Final notes: Re: 2.4.23-preX oops and strange swappiness (nforce2mobo)

From: Disconnect
Date: Mon Nov 03 2003 - 11:37:43 EST

...then it oopsed again, roughly the same way. (decodes available,
but..)

It looks like 2 possibly-related problems: swap didn't appear to be
reclaimed when I had a 2G swap partition, and under memory pressure the
system would oops. (With the 512M swap it would usually go OOM first and
recover.. with 2G it went oops and never made it to OOM. Both times the
system was generally ok after, with possible hung processes if it
oopsed. And reboots didn't work due to swapoff hanging.)

I'm going to go up to pre9 and keep the memory pressure down unless
someone wants more details. When we get into -rc I'll run it again and
see if it still breaks.

On Thu, 2003-10-30 at 11:44, Disconnect wrote:
> Updated info: I dropped the swap partition to 512M instead of 2G and it
> seems to have stablized nicely.
>
> At least, its been a day and a half without an oops, which is a record
> for this machine...
>
> On Mon, 2003-10-27 at 10:56, Disconnect wrote:
> > Unfortunately, I really shouldn't run without nvnet so if thats a known
> > issue then its not in google ;) (If its highly likely to be related I
> > can try it this week; it means I end up downing my wlan..)
> >
> > It doesn't take long - couple of days at most - before there are a
> > couple of oopses in the logs (decodes below) and swap usage begins
> > climbing without ever being reclaimed. (Similar problems occurred in
> > 2.4.23-pre5, with the added bonus of processes cratering out in D state
> > and never going away.)
> >
> > These are from an unpatched 2.4.23-pre8:
> > [10:49:24] [dis@floyd] $ uptime
> > 10:49:26 up 2 days, 30 min, 2 users, load average: 1.56, 1.57, 1.49
> > [10:49:26] [dis@floyd] $ free
> > total used free shared buffers cached
> > Mem: 256932 254736 2196 0 13804 64932
> > -/+ buffers/cache: 176000 80932
> > Swap: 2000084 443248 1556836
> > [10:49:27] [dis@floyd] $ cat /proc/slabinfo
> > slabinfo - version: 1.1
> > kmem_cache 62 72 108 2 2 1
> > ip_conntrack 255 804 320 67 67 1
> > fib6_nodes 12 112 32 1 1 1
> > ip6_dst_cache 13 20 192 1 1 1
> > ndisc_cache 3 30 128 1 1 1
> > tcp_tw_bucket 6 30 128 1 1 1
> > tcp_bind_bucket 72 448 32 4 4 1
> > tcp_open_request 0 30 128 0 1 1
> > inet_peer_cache 3 59 64 1 1 1
> > ip_fib_hash 13 112 32 1 1 1
> > ip_dst_cache 1850 1920 192 96 96 1
> > arp_cache 22 30 128 1 1 1
> > blkdev_requests 3072 3090 128 103 103 1
> > journal_head 1468 5236 48 55 68 1
> > revoke_table 3 250 12 1 1 1
> > revoke_record 0 112 32 0 1 1
> > dnotify_cache 0 0 20 0 0 1
> > file_lock_cache 20 42 92 1 1 1
> > fasync_cache 0 0 16 0 0 1
> > uid_cache 12 112 32 1 1 1
> > skbuff_head_cache 225 780 192 39 39 1
> > sock 240 336 896 83 84 1
> > sigqueue 0 29 132 0 1 1
> > kiobuf 0 0 64 0 0 1
> > cdev_cache 16 118 64 2 2 1
> > bdev_cache 5 59 64 1 1 1
> > mnt_cache 13 59 64 1 1 1
> > inode_cache 6821 6825 512 975 975 1
> > dentry_cache 5297 5310 128 177 177 1
> > filp 2374 2400 128 80 80 1
> > names_cache 0 2 4096 0 2 1
> > buffer_head 17182 17790 128 592 593 1
> > mm_struct 76 90 128 3 3 1
> > vm_area_struct 3811 3930 128 131 131 1
> > fs_cache 73 118 64 2 2 1
> > files_cache 74 81 448 9 9 1
> > signal_act 89 90 1344 30 30 1
> > size-131072(DMA) 0 0 131072 0 0 32
> > size-131072 0 0 131072 0 0 32
> > size-65536(DMA) 0 0 65536 0 0 16
> > size-65536 1 1 65536 1 1 16
> > size-32768(DMA) 0 0 32768 0 0 8
> > size-32768 4 4 32768 4 4 8
> > size-16384(DMA) 0 0 16384 0 0 4
> > size-16384 8 8 16384 8 8 4
> > size-8192(DMA) 0 0 8192 0 0 2
> > size-8192 2 2 8192 2 2 2
> > size-4096(DMA) 0 0 4096 0 0 1
> > size-4096 150 201 4096 150 201 1
> > size-2048(DMA) 0 0 2048 0 0 1
> > size-2048 63 78 2048 35 39 1
> > size-1024(DMA) 0 0 1024 0 0 1
> > size-1024 85 88 1024 22 22 1
> > size-512(DMA) 0 0 512 0 0 1
> > size-512 146 152 512 19 19 1
> > size-256(DMA) 0 0 256 0 0 1
> > size-256 517 525 256 35 35 1
> > size-128(DMA) 0 0 128 0 0 1
> > size-128 1810 1830 128 61 61 1
> > size-64(DMA) 0 0 64 0 0 1
> > size-64 3467 3481 64 59 59 1
> > size-32(DMA) 0 0 64 0 0 1
> > size-32 1612 2301 64 39 39 1
> > [10:51:34] [dis@floyd] $ lsmod
> > Module Size Used by Tainted: P
> > ipt_mark 472 46
> > ipt_state 568 0 (unused)
> > ipt_mac 696 2
> > ip_nat_irc 2192 0 (unused)
> > ip_conntrack_irc 3056 1
> > ip_conntrack_ftp 4016 1 (autoclean)
> > ip_nat_ftp 2864 0 (unused)
> > iptable_filter 1772 1
> > iptable_mangle 2168 1
> > ipt_LOG 3416 0 (unused)
> > ipt_TOS 1048 0 (unused)
> > ipt_REJECT 3544 0 (unused)
> > ipt_MARK 792 3
> > ipt_MASQUERADE 1464 8
> > ipt_REDIRECT 824 4
> > iptable_nat 16014 3 [ip_nat_irc ip_nat_ftp ipt_MASQUERADE ipt_REDIRECT]
> > ip_conntrack 19908 4 [ipt_state ip_nat_irc ip_conntrack_irc ip_conntrack_ftp ip_nat_ftp ipt_MASQUERADE ipt_REDIRECT iptable_nat]
> > ip_tables 12288 14 [ipt_mark ipt_state ipt_mac iptable_filter iptable_mangle ipt_LOG ipt_TOS ipt_REJECT ipt_MARK ipt_MASQUERADE ipt_REDIRECT iptable_nat]
> > nfsd 72880 8 (autoclean)
> > lockd 50928 1 (autoclean) [nfsd]
> > sunrpc 68736 1 (autoclean) [nfsd lockd]
> > sd_mod 11084 0 (autoclean) (unused)
> > sg 27516 0 (autoclean) (unused)
> > sr_mod 15576 0 (autoclean) (unused)
> > scsi_mod 85760 3 (autoclean) [sd_mod sg sr_mod]
> > ide-cd 32032 0 (autoclean)
> > cdrom 28320 0 (autoclean) [sr_mod ide-cd]
> > hid 21572 0 (unused)
> > usbmouse 2264 0 (unused)
> > ehci-hcd 17260 0 (unused)
> > nvnet 26272 1
> > ipv6 167124 -1
> > tulip 40608 1
> > crc32 2880 0 [tulip]
> > usbkbd 3640 0 (unused)
> > usb-ohci 18888 0 (unused)
> > usbcore 62892 1 [hid usbmouse ehci-hcd usbkbd usb-ohci]
> >
> > Crash dump and config attached - if you need anything else let me know.
> >
> > (I'm on the list, no need to CC)
> >
> > Thanks!
--
Disconnect <lkml@xxxxxxxxxxx>

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