add initial guard slabs implementation

This commit is contained in:
Daniel Micay 2018-09-06 18:53:06 -04:00
parent cc1e79fdba
commit 1be74ec40d
2 changed files with 9 additions and 4 deletions

View File

@ -83,8 +83,8 @@ allocation and then unmapped on free.
* High entropy per-slab random values
* [in-progress] Mangled into a unique value per slab slot (although not
with a strong keyed hash due to performance limitations)
* [in-progress] Some slab locations are skipped and remain memory protected,
leaving slab size class regions interspersed with guard pages
* Possible slab locations are skipped and remain memory protected, leaving slab
size class regions interspersed with guard pages
* Zero size allocations are memory protected
* Protected allocator metadata
* Address space for metadata is never used for allocations and vice versa

View File

@ -19,6 +19,8 @@
static_assert(sizeof(void *) == 8, "64-bit only");
static const bool guard_slabs = true;
// either sizeof(uint64_t) or 0
static const size_t canary_size = sizeof(uint64_t);
@ -140,9 +142,9 @@ static size_t get_metadata_max(size_t slab_size) {
}
static struct slab_metadata *alloc_metadata(struct size_class *c, size_t slab_size, bool non_zero_size) {
if (unlikely(c->metadata_count == c->metadata_allocated)) {
if (unlikely(c->metadata_count >= c->metadata_allocated)) {
size_t metadata_max = get_metadata_max(slab_size);
if (c->metadata_count == metadata_max) {
if (c->metadata_count >= metadata_max) {
errno = ENOMEM;
return NULL;
}
@ -162,6 +164,9 @@ static struct slab_metadata *alloc_metadata(struct size_class *c, size_t slab_si
return NULL;
}
c->metadata_count++;
if (guard_slabs) {
c->metadata_count++;
}
return metadata;
}