Closed Bug 5164 Opened 25 years ago Closed 25 years ago

Apprunner Crashes when closed

Categories

(Core Graveyard :: Tracking, defect, P2)

Product:
Core Graveyard
Component:
Tracking
Platform:
x86
Windows 95
Type:
defect

Tracking

(Not tracked)

Status:
VERIFIED WORKSFORME

People

(Reporter: basic, Assigned: law)

Details

(Whiteboard: we're having trouble reproducing internally) 

Build: M4 with Fullcircle
Causes keyboard to stop responding.

run apprunner
close apprunner
watch apprunner crash
keyboard stops responding
APPRUNNER caused an invalid page fault in
module RPCRT4.DLL at 015f:7fbd1052.
Registers:
EAX=00000000 CS=015f EIP=7fbd1052 EFLGS=00010246
EBX=816176c4 SS=0167 ESP=015dfd8c EBP=015dfdc0
ECX=c65b35c0 DS=0167 ESI=7fbd0000 FS=2367
EDX=c00300f4 ES=0167 EDI=00000000 GS=0000
Bytes at CS:EIP:
ff 70 28 ff 15 78 a0 c1 7f c7 05 ec 91 c1 7f 01
Stack dump:
00000000 00000000 7fbd0000 816176c4 7fd684cc 7fd03f2c bff741fb 015dfd90 015dfbbc
015dff78 7fbd90ce 7fc124d0 ffffffff 015dff88 bff7ddcd 7fbd0000
I have to terminate the apprunner dos box instead of closing the mozilla window,
otherwise windows the apprunner dos box will freeze and windows will freeze when
it shut down.
Status: NEW → RESOLVED
Closed: 25 years ago
Resolution: --- → WORKSFORME
Resolved as WORKSFORME.

I just can't reporduce this problem.  Do you attmept to close Apprunner as soon
as the window appears?
Status: RESOLVED → REOPENED
OK, I made a mistake, I meant when I close apprunner, it closes but windows
give me a message saying that it crashed with rpcrt4.dll and the apprunner's
dos box is left open. If I try to close that, the dos box freezes and my
keyboard does not work except for the numlock key.
This also happen in the 4/22 nightly build
Assignee: don → radha
Status: REOPENED → NEW
Priority: P3 → P2
Target Milestone: M5
Re-assigned to radha; set target mileston to M5; changed priority to P2.

OK, I still can't reporduce this, but let's see if Radha can ...
OS: Windows 98 → Windows 95
Well, I had somewhat of a problem on shutdown.  If I close the browser
window before the DOS window, the DOS window stays.  If I click the X
on the DOS window, and tell Win95 to terminate the DOS window, the
cursor then freezes inside the DOS window.  It even prevents the
CTRL-ALT-DEL window from popping up.  It requires a restart (hitting the
reset button) to get rid of it, although it doesn't seem to affect any
other apps.  If I close the DOS window before the browser window, then I
don't get this problem...this is all on M4
M4 Win32 that is
M4 build of Viewer also crashes in the same way when I try to close it.

Neoplanet's preview is also crashes when close but the keyboard still works
( but the ctrl-alt-del combination does not)
Status: NEW → RESOLVED
Closed: 25 years ago → 25 years ago
QA Contact: 3853 → 4130
claudis,can you verify that this is ok with most recent Apprunner Win32 build
please?  Otherwise, re-open.
Status: RESOLVED → VERIFIED
I cannot reproduce this at all on any Win32 systems with the 1999042608 builds. Marking verified.
Build dated 04/26/1999
Still crashes on my machine.
ran M3 to make sure it wasn't windows problem. M3 did not crash.
Could it be because I'm on a low RAM system? (32MB Win98)
build 1999042908

When closing apprunner the dos box shows:

Creating browser app core
BrowserAppCore has been created.
Setting content window
Init!
Reading file...
Reading file...Done
adding "resource:/res/rdf/remote-flash-1.rdf" to the tree
Reading file...
Reading file...Done
adding "resource:/res/rdf/remote-flash-2.rdf" to the tree
Reading file...
Reading file...Done
adding "resource:/res/rdf/remote-flash-3.rdf" to the tree
Reading file...
Reading file...Done

runtime error R6016
- not enough space for thread data

runtime error R6016
- not enough space for thread data

runtime error R6016
- not enough space for thread data

3 error messages pops up successively saying:

APPRUNNER caused an invalid page fault in
module KERNEL32.DLL at 015f:bff7b997.
Registers:
EAX=00000000 CS=015f EIP=bff7b997 EFLGS=00000246
EBX=00000008 SS=0167 ESP=0142ff28 EBP=0142ff38
ECX=7802df7a DS=0167 ESI=78034120 FS=121f
EDX=bffc9490 ES=0167 EDI=815fe8ec GS=0000
Bytes at CS:EIP:
ff 76 04 e8 26 89 ff ff 5e c2 04 00 56 8b 74 24
Stack dump:
78034048 78001020 78034120 00cc4300 0142ff98 78002b70 0000000d 78003a07 815fe8ec
7800ba71 000000ff 00000001 00000000 7800bcfe 000000ff 0000002b

I suspect this is a different bug that could be blocking the original bug as
this one doesn't disable my keyboard.
I am downloading the build before this one to test.
Severity: normal → major
Status: VERIFIED → REOPENED
Whiteboard: we're having trouble reproducing internally
Reopening as this bug deserves to get looked at since two separate folks have repro'd it and is potentially a crasher or at least
very debilitating.
Status: REOPENED → RESOLVED
Closed: 25 years ago → 25 years ago
I've never seen this happen.  A lot of memory leaks have been fixed today. so
can we test this again Monday's build?
hey Basic, PyroGuy could you guys take a stab at reproducing this with the 1999-05-03-17 builds? Try to generate a talkback
report maybe? Otherwise this one is gonna get verified as WFM and fall by the wayside. Thanx.
I'll try but problem with this bug is that I have tried to generate a talkback
in M4 but it could not. I suspect it is because apprunner has unload(?) before
the crash. I'll try anyway.
hmm... I can't seem to produce this bug in M5 but I suspect it could be related
to bug #5896
Just searched through news.mozilla.org and found this message entitled leak
status report:
http://www.dejanews.com/[ST_rn=ps]/getdoc.xp?AN=474369212

Is this related?
Status: RESOLVED → REOPENED
Target Milestone: M5
Who set the target milestone of the bug to M5?  Please don't do that.
Status: REOPENED → RESOLVED
Closed: 25 years ago → 25 years ago
I'm resolving this again as WORKSFORME until we can get a reproducible case
here.  If it's related to bug #5896, then Bill should look at it for M6.
cc'ing law(assuming that's who Don was referring to) so he can take a looksee and maybe dupe it? Or at the very least shine a
little light on it?
I get the same problem that basic@yahoo.com describes above.  This bug happens
even if you delete the mozregistry.dat file before loading apprunner.  Sometimes
after rebooting my computer when apprunner crashes scandisk reports that the
size of mozregistry.dat has been misreported.  Although this same bug has caused
problems for my computer in several builds, have most recently experienced this
bug on the May 25 build.
=============================
tallison@mediaone.net said:
I started apprunner.exe, loaded the browser window and then quit.
I used File - Quit to shut down the browser and received the following:
APPRUNNER caused an invalid page fault in
module RPCRT4.DLL at 014f:70101052.
Registers:
EAX=00000000 CS=014f EIP=70101052 EFLGS=00010246


EBX=816a0a10 SS=0157 ESP=033bfd80 EBP=033bfdb4
ECX=c6b28c84 DS=0157 ESI=816a0a54 FS=2137
EDX=c0016228 ES=0157 EDI=70100000 GS=0000
Bytes at CS:EIP:
ff 70 28 ff 15 78 a0 14 70 c7 05 ec 91 14 70 01
Stack dump:
00000000 70100000 816a0a54 816a0a10 00000002 00000001 bff74277 033bfd84 033bfbac
033bff70 701090ce 701424d0 ffffffff 033bff80 bff7b9ce 70100000

The apprunner dos box had the following dialog/echo:
BrowserAppCore has been created.
Setting content window
Reading file...
Reading file...Done
Document: Done (26.91 secs)
Got a handle to forward menu item
Setting forward menu item disabled
Obtained MenuItem Back
Setting Back menuitem to enabled
Document http://www.mozilla.org/quality/smoketests/ loaded successfully
Reading file...
Reading file...Done
Init!
Reading file...
Reading file...Done
adding "http://cvs-mirror.mozilla.org/webtools/tinderbox/SeaMonkey/flash.rdf" to
 the tree
Connect: Contacting host: cvs-mirror.mozilla.org...
Connect: Host cvs-mirror.mozilla.org contacted. Waiting for reply...
Transferring data from cvs-mirror.mozilla.org
poll interval = "[xpconnect wrapped nsIRDFNode]"
http://cvs-mirror.mozilla.org/webtools/tinderbox/SeaMonkey/flash.rdf: setting po
ll interval to 300sec.
Exiting

------------
keyboard is responding, apprunner is stuck - when I try to shut down the dos box
the system will hang and I will have to power off.
Status: RESOLVED → REOPENED
Assignee: radha → law
Status: REOPENED → NEW
Target Milestone: M8
Bill,

With the additional information, are you able to reproduce this yet?
Status: NEW → RESOLVED
Closed: 25 years ago → 25 years ago
Status: RESOLVED → VERIFIED
Tracking all Crash on exit bug to
http://bugzilla.mozilla.org/show_bug.cgi?id=7799
Moving from Apprunner to Other component temporarily whilst don and I set proper
component.  Apprunner component will be retired/deleted shortly.
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.