> I would guess it's some sort of race condition then. Did you give the
> upstream bugzilla a once over to see if anyone had reported something
> similar?
Good suggestion. See https://bugzilla.mozilla.org/show_bug.cgi?id=227168
that describes what i observed.
Greetings
Helmut Grohne
signature
forwarded 288554 https://bugzilla.mozilla.org/show_bug.cgi?id=227168
thanks
* Helmut Grohne ([EMAIL PROTECTED]) wrote:
> > I would guess it's some sort of race condition then. Did you give the
> > upstream bugzilla a once over to see if anyone had reported something
> > similar?
>
> Good suggest
This bug seems to be fixed in 1.0.5-1. I cannot reproduce the described
behaviour anymore. The bug should therefore be marked as closed.
Helmut
signature.asc
Description: Digital signature
> Did you say you were running on not very powerful hardware? Opening 30
> tabs is certainly going to put a strain on an older machine.
Yes, but 30 tabs *should* consume a lot of cpu. I'd said nothing if it
just took very long. But firefox sometimes even fails to redraw within 5
seconds *although
* Helmut Grohne ([EMAIL PROTECTED]) wrote:
> > > Firefox (among others like mozilla, some other curses applications) hang
> > > within a the following system call (obtained by strace) for 10 seconds
> > > to 3 minutes:
> > Hang and then resume functioning?
>
> Yes, the futex call somewhen returns.
> Give it a shot, try uninstalling them.
As I told you it doesn't matter whether plugins or extensions are
installed. I removed all of them and reproduced the described behaviour.
To start with a "clean" firefox i moved .mozilla out of the way. To
browse the web I had to enter a proxy autoconfigu
* Helmut Grohne ([EMAIL PROTECTED]) wrote:
> > Did you say you were running on not very powerful hardware? Opening 30
> > tabs is certainly going to put a strain on an older machine.
>
> Yes, but 30 tabs *should* consume a lot of cpu. I'd said nothing if it
> just took very long. But firefox some
> Is it swapping heavily? That doesn't use CPU.
No. It simply hangs.
One workaround for this problem is sending a lot of SIGCONTs to
firefox-bin using a little perl one-liner. As far as I know futex()
calls get interrupted when signals arrive.
If I remember correctly I had some similar problems
* Helmut Grohne ([EMAIL PROTECTED]) wrote:
> > Is it swapping heavily? That doesn't use CPU.
>
> No. It simply hangs.
>
> One workaround for this problem is sending a lot of SIGCONTs to
> firefox-bin using a little perl one-liner. As far as I know futex()
> calls get interrupted when signals arr
9 matches
Mail list logo
