Closed Bug 156882 Opened 22 years ago Closed 20 years ago

mozilla crashes when loading site

Categories

(Core :: Internationalization, defect)

Product:
Core
Component:
Internationalization
Platform:
x86
Linux
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

()

Status:
RESOLVED WORKSFORME

People

(Reporter: niels, Assigned: shanjian)

References

(
URL
)

Details

(4 keywords)

Attachments

(1 file, 1 obsolete file)

Crash improved Testcase (Now crashes better than testcase #1)
1.67 KB, text/html
Details 
just that. :-)
wfm with win2k build 20020710..

Reporter:
Do you have a talkback ID from that crash ?
And always add the build ID/Milestone in a bug report.

hi,

sorry first time.

I'm using mozilla Build ID 2002062608; same behavior with mozilla Build ID
2002061911-SuSE. Mozilla starts to display the titlebar "Jobs.ge ..." and than
hangs up.

niels
crashes 20020711 .. talkback doesn't show up..

/dev/sda says 'trrrr..' -> ..swap goes up to 100% and then mozilla is gone
(killed from kernel i think..)

Mozilla crashed for me many times like this.. started imho about the 7. 5..
(most  times on pr0n pages.. so i did not file a bugreport.. just joking ;) )
Attached file Testcase (CRASHES MOZILLA!!) (obsolete) — 
<META http-equiv="Content-Type" content="charset=x-user-defined">
+ &?uml; crashes mozilla..

charset=ISO-blabla works..
Newsflash:
----------
Oh.. it seems that many &?uml; will use much memory..

So on a Computer with much memory, mozilla will not crash/hang on testcase #1
(has only one &ouml;)..
Testcase 2 has many &ouml;'s and should crash/hang even on computers with much
memory..

Seems to be a new bug, could not find a dup.. whats the component for this?

..i was wondering, why newer builds crashed so often on japanese pages, so i
think that caused it.. imho very for annoing japanese users..
Attachment #90961 - Attachment is obsolete: true
WFM (2002071008+WinXP+SP1) 128 MB RAM ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö ö is what i get.
Adrian and Niels: what kind of font settings do you have? specifically for
user-defined, since that seems relevant to this bug.  Do you crash with a clean
profile?

WFM with linux build 20020710
Oof... locked up my machine* for about 60 seconds with 10 mozilla-bin processes
using about 20% of memory each, according to `top`.

Even now, after I closed the offending tab, Mozilla is using 78% of CPU on
average.  I guess it will continue to do so until I close the whole application.

-----
* Machine: AMD 1700+ XP, 512 MB of DDR SDRAM with custom kernel 2.4.18 running
KDE 3.0.0-10.
New profile doesn't help.. i didn't touch anything in the font settings..

I'm using XFree 4.1.0.1 and linux 2.4.19-rc1 (with crypto support..)
Debian Woody
PC has 128 mb Ram & ~ 200 mb Swap


Mozilla doesn't really crash.
It just uses MUCH memory (testcase 1 uses about 80% of my swap space and has
about 1 minute to render, with testcase 2, mozilla eats up all memory and got
killed from linux)
Mac os X 20020710 (600mhz G3)
Works fine ~ 1 second to render both testcases

Linux 2.4.18 - Slack 8 20020711 (x86 - 400mhz) XFree 4.2
Over 1 minute to render testcase 1.. testcase 2 seems to hang forever..

Linux 2.4.19-rc1 - Debian Woody(ppc) 20020703 (600mhz G3) XFree 4.1.0.1
Hangs a VERY LONG time, like on Slack 8.. you want to kill mozilla..


.. All builds got a clean profile, didn't change any font-settings..

But now it seems, that mozilla doesn't always eat up the swapspace.. but in
every case it uses 100% of the CPU..
Keywords: testcase
oh ya.  It Mozilla banged my CPU for a good long while, but did not
significantly increase in mem usage, and certainly didn't crash for me.  We need
to know what triggers the mem usage...
Maybe we got a dup of this..

does the testcase of bug 157320 also hang/crahs for people wich could reproduce
the bug?
..and does it work for people wich couldn't reproduce this bug?
Do you see this with builds from mozilla.org or only with other builds ?

mozilla 1.0 rc2 on GNU/Linux i386, XFree-4.0.2

Loading the page happens to be very slow, and xfs' memory footprint grows from
25 MB to more than 65MB.
the CPU usage is from Mozilla desparately trying to find a font.

with a mozilla build (clean profile), testcase hangs for about a minute,
displays question marks (?)
optimized CVS (same profile!), testcase renders quickly and properly (ö)
debug CVS (same profile!!!), testcase behaves as the mozilla.org build.  much
CPU usage and displays only "?"

debug build also does a lot of:
charset = xxx, file nsFontMetricsGTK.cpp, line 2025
xxx=ISO 50% of the time and other things (jis, gb_, KOI...), the other 50%
and
###!!! ASSERTION: mFontHolder should not be loaded: '!mFontHolder', file
nsFontMetricsGTK.cpp, line 2237

gdb showed the debug build always doing 
nsFontMetricsGTK::SearchNode at nsFontMetricsGTK.cpp:3430
nsFontMetricsGTK::FindAnyFont at nsFontMetricsGTK.h:190
nsFontMetricsGTK::FindFont at nsRenderingContextGTK.cpp:1617

==> Intl
marking NEW, although I still don't know why it crashes.  hopefully, it's
related to what I'm seeing.
Assignee: Matti → yokoyama
Status: UNCONFIRMED → NEW
Component: Browser-General → Internationalization
Ever confirmed: true
QA Contact: asa → ruixu
WFM, current CVS, Linux. Allthough the fonts in this case do not show correct as
i don't have the charset installed
font related -> shanjian
Assignee: yokoyama → shanjian
Keywords: intl
QA Contact: ruixu → ylong
Tried this on 07-15 1.0.1 branch build/linux RH 7.2:
1. No crash.
2. Hang with the testcase attached in comment #5.
3. Load fine with testcase in comment #4 and page www.job.ge.
see also bug 153688
WFM

	
Mozilla 1.6
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.6) Gecko/2004011

Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7b) Gecko/20040310 Firefox/0.8.0+

on FC1

WFM, official Moz 1.7b and my Firefox-20040330 build on Linux.

Reopen if this problem appeas for you.
Status: NEW → RESOLVED
Closed: 20 years ago
Resolution: --- → WORKSFORME
Crashtest added as part of http://hg.mozilla.org/mozilla-central/rev/afc662d52ab1
Flags: in-testsuite+
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: