/* $NetBSD: kern_ksyms.c,v 1.107.4.1 2023/02/24 14:12:51 martin Exp $ */ /*- * Copyright (c) 2008 The NetBSD Foundation, Inc. * All rights reserved. * * This code is derived from software developed for The NetBSD Foundation * by Andrew Doran. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ /* * Copyright (c) 2001, 2003 Anders Magnusson (ragge@ludd.luth.se). * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. The name of the author may not be used to endorse or promote products * derived from this software without specific prior written permission * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ /* * Code to deal with in-kernel symbol table management + /dev/ksyms. * * For each loaded module the symbol table info is kept track of by a * struct, placed in a circular list. The first entry is the kernel * symbol table. */ /* * TODO: * * Add support for mmap, poll. * Constify tables. * Constify db_symtab and move it to .rodata. */ #include __KERNEL_RCSID(0, "$NetBSD: kern_ksyms.c,v 1.107.4.1 2023/02/24 14:12:51 martin Exp $"); #if defined(_KERNEL) && defined(_KERNEL_OPT) #include "opt_copy_symtab.h" #include "opt_ddb.h" #include "opt_dtrace.h" #endif #define _KSYMS_PRIVATE #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #ifdef DDB #include #endif #include "ksyms.h" #if NKSYMS > 0 #include "ioconf.h" #endif struct ksyms_snapshot { uint64_t ks_refcnt; uint64_t ks_gen; struct uvm_object *ks_uobj; size_t ks_size; dev_t ks_dev; int ks_maxlen; }; #define KSYMS_MAX_ID 98304 #ifdef KDTRACE_HOOKS static uint32_t ksyms_nmap[KSYMS_MAX_ID]; /* sorted symbol table map */ #else static uint32_t *ksyms_nmap = NULL; #endif static int ksyms_maxlen; static bool ksyms_initted; static bool ksyms_loaded; static kmutex_t ksyms_lock __cacheline_aligned; static struct ksyms_symtab kernel_symtab; static kcondvar_t ksyms_cv; static struct lwp *ksyms_snapshotting; static struct ksyms_snapshot *ksyms_snapshot; static uint64_t ksyms_snapshot_gen; static pserialize_t ksyms_psz __read_mostly; static void ksyms_hdr_init(const void *); static void ksyms_sizes_calc(void); static struct ksyms_snapshot *ksyms_snapshot_alloc(int, size_t, dev_t, uint64_t); static void ksyms_snapshot_release(struct ksyms_snapshot *); #ifdef KSYMS_DEBUG #define FOLLOW_CALLS 1 #define FOLLOW_MORE_CALLS 2 #define FOLLOW_DEVKSYMS 4 static int ksyms_debug; #endif #define SYMTAB_FILLER "|This is the symbol table!" #ifdef makeoptions_COPY_SYMTAB extern char db_symtab[]; extern int db_symtabsize; #endif /* * used by savecore(8) so non-static */ struct ksyms_hdr ksyms_hdr; int ksyms_symsz; int ksyms_strsz; int ksyms_ctfsz; /* this is not currently used by savecore(8) */ TAILQ_HEAD(ksyms_symtab_queue, ksyms_symtab) ksyms_symtabs = TAILQ_HEAD_INITIALIZER(ksyms_symtabs); static struct pslist_head ksyms_symtabs_psz = PSLIST_INITIALIZER; static int ksyms_verify(const void *symstart, const void *strstart) { #if defined(DIAGNOSTIC) || defined(DEBUG) if (symstart == NULL) printf("ksyms: Symbol table not found\n"); if (strstart == NULL) printf("ksyms: String table not found\n"); if (symstart == NULL || strstart == NULL) printf("ksyms: Perhaps the kernel is stripped?\n"); #endif if (symstart == NULL || strstart == NULL) return 0; return 1; } /* * Finds a certain symbol name in a certain symbol table. */ static Elf_Sym * findsym(const char *name, struct ksyms_symtab *table, int type) { Elf_Sym *sym, *maxsym; int low, mid, high, nglob; char *str, *cmp; sym = table->sd_symstart; str = table->sd_strstart - table->sd_usroffset; nglob = table->sd_nglob; low = 0; high = nglob; /* * Start with a binary search of all global symbols in this table. * Global symbols must have unique names. */ while (low < high) { mid = (low + high) >> 1; cmp = sym[mid].st_name + str; if (cmp[0] < name[0] || strcmp(cmp, name) < 0) { low = mid + 1; } else { high = mid; } } KASSERT(low == high); if (__predict_true(low < nglob && strcmp(sym[low].st_name + str, name) == 0)) { KASSERT(ELF_ST_BIND(sym[low].st_info) == STB_GLOBAL); return &sym[low]; } /* * Perform a linear search of local symbols (rare). Many local * symbols with the same name can exist so are not included in * the binary search. */ if (type != KSYMS_EXTERN) { maxsym = sym + table->sd_symsize / sizeof(Elf_Sym); for (sym += nglob; sym < maxsym; sym++) { if (strcmp(name, sym->st_name + str) == 0) { return sym; } } } return NULL; } /* * The "attach" is in reality done in ksyms_init(). */ #if NKSYMS > 0 /* * ksyms can be loaded even if the kernel has a missing "pseudo-device ksyms" * statement because ddb and modules require it. Fixing it properly requires * fixing config to warn about required, but missing preudo-devices. For now, * if we don't have the pseudo-device we don't need the attach function; this * is fine, as it does nothing. */ void ksymsattach(int arg) { } #endif void ksyms_init(void) { #ifdef makeoptions_COPY_SYMTAB if (!ksyms_loaded && strncmp(db_symtab, SYMTAB_FILLER, sizeof(SYMTAB_FILLER))) { ksyms_addsyms_elf(db_symtabsize, db_symtab, db_symtab + db_symtabsize); } #endif if (!ksyms_initted) { mutex_init(&ksyms_lock, MUTEX_DEFAULT, IPL_NONE); cv_init(&ksyms_cv, "ksyms"); ksyms_psz = pserialize_create(); ksyms_initted = true; } } /* * Are any symbols available? */ bool ksyms_available(void) { return ksyms_loaded; } /* * Add a symbol table. * This is intended for use when the symbol table and its corresponding * string table are easily available. If they are embedded in an ELF * image, use addsymtab_elf() instead. * * name - Symbol's table name. * symstart, symsize - Address and size of the symbol table. * strstart, strsize - Address and size of the string table. * tab - Symbol table to be updated with this information. * newstart - Address to which the symbol table has to be copied during * shrinking. If NULL, it is not moved. */ static const char *addsymtab_strstart; static int addsymtab_compar(const void *a, const void *b) { const Elf_Sym *sa, *sb; sa = a; sb = b; /* * Split the symbol table into two, with globals at the start * and locals at the end. */ if (ELF_ST_BIND(sa->st_info) != ELF_ST_BIND(sb->st_info)) { if (ELF_ST_BIND(sa->st_info) == STB_GLOBAL) { return -1; } if (ELF_ST_BIND(sb->st_info) == STB_GLOBAL) { return 1; } } /* Within each band, sort by name. */ return strcmp(sa->st_name + addsymtab_strstart, sb->st_name + addsymtab_strstart); } static void addsymtab(const char *name, void *symstart, size_t symsize, void *strstart, size_t strsize, struct ksyms_symtab *tab, void *newstart, void *ctfstart, size_t ctfsize, uint32_t *nmap) { Elf_Sym *sym, *nsym, ts; int i, j, n, nglob; char *str; int nsyms = symsize / sizeof(Elf_Sym); int s; /* Sanity check for pre-allocated map table used during startup. */ if ((nmap == ksyms_nmap) && (nsyms >= KSYMS_MAX_ID)) { printf("kern_ksyms: ERROR %d > %d, increase KSYMS_MAX_ID\n", nsyms, KSYMS_MAX_ID); /* truncate for now */ nsyms = KSYMS_MAX_ID - 1; } tab->sd_symstart = symstart; tab->sd_symsize = symsize; tab->sd_strstart = strstart; tab->sd_strsize = strsize; tab->sd_name = name; tab->sd_minsym = UINTPTR_MAX; tab->sd_maxsym = 0; tab->sd_usroffset = 0; tab->sd_ctfstart = ctfstart; tab->sd_ctfsize = ctfsize; tab->sd_nmap = nmap; tab->sd_nmapsize = nsyms; #ifdef KSYMS_DEBUG printf("newstart %p sym %p ksyms_symsz %zu str %p strsz %zu send %p\n", newstart, symstart, symsize, strstart, strsize, tab->sd_strstart + tab->sd_strsize); #endif if (nmap) { memset(nmap, 0, nsyms * sizeof(uint32_t)); } /* Pack symbol table by removing all file name references. */ sym = tab->sd_symstart; nsym = (Elf_Sym *)newstart; str = tab->sd_strstart; nglob = 0; for (i = n = 0; i < nsyms; i++) { /* * This breaks CTF mapping, so don't do it when * DTrace is enabled. */ #ifndef KDTRACE_HOOKS /* * Remove useless symbols. * Should actually remove all typeless symbols. */ if (sym[i].st_name == 0) continue; /* Skip nameless entries */ if (sym[i].st_shndx == SHN_UNDEF) continue; /* Skip external references */ if (ELF_ST_TYPE(sym[i].st_info) == STT_FILE) continue; /* Skip filenames */ if (ELF_ST_TYPE(sym[i].st_info) == STT_NOTYPE && sym[i].st_value == 0 && strcmp(str + sym[i].st_name, "*ABS*") == 0) continue; /* XXX */ if (ELF_ST_TYPE(sym[i].st_info) == STT_NOTYPE && strcmp(str + sym[i].st_name, "gcc2_compiled.") == 0) continue; /* XXX */ #endif /* Save symbol. Set it as an absolute offset */ nsym[n] = sym[i]; #ifdef KDTRACE_HOOKS if (nmap != NULL) { /* * Save the size, replace it with the symbol id so * the mapping can be done after the cleanup and sort. */ nmap[i] = nsym[n].st_size; nsym[n].st_size = i + 1; /* zero is reserved */ } #endif if (sym[i].st_shndx != SHN_ABS) { nsym[n].st_shndx = SHBSS; } else { /* SHN_ABS is a magic value, don't overwrite it */ } j = strlen(nsym[n].st_name + str) + 1; if (j > ksyms_maxlen) ksyms_maxlen = j; nglob += (ELF_ST_BIND(nsym[n].st_info) == STB_GLOBAL); /* Compute min and max symbols. */ if (strcmp(str + sym[i].st_name, "*ABS*") != 0 && ELF_ST_TYPE(nsym[n].st_info) != STT_NOTYPE) { if (nsym[n].st_value < tab->sd_minsym) { tab->sd_minsym = nsym[n].st_value; } if (nsym[n].st_value > tab->sd_maxsym) { tab->sd_maxsym = nsym[n].st_value; } } n++; } /* Fill the rest of the record, and sort the symbols. */ tab->sd_symstart = nsym; tab->sd_symsize = n * sizeof(Elf_Sym); tab->sd_nglob = nglob; addsymtab_strstart = str; if (kheapsort(nsym, n, sizeof(Elf_Sym), addsymtab_compar, &ts) != 0) panic("addsymtab"); #ifdef KDTRACE_HOOKS /* * Build the mapping from original symbol id to new symbol table. * Deleted symbols will have a zero map, indices will be one based * instead of zero based. * Resulting map is sd_nmap[original_index] = new_index + 1 */ if (nmap != NULL) { int new; for (new = 0; new < n; new++) { uint32_t orig = nsym[new].st_size - 1; uint32_t size = nmap[orig]; nmap[orig] = new + 1; /* restore the size */ nsym[new].st_size = size; } } #endif KASSERT(strcmp(name, "netbsd") == 0 || mutex_owned(&ksyms_lock)); KASSERT(cold || mutex_owned(&ksyms_lock)); /* * Publish the symtab. Do this at splhigh to ensure ddb never * witnesses an inconsistent state of the queue, unless memory * is so corrupt that we crash in PSLIST_WRITER_INSERT_AFTER or * TAILQ_INSERT_TAIL. */ PSLIST_ENTRY_INIT(tab, sd_pslist); s = splhigh(); if (TAILQ_EMPTY(&ksyms_symtabs)) { PSLIST_WRITER_INSERT_HEAD(&ksyms_symtabs_psz, tab, sd_pslist); } else { struct ksyms_symtab *last; last = TAILQ_LAST(&ksyms_symtabs, ksyms_symtab_queue); PSLIST_WRITER_INSERT_AFTER(last, tab, sd_pslist); } TAILQ_INSERT_TAIL(&ksyms_symtabs, tab, sd_queue); splx(s); ksyms_sizes_calc(); ksyms_loaded = true; } /* * Setup the kernel symbol table stuff. */ void ksyms_addsyms_elf(int symsize, void *start, void *end) { int i, j; Elf_Shdr *shdr; char *symstart = NULL, *strstart = NULL; size_t strsize = 0; Elf_Ehdr *ehdr; char *ctfstart = NULL; size_t ctfsize = 0; if (symsize <= 0) { printf("[ Kernel symbol table missing! ]\n"); return; } /* Sanity check */ if (ALIGNED_POINTER(start, long) == 0) { printf("[ Kernel symbol table has bad start address %p ]\n", start); return; } ehdr = (Elf_Ehdr *)start; /* check if this is a valid ELF header */ /* No reason to verify arch type, the kernel is actually running! */ if (memcmp(ehdr->e_ident, ELFMAG, SELFMAG) || ehdr->e_ident[EI_CLASS] != ELFCLASS || ehdr->e_version > 1) { printf("[ Kernel symbol table invalid! ]\n"); return; /* nothing to do */ } /* Loaded header will be scratched in addsymtab */ ksyms_hdr_init(start); /* Find the symbol table and the corresponding string table. */ shdr = (Elf_Shdr *)((uint8_t *)start + ehdr->e_shoff); for (i = 1; i < ehdr->e_shnum; i++) { if (shdr[i].sh_type != SHT_SYMTAB) continue; if (shdr[i].sh_offset == 0) continue; symstart = (uint8_t *)start + shdr[i].sh_offset; symsize = shdr[i].sh_size; j = shdr[i].sh_link; if (shdr[j].sh_offset == 0) continue; /* Can this happen? */ strstart = (uint8_t *)start + shdr[j].sh_offset; strsize = shdr[j].sh_size; break; } #ifdef KDTRACE_HOOKS /* Find the CTF section */ shdr = (Elf_Shdr *)((uint8_t *)start + ehdr->e_shoff); if (ehdr->e_shstrndx != 0) { char *shstr = (uint8_t *)start + shdr[ehdr->e_shstrndx].sh_offset; for (i = 1; i < ehdr->e_shnum; i++) { #ifdef KSYMS_DEBUG printf("ksyms: checking %s\n", &shstr[shdr[i].sh_name]); #endif if (shdr[i].sh_type != SHT_PROGBITS) continue; if (strncmp(".SUNW_ctf", &shstr[shdr[i].sh_name], 10) != 0) continue; ctfstart = (uint8_t *)start + shdr[i].sh_offset; ctfsize = shdr[i].sh_size; ksyms_ctfsz = ctfsize; #ifdef DEBUG aprint_normal("Found CTF at %p, size 0x%zx\n", ctfstart, ctfsize); #endif break; } #ifdef DEBUG } else { printf("ksyms: e_shstrndx == 0\n"); #endif } #endif if (!ksyms_verify(symstart, strstart)) return; addsymtab("netbsd", symstart, symsize, strstart, strsize, &kernel_symtab, symstart, ctfstart, ctfsize, ksyms_nmap); #ifdef DEBUG aprint_normal("Loaded initial symtab at %p, strtab at %p, # entries %ld\n", kernel_symtab.sd_symstart, kernel_symtab.sd_strstart, (long)kernel_symtab.sd_symsize/sizeof(Elf_Sym)); #endif /* Should be no snapshot to invalidate yet. */ KASSERT(ksyms_snapshot == NULL); } /* * Setup the kernel symbol table stuff. * Use this when the address of the symbol and string tables are known; * otherwise use ksyms_init with an ELF image. * We need to pass a minimal ELF header which will later be completed by * ksyms_hdr_init and handed off to userland through /dev/ksyms. We use * a void *rather than a pointer to avoid exposing the Elf_Ehdr type. */ void ksyms_addsyms_explicit(void *ehdr, void *symstart, size_t symsize, void *strstart, size_t strsize) { if (!ksyms_verify(symstart, strstart)) return; ksyms_hdr_init(ehdr); addsymtab("netbsd", symstart, symsize, strstart, strsize, &kernel_symtab, symstart, NULL, 0, ksyms_nmap); /* Should be no snapshot to invalidate yet. */ KASSERT(ksyms_snapshot == NULL); } /* * Get the value associated with a symbol. * "mod" is the module name, or null if any module. * "sym" is the symbol name. * "val" is a pointer to the corresponding value, if call succeeded. * Returns 0 if success or ENOENT if no such entry. * * If symp is nonnull, caller must hold ksyms_lock or module_lock, have * ksyms_opencnt nonzero, be in a pserialize read section, be in ddb * with all other CPUs quiescent. */ int ksyms_getval_unlocked(const char *mod, const char *sym, Elf_Sym **symp, unsigned long *val, int type) { struct ksyms_symtab *st; Elf_Sym *es; int s, error = ENOENT; #ifdef KSYMS_DEBUG if (ksyms_debug & FOLLOW_CALLS) printf("%s: mod %s sym %s valp %p\n", __func__, mod, sym, val); #endif s = pserialize_read_enter(); PSLIST_READER_FOREACH(st, &ksyms_symtabs_psz, struct ksyms_symtab, sd_pslist) { if (mod != NULL && strcmp(st->sd_name, mod)) continue; if ((es = findsym(sym, st, type)) != NULL) { *val = es->st_value; if (symp) *symp = es; error = 0; break; } } pserialize_read_exit(s); return error; } int ksyms_getval(const char *mod, const char *sym, unsigned long *val, int type) { if (!ksyms_loaded) return ENOENT; /* No locking needed -- we read the table pserialized. */ return ksyms_getval_unlocked(mod, sym, NULL, val, type); } /* * ksyms_get_mod(mod) * * Return the symtab for the given module name. Caller must ensure * that the module cannot be unloaded until after this returns. */ struct ksyms_symtab * ksyms_get_mod(const char *mod) { struct ksyms_symtab *st; int s; s = pserialize_read_enter(); PSLIST_READER_FOREACH(st, &ksyms_symtabs_psz, struct ksyms_symtab, sd_pslist) { if (mod != NULL && strcmp(st->sd_name, mod)) continue; break; } pserialize_read_exit(s); return st; } /* * ksyms_mod_foreach() * * Iterate over the symbol table of the specified module, calling the callback * handler for each symbol. Stop iterating if the handler return is non-zero. * */ int ksyms_mod_foreach(const char *mod, ksyms_callback_t callback, void *opaque) { struct ksyms_symtab *st; Elf_Sym *sym, *maxsym; char *str; int symindx; if (!ksyms_loaded) return ENOENT; mutex_enter(&ksyms_lock); /* find the module */ TAILQ_FOREACH(st, &ksyms_symtabs, sd_queue) { if (mod != NULL && strcmp(st->sd_name, mod)) continue; sym = st->sd_symstart; str = st->sd_strstart - st->sd_usroffset; /* now iterate through the symbols */ maxsym = sym + st->sd_symsize / sizeof(Elf_Sym); for (symindx = 0; sym < maxsym; sym++, symindx++) { if (callback(str + sym->st_name, symindx, (void *)sym->st_value, sym->st_size, sym->st_info, opaque) != 0) { break; } } } mutex_exit(&ksyms_lock); return 0; } /* * Get "mod" and "symbol" associated with an address. * Returns 0 if success or ENOENT if no such entry. * * Caller must hold ksyms_lock or module_lock, have ksyms_opencnt * nonzero, be in a pserialize read section, or be in ddb with all * other CPUs quiescent. */ int ksyms_getname(const char **mod, const char **sym, vaddr_t v, int f) { struct ksyms_symtab *st; Elf_Sym *les, *es = NULL; vaddr_t laddr = 0; const char *lmod = NULL; char *stable = NULL; int type, i, sz; if (!ksyms_loaded) return ENOENT; PSLIST_READER_FOREACH(st, &ksyms_symtabs_psz, struct ksyms_symtab, sd_pslist) { if (v < st->sd_minsym || v > st->sd_maxsym) continue; sz = st->sd_symsize/sizeof(Elf_Sym); for (i = 0; i < sz; i++) { les = st->sd_symstart + i; type = ELF_ST_TYPE(les->st_info); if ((f & KSYMS_PROC) && (type != STT_FUNC)) continue; if (type == STT_NOTYPE) continue; if (((f & KSYMS_ANY) == 0) && (type != STT_FUNC) && (type != STT_OBJECT)) continue; if ((les->st_value <= v) && (les->st_value > laddr)) { laddr = les->st_value; es = les; lmod = st->sd_name; stable = st->sd_strstart - st->sd_usroffset; } } } if (es == NULL) return ENOENT; if ((f & KSYMS_EXACT) && (v != es->st_value)) return ENOENT; if (mod) *mod = lmod; if (sym) *sym = stable + es->st_name; return 0; } /* * Add a symbol table from a loadable module. */ void ksyms_modload(const char *name, void *symstart, vsize_t symsize, char *strstart, vsize_t strsize) { struct ksyms_symtab *st; struct ksyms_snapshot *ks; void *nmap; st = kmem_zalloc(sizeof(*st), KM_SLEEP); nmap = kmem_zalloc(symsize / sizeof(Elf_Sym) * sizeof (uint32_t), KM_SLEEP); mutex_enter(&ksyms_lock); addsymtab(name, symstart, symsize, strstart, strsize, st, symstart, NULL, 0, nmap); ks = ksyms_snapshot; ksyms_snapshot = NULL; mutex_exit(&ksyms_lock); if (ks) ksyms_snapshot_release(ks); } /* * Remove a symbol table from a loadable module. */ void ksyms_modunload(const char *name) { struct ksyms_symtab *st; struct ksyms_snapshot *ks; int s; mutex_enter(&ksyms_lock); TAILQ_FOREACH(st, &ksyms_symtabs, sd_queue) { if (strcmp(name, st->sd_name) != 0) continue; break; } KASSERT(st != NULL); /* Wait for any snapshot in progress to complete. */ while (ksyms_snapshotting) cv_wait(&ksyms_cv, &ksyms_lock); /* * Remove the symtab. Do this at splhigh to ensure ddb never * witnesses an inconsistent state of the queue, unless memory * is so corrupt that we crash in TAILQ_REMOVE or * PSLIST_WRITER_REMOVE. */ s = splhigh(); TAILQ_REMOVE(&ksyms_symtabs, st, sd_queue); PSLIST_WRITER_REMOVE(st, sd_pslist); splx(s); /* * And wait a grace period, in case there are any pserialized * readers in flight. */ pserialize_perform(ksyms_psz); PSLIST_ENTRY_DESTROY(st, sd_pslist); /* Recompute the ksyms sizes now that we've removed st. */ ksyms_sizes_calc(); /* Invalidate the global ksyms snapshot. */ ks = ksyms_snapshot; ksyms_snapshot = NULL; mutex_exit(&ksyms_lock); /* * No more references are possible. Free the name map and the * symtab itself, which we had allocated in ksyms_modload. */ kmem_free(st->sd_nmap, st->sd_nmapsize * sizeof(uint32_t)); kmem_free(st, sizeof(*st)); /* Release the formerly global ksyms snapshot, if any. */ if (ks) ksyms_snapshot_release(ks); } #ifdef DDB /* * Keep sifting stuff here, to avoid export of ksyms internals. * * Systems is expected to be quiescent, so no locking done. */ int ksyms_sift(char *mod, char *sym, int mode) { struct ksyms_symtab *st; char *sb; int i, sz; if (!ksyms_loaded) return ENOENT; TAILQ_FOREACH(st, &ksyms_symtabs, sd_queue) { if (mod && strcmp(mod, st->sd_name)) continue; sb = st->sd_strstart - st->sd_usroffset; sz = st->sd_symsize/sizeof(Elf_Sym); for (i = 0; i < sz; i++) { Elf_Sym *les = st->sd_symstart + i; char c; if (strstr(sb + les->st_name, sym) == NULL) continue; if (mode == 'F') { switch (ELF_ST_TYPE(les->st_info)) { case STT_OBJECT: c = '+'; break; case STT_FUNC: c = '*'; break; case STT_SECTION: c = '&'; break; case STT_FILE: c = '/'; break; default: c = ' '; break; } db_printf("%s%c ", sb + les->st_name, c); } else db_printf("%s ", sb + les->st_name); } } return ENOENT; } #endif /* DDB */ /* * In case we exposing the symbol table to the userland using the pseudo- * device /dev/ksyms, it is easier to provide all the tables as one. * However, it means we have to change all the st_name fields for the * symbols so they match the ELF image that the userland will read * through the device. * * The actual (correct) value of st_name is preserved through a global * offset stored in the symbol table structure. * * Call with ksyms_lock held. */ static void ksyms_sizes_calc(void) { struct ksyms_symtab *st; int i, delta; KASSERT(cold || mutex_owned(&ksyms_lock)); ksyms_symsz = ksyms_strsz = 0; TAILQ_FOREACH(st, &ksyms_symtabs, sd_queue) { delta = ksyms_strsz - st->sd_usroffset; if (delta != 0) { for (i = 0; i < st->sd_symsize/sizeof(Elf_Sym); i++) st->sd_symstart[i].st_name += delta; st->sd_usroffset = ksyms_strsz; } ksyms_symsz += st->sd_symsize; ksyms_strsz += st->sd_strsize; } } static void ksyms_fill_note(void) { int32_t *note = ksyms_hdr.kh_note; note[0] = ELF_NOTE_NETBSD_NAMESZ; note[1] = ELF_NOTE_NETBSD_DESCSZ; note[2] = ELF_NOTE_TYPE_NETBSD_TAG; memcpy(¬e[3], "NetBSD\0", 8); note[5] = __NetBSD_Version__; } static void ksyms_hdr_init(const void *hdraddr) { /* Copy the loaded elf exec header */ memcpy(&ksyms_hdr.kh_ehdr, hdraddr, sizeof(Elf_Ehdr)); /* Set correct program/section header sizes, offsets and numbers */ ksyms_hdr.kh_ehdr.e_phoff = offsetof(struct ksyms_hdr, kh_phdr[0]); ksyms_hdr.kh_ehdr.e_phentsize = sizeof(Elf_Phdr); ksyms_hdr.kh_ehdr.e_phnum = NPRGHDR; ksyms_hdr.kh_ehdr.e_shoff = offsetof(struct ksyms_hdr, kh_shdr[0]); ksyms_hdr.kh_ehdr.e_shentsize = sizeof(Elf_Shdr); ksyms_hdr.kh_ehdr.e_shnum = NSECHDR; ksyms_hdr.kh_ehdr.e_shstrndx = SHSTRTAB; /* Text/data - fake */ ksyms_hdr.kh_phdr[0].p_type = PT_LOAD; ksyms_hdr.kh_phdr[0].p_memsz = (unsigned long)-1L; ksyms_hdr.kh_phdr[0].p_flags = PF_R | PF_X | PF_W; #define SHTCOPY(name) strlcpy(&ksyms_hdr.kh_strtab[offs], (name), \ sizeof(ksyms_hdr.kh_strtab) - offs), offs += sizeof(name) uint32_t offs = 1; /* First section header ".note.netbsd.ident" */ ksyms_hdr.kh_shdr[SHNOTE].sh_name = offs; ksyms_hdr.kh_shdr[SHNOTE].sh_type = SHT_NOTE; ksyms_hdr.kh_shdr[SHNOTE].sh_offset = offsetof(struct ksyms_hdr, kh_note[0]); ksyms_hdr.kh_shdr[SHNOTE].sh_size = sizeof(ksyms_hdr.kh_note); ksyms_hdr.kh_shdr[SHNOTE].sh_addralign = sizeof(int); SHTCOPY(".note.netbsd.ident"); ksyms_fill_note(); /* Second section header; ".symtab" */ ksyms_hdr.kh_shdr[SYMTAB].sh_name = offs; ksyms_hdr.kh_shdr[SYMTAB].sh_type = SHT_SYMTAB; ksyms_hdr.kh_shdr[SYMTAB].sh_offset = sizeof(struct ksyms_hdr); /* ksyms_hdr.kh_shdr[SYMTAB].sh_size = filled in at open */ ksyms_hdr.kh_shdr[SYMTAB].sh_link = STRTAB; /* Corresponding strtab */ ksyms_hdr.kh_shdr[SYMTAB].sh_addralign = sizeof(long); ksyms_hdr.kh_shdr[SYMTAB].sh_entsize = sizeof(Elf_Sym); SHTCOPY(".symtab"); /* Third section header; ".strtab" */ ksyms_hdr.kh_shdr[STRTAB].sh_name = offs; ksyms_hdr.kh_shdr[STRTAB].sh_type = SHT_STRTAB; /* ksyms_hdr.kh_shdr[STRTAB].sh_offset = filled in at open */ /* ksyms_hdr.kh_shdr[STRTAB].sh_size = filled in at open */ ksyms_hdr.kh_shdr[STRTAB].sh_addralign = sizeof(char); SHTCOPY(".strtab"); /* Fourth section, ".shstrtab" */ ksyms_hdr.kh_shdr[SHSTRTAB].sh_name = offs; ksyms_hdr.kh_shdr[SHSTRTAB].sh_type = SHT_STRTAB; ksyms_hdr.kh_shdr[SHSTRTAB].sh_offset = offsetof(struct ksyms_hdr, kh_strtab); ksyms_hdr.kh_shdr[SHSTRTAB].sh_size = SHSTRSIZ; ksyms_hdr.kh_shdr[SHSTRTAB].sh_addralign = sizeof(char); SHTCOPY(".shstrtab"); /* Fifth section, ".bss". All symbols reside here. */ ksyms_hdr.kh_shdr[SHBSS].sh_name = offs; ksyms_hdr.kh_shdr[SHBSS].sh_type = SHT_NOBITS; ksyms_hdr.kh_shdr[SHBSS].sh_offset = 0; ksyms_hdr.kh_shdr[SHBSS].sh_size = (unsigned long)-1L; ksyms_hdr.kh_shdr[SHBSS].sh_addralign = PAGE_SIZE; ksyms_hdr.kh_shdr[SHBSS].sh_flags = SHF_ALLOC | SHF_EXECINSTR; SHTCOPY(".bss"); /* Sixth section header; ".SUNW_ctf" */ ksyms_hdr.kh_shdr[SHCTF].sh_name = offs; ksyms_hdr.kh_shdr[SHCTF].sh_type = SHT_PROGBITS; /* ksyms_hdr.kh_shdr[SHCTF].sh_offset = filled in at open */ /* ksyms_hdr.kh_shdr[SHCTF].sh_size = filled in at open */ ksyms_hdr.kh_shdr[SHCTF].sh_link = SYMTAB; /* Corresponding symtab */ ksyms_hdr.kh_shdr[SHCTF].sh_addralign = sizeof(char); SHTCOPY(".SUNW_ctf"); } static struct ksyms_snapshot * ksyms_snapshot_alloc(int maxlen, size_t size, dev_t dev, uint64_t gen) { struct ksyms_snapshot *ks; ks = kmem_zalloc(sizeof(*ks), KM_SLEEP); ks->ks_refcnt = 1; ks->ks_gen = gen; ks->ks_uobj = uao_create(size, 0); ks->ks_size = size; ks->ks_dev = dev; ks->ks_maxlen = maxlen; return ks; } static void ksyms_snapshot_release(struct ksyms_snapshot *ks) { uint64_t refcnt; mutex_enter(&ksyms_lock); refcnt = --ks->ks_refcnt; mutex_exit(&ksyms_lock); if (refcnt) return; uao_detach(ks->ks_uobj); kmem_free(ks, sizeof(*ks)); } static int ubc_copyfrombuf(struct uvm_object *uobj, struct uio *uio, const void *buf, size_t n) { struct iovec iov = { .iov_base = __UNCONST(buf), .iov_len = n }; uio->uio_iov = &iov; uio->uio_iovcnt = 1; uio->uio_resid = n; return ubc_uiomove(uobj, uio, n, UVM_ADV_SEQUENTIAL, UBC_WRITE); } static int ksyms_take_snapshot(struct ksyms_snapshot *ks, struct ksyms_symtab *last) { struct uvm_object *uobj = ks->ks_uobj; struct uio uio; struct ksyms_symtab *st; int error; /* Caller must have initiated snapshotting. */ KASSERT(ksyms_snapshotting == curlwp); /* Start a uio transfer to reuse incrementally. */ uio.uio_offset = 0; uio.uio_rw = UIO_WRITE; /* write from buffer to uobj */ UIO_SETUP_SYSSPACE(&uio); /* * First: Copy out the ELF header. */ error = ubc_copyfrombuf(uobj, &uio, &ksyms_hdr, sizeof(ksyms_hdr)); if (error) return error; /* * Copy out the symbol table. The list of symtabs is * guaranteed to be nonempty because we always have an entry * for the main kernel. We stop at last, not at the end of the * tailq or NULL, because entries beyond last are not included * in this snapshot (and may not be fully initialized memory as * we witness it). */ KASSERT(uio.uio_offset == sizeof(struct ksyms_hdr)); for (st = TAILQ_FIRST(&ksyms_symtabs); ; st = TAILQ_NEXT(st, sd_queue)) { error = ubc_copyfrombuf(uobj, &uio, st->sd_symstart, st->sd_symsize); if (error) return error; if (st == last) break; } /* * Copy out the string table */ KASSERT(uio.uio_offset == sizeof(struct ksyms_hdr) + ksyms_hdr.kh_shdr[SYMTAB].sh_size); for (st = TAILQ_FIRST(&ksyms_symtabs); ; st = TAILQ_NEXT(st, sd_queue)) { error = ubc_copyfrombuf(uobj, &uio, st->sd_strstart, st->sd_strsize); if (error) return error; if (st == last) break; } /* * Copy out the CTF table. */ KASSERT(uio.uio_offset == sizeof(struct ksyms_hdr) + ksyms_hdr.kh_shdr[SYMTAB].sh_size + ksyms_hdr.kh_shdr[STRTAB].sh_size); st = TAILQ_FIRST(&ksyms_symtabs); if (st->sd_ctfstart != NULL) { error = ubc_copyfrombuf(uobj, &uio, st->sd_ctfstart, st->sd_ctfsize); if (error) return error; } KASSERT(uio.uio_offset == sizeof(struct ksyms_hdr) + ksyms_hdr.kh_shdr[SYMTAB].sh_size + ksyms_hdr.kh_shdr[STRTAB].sh_size + ksyms_hdr.kh_shdr[SHCTF].sh_size); KASSERT(uio.uio_offset == ks->ks_size); return 0; } static const struct fileops ksyms_fileops; static int ksymsopen(dev_t dev, int flags, int devtype, struct lwp *l) { struct file *fp = NULL; int fd = -1; struct ksyms_snapshot *ks = NULL; size_t size; struct ksyms_symtab *last; int maxlen; uint64_t gen; int error; if (minor(dev) != 0 || !ksyms_loaded) return ENXIO; /* Allocate a private file. */ error = fd_allocfile(&fp, &fd); if (error) return error; mutex_enter(&ksyms_lock); /* * Wait until we have a snapshot, or until there is no snapshot * being taken right now so we can take one. */ while ((ks = ksyms_snapshot) == NULL && ksyms_snapshotting) { error = cv_wait_sig(&ksyms_cv, &ksyms_lock); if (error) goto out; } /* * If there's a usable snapshot, increment its reference count * (can't overflow, 64-bit) and just reuse it. */ if (ks) { ks->ks_refcnt++; goto out; } /* Find the current length of the symtab object. */ size = sizeof(struct ksyms_hdr); size += ksyms_strsz; size += ksyms_symsz; size += ksyms_ctfsz; /* Start a new snapshot. */ ksyms_hdr.kh_shdr[SYMTAB].sh_size = ksyms_symsz; ksyms_hdr.kh_shdr[SYMTAB].sh_info = ksyms_symsz / sizeof(Elf_Sym); ksyms_hdr.kh_shdr[STRTAB].sh_offset = ksyms_symsz + ksyms_hdr.kh_shdr[SYMTAB].sh_offset; ksyms_hdr.kh_shdr[STRTAB].sh_size = ksyms_strsz; ksyms_hdr.kh_shdr[SHCTF].sh_offset = ksyms_strsz + ksyms_hdr.kh_shdr[STRTAB].sh_offset; ksyms_hdr.kh_shdr[SHCTF].sh_size = ksyms_ctfsz; last = TAILQ_LAST(&ksyms_symtabs, ksyms_symtab_queue); maxlen = ksyms_maxlen; gen = ksyms_snapshot_gen++; /* * Prevent ksyms entries from being removed while we take the * snapshot. */ KASSERT(ksyms_snapshotting == NULL); ksyms_snapshotting = curlwp; mutex_exit(&ksyms_lock); /* Create a snapshot and write the symtab to it. */ ks = ksyms_snapshot_alloc(maxlen, size, dev, gen); error = ksyms_take_snapshot(ks, last); /* * Snapshot creation is done. Wake up anyone waiting to remove * entries (module unload). */ mutex_enter(&ksyms_lock); KASSERTMSG(ksyms_snapshotting == curlwp, "lwp %p stole snapshot", ksyms_snapshotting); ksyms_snapshotting = NULL; cv_broadcast(&ksyms_cv); /* If we failed, give up. */ if (error) goto out; /* Cache the snapshot for the next reader. */ KASSERT(ksyms_snapshot == NULL); ksyms_snapshot = ks; ks->ks_refcnt++; KASSERT(ks->ks_refcnt == 2); out: mutex_exit(&ksyms_lock); if (error) { if (fp) fd_abort(curproc, fp, fd); if (ks) ksyms_snapshot_release(ks); } else { KASSERT(fp); KASSERT(ks); error = fd_clone(fp, fd, flags, &ksyms_fileops, ks); KASSERTMSG(error == EMOVEFD, "error=%d", error); } return error; } static int ksymsclose(struct file *fp) { struct ksyms_snapshot *ks = fp->f_data; ksyms_snapshot_release(ks); return 0; } static int ksymsread(struct file *fp, off_t *offp, struct uio *uio, kauth_cred_t cred, int flags) { const struct ksyms_snapshot *ks = fp->f_data; size_t count; int error; /* * Since we don't have a per-object lock, we might as well use * the struct file lock to serialize access to fp->f_offset -- * but if the caller isn't relying on or updating fp->f_offset, * there's no need to do even that. We could use ksyms_lock, * but why bother with a global lock if not needed? Either * way, the lock we use here must agree with what ksymsseek * takes (nothing else in ksyms uses fp->f_offset). */ if (offp == &fp->f_offset) mutex_enter(&fp->f_lock); /* Refuse negative offsets. */ if (*offp < 0) { error = EINVAL; goto out; } /* Return nothing at or past end of file. */ if (*offp >= ks->ks_size) { error = 0; goto out; } /* * 1. Set up the uio to transfer from offset *offp. * 2. Transfer as many bytes as we can (at most uio->uio_resid * or what's left in the ksyms). * 3. If requested, update *offp to reflect the number of bytes * transferred. */ uio->uio_offset = *offp; count = uio->uio_resid; error = ubc_uiomove(ks->ks_uobj, uio, MIN(count, ks->ks_size - *offp), UVM_ADV_SEQUENTIAL, UBC_READ|UBC_PARTIALOK); if (flags & FOF_UPDATE_OFFSET) *offp += count - uio->uio_resid; out: if (offp == &fp->f_offset) mutex_exit(&fp->f_lock); return error; } static int ksymsstat(struct file *fp, struct stat *st) { const struct ksyms_snapshot *ks = fp->f_data; memset(st, 0, sizeof(*st)); st->st_dev = NODEV; st->st_ino = 0; st->st_mode = S_IFCHR; st->st_nlink = 1; st->st_uid = kauth_cred_geteuid(fp->f_cred); st->st_gid = kauth_cred_getegid(fp->f_cred); st->st_rdev = ks->ks_dev; st->st_size = ks->ks_size; /* zero time */ st->st_blksize = MAXPHYS; /* XXX arbitrary */ st->st_blocks = 0; st->st_gen = ks->ks_gen; return 0; } static int ksymsmmap(struct file *fp, off_t *offp, size_t nbytes, int prot, int *flagsp, int *advicep, struct uvm_object **uobjp, int *maxprotp) { const struct ksyms_snapshot *ks = fp->f_data; /* uvm_mmap guarantees page-aligned offset and size. */ KASSERT(*offp == round_page(*offp)); KASSERT(nbytes == round_page(nbytes)); KASSERT(nbytes > 0); /* Refuse negative offsets. */ if (*offp < 0) return EINVAL; /* Refuse mappings that pass the end of file. */ if (nbytes > round_page(ks->ks_size) || *offp > round_page(ks->ks_size) - nbytes) return EINVAL; /* XXX ??? */ /* Success! */ uao_reference(ks->ks_uobj); *advicep = UVM_ADV_SEQUENTIAL; *uobjp = ks->ks_uobj; *maxprotp = prot & VM_PROT_READ; return 0; } static int ksymsseek(struct file *fp, off_t delta, int whence, off_t *newoffp, int flags) { const off_t OFF_MAX = __type_max(off_t); struct ksyms_snapshot *ks = fp->f_data; off_t base, newoff; int error; mutex_enter(&fp->f_lock); switch (whence) { case SEEK_CUR: base = fp->f_offset; break; case SEEK_END: base = ks->ks_size; break; case SEEK_SET: base = 0; break; default: error = EINVAL; goto out; } /* Check for arithmetic overflow and reject negative offsets. */ if (base < 0 || delta > OFF_MAX - base || base + delta < 0) { error = EINVAL; goto out; } /* Compute the new offset. */ newoff = base + delta; /* Success! */ if (newoffp) *newoffp = newoff; if (flags & FOF_UPDATE_OFFSET) fp->f_offset = newoff; error = 0; out: mutex_exit(&fp->f_lock); return error; } __CTASSERT(offsetof(struct ksyms_ogsymbol, kg_name) == offsetof(struct ksyms_gsymbol, kg_name)); __CTASSERT(offsetof(struct ksyms_gvalue, kv_name) == offsetof(struct ksyms_gsymbol, kg_name)); static int ksymsioctl(struct file *fp, u_long cmd, void *data) { struct ksyms_snapshot *ks = fp->f_data; struct ksyms_ogsymbol *okg = (struct ksyms_ogsymbol *)data; struct ksyms_gsymbol *kg = (struct ksyms_gsymbol *)data; struct ksyms_gvalue *kv = (struct ksyms_gvalue *)data; struct ksyms_symtab *st; Elf_Sym *sym = NULL, copy; unsigned long val; int error = 0; char *str = NULL; int len, s; /* Read cached ksyms_maxlen. */ len = ks->ks_maxlen; if (cmd == OKIOCGVALUE || cmd == OKIOCGSYMBOL || cmd == KIOCGVALUE || cmd == KIOCGSYMBOL) { str = kmem_alloc(len, KM_SLEEP); if ((error = copyinstr(kg->kg_name, str, len, NULL)) != 0) { kmem_free(str, len); return error; } } switch (cmd) { case OKIOCGVALUE: /* * Use the in-kernel symbol lookup code for fast * retreival of a value. */ error = ksyms_getval(NULL, str, &val, KSYMS_EXTERN); if (error == 0) error = copyout(&val, okg->kg_value, sizeof(long)); kmem_free(str, len); break; case OKIOCGSYMBOL: /* * Use the in-kernel symbol lookup code for fast * retreival of a symbol. */ s = pserialize_read_enter(); PSLIST_READER_FOREACH(st, &ksyms_symtabs_psz, struct ksyms_symtab, sd_pslist) { if ((sym = findsym(str, st, KSYMS_ANY)) == NULL) continue; #ifdef notdef /* Skip if bad binding */ if (ELF_ST_BIND(sym->st_info) != STB_GLOBAL) { sym = NULL; continue; } #endif break; } if (sym != NULL) { memcpy(©, sym, sizeof(copy)); pserialize_read_exit(s); error = copyout(©, okg->kg_sym, sizeof(Elf_Sym)); } else { pserialize_read_exit(s); error = ENOENT; } kmem_free(str, len); break; case KIOCGVALUE: /* * Use the in-kernel symbol lookup code for fast * retreival of a value. */ error = ksyms_getval(NULL, str, &val, KSYMS_EXTERN); if (error == 0) kv->kv_value = val; kmem_free(str, len); break; case KIOCGSYMBOL: /* * Use the in-kernel symbol lookup code for fast * retreival of a symbol. */ s = pserialize_read_enter(); PSLIST_READER_FOREACH(st, &ksyms_symtabs_psz, struct ksyms_symtab, sd_pslist) { if ((sym = findsym(str, st, KSYMS_ANY)) == NULL) continue; #ifdef notdef /* Skip if bad binding */ if (ELF_ST_BIND(sym->st_info) != STB_GLOBAL) { sym = NULL; continue; } #endif break; } if (sym != NULL) { kg->kg_sym = *sym; } else { error = ENOENT; } pserialize_read_exit(s); kmem_free(str, len); break; case KIOCGSIZE: /* * Get total size of symbol table. */ *(int *)data = ks->ks_size; break; default: error = ENOTTY; break; } return error; } const struct cdevsw ksyms_cdevsw = { .d_open = ksymsopen, .d_close = noclose, .d_read = noread, .d_write = nowrite, .d_ioctl = noioctl, .d_stop = nostop, .d_tty = notty, .d_poll = nopoll, .d_mmap = nommap, .d_kqfilter = nokqfilter, .d_discard = nodiscard, .d_flag = D_OTHER | D_MPSAFE }; static const struct fileops ksyms_fileops = { .fo_name = "ksyms", .fo_read = ksymsread, .fo_write = fbadop_write, .fo_ioctl = ksymsioctl, .fo_fcntl = fnullop_fcntl, .fo_poll = fnullop_poll, .fo_stat = ksymsstat, .fo_close = ksymsclose, .fo_kqfilter = fnullop_kqfilter, .fo_restart = fnullop_restart, .fo_mmap = ksymsmmap, .fo_seek = ksymsseek, };