Closed
Bug 1263763
Opened 9 years ago
Closed 9 years ago
Browser hang in IPC::Channel::ChannelImpl::ProcessIncomingMessages
Categories
(Core :: IPC, defect)
Core
IPC
Tracking
()
RESOLVED
DUPLICATE
of bug 1264398
People
(Reporter: mccr8, Unassigned)
References
Details
(Keywords: hang, Whiteboard: btpp-followup-2016-04-20)
Attachments
(1 file)
3.81 KB,
text/plain
|
Details |
The_8472 in IRC reported that after using the browser for an hour or so, it would become completely unresponsive. Using WinDBG, they were able to get a stack which looks intriguingly related to bug 1235633. Off-hand, I'd guess this is a dupe of bug 1263457, but I figured I'd file this separately in case it is not.
ChannelImpl::ProcessIncomingMessages is calling realloc, and it looks like the new size is 0x81e958, which Google tells me is 8513880.
Reporter | ||
Updated•9 years ago
|
tracking-e10s:
--- → ?
Keywords: hang
Some more details:
* win64 nightly
* main thread was more or less idle during the hang
* the attached thread was the busy one. it spent most of its time in sys and not in user
Dragging large images also consumes huge amounts of system CPU time.
E.g. opening https://upload.wikimedia.org/wikipedia/commons/d/db/Greatwall_large.jpg in a new tab and then starting to drag the image causes a background thread to consume 100% CPU and the drag preview to show up with a large delay.
Reporter | ||
Comment 5•9 years ago
|
||
That makes some sense. Drag and drop can involve some very large messages, as seen here in Msg_InvokeDragSession: http://mzl.la/25ZRaV5
This is almost certainly a dupe.
Status: NEW → RESOLVED
Closed: 9 years ago
Flags: needinfo?(wmccloskey)
Resolution: --- → DUPLICATE
You need to log in
before you can comment on or make changes to this bug.
Description
•