upstream/haproxy
1
0
Fork 0
mirror of https://github.com/haproxy/haproxy.git synced 2026-06-10 09:21:35 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 13

BUG/MEDIUM: cache: fix a refcount leak for missed secondary entries

Browse source 
When a primary cache hit has a Vary secondary_key_signature, the code calls
retain_entry() and shctx_row_detach() before performing the secondary lookup.
If get_secondary_entry() returns NULL (no stored variant matches), res is set
to NULL and the function falls through to return ACT_RET_CONT without calling
release_entry() or shctx_row_reattach(). Each such request leaks one refcount
and pins one shctx row permanently, eventually exhausting the cache.

In order to fix this we must do like when no secondary key could be built
and release everything. We only reattach to the row if we previously
detached.

The issue was introduced in 2.4 with commit 1785f3dd9 ("MEDIUM: cache: Add
the Vary header support"). The code changed a bit in 2.9 with commit
48f81ec09 ("MAJOR: cache: Delay cache entry delete in reserve_hot function"),
so in order to backport to 2.8 and older, the patch will have to be manually
applied (no test on detached).
This commit is contained in:
Willy Tarreau 2026-05-23 22:31:58 +02:00
parent c7c71cb5d2
commit 08eb3d7cc3
1 changed files with 11 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
src/cache.c
View file

@ -2180,7 +2180,17 @@ enum act_return http_action_req_cache_use(struct act_rule *rule, struct proxy *p
sec_entry = get_secondary_entry(cache_tree, res,
s->txn.http->cache_secondary_hash,
0);
if (sec_entry && sec_entry != res) {
if (!sec_entry) {
/* Secondary key miss: release the retained primary entry
* and reattach the detached row before returning.
*/
release_entry(cache_tree, res, 0);
shctx_wrlock(shctx);
if (detached)
shctx_row_reattach(shctx, entry_block);
shctx_wrunlock(shctx);
}
else if (sec_entry != res) {
/* The wrong row was added to the hot list. */
release_entry(cache_tree, res, 0);
retain_entry(sec_entry);