| Message ID | 80fca3d5-7c21-bce1-11db-2010a459a774@fb.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Sep 15, 2016 at 03:16:49PM -0600, Jens Axboe wrote: > We currently account a '0' dispatch, and anything above that still falls > below the range set by BLK_MQ_MAX_DISPATCH_ORDER. If we dispatch more, > we don't account it. > > Change the last bucket to be inclusive of anything above the range we > track, and have the sysfs file reflect that by including a '+' in the > output: > > $ cat /sys/block/nvme0n1/mq/0/dispatched > 0 1006 > 1 20229 > 2 1 > 4 0 > 8 0 > 16 0 > 32+ 0 > > Signed-off-by: Jens Axboe <axboe@fb.com> Reviewed-by: Omar Sandoval <osandov@fb.com>
diff --git a/block/blk-mq-sysfs.c b/block/blk-mq-sysfs.c index 885a4a59acd9..39840c37b7ce 100644 --- a/block/blk-mq-sysfs.c +++ b/block/blk-mq-sysfs.c @@ -208,12 +208,14 @@ static ssize_t blk_mq_hw_sysfs_dispatched_show(struct blk_mq_hw_ctx *hctx, page += sprintf(page, "%8u\t%lu\n", 0U, hctx->dispatched[0]); - for (i = 1; i < BLK_MQ_MAX_DISPATCH_ORDER; i++) { - unsigned long d = 1U << (i - 1); + for (i = 1; i < BLK_MQ_MAX_DISPATCH_ORDER - 1; i++) { + unsigned int d = 1U << (i - 1); - page += sprintf(page, "%8lu\t%lu\n", d, hctx->dispatched[i]); + page += sprintf(page, "%8u\t%lu\n", d, hctx->dispatched[i]); } + page += sprintf(page, "%8u+\t%lu\n", 1U << (i - 1), + hctx->dispatched[i]); return page - start_page; } diff --git a/block/blk-mq.c b/block/blk-mq.c index 4b8034e54478..fb4b6795ec57 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -802,6 +802,14 @@ static void flush_busy_ctxs(struct blk_mq_hw_ctx *hctx, struct list_head *list) } } +static inline unsigned int queued_to_index(unsigned int queued) +{ + if (!queued) + return 0; + + return min(BLK_MQ_MAX_DISPATCH_ORDER - 1, ilog2(queued) + 1); +} + /* * Run this hardware queue, pulling any software queues mapped to it in.
We currently account a '0' dispatch, and anything above that still falls below the range set by BLK_MQ_MAX_DISPATCH_ORDER. If we dispatch more, we don't account it. Change the last bucket to be inclusive of anything above the range we track, and have the sysfs file reflect that by including a '+' in the output: $ cat /sys/block/nvme0n1/mq/0/dispatched 0 1006 1 20229 2 1 4 0 8 0 16 0 32+ 0 Signed-off-by: Jens Axboe <axboe@fb.com> * Note that this function currently has various problems around ordering @@ -890,10 +898,7 @@ static void __blk_mq_run_hw_queue(struct blk_mq_hw_ctx *hctx) dptr = &driver_list; } - if (!queued) - hctx->dispatched[0]++; - else if (queued < (1 << (BLK_MQ_MAX_DISPATCH_ORDER - 1))) - hctx->dispatched[ilog2(queued) + 1]++; + hctx->dispatched[queued_to_index(queued)]++; /* * Any items that need requeuing? Stuff them into hctx->dispatch,