From: Oleksandr Natalenko <pfactum@gmail.com>
To: linux-kernel@vger.kernel.org
Subject: Re: [PATCH]RIFS-V3-Test For 3.4.x kernel.
Date: Sun, 08 Jul 2012 12:21:05 +0300 [thread overview]
Message-ID: <jtbje0$iiv$1@dough.gmane.org> (raw)
In-Reply-To: <CANQmPXjXt4KsYix5syDrRo_qt3DD7tMtHpdNa3Sh-fWhi53Lnw@mail.gmail.com>
The funniest thing is that desktop users started to use systemd (e.g. in
Fedora) which means they cannot even boot without cgroups. Note that :).
08.07.12 06:37, Chen написав(ла):
> I haven't made any support for Cgroup yet. After I finished
> translating the scheduler in modular form it will support Cgroup
> naturally since the modular scheduler support Cgroup.
>
> Anyway it is for desktop users and I don't think I have to support
> Cgroup in short term.
>
> I am going to post the graphical benchmark between RIFS-V3-Test and
> CFS. On my box, The latency of CFS is 8 times of RIFS
> averagely.However I am not using my computer right now so I can't post
> my benchmark now.
>
>
> On Sun, Jul 8, 2012 at 6:01 AM, Oleksandr Natalenko <pfactum@gmail.com> wrote:
>> Could you please make some chart to visually compare latencies with CFS?
>>
>> Also, has RIFS got cgroups support?
>>
>> 07.07.12 23:58, Chen написав(ла):
>>> 1. Benchmark:
>>> [admin@localhost ~]$ latt -c255 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=255
>>> Entries logged: 1020
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 106549 usec
>>> Avg 1446 usec
>>> Stdev 6182 usec
>>> Stdev mean 194 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 2793229 usec
>>> Avg 2189141 usec
>>> Stdev 351389 usec
>>> Stdev mean 11002 usec
>>> [admin@localhost ~]$ latt -c128 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=128
>>> Entries logged: 768
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 70824 usec
>>> Avg 1761 usec
>>> Stdev 5074 usec
>>> Stdev mean 183 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 1464295 usec
>>> Avg 1163262 usec
>>> Stdev 210801 usec
>>> Stdev mean 7607 usec
>>> [admin@localhost ~]$ latt -c64 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=64
>>> Entries logged: 640
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 53780 usec
>>> Avg 1375 usec
>>> Stdev 4772 usec
>>> Stdev mean 189 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 797045 usec
>>> Avg 596825 usec
>>> Stdev 111695 usec
>>> Stdev mean 4415 usec
>>> [admin@localhost ~]$ latt -c32 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=32
>>> Entries logged: 480
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 86032 usec
>>> Avg 2147 usec
>>> Stdev 7659 usec
>>> Stdev mean 350 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 374303 usec
>>> Avg 309004 usec
>>> Stdev 43155 usec
>>> Stdev mean 1970 usec
>>> [admin@localhost ~]$ latt -c16 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=16
>>> Entries logged: 320
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 41166 usec
>>> Avg 1150 usec
>>> Stdev 4706 usec
>>> Stdev mean 263 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 178917 usec
>>> Avg 155367 usec
>>> Stdev 16074 usec
>>> Stdev mean 899 usec
>>> [admin@localhost ~]$ latt -c8 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=8
>>> Entries logged: 184
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 20256 usec
>>> Avg 585 usec
>>> Stdev 2306 usec
>>> Stdev mean 170 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 88262 usec
>>> Avg 75957 usec
>>> Stdev 7102 usec
>>> Stdev mean 524 usec
>>> [admin@localhost ~]$ latt -c4 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=4
>>> Entries logged: 104
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 7950 usec
>>> Avg 663 usec
>>> Stdev 1719 usec
>>> Stdev mean 169 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 50647 usec
>>> Avg 38685 usec
>>> Stdev 4053 usec
>>> Stdev mean 397 usec
>>> [admin@localhost ~]$ latt -c2 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=2
>>> Entries logged: 54
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 33 usec
>>> Avg 9 usec
>>> Stdev 5 usec
>>> Stdev mean 1 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 21700 usec
>>> Avg 20590 usec
>>> Stdev 258 usec
>>> Stdev mean 35 usec
>>> [admin@localhost ~]$ latt -c1 sleep 10
>>>
>>> Parameters: min_wait=100ms, max_wait=500ms, clients=1
>>> Entries logged: 27
>>>
>>> Wakeup averages
>>> -------------------------------------
>>> Max 22 usec
>>> Avg 9 usec
>>> Stdev 3 usec
>>> Stdev mean 1 usec
>>>
>>> Work averages
>>> -------------------------------------
>>> Max 20614 usec
>>> Avg 20162 usec
>>> Stdev 125 usec
>>> Stdev mean 24 usec
>>>
>>>
>>>
>>> RIFS-V3 is the new name of RIFS-ES. It looks like CFS. but with RIFS,
>>> the latency is much lower.
>>>
>>
>>
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
>> Please read the FAQ at http://www.tux.org/lkml/
next prev parent reply other threads:[~2012-07-08 9:21 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-07-07 20:58 [PATCH]RIFS-V3-Test For 3.4.x kernel Chen
2012-07-07 22:01 ` Oleksandr Natalenko
2012-07-08 3:37 ` Chen
2012-07-08 9:21 ` Oleksandr Natalenko [this message]
2012-07-08 12:16 ` Chen
2012-07-08 12:23 ` richard -rw- weinberger
[not found] ` <CANQmPXiMhahyT5BQAaw5tSX9fzJUp6mqjHz7FgEy=9nLqy2-9A@mail.gmail.com>
2012-07-08 13:32 ` Chen
2012-07-08 14:14 ` Oleksandr Natalenko
2012-07-09 12:37 ` Hillf Danton
2012-07-09 13:07 ` richard -rw- weinberger
2012-07-09 13:18 ` Hillf Danton
2012-07-09 15:52 ` Jan Ceuleers
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='jtbje0$iiv$1@dough.gmane.org' \
--to=pfactum@gmail.com \
--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 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.