ISAR/libmdbx
3
0
Fork 0
mirror of https://github.com/isar/libmdbx.git synced 2025-01-02 02:24:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

mdbx: partial workaround for ASAN false-positive (due internal bug).

Browse Source 
Change-Id: Ib86e187e58f3efbb5e5da95219f1b6e990022b9a
This commit is contained in:
Leonid Yuriev 2019-10-11 09:12:59 +03:00
parent 456f77abec
commit 4d692821cb
1 changed files with 28 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

39
src/elements/core.c
View File

@ -2077,6 +2077,7 @@ static size_t bytes_align2os_bytes(const MDBX_env *env, size_t bytes) {
}
static void __cold mdbx_kill_page(MDBX_env *env, MDBX_page *mp) {
mdbx_assert(env, mp->mp_pgno >= NUM_METAS);
const size_t len = env->me_psize - PAGEHDRSZ;
void *ptr = (env->me_flags & MDBX_WRITEMAP)
? &mp->mp_data
@ -3719,9 +3720,9 @@ static void mdbx_txn_valgrind(MDBX_env *env, MDBX_txn *txn) {
}
last = mdbx_find_largest_this(env, last);
const pgno_t edge = env->me_poison_edge ? env->me_poison_edge
: bytes2pgno(env, env->me_mapsize);
const pgno_t edge = env->me_poison_edge;
if (edge > last) {
mdbx_assert(env, last >= NUM_METAS);
env->me_poison_edge = last;
VALGRIND_MAKE_MEM_NOACCESS(env->me_map + pgno2bytes(env, last),
pgno2bytes(env, edge - last));
@ -6153,10 +6154,17 @@ static int mdbx_sync_locked(MDBX_env *env, unsigned flags,
const pgno_t largest_pgno = mdbx_find_largest(
env, (head->mm_geo.next > pending->mm_geo.next) ? head->mm_geo.next
: pending->mm_geo.next);
VALGRIND_MAKE_MEM_NOACCESS(env->me_map + pgno2bytes(env, largest_pgno),
env->me_mapsize - pgno2bytes(env, largest_pgno));
ASAN_POISON_MEMORY_REGION(env->me_map + pgno2bytes(env, largest_pgno),
env->me_mapsize - pgno2bytes(env, largest_pgno));
mdbx_assert(env, largest_pgno >= NUM_METAS);
#if defined(MDBX_USE_VALGRIND) || defined(__SANITIZE_ADDRESS__)
const pgno_t edge = env->me_poison_edge;
if (edge > largest_pgno) {
env->me_poison_edge = largest_pgno;
VALGRIND_MAKE_MEM_NOACCESS(env->me_map + pgno2bytes(env, largest_pgno),
pgno2bytes(env, edge - largest_pgno));
ASAN_POISON_MEMORY_REGION(env->me_map + pgno2bytes(env, largest_pgno),
pgno2bytes(env, edge - largest_pgno));
}
#endif /* MDBX_USE_VALGRIND */
const size_t largest_aligned2os_bytes =
pgno_align2os_bytes(env, largest_pgno);
const pgno_t largest_aligned2os_pgno =
@ -6503,16 +6511,11 @@ bailout:
static int __cold mdbx_env_map(MDBX_env *env, const int is_exclusive,
const size_t usedsize) {
mdbx_assert(env, usedsize >= pgno2bytes(env, NUM_METAS));
int rc = mdbx_mmap(env->me_flags, &env->me_dxb_mmap, env->me_dbgeo.now,
env->me_dbgeo.upper);
if (unlikely(rc != MDBX_SUCCESS))
return rc;
VALGRIND_MAKE_MEM_NOACCESS(env->me_map + usedsize,
env->me_mapsize - usedsize);
ASAN_POISON_MEMORY_REGION(env->me_map + usedsize, env->me_mapsize - usedsize);
#ifdef MADV_DONTFORK
if (unlikely(madvise(env->me_map, env->me_mapsize, MADV_DONTFORK) != 0))
return errno;
@ -7143,6 +7146,20 @@ static int __cold mdbx_setup_dxb(MDBX_env *env, const int lck_rc) {
if (err != MDBX_SUCCESS)
return err;
mdbx_assert(env, used_bytes >= pgno2bytes(env, NUM_METAS) &&
used_bytes <= env->me_mapsize);
#if defined(MDBX_USE_VALGRIND) || defined(__SANITIZE_ADDRESS__)
VALGRIND_MAKE_MEM_NOACCESS(env->me_map + used_bytes,
env->me_mapsize - used_bytes);
ASAN_POISON_MEMORY_REGION(env->me_map + used_bytes,
env->me_mapsize - used_bytes);
env->me_poison_edge = bytes2pgno(env, env->me_mapsize);
#endif /* MDBX_USE_VALGRIND */
/* NOTE: AddressSanitizer (at least GCC 7.x, 8.x) could generate
* false-positive alarm here. I have no other explanation for this
* except due to an internal ASAN error, as the problem is reproduced
* in a single-threaded application under the active assert() above. */
const unsigned meta_clash_mask = mdbx_meta_eq_mask(env);
if (meta_clash_mask) {
mdbx_error("meta-pages are clashed: mask 0x%d", meta_clash_mask);