From: Arnd Bergmann <arnd@arndb.de>
To: Andrei Warkentin <andreiw@motorola.com>
Cc: linux-arm-kernel@lists.infradead.org,
Linus Walleij <linus.walleij@linaro.org>,
linux-mmc@vger.kernel.org
Subject: Re: MMC quirks relating to performance/lifetime.
Date: Sat, 19 Feb 2011 12:20:54 +0100 [thread overview]
Message-ID: <201102191220.54815.arnd@arndb.de> (raw)
In-Reply-To: <AANLkTimWAnPg2sYiLbfZXAAUiu7P5T3LmNoVbPi9RTEG@mail.gmail.com>
On Saturday 19 February 2011 00:17:51 Andrei Warkentin wrote:
> # echo 0 > /sys/block/mmcblk0/device/page_size
> # ./flashbench -A -b 1024 /dev/block/mmcblk0p9
> write align 8388608 pre 3.59ms on 6.54ms post 3.65ms diff 2.92ms
> write align 4194304 pre 4.13ms on 7.37ms post 4.27ms diff 3.17ms
> write align 2097152 pre 3.62ms on 6.81ms post 3.94ms diff 3.03ms
> write align 1048576 pre 3.62ms on 6.53ms post 3.55ms diff 2.95ms
> write align 524288 pre 3.62ms on 6.51ms post 3.63ms diff 2.88ms
> write align 262144 pre 3.62ms on 6.51ms post 3.63ms diff 2.89ms
> write align 131072 pre 3.62ms on 6.5ms post 3.63ms diff 2.88ms
> write align 65536 pre 3.61ms on 6.49ms post 3.62ms diff 2.88ms
> write align 32768 pre 3.61ms on 6.49ms post 3.61ms diff 2.88ms
> write align 16384 pre 3.68ms on 107ms post 3.51ms diff 103ms
> write align 8192 pre 3.74ms on 121ms post 3.91ms diff 117ms
> write align 4096 pre 3.88ms on 3.87ms post 3.87ms diff -2937ns
> write align 2048 pre 3.89ms on 3.88ms post 3.88ms diff -8734ns
> # fjnh84@fjnh84-desktop:~/src/n/src/flash$ adb -s 17006185428011d7 shell
> # echo 8192 > /sys/block/mmcblk0/device/page_size
> # cd data
> # ./flashbench -A -b 1024 /dev/block/mmcblk0p9
> write align 8388608 pre 3.33ms on 6.8ms post 3.65ms diff 3.31ms
> write align 4194304 pre 4.34ms on 8.14ms post 4.53ms diff 3.71ms
> write align 2097152 pre 3.64ms on 7.31ms post 4.09ms diff 3.44ms
> write align 1048576 pre 3.65ms on 7.52ms post 3.65ms diff 3.87ms
> write align 524288 pre 3.62ms on 6.8ms post 3.63ms diff 3.17ms
> write align 262144 pre 3.62ms on 6.84ms post 3.63ms diff 3.22ms
> write align 131072 pre 3.62ms on 6.85ms post 3.44ms diff 3.32ms
> write align 65536 pre 3.39ms on 6.8ms post 3.66ms diff 3.28ms
> write align 32768 pre 3.64ms on 6.86ms post 3.66ms diff 3.21ms
> write align 16384 pre 3.67ms on 6.86ms post 3.65ms diff 3.2ms
> write align 8192 pre 3.66ms on 6.84ms post 3.64ms diff 3.19ms
> write align 4096 pre 3.71ms on 3.71ms post 3.64ms diff 38.6µs
> write align 2048 pre 3.71ms on 3.71ms post 3.72ms diff -656ns
>
> This was with the split unaligned accesses patch... Which I am
> attaching for comments.
I agree, this is very fascinating behavior. 100ms second latency for a
single 2KB access is definitely something we should try to avoid, and I
wonder why the drive decides to do that. It must get into a state where
it requires an extra garbage collection (you mentioned that earlier).
The numbers you see here are taken over multiple runs. Do you see a lot
of fluctuation when doing this with --count=1?
Also, does the same happen with other blocksizes, e.g. 4096 or 8192, passed
to flashbench?
Arnd
WARNING: multiple messages have this Message-ID (diff)
From: arnd@arndb.de (Arnd Bergmann)
To: linux-arm-kernel@lists.infradead.org
Subject: MMC quirks relating to performance/lifetime.
Date: Sat, 19 Feb 2011 12:20:54 +0100 [thread overview]
Message-ID: <201102191220.54815.arnd@arndb.de> (raw)
In-Reply-To: <AANLkTimWAnPg2sYiLbfZXAAUiu7P5T3LmNoVbPi9RTEG@mail.gmail.com>
On Saturday 19 February 2011 00:17:51 Andrei Warkentin wrote:
> # echo 0 > /sys/block/mmcblk0/device/page_size
> # ./flashbench -A -b 1024 /dev/block/mmcblk0p9
> write align 8388608 pre 3.59ms on 6.54ms post 3.65ms diff 2.92ms
> write align 4194304 pre 4.13ms on 7.37ms post 4.27ms diff 3.17ms
> write align 2097152 pre 3.62ms on 6.81ms post 3.94ms diff 3.03ms
> write align 1048576 pre 3.62ms on 6.53ms post 3.55ms diff 2.95ms
> write align 524288 pre 3.62ms on 6.51ms post 3.63ms diff 2.88ms
> write align 262144 pre 3.62ms on 6.51ms post 3.63ms diff 2.89ms
> write align 131072 pre 3.62ms on 6.5ms post 3.63ms diff 2.88ms
> write align 65536 pre 3.61ms on 6.49ms post 3.62ms diff 2.88ms
> write align 32768 pre 3.61ms on 6.49ms post 3.61ms diff 2.88ms
> write align 16384 pre 3.68ms on 107ms post 3.51ms diff 103ms
> write align 8192 pre 3.74ms on 121ms post 3.91ms diff 117ms
> write align 4096 pre 3.88ms on 3.87ms post 3.87ms diff -2937ns
> write align 2048 pre 3.89ms on 3.88ms post 3.88ms diff -8734ns
> # fjnh84 at fjnh84-desktop:~/src/n/src/flash$ adb -s 17006185428011d7 shell
> # echo 8192 > /sys/block/mmcblk0/device/page_size
> # cd data
> # ./flashbench -A -b 1024 /dev/block/mmcblk0p9
> write align 8388608 pre 3.33ms on 6.8ms post 3.65ms diff 3.31ms
> write align 4194304 pre 4.34ms on 8.14ms post 4.53ms diff 3.71ms
> write align 2097152 pre 3.64ms on 7.31ms post 4.09ms diff 3.44ms
> write align 1048576 pre 3.65ms on 7.52ms post 3.65ms diff 3.87ms
> write align 524288 pre 3.62ms on 6.8ms post 3.63ms diff 3.17ms
> write align 262144 pre 3.62ms on 6.84ms post 3.63ms diff 3.22ms
> write align 131072 pre 3.62ms on 6.85ms post 3.44ms diff 3.32ms
> write align 65536 pre 3.39ms on 6.8ms post 3.66ms diff 3.28ms
> write align 32768 pre 3.64ms on 6.86ms post 3.66ms diff 3.21ms
> write align 16384 pre 3.67ms on 6.86ms post 3.65ms diff 3.2ms
> write align 8192 pre 3.66ms on 6.84ms post 3.64ms diff 3.19ms
> write align 4096 pre 3.71ms on 3.71ms post 3.64ms diff 38.6?s
> write align 2048 pre 3.71ms on 3.71ms post 3.72ms diff -656ns
>
> This was with the split unaligned accesses patch... Which I am
> attaching for comments.
I agree, this is very fascinating behavior. 100ms second latency for a
single 2KB access is definitely something we should try to avoid, and I
wonder why the drive decides to do that. It must get into a state where
it requires an extra garbage collection (you mentioned that earlier).
The numbers you see here are taken over multiple runs. Do you see a lot
of fluctuation when doing this with --count=1?
Also, does the same happen with other blocksizes, e.g. 4096 or 8192, passed
to flashbench?
Arnd
next prev parent reply other threads:[~2011-02-19 11:20 UTC|newest]
Thread overview: 117+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-02-08 21:22 MMC quirks relating to performance/lifetime Andrei Warkentin
2011-02-08 21:38 ` Wolfram Sang
2011-02-08 21:38 ` Wolfram Sang
2011-02-08 22:42 ` Russell King - ARM Linux
2011-02-09 8:37 ` Linus Walleij
2011-02-09 8:37 ` Linus Walleij
2011-02-09 9:13 ` Arnd Bergmann
2011-02-09 9:13 ` Arnd Bergmann
2011-02-11 22:33 ` Andrei Warkentin
2011-02-11 22:33 ` Andrei Warkentin
2011-02-12 17:05 ` Arnd Bergmann
2011-02-12 17:05 ` Arnd Bergmann
2011-02-12 17:33 ` Andrei Warkentin
2011-02-12 17:33 ` Andrei Warkentin
2011-02-12 18:22 ` Arnd Bergmann
2011-02-12 18:22 ` Arnd Bergmann
2011-02-18 1:10 ` Andrei Warkentin
2011-02-18 1:10 ` Andrei Warkentin
2011-02-18 13:44 ` Arnd Bergmann
2011-02-18 13:44 ` Arnd Bergmann
2011-02-18 19:47 ` Andrei Warkentin
2011-02-18 19:47 ` Andrei Warkentin
2011-02-18 22:40 ` Andrei Warkentin
2011-02-18 22:40 ` Andrei Warkentin
2011-02-18 23:17 ` Andrei Warkentin
2011-02-18 23:17 ` Andrei Warkentin
2011-02-19 11:20 ` Arnd Bergmann [this message]
2011-02-19 11:20 ` Arnd Bergmann
2011-02-20 5:56 ` Andrei Warkentin
2011-02-20 5:56 ` Andrei Warkentin
2011-02-20 15:23 ` Arnd Bergmann
2011-02-20 15:23 ` Arnd Bergmann
2011-02-22 7:05 ` Andrei Warkentin
2011-02-22 7:05 ` Andrei Warkentin
2011-02-22 16:49 ` Arnd Bergmann
2011-02-22 16:49 ` Arnd Bergmann
2011-02-19 9:54 ` Arnd Bergmann
2011-02-19 9:54 ` Arnd Bergmann
2011-02-20 4:39 ` Andrei Warkentin
2011-02-20 4:39 ` Andrei Warkentin
2011-02-20 15:03 ` Arnd Bergmann
2011-02-20 15:03 ` Arnd Bergmann
2011-02-22 6:42 ` Andrei Warkentin
2011-02-22 6:42 ` Andrei Warkentin
2011-02-22 16:42 ` Arnd Bergmann
2011-02-22 16:42 ` Arnd Bergmann
2011-02-11 23:23 ` Linus Walleij
2011-02-11 23:23 ` Linus Walleij
2011-02-12 10:45 ` Arnd Bergmann
2011-02-12 10:45 ` Arnd Bergmann
2011-02-12 10:59 ` Russell King - ARM Linux
2011-02-12 10:59 ` Russell King - ARM Linux
2011-02-12 16:28 ` Arnd Bergmann
2011-02-12 16:28 ` Arnd Bergmann
2011-02-12 16:37 ` Russell King - ARM Linux
2011-02-12 16:37 ` Russell King - ARM Linux
2011-02-11 22:27 ` Andrei Warkentin
2011-02-11 22:27 ` Andrei Warkentin
2011-02-12 18:37 ` Arnd Bergmann
2011-02-12 18:37 ` Arnd Bergmann
2011-02-13 0:10 ` Andrei Warkentin
2011-02-13 0:10 ` Andrei Warkentin
2011-02-13 17:39 ` Arnd Bergmann
2011-02-13 17:39 ` Arnd Bergmann
2011-02-14 19:29 ` Andrei Warkentin
2011-02-14 19:29 ` Andrei Warkentin
2011-02-14 20:22 ` Arnd Bergmann
2011-02-14 20:22 ` Arnd Bergmann
2011-02-14 22:25 ` Andrei Warkentin
2011-02-14 22:25 ` Andrei Warkentin
2011-02-15 17:16 ` Arnd Bergmann
2011-02-15 17:16 ` Arnd Bergmann
2011-02-17 2:08 ` Andrei Warkentin
2011-02-17 2:08 ` Andrei Warkentin
2011-02-17 15:47 ` Arnd Bergmann
2011-02-17 15:47 ` Arnd Bergmann
2011-02-20 11:27 ` Andrei Warkentin
2011-02-20 11:27 ` Andrei Warkentin
2011-02-20 14:39 ` Arnd Bergmann
2011-02-20 14:39 ` Arnd Bergmann
2011-02-22 7:46 ` Andrei Warkentin
2011-02-22 7:46 ` Andrei Warkentin
2011-02-22 17:00 ` Arnd Bergmann
2011-02-22 17:00 ` Arnd Bergmann
2011-02-23 10:19 ` Andrei Warkentin
2011-02-23 10:19 ` Andrei Warkentin
2011-02-23 16:09 ` Arnd Bergmann
2011-02-23 16:09 ` Arnd Bergmann
2011-02-23 22:26 ` Andrei Warkentin
2011-02-23 22:26 ` Andrei Warkentin
2011-02-24 9:24 ` Arnd Bergmann
2011-02-24 9:24 ` Arnd Bergmann
2011-02-25 11:02 ` Andrei Warkentin
2011-02-25 11:02 ` Andrei Warkentin
2011-02-25 12:21 ` Arnd Bergmann
2011-02-25 12:21 ` Arnd Bergmann
2011-03-01 18:48 ` Jens Axboe
2011-03-01 18:48 ` Jens Axboe
2011-03-01 19:11 ` Arnd Bergmann
2011-03-01 19:11 ` Arnd Bergmann
2011-03-01 19:15 ` Jens Axboe
2011-03-01 19:15 ` Jens Axboe
2011-03-01 19:51 ` Arnd Bergmann
2011-03-01 19:51 ` Arnd Bergmann
2011-03-01 21:33 ` Andrei Warkentin
2011-03-01 21:33 ` Andrei Warkentin
2011-03-02 10:34 ` Andrei Warkentin
2011-03-02 10:34 ` Andrei Warkentin
2011-03-05 9:23 ` Andrei Warkentin
2011-03-05 9:23 ` Andrei Warkentin
2011-02-11 14:41 ` Pavel Machek
2011-02-11 14:51 ` Arnd Bergmann
2011-02-11 15:20 ` Lei Wen
2011-02-11 15:25 ` Arnd Bergmann
2011-03-08 6:59 ` Pavel Machek
2011-03-08 14:03 ` Arnd Bergmann
2011-03-08 14:03 ` Arnd Bergmann
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=201102191220.54815.arnd@arndb.de \
--to=arnd@arndb.de \
--cc=andreiw@motorola.com \
--cc=linus.walleij@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-mmc@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.