* [ext4] 71d4f7d0321: -49.6% xfstests.generic.274.seconds
@ 2014-08-21 14:30 Fengguang Wu
2014-08-22 1:52 ` Fengguang Wu
0 siblings, 1 reply; 2+ messages in thread
From: Fengguang Wu @ 2014-08-21 14:30 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: ext4, LKML, lkp
[-- Attachment #1: Type: text/plain, Size: 2034 bytes --]
Hi Ted,
We noticed increased xfstests 274's test speed and the first good
commit is 71d4f7d032149b935a26eb3ff85c6c837f3714e1 ("ext4: remove
metadata reservation checks").
test case: snb-drag/xfstests/4HDD-ext4-generic-slow2
snb-drag is a Sandy Bridge PC with 6G memory.
d5e03cbb0c88cd1 71d4f7d032149b935a26eb3ff
--------------- -------------------------
51 ± 2% -49.6% 25 ± 1% TOTAL xfstests.generic.274.seconds
817 ± 1% -3.0% 792 ± 1% TOTAL time.elapsed_time
xfstests.generic.274.seconds
60 ++---------------------------------------------------------------------+
| |
55 ++ .*..*.. *.. *.. |
| .*. . .. .. |
50 *+.*..*. *..*..*..* *..*..*..*..* * |
| |
45 ++ |
| |
40 ++ |
| |
35 ++ |
| |
30 ++ |
| |
25 O+-O--O--O--O--O---O--O--O--O--O--O--O--O--O--O--O--O---O--O--O--O--O--O
[*] bisect-good sample
[O] bisect-bad sample
Disclaimer:
Results have been estimated based on internal Intel analysis and are provided
for informational purposes only. Any difference in system hardware or software
design or configuration may affect actual performance.
Thanks,
Fengguang
[-- Attachment #2: reproduce --]
[-- Type: text/plain, Size: 733 bytes --]
echo performance > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
mkfs -t ext4 -q /dev/sda7
mkfs -t ext4 -q /dev/sda10
mkfs -t ext4 -q /dev/sda6
mkfs -t ext4 -q /dev/sda5
mount -t ext4 /dev/sda10 /fs/sda10
mount -t ext4 /dev/sda5 /fs/sda5
mount -t ext4 /dev/sda6 /fs/sda6
mount -t ext4 /dev/sda7 /fs/sda7
export TEST_DIR=/fs/sda10
export TEST_DEV=/dev/sda10
export SCRATCH_MNT=/fs/scratch
mkdir /fs/scratch -p
export SCRATCH_DEV=/dev/sda7
./check generic/224 generic/231 generic/274 generic/299 generic/300
^ permalink raw reply [flat|nested] 2+ messages in thread* Re: [ext4] 71d4f7d0321: -49.6% xfstests.generic.274.seconds
2014-08-21 14:30 [ext4] 71d4f7d0321: -49.6% xfstests.generic.274.seconds Fengguang Wu
@ 2014-08-22 1:52 ` Fengguang Wu
0 siblings, 0 replies; 2+ messages in thread
From: Fengguang Wu @ 2014-08-22 1:52 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: linux-ext4, LKML, lkp
On Thu, Aug 21, 2014 at 10:30:06PM +0800, Fengguang Wu wrote:
> Hi Ted,
>
> We noticed increased xfstests 274's test speed and the first good
> commit is 71d4f7d032149b935a26eb3ff85c6c837f3714e1 ("ext4: remove
> metadata reservation checks").
>
> test case: snb-drag/xfstests/4HDD-ext4-generic-slow2
> snb-drag is a Sandy Bridge PC with 6G memory.
>
> d5e03cbb0c88cd1 71d4f7d032149b935a26eb3ff
> --------------- -------------------------
> 51 ± 2% -49.6% 25 ± 1% TOTAL xfstests.generic.274.seconds
> 817 ± 1% -3.0% 792 ± 1% TOTAL time.elapsed_time
>
> xfstests.generic.274.seconds
>
> 60 ++---------------------------------------------------------------------+
> | |
> 55 ++ .*..*.. *.. *.. |
> | .*. . .. .. |
> 50 *+.*..*. *..*..*..* *..*..*..*..* * |
> | |
> 45 ++ |
> | |
> 40 ++ |
> | |
> 35 ++ |
> | |
> 30 ++ |
> | |
> 25 O+-O--O--O--O--O---O--O--O--O--O--O--O--O--O--O--O--O---O--O--O--O--O--O
>
>
> [*] bisect-good sample
> [O] bisect-bad sample
This is sweet, got another improvement:
testbox/testcase/testparams: snb-drag/xfstests/4HDD-ext4-generic-mid
d5e03cbb0c88cd1 71d4f7d032149b935a26eb3ff
--------------- -------------------------
45 ± 4% -50.2% 22 ± 7% xfstests.generic.256.seconds
7044859 ± 2% -7.0% 6553446 ± 0% time.minor_page_faults
36156 ± 2% -7.0% 33638 ± 0% time.involuntary_context_switches
171551 ± 1% -4.8% 163312 ± 0% time.voluntary_context_switches
436 ± 1% -5.7% 411 ± 0% time.elapsed_time
time.voluntary_context_switches
176000 ++-----------------------------------------------------------------+
| * *.. *.. |
174000 ++ *.. .. : : *..* : * |
172000 ++ .. .*.* : : + : + |
*..*..* *. : : + : + .* |
170000 ++ : .* * *. |
168000 ++ *. |
| |
166000 ++ |
164000 ++ O O |
| O O O O O O O O O O O
162000 ++ O O O O O O |
160000 O+ O O |
| O O |
158000 ++-----------------------------------------------------------------+
xfstests.generic.256.seconds
60 ++-----------------------------*---------------------------------------+
| : : |
55 ++ : : |
50 ++ : : |
| .*.. .* : .*.. |
45 *+ *.. .*..*...*..*..*. : .*. .*.. .* |
| *. *. *. *. |
40 ++ |
| |
35 ++ |
30 ++ |
| |
25 ++ O O |
O O O O O O O O O O O O O |
20 ++----O--------O---------O--O--O-----O-----O-----O---------------------O
Thanks,
Fengguang
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2014-08-22 1:52 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-08-21 14:30 [ext4] 71d4f7d0321: -49.6% xfstests.generic.274.seconds Fengguang Wu
2014-08-22 1:52 ` Fengguang Wu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox