Re: 2.6.7-mm7

From: Andrew Morton
Date: Sat Jul 10 2004 - 15:29:32 EST


Marcin Gibu__a <mg@xxxxxxxx> wrote:
>
> On Friday 09 July 2004 08:50, Andrew Morton wrote:
> > ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.7/2.6.7-m
> >m7/
>
> Hi,
> I've encountered a process locked up in D state while using this kernel. The
> process is mpg123 playing mp3 from ntfs partition. I'm using cfq elevator.
>
> Alt-SysRq-T reveals that:
>
> mpg123 D 00000D33 0 18012 18592 (NOTLB)
> d8c11e28 00200086 37c2b380 00000d33 00000000 37c2b380 00000d33 c8761190
> c8761338 c1253760 c14011e8 00000275 d8c11e30 c02e9c27 00000275 c0127522
> c1253760 00000000 00000001 c8761190 c01272e1 d8c11e6c d8c11e6c c153d480
> Call Trace:
> [<c02e9c27>] io_schedule+0xe/0x16
> [<c0127522>] __lock_page+0xa7/0xba
> [<c01272e1>] page_wake_function+0x0/0x36
> [<c01272e1>] page_wake_function+0x0/0x36
> [<c0127966>] do_generic_mapping_read+0x15d/0x2f4
> [<c0127d52>] __generic_file_aio_read+0x198/0x1b6
> [<c0127afd>] file_read_actor+0x0/0xbd
> [<c0127e1f>] generic_file_read+0x68/0x7c
> [<c0260014>] snd_pcm_oss_write1+0x131/0x165
> [<c013e27c>] vfs_read+0x9d/0xc9
> [<c013e436>] sys_read+0x2c/0x42
> [<c01039e1>] sysenter_past_esp+0x52/0x71

Is it repeatable at all? If so, does using a different I/O scheduler fix
it up?
-
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/