Messages in this thread Patch in this message | | | Date | Thu, 11 Mar 2004 01:05:07 +0100 | From | Miquel van Smoorenburg <> | Subject | Re: [PATCH] backing dev unplugging |
| |
On Thu, 11 Mar 2004 00:05:42, Andrew Morton wrote: > "Miquel van Smoorenburg" <miquels@cistron.nl> wrote: > > > > > > With the latest patches from Joe it would be more like > > > > map = dm_get_table(md); > > if (map) { > > dm_table_unplug_all(map); > > dm_table_put(map); > > } > > > > No lock ranking issues, you just get a refcounted map (table, really). > > Ah, OK. Jens, you'll be needing this (on rc2-mm1): > > dm.c: protect md->map with a rw spin lock rather than the md->lock > semaphore. Also ensure that everyone accesses md->map through > dm_get_table(), rather than directly. > > 25-akpm/drivers/md/dm-table.c | 3 + > 25-akpm/drivers/md/dm.c | 88 +++++++++++++++++++++++++-----------------
.. and this final one on top of it, presumably.
See https://www.redhat.com/archives/dm-devel/2004-March/msg00036.html
dm.c: remove __dm_request (merge with previous patch). --- diff/drivers/md/dm.c 2004-03-08 15:48:05.000000000 +0000 +++ source/drivers/md/dm.c 2004-03-09 09:40:37.000000000 +0000 @@ -506,8 +506,13 @@ static void __split_bio(struct mapped_de { struct clone_info ci; - ci.md = md; ci.map = dm_get_table(md); + if (!ci.map) { + bio_io_error(bio, bio->bi_size); + return; + } + + ci.md = md; ci.bio = bio; ci.io = alloc_io(md); ci.io->error = 0; @@ -530,17 +535,6 @@ static void __split_bio(struct mapped_de * CRUD END *---------------------------------------------------------------*/ - -static inline void __dm_request(struct mapped_device *md, struct bio *bio) -{ - if (!md->map) { - bio_io_error(bio, bio->bi_size); - return; - } - - __split_bio(md, bio); -} - /* * The request function that just remaps the bio built up by * dm_merge_bvec. @@ -579,7 +573,7 @@ static int dm_request(request_queue_t *q down_read(&md->lock); } - __dm_request(md, bio); + __split_bio(md, bio); up_read(&md->lock); return 0; } @@ -591,7 +585,6 @@ static int dm_any_congested(void *conges struct dm_table *map = dm_get_table(md); if (!map || test_bit(DMF_BLOCK_IO, &md->flags)) - /* FIXME: shouldn't suspended count a congested ? */ r = bdi_bits; else r = dm_table_any_congested(map, bdi_bits); @@ -850,7 +843,7 @@ static void __flush_deferred_io(struct m while (c) { n = c->bi_next; c->bi_next = NULL; - __dm_request(md, c); + __split_bio(md, c); c = n; } }
Mike. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |