Closed Bug 607514 Opened 14 years ago Closed 13 years ago

Crash when available RAM is low

Categories

(Firefox :: General, defect)

3.6 Branch
x86
Windows 7
defect
Not set
normal

Tracking

()

VERIFIED INCOMPLETE

People

(Reporter: lucario365, Unassigned)

Details

(Keywords: crash)

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 6.1; es-ES; rv:1.9.2.11) Gecko/20101012 Firefox/3.6.11
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 6.1; es-ES; rv:1.9.2.11) Gecko/20101012 Firefox/3.6.11

Firefox crashes when available RAM decreases under 350MB.

Reproducible: Sometimes

Steps to Reproduce:
Make any application "eat" all available RAM.
Actual Results:  
Firefox closes without warning.

Expected Results:  
Firefox shouldn't close. If it needs more memory and there's not enough available, it should alert user can't continue using it due "RAM problems", alowing the user close some tabs, close saving actual state, disable add-ons or plugins, etc.
Did you see a crash ? See <about:crashes> for a list of recent crash-ids.

If a memory allocation doesn't work, we need to know where the fault lies. If your report doesn't mention anything about that, it's useless.

Firefox has a low-memory handler, which tries to survive this OOM situation (clearing memory cache etc ...). On the other hand, this becomes so rare (users have enough memory and swapspace now), that future code might not care anymore, and assume that a memory allocation might never fail.
Many times when reopening Firefox a yellow bar tells something about Crash Reporter Helper (I don't remember exactly what) and that the last close was "bad". The last time, the Crash Reporter window appeared when crashed Firefox, but it failed to send the report (I'm going to see if my firewall is blocking it, but I don't think so).
<about:crashes> doesn't show anything about this unexpected closes.
I'm going to crash firefox one more time and send the report if crash reporter appears. And I'll try to replicate the results in a new profile.



(In reply to comment #1)
> Did you see a crash ? See <about:crashes> for a list of recent crash-ids.
> 
> If a memory allocation doesn't work, we need to know where the fault lies. If
> your report doesn't mention anything about that, it's useless.
> 
> Firefox has a low-memory handler, which tries to survive this OOM situation
> (clearing memory cache etc ...). On the other hand, this becomes so rare (users
> have enough memory and swapspace now), that future code might not care anymore,
> and assume that a memory allocation might never fail.
Ok. In a new and empty profile the crash also happens.
The last crash is now in my <about:crashes> with ID: bp-706f6927-a42c-47d1-a951-cbc732101027.
Version: unspecified → 3.6 Branch
I'm not able to load that report.
First of all, I upgraded to 3.6.12. As you can imagine, the bug is still there.
I don't know why, but with this kind of crash the Crash Reporter always fails sending the report (it works with other crashes. Recently FF crashed by clicking a link in Kingston's web site, and the CRH submitted the report successfuly).
I have another ID for you (bp-6efed0e5-b346-4d17-aa4c-c7eaa2101102), but probably you won't be able to load it.
Firefox says in <about:crashes> the report is "waiting" (or something similar. I'm not English and I don't know what appears exactly in the English version).

If there's another way to obtain the report data or see where is the problem, tell me what I must do (I'm an experimented programmer in Visual Basic 6 and a begginer in VB.NET and Borland C++Builder 6, but I'm not too newbie in C/C++ for microcontrolers). I didn't try debug the code because I found it a bit complicated for me.
You could also try to reproduce it. If you have Win7, the quickest way is disable the pagefile and scan an external hard-drive with the chkdsk utility, if you don't have any application which needs a lot of RAM.
Keywords: crash
Reporter, Firefox 4.0.1 has been released, and it features significant improvements over previous releases. Can you please update to Firefox 4.0.1 or later, and retest your bug? Please also create a fresh profile (
http://support.mozilla.com/kb/Managing+profiles), update your plugins (Flash, Java, Quicktime, Reader, etc) and update your graphics driver and Operating system to the latest versions available. 

If you still continue to see this issue, please comment. If you do not, please close this bug as RESOLVED > WORKSFORME

filter: prefirefox4uncobugs
Thanks, but I'm not interested on Firefox 4.

(In reply to comment #6)
> Reporter, Firefox 4.0.1 has been released, and it features significant
> improvements over previous releases. Can you please update to Firefox 4.0.1
> or later, and retest your bug? Please also create a fresh profile (
> http://support.mozilla.com/kb/Managing+profiles), update your plugins
> (Flash, Java, Quicktime, Reader, etc) and update your graphics driver and
> Operating system to the latest versions available. 
> 
> If you still continue to see this issue, please comment. If you do not,
> please close this bug as RESOLVED > WORKSFORME
> 
> filter: prefirefox4uncobugs
Firefox 3.6 is nearing it's end of life, and will not be receiving any more updates. and an issue like this would probably only be fixed in Firefox 5 or later. By using old software you are opening yourself up to security holes and using slow software with known bugs. Please update to Firefox 5 and test. without that information, there is nothing we can do on this bug.
Status: UNCONFIRMED → RESOLVED
Closed: 13 years ago
Resolution: --- → INCOMPLETE
I tested FF4 and 5 and they are slower than 3.6! Versions 4 and 5 are faster if you compare a clean installation of them with a full-of-addons version 3.6 installation (not to mention the issue with resource consumption, which is enormous in newer versions). But this is not my case.
Security holes? Since last reinstall of windows I didn't updated it, and I never had problems (in any case, Windows is faster without updates). People have paranoia with security. Be careful and you don't need patches to cover holes or antivirus. You'll need only a firewall to control net connections, what enters and what goes out.

You can consider this bug settled. You didn't resolved the issue at time and I was forced to setup an enormous pagefile. As soon as FF 3.6 loses all support in the web, I'll migrate to Opera. I don't like Chrome nor its clones (FF4 and higher).

And think that I admired FF team and all the people who poured their efforts into it... Now that team is just an imitator of the big G.

(In reply to comment #8)
> Firefox 3.6 is nearing it's end of life, and will not be receiving any more
> updates. and an issue like this would probably only be fixed in Firefox 5 or
> later. By using old software you are opening yourself up to security holes
> and using slow software with known bugs. Please update to Firefox 5 and
> test. without that information, there is nothing we can do on this bug.
And thus is how botnets are created...
Status: RESOLVED → VERIFIED
And if you see zombie-PCs owners you'll see extremely low level users which were infected using psychological tricks, with or without security holes. Another thing is business area.

(In reply to comment #10)
> And thus is how botnets are created...
You need to log in before you can comment on or make changes to this bug.