ibook g4 hard drive failures

[Paul Brossier <piem@altern.org>]

hello,

i have had more and more problems in the last two months with the hard disk of
my ibook g4 (PowerBook6,5). it all started one morning, while reading emails,
after a few days uptime, heavy disc activity the day before. the system was
being more and more unresponsive. trying to quit mutt, it told me 'read only
filesystem'. dmesg replied 'Input/Output error'.

after struggling for a week with this weird disk to get the few bits of data
that i didn't backup, the machine started working again as normal for 7 or 10
days. then the problem came back.

since, i had no filesystem corruptions, but never got the disk behaving normaly
for more than 4 hours.  turning off dma does not seem to help much, but it
seems the temperature of the machine changes the time it takes to reproduce.
if i wait longer before rebooting it, it takes longer to lock up.  if insist on
trying to reboot it, the hard drive is not found at boot time.

so far, it looked to me like a hardware issue, but i found this thread:
http://lists.debian.org/debian-powerpc/2005/08/msg00182.html and i have this
line at boot time: 'hda: Enabling Ultra DMA 5' (shouldn't it be a 2?)

so, hardware or software? is there anything i could try before sending it for
repair?

cheers, piem

$ sudo hdparm /dev/hda
ide-pmac lost interrupt, dma status: 8480
hda: lost int
hda: dma_intr: status=0xd0 { Busy }

ide: failed opcode was: unknown
hda: DMA disabled
ide0: reset timed-out, status=0xd0
hda: status timeout: status=0xd0 { Busy }

ide: failed opcode was: unknown
hda: drive not ready for command
ide0: reset timed-out, status=0xd0
end_request: I/O error, dev hda, sector 38799338
end_request: I/O error, dev hda, sector 3937810
EXT3-fs error (device hda3): ext3_get_inode_loc: unable to read inode block - inode=260133, block=491947
Remounting filesystem read-only
end_request: I/O error, dev hda, sector 2018
Buffer I/O error on device hda3, logical block 0
lost page write due to I/O error on hda3
-bash: sudo: command not found
$ EXT3-fs error (device hda3): read_inode_bitmap: Cannot read inode bitmap - block_group = 148, inodde_bitmap = 4849665
EXT3-fs error (device hda3) in ext3_new_inode: IO Failure
end_request: I/O error, dev hda, sector 24680010
end_request: I/O error, dev hda, sector 27824842
end_request: I/O error, dev hda, sector 28624498
Aborting journal on device hda3.
__journal_remove_journal_head: freeing b_commited_data
__journal_remove_journal_head: freeing b_commited_data
__journal_remove_journal_head: freeing b_commited_data
end_request: I/O error, dev hda, sector 2018
end_request: I/O error, dev hda, sector 2026
end_request: I/O error, dev hda, sector 2034
end_request: I/O error, dev hda, sector 3935010
end_request: I/O error, dev hda, sector 3935394
end_request: I/O error, dev hda, sector 3936202
end_request: I/O error, dev hda, sector 3936.... 
  .... ...
  .... (many more like this)
  .... ...
end_request: I/O error, dev hda, sector 40224810
end_request: I/O error, dev hda, sector 3936202
EXT3-fs error (device hda3): ext3_readdir: directory #1999090 contains a hole at offset 0