Bad, bad, bad VM behaviour in 2.4.10

Bad, bad, bad VM behaviour in 2.4.10

Post by Marcelo Tosatt » Thu, 27 Sep 2001 06:30:17



Juan,

It is a known problem which we are looking into.

I need some information which may help confirm a guess of mine:

Do you have swap available ?

If so, there was available anonymous memory to be swapped out ?


> Hi!

> My test is very simple. I have started X-Window and XMMS in order to
> listen to some songs. Then, I have executed

>    dd if=/dev/hdc1 of=/dev/null

> as root within a terminal, and I have got the following a few seconds
> later:

> Sep 25 22:05:55 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:00 localhost last message repeated 2 times
> Sep 25 22:06:00 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0xf0/0) from c012ff60
> Sep 25 22:06:00 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:00 localhost last message repeated 2 times
> Sep 25 22:06:00 localhost kernel: VM: killing process xmms
> Sep 25 22:06:00 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:04 localhost last message repeated 11 times
> Sep 25 22:06:04 localhost kernel: VM: killing process xmms
> Sep 25 22:06:04 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:04 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:04 localhost kernel: VM: killing process kmix
> Sep 25 22:06:04 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:04 localhost last message repeated 2 times
> Sep 25 22:06:04 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0xf0/0) from c012ff60
> Sep 25 22:06:04 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:04 localhost last message repeated 3 times
> Sep 25 22:06:04 localhost kernel: VM: killing process gpm
> Sep 25 22:06:05 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:06 localhost last message repeated 6 times
> sep 25 22:06:06 localhost su(pam_unix)[2548]: session closed for user
> root
> Sep 25 22:06:06 localhost kernel: VM: killing process sendmail
> Sep 25 22:06:07 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:07 localhost kernel: VM: killing process konsole
> Sep 25 22:06:07 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:07 localhost kernel: VM: killing process named
> Sep 25 22:06:07 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:07 localhost kernel: VM: killing process xmms
> Sep 25 22:06:07 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:07 localhost last message repeated 3 times
> Sep 25 22:06:07 localhost kernel: VM: killing process ksmserver
> Sep 25 22:06:07 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:07 localhost kernel: VM: killing process kdeinit
> Sep 25 22:06:08 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:08 localhost kernel: VM: killing process X
> Sep 25 22:06:08 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0xf0/0) from c012ff60
> Sep 25 22:06:08 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> f60
> Sep 25 22:06:08 localhost last message repeated 2 times
> Sep 25 22:06:08 localhost kernel: VM: killing process kdeinit
> Sep 25 22:06:08 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:08 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:08 localhost kernel: VM: killing process startkde
> Sep 25 22:06:09 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:09 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60
> Sep 25 22:06:09 localhost kernel: VM: killing process named
> Sep 25 22:06:09 localhost kernel: __alloc_pages: 0-order allocation
> failed (gfp=0x1d2/0) from c012ff60

> The /dev/hdc1 partition capacity is 6 GB. My root partition is on
> /dev/hda5. My computer is a Pentium III with 384 MB of RAM.

> BTW, the same test in 2.4.6 works fine without any problem.

> Regards.

> --
> D. Juan Piernas Cnovas
> Departamento de Ingeniera y Tecnologa de Computadores
> Facultad de Informtica. Universidad de Murcia
> Campus de Espinardo - 30080 Murcia (SPAIN)
> Tel.: +34968367657    Fax: +34968364151

> PGP public key:
> http://pgp.rediris.es:11371/pks/lookup?search=piernas%40ditec.um.es&o...
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
 
 
 

Bad, bad, bad VM behaviour in 2.4.10

Post by Marcelo Tosatt » Thu, 27 Sep 2001 07:50:10



> Marcelo Tosatti escribi:

> > Juan,

> > It is a known problem which we are looking into.

> > I need some information which may help confirm a guess of mine:

> > Do you have swap available ?
> Yes, the /dev/hda6 partition, that is 257000 KB in size.

> > If so, there was available anonymous memory to be swapped out ?

> Anonymous memory? Sorry, but I do not understand this question. Could
> you redo it?

By anynomymous memory I mean memory which is not disk cache (ie is not
data which is going to be written to the filesystem): Program data.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

 
 
 

1. VM: 2.4.10 vs. 2.4.10-ac2 and qsort()

Just rip it off NetBSD and there you go. (FreeBSD's breaks on machines
like SPARC, NetBSD's does not.)

http://www.de.freebsd.org/cgi/cvsweb.cgi/basesrc/lib/libc/stdlib/merg...
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

2. caldera preview?

3. Bad interaction of SO_BINDTODEVICE and ARP in Linux 2.4.10

4. Searching Local Documents and displaying results in HTML

5. RH 7.1 + Kernel 2.4.10 bad tcp-flags

6. (subject: mips libc & no fpu)

7. 014 Bad Bad Bad !!! for Linux

8. TELNET Program

9. Bad driver...Bad bad driver

10. Bad, bad, bad error...

11. SoftwarBuyLine.com is bad, bad, bad...

12. vm rewrite ready [Re: broken VM in 2.4.10-pre9]

13. Bad.. Win95... Bad, Bad, Bad... WHACK!!!