Re: Linux 2.6.10-ac9

[Sami Farin]

On Sat, Jan 15, 2005 at 12:32:59AM +0000, Alan Cox wrote:
> On Gwe, 2005-01-14 at 03:01, Sami Farin wrote:
> > at this spot I have no /dev/dsp etc.
> > then I reload snd_ens1371:
> 
> That sounds like the new udev rather than kernel side. The changes from
> ac8 to ac9 are tiny on the audio side and don't involve driver setup
> stuff.
>
> > also, with 2.6.10 I can't disable write cache...
> > I could do it in 2.6.9.
> 
> Works for me in 2.6.10-ac. Are there any diagnostics on dmesg when you
> try and turn the cache off ?

no, nothing.

how do I know now is hdparm -I or -W lying?
(-I says it's enabled and -W0 says it was disabled just fine.)

my drives do not support cache flushes, I guess your drives do?

drivers/ide/ide-disk.c:write_cache()
if (!ide_id_has_flush_cache(drive->id))
  return 1;

cat /proc/ide/hda/settings does not work, either

cat           D C0572788     0  8016   5187                6044 (NOTLB)
cdc97eb0 00000046 cf7f00a0 c0572788 00005699 c01fe93f 000001ad 00000023 
       01078345 00005699 cf7f00a0 00000bbb 02ed12c3 00005699 cfec51d8 c04de020 
       cfec5080 00000246 cdc97ee8 c0409764 c04de028 00000001 cfec5080 c011ab70 
Call Trace:
 [<c0409764>] __down+0x64/0xc0
 [<c04098ba>] __down_failed+0xa/0x10
 [<c0302acf>] .text.lock.ide_proc+0x8b/0x1fc
 [<c018c594>] proc_file_read+0xc4/0x260
 [<c0158fbf>] vfs_read+0xcf/0x150
 [<c01592db>] sys_read+0x4b/0x80
 [<c0103163>] syscall_call+0x7/0xb

-- 
-
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/