| Issue 31716: | Crash under RenderWidgetHostViewMac::ShowPopupWithItems(). | |
| 8 people starred this issue and may be notified of changes. | Back to list |
Sign in to add a comment
|
Product: Chrome_Mac Stack Signature: objc_msgSend-D73B89 New Signature Label: objc_msgSend New Signature Hash: d88c4a37_496b08fd_ba1b9384_7d4423f4_edd2720d Report link: http://go/crash/reportdetail?reportid=05b5e8e1c8a58b1b Meta information: Product Name: Chrome_Mac Product Version: 4.0.288.1 Report ID: 05b5e8e1c8a58b1b Report Time: 2010/01/06 23:17:44, Wed Uptime: 4 sec Cumulative Uptime: 0 sec OS Name: Mac OS X OS Version: 10.6.2 10C540 CPU Architecture: x86 CPU Info: GenuineIntel family 6 model 23 stepping 6 0x98bd9edb [libobjc.A.dylib + 0x00005edb] objc_msgSend 0x003704e2 [Google Chrome Framework - render_widget_host_view_mac.mm:425] RenderWidgetHostViewMac::ShowPopupWithItems(gfx::Rect, int, int, std::vector<WebMenuItem, std::allocator<WebMenuItem> > const&) 0x00369461 [Google Chrome Framework - render_widget_host.cc:802] RenderWidgetHost::OnMsgShowPopup(ViewHostMsg_ShowPopup_Params const&) 0x0036c1b9 [Google Chrome Framework - ../base/tuple.h:422] RenderWidgetHost::OnMessageReceived(IPC::Message const&) 0x003441b2 [Google Chrome Framework - browser_render_process_host.cc:777] BrowserRenderProcessHost::OnMessageReceived(IPC::Message const&) 0x00a1668a [Google Chrome Framework - ../base/tuple.h:422] RunnableMethod<IPC::ChannelProxy::Context, void (IPC::ChannelProxy::Context::*)(IPC::Message const&), Tuple1<IPC::Message> >::Run() 0x0053fd5a [Google Chrome Framework - message_loop.cc:320] MessageLoop::DeferOrRunPendingTask(MessageLoop::PendingTask const&) 0x005406aa [Google Chrome Framework - message_loop.cc:435] MessageLoop::DoWork() 0x0051ccd3 [Google Chrome Framework - message_pump_mac.mm:291] base::MessagePumpCFRunLoopBase::RunWorkSource(void*) 0x92d608ca [CoreFoundation + 0x0003e8ca] __CFRunLoopDoSources0 0x92d5e38e [CoreFoundation + 0x0003c38e] __CFRunLoopRun 0x92d5d863 [CoreFoundation + 0x0003b863] CFRunLoopRunSpecific 0x92d5d690 [CoreFoundation + 0x0003b690] CFRunLoopRunInMode 0x98084f0b [HIToolbox + 0x00034f0b] RunCurrentEventLoopInMode 0x98084cc2 [HIToolbox + 0x00034cc2] ReceiveNextEventCommon 0x98084b47 [HIToolbox + 0x00034b47] BlockUntilNextEventMatchingListInMode 0x951b3ac4 [AppKit + 0x00048ac4] _DPSNextEvent 0x951b3305 [AppKit + 0x00048305] -[NSApplication nextEventMatchingMask:untilDate:inMode:dequeue:] 0x9517549e [AppKit + 0x0000a49e] -[NSApplication run] 0x0051c77c [Google Chrome Framework - message_pump_mac.mm:677] base::MessagePumpNSApplication::DoRun(base::MessagePump::Delegate*) 0x0051bf05 [Google Chrome Framework - message_pump_mac.mm:213] base::MessagePumpCFRunLoopBase::Run(base::MessagePump::Delegate*) 0x0053fca3 [Google Chrome Framework - message_loop.cc:205] MessageLoop::Run() 0x000e6e42 [Google Chrome Framework - browser_main.cc:163] BrowserMain(MainFunctionParams const&) 0x0000ae03 [Google Chrome Framework - chrome_dll_main.cc:744] ChromeMain 0x00001fc5 [Google Chrome + 0x00000fc5] 0x00000001
Comment
1
by
shess@chromium.org,
Jan 11, 2010
Labels: OS-Mac
,
Jan 11, 2010
See also Issue 31856
,
Jan 13, 2010
Is a heavy crasher.
Status: Untriaged
,
Jan 13, 2010
Perhaps the same: http://crash/reportdetail?reportid=78eee210794a421e Additional stack level: 0x0036affb [Google Chrome Framework - render_widget_host.cc:380] RenderWidgetHost::ForwardKeyboardEvent(NativeWebKeyboardEvent const&) 0x00370543 [Google Chrome Framework - render_widget_host_view_mac.mm:430] RenderWidgetHostViewMac::ShowPopupWithItems(gfx::Rect, int, int, std::vector<WebMenuItem, std::allocator<WebMenuItem> > const&) 0x00369461 [Google Chrome Framework - render_widget_host.cc:802] RenderWidgetHost::OnMsgShowPopup(ViewHostMsg_ShowPopup_Params const&) 0x0036c1b9 [Google Chrome Framework - ../base/tuple.h:422] RenderWidgetHost::OnMessageReceived(IPC::Message const&)
,
Jan 14, 2010
There is a collection of crashes which all seem to relate to RenderWidgetHostViewMac::ShowPopupWithItems(). As a group, these are our biggest browser crasher. Some of the crashes have signatures which don't work with the crash-server dashboard, and I think this is why they don't show up in the Mac-crash summary dashboard. I've been unsuccessfully trying to find some time to look at it, but haven't managed it.
Cc: mikesm...@chromium.org pinker...@chromium.org j...@chromium.org
,
Jan 14, 2010
(No comment was entered for this change.)
Labels: Crash-TopCrashers
,
Jan 14, 2010
(No comment was entered for this change.)
Labels: -Crash-TopCrashers Crash-TopCrasher
,
Jan 14, 2010
Scott, do you have any repro info, or something you'd like test to bang on?
,
Jan 14, 2010
Nope. Haven't had time to look into it much, either. It isn't happening on 249.x channel, so I'm wondering if it's something new that's been added (actually, could extensions poke into this for their UI?).
,
Jan 15, 2010
extension related, perhaps?
Status: Available
Cc: andyb...@chromium.org m...@chromium.org Labels: Mstone-5 ReleaseBlock-Beta
,
Jan 15, 2010
The only thing I can think of is a SELECT crasher I fixed a while ago: http://src.chromium.org/viewvc/chrome?view=rev&revision=34136 But that fix has been in for a while.
,
Jan 19, 2010
This continues to happen in 295.0. It's likely our largest browser crasher (the crash server is sketchy, the template arguments appear to conflict with doing searches in some cases). I'm OOT this week and definitely will not get around to looking at it.
,
Jan 19, 2010
I'll take a look.
,
Jan 19, 2010
Promoting to P1
Labels: -Pri-2 Pri-1
,
Jan 20, 2010
We should fix this for *this* beta, so it's P0.
Labels: -Pri-1 Pri-0
,
Jan 20, 2010
I haven't been able to repro, but I suspect it's extension-related. Timing-related, anyway, with a page navigating or closing just as its <select> control is being closed. Checking for invalid render_widget_host_ before using it near the affected lines (425 and 430) will probably prevent the crash. Wish I could repro, to be certain there wasn't something more subtle going on. Anyway, when I get back tomorrow I'll put that much in. More eyes also welcome.
,
Jan 20, 2010
(No comment was entered for this change.)
Labels: ForMerge
,
Jan 21, 2010
No longer so convinced that the problem is an invalid render_widget_host_ here. Might be chrome_view_, although AFAICT that's not used in the second crasher shess posted above. I'm reluctant to just paper over what *might* be the superficial problem without understanding the underlying cause, so in go some diagnostics in the hopes that they'll shed more light on this.
,
Jan 22, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=36790
------------------------------------------------------------------------
r36790 | pam@chromium.org | 2010-01-21 13:55:40 -0800 (Thu, 21 Jan 2010) | 6 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.h?r1=36790&r2=36789
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=36790&r2=36789
Insert diagnostics in an attempt to track down a crash in popup menus
in RenderWidgetHostViewMac::ShowPopupWithItems().
BUG=31716
TEST=should have no user-visible effect; see crash reports
Review URL: http://codereview.chromium.org/545159
------------------------------------------------------------------------
,
Jan 22, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=36894
------------------------------------------------------------------------
r36894 | pam@chromium.org | 2010-01-22 12:59:29 -0800 (Fri, 22 Jan 2010) | 9 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/branches/302/src/chrome/browser/renderer_host/render_widget_host_view_mac.h?r1=36894&r2=36893
M http://src.chromium.org/viewvc/chrome/branches/302/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=36894&r2=36893
Merge 36790 - Insert diagnostics in an attempt to track down a crash in popup menus
in RenderWidgetHostViewMac::ShowPopupWithItems().
BUG=31716
TEST=should have no uservisible effect; see crash reports
Review URL: http://codereview.chromium.org/545159
TBR=pam@chromium.org
Review URL: http://codereview.chromium.org/548119
------------------------------------------------------------------------
,
Jan 25, 2010
Tracking note: these diagnostics will be in the next dev release at or after 4.0.302.4.
,
Jan 25, 2010
Presumed fixed and merged. Reopen if that's not the case.
Status: Fixed
,
Jan 25, 2010
Nope, not fixed. The patch only added diagnostics, since we're still getting crashes reported automatically but haven't been able to reproduce them.
Status: Started
,
Jan 25, 2010
Unassigning, at least for now. I'll be out starting the day after tomorrow, and we likely won't have crash data from 302.4+ before then.
Status: Available
Owner: --- Cc: p...@chromium.org
,
Jan 25, 2010
(No comment was entered for this change.)
Cc: viettrun...@chromium.org
,
Jan 26, 2010
Trung, can you look into this?
Status: Assigned
Owner: viettrun...@chromium.org Cc: sh...@chromium.org hbri...@google.com
,
Jan 26, 2010
Wait, we were waiting for the diagnostics that pam put in to make it out to the dev channel and into the crash logs (per her comments above). Is there anything we can do until then?
,
Jan 26, 2010
Sure, we just wanted to make sure that there's an owner for looking at the info when it comes back since Pam will be gone.
Labels: -Area-Undefined Area-Internals
,
Jan 27, 2010
This crash, or a similar one, can be caused by an extension killing the tab while the menu is showing. I'll investigate whether it can also happen with JS killing the window.
,
Jan 27, 2010
A JavaScript window.close() while the menu is showing will also cause a crash.
,
Jan 27, 2010
shess "volunteered" to take care of this, since he took care of a bunch of similar cases before.
Owner: sh...@chromium.org
,
Jan 27, 2010
(No comment was entered for this change.)
Status: Started
,
Jan 27, 2010
Lovely. I broke this with Issue 30147 . To repro, pull up attached test.html, and bring up the select menu. When the tab goes away, anything you do should crash (if it doesn't, select something).
,
Jan 27, 2010
shess mentioned that pink suggested that things would also break (i.e., crash) if a navigation to another page occurred while the menu was up. I verified that this is in fact the case.
,
Jan 27, 2010
Issue 33250 , which I'm about to mark as a dup of this, has some very simple steps to reproduce (which I have not tried myself): "1. Go to amazon.com 2. Submit the search form at the top. 3. Before the next page loads, click on the drop-down to the left to open it. 4. Once the next page loads, click on the page to close the drop-down."
,
Jan 27, 2010
Issue 33250 has been merged into this issue.
Cc: kr...@chromium.org rohi...@chromium.org
,
Jan 27, 2010
Hoo boy, those steps above certainly work like a charm (he said, waving away the clouds of choking black smoke...)
,
Jan 28, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=37417
------------------------------------------------------------------------
r37417 | shess@chromium.org | 2010-01-28 10:53:08 -0800 (Thu, 28 Jan 2010) | 6 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=37417&r2=37416
[Mac] Prevent script-initiated tab close while a content select menu is up.
BUG=31716
TEST=test.html from the bug, page shouldn't close while select popup is up.
Review URL: http://codereview.chromium.org/555144
------------------------------------------------------------------------
,
Jan 28, 2010
(No comment was entered for this change.)
Status: Fixed
,
Jan 29, 2010
Sigh. I thought I had merged this over to branch 307, but apparently something went wrong. I think Jens' fix will prevent this crash from happening, so I don't think it's at all worth re-spinning the build. But I'm going to merge it over to 307, so that if we end up re-spinning for other reasons it comes along.
,
Jan 31, 2010
Still seeing this in 5.0.307.1 : http://crash/reportdetail?reportid=10c9cc09d035e23e
Status: Available
,
Feb 1, 2010
The fix is merged over to 307, but isn't present in 307.1, so moving this back to "fixed", assuming that the next step is "verified". Would it be reasonable to have a tracking option on the order of "upstream"?
Status: Fixed
,
Feb 1, 2010
Since the bot isn't updating the bug, fix is merged at: http://src.chromium.org/viewvc/chrome?view=rev&revision=37549
,
Feb 1, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=37438
------------------------------------------------------------------------
r37438 | snej@chromium.org | 2010-01-28 13:24:12 -0800 (Thu, 28 Jan 2010) | 6 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=37438&r2=37437
Fix Mac crash when page goes away while a pop-up menu is active.
This may also fix the older related bug 31716 .
BUG=33250
TEST=see steps to reproduce in the bug report
Review URL: http://codereview.chromium.org/558021
------------------------------------------------------------------------
,
Feb 1, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=37440
------------------------------------------------------------------------
r37440 | snej@chromium.org | 2010-01-28 13:30:32 -0800 (Thu, 28 Jan 2010) | 9 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/branches/307/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=37440&r2=37439
Merge 37438 - Fix Mac crash when page goes away while a popup menu is active.
This may also fix some cases of the older related bug 31716 .
BUG=33250
TEST=see steps to reproduce in the bug report
Review URL: http://codereview.chromium.org/558021
TBR=snej@chromium.org
Review URL: http://codereview.chromium.org/548191
------------------------------------------------------------------------
,
Feb 1, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=37549
------------------------------------------------------------------------
r37549 | shess@chromium.org | 2010-01-29 14:05:34 -0800 (Fri, 29 Jan 2010) | 9 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/branches/307/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=37549&r2=37548
Merge 37417 - [Mac] Prevent scriptinitiated tab close while a content select menu is up.
BUG=31716
TEST=test.html from the bug, page shouldn't close while select popup is up.
Review URL: http://codereview.chromium.org/555144
TBR=shess@chromium.org
Review URL: http://codereview.chromium.org/549212
------------------------------------------------------------------------
,
Feb 2, 2010
Issue 33717 has been merged into this issue.
,
Feb 2, 2010
The following revision refers to this bug:
http://src.chromium.org/viewvc/chrome?view=rev&revision=37866
------------------------------------------------------------------------
r37866 | shess@chromium.org | 2010-02-02 12:11:04 -0800 (Tue, 02 Feb 2010) | 11 lines
Changed paths:
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.h?r1=37866&r2=37865
M http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/renderer_host/render_widget_host_view_mac.mm?r1=37866&r2=37865
[Mac] Revert diagnostics for content popup bug.
Revert pamg's commit.
> Insert diagnostics in an attempt to track down a crash in popup menus
> in RenderWidgetHostViewMac::ShowPopupWithItems().
BUG=31716
TEST=none
Review URL: http://codereview.chromium.org/566018
------------------------------------------------------------------------
,
Feb 3, 2010
Verified in 5.0.307.5 (Official Build 37950) dev, Steps followed: 1. Go to amazon.com 2. Submit the search form at the top. 3. Before the next page loads, click on the drop-down to the left to open it. 4. Once the next page loads, click on the page to close the drop-down.
Status: Verified
,
Feb 9, 2010
(No comment was entered for this change.)
Labels: -Crash-TopCrasher Crash-TopFixed
,
Feb 14, 2010
Merged, verified: removing formerge label.
Labels: -ForMerge
|
||||||||||
| ► Sign in to add a comment | |||||||||||