Closed Bug 306809 Opened 20 years ago Closed 20 years ago

oom mlk in txVariable::Convert

Categories

(Core :: XSLT, defect, P1)

x86
Windows XP
defect

Tracking

()

RESOLVED FIXED
mozilla1.8beta4

People

(Reporter: timeless, Assigned: peterv)

References

()

Details

(Keywords: fixed1.8, memory-leak)

Attachments

(1 file)

first the easy mlk: 1108 txXPathNode *xpathNode = 1109 txXPathNativeNode::createXPathNode(node); 1110 if (!xpathNode) { 1111 return NS_ERROR_FAILURE; 1112 peterv 1.39 } 1113 1114 nodeSet->add(*xpathNode); add returns nsresult (NS_ERROR_OUT_OF_MEMORY). you leak xpathNode; this also happens near 1202. i think there's a more fun crash case that could happen involving spidermonkey, but i'm not sure about the rules for js dependent strings (if i'm right, you could try to construct a dependentstring from a null pointer or a fmr length).
Attached patch v1Splinter Review
Attachment #194673 - Flags: superreview?(bzbarsky)
Attachment #194673 - Flags: review?(bzbarsky)
Comment on attachment 194673 [details] [diff] [review] v1 I'm assuming add() just copies the data in the node and doesn't hold a ptr to it, right?
Attachment #194673 - Flags: superreview?(bzbarsky)
Attachment #194673 - Flags: superreview+
Attachment #194673 - Flags: review?(bzbarsky)
Attachment #194673 - Flags: review+
Yes. We're going to clean up the whole txXPathNativeNode mess at some point so that we won't run into these leaks anymore.
Status: NEW → ASSIGNED
Priority: -- → P1
Target Milestone: --- → mozilla1.8beta4
Status: ASSIGNED → RESOLVED
Closed: 20 years ago
Resolution: --- → FIXED
Comment on attachment 194673 [details] [diff] [review] v1 Trivial leak fix.
Attachment #194673 - Flags: approval1.8b4?
Attachment #194673 - Flags: approval1.8b4? → approval1.8b4+
Time is short for 1.8b4. If this isn't landed today, it's not going to make the train.
peterv landed this on branch on the 4th.
Keywords: fixed1.8
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: