dm raid1: handle log failure (fc1ff958) · Commits · 戴 / test

When writing to a mirror, the log must be updated first. Failure to update the log could result in the log not properly reflecting the state of the mirror if the machine should crash. We change the return type of the rh_flush function to give us the ability to check if a log write was successful. If the log write was unsuccessful, we fail the writes to avoid the case where the log does not properly reflect the state of the mirror. A follow-up patch - which is dependent on the ability to requeue I/O's to core device-mapper - will requeue the I/O's for retry (allowing the mirror to be reconfigured.) Signed-off-by:

Jonathan Brassow <jbrassow@redhat.com> Signed-off-by:

Alasdair G Kergon <agk@redhat.com> Signed-off-by:

Linus Torvalds <torvalds@linux-foundation.org>

drivers/md/dm-raid1.c

+8 −4

Original line number	Diff line number	Diff line
		@@ -134,6 +134,7 @@ struct mirror_set {
		/* recovery */
		region_t nr_regions;
		int in_sync;
		int log_failure;

		struct mirror default_mirror; / Default mirror */

		@@ -589,9 +590,9 @@ static void rh_recovery_end(struct region *reg, int success)
		wake(rh->ms);
		}

		static void rh_flush(struct region_hash *rh)
		static int rh_flush(struct region_hash *rh)
		{
		rh->log->type->flush(rh->log);
		return rh->log->type->flush(rh->log);
		}

		static void rh_delay(struct region_hash rh, struct bio bio)
		@@ -892,12 +893,15 @@ static void do_writes(struct mirror_set ms, struct bio_list writes)
		*/
		rh_inc_pending(&ms->rh, &sync);
		rh_inc_pending(&ms->rh, &nosync);
		rh_flush(&ms->rh);
		ms->log_failure = rh_flush(&ms->rh) ? 1 : 0;

		/*
		* Dispatch io.
		*/
		if (unlikely(ms->log_failure))
		while ((bio = bio_list_pop(&sync)))
		bio_endio(bio, bio->bi_size, -EIO);
		else while ((bio = bio_list_pop(&sync)))
		do_write(ms, bio);

		while ((bio = bio_list_pop(&recover)))

Admin message