From: "Darrick J. Wong" <darrick.wong@oracle.com>
To: Eric Sandeen <sandeen@sandeen.net>
Cc: xfs@oss.sgi.com
Subject: Re: [PATCH] xfs: check sizes of XFS on-disk structures at compile time
Date: Fri, 15 Jan 2016 13:05:56 -0800 [thread overview]
Message-ID: <20160115210556.GC5757@birch.djwong.org> (raw)
In-Reply-To: <56969696.2040206@sandeen.net>
On Wed, Jan 13, 2016 at 12:25:26PM -0600, Eric Sandeen wrote:
> On 1/11/16 5:26 PM, Darrick J. Wong wrote:
>
> > diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c
> > index 36bd882..31b69d1 100644
> > --- a/fs/xfs/xfs_super.c
> > +++ b/fs/xfs/xfs_super.c
> > @@ -1812,11 +1812,56 @@ xfs_destroy_workqueues(void)
> > destroy_workqueue(xfs_alloc_wq);
> > }
> >
> > +static void __init
> > +xfs_check_ondisk_structs(void)
> > +{
> > + /* on-disk structures */
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_dsb, 264);
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_agf, 224);
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_agi, 336);
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_agfl, 36);
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_timestamp, 8);
> > + XFS_CHECK_STRUCT_SIZE(struct xfs_dinode, 176);
>
> I'm not sure why, but all this hard-coded stuff in the codebase bugs me.
>
> What if we did it as part of the build, but outside the code, something
> like (pardon my bad Make-fu):
>
> diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile
> index f646391..da9ec6d 100644
> --- a/fs/xfs/Makefile
> +++ b/fs/xfs/Makefile
> @@ -122,3 +122,8 @@ xfs-$(CONFIG_XFS_POSIX_ACL) += xfs_acl.o
> xfs-$(CONFIG_SYSCTL) += xfs_sysctl.o
> xfs-$(CONFIG_COMPAT) += xfs_ioctl32.o
> xfs-$(CONFIG_NFSD_PNFS) += xfs_pnfs.o
> +
> +check_structures: $(obj)/xfs.o
> + $(src)/scripts/check_structures.sh $(obj)/xfs.o
> +
> +$(obj)/built-in.o: check_structures
>
> where the script executes pahole or whatever we want, to check the
> structure sizes and possibly even alignment explicitly? Presumably
> we could make running the check a Kconfig option as well...
We certainly /could/ do that, but now the build depends on pahole or whatever
being installed, whereas this macroey doesn't add more tool dependencies.
I think between this and xfs/122 we're probably fine?
--D
>
> similar tricks could surely be done for userspace as well.
>
> -Eric
>
> _______________________________________________
> xfs mailing list
> xfs@oss.sgi.com
> http://oss.sgi.com/mailman/listinfo/xfs
_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs
next prev parent reply other threads:[~2016-01-15 21:06 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-01-11 23:26 [PATCH] xfs: check sizes of XFS on-disk structures at compile time Darrick J. Wong
2016-01-12 14:01 ` Brian Foster
2016-01-13 18:25 ` Eric Sandeen
2016-01-15 21:05 ` Darrick J. Wong [this message]
2016-01-15 20:58 ` Darrick J. Wong
2016-01-15 22:52 ` Dave Chinner
2016-01-20 5:12 ` Darrick J. Wong
2016-01-20 15:40 ` Josef 'Jeff' Sipek
2016-01-22 22:01 ` Darrick J. Wong
2016-02-01 5:06 ` Dave Chinner
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=20160115210556.GC5757@birch.djwong.org \
--to=darrick.wong@oracle.com \
--cc=sandeen@sandeen.net \
--cc=xfs@oss.sgi.com \
/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