kswapd failed under 2.1.63

Nicholas J. Leon (nicholas@binary9.net)
Sat, 15 Nov 1997 17:31:12 -0500 (EST)


Whilst running 'badblocks -vw -o ....' on an 2.4GB RAID5 device (across
3 drives) on

Linux 2.1.63 (root@neko) (gcc 2.7.2.3) #25 Fri Nov 14 10:50:08 EST 1997 [neko]

Memory: Total Used Free Shared Buffers Cached
Mem: 46400 32456 13944 9556 11212 6816
Swap: 130716 140 130576

Bootup: Sat Nov 15 16:34:09 1997 Load average: 0.01 0.02 0.25 2/31 325

user : 0:08:07.83 15.9% page in : 1219264 disk 1: 33977r 38075w
nice : 0:00:00.00 0.0% page out: 2626370
system: 0:08:13.39 16.0% swap in : 10 disk 3: 4996r 17046w
idle : 0:34:54.56 68.1% swap out: 53 disk 4: 4240r 16578w
uptime: 0:51:15.76 context : 857898

irq 0: 307578 ide1 irq 8: 0
irq 1: 1794 keyboard irq 9: 0
irq 2: 0 cascade [4] irq 10: 1235 3c509
irq 3: 0 irq 11: 0
irq 4: 40798 serial irq 12: 0
irq 5: 0 irq 13: 1 fpu
irq 6: 3 irq 14: 71907 ide0
irq 7: 0 irq 15: 1400897 ide1

I got loads of these

Kswapd failed: got 27/96

Out of memory for kswapd<c3800fc8/bf7ff90>

...

The machine ran long enough to switch vcs, type a few commands and watch
them fail in the same way. After that, it was hard reset time except for
the ability to change vc's.

Practically everything is via modules/kerneld 'cept for ext2fs.

I also notice that the <x/x> notation looks interestingly similar to the
output from catching the f00f bug.

___________________________________________________________________________

simple is elegant nicholas@binary9.net
___________________________________________________________________________