Linux Btrfs filesystem development
 help / color / mirror / Atom feed
From: David Sterba <dsterba@suse.cz>
To: Qu Wenruo <wqu@suse.de>
Cc: dsterba@suse.cz, Qu Wenruo <wqu@suse.com>, linux-btrfs@vger.kernel.org
Subject: Re: [PATCH] btrfs: Introduce compile time structure size check
Date: Fri, 13 Jul 2018 16:46:38 +0200	[thread overview]
Message-ID: <20180713144638.GY3126@twin.jikos.cz> (raw)
In-Reply-To: <5f2369c6-c882-7ae6-04f5-6de98bf10f58@suse.de>

On Fri, Jul 13, 2018 at 10:36:42PM +0800, Qu Wenruo wrote:
> 
> 
> On 2018年07月13日 22:34, David Sterba wrote:
> > On Thu, Jul 12, 2018 at 02:19:07PM +0800, Qu Wenruo wrote:
> >> Introduce a new macro based compile time check for ioctl structures.
> >>
> >> The new macro is BTRFS_ASSERT_SIZE(), which is mostly copied from
> >> VMMDEV_ASSERT_SIZE().
> > 
> > The macro should be generic, there's nothing specific to btrfs. There's
> > a similar one in the progs.
> > 
> >> Such check is only added to structure pended to power of 2.
> > 
> > There's no constraint about power of 2 sizes, it works for any size.
> 
> While some structure doesn't do padding at all, and it looks like they
> may get expanded later.
> Do we really need to limit the size right now?

The size of an ioctl structure is part of kernel<->userspace API and
must not change once public. The ioctl number is calculated using the
structure size.

> >> And exposed one structure, btrfs_ioctl_get_dev_stats() is not aligned
> >> well.
> >> The misalign is introduced by commit b27f7c0c150f ("btrfs: join DEV_STATS
> >> ioctls to one").
> > 
> > Yeah, that was an oversight, so the correct value to check against is
> > 1024 + 8 = 1032, see ioctl.h in progs.
> 
> Can't we just revert to 1024?

Answered by the above, we absolutely cannot change the numbers now.

> That plus 8 doesn't really look well, and shrink the size shouldn't
> bring any problem AFAIK.

See

https://elixir.bootlin.com/linux/latest/source/include/uapi/asm-generic/ioctl.h#L88

for _IOWR definition,

btrfs-progs:ioctl.h

817 #define BTRFS_IOC_GET_DEV_STATS _IOWR(BTRFS_IOCTL_MAGIC, 52, \
818                                       struct btrfs_ioctl_get_dev_stats)

kernel:fs/btrfs/ioctl.c

5914         case BTRFS_IOC_GET_DEV_STATS:
5915                 return btrfs_ioctl_get_dev_stats(fs_info, argp);

The values must match, otherwise dev stats will randomly break on
old/new kernel and progs combinations.

  reply	other threads:[~2018-07-13 15:01 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-12  6:19 [PATCH] btrfs: Introduce compile time structure size check Qu Wenruo
2018-07-12  6:27 ` Nikolay Borisov
2018-07-12  6:56   ` Qu Wenruo
2018-07-13 14:34     ` David Sterba
2018-07-13 14:34 ` David Sterba
2018-07-13 14:36   ` Qu Wenruo
2018-07-13 14:46     ` David Sterba [this message]
2018-07-13 14:57       ` Qu Wenruo
2018-07-13 15:25         ` David Sterba

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=20180713144638.GY3126@twin.jikos.cz \
    --to=dsterba@suse.cz \
    --cc=linux-btrfs@vger.kernel.org \
    --cc=wqu@suse.com \
    --cc=wqu@suse.de \
    /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