Very Computer

       I`ve experienced strange gcc sig11 crashes.
    prerequisities: gcc-2.95.3-vanille, 2.4.7-vanille

    recently i`ve dloaded arachne, a svgalib-based
    web browser. It has quite a number of various problems,
    with console-switching screen/terminal corruptions
    being most significant. so as i said, it crashed
    alot, and i recovered from these crashes w/o reboot.

    then i get libggi-2.0.4b4. (that is all happening
    w/o reboot between actions)

    i`ve started compile process. bah! - sig11.

    Although i quite often have those, but
    this time i had about _100_ of them!!!
    all these in libggi. I `ve got tired of tweakin`
    cflags of different makfiles, because it helps
    to go thru prblematic .c`s, and i rebooted.

    Hm the /most/ mystic of all this is the fact, that
    after reboot sig11 rate dropped from sig11 per ~4
    files to one per ~120 of them. To make clear
    the situation, i`ll tell that libggi is VERY
    broken for me in the scope of gcc`s sig11.

    also i can say, that the sig11/file rate had been
    constantly increasing with the compilation process.

    so it seems to me like kernel problem...

    further information upon request.

---

cheers,

   Samium Gromoff
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Sounds like flaky hardware. Something in your system is probably overheating
and causing problems. Try swapping the components in your system one at a
time, starting with memory.

--

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[random sig11s]

Time to read http://www.bitwizard.nl/sig11/ .

HTH,
Ray
--
The "Penguin Powered" logos people love should really be "Penguin
Empowered". That, I think, is the best thing about Linux, too. We've given
the computer back to the user.
        Alan Cox in http://www2.linuxjournal.com/articles/culture/012.html

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Quote:> Sounds like flaky hardware. Something in your system is probably overheating
> and causing problems. Try swapping the components in your system one at a
> time, starting with memory.

First check that the fans on the CPU and elsewhere work.
--

Casilla 9G, Vin~a del Mar, Chile                               +56 32 672616
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Quote:>>     so it seems to me like kernel problem...

>why is that?  I've never seen a sig11 from production >code
>that wasn't caused by flakey ram.  in fact, your >descriptions
>are a perfect example of similar hardware problems.

and some other people told me about cpu overheating...

but 55 minutes long memtest run showed no problems et al
 with cpu (P166-nonMMX-oc`ed to 180) staying warm, not by any means hot
 on mobo Zida 5DVX.

 maybe 55 min is not enough for proper mem testing?

---

cheers,

   Samium Gromoff
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[...]

Quote:>  maybe 55 min is not enough for proper mem testing?

Not enough to be sure. I have had memtest86 report problems
only after 11 hours and only in a specific test. Others on
the list have reported memtest runs much longer that did
not find anything but still their gcc problems went away
after putting in new RAM.

For some reason gcc seems to be an excellent RAM tester ;)

Regards,
  Rasmus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Synthetic tests are never as good as a real good gcc run, I'd *never*
trust them over the indications given by attempting to compile the kernel
or something big like XFree86.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Quote:> > >>     so it seems to me like kernel problem...

> > >why is that?  I've never seen a sig11 from production >code
> > >that wasn't caused by flakey ram.  in fact, your >descriptions
> > >are a perfect example of similar hardware problems.

> Synthetic tests are never as good as a real good gcc run, I'd *never*
> trust them over the indications given by attempting to compile the kernel
> or something big like XFree86.

I agree.  I had my computer mildy overclocked for a little while.  Everything
ran just great, I could compile kernels, quake3 timedemo for two days, etc.  
All was well until I tried to compile gcc itself and I kept getting random
errors.  I finally opened up the case and bumped it back down, and like magic
it compiled just fine.  

So my advice would be to don't overclock.

Josh
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/