also sprach Antony Gelberg <antony@antgel.co.uk> [2004.04.12.1611 +0200]:
> Try turning APIC off. It's in the archives - nForce chipsets have
> issues with APIC.
Per your request, I compiled a new kernel without APIC/IOAPIC
(realising only afterwards that I could have used kernel
parameters...).
However, the result is still the same:
embryo:~# memtest all [224]
memtest v. 2.93.1
(C) 2000 Charles Cazabon <memtest@discworld.dyndns.org>
Original v.1 (C) 1999 Simon Kirby <sim@stormix.com> <sim@neato.org>
Current limits:
RLIMIT_RSS 0xffffffff
RLIMIT_VMEM 0xffffffff
Raising limits...
Unable to malloc 4293918720 bytes.
[...]
Allocated 753926144 bytes...trying mlock...
--> hard lockup
Yesterday, during random checking from within a 2.4.25-1-k7 boot,
I found 3 bad blocks in the swap space. So I turned the swap space
off. Without swap, the memtest in a custom 2.6.5 kernel looks like
so:
embryo:~# memtest all | less
memtest v. 2.93.1
(C) 2000 Charles Cazabon <memtest@discworld.dyndns.org>
Original v.1 (C) 1999 Simon Kirby <sim@stormix.com> <sim@neato.org>
Current limits:
RLIMIT_RSS 0xffffffff
RLIMIT_VMEM 0xffffffff
Raising limits...
Unable to malloc 4293918720 bytes.
[...]
Unable to malloc 246415360 bytes.
Allocated 245366784 bytes...trying mlock...
Received signal 15 (Terminated)
munlock'ed memory.
4294967295 runs completed. 0 errors detected. Total runtime: 1081804768 seconds.
Exiting...
And kern.log will report this:
Out of Memory: Killed process 562 (memtest).
memtest: page allocation failure. order:0, mode:0xd2
Call Trace:
[<c012dc6b>] __alloc_pages+0x2eb/0x330
[<c013638e>] do_anonymous_page+0x6e/0x1a0
[<c0136521>] do_no_page+0x61/0x2c0
[<c0136955>] handle_mm_fault+0xd5/0x140
[<c01355fa>] get_user_pages+0xaa/0x2b0
[<c0136a5c>] make_pages_present+0x8c/0xb0
[<c0136f37>] mlock_fixup+0xb7/0xd0
[<c0137022>] do_mlock+0xd2/0x100
[<c01370eb>] sys_mlock+0x9b/0xb0
[<c0106c5f>] syscall_call+0x7/0xb
Furthermore, the three bad blocks are gone from within 2.6.5
(custom).
If I boot back into 2.4.25-1-k7, the three bad blocks are at the
same spot as they were yesterday. But they only exist with 2.4.25.
Can anyone help me make sense of this?
--
Please do not CC me when replying to lists; I read them!
.''`. martin f. krafft <madduck@debian.org>
: :' : proud Debian developer, admin, and user
`. `'`
`- Debian - when you have better things to do than fixing a system
Invalid/expired PGP subkeys? Use subkeys.pgp.net as keyserver!
Attachment:
signature.asc
Description: Digital signature