From: Tejun Heo <tj@kernel.org>
To: Frederic Weisbecker <fweisbec@gmail.com>
Cc: torvalds@linux-foundation.org, mingo@elte.hu,
linux-kernel@vger.kernel.org, jeff@garzik.org,
akpm@linux-foundation.org, rusty@rustcorp.com.au,
cl@linux-foundation.org, dhowells@redhat.com,
arjan@linux.intel.com, oleg@redhat.com, axboe@kernel.dk,
dwalker@codeaurora.org, stefanr@s5r6.in-berlin.de,
florian@mickler.org, andi@firstfloor.org, mst@redhat.com,
randy.dunlap@oracle.com
Subject: Re: [PATCH 12/35] workqueue: update cwq alignement
Date: Tue, 29 Jun 2010 18:09:35 +0200 [thread overview]
Message-ID: <4C2A1ABF.8050504@kernel.org> (raw)
In-Reply-To: <20100629160100.GL5318@nowhere>
Hello,
On 06/29/2010 06:01 PM, Frederic Weisbecker wrote:
> So, imagine you allocate your struct with alloc_percpu(align).
>
> The per cpu pointer is 0x400 (purely imagination).
>
> Now you have two cpus and they have the following base offsets for
> per cpu allocations:
>
> CPU 0 = 0xf1000000
> CPU 1 = 0xf2000000
>
> So, the true pointers for your cpu workqueue structs will be:
>
> CPU 0 = 0xf1000400
> CPU 1 = 0xf2000400
>
> These addresses are aligned like you wanted to, and it seems it is what
> matters, to store these addresses in the work flags.
Yes.
> So why does the size of the struct need to be aligned too?
Where am I doing that?
> All you want is that the two above addresses are aligned. Now why
> the size of the struct itself needs this alignment too. That's the
> obscure point for me. If it's useless, this could avoid all this
> alignment maintainance, except during the allocation itself.
What alignment maintenance? Are you talking about the UP code? If
you're talking about the UP code, the ugliness there is because the
current UP __alloc_percpu() can't honor the alignment parameter.
Heh, it seems I'm still lost. Care to give one more shot at it? :-)
Thanks.
--
tejun
next prev parent reply other threads:[~2010-06-29 16:11 UTC|newest]
Thread overview: 98+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-06-28 21:03 [PATCHSET] workqueue: concurrency managed workqueue, take#6 Tejun Heo
2010-06-28 21:03 ` [PATCH 01/35] kthread: implement kthread_worker Tejun Heo
2010-06-28 21:03 ` [PATCH 02/35] ivtv: use kthread_worker instead of workqueue Tejun Heo
2010-07-05 17:11 ` Andy Walls
2010-07-06 7:01 ` Tejun Heo
2010-07-09 13:15 ` Andy Walls
2010-06-28 21:03 ` [PATCH 03/35] kthread: implement kthread_data() Tejun Heo
2010-06-28 21:03 ` [PATCH 04/35] acpi: use queue_work_on() instead of binding workqueue worker to cpu0 Tejun Heo
2010-06-28 21:03 ` [PATCH 05/35] workqueue: kill RT workqueue Tejun Heo
2010-06-28 21:03 ` [PATCH 06/35] workqueue: misc/cosmetic updates Tejun Heo
2010-06-28 21:03 ` [PATCH 07/35] workqueue: merge feature parameters into flags Tejun Heo
2010-06-28 21:03 ` [PATCH 08/35] workqueue: define masks for work flags and conditionalize STATIC flags Tejun Heo
2010-06-28 21:03 ` [PATCH 09/35] workqueue: separate out process_one_work() Tejun Heo
2010-06-28 21:03 ` [PATCH 10/35] workqueue: temporarily remove workqueue tracing Tejun Heo
2010-06-28 21:03 ` [PATCH 11/35] workqueue: kill cpu_populated_map Tejun Heo
2010-06-28 21:04 ` [PATCH 12/35] workqueue: update cwq alignement Tejun Heo
2010-06-28 22:47 ` Frederic Weisbecker
2010-06-29 7:39 ` Tejun Heo
2010-06-29 12:36 ` Frederic Weisbecker
2010-06-29 15:42 ` Tejun Heo
2010-06-29 15:47 ` Frederic Weisbecker
2010-06-29 15:51 ` Tejun Heo
2010-06-29 16:01 ` Frederic Weisbecker
2010-06-29 16:09 ` Tejun Heo [this message]
2010-06-29 16:17 ` Frederic Weisbecker
2010-07-06 14:22 ` Christoph Lameter
2010-07-06 14:26 ` Tejun Heo
2010-06-29 8:12 ` [PATCH UPDATED " Tejun Heo
2010-06-29 13:39 ` Frederic Weisbecker
2010-06-28 21:04 ` [PATCH 13/35] workqueue: reimplement workqueue flushing using color coded works Tejun Heo
2010-06-28 21:04 ` [PATCH 14/35] workqueue: introduce worker Tejun Heo
2010-06-28 21:04 ` [PATCH 15/35] workqueue: reimplement work flushing using linked works Tejun Heo
2010-06-28 21:04 ` [PATCH 16/35] workqueue: implement per-cwq active work limit Tejun Heo
2010-06-28 21:04 ` [PATCH 17/35] workqueue: reimplement workqueue freeze using max_active Tejun Heo
2010-06-28 21:04 ` [PATCH 18/35] workqueue: introduce global cwq and unify cwq locks Tejun Heo
2010-06-28 21:04 ` [PATCH 19/35] workqueue: implement worker states Tejun Heo
2010-06-28 21:04 ` [PATCH 20/35] workqueue: reimplement CPU hotplugging support using trustee Tejun Heo
2010-06-28 21:04 ` [PATCH 21/35] workqueue: make single thread workqueue shared worker pool friendly Tejun Heo
2010-06-28 21:04 ` [PATCH 22/35] workqueue: add find_worker_executing_work() and track current_cwq Tejun Heo
2010-06-28 21:04 ` [PATCH 23/35] workqueue: carry cpu number in work data once execution starts Tejun Heo
2010-06-28 21:04 ` [PATCH 24/35] workqueue: implement WQ_NON_REENTRANT Tejun Heo
2010-06-28 21:04 ` [PATCH 25/35] workqueue: use shared worklist and pool all workers per cpu Tejun Heo
2010-06-28 21:04 ` [PATCH 26/35] workqueue: implement worker_{set|clr}_flags() Tejun Heo
2010-06-28 21:04 ` [PATCH 27/35] workqueue: implement concurrency managed dynamic worker pool Tejun Heo
2010-07-09 9:11 ` Yong Zhang
2010-07-12 8:53 ` [PATCH] workqueue: fix locking in retry path of maybe_create_worker() Tejun Heo
2010-07-12 13:23 ` Yong Zhang
2010-07-14 9:37 ` Tejun Heo
2010-06-28 21:04 ` [PATCH 28/35] workqueue: increase max_active of keventd and kill current_is_keventd() Tejun Heo
2010-06-28 21:04 ` [PATCH 29/35] workqueue: s/__create_workqueue()/alloc_workqueue()/, and add system workqueues Tejun Heo
2010-06-28 21:04 ` [PATCH 30/35] workqueue: implement several utility APIs Tejun Heo
2010-06-28 21:04 ` [PATCH 31/35] workqueue: implement high priority workqueue Tejun Heo
2010-06-28 21:04 ` [PATCH 32/35] workqueue: implement cpu intensive workqueue Tejun Heo
2010-06-28 21:04 ` [PATCH 33/35] libata: take advantage of cmwq and remove concurrency limitations Tejun Heo
2010-06-28 22:32 ` Jeff Garzik
2010-06-29 7:00 ` Tejun Heo
2010-06-28 21:04 ` [PATCH 34/35] async: use workqueue for worker pool Tejun Heo
2010-06-28 22:55 ` Frederic Weisbecker
2010-06-29 7:25 ` Tejun Heo
2010-06-29 12:18 ` Frederic Weisbecker
2010-06-29 15:46 ` Tejun Heo
2010-06-29 15:52 ` Frederic Weisbecker
2010-06-29 15:55 ` Tejun Heo
2010-06-29 16:40 ` Arjan van de Ven
2010-06-29 16:59 ` Tejun Heo
2010-06-29 17:12 ` Tejun Heo
2010-06-29 18:08 ` Arjan van de Ven
2010-06-29 18:07 ` Arjan van de Ven
2010-06-29 18:15 ` Tejun Heo
2010-06-29 18:22 ` Arjan van de Ven
2010-06-29 18:34 ` Tejun Heo
2010-06-29 18:41 ` Arjan van de Ven
2010-06-29 18:59 ` Tejun Heo
2010-06-29 21:37 ` David Howells
2010-07-02 9:17 ` [PATCHSET] workqueue: implement and use WQ_UNBOUND Tejun Heo
2010-07-02 9:19 ` [PATCH 1/4] workqueue: prepare for WQ_UNBOUND implementation Tejun Heo
2010-07-02 9:24 ` [PATCH 3/4] workqueue: remove WQ_SINGLE_CPU and use WQ_UNBOUND instead Tejun Heo
2010-07-02 9:25 ` [PATCH 4/4] async: use workqueue for worker pool Tejun Heo
2010-07-02 15:09 ` Stefan Richter
2010-07-02 16:26 ` Tejun Heo
2010-07-02 16:25 ` [PATCH UPDATED " Tejun Heo
2010-07-02 9:28 ` [PATCH 2/4] workqueue: implement unbound workqueue Tejun Heo
2010-07-02 9:32 ` [PATCHSET] workqueue: implement and use WQ_UNBOUND Tejun Heo
2010-07-07 5:41 ` Tejun Heo
2010-07-14 9:39 ` Tejun Heo
2010-07-02 9:20 ` [PATCH 2/4] workqueue: implement unbound workqueue Tejun Heo
2010-07-20 22:01 ` [PATCHSET] workqueue: implement and use WQ_UNBOUND David Howells
2010-06-28 21:04 ` [PATCH 35/35] pcrypt: use HIGHPRI and CPU_INTENSIVE workqueues for padata Tejun Heo
2010-06-28 23:18 ` [PATCHSET] workqueue: concurrency managed workqueue, take#6 Frederic Weisbecker
2010-06-29 7:05 ` Tejun Heo
2010-07-02 8:32 ` [PATCHSET] workqueue: fixes on top of cmwq take#6 Tejun Heo
2010-07-02 8:33 ` [PATCH 1/4] workqueue: use worker_set/clr_flags() only from worker itself Tejun Heo
2010-07-02 8:34 ` [PATCH 2/4] workqueue: fix race condition in flush_workqueue() Tejun Heo
2010-07-02 8:35 ` [PATCH 3/4] workqueue: fix incorrect cpu number BUG_ON() in get_work_gcwq() Tejun Heo
2010-07-02 8:35 ` [PATCH 4/4] workqueue: fix worker management invocation without pending works Tejun Heo
2010-07-19 14:51 ` [PATCHSET] workqueue: concurrency managed workqueue, take#6 Tejun Heo
2010-07-21 13:23 ` David Howells
2010-07-21 14:52 ` Tejun Heo
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=4C2A1ABF.8050504@kernel.org \
--to=tj@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=arjan@linux.intel.com \
--cc=axboe@kernel.dk \
--cc=cl@linux-foundation.org \
--cc=dhowells@redhat.com \
--cc=dwalker@codeaurora.org \
--cc=florian@mickler.org \
--cc=fweisbec@gmail.com \
--cc=jeff@garzik.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=mst@redhat.com \
--cc=oleg@redhat.com \
--cc=randy.dunlap@oracle.com \
--cc=rusty@rustcorp.com.au \
--cc=stefanr@s5r6.in-berlin.de \
--cc=torvalds@linux-foundation.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;
as well as URLs for NNTP newsgroup(s).