Alan wrote: >> Would it be possible for the kernel to realize that the drive stopped >> responding? And make sure that it doesn't make the whole computer >> unresponsive... > > libata should eventually get rid of the drive. The old ide isn't quite so > smart. > I see the same happening with a non-udma capable cdrw drive since 2.6.19-rc1. 2.6.18.x was ok. Unfortunately libata was not a solution (at least in 2.6.19). It took the hard drive sharing the bus with the mwdma cdrw down from UDMA100 to PIO0 the entire chain and finally offline in steps each time it detected trouble on the bus. I suspect the genirq changes (I'm on x86_64) which went in at that point to be the cause. The ide irq's are edge triggered, and changes to the fragile timing there might have made it miss. I solved this by replacing the drive against an UDMA capable on (but still have the old one around for testing, if need be). In 2.6.19 (after trying libata on it) I saw errors of command time outs etc. 2.6.20 was even worse as it failed silently. Andy - To unsubscribe from this list: send the line "unsubscribe linux-ide" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
