[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <alpine.LSU.2.00.1205141242530.1623@eggly.anvils>
Date: Mon, 14 May 2012 12:48:16 -0700 (PDT)
From: Hugh Dickins <hughd@...gle.com>
To: Cong Wang <xiyou.wangcong@...il.com>
cc: Andrew Morton <akpm@...ux-foundation.org>,
Christoph Hellwig <hch@...radead.org>,
Andi Kleen <andi@...stfloor.org>,
Al Viro <viro@...iv.linux.org.uk>, linux-mm@...ck.org,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 2/10] tmpfs: enable NOSEC optimization
On Mon, 14 May 2012, Cong Wang wrote:
> On 05/12/2012 08:02 PM, Hugh Dickins wrote:
> > Let tmpfs into the NOSEC optimization (avoiding file_remove_suid()
> > overhead on most common writes): set MS_NOSEC on its superblocks.
> >
> > Signed-off-by: Hugh Dickins<hughd@...gle.com>
> > ---
> > mm/shmem.c | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > --- 3045N.orig/mm/shmem.c 2012-05-05 10:45:17.888060878 -0700
> > +++ 3045N/mm/shmem.c 2012-05-05 10:46:05.732062006 -0700
> > @@ -2361,6 +2361,7 @@ int shmem_fill_super(struct super_block
> > }
> > }
> > sb->s_export_op =&shmem_export_ops;
> > + sb->s_flags |= MS_NOSEC;
>
> Isn't setting the flag on inode better? Something like:
I don't think so. The MS_NOSEC S_NOSEC business is fairly subtle,
and easy to miss if it's gone wrong, so I would much rather follow
the established pattern in local block filesystems: which is to set
MS_NOSEC in superblock flags, and leave S_NOSEC to file_remove_suid().
Hugh
>
> diff --git a/mm/shmem.c b/mm/shmem.c
> index f99ff3e..7d98fb5 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -2325,6 +2325,7 @@ static void shmem_init_inode(void *foo)
> {
> struct shmem_inode_info *info = foo;
> inode_init_once(&info->vfs_inode);
> + info->vfs_inode.i_flags |= S_NOSEC;
> }
>
> static int shmem_init_inodecache(void)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists