Re: [e-users] Randomg segfaults

[The Rasterman]

On Fri, 16 Jun 2017 17:55:34 + Wido  said:

> heap is corrupted!??!?! should I check the physical ram? because I've
> powered off the laptop a bunch of times and the error still happend (power
> off was not to test WSOD)

that would be unlikely... unless you are also seeing other random segv's of
processes and sometimes kernel oopses as well...

> I have placed those EINA vars in my .profile, I expect that to work,
> because the packages I'm using have the xsession.destop file and I don't
> think I could place the vars in there

this might make segv's happen more often - but hopefully be able to catch the
"it would have done something wrong here... but worked anyway" cases so maybe
tell us more...

but without debug symbols that something more is... nothing really... :)

> Regarding valgrind, debug symbols and libunwind. I haven't used c debug
> tools in a long LONG time, and I haven't compiled E since e18. I "could"
> try, but it's going to take some (lot?) time to prepare that.
> 
> 
> Thanks Raster for the feedback =)
> 
> El jue., 15 de jun. de 2017 a la(s) 23:30, Carsten Haitzler <
> ras...@rasterman.com> escribió:
> 
> > On Thu, 15 Jun 2017 13:44:55 + Wido  said:
> >
> > > Hi all,
> > >
> > > I'm having some random segfaults. I'm using debian testing (stretch) +
> > > sparky repos. Sparky compiles the latest releases, so I'm using, at the
> > > moment of this writting, elf 1.19.1-1 and e 0.21.8-1
> > >
> > > The hardware is a lenovo e440, with ssd disk (encrypted home), core i5,
> > > intel integrated gpu (driver i915...i think. xorg.log wasn't exactly
> > clear
> > > on this), and I usually have 3 screens (laptpot + 2 monitors)
> > >
> > > I haven't detected what triggers the segfault, sometimes are 3 in half an
> > > hour, sometimes nothing happens for long time. Most of the times I get
> > back
> > > with F1, but sometimes the white-screen-of-dead doesn't show and I have
> > to
> > > restart X
> > >
> > > Here's a link to the .xsession-errors, with the memmap and backtrace.
> > > There's nothing else in the logs (dmesg, messages, kern.log, xorg.log).
> > And
> > > I don't have any debug tools, as I'm just a user and not a dev :P
> > >
> > > https://pastebin.com/yxmeY298
> >
> > well that tells me "heap is corrupted and malloc is now unhappy". why - no
> > idea. no info on why. somethnig has scribbled over the heap memory outside
> > an
> > allocation tho... to catch this you need valgrind. and valgrind will make
> > everything super-slow.
> >
> > you COULD try setting these env vars before running e:
> >
> > export EINA_LOG_BACKTRACE=999
> > export EINA_FREEQ_TOTAL_MAX=1024
> > export EINA_FREEQ_FILL_MAX=4096
> >
> > now ... maybe if you also have debug symbols (you don't) and libunwind
> > installed on efl compile... then you'll get backtraces for error logs
> > AND...
> > efl will fill freed memory with patterns and have a purgatory space with
> > "to be
> > freed memory on a queue" filled with other patterns... and it might give us
> > more info on what is happening with a live gdb session...
> >
> > --
> > - Codito, ergo sum - "I code, therefore I am" --
> > The Rasterman (Carsten Haitzler)ras...@rasterman.com
> >
> > --
> Wido
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> enlightenment-users mailing list
> enlightenment-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/enlightenment-users


-- 
- Codito, ergo sum - "I code, therefore I am" --
The Rasterman (Carsten Haitzler)ras...@rasterman.com


--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
enlightenment-users mailing list
enlightenment-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-users

Re: [e-users] Randomg segfaults

[Wido]

heap is corrupted!??!?! should I check the physical ram? because I've
powered off the laptop a bunch of times and the error still happend (power
off was not to test WSOD)

I have placed those EINA vars in my .profile, I expect that to work,
because the packages I'm using have the xsession.destop file and I don't
think I could place the vars in there

Regarding valgrind, debug symbols and libunwind. I haven't used c debug
tools in a long LONG time, and I haven't compiled E since e18. I "could"
try, but it's going to take some (lot?) time to prepare that.


Thanks Raster for the feedback =)

El jue., 15 de jun. de 2017 a la(s) 23:30, Carsten Haitzler <
ras...@rasterman.com> escribió:

> On Thu, 15 Jun 2017 13:44:55 + Wido  said:
>
> > Hi all,
> >
> > I'm having some random segfaults. I'm using debian testing (stretch) +
> > sparky repos. Sparky compiles the latest releases, so I'm using, at the
> > moment of this writting, elf 1.19.1-1 and e 0.21.8-1
> >
> > The hardware is a lenovo e440, with ssd disk (encrypted home), core i5,
> > intel integrated gpu (driver i915...i think. xorg.log wasn't exactly
> clear
> > on this), and I usually have 3 screens (laptpot + 2 monitors)
> >
> > I haven't detected what triggers the segfault, sometimes are 3 in half an
> > hour, sometimes nothing happens for long time. Most of the times I get
> back
> > with F1, but sometimes the white-screen-of-dead doesn't show and I have
> to
> > restart X
> >
> > Here's a link to the .xsession-errors, with the memmap and backtrace.
> > There's nothing else in the logs (dmesg, messages, kern.log, xorg.log).
> And
> > I don't have any debug tools, as I'm just a user and not a dev :P
> >
> > https://pastebin.com/yxmeY298
>
> well that tells me "heap is corrupted and malloc is now unhappy". why - no
> idea. no info on why. somethnig has scribbled over the heap memory outside
> an
> allocation tho... to catch this you need valgrind. and valgrind will make
> everything super-slow.
>
> you COULD try setting these env vars before running e:
>
> export EINA_LOG_BACKTRACE=999
> export EINA_FREEQ_TOTAL_MAX=1024
> export EINA_FREEQ_FILL_MAX=4096
>
> now ... maybe if you also have debug symbols (you don't) and libunwind
> installed on efl compile... then you'll get backtraces for error logs
> AND...
> efl will fill freed memory with patterns and have a purgatory space with
> "to be
> freed memory on a queue" filled with other patterns... and it might give us
> more info on what is happening with a live gdb session...
>
> --
> - Codito, ergo sum - "I code, therefore I am" --
> The Rasterman (Carsten Haitzler)ras...@rasterman.com
>
> --
Wido
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
enlightenment-users mailing list
enlightenment-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-users

[e-users] interaction question

[Christopher Barry]

E folk, 

I've always wondered why on the desktop I cannot click and drag a
shaded window selector across several icons, then simply hit delete to
delete them all. It is a standard interaction methodology on lots of
various desktop systems. It seems like this was a conscious design
decision to implement this, and I guess I'm not understanding the
rationale there. Is this really on purpose, or do I simply have
something mis-configured?

Thanks

-- 
Regards,
Christopher

--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
enlightenment-users mailing list
enlightenment-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/enlightenment-users