From: Ravikiran G Thirumalai <kiran@in.ibm.com>
To: akpm@osdl.org
Cc: linux-kernel@vger.kernel.org
Subject: [Patch] Remove percpufication of in_flight counter in diskstats
Date: Tue, 2 Sep 2003 16:24:11 +0530 [thread overview]
Message-ID: <20030902105409.GC1272@llm08.in.ibm.com> (raw)
Hi Andrew,
The routine disk_round_stats showed up considerably under oprofile
for high disk io load (four processes doing dd to the same disk
(different partitions) on a 4 way). This is because the counter
in_flight which is per-cpu right now gets read every time disk_round_stats
gets called. Per cpu counters like disk statistics improve write speed,
but reads are slow (since all cpus' local counter values have to be
read and summed up).
Considering the fact that in_flight counter is modified post disk_round_stats
(which reads the in_flight counter) it is better not to per-cpu this counter.
Following patch does just that. Below is the profile comparison before
and after the change. This was on a 4 way PIII Xeon, 1G ram, 2.6.0-test4-mm2.
Please apply.
Thanks,
Kiran
Before:
c010aa60 2910109 92.2249 poll_idle
c0275340 23208 0.73549 __copy_to_user_ll
c02753b0 11191 0.354657 __copy_from_user_ll
c0114aa0 7168 0.227163 mark_offset_tsc
c011ad10 6767 0.214455 schedule
c011a2b0 6741 0.213631 load_balance
c0138890 6710 0.212648 __generic_file_aio_write_nolock
c011d302 4683 0.14841 .text.lock.sched
c02e4b50 4533 0.143656 ahc_linux_isr
c029cec0 3582 0.113518 disk_round_stats
c0119b40 3509 0.111205 try_to_wake_up
c029d320 3306 0.104771 __make_request
c01567d0 3300 0.104581 __block_write_full_page
c0156c00 3299 0.104549 __block_prepare_write
After:
c010aa60 2777940 92.1302 poll_idle
c0275340 23479 0.778679 __copy_to_user_ll
c02753b0 10943 0.362924 __copy_from_user_ll
c0114aa0 7022 0.232884 mark_offset_tsc
c0138890 6988 0.231757 __generic_file_aio_write_nolock
c011ad10 6607 0.219121 schedule
c011d302 5771 0.191395 .text.lock.sched
c02e4a60 4458 0.147849 ahc_linux_isr
c011a2b0 3921 0.13004 load_balance
c01567d0 3569 0.118366 __block_write_full_page
c029d2a0 3540 0.117404 __make_request
...
c029ceb0 311 0.0103143 disk_round_stats
c011d5b0 299 0.00991631 remove_wait_queue
diff -ruN -X dontdiff linux-2.6.0-test4/drivers/block/genhd.c diskstat-mod-2.6.0-test4/drivers/block/genhd.c
--- linux-2.6.0-test4/drivers/block/genhd.c 2003-08-23 05:25:38.000000000 +0530
+++ diskstat-mod-2.6.0-test4/drivers/block/genhd.c 2003-09-01 21:36:43.000000000 +0530
@@ -372,7 +372,7 @@
disk_stat_read(disk, write_merges),
(unsigned long long)disk_stat_read(disk, write_sectors),
jiffies_to_msec(disk_stat_read(disk, write_ticks)),
- disk_stat_read(disk, in_flight),
+ disk->in_flight,
jiffies_to_msec(disk_stat_read(disk, io_ticks)),
jiffies_to_msec(disk_stat_read(disk, time_in_queue)));
}
@@ -492,7 +492,7 @@
disk_stat_read(gp, writes), disk_stat_read(gp, write_merges),
(unsigned long long)disk_stat_read(gp, write_sectors),
jiffies_to_msec(disk_stat_read(gp, write_ticks)),
- disk_stat_read(gp, in_flight),
+ gp->in_flight,
jiffies_to_msec(disk_stat_read(gp, io_ticks)),
jiffies_to_msec(disk_stat_read(gp, time_in_queue)));
diff -ruN -X dontdiff linux-2.6.0-test4/drivers/block/ll_rw_blk.c diskstat-mod-2.6.0-test4/drivers/block/ll_rw_blk.c
--- linux-2.6.0-test4/drivers/block/ll_rw_blk.c 2003-08-23 05:22:13.000000000 +0530
+++ diskstat-mod-2.6.0-test4/drivers/block/ll_rw_blk.c 2003-09-01 21:51:30.000000000 +0530
@@ -1652,7 +1652,7 @@
}
if (new_io) {
disk_round_stats(rq->rq_disk);
- disk_stat_inc(rq->rq_disk, in_flight);
+ rq->rq_disk->in_flight++;
}
}
@@ -1693,10 +1693,10 @@
unsigned long now = jiffies;
disk_stat_add(disk, time_in_queue,
- disk_stat_read(disk, in_flight) * (now - disk->stamp));
+ disk->in_flight * (now - disk->stamp));
disk->stamp = now;
- if (disk_stat_read(disk, in_flight))
+ if (disk->in_flight)
disk_stat_add(disk, io_ticks, (now - disk->stamp_idle));
disk->stamp_idle = now;
}
@@ -1808,7 +1808,7 @@
if (req->rq_disk) {
disk_round_stats(req->rq_disk);
- disk_stat_dec(req->rq_disk, in_flight);
+ req->rq_disk->in_flight--;
}
__blk_put_request(q, next);
@@ -2470,7 +2470,7 @@
break;
}
disk_round_stats(disk);
- disk_stat_dec(disk, in_flight);
+ disk->in_flight--;
}
__blk_put_request(req->q, req);
/* Do this LAST! The structure may be freed immediately afterwards */
diff -ruN -X dontdiff linux-2.6.0-test4/include/linux/genhd.h diskstat-mod-2.6.0-test4/include/linux/genhd.h
--- linux-2.6.0-test4/include/linux/genhd.h 2003-08-23 05:22:28.000000000 +0530
+++ diskstat-mod-2.6.0-test4/include/linux/genhd.h 2003-09-01 14:45:50.000000000 +0530
@@ -75,7 +75,6 @@
unsigned read_merges, write_merges;
unsigned read_ticks, write_ticks;
unsigned io_ticks;
- int in_flight;
unsigned time_in_queue;
};
@@ -101,6 +100,7 @@
unsigned sync_io; /* RAID */
unsigned long stamp, stamp_idle;
+ int in_flight;
#ifdef CONFIG_SMP
struct disk_stats *dkstats;
#else
reply other threads:[~2003-09-02 10:47 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20030902105409.GC1272@llm08.in.ibm.com \
--to=kiran@in.ibm.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.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
all inboxes | Powered by JetHome®