Hello!
On Jan 28, 2008, at 3:17 PM, Chuck Lever wrote:
@@ -501,25 +501,29 @@ nlmsvc_testlock(struct svc_rqst *rqstp,
struct nlm_file *file,
block, block->b_flags, block->b_fl);
if (block->b_flags & B_TIMED_OUT) {
nlmsvc_unlink_block(block);
- return nlm_lck_denied;
+ ret = nlm_lck_denied;
+ goto out;
Aren't we also leaking the memory pointed to by block->b_fl that
nlmsvc_testlock just allocated?
When last reference on a block goes away, nlmsvc_free_block is called
that frees block->b_fl, so we are good here, I think.
nlmsvc_unlink_block() already invokes nlmsvc_release_block() in some
cases. So now sometimes we have a double release.
It in fact always calls release_block, but this is not a free
function, this is sort of refput function instead.
We got one reference on a block (from lookup_block/create_block) that
we release after out label, and one that i consumed
by a list membership. When we unlink a block from a list, we drop that
reference too.
Bye,
Oleg
-
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
