Re: Unkillable processes?

[Bradley M Alexander <storm@tux.org>]

On Wed, Jul 16, 2003 at 11:51:18PM +0200, Nicos Gollan wrote:
> 
> Assuming that everything else works, I'd think it's dud memory. Try running 
> memtest86 for some time. That's the only thing that made my always-on boxes 
> misbehave up to now.

You know, I rebooted last night (it felt very MCSE-ish of me :), and got a
kernel trap, which I didn't have time to analyze yet, but I'm inclined to
agree. The trap is as follows. Is there any way to tell which SIMM to
replace?

Jul 16 23:02:04 lakota kernel: Unable to handle kernel paging request at
virtual address 83dd7120
 Jul 16 23:02:04 lakota kernel:  printing eip:
 Jul 16 23:02:04 lakota kernel: c01d82d0
 Jul 16 23:02:04 lakota kernel: *pde = 00000000
 Jul 16 23:02:04 lakota kernel: Oops: 0002
 Jul 16 23:02:04 lakota kernel: CPU:    0
 Jul 16 23:02:04 lakota kernel: EIP:    0010:[dput+32/352]    Not tainted
 Jul 16 23:02:04 lakota kernel: EFLAGS: 00010282
 Jul 16 23:02:04 lakota kernel: eax: 0000001b   ebx: 83dd7120   ecx:
 00000002   edx: 0000001c
 Jul 16 23:02:04 lakota kernel: esi: c3d9f1c0   edi: c3d88740   ebp:
 0000074a   esp: c1111f34
 Jul 16 23:02:04 lakota kernel: ds: 0018   es: 0018   ss: 0018
 Jul 16 23:02:04 lakota kernel: Process kswapd (pid: 4, stackpage=c1111000)
 Jul 16 23:02:04 lakota kernel: Stack: 00000292 c5ffef18 83dd7120 c3d9f1c0
 c01d869d 83dd7120 c3d9f1c0 0000000c 
 Jul 16 23:02:04 lakota kernel:        000001d0 00000020 00000006 c01d89c4
 000015f7 c01bce07 00000006 000001d0 
 Jul 16 23:02:04 lakota kernel:        c0103230 00000006 000001d0 c0103230
 00000000 c01bce66 00000020 c0103230 
 Jul 16 23:02:04 lakota kernel: Call Trace:    [prune_dcache+253/352]
 [shrink_dcache_memory+36/64] [shrink_caches+119/160] [try_to_free_pages_zone+54/80]
 [kswapd_balance_pgdat+92/176]
 Jul 16 23:02:04 lakota kernel:   [kswapd_balance+40/64] [kswapd+157/192]
 [arch_ ernel_thread+46/64] [kswapd+0/192]
 Jul 16 23:02:04 lakota kernel: 
 Jul 16 23:02:04 lakota kernel: Code: ff 0b 0f 94 c0 84 c0 0f 84 c3 00 00
 00 8d 7
 3 18 39 73 18 74 
 

> An alternative idea would be that you're running an extremely funky kernel 
> (early 2.4 series perhaps?) that could need upgrading.

Nope. Got 2.4.21 as soon as it was released.