From: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
To: axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org
Cc: jack-AlSwsSmVLrQ@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
kernel-team-b10kYP2dOMg@public.gmane.org,
vgoyal-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org,
Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
Arianna Avanzini
<avanzini.arianna-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Subject: [PATCH 8/9] cfq-iosched: fold cfq_find_alloc_queue() into cfq_get_queue()
Date: Tue, 7 Jul 2015 11:36:00 -0400 [thread overview]
Message-ID: <1436283361-3889-9-git-send-email-tj@kernel.org> (raw)
In-Reply-To: <1436283361-3889-1-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
cfq_find_alloc_queue() checks whether a queue actually needs to be
allocated, which is unnecessary as its sole caller, cfq_get_queue(),
only calls it if so. Also, the oom queue fallback logic is scattered
between cfq_get_queue() and cfq_find_alloc_queue(). There really
isn't much going on in the latter and things can be made simpler by
folding it into cfq_get_queue().
This patch collapses cfq_find_alloc_queue() into cfq_get_queue(). The
change is fairly straight-forward with one exception - async_cfqq is
now initialized to NULL and the "!is_sync" test in the last if
conditional is replaced with "async_cfqq" test. This is because gcc
(5.1.1) gets confused for some reason and warns that async_cfqq may be
used uninitialized otherwise. Oh well, the code isn't necessarily
worse this way.
This patch doesn't cause any functional difference.
v2: Updated to reflect GFP_ATOMIC -> GPF_NOWAIT.
Signed-off-by: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Reviewed-by: Jeff Moyer <jmoyer-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Vivek Goyal <vgoyal-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Arianna Avanzini <avanzini.arianna-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
---
block/cfq-iosched.c | 49 +++++++++++++++----------------------------------
1 file changed, 15 insertions(+), 34 deletions(-)
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index f3ea8a1..fa2a2f7 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -3640,33 +3640,6 @@ static void check_blkcg_changed(struct cfq_io_cq *cic, struct bio *bio)
static inline void check_blkcg_changed(struct cfq_io_cq *cic, struct bio *bio) { }
#endif /* CONFIG_CFQ_GROUP_IOSCHED */
-static struct cfq_queue *
-cfq_find_alloc_queue(struct cfq_data *cfqd, struct cfq_group *cfqg, bool is_sync,
- struct cfq_io_cq *cic, struct bio *bio)
-{
- struct cfq_queue *cfqq;
-
- cfqq = cic_to_cfqq(cic, is_sync);
-
- /*
- * Always try a new alloc if we fell back to the OOM cfqq
- * originally, since it should just be a temporary situation.
- */
- if (!cfqq || cfqq == &cfqd->oom_cfqq) {
- cfqq = kmem_cache_alloc_node(cfq_pool,
- GFP_NOWAIT | __GFP_ZERO,
- cfqd->queue->node);
- if (cfqq) {
- cfq_init_cfqq(cfqd, cfqq, current->pid, is_sync);
- cfq_init_prio_data(cfqq, cic);
- cfq_link_cfqq_cfqg(cfqq, cfqg);
- cfq_log_cfqq(cfqd, cfqq, "alloced");
- } else
- cfqq = &cfqd->oom_cfqq;
- }
- return cfqq;
-}
-
static struct cfq_queue **
cfq_async_queue_prio(struct cfq_data *cfqd, int ioprio_class, int ioprio)
{
@@ -3691,7 +3664,7 @@ cfq_get_queue(struct cfq_data *cfqd, bool is_sync, struct cfq_io_cq *cic,
{
int ioprio_class = IOPRIO_PRIO_CLASS(cic->ioprio);
int ioprio = IOPRIO_PRIO_DATA(cic->ioprio);
- struct cfq_queue **async_cfqq;
+ struct cfq_queue **async_cfqq = NULL;
struct cfq_queue *cfqq;
struct cfq_group *cfqg;
@@ -3714,12 +3687,20 @@ cfq_get_queue(struct cfq_data *cfqd, bool is_sync, struct cfq_io_cq *cic,
goto out;
}
- cfqq = cfq_find_alloc_queue(cfqd, cfqg, is_sync, cic, bio);
+ cfqq = kmem_cache_alloc_node(cfq_pool, GFP_NOWAIT | __GFP_ZERO,
+ cfqd->queue->node);
+ if (!cfqq) {
+ cfqq = &cfqd->oom_cfqq;
+ goto out;
+ }
+
+ cfq_init_cfqq(cfqd, cfqq, current->pid, is_sync);
+ cfq_init_prio_data(cfqq, cic);
+ cfq_link_cfqq_cfqg(cfqq, cfqg);
+ cfq_log_cfqq(cfqd, cfqq, "alloced");
- /*
- * pin the queue now that it's allocated, scheduler exit will prune it
- */
- if (!is_sync && cfqq != &cfqd->oom_cfqq) {
+ if (async_cfqq) {
+ /* a new async queue is created, pin and remember */
cfqq->ref++;
*async_cfqq = cfqq;
}
@@ -4469,7 +4450,7 @@ static int cfq_init_queue(struct request_queue *q, struct elevator_type *e)
cfqd->prio_trees[i] = RB_ROOT;
/*
- * Our fallback cfqq if cfq_find_alloc_queue() runs into OOM issues.
+ * Our fallback cfqq if cfq_get_queue() runs into OOM issues.
* Grab a permanent reference to it, so that the normal code flow
* will not attempt to free it. oom_cfqq is linked to root_group
* but shouldn't hold a reference as it'll never be unlinked. Lose
--
2.4.3
next prev parent reply other threads:[~2015-07-07 15:36 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-07-07 15:35 [PATCHSET v2 block/for-4.3] block, cgroup: make cfq charge async IOs to the appropriate blkcgs Tejun Heo
2015-07-07 15:35 ` [PATCH 2/9] cfq-iosched: fix async oom queue handling Tejun Heo
2015-07-07 15:35 ` [PATCH 3/9] cfq-iosched: fix oom cfq_queue ref leak in cfq_set_request() Tejun Heo
[not found] ` <1436283361-3889-1-git-send-email-tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
2015-07-07 15:35 ` [PATCH 1/9] cfq-iosched: simplify control flow in cfq_get_queue() Tejun Heo
2015-07-07 15:35 ` [PATCH 4/9] cfq-iosched: minor cleanups Tejun Heo
2015-07-07 15:36 ` Tejun Heo [this message]
2015-07-07 15:36 ` [PATCH 9/9] cfq-iosched: charge async IOs to the appropriate blkcg's instead of the root Tejun Heo
2015-07-07 15:35 ` [PATCH 5/9] blkcg, cfq-iosched: use GFP_NOWAIT instead of GFP_ATOMIC for non-critical allocations Tejun Heo
2015-07-07 15:35 ` [PATCH 6/9] cfq-iosched: remove @gfp_mask from cfq_find_alloc_queue() Tejun Heo
2015-07-07 15:35 ` [PATCH 7/9] cfq-iosched: move cfq_group determination from cfq_find_alloc_queue() to cfq_get_queue() Tejun Heo
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1436283361-3889-9-git-send-email-tj@kernel.org \
--to=tj-dgejt+ai2ygdnm+yrofe0a@public.gmane.org \
--cc=avanzini.arianna-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=axboe-tSWWG44O7X1aa/9Udqfwiw@public.gmane.org \
--cc=cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=jack-AlSwsSmVLrQ@public.gmane.org \
--cc=kernel-team-b10kYP2dOMg@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=vgoyal-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).