Skip to contentSkip to conversation

Athlon optimisation bug (was Re: Duron kernel crash)

GOLDMINE!!! Dmitry Volkoff 4 participants 4 messages Sep 12, 2001

lore
Share & tools
lore
#0 lore raw
Quoted 6 lines

I don't have any problems with the kernel, but if I'm compiling (say) the kernel or any large program, user-mode apps like gcc, make, etc all randomly segfault on me.

I'm running a Duron 800 on a Asus A7V133 mobo - and no bios version I've tried has made a bit of difference yet :p

Same things here. I'm also experiencing random segfaults with kernels 2.4.10pre[47]. My motherbord is Chaintech CT-7AIA2 (kt133A/686A), Duron 600. I did'nt see such things with kernels 2.4.[78]. The problem persists even after recompiling kernel with K6-III optimisation. Note, this motherboard was rock solid and I never had any problems with athlon-optimised kernels prior to 2.4.9.

Signature · 1 line
    DV
#1 lore raw
Quoted 7 lines · Dmitry Volkoff
Quoted 6 lines

I don't have any problems with the kernel, but if I'm compiling (say) the kernel or any large program, user-mode apps like gcc, make, etc all randomly segfault on me.

I'm running a Duron 800 on a Asus A7V133 mobo - and no bios version I've tried has made a bit of difference yet :p

Quoted 7 lines

Same things here. I'm also experiencing random segfaults with kernels 2.4.10pre[47]. My motherbord is Chaintech CT-7AIA2 (kt133A/686A), Duron 600. I did'nt see such things with kernels 2.4.[78]. The problem persists even after recompiling kernel with K6-III optimisation. Note, this motherboard was rock solid and I never had any problems with athlon-optimised kernels prior to 2.4.9.

A colleague purchased a number of A7V133s all of which exhibited seg faults in make and gcc. make reproducibly seg faulted in vfork(). This was with 2.2.18 kernels compiled for i586. We next tried a 2.4.8 kernel built for i686 and the problem persisted, though less reproducibly. The kernel reports 'Applying VIA southbridge workaround', but it looks like we need another fix.

Most of the BIOS options had no effect, but changing the 'PCI latency' setting from 32 to 64 seems to have fixed it, fingers crossed.

P.

Signature · 4 lines
+--------------------------------------------------+
| Peter Horton      | [email protected]        |
| Software Engineer | http://www.colonel-panic.org |
+--------------------------------------------------+
#2 lore raw

Intresting. I'll try that, because it sounds exactly what I had (constant segfaults in 2.2.18, occasional and more random in 2.4.x).

- Ender

Quoted 23 lines

A colleague purchased a number of A7V133s all of which exhibited seg faults in make and gcc. make reproducibly seg faulted in vfork(). This was with 2.2.18 kernels compiled for i586. We next tried a 2.4.8 kernel built for i686 and the problem persisted, though less reproducibly. The kernel reports 'Applying VIA southbridge workaround', but it looks like we need another fix.

Most of the BIOS options had no effect, but changing the 'PCI latency' setting from 32 to 64 seems to have fixed it, fingers crossed.

P.

--

+--------------------------------------------------+
| Peter Horton      | [email protected]        |
| Software Engineer | http://www.colonel-panic.org |
+--------------------------------------------------+

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#3 lore raw
In reply to Dmitry Volkoff
Quoted 3 lines

2.4.10pre[47]. My motherbord is Chaintech CT-7AIA2 (kt133A/686A), Duron 600. I did'nt see such things with kernels 2.4.[78]. The problem persists even after recompiling kernel with K6-III optimisation. Note, this motherboard

Thats unrelated. Thats 2.4.10pre being exactly what it sounds - a buggy pre-release snapshot