify DCB 1.2 weirdness, hopefully fixup G5 and use dcb_table more simply commit 38829fa5cdac1469a6a71a092a42885ea048a1ba Author: Stuart Bennett Date: Tue Nov 6 19:41:47 2007 +0000 Implement lvds_output_{get_modes, mode_valid, mode_fixup} commit 29c578bfe9415aa86910cf6d5cd44787b4fdf44d Author: Stuart Bennett Date: Fri Oct 26 03:01:17 2007 +0100 Reindentation, renaming, adding debug, and assume LVDS always connected commit 7aeae6cdc99ab633361f107752ef1bfa3e0492e0 Author: Stuart Bennett Date: Wed Oct 31 17:29:16 2007 +0000 LVDS VBIOS FP mode finding commit cd806f92e29573ea968ab2fddb454eb1513e37dd Author: Maarten Maathuis Date: Wed Nov 7 00:16:33 2007 +0100 randr12: Fix ugly alligment. commit 88faf35179b9af8060c35a7a95e9a6baaae552be Author: Maarten Maathuis Date: Wed Nov 7 00:11:51 2007 +0100 randr12: Improve hotplugging of outputs. commit 4aa3acbd5410c8dda57a483cf7967258440888dc Author: Peter Winters Date: Tue Nov 6 23:40:07 2007 +0100 NV10 EXA: Implement a8 1x1 repeat commit 524e555c468fbab32b515ecd2c1e9cccf6c48443 Author: Peter Winters Date: Tue Nov 6 23:29:03 2007 +0100 NV10 EXA: setup viewport like nv40 exa commit 2f2298f0ae3777c775b3ab81356adebf4f52ed77 Author: Maarten Maathuis Date: Tue Nov 6 22:15:20 2007 +0100 randr12: Readd crt load detect + misc changes + restrictions (see full log) - Crt load detect has been readded, but only for the primary output (the other gives false positives). - Misc changes to ramdac management. - Until i know how to trigger a full mode change, i cannot allow adding an output which requires ramdac0 during runtime. commit a6244c25a5c7efc633c74f506ca121e10a5642c0 Author: Maarten Maathuis Date: Tue Nov 6 16:11:43 2007 +0100 randr12: Reasonably complete TMDS programming, see notes. - Please file a bug and make a mmio-trace when it fails for you. - I can now hotplug TMDS, so that makes both types hotpluggable on nv4x. - I have no clue what happens on pre-nv40 hardware, so please try. commit 5f6912f083048f01a33582d45bc6b22f0ac27d4e Author: Maarten Maathuis Date: Tue Nov 6 10:56:22 2007 +0100 randr12: Associate VPLL's with ramdac instead of crtc. commit 8241710c94f6df0bf683bc3c93f7ea1ca14d118c Author: Maarten Maathuis Date: Tue Nov 6 10:34:56 2007 +0100 randr12: Put TEST_CONTROL into neutral position. commit 19533d8ba549965dedd9083330b217270efaa49b Author: Maarten Maathuis Date: Mon Nov 5 23:32:16 2007 +0100 NV50: Don't hit pre-NV50 codepath. commit 5ca10da0c7dd2fd689cc4369e37ed385a0c10d21 Author: Maarten Maathuis Date: Mon Nov 5 21:28:22 2007 +0100 randr12: Remove incorrect (and hopefully redundant) output initialisation check. commit fcacad59d9233a66111d1c434e14cc4b2a08961e Author: Maarten Maathuis Date: Mon Nov 5 21:18:13 2007 +0100 randr12: Do not remove any bits from crtc register 580. commit e58c6735c73cc3a94a7a139e235b1dff9220e507 Author: Maarten Maathuis Date: Mon Nov 5 20:30:41 2007 +0100 randr12: Only use VPLL as timing source. commit 1071aeef24b349bac08a2ac091733c23cad1cb44 Author: Maarten Maathuis Date: Mon Nov 5 20:04:26 2007 +0100 randr12: Only enable crtc at the very last stage, should avoid mode setting "blinking". commit 1ac7798bdec75d9336afd1cfbe7fa177f5fe1465 Author: Maarten Maathuis Date: Mon Nov 5 10:03:15 2007 +0100 randr12: Add a native mode for TMDS and make that preferred + some ddc related fixes. commit bfb75679e504b335be0de4384830cc8129f92c1b Author: Ben Skeggs Date: Mon Nov 5 11:44:58 2007 +1100 Fix a thinko from commit 62294d9be8ec6f1159e3c2ad273649f937b7aa7e. commit bf8a2dfa7f83956d100316ed7450669795ea47d2 Author: Maarten Maathuis Date: Sat Nov 3 22:03:18 2007 +0100 randr12; Forgot a small thing. commit 19b0971d0c1d03896bfaee2b833194ef1e4d7566 Author: Maarten Maathuis Date: Sat Nov 3 21:54:48 2007 +0100 randr12: A legitimate DPMS call could reset the ramdac and effectively disable the mouse. commit a29839c718fd6bf3003081299017536dd9f8223e Author: Maarten Maathuis Date: Sat Nov 3 10:00:06 2007 +0100 randr12: Hopefully improve the native flatpanel mode finder. commit bd20ff5b8c9467ca913c00d3ec946e7e975ad3d4 Author: Maarten Maathuis Date: Fri Nov 2 23:17:16 2007 +0100 randr12: We cannot reset ramdacs during detect, there may not be a modesetting after it. commit 7054de29c589a7be0a6b027f3067ae2c02e2fc45 Author: Maarten Maathuis Date: Fri Nov 2 21:44:40 2007 +0100 randr12: Futher refine crtc timings, hoping to solve the "pink line" issue. commit 39ced4abb536993249e55a72eceeba26d577724c Author: Maarten Maathuis Date: Thu Nov 1 17:54:26 2007 +0100 randr12: Improved ramdac choosing, based on bios information (or value). commit 28ba9ed765d9956b2f07c6a405fca675a71012db Author: Maarten Maathuis Date: Thu Nov 1 17:01:53 2007 +0100 randr12: Remove redundant output setup + don't print full edid. commit 609f7b9c2962ccf04861d2164a0ff12c58e84010 Author: Maarten Maathuis Date: Thu Nov 1 13:00:45 2007 +0100 randr12: Cursor fixes(probably only affects nv4x). commit 6945dfdc5e6b2540432805e845cb8540dc00d326 Author: Maarten Maathuis Date: Thu Nov 1 11:10:24 2007 +0100 randr12: Improve analog monitor allignment + misc fixes. commit 5bd8ba52788b3b3b8f91ba59c29e154e38482481 Author: Ben Skeggs Date: Thu Nov 1 18:35:33 2007 +1100 nv40: what can I say? oops! commit f59e596bcd90ef824cd41e0c37952e574d6914bb Author: Ben Skeggs Date: Thu Nov 1 18:17:20 2007 +1100 nv40: completely untested fix for composite on ppc Thanks go to ps3rsx guys for finding the issue. commit 59c4fdb650ff2e6b6948d1bdd6bb22aa94195a32 Author: Maarten Maathuis Date: Wed Oct 31 22:59:54 2007 +0100 randr12: Some misc cleanup to make malc0's patches work. commit f165d4f2ca90916b1334cdc18d62b82ea71d9f62 Author: Maarten Maathuis Date: Wed Oct 31 21:30:36 2007 +0100 randr12: The "easy" TMDS registers, please report regressions. commit 860c4bdd38fd259ae51e32e4873a2aa8b1e6883e Author: Maarten Maathuis Date: Wed Oct 31 19:00:28 2007 +0100 randr12: Misc register fixes. commit d758e8f4c1a7d4effe1298a0a41d4e0b96496ab9 Author: Maarten Maathuis Date: Wed Oct 31 13:22:29 2007 +0100 randr12: Some changes to VCLK ratio handling. commit 899559921b0e46db010fa1e771203c2edefb7ade Author: Maarten Maathuis Date: Wed Oct 31 09:18:59 2007 +0100 randr12: some blob copying commit b5ecadb73335d934a8cb1c106983d219c9d8d399 Author: Stephane Marchesin Date: Wed Oct 31 02:09:08 2007 +0100 Fix bugs in nv30 exa (kde text seems back) commit f7b553a87e7f67138e06e74877145a1f83e9d188 Author: Peter Winters Date: Tue Oct 30 23:04:28 2007 +0100 NV10 EXA: implement repeat for 1x1 textures commit bf7b0231db1f7e45cee60d7d2e04b1b006aec2b0 Author: Peter Winters Date: Sun Oct 28 23:52:22 2007 +0100 NV10 EXA: Fix viewport setup commit f8f9c8789b89a02886702c46c5feafec169ae78a Author: Maarten Maathuis Date: Tue Oct 30 21:00:54 2007 +0100 randr12: Flatpanel scaling is finally working, read full message. - The option FPScale is now obeyed and will cause scaling to happen in the monitor. - The default is scaling is using the GPU. - When the aspect ratio is very different(>0.10 diff), it will obey aspect ratio and create bands around it. commit e73aa148d965abbf73392c6ffef3993dc26c31fd Author: Maarten Maathuis Date: Tue Oct 30 13:54:25 2007 +0100 randr12: Some copyright stuff. commit c1508e1448e3822e8b6bb2604c09a9d766a64a93 Author: Maarten Maathuis Date: Tue Oct 30 13:44:44 2007 +0100 randr12: Disable some panel scaling stuff, because it is broken. commit 3bf8110b61ef854af803a45f0faf1eaffbbcfac9 Author: Maarten Maathuis Date: Tue Oct 30 11:44:58 2007 +0100 randr12: Lots of misc fixes. - This does not fix the issue with -hsync, +vsync modelines (or the other way around) yet. - There are two PVIO ranges, i'm surprised noone noticed this. - Some misc fixes. commit 894f9aacce3d685b811f8961275a9f225b964159 Author: Stephane Marchesin Date: Tue Oct 30 01:49:38 2007 +0100 Disable nv30 exa no big endian until we fix the endianness bugs related to the 3D engine. commit 482cbcb9efdfbdd4bbe998c09f7d29033aeca09b Author: Stephane Marchesin Date: Tue Oct 30 01:37:45 2007 +0100 Use the new header with license. commit a9908dbb70df0d0abb586b9774efdc10adf4d74f Author: Stephane Marchesin Date: Mon Oct 29 23:00:14 2007 +0100 More copyrights I remember... commit 67e142394edcb553a40c6c8b0ef091bd4dad2bdd Author: Stephane Marchesin Date: Mon Oct 29 22:57:30 2007 +0100 Add some copyrights... not all files are covered yet, but I did the ones I recalled. commit f75e36a901fd47bbd9164242c408e9f8b0e03e2c Author: Maarten Maathuis Date: Mon Oct 29 18:37:34 2007 +0100 randr12: My suspicion was incorrect, i2c buses are swapped for all nv4x it seems. commit e741ac76ac45e8ea959792810670abdf2afbc6bd Author: Maarten Maathuis Date: Mon Oct 29 13:40:30 2007 +0100 randr12: Show NV_RAMDAC_FP_DEBUG_0 of both ramdac's when output is refused. commit c53754162087a7b66d40b7c3982cdc5b6d160ea1 Author: Maarten Maathuis Date: Mon Oct 29 13:34:20 2007 +0100 randr12: Minor change, turning on bit 18 of NV_RAMDAC_SEL_CLK. commit fef872bd999b430a41bdfae8d408f8dfed0010a2 Author: Stephane Marchesin Date: Fri Oct 26 15:50:40 2007 +0200 nv30 exa : add A8 dst (requires new drm), add some dst formats, handle 1x1 repeat, enable by default. commit 1d7354cbb5875eb13ed88e6aaab69aff615f82b2 Author: Stephane Marchesin Date: Fri Oct 26 04:37:51 2007 +0200 Fix nv30 exa. commit 2cd3d47713cf3eb0d36124902a2c66b689acba51 Author: Arthur Huillet Date: Thu Oct 25 16:55:18 2007 +0200 s/#define ABC ;/#define ABC do { ; } while(0);/ commit 228b91d3da445fc58bfe6b13e792bbdff3728301 Author: Arthur Huillet Date: Thu Oct 25 16:40:34 2007 +0200 NV10 EXA: implemented masking via register combiners and added (disabled by default) statistics displaying masking works for ARGB, XRGB (it's a nop), A8 - no component alpha for now, but it will come soon hopefully commit 7d77ab0692126cea9248b7ee61b001e138b4e47c Author: Arthur Huillet Date: Thu Oct 25 10:52:59 2007 +0200 NV10 EXA: fixed A8 as pointed out by p0g commit cdb54271837564bb4a5afd2dc4cf7e7169138975 Author: Arthur Huillet Date: Thu Oct 25 10:38:42 2007 +0200 NV10 EXA: added support for A8 as source commit 220acbe07e84015e489257a9b9dae8e8f1be8268 Author: Peter Winters Date: Thu Oct 25 10:03:39 2007 +0200 NV10 EXA: use proper XRGB format and use correct RC setup commit 6034387ea96816d34b627c306998342136ecb01b Author: Peter Winters Date: Thu Oct 25 09:59:01 2007 +0200 NV10 EXA: always use an even width for NPOT_SIZE commit f7f17b789e89aadafa2734cba0f4464c5f317a68 Author: Jeremy Kolb Date: Wed Oct 24 19:34:49 2007 -0400 nv30: Fix some corruption. Give us some semblance of icons (still messed up though) commit bafa34a7bd7a9136a95c8ced25c5b80556c78c0c Author: Maarten Maathuis Date: Wed Oct 24 20:28:46 2007 +0200 Fix a few NV_IMAGE_BLIT offsets. commit 09f4b66d35543037c4f50910d0265751dded66c0 Author: Arthur Huillet Date: Wed Oct 24 10:40:17 2007 +0200 NV10EXA : enabled by default for NV1x. Waiting for pq before doing the same for NV2x. commit 6c978164a212ddf4f60f82cdc72b11ebbc96efbb Author: Arthur Huillet Date: Wed Oct 24 09:37:34 2007 +0200 Xv: disabled YV12 overlay for NV20 chipsets that do not seem to have it commit 9f48bd410f8cf9aa700e03db807cef2e7cc07e34 Author: Maarten Maathuis Date: Wed Oct 24 09:08:41 2007 +0200 randr12: The last step to make lower flatpanel resolutions working. - Note: the previous randr12 commit also contained some fixes in this area. commit 2bf8bbd1764979d194ff2dbac4a6529cbcbdbc1d Author: Maarten Maathuis Date: Wed Oct 24 00:00:59 2007 +0200 randr12: Misc changes. commit 5841664c82126058e582f215af2323cc80a6150d Author: Arthur Huillet Date: Tue Oct 23 22:53:46 2007 +0200 x8r8g8b8 now seems to work (it passes rendercheck) with format 0x380 commit 27a713057f0fd2d3c8a77f5eff86393059b60511 Author: Arthur Huillet Date: Tue Oct 23 22:53:21 2007 +0200 Remove NV10 texture formats that are all wrong. commit 78a0d5a42f96816889f1c9294f11c637dfcba486 Author: Arthur Huillet Date: Tue Oct 23 19:12:19 2007 +0200 NV10 EXA: A8R8G8B8 without mask works with rendercheck now commit 7543e65cecab510f6bea8b6efc5cdf3cc9d261ae Author: Maarten Maathuis Date: Tue Oct 23 18:41:38 2007 +0200 randr12: actually add scaling factor's - This was forgotten when i first added this from the haiku driver. commit a11faa794c5dd24cd74d949006c8e86c6c3f0310 Author: Maarten Maathuis Date: Tue Oct 23 16:22:56 2007 +0200 randr12: Add a few debug statements that should explain why an output is refused. commit 2d842c39bf210b4b8294a447ec9ac194ec3e2e47 Author: Maarten Maathuis Date: Tue Oct 23 16:19:10 2007 +0200 randr12: Check if output is needed before creating it. - This hopefully fixes a problem some NV30 people have. commit 5c884a188296c9a026f110f272dff5d88d58c0b6 Author: Maarten Maathuis Date: Tue Oct 23 15:47:49 2007 +0200 Revert a G70 workaround, this may BREAK your setup. - Please report if you have issues, but be prepared to help out to find a real fix. commit 0fb93b45378c9d0fbee9bb96bd60d7a995c667b9 Author: Arthur Huillet Date: Mon Oct 22 19:02:17 2007 +0200 added (disabled) NV10 EXA hooks commit 78af60e798f5b91cc128cd2f84ddae7d0a1523cd Author: Arthur Huillet Date: Mon Oct 22 19:00:20 2007 +0200 added NV10 EXA functions to proto header commit cf053adacabaee887ecedaa9967b07b3185095b5 Author: Peter Winters Date: Mon Oct 22 17:04:23 2007 +0200 Fix some blending stuff for NV10 EXA commit b0605fbeef0765e39554a15ea9f871c19258f72c Author: Peter Winters Date: Mon Oct 22 17:01:26 2007 +0200 Add basic RC setup for NV10 EXA commit 8dea36f6533c5fc35dcb3c24e2fe0c1d009b6693 Author: Peter Winters Date: Mon Oct 22 16:50:47 2007 +0200 Fix NV10 EXA transform commit b9f97583979d7dbbf0058de9f807a97812362c71 Author: Peter Winters Date: Mon Oct 22 16:48:53 2007 +0200 Fix NV10 EXA list of supported formats commit 3dc9778902d9412c469d79e3d437838a6c9da7a5 Author: Peter Winters Date: Mon Oct 22 16:44:08 2007 +0200 Various NV10 EXA fixes commit 391e2dd1216f598538524fdc52a14bfdd4451567 Author: Ben Skeggs Date: Mon Oct 22 22:48:13 2007 +1100 Fix bad s/ commit 162edc70769b24b39f432cb79f5959eff363c564 Author: Pekka Paalanen Date: Sun Oct 21 19:55:41 2007 +0300 Revert "EXA: now caching BLIT_OPERATION and RECT_OPERATION..." This reverts commit 6d8caf5e0dd915809152c52c2c56a39d76e2ed8c. Conflicts: src/nv_exa.c The above mentioned commit broke EXA copy and solid operations on nv20 after display mode change. Triggered by running freedroid in fullscreen. (Not freedroidRPG) commit cc04b651348768dc214f59211ba396343aaa6a17 Author: Maarten Maathuis Date: Sat Oct 20 20:30:02 2007 +0200 Add back comment. commit 6fd262027c74d5008e484bb5f09edbf5f657470b Author: Maarten Maathuis Date: Sat Oct 20 00:26:02 2007 +0200 randr12: Fix critical bug and minor change. - The previous code could allow the TMDS[0x4] register of both heads to be set to 0x88. - This caused problems and freaked out my monitor and scared me as well :-( - Until further notice, the usage of this swap bit is restricted to very specific situations. commit 377f69172b4935372538d7bbf26852bf8a1fe1c4 Author: Maarten Maathuis Date: Fri Oct 19 20:57:58 2007 +0200 randr12: Some changes to crtc regs. commit 85dac1041b38127d41832a6a322d01f87d70f28f Author: Maarten Maathuis Date: Fri Oct 19 08:06:33 2007 +0200 randr12: Some changes to output detection. - I'm aware that we still can't fully program modes. commit 8320f4f61ca72ee0101c1d552ea3ef28e573d21d Author: Stephane Marchesin Date: Thu Oct 18 01:56:17 2007 +0200 Remove xaa includes. commit 86855f03945d3bad90da0b28311bece484cbfce7 Author: Stephane Marchesin Date: Thu Oct 18 01:55:06 2007 +0200 Handle -minline-all-stringops properly commit 53ee9726eeda8cf844d32a9563c82b7549ba6c03 Author: Ben Skeggs Date: Wed Oct 17 23:28:50 2007 +1100 re-align nv40exa after ring macro changes commit a5c4786aaeadce6103fcc9b3a6f4e6f25a66730f Author: Ben Skeggs Date: Wed Oct 17 23:22:05 2007 +1100 Switch to *_RING instead of NVDma* commit 0d1835d67e38ca12347a7965ac43fbfc29063b87 Author: Ben Skeggs Date: Wed Oct 17 16:29:01 2007 +1100 Switch to header generated from XML + other cleanups. - Renamed object header to nouveau_class.h, so it's not confused with mmio - Removed usage of method defines in nv_dma.h - Removed 2D engine EXA composite hook, it was broken, useless and soon to be obsoleted by card-specific versions. - Use names in a few places that magic values were previously used. commit b36fbf174849b04928efda2909c461b6c1688cdc Author: Dave Airlie Date: Tue Oct 16 09:28:58 2007 +1000 nouveau: missing include commit 8425d93f7f932281a1b9b420e676acba2a2b5ecd Author: Stephane Marchesin Date: Sun Oct 14 17:10:59 2007 +0200 more changes to nv30 exa... displays the pictures at their correct postition and size but doesn't obey the blending parameters (i.e. no blending takes place). commit 212a4cd8884ff9276f64bdd0a0115782d146e36a Author: Maarten Maathuis Date: Sat Oct 13 23:02:27 2007 +0200 randr12: Some fallback defaults for dual-dvi cards, please submit dumps when you can. commit 6d8caf5e0dd915809152c52c2c56a39d76e2ed8c Author: Arthur Huillet Date: Sat Oct 13 22:28:39 2007 +0200 EXA: now caching BLIT_OPERATION and RECT_OPERATION (0x2FC) to limit the number of IRQs on NV04 commit ceaf8df424651e0f20d63f817864f3e8657f2c21 Author: Maarten Maathuis Date: Sat Oct 13 21:45:50 2007 +0200 randr12: Revert a small unnecesary change. commit 7171acbe7f174dcbe8c28c7f1fcb7481f2fafd90 Author: Maarten Maathuis Date: Sat Oct 13 21:33:07 2007 +0200 randr12: Hopefully fix panel size detection and allow negative hsync/vsync polarities. commit be08f09ad1ba6aeebf4491727052d246d5568fd2 Author: Maarten Maathuis Date: Sat Oct 13 17:26:32 2007 +0200 randr12: Some debug stuff. commit 418e6117a25e51d5b0dbcd87c17f78f9b04fa588 Author: Maarten Maathuis Date: Sat Oct 13 12:52:10 2007 +0200 randr12: Remove a double reg and misc changes. commit e30e21ed0ee8b1bd112780badc933dc4552a2e81 Author: Maarten Maathuis Date: Fri Oct 12 19:49:32 2007 +0200 randr12: Some misc changes. commit 9425f7a3ea2d80f4df79239e1ca8000bd729c772 Author: Jeremy Kolb Date: Thu Oct 11 20:32:47 2007 -0400 Remove division in NV30EXATransformCoord. We want coords between 0 and texture width, height. commit 879e653867552f6bb595160670c8325bcb5a0d3b Author: Maarten Maathuis Date: Wed Oct 10 21:40:13 2007 +0200 randr12: Misc fix. commit 99dd9da140a5750b7c4d6030a0e8e1f7fca44033 Author: Maarten Maathuis Date: Wed Oct 10 20:08:28 2007 +0200 randr12: Fix switch mode function, so fullscreen apps don't screw up your screen by definition ;-) commit 0c80565e2cb7fa5ff359f57cc3d7016a78a43ac2 Author: Stephane Marchesin Date: Wed Oct 10 14:43:34 2007 +0200 Oops. commit 761ef4bd32f42fb77a3e48985fd6dda81aa081f5 Author: Stephane Marchesin Date: Wed Oct 10 14:42:46 2007 +0200 Try to do something for our desastrous UTS/DFS performance. My PCI card is really hurting. commit 254f9b075034c86e5936954b52f499d573880039 Author: Stephane Marchesin Date: Wed Oct 10 02:34:06 2007 +0200 Still more work on nv30 exa... now starts but nothing works. commit 79dc3ad2c89eef74596a4c8e5714c2c661f56fee Author: Stephane Marchesin Date: Wed Oct 10 00:48:46 2007 +0200 Don't swap the registers from user space. This is very wrong, and the drm will handle that soon. commit 2c08930f0b2bf06282976d6a357daf7f45405cdc Author: Stephane Marchesin Date: Wed Oct 10 00:48:06 2007 +0200 Revert back to the other shaders. The terminator instruction doesn't seem to help, and breaks the A8 hack. commit 90d2457125b059916f9b9dd5ddb7cec73a554c55 Author: Stephane Marchesin Date: Tue Oct 9 04:38:29 2007 +0200 Cleanup some nv30 exa issues, add debbuging info to the composite hook. commit e184fc802c7f9103bbc8ee8c036cfe5a9ddd9513 Author: Stephane Marchesin Date: Tue Oct 9 03:28:39 2007 +0200 Some fixes to nv30_exa. Now goes through a couple of composite() calls without crashing. commit 708f28dc8175375713cd9ae4d07b2cfcea3b106c Author: Stephane Marchesin Date: Tue Oct 9 03:28:11 2007 +0200 Use terminators on the nv30/40 shaders. commit 43d5f747103721b3e0ccf4fe73494729ed34986c Author: Maarten Maathuis Date: Mon Oct 8 19:53:56 2007 +0200 randr12: Small changes makes finding regressions easy ;-) commit 62294d9be8ec6f1159e3c2ad273649f937b7aa7e Author: Maarten Maathuis Date: Sat Oct 6 19:04:51 2007 +0200 EXA: Starting from version 2.4 we can't use devPrivate.ptr anymore. commit 721de8da8fccd6f7b6e5a3044ce82616cdb689e8 Author: Stephane Marchesin Date: Fri Oct 5 23:50:18 2007 +0200 some more work on nv30 exa... commit 80bcbf493b6b2ad927c7a3826cc6e5fa46bca76e Author: Maarten Maathuis Date: Fri Oct 5 22:29:07 2007 +0200 randr12: Misc changes. commit 63ad301baab5803329ba0f192f12c917b435594e Author: Stephane Marchesin Date: Fri Oct 5 17:20:05 2007 +0200 mandatory oops commit. commit 3773e0e5045a6c347bf26eff71e6bdd4c592cebc Author: Stephane Marchesin Date: Fri Oct 5 17:18:40 2007 +0200 Split nv30 exa and nv40 exa. commit 37900537c0f8b8dfa3dc80300d6809f1f4ac9423 Author: Jeremy Kolb Date: Thu Oct 4 23:15:38 2007 -0400 NV30 texture changes. Commented out table that appears to work for nv35 at least. commit 44eb015e88caba82c8d84fee4ea71768c9a6b2a3 Author: Jeremy Kolb Date: Thu Oct 4 20:37:35 2007 -0400 Fix NV30 arch check. commit 5a76e5a751cd66021565ab92b2fc3c61204c7d73 Author: Stephane Marchesin Date: Thu Oct 4 23:46:43 2007 +0200 fix nv30_exa for nv40. commit d7b767fc951287a1c87965db3c236e05cc94d8f6 Author: Maarten Maathuis Date: Thu Oct 4 21:49:33 2007 +0200 NV50: Fix a minor mistake. commit 926e887a4b374768648fe772b1b3d7730096e45f Author: Maarten Maathuis Date: Wed Oct 3 22:57:56 2007 +0200 randr12: Some reindenting. commit 7ee3f670077641e14a0ff024443805564d5be599 Author: Patrice Mandin Date: Thu Oct 4 19:54:40 2007 +0200 nv10exa: init link between tcl and image blit for >=nv11 commit 6510e75ada6fb921e417588d98afdc58b2c14972 Author: Jeremy Kolb Date: Wed Oct 3 23:34:23 2007 -0400 nv30_exa: send down pitch correctly commit b7fc7b63eec4a5e23879120284c882e19e095471 Author: Jeremy Kolb Date: Wed Oct 3 20:54:28 2007 -0400 Fill in some regs. commit 46facb7c29d2737f1cc414655de8e09ee0aea22f Author: Maarten Maathuis Date: Wed Oct 3 22:16:13 2007 +0200 randr12: Actually borrow some timing stuff from Haiku correctly ;-) + new crtc reg commit 9d6b3216ed519e89386480d04070b1c1e4702921 Author: Matthieu Castet Date: Wed Oct 3 21:21:41 2007 +0200 abort on Init table command not found when parsing bios commit b6351c0efa3629a448557fc7e597ba14fbf057fb Author: Maarten Maathuis Date: Wed Oct 3 19:54:40 2007 +0200 NV50: Fix small mistake. commit f00ea962e549934965100dc6d2806709129995e8 Author: Maarten Maathuis Date: Wed Oct 3 19:03:38 2007 +0200 randr12 + NV50: Misc changes. - Switched everything back to head variable instead of pcio. - Removed ugly C() macro for NV50. - Probably some other stuff too. - Please test NV5x users ;-) commit 5d7f06c2a205004b78692b4dfad477ade3947fd5 Author: Maarten Maathuis Date: Tue Oct 2 19:31:10 2007 +0200 NV50: Some infrastructure and minor changes. commit 7680648c9ddb81e2a1c6295431d7074366fb7b9a Author: Maarten Maathuis Date: Tue Oct 2 16:55:07 2007 +0200 NV50: The beginning of moving some code around. - Please contact me if something breaks, i don't have an NV5x card. commit b812ecaef75216cc76557e6a3a080efa870d189f Author: Maarten Maathuis Date: Tue Oct 2 10:35:37 2007 +0200 NV30EXA: Minor code cleanup. commit 5b974b3bed86d53ccf6915c46aeb310a2bba3519 Author: Stephane Marchesin Date: Mon Oct 1 17:19:41 2007 +0200 Remove those useless casts. commit 734785c05357455a7ae7181fd7326f02e0bd0478 Author: Stephane Marchesin Date: Mon Oct 1 02:01:30 2007 +0200 Some work on nv10 exa. commit 0cfcbcc9d64bfd47b765f4f2ff41d7ad467990bd Author: Matthieu Castet Date: Sun Sep 30 22:41:15 2007 +0200 enable bios PINS parsing on post nv10 cards commit 9ac6d0e9d46abdf867fd06f7ca24b718a340bb74 Author: Pekka Paalanen Date: Sun Sep 30 23:09:33 2007 +0300 nv20 inits nv10tcl. commit c3439aebe000a3c6dd716bd4739ce01321bee26b Author: Maarten Maathuis Date: Sun Sep 30 21:43:41 2007 +0200 Fix detection of NV04 and NV05 when using pci-rework. commit 9e070c09f68b0e3e41109360c455b3e2cc1dbb9c Author: Maarten Maathuis Date: Sun Sep 30 19:37:53 2007 +0200 randr12: Switch to proper cursor api + minor changes + small nv50 fix - Remove the ugly cursor hack. - A function name on nv50 was still g80, this caused an implicit decleration. commit c7de03556a51009ce69b073d7ecfd01919d24c5f Author: Maarten Maathuis Date: Sun Sep 30 12:40:50 2007 +0200 NV30EXA: Fix rendering issues. - Full non-composite tests using rendercheck. - The full test suite of composite has been run (it takes ~15 hours). - A partial cacomposite has been run (a8, a8r8g8b8, x8r8g8b8). - All the rendering tests work fine, as does the cairo-knockout demo. - Thanks to doener for reporting this issue. commit 0574dc5a8bd09216e41656c943c92a437bcf020b Author: Stephane Marchesin Date: Sat Sep 29 20:20:47 2007 +0200 A little more work on nv10 exa. commit aae155f149af522e1a83168ed4ec5b7b5a40bb73 Author: Maarten Maathuis Date: Sat Sep 29 17:28:10 2007 +0200 randr12: Some more CRTC regs. commit 8b03086ee36ca9c3e77f9df9eb4a1805072d2103 Author: Maarten Maathuis Date: Sat Sep 29 14:33:38 2007 +0200 randr12: make crtc owner beheaviour consistent with blob + minor changes commit c572315a974746224dba1213e956a1b6189f175d Author: Maarten Maathuis Date: Fri Sep 28 19:58:39 2007 +0200 randr12: Minor changes and more debugging output. commit db36a1df4ba3dddf3669f491bfe6461e4fd490b6 Author: Maarten Maathuis Date: Fri Sep 28 18:52:47 2007 +0200 randr12: Some more debugging output and some other minor changes. commit 2d21405c9c0bff36f47e28a5e4086dacae557d17 Author: Maarten Maathuis Date: Fri Sep 28 13:02:51 2007 +0200 randr12: Preperation to allow differentation between crtc and pcio in the future. commit f235cf896c1da098ff635d5a882561f16f5fb4f7 Author: Stephane Marchesin Date: Thu Sep 27 18:21:25 2007 +0200 Import nv10 exa code. Does not work yet. commit 74a243d8e188432854c15849ad1ab6e893e48f0a Author: Maarten Maathuis Date: Wed Sep 26 09:04:43 2007 +0200 randr12: Some registers were not constants, so changed it back with a note about the values i had seen. commit a3c71495097f8d745110aad06c9f8e2f6f03baad Author: Maarten Maathuis Date: Tue Sep 25 22:05:36 2007 +0200 randr12: Getting closer, doing a decent attempt at detecting the proper ramdac. - crtc detection may need to be added - When i boot up with both analog and dvi attached, my dvi is misprogrammed, maybe some sharing is happening somewhere? commit 4a57ebb9c2098ce5d568a2da6e0640cdcc18d3f0 Author: Maarten Maathuis Date: Tue Sep 25 21:06:03 2007 +0200 randr12: detection changes, more to come later - Detection of crtc's (or a way to program them effectively) still has to be done/fixed. - It now detects if tmds controls is on another ramdac, for a non-dual dvi setup. - Disabled TMDS stuff for the moment, since it's incorrect somehow. - Dumps for single dvi, dual dvi, etc are appreciated. commit 41df1ff6ce348dc6a38e34ce63887c66b0300c09 Author: Maarten Maathuis Date: Tue Sep 25 16:22:38 2007 +0200 randr12: Some reindenting. commit 334a9b065c4973876b894eadf05ce86ca439e08f Author: Maarten Maathuis Date: Tue Sep 25 16:17:03 2007 +0200 randr12: simplification of the swapped i2c buses "workaround" for nv4x and some code reindenting commit c40559dacd6fe20f35710111330dee37fb2b9a58 Author: Maarten Maathuis Date: Sun Sep 23 01:16:55 2007 +0200 randr12: A frail attempt at some TMDS programming, i have not figured out yet what switches DVI from one RAMDAC to the other. commit 50bc4dbff6df80f5d80f731fe42b06768c429dc0 Author: Maarten Maathuis Date: Sun Sep 23 00:57:18 2007 +0200 randr12: Reindenting and minor changes. commit 1a3303a5b6f261470c1c56f84f9b89af950a0b35 Author: Maarten Maathuis Date: Sat Sep 22 22:21:37 2007 +0200 randr12: Assign ramdac's in more sane way, otherwise two outputs may end up on the same ramdac. commit 9b62853dfe1664f3624f6be8aa86f6ca15ca8392 Author: Maarten Maathuis Date: Sat Sep 22 20:37:32 2007 +0200 randr12: Add some TMDS register infrastructure, so i can try some stuff in the future. commit a445cbcd71ee51d7680a6eb13cd037e06d59b0ba Author: Maarten Maathuis Date: Sat Sep 22 15:55:01 2007 +0200 randr12: hardcoding a base value for debug_0 register. commit f89a3217a18261424e212b355628ce5f5b8b9142 Author: Maarten Maathuis Date: Fri Sep 21 23:16:40 2007 +0200 randr12: Some minor changes. commit 5ea59f8f01dbbd4211aee1a0e46a6529cddc016c Author: Maarten Maathuis Date: Fri Sep 21 20:37:50 2007 +0200 randr12: hardcode more stuff + actually use a few previous regs - I'm going to try and do full modesetting, before tackling the problems that dual monitor setups face (i tried, it wasn't pretty). commit 0f66c725cb8806041faddf94cceb88009f7e9f06 Author: Dave Airlie Date: Thu Sep 20 08:29:43 2007 +1000 make distcheck pass commit bf3cceeb9c72c16a1cf601c8bd0dedd8c676e3cd Author: Maarten Maathuis Date: Mon Sep 17 18:58:31 2007 +0200 randr12: Add some CRTC regs which the blob sets. commit e3346c17c1ead9619206a5921994b5de7ed72739 Author: Maarten Maathuis Date: Mon Sep 17 18:22:53 2007 +0200 randr12: Change the way CRTC register 59 is set. - Small change, i suspect the original author had a dfp on the second crtc. - Information taken from a mmio trace of a nv28. commit 6c3aee9d68abdb9303cf03d9024f4f0deb5f1e3e Author: Ben Skeggs Date: Mon Sep 17 00:01:22 2007 +1000 Make use of exaGetPixmapOffset(). commit 4d9e5d451a516aac64c4b95a4478f252c0c9eafb Author: Ben Skeggs Date: Sun Sep 16 23:54:14 2007 +1000 Fix randr1.2 with driver-provided modes (ie. LVDS native mode) Fixes #12368 commit 5e58e74435da2071aab9f21386ac6dabdf3f8e7d Author: Ben Skeggs Date: Sun Sep 16 19:44:42 2007 +1000 NV30EXA: Support PICT_a8 dest surfaces Quick hack for now, shader handling will get a rework one day and clean all this up. commit d6d3ddfafaece2d153942ec37e3a8c3127c7c3fa Author: Ben Skeggs Date: Sun Sep 16 19:01:21 2007 +1000 Bring G8x modesetting into sync with xf86-video-nv. commit 53e35a1d42070fd683eb6d0f169ac3ce482a45b0 Author: Ben Skeggs Date: Sun Sep 16 17:27:22 2007 +1000 Move G80 VBIOS download into nv_bios.c. commit 02d5443d650e628256ebed1f331ad32b0f642a38 Author: Arthur Huillet Date: Fri Sep 14 21:13:26 2007 +0200 Xv: NV04 overlay now has one "colorkey" attribute It is technically possible to add other attributes, by copypasting vidix code from mplayer, but it requires a bit of work especially on testing, so I am going to do it only if someone actually needs it. commit 0fa042e5123066ca2d9f36ff05cf64d0f68289a8 Author: Arthur Huillet Date: Fri Sep 14 14:17:32 2007 +0200 Xv: NV04 overlay now almost functionnal commit a7ff07f4ab63c5a574e549326f022f84c2b9bab4 Author: Arthur Huillet Date: Fri Sep 14 12:27:45 2007 +0200 Xv: quick imple of NV04 overlay commit 77076a391647f6111f38b10192d78ea7560ca9e8 Author: Stephane Marchesin Date: Wed Sep 12 13:10:37 2007 +0200 Try to fix nv4/tnt1 commit e03fabf7e14599c1619673430ec3d81db90ffd1e Author: Maarten Maathuis Date: Tue Sep 11 20:15:11 2007 +0200 Restore PRAMDAC1 and PCRTC1 offsets, in proper format (0x2000/4) and move them to their proper place. commit 564043d9c1c9b5023f1b69a79d80742338256b83 Author: Maarten Maathuis Date: Mon Sep 10 21:14:24 2007 +0200 Fix the memory adresses of the secondary crtc and ramdac + minor change. commit 5672842381a0e8f605cd482563c5d9ea02118647 Author: Patrice Mandin Date: Mon Sep 10 18:29:10 2007 +0200 make randr 1.2 disableable for xorg server < 1.3 commit 2778d3a9041ec62462585f623f84b1ed05b4ebad Author: Patrice Mandin Date: Mon Sep 10 18:28:03 2007 +0200 make randr 1.2 disableable for xorg server < 1.3 commit 9bb76f73f0e648fda79a9c84d6e78469752a4eac Author: Maarten Maathuis Date: Mon Sep 10 07:57:01 2007 +0200 randr12: Apply the same pll restrictions to a two stage design. commit b46bbab81cc9f2093406fa8b53f9ffaf6f7fd0fc Author: Maarten Maathuis Date: Mon Sep 10 07:25:25 2007 +0200 randr12: Some more pll timer restrictions from Haiku. commit 77b424415d31428e113688ec27a405966d5080e6 Author: Maarten Maathuis Date: Sun Sep 9 23:11:18 2007 +0200 randr12: Some new pll clock restrictions. commit a1bbfe17cc863497644b58a9206ae3876a654543 Author: Maarten Maathuis Date: Sun Sep 9 22:50:19 2007 +0200 Some reindenting. commit dedb80d1237ebacf7531938d627328a20453d5d0 Author: Maarten Maathuis Date: Sun Sep 9 22:12:38 2007 +0200 randr12: Modify dcb table for nv40 to make it look more like older hardware. Remove i2c bus reverse. commit 7aa96706ef428f1390d8e938d0bf26b8c046a67d Author: Maarten Maathuis Date: Sun Sep 9 20:40:20 2007 +0200 Removed a check for > 1 active crtc's, which seems really odd, no regressions occured for me. - I cannot plug in both the vga and dvi connector of my monitor, so i cannot verify that. - The seperate monitor, on crtc0 and crtc1 work as they did (the afp being slightly shifted). - Also did a little reindenting. commit 228cb177242be053595b959dedfdfbfa77b1c54f Author: Maarten Maathuis Date: Sun Sep 9 18:49:08 2007 +0200 randr12: Fix a nasty regression from 6ba7e83f8cf2da84dd796ee59cb24c15bc36b746. - For non-flatpanels the dac was never initialized. commit 0a56cb53d7e2a58ef044c9090f48fd46a2b251d9 Author: Maarten Maathuis Date: Sun Sep 9 00:39:41 2007 +0200 randr12: Some missing init. commit 80bfb59a1f1c3968f8e116de611975df4acf5618 Author: Maarten Maathuis Date: Sat Sep 8 23:58:39 2007 +0200 Some much needed indenting and readability improvements. commit 20aaacdd69c427f6c75b80f5f206b3187262f3a1 Author: Maarten Maathuis Date: Sat Sep 8 21:00:45 2007 +0200 Hopefully fix bridged chip detection, such as the 7800GS AGP. commit b99549b5cb4f4cba2d60292994b2db483f8d4a8c Author: Maarten Maathuis Date: Sat Sep 8 14:25:39 2007 +0200 Remove list that is incorrect and no longer needed. commit b104314a11d3cf3bbb23095ff5805c1d5d8d9775 Author: Maarten Maathuis Date: Sat Sep 8 14:24:54 2007 +0200 Fix previous commit, change the way Digital Flat Panels are programmed. - This is mostly taken from the Haiku driver. - It works for me, please report any issues you have. - Added scaling, for running 4:3 resolutions on widescreen for example (untested). commit a867d6912456a5e15a965050bf784687e60634ee Author: Maarten Maathuis Date: Sat Sep 8 00:46:20 2007 +0200 Hopefully this will improve flatpanel beheaviour. - I looked at Haiku and copied as much as i could. - I removed the htiming and vtiming hack. - Let's hope this works. commit 6ba7e83f8cf2da84dd796ee59cb24c15bc36b746 Author: Maarten Maathuis Date: Fri Sep 7 23:11:51 2007 +0200 Some more indenting and such. commit d8d835dfb61f437b3eaede0a6ec188835f6eb5ff Author: Maarten Maathuis Date: Fri Sep 7 22:58:56 2007 +0200 Removing a hack in the hopes of finding a better way and some minor changes. commit dd7e568bd101ed34dcdd7d3b6d8169740499d593 Author: Maarten Maathuis Date: Fri Sep 7 22:27:18 2007 +0200 Indenting and minor code style changes. commit 8ed256e94dae884b723c172320b15b0d6ae00a1b Author: Maarten Maathuis Date: Fri Sep 7 14:37:40 2007 +0200 randr12: Fix Xv blitter, please see full changelog for other implications. - Removed the CRTCNumber variable, instead just store if crtc0 and crtc1 are active. - This coule break old style setups (picking the wrong monitor for example). - CRTCNumber option temporarily removed, because the piece of code surrounding it is ugly and needs a rework. commit eb595b2e06ec33d5fe2af43bc7ad107f551d99c8 Author: Maarten Maathuis Date: Thu Sep 6 23:26:04 2007 +0200 randr-1.2: Only the NV4x have these busses swapped, revert for other architectures. - If this is also true for NV3x, then please contact me. commit dd146a074fff575e18a41fd386e2a2ce5a1ec847 Author: Maarten Maathuis Date: Thu Sep 6 22:55:16 2007 +0200 Update to latest pciaccess api, this now requires libpciaccess 0.10.0. commit 2f37d249605d6f0b80c290683ca35a9431b1078c Author: Maarten Maathuis Date: Thu Sep 6 22:00:30 2007 +0200 NV03 is dead. commit 2b2c99f2581e14d3f1726bc503951b1fe7760c95 Author: Maarten Maathuis Date: Thu Sep 6 21:43:55 2007 +0200 randr12: A prettier hack for DFP, that uses several known htiming and vtiming modes. - I'm fairly sure these modes are tied to resolution. - You are welcome to supply additional modes. - A final solution would be even better ;-) commit ab7d37d61a01afea218321d336664896eaec5651 Author: Arthur Huillet Date: Thu Sep 6 15:09:59 2007 +0200 Xv: work around some overlay bandwidth problems - may break things though extended nv workaround from bug 4686 in nv_dac.c to NV1x and NV2x cards, as it fixed problems on someone's NV17. commit f46bc9ad5558e9ecf6b174e24658e812811e94ce Author: Maarten Maathuis Date: Wed Sep 5 23:15:57 2007 +0200 Workaround for flatpanels. - Remove the hardcoded values. - This should probably work for monitors whose mode is already set. commit 1825031465fc614a2a972ea7a8b44ac32aed8ea3 Author: Maarten Maathuis Date: Wed Sep 5 21:23:25 2007 +0200 Fix minor issue. commit c0eefaba3f3706eb62e172bc57ac2795bd16bd56 Author: Maarten Maathuis Date: Wed Sep 5 21:22:05 2007 +0200 Fix DPMS, patch by Bernhard Kaindl . commit d4dda6bb972003394dd3326f8042fa5b759fc063 Author: Maarten Maathuis Date: Wed Sep 5 21:18:51 2007 +0200 Reverse DDC1 and DDC2. - I need them reversed. - I also found this: http://www.mail-archive.com/git-commits-head@vger.kernel.org/msg12009.html - If you need them otherwise, then please consult and hopefully we can solve it. commit da89d5a4e0263b8a89c549496d10ab43c7f1f230 Author: Maarten Maathuis Date: Wed Sep 5 21:16:11 2007 +0200 randr-1.2 clients have their DDC's initialized elsewhere. commit 2b209d9b8907fad3bbaaf7297ff3c340f086bd24 Author: Maarten Maathuis Date: Wed Sep 5 21:12:18 2007 +0200 LockUnlock CRTC's when going back to X (EnterVT), patch by Bernhard Kaindl . commit ba7aa5a1525bb3fbc2ae7fbfce3e42a2e945626e Author: Maarten Maathuis Date: Wed Sep 5 18:34:27 2007 +0200 Xv: improved behavior with NoAccel by disabling blitter and overlay commit 42fd07b25d9abece034e972d1413b31fe78eba4c Author: Ben Skeggs Date: Wed Sep 5 11:58:19 2007 +1000 Avoid segv. commit e74d58cdef0a0ee561e54f3d7d8afdfac030e04a Author: Ben Skeggs Date: Wed Sep 5 11:53:48 2007 +1000 Add back DCB table parsing from randr-1.2 branch. commit 8a12758a6b7e4313e0269585b6f64eaffe54c74c Author: Ben Skeggs Date: Wed Sep 5 11:51:35 2007 +1000 Add LVDS panels when found in DCB. commit b01fbbfa2b6be793dbd7695624699ed703ff4424 Author: Ben Skeggs Date: Wed Sep 5 11:34:14 2007 +1000 Nuke "comparison is always false..." warning. commit c53f495ec2994f013406d9738764ab124ed32061 Author: Ben Skeggs Date: Wed Sep 5 10:59:33 2007 +1000 Fix pciid detect on non-8800 cards (they identify as NV8x, not NV5x) Actually, use 0x88000 to get pciid on all chipsets since NV40. Evidence suggests that NV4x has both the 0x88xxx regs and 0x18xx regs. commit 17d1a65296ae4fc524a31cd2fe4b9fe9484c0452 Author: Ben Skeggs Date: Wed Sep 5 10:53:32 2007 +1000 Neither overlay or blit adaptor work on G8x. commit 9ce1115cf26a7cf3212fdf446b83907f5795913a Author: Ben Skeggs Date: Wed Sep 5 10:51:02 2007 +1000 Nuke riva_* files + bits of NV03 that got reincarnated. commit 51e3855104ff7b68559d37e774cde306424399d8 Author: Ben Skeggs Date: Wed Sep 5 10:46:54 2007 +1000 Nuke XAA. commit a0cb03d32b0da3ad2b4e03515926a3d616f487bc Author: Maarten Maathuis Date: Tue Sep 4 23:04:59 2007 +0200 There was some redundant arch checking code left, must have been missed after the randr merge. commit 13aeb83d7e5fcb3f051f95faa48a2ef636d67260 Author: Maarten Maathuis Date: Tue Sep 4 22:54:28 2007 +0200 I unknowingly broke CRT's i think. commit a57fe870e49b2e7a3a6caf17fea8c8c878b1024c Author: Maarten Maathuis Date: Tue Sep 4 22:17:52 2007 +0200 Fix several problems. - Implement a temporary NVSetMode(), suggestion and code by: Bernhard Kaindl . Note: I've left his comment intact, just a minor change. - Fix last commit, i forgot to store the last mode. - Move some header stuff around, which i created without knowing the true purpose of nv_proto.h ;-) commit be468069b752a3a3c595f62eda40e90719496c90 Author: Maarten Maathuis Date: Tue Sep 4 20:45:25 2007 +0200 Fix DVI/Digital alligment with some (magic?) values discovered by Matthew Garrett . commit 9b54aac1641c35ca3edb2d208d8a9a9c55737894 Author: Maarten Maathuis Date: Tue Sep 4 20:20:02 2007 +0200 Add suggestion/patch by Bernhard Kaindl and fix NVRestore(). - When a monitor comes out of DPMS sleep the old state is restored. - NVRestore() was not ported from randr-1.2 branch, meaning that going from X to console resulted in a unacceptable mode. commit 186e3b551da6e37e0860d06e9265c3c58b763e27 Author: Maarten Maathuis Date: Tue Sep 4 18:39:44 2007 +0200 Add pci-rework compatibility and remove the pci-id list. - Detection is now done by probing registers. - Nforce/Nforce2 have special "hacks", they have been converted, but are untested. - NV5x detection should work, but has seen no testing. commit 6f860d6d7637f4eba49636d2bab11ac62da279cf Author: Ben Skeggs Date: Tue Sep 4 11:38:39 2007 +1000 Badness. commit 54ca35e7379fdb375c0d7be2f8c3ad1513be2a59 Author: Ben Skeggs Date: Mon Sep 3 22:06:27 2007 +1000 Hook up G8x accel code. commit 8c6fd43f63e450acbfd8de563580db715e72f90b Author: Ben Skeggs Date: Mon Sep 3 22:06:09 2007 +1000 Fix G8x hwcursor commit f6d0e2dde750436090f50f442e3e0f1aa226a7f3 Author: Ben Skeggs Date: Mon Sep 3 21:35:06 2007 +1000 Fix up M2MF code for G8x. commit fe8788da8cd5aca90bbe85e1c975b79ff02682ba Author: Ben Skeggs Date: Mon Sep 3 21:30:16 2007 +1000 Init grobjs we need on G8x. commit f450a4d0ab9c199db0a99a5e78642a0d66ca2915 Author: Ben Skeggs Date: Mon Sep 3 21:14:59 2007 +1000 Allocate some VRAM for G8x palette. commit dfcfc9a8dac7f1fdf7b31ebdd2bca4c5f33313e0 Author: Ben Skeggs Date: Mon Sep 3 21:06:15 2007 +1000 Don't pound PMC_ENABLE on startup, this is DRM's job. Fixes resetting parts of G8x that shouldn't be. commit c8a32bab34ffe74a4a323de3885fae87bd916218 Author: Ben Skeggs Date: Mon Sep 3 21:05:08 2007 +1000 Detect unknown G8x chips in NVProbe() + Force randr12 on G8x commit 087fd31d6a5d6a835fc318549a12faa916cf5ed0 Author: Dave Airlie Date: Mon Sep 3 16:46:34 2007 +1000 fix distcheck after randr-1.2 brought in commit 423f95c02c5b018d6fc78f501ea8b546d73f33fb Author: Dave Airlie Date: Mon Sep 3 16:42:32 2007 +1000 merge in the nv50 code from the randr 1.2 branch commit 39e8b6c7c0a159f7a06d786abc0f9c5c30ae6793 Author: Dave Airlie Date: Mon Sep 3 16:22:47 2007 +1000 randr12: bring over cursor code commit ffd87b9bf2712311e197063a7c4b65b7c06a5ebb Author: Dave Airlie Date: Mon Sep 3 16:16:43 2007 +1000 randr12: try and bring over most of the randr12 code into master This currently shouldn't break master, and randr-1.2 more than likely doesn't actually work commit 708cb3f9d9b66030dd0bfc4beaa657fcdfded784 Author: Arthur Huillet Date: Fri Aug 31 21:42:51 2007 +0200 Xv: now "powercycling" the overlay using PMC_ENABLE in order to workaround some lock up problems commit 3a5095db9fd33d2feeb4bcfcb8453848998ce04b Author: Stephane Marchesin Date: Fri Aug 31 01:36:11 2007 +0200 Cut remains of NV03 support. commit 54a922dffaa1a4a7c29c2cf8d73449e6fe9058ed Author: Stephane Marchesin Date: Fri Aug 31 01:35:52 2007 +0200 Good ol' nv04 doesn't seem to support dithering type for SIFM. commit b680a9652508e0d0e660fd924937870cd4ad3530 Author: Arthur Huillet Date: Thu Aug 30 23:19:13 2007 +0200 Xv: nv NV30 arbitration patch for DFPs - fixes (some?) bw problems commit f29c9d3fc95f160ded64b911266a479ad0a2fc1a Author: Stephane Marchesin Date: Wed Aug 29 00:56:05 2007 +0200 Silence more warnings. commit 663bf3cbc881631e54e7d63932f59e1b4289979f Author: Stephane Marchesin Date: Wed Aug 29 00:20:19 2007 +0200 Silence a warning. commit 5937ace8e2376f0b38249e38d1f97a88f4aca99f Author: Stephane Marchesin Date: Tue Aug 28 11:18:54 2007 +0200 Some small fixes. commit 30d2b3f0dd4a9b6be5d3ad7a38993e5b84baf90a Author: Stephane Marchesin Date: Thu Aug 23 22:54:53 2007 +0200 Oops I forgot to enable the nv35 object on nv35... commit 761d25eb9afd4be34f00ff7eff44dac0c2f9d9a2 Author: Stephane Marchesin Date: Thu Aug 23 22:39:34 2007 +0200 Don't enable that non-working nv30 EXA just yet. commit 7e7dec68969db5dfc15a96f6dbe78edbcd1ab5ab Author: Stephane Marchesin Date: Thu Aug 23 22:33:58 2007 +0200 Add nv30 exa code, doesn't work yet. commit e1789ac879d474015c87c36847f0e2777486d6ad Author: Stephane Marchesin Date: Wed Aug 22 22:38:08 2007 +0200 Update to the latest nouveau_reg.h file. commit 4fc38aeb54da1cba9d5e231a5151e6429af0ea7e Author: Matthieu Castet Date: Tue Aug 21 20:03:15 2007 +0200 init SCALED_IMAGE_FROM_MEMORY_COLOR_CONVERSION to COLOR_CONVERSION_TYPE_DITHER. riva and blob init it. commit c63b5a3740dfe3ff5c6b77412e3666a79137c120 Author: Arthur Huillet Date: Tue Aug 21 14:55:07 2007 +0200 Fixed blending in NVComposite by actually creating ContextBeta1 and 4 objects commit 4454b3d0104c14890b9b5ead33823ffd0362315c Author: Arthur Huillet Date: Mon Aug 20 23:50:43 2007 +0200 Workaround for NVComposite bug #11949 Accepting only PictOpSrc now, PictOpOver causing crashes. commit 450cf465bd850d70ba00711e7d85e1e8eaeff502 Author: Arthur Huillet Date: Mon Aug 20 13:18:15 2007 +0200 Xv: fixed wrong fix of big endian NV12 copy, as explained by Daniel Chassot No sleepless coding ever again. <- big lie commit b8780f3d9a7e1bd596f19cbb4867908dd6be53db Author: Arthur Huillet Date: Mon Aug 20 00:46:00 2007 +0200 Xv: patch from Daniel Chassot to do 32bit memory writes in NVCopyNV12ColorPlanes, with minor reworking of mine commit 690b4b7adc02e7170451bd5bc6616bc03823d2e5 Author: Arthur Huillet Date: Sun Aug 19 21:04:00 2007 +0200 Xv: this time a correct matching ? :) commit 6b6505bc54fc29fc22d2be995f34b24ddb9dbaa0 Author: Arthur Huillet Date: Sun Aug 19 20:18:49 2007 +0200 Xv: stupid build error. blah. commit 28b66c26f85f7730ba9a54db3f778bfd5b24915d Author: Arthur Huillet Date: Sun Aug 19 20:14:38 2007 +0200 Xv: now catching all chips < NV17 commit 94ae23e2a1ebe9fe3f4bf2039ad3b2e5f0e3e55e Author: Arthur Huillet Date: Sun Aug 19 19:50:57 2007 +0200 Xv: disabled YV12 overlay on < NV17 commit 1f83dbfa6ef774276e0da3dc4f74cbabe55f76fa Author: Arthur Huillet Date: Fri Aug 17 00:45:47 2007 +0200 Xv: now only uploading the part of source image that is used to workaround bandwidth problems commit ffa62dc9f573448c8a832324973513f7ba985b33 Author: Arthur Huillet Date: Wed Aug 15 02:27:15 2007 +0200 Cleaner DMA and CPU copy upload of color plane for overlay. Fixes a few bugs and is necessary for the bandwidth problem workaround. commit 72c62c48faf2acffae1bc311051537f46972ae58 Author: Patrice Mandin Date: Tue Aug 14 22:26:02 2007 +0200 Can not setup dfp modes with xorg-server-1.1 also commit 54eadd624282ae6d0493a3cb54afda7600c7e12b Author: Patrice Mandin Date: Tue Aug 14 22:12:37 2007 +0200 Make NV30 EXA disablable commit da6522dbfcff10081ce2ad866ba4afdd18b413d7 Author: Arthur Huillet Date: Tue Aug 14 15:01:58 2007 +0200 Implemented CPU copy fallback for native YV12 overlay as well. commit e417307bfc8610f010728d82bb80fced7d0d7aa7 Author: Arthur Huillet Date: Tue Aug 14 14:47:44 2007 +0200 Native YV12 overlay support, cleaner Xv code Added support for a native YV12 overlay for NV10+ cards. Cleaned up Xv code. CPU copy fallback does not do native YV12 yet. Bandwidth problem with YV12. commit b3e4da61e8c846d8d3f01b015cb7155da4e266e9 Author: Carlos Martín Nieto Date: Sat Aug 11 12:27:56 2007 +0200 Don't dereference the uninitialised pNv->FB pointer. Opps, mea culpa. I accidentally commited a debug statement that dereferences an uninitialised pointer. commit a7d29fd932fe14ca5b82cda64341704164fb7e09 Author: Carlos Martín Nieto Date: Fri Aug 10 20:01:27 2007 +0200 Move DRI's FB to the higher half of VRAM If we don't, DRM's push buffer takes the map away from us. This is just a workaround until that useless DRI mapping goes away. commit f9e7849e8736d1477e867a3a1262f0f82baefbcb Author: Arthur Huillet Date: Fri Aug 10 20:59:12 2007 +0200 of course, we don't need Xv debug messages upstream commit ee098308c6aca7615bc6ee597e9ca531520f4119 Author: Arthur Huillet Date: Fri Aug 10 19:28:22 2007 +0200 "smart" overlay with composite, and a tentative fix of flickering when dropping frames added "smart overlay" when composite is enabled. not extensively tested now not changing the buffer when skipping frames - not sure it even has an impact. commit 9cc11cbc9ce10b99f72d7c08584bf61d179cdbda Author: Ben Skeggs Date: Thu Aug 9 01:46:46 2007 +1000 Fix blit adaptor + NVWaitVBlank() (ie. replace old magic 0x400710 hack) commit 22e965038bbc64c7b5118fc9a03ecf5f467d1665 Author: Stephane Marchesin Date: Wed Aug 8 12:45:59 2007 +0200 And of course, we don't really need these ErrorFs. Not any more than last time. commit 1c5a3e271ca628c61fce0bcef24d0bda15d06d7e Author: Stephane Marchesin Date: Wed Aug 8 12:38:14 2007 +0200 Oh, and of course we need a new object. commit a9b44880924ee814f0bd4b0015ab152f2d0b9051 Author: Stephane Marchesin Date: Wed Aug 8 12:37:31 2007 +0200 Use IMAGE_FROM_CPU for smaller UTS. This is async and helps a lot with glyphs especially. commit 9cb4c95a4fbf38fcb8249e765ff71b2e24912244 Author: Ben Skeggs Date: Wed Aug 8 11:48:48 2007 +1000 Add forgotten prototype commit 089ff874e0a798c1b2693b5ab01cfa04c939f758 Author: Ben Skeggs Date: Wed Aug 8 10:32:08 2007 +1000 Timeouts got lost somewhere along the way.. commit 15a680a6d715f87910b07133e1b98e2d4be9a108 Author: Stephane Marchesin Date: Tue Aug 7 23:42:09 2007 +0200 Fix some overlooked hardcoded subchannels... commit 3f0a52c44697787d84d4f2cf61872e13d630f62a Author: Stephane Marchesin Date: Mon Aug 6 23:30:04 2007 +0200 Fix explicit usage of subchannels, and remove subchannel defines altogether. commit 17b65f1c2be3e609d387066d72dd8f5f7e2a6d92 Author: Matthieu Castet Date: Mon Aug 6 19:55:52 2007 +0200 make XAA work on nv17 commit bf5684ba05e50fe48cce1e634caf03725fa907e6 Author: Stephane Marchesin Date: Mon Aug 6 18:27:19 2007 +0200 oops remove the debugging ErrorFs. commit 2eab8453d73b4e7fd6c2d772d595ca5e8fa8c975 Author: Stephane Marchesin Date: Mon Aug 6 18:13:13 2007 +0200 Fix UTS/DFS when gart is smaller than transfer size. commit 71076f7947ec71b7a0f363c23f5a51ef09cdd9c0 Author: Ben Skeggs Date: Tue Aug 7 02:09:30 2007 +1000 NVDmaSetObjectFromSubchannel doesn't exist anymore! commit 4a6c2bd2bedcf9f67c3e498b1411ad58cd66199c Author: Stephane Marchesin Date: Mon Aug 6 17:29:41 2007 +0200 Add subchannel handling code commit c7b16f6fff96ffcff8049feed2a0a9ef6de4209f Author: Ben Skeggs Date: Mon Aug 6 21:45:50 2007 +1000 Adapt to drm 0.0.10 commit 59d073c9b01a8f61675b8d74f5c55f134ddfb8bf Author: Arthur Huillet Date: Mon Aug 6 01:37:17 2007 +0200 Xv: waiting for last transfers to finish before freeing notifiers commit 047aa7e0a6ecce59c9be8d36c51f082c0ddaafe8 Author: Ben Skeggs Date: Mon Aug 6 00:13:35 2007 +1000 Clamp NV_MEMORY_TO_MEMORY_FORMAT_LINE_COUNT to 2047 lines. At least NV40 can't handle values larger than this. Patch fixes hang when visiting opengl.org. commit 68070ff7309c8b1cf149490dcf537892b0bf44bc Author: Arthur Huillet Date: Mon Jul 30 03:27:55 2007 +0200 Xv now manages a pool of 6 notifiers for its double buffering. Also improved the memory allocation logic. commit bceef7210e44dc643a5f7a5b5683d5c434a07565 Author: Arthur Huillet Date: Sun Jul 29 01:54:42 2007 +0200 Xv: fixed bugs when displayed only part of the source image (tvtime with overscan) Xv: fixed bug with overlay behaving badly when window is partly out of screen Xv: clean up NVPutImage code commit a434cc6dd38ffa658bacf5547ec0384ec5351e9c Author: Dave Airlie Date: Sat Jul 28 16:34:49 2007 +1000 exa: disable compositing on big endian machines This at least allows me to see what I'm typing for now :-) commit 175c2b7f7b9e45924e1de7f7ea1130fda76d02fd Author: Arthur Huillet Date: Fri Jul 27 19:50:48 2007 +0200 no overlay with composite commit 200c82634971b32bcc308b5e07c113048cf37aa4 Author: Arthur Huillet Date: Fri Jul 27 16:28:48 2007 +0200 Xv: gotos are evil commit 8544c0ce64780be1bb5f66550d9d6f1e3c87865c Author: Arthur Huillet Date: Fri Jul 27 15:11:31 2007 +0200 Removed comments in preparation for upstream push commit 2c60f6e9fc5260eba5b7ddaf67085781f35bfc58 Author: Arthur Huillet Date: Thu Jul 26 23:00:02 2007 +0200 re-added CPU copy code as a second fallback option (first being to use X GARTScratch) commit c4dfad770d380dacddd3be025e96f867047327cc Author: Arthur Huillet Date: Thu Jul 26 16:08:40 2007 +0200 Completely removed notifier wait, when using double buffered stuff. I may have to re-add a wait in the future, but I'm not sure yet how to use two notifiers at a time. commit 8ba6a9c7494ebd53a3fe775d4d9b402e918a6598 Author: Arthur Huillet Date: Thu Jul 26 15:21:08 2007 +0200 Preliminary implementation of "host-side double buffering" to improve performance of Xv commit 4f4721f1b8a76a0bf9938e3c614ba7166d5f3dab Author: Arthur Huillet Date: Wed Jul 25 22:23:37 2007 +0200 removed pointless sleep in notifier wait commit 290e71edfbc6297c63989d5c21a67717b4ca5f76 Author: Arthur Huillet Date: Fri Jul 20 17:46:18 2007 +0200 cleaner Xv code - fallback on CPU copy yet to be implemented commit e760235a0596895b726220366c50d4e7c244006e Author: Arthur Huillet Date: Fri Jul 20 01:33:49 2007 +0200 Now using GART as the source for the blitter - preliminary implementation, gotta clean it up commit a97181f95a1f5da6180b8a72f3ca307079b92079 Author: Arthur Huillet Date: Fri Jul 20 01:22:11 2007 +0200 now yielding in the wait for notifier for Xv only commit 81cc744eb31343a509f9ca0d37993d462f162d80 Author: Arthur Huillet Date: Sun Jul 15 01:00:19 2007 +0200 VERY EXPERIMENTAL now EXA waits for sync before its operations - this removed the text corruption and didn't affect performance here, please report. commit 74ed14e1d27903f0647fe650da666cacb1e2e940 Author: Arthur Huillet Date: Sun Jul 15 00:35:46 2007 +0200 VERY EXPERIMENTAL moved the notifier wait at the beginning of NvPutImage commit 590e385cb5e4f26532ec5ee380f2aaae1cd69894 Author: Arthur Huillet Date: Thu Jul 12 03:19:21 2007 +0200 new structure for PutImage, and DMA transfers commit d4abe1fb011d414fda1871f052dda5b860f0820b Author: Ben Skeggs Date: Tue Jul 17 13:09:34 2007 +1000 Emit NOTIFY+NOP sequence after M2MF. commit 36396aeaeabfd00c7014fa52c17261c3da8b6a82 Author: Ben Skeggs Date: Fri Jul 13 15:28:56 2007 +1000 Play "nuke the typedef" commit 99e0fbc1a34f535abf41c6f4566901c326dfa4b5 Author: Ben Skeggs Date: Thu Jul 12 10:47:53 2007 +1000 Match drm 0.0.9 interface changes. commit e88b20b392c6972088d88f52b2da83b82b821be5 Author: Arthur Huillet Date: Wed Jul 11 02:42:24 2007 +0200 added support for PCI GART when no AGP is available - requires DRM interface version 8 commit dff1cce299df636813e8f85109f4556f520620c0 Author: Ben Skeggs Date: Sun Jul 1 18:22:37 2007 +1000 Always provide UTS, fallback to memcpy if no GART. commit 2e0e7e3856573fbbcb5fe5c96cadcde4d33db68f Author: Ben Skeggs Date: Sun Jul 1 18:19:44 2007 +1000 Always provide DFS, use memcpy fallback if no GART. commit ba0d3fbba7c1a83ddc509414772199862a4a1d03 Author: Ben Skeggs Date: Sat Jun 30 15:27:07 2007 +1000 Remove some unused cruft commit 3e2677040937fda1cd6e6ea26c238049bd9f99ba Author: Ben Skeggs Date: Sat Jun 30 04:26:48 2007 +1000 No need to call exaMarkSync() in UTS/DFS. commit e81c9a29e6d1c6573892df8601f0361c2b09c659 Author: Ben Skeggs Date: Thu Jun 28 15:16:35 2007 +1000 Bail if GlxSetVisualConfigs isn't found. commit fce3edda42d10a9c3249c6eae4218a51a981493f Author: Ben Skeggs Date: Mon Jun 25 12:45:31 2007 +1000 Bump required xorg-server version to 1.3 commit 8f1c9af02651404f8b4800b150a86bb47062fd68 Author: Ben Skeggs Date: Mon Jun 25 12:44:12 2007 +1000 Check PRAMIN for BIOS image if not found in PROM. commit 824331e3fce36098e8dec31746204f7c13038abb Author: Ben Skeggs Date: Sun Jun 24 17:48:33 2007 +1000 Implement NV40 EXA composite hooks with 3D engine. Named NV30EXA, but won't work on NV30 just yet. This should change at some point in the future. Will likely break non-voodoo'd NV4X cards (if the latest drm changes didn't already break them). Use EXANoComposite xorg.conf option to disable the hook if that's the case. commit 2e2e95fc3b650ca633294bf4c11c9ba28f45b743 Author: Ben Skeggs Date: Sun Jun 24 17:34:10 2007 +1000 Slightly saner UploadToScreen implementation commit 35806975047aa8c0dbc059709bb823f475ee7368 Author: Ben Skeggs Date: Sun Jun 24 17:27:42 2007 +1000 Wait on notifier instead of PGRAPH_STATUS When multiple channels are in use, PGRAPH_STATUS may report busy even if the ddx's channel is idle. commit c39089b3d686eecef11db2b5a2d606ba1ca2dc0c Author: Ben Skeggs Date: Sun Jun 24 17:21:27 2007 +1000 Match drm 0.0.7 interface changes commit 45dccb99a541a0e1e7be7b4e5aab12c3d8d9f30d Author: Ben Skeggs Date: Sun Jun 24 16:47:36 2007 +1000 Hide the allocator's use of physical addresses commit d41cf4febd555cca1c65df7aa981be64a6ae4219 Author: Ben Skeggs Date: Sun Jun 24 16:46:40 2007 +1000 Pass pScrn instead of pNv in a few places commit d59fbb4725602c409beee4ab96dca1eab5d73b2a Author: Ben Skeggs Date: Mon Jun 11 20:36:28 2007 +1000 don't nuke user-specified CFLAGS commit 7af7d95eb8826fa3fb9914a8c6ad4b6f4867095e Author: Henry Zhao Date: Mon Apr 23 01:38:09 2007 +0200 Import from nv : Need to inject a mode corresponding panel width/height for validation commit 896fe62233f62d5e1f874ed0eba3a200fa5ef14b Author: Richard Hughes Date: Sun Apr 15 23:01:52 2007 +0200 Fix make dist. commit e0e90f427581929fd01c96f1958a1292982dee83 Author: Ben Skeggs Date: Mon Mar 26 01:04:15 2007 +1000 nouveau: rip out card init, everything except modesetting. commit 23fdec37c580c3b5de263dfd48bd80fa9f21bc19 Author: Ben Skeggs Date: Wed Mar 21 17:55:38 2007 +1100 nouveau: bump for drm interface changes (v0.0.5) commit 1fddcfe6680148ed2edd8017217734a1f812298c Author: Matthieu Castet Date: Sun Mar 18 11:30:54 2007 +0100 Fix "Fix NoAccel mode" commit : NVAccelCommonInit return a bool commit 113c6837a6515d533bfa15daf49f6cd2c4c95ac4 Author: Matthieu Castet Date: Sun Mar 18 11:28:04 2007 +0100 Fix NoAccel mode commit d1b75251a3f225890f1e25b11038389d71b67611 Author: Ben Skeggs Date: Wed Mar 7 21:41:22 2007 +1100 Rip out nv30_exa.c, we have other means to play with the 3D engine now. commit adb2df9a9c88bc9646b65a235aa36416c3064c9f Author: Ben Skeggs Date: Wed Mar 7 16:46:27 2007 +1100 Xv: Make the blit adaptor play nice with EXA Previously the blit adaptor assumed that all drawing (ie. XAA) would occur to the screen pixmap, so it didn't bother touching surface offsets. Now, with EXA this isn't true and the card complained about exceeding the colour buffer limits when attempting to play video. This patch kills 2 birds with one stone; instead of pointing the GPU back at the screen pixmap when playing video, we point it at the destination drawable. So, Xv will also hopefully play nicely with redirected windows now. commit f28abbedfc7deeb90317dd24c7ec5dfe1ae94fdf Author: Ben Skeggs Date: Wed Mar 7 13:58:48 2007 +1100 Only use Y32 for Solid(). It's the only place the workaround is needed at this point, and in some places we need to use the "proper" ARGB format (Xv for instance). commit 6441253d34d118befe8950781ba012fc58bb5158 Author: Ben Skeggs Date: Wed Mar 7 04:33:44 2007 +1100 Fail configure if xorg-server isn't at least version 1.1 The minimum requirement has been 1.1 for a while now, but would only fail at compile time when compiling nv_exa.c. This patch makes configure fail, and removes the remaining conditional support for older servers. Hopefully the configure.ac is ok, I suck at autofoo. commit e568c2d8e82b1bcf25e495f4a82331803d954234 Author: Ben Skeggs Date: Wed Mar 7 04:22:24 2007 +1100 reindent and move nv_video.c around a bit commit a0a5586836eb87a8c21abb06f9b8afd64eec714a Author: Ben Skeggs Date: Tue Mar 6 05:42:29 2007 +1100 Use Y32 format instead of A8R8G8B8 for 32-bpp pixmaps fixes 32-bpp EXA solid-fills, previously alpha was forced to 0xff. commit 2e9976c034b6c310b7ad0fc858250c5e54781a9e Author: Ben Skeggs Date: Mon Mar 5 14:45:04 2007 +1100 move NVAccel* funcs into nv_accel_common.c commit add6463b848fab274cb5cc190a78aeafc1389203 Author: Stephane Marchesin Date: Sun Mar 4 22:40:48 2007 +0100 Joint patch from Ben Skeggs and me that fixes some (all ?) composite issues. commit 83868bd0a36520da49e40e2b8404e4652de39824 Author: Ben Skeggs Date: Thu Mar 1 08:39:23 2007 +1100 nouveau: avoid attempting an AGP-only alloc will fail on non-AGP cards, and leave nasty messages in Xorg.log... commit da1dd87acb7061b4772c271dc3c9071a3b160e8a Author: Ben Skeggs Date: Wed Feb 28 15:30:52 2007 +1100 match drm v4 interface changes context (gr/dma) object handling: - remove use of flags/dmaobj when creating objects - move object creation/setup into nv_accel_common.c - NV04: make PGRAPH complain a bit more if we program the hardware incorrectly. - NV04/NV10: enable PGRAPH_DEBUG_3_CTX_METHODS This bit seems to enable the use of SET_* methods. - NV04/NV10/NV40: enable PGRAPH_DEBUG_3_IGNORE_PATCHVALID Objects no longer get PATCH_CONFIG_VALID set in their context. Not sure how to get this set other than to intercept the PATCH_EXCEPTION error and enable the flag. I opted for the same method NVIDIA use, which makes PGRAPH ignore it. misc: - use NV04_SCALED_IMAGE_FROM_MEMORY on all NV_ARCH_04 cards - fix offScreenBase calculation (EXA) - move mem alloc/free helpers into nv_mem.c commit 7e5f90b264fdb3081250ff16748164998b6461a6 Author: Adam Jackson Date: Tue Feb 27 14:46:20 2007 -0500 Typo fix. commit c85da5d8b346e17e68da68e57d62c3f77029b680 Author: Damien Thebault Date: Wed Feb 14 00:35:50 2007 +0100 Don't output the chipset list to the log file, silence some warnings. commit f650cb126e042b44126df6a298e996726356ec5b Author: Stephane Marchesin Date: Wed Feb 14 00:29:52 2007 +0100 Fix the workaround (or workaround the fix). commit ac2784d38634bc6d8aec3a94d609b60cd49b21b6 Author: Stephane Marchesin Date: Wed Jan 24 20:39:40 2007 +0100 Make NV_DMA_DEBUG print the fifo position. commit d54d441df4294bf4fed26635a738f4111087e06d Author: Stephane Marchesin Date: Wed Jan 24 19:13:05 2007 +0100 Increase the fifo dump size. commit b21dbb170aa49c83ef450c2ab8c34614277e7fbc Author: B. Rathmann Date: Wed Jan 24 13:46:16 2007 +0100 Fix the G70 issues. commit e683979666a4c07590b7bade2f62cca33c19a3b2 Author: Ben Skeggs Date: Fri Jan 19 15:40:22 2007 +1100 bump drm patchlevel commit feff7cb67ae61987146a21dad535702f3db2264c Author: Jeremy Kolb Date: Thu Jan 18 21:28:51 2007 -0500 Fill in some more nv30 magic values. Not sure what they do yet but they don't hurt anything. Apply names to other registers. commit 31462c30bb79e8245edee0cf57c46b889d2f259b Author: Jeremy Kolb Date: Wed Jan 17 08:34:56 2007 -0500 Fix PGRAPH_DEBUG* regs to match what's done on the nv35. Specifically change DEBUG_3 | 0x00de0000 for data/format/dma checking. This is consistent with other cards and dumps so it should be correct. commit 4583312ba72047e6bc65bfc576a0bcf90e04b2de Author: David Jeannot Date: Tue Jan 16 00:03:56 2007 +0100 Indent nv_bios.c properly. commit 6246fba5a54dc7af9a8b16c5a14131180831562d Author: Patrice Mandin Date: Fri Jan 12 20:25:16 2007 +0100 Add missing include for malloc/free commit 37937980ec7b21c40539daee029886c4f4a0f0cd Author: Jeremy Kolb Date: Fri Jan 12 00:26:20 2007 -0500 NV30: Set DEBUG3 reg bit 1. Turns on automatic hw context switching. commit d5617e91ff13f9e19d9e5d421fd081f59dec685c Author: jkolb@brandeis.edu Date: Thu Nov 2 20:23:36 2006 -0500 Just a superficial cleaning up. Nothing exciting. commit 1f6239de8d7618de6aee402eaf1d8552524a996e Author: Ben Skeggs Date: Tue Jan 2 15:08:36 2007 +1100 NV4x changes to work with latest drm. commit f9ba29a07581ede8c409b0eb8b549db3b446f490 Author: Stephane Marchesin Date: Wed Dec 27 21:52:15 2006 +0100 Output the driver date to the logs on startup. commit 2f1d14fe1c03c32ad60776e0d0c2ae41a42ea4be Author: Ben Skeggs Date: Tue Dec 12 00:13:05 2006 +1100 Remove final remaining PRAMIN access (NV04-specific), the drm handles it now commit de7255d5420c64a2707f5e78a4157d93618e6c17 Author: Ben Skeggs Date: Fri Dec 8 18:09:26 2006 +1100 Provide some visuals with a stencil buffer, and actually call NVDRIInitVisualConfigs. commit e64dcc59b607ae5ad7049c35fed6d6a300cd5f6e Merge: 0f7957d 656dfcd Author: Matthieu Castet Date: Wed Dec 6 22:09:23 2006 +0100 Merge branch 'master' of git+ssh://matc@git.freedesktop.org/git/nouveau/xf86-video-nouveau/ commit 0f7957da9ce856ef0accc922ab368ac10171b5e6 Author: Matthieu Castet Date: Wed Dec 6 22:07:26 2006 +0100 More pgraph regs commit 656dfcd635551245b2f58835c5ea87474d3eb60d Merge: ed307d5 224578b Author: Stephane Marchesin Date: Sun Dec 3 10:20:04 2006 +0100 Merge branch 'master' of git+ssh://marcheu@git.freedesktop.org/git/nouveau/xf86-video-nouveau commit ed307d559be72e06f64566574beb36200ebe47e2 Author: Stephane Marchesin Date: Sun Dec 3 10:14:04 2006 +0100 Introduce lock step versioning with the drm. Fix the MEM_ALLOC issue. commit 224578b74645fe3624669dc3521fe6deb1cc4aae Author: Matthieu Castet Date: Tue Nov 28 19:12:55 2006 +0100 Don't alloc fifo if it is not needed. commit 146eb74cb7d1da1da55247a04e7a969d98601e42 Author: Stephane Marchesin Date: Mon Nov 27 00:03:11 2006 +0100 Mandatory oops commit. commit bdfa6776128f4bc7e2f3caa57296571acb9bd661 Author: Andrew Mitchell Date: Sun Nov 26 23:59:06 2006 +0100 Remove XF86DRI defines from the driver. It's already mandatory anyway. commit 7ebc8c4327d0b50f60c4cd4423414d4cfbb8b7d8 Author: airlied Date: Fri Nov 17 16:23:38 2006 +1100 add VGA CRTC registers from haiku commit 7150a4357dff0a2e4468190e2fd228ed07d367f9 Author: Dave Airlie Date: Fri Nov 17 12:55:37 2006 +1100 remove nvvga.h unused commit 64a1c5bb90fd13ce1c191873a4f4afc2a9ede09b Author: Dave Airlie Date: Fri Nov 17 12:51:11 2006 +1100 find and version DRM earlier in startup sequence so it is loaded firstup commit 659b5de8200d9022e7616f03b20a306fff2af0fe Author: Matthieu Castet Date: Fri Nov 17 00:59:58 2006 +0100 Cleanup of PVIDEO. commit 370b83eb3347d14003f2878737d25de7dacaeec1 Author: Matthieu Castet Date: Wed Nov 15 22:10:27 2006 +0100 Use objects from nouveau_reg.h. commit 2ffeb45a953a2ff52abe6f78dacfb9c53f873dbb Author: Ben Skeggs Date: Wed Nov 15 01:22:18 2006 +1100 Set SURFACE_FORMAT in NVExaPrepareCopy. Fixes one of the many issues with xorg-server-1.2 commit 85bda095107558846fe6fb803b6df412213b27f4 Author: Ben Skeggs Date: Wed Nov 15 00:47:04 2006 +1100 nv_exa.c tab goodness commit 8b733a25c30fb9e499d678c26b777d072d1ac187 Author: Ben Skeggs Date: Tue Nov 14 08:31:21 2006 +1100 Fix type. 0x84 is DEBUG_1 commit 030eb38dd1fe767807aab0447ed4fab4ec8c368a Author: Ben Skeggs Date: Tue Nov 14 03:09:05 2006 +1100 Fill in front_offset/front_pitch commit 589bb37ff37bb7d4775f5dca0301c5d97ea73b06 Author: Ben Skeggs Date: Sun Nov 12 16:04:32 2006 +1100 Add NVDRIFinishScreenInit. commit d0c09a1bd0e1c26cc591a3d8738b94b1efe576ec Author: Dave Airlie Date: Sun Nov 12 14:08:45 2006 +1100 add some names to registers commit 84928fd946b4f02d8a943226481304417475c4f9 Author: Ben Skeggs Date: Sun Nov 12 07:43:24 2006 +1100 Import nouveau_dri.h from the 3D driver as nv_dri.h, and setup DRIInfo->devPrivate. commit cdb9e3acf0aa5cdf6e5aa33fb499ea8e93f7f1f2 Author: Ben Skeggs Date: Sun Nov 12 07:33:53 2006 +1100 Temporary workaround for DRI drmMap problem, not sure of what we actually need to do here yet. commit d3babbfbed898921e1a167df08eef92d381d6aa0 Author: Ben Skeggs Date: Sun Nov 12 07:28:31 2006 +1100 nv_dri.c still had some space-indented functions, not anymore :) commit bad4e6c674ac4fdc0aebbbc6d36fc2854ee8f885 Merge: 1b14821 c096565 Author: Stephane Marchesin Date: Mon Nov 6 03:14:07 2006 +0100 Merge branch 'master' of git+ssh://marcheu@git.freedesktop.org/git/nouveau/xf86-video-nouveau commit 1b14821ecd67a62a34a902e62c0b229bb9ed97ca Author: Stephane Marchesin Date: Mon Nov 6 03:13:50 2006 +0100 Added a DRI visual configs functions. It should work but I can't test it... commit c096565a0f6e6c12255e57eaefb878904ce04971 Author: Ben Skeggs Date: Mon Nov 6 11:54:19 2006 +1100 Fix TEX0_IN_TEX1 fragprog.. I must've been asleep when I did that.. commit 47f32cfe6e2f513f86b7a8f02a8872c134e2327d Author: Dave Airlie Date: Mon Nov 6 11:45:53 2006 +1100 use pagesize to work out the sarea size commit e6c4d40e247939f677e285db15a78ccc33b92b41 Author: Dave Airlie Date: Mon Nov 6 08:39:49 2006 +1100 don't fail dri module load if loaded already commit 4cc4a1f7ef559f1ce56706f16ee24a3b6879d28d Merge: b95f8aa b4e4307 Author: Dave Airlie Date: Sun Nov 5 18:00:29 2006 +1100 Merge ssh://10.1.1.200/home/airlied/git/xf86-video-nouveau commit b4e4307af982b50aff191983d209cb855a377f80 Author: Dave Airlie Date: Sun Nov 5 17:59:17 2006 +1100 fix G5 endianness issue with panel size regs commit b95f8aac9e6dcbf4d10a5e26bde6a0f3595608cb Merge: ba71154 e6b1f68 Author: Dave Airlie Date: Sun Nov 5 17:53:53 2006 +1100 Merge branch 'master' of ssh://git.freedesktop.org/git/nouveau/xf86-video-nouveau commit ba711545a0fd091c05b514f30c1e904370026499 Author: Dave Airlie Date: Sun Nov 5 17:51:10 2006 +1100 set the PPC use BE MMIO commit e6b1f687e3bf127d8dba1871d67eff9d83ac284e Author: Stephane Marchesin Date: Sat Nov 4 16:57:30 2006 +0100 Remove the context object flags handling from the DDX since it's now in the DRM. commit ead77e0ec1d0c62829fa973db3d62937b44781e5 Author: Stephane Marchesin Date: Sat Nov 4 04:47:57 2006 +0100 Add the NV44A 3D object class. commit 87463666ab23e047f355cfcd7016aa9c7cd8ad76 Author: Ben Skeggs Date: Fri Nov 3 04:00:09 2006 +1100 Use "do { blah; } while(0)" instead of just braces. (suggestion from megari on irc) commit 68682238b36292409ce7f41ed90bd3be860e77f3 Author: Ben Skeggs Date: Thu Nov 2 16:07:26 2006 +1100 Install NV30_TCL EXA hooks if NV_ENABLE_3D is defined. Ok.. Firstly I don't recommend using this code, it is not ready for general usage. We probably only want to keep the Composite hook, but the others are useful to test things with. 1. You will most likely (read: definitely) need to start X with the binary driver before starting X with nouveau/NV30_TCL.. If you don't PGRAPH will yell at you and X will crash. 2. Copy() is bad when the src and dest are the same, so scrolling and moving windows around will fall back to software rendering.. 3. Composite() is slightly better than the one in nv_exa.c, but it's buggy in a few cases and not complete yet. Particularly, transforms are missing. 4. There are some correctness problems that I haven't bothered to fix yet, so you'll probably see bits of windows left behind when moving them around.. 5. Don't be surprised if it's horribly broken :) commit c371d90935d6baf2f1dd0073741637d14b506f59 Author: Ben Skeggs Date: Thu Nov 2 15:57:47 2006 +1100 Put braces around NVDmaSetObjectOnSubchannel... commit b45aea22a8da69ad4a2493f7f33fa65186a83ba3 Author: Ben Skeggs Date: Thu Nov 2 14:52:24 2006 +1100 Merge >=NV30-specific EXA code.. hard-disabled pending testing on my C51, and a proper commit message.. commit e79b0befe92110c1cd2e0e1ae69fced1cd5acba3 Author: Ben Skeggs Date: Thu Nov 2 01:29:34 2006 +1100 Reformat NVExaInit commit 6a77e5bf2e58b19df67c029ad7bffd92f8cf652f Author: Ben Skeggs Date: Thu Nov 2 01:18:14 2006 +1100 First part of preparation for NV30_TCL EXA code merge. commit 1d6e911ab5391277b403ef6f7b095203cfd19a53 Author: Ben Skeggs Date: Thu Nov 2 00:38:53 2006 +1100 Reindent nv_dma.c with tabs.. commit 06748f74f639242fd690a3dda7e685ae4fa1f95e Author: Ben Skeggs Date: Wed Oct 11 11:06:16 2006 +1100 s/init/alloc/ commit a2d55603db8c01cc4b9f3404c282b1e4963a152c Author: Stephane Marchesin Date: Wed Oct 11 00:39:36 2006 +0200 Remove the PFIFO_REINIT hack, and enable the irq by default. This commit requires the matching drm commit, and will probably break stuff. commit 78537b3342bbf1c16dc78f8f06cb3f989ce8f03f Author: Ben Skeggs Date: Thu Sep 14 12:56:28 2006 +1000 It usually helps to wait on a notifier, and not on a NVAllocRec.. Oops. commit 01dde28611238cf2d81b087b371cc67296d1b581 Author: Ben Skeggs Date: Thu Sep 14 12:17:32 2006 +1000 Allocate notifiers with NVAllocateMemory. Fix some warnings. Return success/fail for DMA and Context object creation. commit 3d5c560f509646853cc863e35ed2412db39cd8a3 Author: Stephane Marchesin Date: Tue Sep 5 07:39:10 2006 +0200 Port some of the least intrusive nv03 code to the nouveau driver. commit 6fb7a63bf5d4f046583f9d14862c16ec5979a318 Author: Ben Skeggs Date: Sun Sep 3 06:41:52 2006 +1000 Config options for cmdbuf location/size. Will default to putting the cmdbuf in AGP (except for PowerPC) if an AGP card is detected. Won't currently be detected because something is wrong DRM-size. If you have an AGP card you can manually specify 'FBLocation "AGP"' in xorg.conf to make it work. Needs an updated DRM to have any effect. commit e49d561d2328eeac7691661e28955a882bba2192 Author: Ben Skeggs Date: Sun Sep 3 01:51:23 2006 +1000 Copy+Paste a workaround for a NV40 hardware bug from NVShowHideCursor() into LoadStateExt(). This should fix the problem of the HW-cursor still being enabled when VT-switching back to a framebuffer console. commit 69ace0c4388cb1a66ea74f627fe02fe89e7d35c0 Author: Ben Skeggs Date: Sat Sep 2 22:17:55 2006 +1000 Kill off one of the M2MF objects, and use the engine's SET_OBJECT_{IN_OUT} to swap between AGP->VRAM and VRAM->AGP. commit 72cf79b9e5dc0d97e933f540d063161129bbebf7 Author: Ben Skeggs Date: Sat Sep 2 21:59:26 2006 +1000 Don't create objects that aren't needed. This frees up a couple of subchannels for EXA, use one of them for MEMORY_TO_MEMORY_FORMAT. commit 6a06d9022f3c0bc845f02091d83faf95f69f81bd Author: Ben Skeggs Date: Sat Sep 2 18:15:15 2006 +1000 Move code out of nv_xaa.c that doesn't belong there. commit 46ac77d524dd183c6b50a7a0e77990514ff28840 Author: Ben Skeggs Date: Sat Sep 2 17:57:38 2006 +1000 Add back some setup that went missing during PFB reg cleanup. Should fix corruption on NV40. commit 03b86f08667a10c44420d60787fe9e17b5d20ab8 Author: Ben Skeggs Date: Thu Aug 31 15:53:25 2006 +1000 Rename module from nv_drv.so to nouveau_drv.so to allow both nv and nouveau to coexist. Also disabled building of riva128.so, as we plan on merging it with the rest of the nv code. commit 17c871a44987af04d1688798e2e7802a5f5ba522 Author: Dave Airlie Date: Wed Aug 30 06:24:23 2006 +1000 clean up FB register names commit 7d3c12424fd79daf831b7c28780586e8bd9c3574 Author: Dave Airlie Date: Wed Aug 30 04:23:43 2006 +1000 add some names for the RAMDAC registers commit 0e70296ce9f7a921c4b971919d12c262202d8d48 Author: Dave Airlie Date: Wed Aug 30 02:18:02 2006 +1000 add some names for some of the CRTC registers - some names not very meaningful yet commit 9a5e2c5be96eaced3aadd153fb2f9b1928173af7 Author: Dave Airlie Date: Wed Aug 30 01:48:07 2006 +1000 remove select head registers PCIO and PDIO commit dec0691db79b1ded2942c1cbfa97434fa8693f9b Author: Dave Airlie Date: Wed Aug 30 01:13:05 2006 +1000 remove unused pointers after macroisation commit 9444b692607895572c8199a60d6968fbc374d51e Author: Dave Airlie Date: Wed Aug 30 01:02:45 2006 +1000 missed 2x PGRAPHs in XAA commit f73cb8ef4f4fbb588bfe9ca767286153b2565a37 Author: Dave Airlie Date: Wed Aug 30 00:59:39 2006 +1000 remove PCRTC via macros commit 6d5b1765e10856a7dfa75c27ad96f547dc26b23a Author: Dave Airlie Date: Wed Aug 30 00:48:23 2006 +1000 remove PEXTDEV and PTIMER using macros commit 6140bdfe6959cd0f6ce8e06d34b6e838d8cef046 Author: Dave Airlie Date: Wed Aug 30 00:43:07 2006 +1000 remove PMC using macros commit 28d9befa6308462a6bc1809b894ebe389873fd6a Author: Ben Skeggs Date: Wed Aug 30 01:26:58 2006 +1000 Make EXA the default commit fa9209f77df3d835ebad43dc82988c000398f650 Merge: ab2b550 2f7caf6 Author: Stephane Marchesin Date: Tue Aug 29 13:42:49 2006 +0200 Merge branch 'master' of ssh://marcheu@git.freedesktop.org/git/nouveau/xf86-video-nouveau commit ab2b5509e82a646ac187b020c91234e33f81065d Author: Stephane Marchesin Date: Tue Aug 29 13:42:34 2006 +0200 Make use of NV5_SCALED_IMAGE_FROM_MEMORY on NV05. commit 9911eb726f99166968165710a9e2f540388cd42a Author: Stephane Marchesin marchesin@icps.u-strasbg.fr Date: Tue Aug 29 03:18:27 2006 +0200 Enable the composite hooks. After changing NV_SCALED_IMAGE_FROM_MEMORY for NV10_SCALED_IMAGE_FROM_MEMORY it seems the composite hook works. Also, add a more complete list of desktop pci ids. commit 2f7caf6eb5eb6a3991720fc413bd20723473212f Author: Dave Airlie Date: Tue Aug 29 08:13:59 2006 +1000 remove PGRAPH direct accesses commit 51f0b67d9064fffc6b357d2923188a030e85eace Author: Dave Airlie Date: Tue Aug 29 07:51:49 2006 +1000 remove PFB usage with macros commit b50736d1583b960ab70f3e113294db8c2da69e31 Author: Ben Skeggs Date: Mon Aug 28 17:07:06 2006 +1000 Use NVAllocateMemory for AGP scratch buffer. commit a5ddd8fcf4278c35d71605794d0d5711e7c0ebb7 Author: Dave Airlie Date: Mon Aug 28 16:41:20 2006 +1000 remove legacy crap from nvreg.h add cursor reg defines commit f035bc63c34bce49e1c2619f14ab98b4b1922beb Author: Dave Airlie Date: Mon Aug 28 16:23:32 2006 +1000 missed some vga routines commit 0b45a9f725442d5bd784ee67b7473929993f6082 Author: Dave Airlie Date: Mon Aug 28 16:19:56 2006 +1000 add RAMDAC register read/write commit 24267d7e39383930c589bc80e2bbd4e5cf66ec2a Merge: 349496c c0154e6 Author: Ben Skeggs Date: Mon Aug 28 14:46:51 2006 +1000 Merge branch 'master' of git+ssh://git.freedesktop.org/git/nouveau/xf86-video-nouveau commit 349496c97fafed04806c01bd9d4386b0fcaa9fd2 Author: Ben Skeggs Date: Mon Aug 28 14:45:42 2006 +1000 More cleanups from DRM memory alloc changes. commit c0154e615c091c7bb483fcfd07e216b21618d0f2 Author: Dave Airlie Date: Mon Aug 28 14:07:05 2006 +1000 use nvReadVGA and nvWriteVGA in more places commit e0f1479773c1a00ec69edbece98e5e2b66c0c2c9 Author: Ben Skeggs Date: Sun Aug 27 22:45:36 2006 +1000 Fix ShadowFB for FB alloc changes, should fix RandR rotation also. commit 0de4d0b8b79e2238f2c4b374b5012a25ae985502 Author: Ben Skeggs Date: Sun Aug 27 22:27:17 2006 +1000 Slash UseFBDev and DGA support. commit e49d08ff482755e9f95c4c3b3cf065e6f0d2bc23 Author: Dave Airlie Date: Sun Aug 27 10:04:50 2006 +1000 add some read/write NV VGA register wrappers commit c64c55b3133221462980b934a308daf2cc0e473d Author: Dave Airlie Date: Sun Aug 27 09:56:36 2006 +1000 remove all U0* types in favour of more standard CARD* types commit 9c294b962379a955d2e8e2cb2c54f845fd4c5740 Author: Ben Skeggs Date: Sat Aug 26 00:43:21 2006 +1000 Wow I suck tonight.. another oops. commit b8411ae92b6833c9336f82ed86dc8755ca8b9dcc Author: Ben Skeggs Date: Sat Aug 26 00:39:36 2006 +1000 Whoa, how did that happen... commit d458cdac152a49eec9acdfe5081f6d426e712bf2 Author: Ben Skeggs Date: Sat Aug 26 00:33:47 2006 +1000 Initial work in porting the DDX to use the DRM for memory allocations. This is enough to make it work, but there's still a *lot* of cleanups to do. commit 83ae039b8f8ea08ae9436201014c7e1ba25a4839 Author: Dave Airlie Date: Fri Aug 25 19:33:08 2006 +1000 fixup compile for git commit ea2bc33d32bf9abeb23b3fe8cb6c9d28ddccef01 Author: Dave Airlie Date: Fri Aug 25 19:17:51 2006 +1000 updated git tree from changes that got lost on way from CVS commit 07b81112ba4678085a148e5ffaadd4684f74d64d Author: marcheu Date: Thu Aug 24 21:00:42 2006 +0000 Make NVInitAGP fail if there is no AGP area. Maybe that function shouldn't be called for PCI(-e) cards in the first place. commit 109b5abed6c0c701a1fb84ec63ff8f5914eee8e3 Author: marcheu Date: Wed Aug 23 21:32:47 2006 +0000 If you stare at nv_hw.c long enough, things start to make sense. I swear ! commit 80ba1cfec290dbb82c4a1042255a21b626512e84 Author: marcheu Date: Wed Aug 23 17:11:22 2006 +0000 Comment out NVDmaKickoff() in NVResetGraphics(). NVResetGraphics is called twice on startup, and flushed to the card in between. Since NVResetGraphics resets the fifo pointer, this sometimes ends up in a situation where the CPU writes at the same place of the fifo (at the beggining) the GPU is currently reading from. Obviously this is bad. Also, add code to dump fifo in case of crashes. commit 9b19daa55c7770a3e66b2374da0212c75ed21260 Author: marcheu Date: Tue Aug 22 01:16:18 2006 +0000 Remove debug message. commit b44b54e4c369c612a692eabb8662b3db97758a83 Author: marcheu Date: Mon Aug 21 13:19:06 2006 +0000 Fill more fields in the DRIInfo struct. commit aaf13d0e7cc5c76ee81b4234dc2c55bbedac1faa Author: darktama_ Date: Sun Aug 20 15:11:06 2006 +0000 Oops, this shouldn't be here still... commit 00fd86dea8d2eaa561cc89a35046d15bcf92bb82 Author: darktama_ Date: Sun Aug 20 14:40:43 2006 +0000 - Allocate notifiers with the DRM - Rename agpMemory and friends to agpScratch and friends commit d5c68d3f83f84ca67ee9df269b778bbed50851a0 Author: darktama_ Date: Sun Aug 20 12:51:49 2006 +0000 Alloc AGP through the DRM commit 6d0924c272954f44b2a9b6b11be8c209dd67de8e Author: marcheu Date: Sat Aug 19 16:04:56 2006 +0000 Begginings of PINS parsing code I had laying around... commit 59e2d852d6f10ba26a394ff7ea5d3e7b30f2be71 Author: darktama_ Date: Mon Jul 31 07:32:57 2006 +0000 Need.To.Pay.More.Attention... commit 9a314361fd004afe7129ae2d41f58ddd7fe6f56a Author: darktama_ Date: Mon Jul 31 03:24:17 2006 +0000 Updates for DRM changes, and a few cleanups commit 11fbb14876e7873a863de16b01fb5778e841b4bd Author: erikwaling Date: Thu Jul 13 04:13:07 2006 +0000 Added more opcodes and some TODOs. commit dd54fea7562b8c9a6d26b645a03c9644c08090cc Author: erikwaling Date: Thu Jul 13 03:09:11 2006 +0000 Changed INIT_MACRO to haiku-like implementation. commit b7e68083f3bf6b0943b58cfed0b82de61eeaac33 Author: marcheu Date: Mon Jul 10 00:36:31 2006 +0000 Add a check that bios writes fall within known areas. Use these areas to setup mappings as well. Add some debug support to the bios code. commit fe2a1360d72385ad0f61f381a296820d02a60e79 Author: marcheu Date: Mon Jul 10 00:30:54 2006 +0000 Use -Wall by default commit ca38f1d9d61393915515e1042204b35aeac863ba Author: darktama_ Date: Sat Jul 8 13:53:00 2006 +0000 Remove FIFO and object handling out of the DDX. This commit makes the DDX dependant on the DRM to operate at all.. commit b9b880b83896defe06ce61c9aef2f5ae4eb4f37e Author: lumag Date: Mon Jun 12 20:21:11 2006 +0000 Fix exa acceleration by filling versioning fields. commit 0bbc0bcfe33c2625ea6175c3b052d231cc3b7c8f Author: darktama_ Date: Sat Jun 10 18:34:14 2006 +0000 Quick hack to make sure CRTC 0x44 contains it's original value when VT switching back to the console (also on server shutdown). Should fix the problem of one display having scrolling issues, and losing it's text-mode cursor. commit 18e65f4e0604c744ea905319348c688141e14bbf Author: marcheu Date: Thu Jun 8 01:10:01 2006 +0000 Remove xf86ReadPciBIOS, use the bios from the reg aperture instead (that will probably work on more archs anyway) commit 2f5fcd131dbea5cb9873cc8a4aad44ae1bcf1340 Author: darktama_ Date: Fri Jun 2 21:54:48 2006 +0000 Remove some XFree86LOADER conditionals commit 7889cc0241450aaa8e9dafc65f9f7976df3cd99b Author: darktama_ Date: Sat Apr 22 02:26:53 2006 +0000 Need to include nv_include.h before checking that XF86DRI is definied. XF86DRI lives in xorg-server.h - included from config.h commit 3083263eb40f57e4c529473c5e9f935109a1719f Author: marcheu Date: Fri Apr 21 22:15:59 2006 +0000 Don't include config.h in all files but in the *_include.h files. That's what they're here for after all. commit 864e2c6936a37005112a3c6f3f7db5a16100b96c Author: darktama_ Date: Fri Apr 21 02:46:31 2006 +0000 Include config.h in all source files commit 74ff8fc182c6f728a1b5e8a0415c25b8b7a4af94 Author: marcheu Date: Thu Apr 20 18:54:12 2006 +0000 Added a fixme about laptop chips commit ad38154a7e075207d8f462ffb90489efe9cf01f1 Author: erikwaling Date: Thu Apr 20 18:50:11 2006 +0000 Ported nv_exa.c to work with EXA >=2.0.0 commit 7c25d338e0c79a288fa192d9a2d4ac6eb59996c1 Author: Dave Airlie Date: Fri Aug 25 18:51:03 2006 +1000 initial import of nouveau code commit 98372d0f94a9d658a7257d2199ef37a2449b9961 Author: Adam Jackson Date: Fri Aug 11 16:02:28 2006 -0400 Replace panel message + MODE_BAD with MODE_PANEL. commit 43ec66396f805aba87017d193d97a03372c29909 Author: Aaron Plattner Date: Fri Jun 30 12:54:44 2006 -0700 Bump version to 1.2.0 for release. commit 6265b77dc8878e36abd76b73e209a2d76d70ba22 Author: Mark Vojkovich Date: Thu Jun 29 18:41:44 2006 -0700 New chip support and new PCI IDs. commit 5114639dff73c22845451acffed94f1da284a7df Author: Aaron Plattner Date: Mon Jun 19 09:32:11 2006 -0700 Generate the ChangeLog from git history for dist. commit 04d344dfcbb9928cf33943a880de333557e14be9 Author: Aaron Plattner Date: Fri Jun 16 18:03:21 2006 -0700 Delete the ChangeLog. The git history is just as good. commit eda3bac20dce2d5584ac07cc6b28a08ec55aa798 Author: Aaron Plattner Date: Wed May 24 15:05:22 2006 -0700 Derive the module version numbers from the package version. This idea was lifted from Michel Dänzer's similar code in xf86-video-ati. Also, make the riva and nv driver versions match. commit 1a40c4a79b79034d191496726a27af8b0805dba6 Author: Aaron Plattner Date: Thu Apr 27 18:10:37 2006 -0700 Add the libcwrapper back, conditionalized on the server version. commit 425e30d58b7b62fff01fbb58585ee69b2df6fec1 Author: Aaron Plattner Date: Thu Apr 27 10:32:10 2006 -0700 Determine whether PutImage hooks take a DrawablePtr by server version. Remove dependency on xorg-server >= 1.0.99.901. commit c34a93265cb42168376ebdf0746b3ccded9f126c Author: Aaron Plattner Date: Tue Apr 25 14:38:43 2006 -0700 Bump to 1.1.2 for new product names. commit f70247ade9eac8fa6115889662abb8e678a182dc Author: Aaron Plattner Date: Tue Apr 25 14:29:44 2006 -0700 Add names for Quadro FX 5500, 3500, 1500, 560, 550, and 350. commit ca4c8af624e87e5c6a2927092a9c230ae0e60c0e Author: Aaron Plattner Date: Wed Apr 12 02:04:52 2006 -0700 Add .gitignores. commit fb3faeb6073cef5a6a1b094f3d3bf020ceba9db6 Author: Adam Jackson Date: Sat Apr 8 01:38:33 2006 +0000 Bump to 1.1.1 for Xv changes. commit f793182898c4d844e9861a60a945f53ab7071adc Author: Aaron Plattner Date: Fri Apr 7 23:11:06 2006 +0000 Add a DrawablePtr argument to the XV functions to pave the way for redirected video. commit cb6b4073df7404822e2b553519c3a101f2ab0524 Author: Adam Jackson Date: Fri Apr 7 19:40:34 2006 +0000 Unlibcwrap. Bump server version requirement. Bump to 1.1.0. commit 3f140052f891a4dc674b82e438c9595e59b4676b Author: Aaron Plattner Date: Mon Mar 27 21:40:24 2006 +0000 Bump version to 1.0.2.0 for release. commit 84c007b426d321786f9aec1d9fda14b5baa35561 Author: Aaron Plattner Date: Fri Mar 10 02:29:03 2006 +0000 Coverity nos. 868 and 869: Free the result of xf86GetEntityInfo. commit 61e237eec7193701ce1236e67ebdfe7465b5db8a Author: Aaron Plattner Date: Tue Jan 24 16:45:29 2006 +0000 Fix typo in PCI ID (reported by Stefan Dirsch). Patch from Mark Vojkovich to XFree86 CVS on 2006-01-23. commit 5064f8a903923ca4b819bdf345f8f4c9254bdf3b Author: Aaron Plattner Date: Mon Jan 23 00:04:41 2006 +0000 New chip support from Mark Vojkovich's 2006-01-20 and 2006-01-21 XFree86 CVS updates. commit e8fd45625f2e09830136bde6bc70d51f25892450 Author: Aaron Plattner Date: Mon Jan 2 01:54:19 2006 +0000 Preliminary RandR rotation support. Bug #4708 commit 5ec53e759c5233a26a6c80901b9f31b51153f756 Author: Kevin E Martin Date: Wed Dec 21 02:30:01 2005 +0000 Update package version for X11R7 release. commit b0c8ca8c03c8ad1648b4597f7825bac7531100a7 Author: Adam Jackson Date: Mon Dec 19 16:25:54 2005 +0000 Stub COPYING files commit 4fd25919c2a217d355419efdfbf0fe148af6c146 Author: Kevin E Martin Date: Thu Dec 15 00:24:18 2005 +0000 Update package version number for final X11R7 release candidate. commit e4194bab2a790af10581ea2842717623db8998a2 Author: Kevin E Martin Date: Tue Dec 6 22:48:36 2005 +0000 Change *man_SOURCES ==> *man_PRE to fix autotools warnings. commit baf60566b505fdf4387ec25789584b3f6ad72ce8 Author: Kevin E Martin Date: Sat Dec 3 05:49:34 2005 +0000 Update package version number for X11R7 RC3 release. commit 4df200ea93b0a072f29057eeaf7a09e5a3a1dfe2 Author: Kevin E Martin Date: Fri Dec 2 02:16:09 2005 +0000 Remove extraneous AC_MSG_RESULT. commit d4d5d7325fab1ea81f71d1ffc1fbbcc2d2253a65 Author: Adam Jackson Date: Tue Nov 29 23:30:00 2005 +0000 Only build dlloader modules by default. commit d76f3f776bdab920d050542da45ec2871d4b84e6 Author: Alan Coopersmith Date: Mon Nov 28 22:04:10 2005 +0000 Change *mandir targets to use new *_MAN_DIR variables set by xorg-macros.m4 update to fix bug #5167 (Linux prefers *.1x man pages in man1 subdir) commit d4e306b585f737b4c1287c520470425b0c11dc32 Author: Kevin E Martin Date: Wed Nov 9 21:15:19 2005 +0000 Update package version number for X11R7 RC2 release. commit dde6dd1c7f95f9dccabfd919ad4ced795fe3fbd1 Author: Kevin E Martin Date: Tue Nov 1 15:08:53 2005 +0000 Update pkgcheck depedencies to work with separate build roots. commit 92b1bfcc8fd0f34e636a1e9955d5236334f78cd3 Author: Kevin E Martin Date: Wed Oct 19 02:48:03 2005 +0000 Update package version number for RC1 release. commit eda1cbc29cb0b2193a5bfb0eb620c740d9aef7c6 Author: Alan Coopersmith Date: Tue Oct 18 00:01:53 2005 +0000 Use @DRIVER_MAN_SUFFIX@ instead of $(DRIVER_MAN_SUFFIX) in macro substitutions to work better with BSD make commit 6c58a05e405cbc87fbe0497123d9b82d2480bdd3 Author: Adam Jackson Date: Mon Oct 17 22:57:34 2005 +0000 More 1.7 braindamage: define EXTRA_DIST in terms of @DRIVER_NAME@ instead of indirectly commit fc20886c28c465578e6bd3ef2d9fe62bd993382b Author: Alan Coopersmith Date: Mon Oct 17 00:09:04 2005 +0000 Use sed & cpp to substitute variables in driver man pages commit 1e3c84f332b7b9a4b320b370b8b2ecafbbdd8f8a Author: Søren Sandmann Pedersen Date: Tue Oct 4 21:36:51 2005 +0000 Link nvidida readme file into nvidia driver, Exclude duplicated neomagic readme commit d8be62ab175a013f78d4944036da284ffad34d96 Author: Aaron Plattner Date: Thu Sep 29 21:47:29 2005 +0000 Add entries for GeForce Go 7800. (Change went into XFree86 on 2005-09-28). Don't hang if j is zero. This should never happen, but it's better to be safe than sorry. commit e39eb8b8c1a79d04f90a195351c4eec3dba0d612 Author: Aaron Plattner Date: Wed Sep 28 17:18:02 2005 +0000 Bug #3333 Patch #3417 Fix a hang during screen-to-screen blits on certain GeForce 6xxx and Quadro FX 540 graphics cards. commit dc5acc2e477edb85f468f8d46862ca138426cfd9 Author: Alan Coopersmith Date: Fri Sep 23 16:25:28 2005 +0000 Bug #4558 Patch #3551 22-Sep-2005 nv driver update from Mark Vojkovich: Fix possible cause of some acceleration instability on some GeForce6xxx parts. commit b0ea43a51d2a6983559eecdb7dbb73c5e585697c Author: Alan Coopersmith Date: Sat Sep 17 01:37:18 2005 +0000 Bugzilla #4480 Patch #3305 13-Sep-2005 nv driver update from Mark Vojkovich: - Adds PCI ids for GeForce 6800 XT cards - Adds names for existing PCI ids for some GeForce 6xxx & 7xxx cards - Fix a potential problem with pixmap cache corruption on GeForce 6xxx and 7xxx parts. (Mark Vojkovich, NVIDIA) commit cdf20033e2cfc9e8634876dd42613f7719340a5a Author: Eric Anholt Date: Fri Sep 16 01:36:45 2005 +0000 Put the riva_* files in the riva128.so subdriver, like monolithic. commit 9b1b3184cd586f6f8bf65ebd6f3c1afb1ae64d64 Author: Daniel Stone Date: Thu Aug 18 09:03:47 2005 +0000 Update autogen.sh to one that does objdir != srcdir. commit 75a0deb6e9d1a5bc698ea67b338c56bc99c1b2c0 Author: Alan Coopersmith Date: Fri Aug 5 03:52:28 2005 +0000 Bug #3987 Patch #3264 Workaround for 512 Meg video cards (can only map 256 Meg of it). (Mark Vojkovich) commit fb4033bd2662c199d7e4f0e3df7c4a0fb9870984 Author: Kevin E Martin Date: Fri Jul 29 21:22:44 2005 +0000 Various changes preparing packages for RC0: - Verify and update package version numbers as needed - Implement versioning scheme - Change bug address to point to bugzilla bug entry form - Disable loadable i18n in libX11 by default (use --enable-loadable-i18n to reenable it) - Fix makedepend to use pkgconfig and pass distcheck - Update build script to build macros first - Update modular Xorg version commit 6ea22f19210448605b9a18bf500c33168a77e77c Author: Søren Sandmann Pedersen Date: Tue Jul 26 18:40:29 2005 +0000 Build system for nv commit 74772a7dc49e34229497441fbbf2631bfb259b39 Author: Adam Jackson Date: Mon Jul 11 02:29:58 2005 +0000 Prep for modular builds by adding guarded #include "config.h" everywhere. commit 80ac36ec4fbbd9a4aae1bde01fe6ebc1088fa8cb Author: Alan Coopersmith Date: Sat Jul 9 16:51:58 2005 +0000 Bug #3740 Patch #3058 08-Jul-2005 nv driver updates from Mark Vojkovich: Change some console restore code for NV11. Hopefully, we can more reliably restore the console for desktop systems using DVI. This may correct a recent regression on NV11. Also, new PCI IDs. Add new nVidia PCI ids to match nv_driver.c changes. commit 6b75bc13462f1cfb735fe1de359ed9e45b5f4415 Author: Daniel Stone Date: Fri Jul 1 22:43:12 2005 +0000 Change all misc.h and os.h references to . commit 270f5269a08da8ed0186e8a2223b673e0433a771 Author: Alan Coopersmith Date: Wed Jun 29 15:56:23 2005 +0000 Bugzilla #3657 Patch #2992 Experimental DPMS support for GeForce4 and later laptops. (Mark Vojkovich) commit bf73c0716e9e4be0540d2ee7340bd5833831a434 Author: Adam Jackson Date: Sun Jun 26 04:38:12 2005 +0000 Bug #3628: Fix video driver submodule visibility. commit 175766a44ceb0a5d4226e9088d8e3e6d73815b4f Author: Adam Jackson Date: Sat Jun 25 21:16:55 2005 +0000 Bug #3626: _X_EXPORT tags for video and input drivers. commit 8759399a3744f0eb1e0cdc83ed23aaaafb9be444 Author: Alan Coopersmith Date: Thu Apr 21 22:25:47 2005 +0000 programs/Xserver/hw/xfree86/drivers/nv/nv_driver.c programs/Xserver/hw/xfree86/drivers/nv/nv_setup.c //bugs.freedesktop.org/show_bug.cgi?id=3095> Patch #2501 nv DPMS panel fixes (Mark Vojkovich) commit 2db214d7cf5336fbea36d83bc9b089f70a494614 Author: Daniel Stone Date: Wed Apr 20 12:25:29 2005 +0000 Fix includes right throughout the Xserver tree: change "foo.h" to for core headers, e.g. X.h, Xpoll.h; change "foo.h", "extensions/foo.h" and "X11/foo.h" to for extension headers, e.g. Xv.h; change "foo.[ch]" to for Xtrans files. commit 3fe983d254f1ffe3c285548013be677ed4a67137 Author: Alan Coopersmith Date: Mon Feb 14 03:37:20 2005 +0000 Bugzilla #2533 (https://bugs.freedesktop.org/show_bug.cgi?id=2533) Feb. 3, 2005 "nv" driver update from Mark Vojkovich (Mainly driver updates for nVidia cards with ((pNv->Chipset & 0xfff0) == 0x0090) ) commit 8c95fdc889b56a76497f5194bafd06edfc962387 Author: Alan Coopersmith Date: Wed Jan 26 01:36:04 2005 +0000 Jan. 2005 nv chipset update from Mark Vojkovich Bug #2380 Patch #1752 commit 67c89797f62ea53b37927c64e0403e7e7c42de56 Author: Alan Coopersmith Date: Fri Dec 10 03:56:42 2004 +0000 Bugzilla #1985 sync to 12/08/04 version of Nvidia driver from Mark Vojkovich, including these fixes since last sync: Fix some DAC/Graphics memory contention issues on newer NVIDIA chips (specifically, NV40). More PCI IDs. Also, I've had trouble getting some panels to work automagically so I expose an FPTweak option to let the user adjust a troublesome register. commit 7c34d051168eba499beb20c40fd95703375f4c7b Author: Alan Coopersmith Date: Wed Nov 10 02:53:44 2004 +0000 (Bugzilla #1777) Import Mark Vojkovich's changes to NV driver since last sync with XFree86, including: Support for new NVIDIA chips. Egbert has a laptop that needs some DPMS ordering reversed. New chip support and workaround for Toshiba M30 laptop issue. PCI-Xpress chip support. DPMS support for DVI. Cursor changes for NV11 so that alpha blended cursors will work in conjunction with dithering (laptop panels). It previously did not. HW support for alpha blended stretch blits on NV5 and higher, though nothing is using them at this point. Add support for VBlank syncing the Xv Blit adaptor on hardware that supports it (NV11 and higher). Make console restoration a little more robust. We were making an assumption that failed on a newer laptop. Workaround for some NV31 DVI artifacts. Also add nVidia license statement to nv man page to comply with requirement that it appear in "user documentation" commit fcd6d7b57edb4e396189a0a2683bcf08974f3b59 Author: Egbert Eich Date: Tue Oct 12 19:13:44 2004 +0000 Set fbdev mode as the default mode on PPC (Olaf Hering). Added support for IBM space saver keyboard (Stefan Dirsch). Added support for Cherry CyMotion Master XPress (Marcus Schaefer). Change order of SetDisplayDevice(), HWRestore(), UnbindGART() and RestoreBIOSMemSize() to be exactly opposite to the Save procedure in EnterVT() (Matthias Hopf, Alan Hourihane). Fix text mode restauration by removing the assumption that the register which determines which head is programmed is set. to the active head by the BIOS (Mark Vojkovich). When I wrote the resource code 5 years ago I made some assumptions which turned out to be false: I've assumed that the bus number of the PCI hostbridge would be the PCI bus the bridge links to. This is not correct. Fixing this assumption is not easy. However I hope that the attached patch will make the system work regardless as it 'ignores' host bridges from which the target bus is not known. This should not matter at all as we really don't care about host bridges (unless we have bridge specific code which retrieves information about the bridge). Fixed server crash on reset when a structure allocated in PreInit() was freed on CloseScreen(). Fixed ring buffer lock ups that happened because the structure that contained ringbuffer data was not zeroed after allocation. Fixed numerous warnings due to signed unsigned comparisons. programs/Xserver/hw/xfree86/drivers/nv/nv_driver.c: (NVBacklightEnable): Changed the order in which the sequencer registers and the backlight control registers are written. The sequencer control register need to be written first otherwise DPMS screen blanking produces vertical bars on a mobile device. lib/Xau/Imakefile: Build libXau static library PIC so it can be linked into toolkits that would like to wrap its functionality. commit 65e13fc61141b36a74c7a0e696d274574641ba20 Author: Adam Jackson Date: Fri Sep 17 03:04:52 2004 +0000 Bug #1192: Remove cfb support from drivers where its use is an option. Delete xf24_32bpp, as s3virge was the last user. Fix up some comments to refer to fb rather than cfb. commit bbafc6d2fcb21671de3cfb4ff973409405162e7a Author: Adam Jackson Date: Mon Aug 16 09:13:14 2004 +0000 Bug #1087: Make sure all the *Weak symbols are documented in the import list, so the loader knows not to complain about unresolved symbols when an optional module isn't loaded. This typically manifests as warnings about fbdevHW symbols when the user isn't using the framebuffer console. commit 8b6de7d388b7ae5169b2e38ad1a40981bc406cfc Author: Adam Jackson Date: Sat Jul 31 01:21:19 2004 +0000 Change several LoaderSymbol calls introduced by the bug #400 patch to *Weak() resolver functions. commit f745157f15befebe01552b103c573ccb94b8f388 Author: Adam Jackson Date: Fri Jul 30 20:30:53 2004 +0000 Bug #400 (partial): Driver fixes for the dlloader. When using dlloader, all framebuffer formats except cfb and the overlay modes should work, and r128 and radeon need to be loaded from the ati driver (both issues to be fixed soon). Tested on i740, s3virge, mach64, tdfx, vesa, and vga drivers. elfloader users shouldn't be affected. commit b9a6776bc04eb19343882bddb371b705677cfc14 Author: Eric Anholt Date: Wed Jun 16 09:23:29 2004 +0000 DRI XFree86-4_3_99_12-merge import commit cbf040deb0ab24425468ed60b3f3edd322ec210f Author: Eric Anholt Date: Wed Jun 16 09:23:29 2004 +0000 Initial revision commit 99a7f7ccce4e012bde57021cafcb132c3d599b96 Author: Egbert Eich Date: Fri Apr 23 19:42:10 2004 +0000 Merging XORG-CURRENT into trunk commit 708bb2eb0db3808e3edf94a5f22b3e3f5bb6488d Author: Egbert Eich Date: Sun Mar 14 08:33:30 2004 +0000 Importing vendor version xf86-4_4_99_1 on Sun Mar 14 00:26:39 PST 2004 commit fe564263d97e370817fc862e5c8ead09bfb70b58 Author: Egbert Eich Date: Wed Mar 3 12:12:23 2004 +0000 Importing vendor version xf86-4_4_0 on Wed Mar 3 04:09:24 PST 2004 commit 1a0be799ce4c479e39f8af86fcca3d192e83f455 Author: Egbert Eich Date: Thu Feb 26 13:35:54 2004 +0000 readding XFree86's cvs IDs commit ee5c9ef51cea88a630ecc652512603c42b39a533 Author: Egbert Eich Date: Thu Feb 26 09:23:23 2004 +0000 Importing vendor version xf86-4_3_99_903 on Wed Feb 26 01:21:00 PST 2004 commit d176fa338d2d83bcf8d7ec68ff1aec205a258fec Author: Kaleb Keithley Date: Mon Feb 23 20:35:04 2004 +0000 Import most of XFree86 4.4RC3. This import excludes files which have the new license. If we want to, later we can import 4.4RC3 again and pick up the files that have the new license, but for now the vendor branch is "pure." commit ea7a90e680a2e9259e471dd6677bcca12aeea79e Author: Egbert Eich Date: Thu Jan 29 08:08:36 2004 +0000 Importing vendor version xf86-012804-2330 on Thu Jan 29 00:06:33 PST 2004 commit a6c83bd98c34aacdf7a6b82ea83d6ae775479502 Author: Kaleb Keithley Date: Fri Dec 19 20:55:14 2003 +0000 XFree86 4.3.99.902 (RC 2) commit 935c2adf58a10e6d3caa75823b4e7a0689555c49 Author: Kaleb Keithley Date: Tue Nov 25 19:28:38 2003 +0000 XFree86 4.3.99.16 Bring the tree up to date for the Cygwin folks commit d84248580d5b05bb47a2ccdce7badfc884efa6da Author: Kaleb Keithley Date: Tue Nov 25 19:28:38 2003 +0000 Initial revision commit dd1a62caaaac6acc35584a2c8c525962d8574c95 Author: Kaleb Keithley Date: Fri Nov 14 16:48:55 2003 +0000 XFree86 4.3.0.1 commit 2f223903fba2bdee1623f3442d7580c809b428cc Author: Kaleb Keithley Date: Fri Nov 14 16:48:55 2003 +0000 Initial revision commit 542d57ea12b5461891a13a339e01ca9e1add124d Author: Kaleb Keithley Date: Fri Nov 14 15:54:48 2003 +0000 R6.6 is the Xorg base-line