AWT
10-08-2006, 07:07 PM
I think that my 2 drive TiVo has a bad sector. Occasionally it will lockup during play and reboot after a short period.
I could reproduce the problem at the exact same point in the recording and I've managed to undelete the problematic recording with a (reluctant) view to keeping it occupying the bad sector(s). TBH though, I'd rather fix the problem and delete the recording.
Here's the log:
Oct 8 22:05:00 (none) kernel: mediaswitch: returning 0 from standin tune after tuning to ch 0 with adjust -1
Oct 8 22:05:05 (none) kernel: IDE: calling handler with dma_running, altstat=0x51
Oct 8 22:05:05 (none) kernel: IDE: read command for sector 4480064 bailed with DMA running
Oct 8 22:05:05 (none) kernel: hda: unexpected_intr: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:05 (none) kernel: hda: unexpected_intr: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:05 (none) kernel: Stopping immediate on Ide
Oct 8 22:05:05 (none) kernel: hda: tivo_dma_intr_direct: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:05 (none) kernel: hda: tivo_dma_intr_direct: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:06 (none) kernel: hda: tivo_dma_intr_direct: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:06 (none) kernel: hda: tivo_dma_intr_direct: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:16 (none) kernel: Stopping immediate on Ide
Oct 8 22:05:16 (none) kernel: IDE: tivo DMA engine aborted
Oct 8 22:05:16 (none) kernel: hdb: irq timeout: status=0x58 { DriveReady SeekComplete DataRequest }
Oct 8 22:05:16 (none) kernel: hdb: irq timeout: error=0x00 { }, secCnt=0, LBAsect=133699136
Oct 8 22:05:16 (none) kernel: IDE: DMA in a strange state
Oct 8 22:05:16 (none) kernel: FPGA callback with DMA not running - delayed interrupt?
The drive is about 1yr old and I'm loathe to replace it 'just to be safe'. Assuming it's simply a bad sector, why hasn't the drive's electronics silently remapped it to one of the good spares? Perhaps the spare sectors are all used up?
Is there a way that I can mark that 'bad' sector as bad at the shell prompt and forget about this? Should I pull the drive and run Spinrite on it (I'd rather not unplug the TiVo if I can help it)?
I could reproduce the problem at the exact same point in the recording and I've managed to undelete the problematic recording with a (reluctant) view to keeping it occupying the bad sector(s). TBH though, I'd rather fix the problem and delete the recording.
Here's the log:
Oct 8 22:05:00 (none) kernel: mediaswitch: returning 0 from standin tune after tuning to ch 0 with adjust -1
Oct 8 22:05:05 (none) kernel: IDE: calling handler with dma_running, altstat=0x51
Oct 8 22:05:05 (none) kernel: IDE: read command for sector 4480064 bailed with DMA running
Oct 8 22:05:05 (none) kernel: hda: unexpected_intr: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:05 (none) kernel: hda: unexpected_intr: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:05 (none) kernel: Stopping immediate on Ide
Oct 8 22:05:05 (none) kernel: hda: tivo_dma_intr_direct: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:05 (none) kernel: hda: tivo_dma_intr_direct: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:06 (none) kernel: hda: tivo_dma_intr_direct: status=0x51 { DriveReady SeekComplete Error }
Oct 8 22:05:06 (none) kernel: hda: tivo_dma_intr_direct: error=0x40 { UncorrectableError }, secCnt=45, LBAsect=4480275
Oct 8 22:05:16 (none) kernel: Stopping immediate on Ide
Oct 8 22:05:16 (none) kernel: IDE: tivo DMA engine aborted
Oct 8 22:05:16 (none) kernel: hdb: irq timeout: status=0x58 { DriveReady SeekComplete DataRequest }
Oct 8 22:05:16 (none) kernel: hdb: irq timeout: error=0x00 { }, secCnt=0, LBAsect=133699136
Oct 8 22:05:16 (none) kernel: IDE: DMA in a strange state
Oct 8 22:05:16 (none) kernel: FPGA callback with DMA not running - delayed interrupt?
The drive is about 1yr old and I'm loathe to replace it 'just to be safe'. Assuming it's simply a bad sector, why hasn't the drive's electronics silently remapped it to one of the good spares? Perhaps the spare sectors are all used up?
Is there a way that I can mark that 'bad' sector as bad at the shell prompt and forget about this? Should I pull the drive and run Spinrite on it (I'd rather not unplug the TiVo if I can help it)?