From: Mikulas Patocka <mpatocka@redhat.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jens Axboe <axboe@kernel.dk>,
Jeff Chua <jeff.chua.linux@gmail.com>,
Lai Jiangshan <laijs@cn.fujitsu.com>, Jan Kara <jack@suse.cz>,
lkml <linux-kernel@vger.kernel.org>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>
Subject: Re: [PATCH] Introduce a method to catch mmap_region (was: Recent kernel "mount" slow)
Date: Wed, 28 Nov 2012 14:50:56 -0500 (EST) [thread overview]
Message-ID: <Pine.LNX.4.64.1211281422360.26389@file.rdu.redhat.com> (raw)
In-Reply-To: <CA+55aFx4zYnEsUyGJurBjB0XZ0AgmyKqW-+KWVDvma3w6csnHQ@mail.gmail.com>
On Wed, 28 Nov 2012, Linus Torvalds wrote:
> No, this is crap.
>
> We don't introduce random hooks like this just because the block layer
> has shit-for-brains and cannot be bothered to do things right.
>
> The fact is, the whole locking in the block layer open routine is
> total and utter crap. It doesn't lock the right thing, even with your
> change *anyway* (or with the change Jens had). Absolutely nothing in
> "mmap_region()" cares at all about the block-size anywhere - it's
> generic, after all - so locking around it is f*cking pointless. There
> is no way in hell that the caller of ->mmap can *ever* care about the
> block size, since it never even looks at it.
>
> Don't do random crap like this.
>
> Why does the code think that mmap matters so much anyway? As you say,
> the mmap itself does *nothing*. It has no impact for the block size.
>
> Linus
mmap_region() doesn't care about the block size. But a lot of
page-in/page-out code does.
The problem is that once the block device is mapped, page faults or page
writeback can happen anytime - so the simplest solution is to not allow
the block device being mapped while we change block size.
The function set_blocksize takes bd_block_size_semaphore for write (that
blocks read/write/mmap), then it calls sync_blockdev (now we are sure that
there is no more writeback), then it changes the block size, then it calls
kill_bdev (now we are sure that there are no more any pages with buffers
with the old blocksize).
If you want to allow to change block size while a block device is mapped,
you'd have to add explicit locks around all mm callbacks (so that the
block size can't change while the callback is in progress) - and even
then, there are some unsolvable cases - i.e. what are you going to do if
the user mlocks a mapped block device and you change block size of that
device? - you can't drop the pages (that would violate mlock semantics)
and you can leave them there (because they have buffers with wrong size).
If you don't like what I sent, propose a different solution.
Mikulas
next prev parent reply other threads:[~2012-11-28 19:51 UTC|newest]
Thread overview: 82+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-11-19 0:33 Recent kernel "mount" slow Jeff Chua
2012-11-20 18:09 ` Jan Kara
2012-11-21 15:46 ` Jeff Chua
2012-11-22 14:30 ` Jeff Chua
2012-11-22 19:21 ` Linus Torvalds
2012-11-23 13:24 ` Jens Axboe
2012-11-23 22:21 ` Jeff Chua
2012-11-23 23:31 ` Jeff Chua
2012-11-23 23:48 ` Jeff Chua
2012-11-24 21:09 ` Mikulas Patocka
2012-11-24 23:23 ` Jeff Chua
2012-11-27 5:57 ` Jeff Chua
2012-11-27 7:38 ` Jens Axboe
2012-11-27 7:44 ` Jens Axboe
2012-11-27 8:45 ` Jeff Chua
2012-11-27 10:06 ` Jeff Chua
2012-11-27 12:33 ` Jens Axboe
2012-11-28 3:57 ` Mikulas Patocka
2012-11-28 8:33 ` Jens Axboe
2012-11-28 13:05 ` Jeff Chua
2012-11-28 17:25 ` [PATCH] Introduce a method to catch mmap_region (was: Recent kernel "mount" slow) Mikulas Patocka
2012-11-28 19:15 ` Linus Torvalds
2012-11-28 19:43 ` Al Viro
2012-11-28 19:53 ` Linus Torvalds
2012-11-28 22:01 ` [PATCH v2] Do a proper locking for mmap and block size change Mikulas Patocka
2012-11-29 17:19 ` Linus Torvalds
2012-11-29 18:23 ` Mikulas Patocka
2012-11-29 18:46 ` Linus Torvalds
2012-11-29 19:02 ` Linus Torvalds
2012-11-29 19:15 ` Chris Mason
2012-11-29 19:26 ` Linus Torvalds
2012-11-29 19:48 ` Chris Mason
2012-11-29 19:55 ` Linus Torvalds
2012-11-29 20:10 ` Linus Torvalds
2012-11-29 20:52 ` Linus Torvalds
2012-11-29 21:29 ` Chris Mason
2012-11-29 22:16 ` Linus Torvalds
2012-11-29 22:36 ` Linus Torvalds
2012-11-30 1:16 ` Chris Mason
2012-11-30 2:13 ` Linus Torvalds
2012-11-30 2:27 ` Chris Mason
2012-11-30 2:49 ` Dave Chinner
2012-11-30 14:31 ` Chris Mason
2012-11-30 16:42 ` Linus Torvalds
2012-11-30 16:36 ` Christoph Hellwig
2012-11-30 22:40 ` Dave Chinner
2012-11-30 23:09 ` Christoph Hellwig
2012-11-29 19:50 ` Linus Torvalds
2012-11-28 19:50 ` Mikulas Patocka [this message]
2012-11-28 20:03 ` [PATCH] Introduce a method to catch mmap_region (was: Recent kernel "mount" slow) Linus Torvalds
2012-11-28 20:13 ` Linus Torvalds
2012-11-28 20:32 ` Linus Torvalds
2012-11-28 20:47 ` Linus Torvalds
2012-11-28 22:10 ` Mikulas Patocka
2012-11-28 21:29 ` Mikulas Patocka
2012-11-28 22:52 ` Linus Torvalds
2012-11-28 23:13 ` Linus Torvalds
2012-11-29 1:20 ` Mikulas Patocka
2012-11-29 0:38 ` Mikulas Patocka
2012-11-29 2:04 ` Linus Torvalds
2012-11-29 2:58 ` Linus Torvalds
2012-11-29 6:16 ` Linus Torvalds
2012-11-29 6:25 ` Al Viro
2012-11-29 6:30 ` Al Viro
2012-11-29 6:37 ` Linus Torvalds
2012-11-29 6:45 ` Al Viro
2012-11-29 10:57 ` Jeff Chua
2012-11-29 6:33 ` Linus Torvalds
2012-11-29 14:12 ` Chris Mason
2012-11-29 17:26 ` Chris Mason
2012-11-29 17:26 ` Linus Torvalds
2012-11-29 17:51 ` Chris Mason
2012-11-29 18:12 ` Linus Torvalds
2012-11-28 3:59 ` [PATCH 1/2] percpu-rwsem: use synchronize_sched_expedited Mikulas Patocka
2012-11-28 4:01 ` [PATCH 2/2] block_dev: don't take the write lock if block size doesn't change Mikulas Patocka
2012-11-28 14:24 ` Jeff Chua
2012-11-28 22:03 ` Mikulas Patocka
2012-11-28 14:19 ` [PATCH 1/2] percpu-rwsem: use synchronize_sched_expedited Jeff Chua
2012-11-30 0:06 ` Andrew Morton
2012-11-30 3:00 ` Mikulas Patocka
2012-11-30 13:42 ` Paul E. McKenney
2012-11-30 18:57 ` Oleg Nesterov
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=Pine.LNX.4.64.1211281422360.26389@file.rdu.redhat.com \
--to=mpatocka@redhat.com \
--cc=axboe@kernel.dk \
--cc=jack@suse.cz \
--cc=jeff.chua.linux@gmail.com \
--cc=laijs@cn.fujitsu.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--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
all inboxes | Powered by JetHome®