[PATCH] blk-throttle: only update last_low_overflow_time when LIMIT_MAX

[Jianchao Wang]

Currently, last_low_overflow_time will be updated whenever bios
are throttled and queued. This is true when LIMIT_MAX, but not
for LIMIT_LOW. what last_low_overflow_time indicates is dispatching
not submitting. When LIMIT_LOW, the dispatch bps/iops should never
be above the low limit, we should not update last_low_overflow_time
at the moment. Otherwise, it will be hard to upgrade limit_index
when there is always bios to be submitted.

Signed-off-by: Jianchao Wang <jianchao.w.w...@oracle.com>
---
 block/blk-throttle.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index 851aa16..b5ba845 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -2214,7 +2214,8 @@ bool blk_throtl_bio(struct request_queue *q, struct 
blkcg_gq *blkg,
 
                /* if above limits, break to queue */
                if (!tg_may_dispatch(tg, bio, NULL)) {
-                       tg->last_low_overflow_time[rw] = jiffies;
+                       if (td->limit_index == LIMIT_MAX)
+                               tg->last_low_overflow_time[rw] = jiffies;
                        if (throtl_can_upgrade(td, tg)) {
                                throtl_upgrade_state(td);
                                goto again;
@@ -2258,7 +2259,8 @@ bool blk_throtl_bio(struct request_queue *q, struct 
blkcg_gq *blkg,
                   tg->io_disp[rw], tg_iops_limit(tg, rw),
                   sq->nr_queued[READ], sq->nr_queued[WRITE]);
 
-       tg->last_low_overflow_time[rw] = jiffies;
+       if (td->limit_index == LIMIT_MAX)
+               tg->last_low_overflow_time[rw] = jiffies;
 
        td->nr_queued[rw]++;
        throtl_add_bio_tg(bio, qn, tg);
-- 
2.7.4