On Tue, Feb 25, 2020 at 08:54:07PM +0100, Daniel Glöckner wrote: > bio_reset will reset too many fields. As you can see in the context of > the diff, dm-integrity expects f.ex. the values modified by bio_advance > to stay intact and the transfer should of course use the same disk and > operation. > > How about doing the atomic_set in bio_remaining_done (in block/bio.c) > where the BIO_CHAIN flag is cleared once __bi_remaining hits zero? > Or is requeuing a bio without bio_reset really a no-go? In that case a > one-liner won't do... That tends to add a overhead to the fast path for a rather exotic case. I'm having a bit of a hard time understanding the dm-integrity code due to it's annoyingly obsfucated code, but it seems like it tries to submit a bio again after it came out of a ->end_io handler. That might have some other problems, but if we only want to paper over the remaining count a isngle call to bio_inc_remaining might be all you need. -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
