On Tue, 7 Apr 2015, Ján Tomko wrote:
On Mon, Mar 30, 2015 at 01:41:01PM +1100, Michael Chapman wrote:
The close callbacks hash are keyed by a UUID-string, but
virCloseCallbacksRun was attempting to remove them by raw UUID. This
patch ensures the callback entries are removed by UUID-string as well.
This bug could cause problems when guest migrations were abnormally aborted:
# timeout --signal KILL 1 \
virsh migrate example qemu+tls://remote/system \
--verbose --compressed --live --auto-converge \
--abort-on-error --unsafe --persistent \
--undefinesource --copy-storage-all --xml example.xml
Killed
# virsh migrate example qemu+tls://remote/system \
--verbose --compressed --live --auto-converge \
--abort-on-error --unsafe --persistent \
--undefinesource --copy-storage-all --xml example.xml
error: Requested operation is not valid: domain 'example' is not being migrated
Signed-off-by: Michael Chapman <mike@xxxxxxxxxxxxxxxxx>
---
src/util/virclosecallbacks.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/util/virclosecallbacks.c b/src/util/virclosecallbacks.c
index 4128057..b0a1a9e 100644
--- a/src/util/virclosecallbacks.c
+++ b/src/util/virclosecallbacks.c
@@ -236,6 +236,7 @@ typedef struct _virCloseCallbacksListEntry virCloseCallbacksListEntry;
typedef virCloseCallbacksListEntry *virCloseCallbacksListEntryPtr;
struct _virCloseCallbacksListEntry {
unsigned char uuid[VIR_UUID_BUFLEN];
+ char uuidstr[VIR_UUID_STRING_BUFLEN];
We shouldn't be storing the uuid twice.
OK.
virCloseCallback callback;
};
@@ -279,6 +280,8 @@ virCloseCallbacksGetOne(void *payload,
memcpy(data->list->entries[data->list->nentries - 1].uuid,
uuid, VIR_UUID_BUFLEN);
+ memcpy(data->list->entries[data->list->nentries - 1].uuidstr,
+ uuidstr, VIR_UUID_STRING_BUFLEN);
data->list->entries[data->list->nentries - 1].callback = closeDef->cb;
}
@@ -332,7 +335,7 @@ virCloseCallbacksRun(virCloseCallbacksPtr closeCallbacks,
for (i = 0; i < list->nentries; i++) {
virHashRemoveEntry(closeCallbacks->list,
- list->entries[i].uuid);
+ list->entries[i].uuidstr);
I think the CPU time wasted by doing virUUIDFormat here again is
negligible and easier than writing a Hash function that works on
VIR_UUID buffers :)
Jan
Yeah, sounds reasonable. I'll send through a v2 patch shortly. I've also
got a few extra patches lined up, but I'll start a new series for those.
Thanks for the review!
- Michael
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list