Bug#561890: lenny kernel memory leak?
-------- Forwarded Message --------
From: Mikko Rapeli <mikko.rapeli@iki.fi>
To: Debian kernel team <debian-kernel@lists.debian.org>
Subject: Re: lenny kernel memory leak?
Date: Wed, 16 Dec 2009 08:56:27 +0200
On Tue, Dec 15, 2009 at 10:03:12PM +0000, Ben Hutchings wrote:
> I'm not aware of such a bug, but that doesn't mean it doesn't exist.
> Please note that the rt2500 driver is not part of the kernel package,
> which instead provides the rt2500pci driver. Have you tried using
> rt2500pci instead?
Actually I'm using rt2500pci. Full module list below.
> The strongest evidence is that the 'slab' (kernel heap) size grows
> hugely. The 'anonpages' size shrinking shows that this is squeezing
> user programs into swap.
Ok, slab did grow big compared to ram size. If I see this again I'll try
compiling the kernel with additional debug options. I think at least
CONFIG_DEBUG_SLAB would help.
-Mikko
Modules loaded on the host:
Module Size Used by
ipt_MASQUERADE 2432 1
iptable_nat 4616 1
nf_nat 15248 2 ipt_MASQUERADE,iptable_nat
nf_conntrack_ipv4 12044 3 iptable_nat,nf_nat
nf_conntrack 54308 4 ipt_MASQUERADE,iptable_nat,nf_nat,nf_conntrack_ipv4
ipt_REJECT 2688 2
xt_tcpudp 2816 6
iptable_filter 2560 1
ip_tables 9616 2 iptable_nat,iptable_filter
x_tables 12932 5 ipt_MASQUERADE,iptable_nat,ipt_REJECT,xt_tcpudp,ip_tables
ipv6 225172 14
sbp2 18700 0
ieee1394 75960 1 sbp2
loop 12428 0
arc4 1792 2
ecb 2560 2
crypto_blkcipher 14724 1 ecb
snd_cs4232 11316 0
snd_cs4231_lib 21376 1 snd_cs4232
ns558 3712 0
gameport 10504 2 ns558
rt2500pci 17152 0
rt2x00pci 7680 1 rt2500pci
rt2x00lib 22272 2 rt2500pci,rt2x00pci
rfkill 5524 1 rt2x00lib
led_class 3844 1 rt2x00lib
input_polldev 3720 1 rt2x00lib
mac80211 137756 2 rt2x00pci,rt2x00lib
evdev 7808 0
cfg80211 21384 2 rt2x00lib,mac80211
eeprom_93cx6 2176 1 rt2500pci
snd_ad1816a 16736 0
snd_opl3_lib 9344 2 snd_cs4232,snd_ad1816a
snd_hwdep 6148 1 snd_opl3_lib
snd_mpu401_uart 6400 2 snd_cs4232,snd_ad1816a
snd_rawmidi 18592 1 snd_mpu401_uart
snd_seq_device 6412 2 snd_opl3_lib,snd_rawmidi
pcmcia 29612 0
firmware_class 6656 2 rt2x00lib,pcmcia
8139too 20224 0
snd_pcsp 8220 0
uhci_hcd 18448 0
i2c_piix4 7184 0
snd_pcm 60680 3 snd_cs4231_lib,snd_ad1816a,snd_pcsp
i2c_core 19728 1 i2c_piix4
snd_timer 17668 4 snd_cs4231_lib,snd_ad1816a,snd_opl3_lib,snd_pcm
parport_pc 22436 0
yenta_socket 20620 2
8139cp 17024 0
rsrc_nonstatic 9472 1 yenta_socket
mii 4864 2 8139too,8139cp
usbcore 117616 2 uhci_hcd
parport 30408 1 parport_pc
snd 45092 11 snd_cs4232,snd_cs4231_lib,snd_ad1816a,snd_opl3_lib,snd_hwdep,snd_mpu401_uart,snd_rawmidi,snd_seq_device,snd_pcsp,snd_pcm,snd_timer
pcmcia_core 31888 3 pcmcia,yenta_socket,rsrc_nonstatic
soundcore 6112 1 snd
snd_page_alloc 7816 2 snd_cs4231_lib,snd_pcm
ext3 103688 4
jbd 35092 1 ext3
mbcache 6656 1 ext3
dm_mirror 14720 0
dm_log 8192 1 dm_mirror
dm_snapshot 14240 0
dm_mod 45384 13 dm_mirror,dm_log,dm_snapshot
thermal_sys 10784 0
ide_disk 10496 7
ide_pci_generic 3844 0 [permanent]
piix 6532 0 [permanent]
ide_core 95144 3 ide_disk,ide_pci_generic,piix
ata_generic 4612 0
libata 141216 1 ata_generic
scsi_mod 130060 2 sbp2,libata
dock 8076 1 libata
--
Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.
Reply to: