From: drivers@0xabadba.be
To: "Fans of the OS Plan 9 from Bell Labs" <9fans@9fans.net>
Subject: Re: [9fans] "Blocks" in C
Date: Fri, 18 Sep 2009 15:50:27 +0000 [thread overview]
Message-ID: <1934194204-1253289082-cardhu_decombobulator_blackberry.rim.net-410526281-@bda790.bisx.prod.on.blackberry> (raw)
Is there some method of lock profiling on plan9? For example when I do work on freebsd and say remove a giant lock from the keyboard subsystem; I run the lock profiler before and after the change to see how long the system was sitting at the in kernel locks.
I am doing my thesis (undergrad) on these kind of topics and am interested in seeing how I may be of use.
=jt
------Original Message------
From: erik quanstrom
Sender: 9fans-bounces@9fans.net
To: 9fans@9fans.net
ReplyTo: Fans of the OS Plan 9 from Bell Labs
Subject: Re: [9fans] "Blocks" in C
Sent: Sep 18, 2009 07:41
> it isn't really a `big kernel lock' in the linux sense.
you're right, technically it is a very different problem.
the effects seem similar. the lock is just in the block allocator
rather than the syscall interface. if you're doing a lot of i/o
in a standard kernel, there's a lot of block allocation.
> with error checking. an alternative allocator that uses quickfit on top of
> a variant of the simple allocator in kernighan and ritchie is
> roughly 30 times faster on some malloc-thrashing tests (i'm just remembering that
> from tests i did last year, so i might have the number wrong, but it's substantial).
i see the same thing. with the allocator in a specific packet-pushing
test, i see 185000 packets/s, with a block pool i see ~450000 packets/s.
all the packets are the same size.
would you be willing to share your allocator? this may help a number
of cases i'm currently fighting. 10gbe and 2×10gbe + quad processors
have a way of putting a sharp point on an otherwise dull problem.
i also wonder if an ilock is really the right protection
mechanism on machines with 8 or 16 logical processors
for malloc.
the reason i mention this is that it turns out that the lock
on block pools can be hot enough that splitting the lock
per controller can have a big positive effect. i need to set
up a faster target, but i've seen significant jumps in bandwidth
for 2 82563 controllers when using 2 pools instead of one
even on a fairly bored machine.
sorry for the lack of real numbers.
- erik
=jt
next reply other threads:[~2009-09-18 15:50 UTC|newest]
Thread overview: 117+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-09-18 15:50 drivers [this message]
2009-09-18 15:56 ` erik quanstrom
-- strict thread matches above, loose matches on Subject: below --
2009-09-17 9:19 Andrew Simmons
2009-09-17 14:45 ` LiteStar numnums
2009-09-17 17:24 ` Daniel Lyons
2009-09-04 1:35 drivers
2009-09-04 1:52 ` erik quanstrom
2009-09-02 8:04 Anant Narayanan
2009-09-02 9:43 ` Francisco J Ballesteros
2009-09-02 11:00 ` Philippe Anel
2009-09-02 11:13 ` Charles Forsyth
2009-09-03 9:52 ` Greg Comeau
2009-09-02 11:40 ` Eris Discordia
2009-09-02 12:32 ` Uriel
2009-09-02 14:20 ` Devon H. O'Dell
2009-09-03 9:52 ` Greg Comeau
2009-09-03 10:48 ` Akshat Kumar
2009-09-03 11:25 ` Charles Forsyth
2009-09-03 13:59 ` Greg Comeau
2009-09-03 15:15 ` Uriel
2009-09-03 15:44 ` David Leimbach
2009-09-03 16:01 ` Roman V Shaposhnik
2009-09-04 9:15 ` Greg Comeau
2009-09-04 16:49 ` Roman Shaposhnik
2009-09-03 16:02 ` erik quanstrom
2009-09-03 18:56 ` David Leimbach
2009-09-03 18:58 ` erik quanstrom
2009-09-03 19:13 ` David Leimbach
2009-09-03 19:36 ` erik quanstrom
2009-09-03 19:50 ` David Leimbach
2009-09-03 20:30 ` Iruata Souza
2009-09-06 22:35 ` Uriel
2009-09-07 0:41 ` David Leimbach
2009-09-03 21:08 ` Roman V Shaposhnik
2009-09-03 21:35 ` erik quanstrom
2009-09-03 21:45 ` David Leimbach
2009-09-03 21:49 ` David Leimbach
2009-09-03 21:51 ` David Leimbach
2009-09-03 22:36 ` Roman V Shaposhnik
2009-09-04 9:16 ` Greg Comeau
2009-09-03 22:10 ` Daniel Lyons
2009-09-03 22:07 ` Daniel Lyons
2009-09-04 9:15 ` Greg Comeau
2009-09-04 3:21 ` Anant Narayanan
2009-09-04 3:52 ` erik quanstrom
2009-09-04 4:18 ` David Leimbach
2009-09-04 4:44 ` erik quanstrom
2009-09-04 5:31 ` David Leimbach
2009-09-04 5:35 ` David Leimbach
2009-09-04 7:11 ` Bakul Shah
2009-09-04 7:47 ` David Leimbach
2009-09-04 14:41 ` Bakul Shah
2009-09-04 15:04 ` David Leimbach
2009-09-04 15:58 ` Bakul Shah
2009-09-04 12:14 ` erik quanstrom
2009-09-04 13:52 ` David Leimbach
2009-09-04 13:59 ` matt
2009-09-04 14:20 ` erik quanstrom
2009-09-04 14:56 ` David Leimbach
2009-09-06 23:03 ` Uriel
2009-09-07 0:45 ` David Leimbach
2009-09-04 16:50 ` Roman Shaposhnik
2009-09-04 6:45 ` Bakul Shah
2009-09-04 16:44 ` Roman Shaposhnik
2009-09-04 16:58 ` Iruata Souza
2009-09-04 17:09 ` Roman Shaposhnik
2009-09-04 17:42 ` erik quanstrom
2009-09-04 18:34 ` erik quanstrom
2009-09-04 21:54 ` Roman V Shaposhnik
2009-09-07 9:06 ` Greg Comeau
2009-09-07 9:05 ` Greg Comeau
2009-09-07 9:40 ` Uriel
2009-09-08 15:31 ` David Leimbach
2009-09-08 15:44 ` Uriel
2009-09-08 16:40 ` Bakul Shah
2009-09-11 18:15 ` Iruata Souza
2009-09-11 18:46 ` David Leimbach
2009-09-11 20:36 ` Roman V Shaposhnik
2009-09-11 21:28 ` David Leimbach
2009-09-14 16:07 ` Lawrence E. Bakst
2009-09-12 11:08 ` Anant Narayanan
2009-09-13 19:03 ` David Leimbach
2009-09-17 1:54 ` Lawrence E. Bakst
2009-09-17 8:43 ` Charles Forsyth
2009-09-17 9:12 ` Daniel Lyons
2009-09-17 15:56 ` David Leimbach
2009-09-17 17:38 ` Jack Norton
2009-09-17 20:23 ` Anant Narayanan
2009-09-17 20:26 ` erik quanstrom
2009-09-17 20:31 ` Anant Narayanan
2009-09-17 20:47 ` Akshat Kumar
2009-09-17 21:02 ` erik quanstrom
2009-09-17 22:26 ` Steve Simon
2009-09-17 22:32 ` Roman V Shaposhnik
2009-09-18 2:05 ` Daniel Lyons
2009-09-18 9:30 ` Charles Forsyth
2009-09-18 11:41 ` erik quanstrom
2009-09-17 21:03 ` Uriel
2009-09-03 9:52 ` Greg Comeau
2009-09-02 14:08 ` Rodolfo (kix)
2009-09-03 9:52 ` Greg Comeau
[not found] ` <C56117D7BD9A097270529AAE@192.168.1.2>
2009-09-02 15:07 ` David Leimbach
2009-09-02 15:26 ` Roman V Shaposhnik
2009-09-02 15:45 ` David Leimbach
2009-09-03 15:32 ` Uriel
2009-09-03 15:49 ` Roman V Shaposhnik
2009-09-03 15:54 ` erik quanstrom
2009-09-03 16:20 ` Roman V Shaposhnik
2009-09-03 16:44 ` erik quanstrom
2009-09-04 1:05 ` Roman V Shaposhnik
2009-09-04 1:30 ` Russ Cox
2009-09-04 1:32 ` erik quanstrom
2009-09-03 21:54 ` James Tomaschke
2009-09-06 22:26 ` Uriel
2009-09-04 9:04 ` Greg Comeau
2009-09-02 15:20 ` Roman V Shaposhnik
2009-09-02 17:51 ` Bakul Shah
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=1934194204-1253289082-cardhu_decombobulator_blackberry.rim.net-410526281-@bda790.bisx.prod.on.blackberry \
--to=drivers@0xabadba.be \
--cc=9fans@9fans.net \
/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).