block: use nanosecond resolution for iostat
Klaus Kusche reported that the I/O busy time in /proc/diskstats was not
updating properly on 4.18. This is because we started using ktime to
track elapsed time, and we convert nanoseconds to jiffies when we update
the partition counter. However, this gets rounded down, so any I/Os that
take less than a jiffy are not accounted for. Previously in this case,
the value of jiffies would sometimes increment while we were doing I/O,
so at least some I/Os were accounted for.
Let's convert the stats to use nanoseconds internally. We still report
milliseconds as before, now more accurately than ever. The value is
still truncated to 32 bits for backwards compatibility.
Fixes: 522a777566f5 ("block: consolidate struct request timestamp fields")
Cc: stable@vger.kernel.org
Reported-by: Klaus Kusche <klaus.kusche@computerix.info>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
diff --git a/block/blk-core.c b/block/blk-core.c
index 4dbc93f..cff0a60 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -2733,17 +2733,15 @@ void blk_account_io_done(struct request *req, u64 now)
* containing request is enough.
*/
if (blk_do_io_stat(req) && !(req->rq_flags & RQF_FLUSH_SEQ)) {
- unsigned long duration;
const int sgrp = op_stat_group(req_op(req));
struct hd_struct *part;
int cpu;
- duration = nsecs_to_jiffies(now - req->start_time_ns);
cpu = part_stat_lock();
part = req->part;
part_stat_inc(cpu, part, ios[sgrp]);
- part_stat_add(cpu, part, ticks[sgrp], duration);
+ part_stat_add(cpu, part, nsecs[sgrp], now - req->start_time_ns);
part_round_stats(req->q, cpu, part);
part_dec_in_flight(req->q, part, rq_data_dir(req));