On Thu, Jul 13, 2017 at 03:09:54PM +0100, [email protected] wrote: > From: Filipe Manana <[email protected]> > > The recent changes to make bio cloning faster (added in the 4.13 merge > window) by using the bio_clone_fast() API introduced a regression on > raid5/6 modes, because cloned bios have an invalid bi_vcnt field > (therefore it can not be used) and the raid5/6 code uses the > bio_for_each_segment_all() API to iterate the segments of a bio, and this > API uses a bio's bi_vcnt field.
Good catch! I'm going to add assertions to all remaining bio_for_each_segment_all calls so that cloned bios do not get accidentally iterated that way. Other raid levels seem to be fine with the usecase you've provided, I'm now running fstests with the assertions. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
