21400 - [DOGFOOD][REGRESSION] Cannot file new bug, crashes. Scrolling bug?

Status: VERIFIED FIXED

(Core :: Layout, defect, P3)

Description

[Chris McAfee]

I crash trying to render/scroll the enter_bug.cgi page

  http://bugzilla.mozilla.org/enter_bug.cgi?product=Browser

Comment 1

[Christopher Blizzard (:blizzard)]

I haven't seen this crash yet.  I pulled this morning.

Comment 2

[chris hofmann]

testing out bug posting

Comment 3

[Chris McAfee]

selmer says, and I concur:
scrolling seems to be the common element.  I've had crashes in a number
of different areas, all when scrolling.

Related to saari's focus changes from yesterday?

Comment 4

[scottputterman]

I've had a number of crashes today when clicking on links, buttons, or messages.
I filed it as 21388 with a stack.  Perhaps others are seeing this?

Comment 5

[leger]

Putting on the PDT+ radar.

Comment 6

[David :Bienvenu]

this is the crash I've seen several times while scrolling web pages:

mparent has been deleted.

nsGenericElement::HandleDOMEvent(nsIPresContext * 0x031cff10, nsEvent *
0x0012f740, nsIDOMEvent * * 0x0012f35c, unsigned int 0x00000001, nsEventStatus *
0x0012f760) line 774 + 33 bytes
nsHTMLAnchorElement::HandleDOMEvent(nsHTMLAnchorElement * const 0x0398231c,
nsIPresContext * 0x031cff10, nsEvent * 0x0012f740, nsIDOMEvent * * 0x00000000,
unsigned int 0x00000001, nsEventStatus * 0x0012f760) line 341 + 31 bytes
nsEventStateManager::SendFocusBlur(nsEventStateManager * const 0x031b4160,
nsIPresContext * 0x031cff10, nsIContent * 0x00000000) line 1936
nsEventStateManager::ChangeFocus(nsIContent * 0x00000000, nsIFrame * 0x01f5adf4,
int 0x00000001) line 1433
nsEventStateManager::PostHandleEvent(nsEventStateManager * const 0x031b4160,
nsIPresContext * 0x031cff10, nsGUIEvent * 0x0012fb68, nsIFrame * 0x01f5adf4,
nsEventStatus * 0x0012fa74, nsIView * 0x031b6d10) line 552
PresShell::HandleEvent(PresShell * const 0x0319b834, nsIView * 0x031b6d10,
nsGUIEvent * 0x0012fb68, nsEventStatus * 0x0012fa74) line 2577 + 43 bytes
nsView::HandleEvent(nsView * const 0x031b6d10, nsGUIEvent * 0x0012fb68, unsigned
int 0x00000008, nsEventStatus * 0x0012fa74, int & 0x00000000) line 841
nsView::HandleEvent(nsView * const 0x031929e0, nsGUIEvent * 0x0012fb68, unsigned
int 0x0000001c, nsEventStatus * 0x0012fa74, int & 0x00000000) line 826
nsViewManager::DispatchEvent(nsViewManager * const 0x03192bb0, nsGUIEvent *
0x0012fb68, nsEventStatus * 0x0012fa74) line 1678
HandleEvent(nsGUIEvent * 0x0012fb68) line 69
nsWindow::DispatchEvent(nsWindow * const 0x031928b4, nsGUIEvent * 0x0012fb68,
nsEventStatus & nsEventStatus_eIgnore) line 421 + 10 bytes
nsWindow::DispatchWindowEvent(nsGUIEvent * 0x0012fb68) line 442
nsWindow::DispatchMouseEvent(unsigned int 0x0000012e, nsPoint * 0x00000000) line
3332 + 21 bytes
ChildWindow::DispatchMouseEvent(unsigned int 0x0000012e, nsPoint * 0x00000000)
line 3550
nsWindow::ProcessMessage(unsigned int 0x00000201, unsigned int 0x00000001, long
0x02cd0360, long * 0x0012fdc8) line 2627 + 24 bytes
nsWindow::WindowProc(HWND__ * 0x01370aee, unsigned int 0x00000201, unsigned int
0x00000001, long 0x02cd0360) line 608 + 27 bytes
USER32! 77e71268()

Comment 7

[troy]

Re-assigning to Tom because it looks to be event related. Looks like another
stack trace I saw, so I suspect it is a DUP

Comment 8

[chris hofmann]

scrolling and filing bugs seems to work ok for me
in the win32 release builds from 12/10 on win95.

Comment 9

[selmer (gone)]

Stack trace from one of the scrolling crashes in edit bookmarks.  Note that the
variable "parent" appears to have been had its nsISupports released (__vfptr =
0xdddddddd)

nsViewManager::DispatchEvent(nsViewManager * const 0x0218d890, nsGUIEvent *
0x0012fb68, nsEventStatus * 0x0012fa74) line 1647 + 13 bytes
HandleEvent(nsGUIEvent * 0x0012fb68) line 69
nsWindow::DispatchEvent(nsWindow * const 0x0218d544, nsGUIEvent * 0x0012fb68,
nsEventStatus & nsEventStatus_eIgnore) line 421 + 10 bytes
nsWindow::DispatchWindowEvent(nsGUIEvent * 0x0012fb68) line 442
nsWindow::DispatchMouseEvent(unsigned int 300, nsPoint * 0x00000000) line 3332 +
21 bytes
ChildWindow::DispatchMouseEvent(unsigned int 300, nsPoint * 0x00000000) line
3550
nsWindow::ProcessMessage(unsigned int 512, unsigned int 1, long 28312104, long *
0x0012fdc8) line 2620 + 24 bytes
nsWindow::WindowProc(HWND__ * 0x01a30376, unsigned int 512, unsigned int 1, long
28312104) line 608 + 27 bytes
USER32! 77e71820()
01b00228()

Comment 10

[Chris McAfee]

This is a regression from yesterday, have we looked at what went in recently to
make this happen?

Comment 11

[joki (gone)]

Well this bug has been fun in that a number of people, including myself, can't
seem to reproduce this at all.  However, hyatt found a fixed a crash problem
caused by null parent issues in focus changes.  I believe (or at least hope)
that was causing this as well.  I'm going to mark this fixed and hope for the
best.

Comment 12

[Chris Petersen]

With the Dec 13 th Linux build (1999121314), I can't reproduce this problem.

Comment 13

[Chris McAfee]

agreed, things are much better now.

Comment 14

[Chris Petersen]

Base on the reporter's last comments, I'm marking this verified fixed in the Dec
13th Linux build.