On Mon, Apr 27 2015 at 8:59pm -0400, Mike Snitzer <snitzer@xxxxxxxxxx> wrote: > From: Christoph Hellwig <hch@xxxxxx> > > Currently dm-multipath has to clone the bios for every request sent > to the lower devices, which wastes cpu cycles and ties down memory. > > This patch instead adds a new REQ_CLONE flag that instructs req_bio_endio > to not complete bios attached to a request, which we set on clone > requests similar to bios in a flush sequence. With this change I/O > errors on a path failure only get propagated to dm-multipath, which > can then either resubmit the I/O or complete the bios on the original > request. > > I've done some basic testing of this on a Linux target with ALUA support, > and it survives path failures during I/O nicely. > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> > Signed-off-by: Mike Snitzer <snitzer@xxxxxxxxxx> > --- > block/blk-core.c | 94 +++---------------------- > drivers/md/dm-table.c | 25 ++++--- > drivers/md/dm.c | 171 +++++++++++----------------------------------- > drivers/md/dm.h | 5 +- > include/linux/blk_types.h | 2 + > include/linux/blkdev.h | 6 +- > 6 files changed, 73 insertions(+), 230 deletions(-) > > v3: really just some line-wrapping, whitespace, and misc small nits Hey Jens, Given the blk-core changes, would you like to pick this up for 4.2 and I'll just base DM ontop of linux-block.git again for the 4.2 cycle? The patch is available here: https://patchwork.kernel.org/patch/6285081/ Mike p.s. I do have a followup DM patch that cleans things up further but I can just stage it via linux-dm.git: https://patchwork.kernel.org/patch/6285091/ -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
