On Sat, Apr 16, 2016 at 2:55 PM, Christoph Hellwig <hch@xxxxxx> wrote: > Or rather nice async discard support :) I have patches pending for XFS > and for the upcoming NVMe target driver to submit discard requests > without waiting for them in caller. This series has one patch to > massively clean up the routines in blk-lib.c by using chained bios, > and a second one to export the guts of blkdev_issue_discard so that > callers can chain them even more (e.g. for range discards from NVMe > DSM commands, or from the XFS transation commit code) and submit them > with the callers own end_io handler instead of waiting for them. Do you have a pointer to either the NVMe or XFS code that is making use of this new interface? I've been carrying __blkdev_issue_discard_async() in drivers/md/dm-thin.c for a while and will clean that up now. But that dm-thin.c code is similar but different from your __blkdev_issue_discard(). I'd just like to make sure dm-thin.c falls in line as it should. For instance I need to get rid of the bio_inc_remaining() duplication that is dm-thin.c:__bio_inc_remaining(). Thanks, Mike -- To unsubscribe from this list: send the line "unsubscribe linux-block" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
