On 12/19/2013 02:18 AM, Peter Krempa wrote:
On 12/19/13 05:38, Eric Blake wrote:
> Commit 6cd60b6 was flat out broken - it tried to print into the
> wrong variable. My testing was obviously too cursory (did the
> name get a slash added?); valgrind would have caught the error.
> Thankfully it didn't hit any release.
>
... you still need to free state->uri->path before you overwrite it with
the previous value that is temporarily stored in 'tmp' a few lines below:
tmp = state->uri->path;
if (virAsprintf(&vol->key, "%s%s", state->uri->path, name)
< 0) {
state->uri->path = tmp;
goto cleanup;
}
if (!(vol->target.path = virURIFormat(state->uri))) {
state->uri->path = tmp;
goto cleanup;
}
state->uri->path = tmp; <--- here
ACK with the memleak resolved.
A demonstration of how NOT to write patches late at night. Thanks for
catching that; pushing with this squashed in.
diff --git i/src/storage/storage_backend_gluster.c
w/src/storage/storage_backend_gluster.c
index 9aa692e..2ec2424 100644
--- i/src/storage/storage_backend_gluster.c
+++ w/src/storage/storage_backend_gluster.c
@@ -232,9 +232,11 @@
virStorageBackendGlusterRefreshVol(virStorageBackendGlusterStatePtr state,
goto cleanup;
}
if (!(vol->target.path = virURIFormat(state->uri))) {
+ VIR_FREE(state->uri->path);
state->uri->path = tmp;
goto cleanup;
}
+ VIR_FREE(state->uri->path);
state->uri->path = tmp;
if (S_ISDIR(st->st_mode)) {
--
Eric Blake eblake redhat com +1-919-301-3266
Libvirt virtualization library
http://libvirt.org