On Tue, 18 Nov 2008, James Smart wrote: > Alan, > > It was probably this change: > http://marc.info/?l=linux-scsi&m=122118359130474&w=2 > > However, I would *not* back it out. It's a silent data corruption if you > do. Past behavior reset good_bytes to indicate all data was transfered > and ignored any residual - which meant if no data was transfered, we > misinform the callee and the callee gets stuck with whatever data is in > the buffer. Current behavior, the patch, uses residual, and likely falls > into an i/o retry path. I agree, that was probably the cause, and I also agree that it should not be removed. That's part of what makes this so difficult; we don't want to just remove the original cause, we want to fix things. Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
