hda errors on iBook G4

Hi,

I noticed hda status errors in my iBook G4's dmesg quite a long time
ago, but never looked any more... Today I just ran tests with hdparm and
this looks like really reproduceable:

Results are different whether I try it only or more, see below.

Any hint about where I should look at, and what kind of bug may cause
this error?

# hdparm -X69 /dev/hda
 
/dev/hda:
 setting xfermode to 69 (UltraDMA mode5)
# dmesg -c
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command

# hdparm -X68 /dev/hda
 
/dev/hda:
 setting xfermode to 68 (UltraDMA mode4)
# dmesg -c
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command
hda: status timeout: status=0xd0 { Busy }
 
ide: failed opcode was: unknown
hda: DMA disabled
hda: drive not ready for command
ide0: reset: success
# hdparm -X68 /dev/hda
 
/dev/hda:
 setting xfermode to 68 (UltraDMA mode4)
[hangs for about 30 seconds]
# dmesg -c
hda: lost interrupt
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command
hda: lost interrupt
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command
hda: lost interrupt
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command
hda: lost interrupt
hda: status error: status=0x58 { DriveReady SeekComplete DataRequest }
 
ide: failed opcode was: unknown
hda: drive not ready for command
ide0: reset: success

-- 
Colin

On 22 Oct 2004 at 17h10, Colin Leroy wrote:

Hi, 

I noticed hda status errors in my iBook G4's dmesg quite a long time
ago, but never looked any more... Today I just ran tests with hdparm
and this looks like really reproduceable:

I found one strange thing in drivers/ide/ppc/pmac.c:

pmac_ide_do_setfeature(ide_drive_t *drive, u8 command) {...}

pmac_ide_udma_enable(ide_drive_t *drive, u16 mode)
{...
	ret = pmac_ide_do_setfeature(drive, mode);
...}

We really want to pass u16 to the function taking an u8 arg?
(just a wild shot, it's probably wanted, but i thought i'd share).

-- 
Colin

On Sat, 2004-10-23 at 01:30, Colin Leroy wrote:

Hi,

I noticed hda status errors in my iBook G4's dmesg quite a long time
ago, but never looked any more... Today I just ran tests with hdparm and
this looks like really reproduceable:

Results are different whether I try it only or more, see below.

Any hint about where I should look at, and what kind of bug may cause
this error?

Just stop messing with the transfer speeds with hdparm ;) This is
bad, the ide-pmac driver will automatically pick the optimal speed
for your drive.

Ben.

On Sat, 2004-10-23 at 01:59, Colin Leroy wrote:

On 22 Oct 2004 at 17h10, Colin Leroy wrote:

Hi, 

quoted

I noticed hda status errors in my iBook G4's dmesg quite a long time
ago, but never looked any more... Today I just ran tests with hdparm
and this looks like really reproduceable:

I found one strange thing in drivers/ide/ppc/pmac.c:

pmac_ide_do_setfeature(ide_drive_t *drive, u8 command) {...}

pmac_ide_udma_enable(ide_drive_t *drive, u16 mode)
{...
	ret = pmac_ide_do_setfeature(drive, mode);
...}

We really want to pass u16 to the function taking an u8 arg?
(just a wild shot, it's probably wanted, but i thought i'd share).

It's a thinko, but it's always a u8 anyway

Ben.