Repository navigation
Conversation
|
Upstream branch: 72d3fcf |
c6537fe to
eebb5bc
Compare
|
Upstream branch: 551c722 |
e9dc365 to
1e1a43e
Compare
eebb5bc to
bbd3af0
Compare
|
Upstream branch: ce1e022 |
1e1a43e to
f8c2578
Compare
bbd3af0 to
a9c0b46
Compare
|
Upstream branch: e767a4e |
f8c2578 to
efc90f4
Compare
a9c0b46 to
72d0f5e
Compare
|
Upstream branch: a74306e |
efc90f4 to
5b6a9ec
Compare
72d0f5e to
d980ad5
Compare
|
Upstream branch: None |
5b6a9ec to
b6da652
Compare
d980ad5 to
83b99cc
Compare
|
Upstream branch: 22430ae |
b6da652 to
bcc8e10
Compare
83b99cc to
9081535
Compare
|
Upstream branch: 69f80fe |
bcc8e10 to
570669c
Compare
9081535 to
6118c5c
Compare
|
Upstream branch: 7b63ef2 |
|
Upstream branch: c03c1a1 |
1a19b9b to
4caba3f
Compare
90ea7e4 to
435b194
Compare
|
Upstream branch: 7be4bc5 |
4caba3f to
296372d
Compare
435b194 to
fd67522
Compare
|
Upstream branch: 17bb0e2 |
296372d to
ba4277f
Compare
fd67522 to
3b53a63
Compare
|
Upstream branch: 21a7f44 |
ba4277f to
ed46d96
Compare
3b53a63 to
5d84576
Compare
|
Upstream branch: 6b4b543 |
ed46d96 to
a5ab4cb
Compare
5d84576 to
6582ed2
Compare
|
Upstream branch: fb2c3f4 |
a5ab4cb to
42b421d
Compare
6582ed2 to
b3a1abc
Compare
|
Upstream branch: d135ab4 |
42b421d to
913deb3
Compare
b3a1abc to
712496f
Compare
|
Upstream branch: c339f40 |
913deb3 to
c436e7b
Compare
712496f to
a186880
Compare
elevator_set_default() runs from blk_register_queue() while the disk is still being added: the block device is not visible yet and uevents are suppressed, so no file system I/O can be issued. elevator_change() freezes the queue and cancels the dispatch work before switching, which is enough to drain dispatch activity from passthrough requests, as the comment in elevator_change() already states. elevator_switch() still quiesces the queue unconditionally, though, so every disk added with a default elevator waits for an RCU grace period. Commit 245a489 ("block: avoid to quiesce queue in elevator_init_mq") had removed this wait from the default elevator setup, and it came back when that setup was folded into elevator_change(). The wait is paid in full by callers that add disks one at a time. ublk adds the disk of a single-queue device from START_DEV, which takes 7.7ms at p50 in a 16 vCPU KVM guest, most of it in blk_mq_quiesce_queue() called from elevator_set_default(). Skip quiescing in elevator_switch() when setting the default elevator. ublk null target, 1000 single-queue devices created one by one: before after START_DEV p50 7.74ms 0.68ms START_DEV p99 14.61ms 1.32ms devices/s 114 971 Fixes: 1e44bed ("block: unifying elevator change") Cc: stable@vger.kernel.org Signed-off-by: Qiliang Yuan <odys.yuan@gmail.com> Reviewed-by: Ming Lei <tom.leiming@gmail.com> Reviewed-by: Nilay Shroff <nilay@linux.ibm.com>
|
Upstream branch: d89e1d1 |
c436e7b to
f93116f
Compare
|
At least one diff in series https://patchwork.kernel.org/series/1177359/ irrelevant now for [{'archived': False, 'project': 241}] search patterns |
Pull request for series with
subject: block: don't quiesce the queue when setting the default elevator
version: 1
url: https://patchwork.kernel.org/series/1177359/