lkml.org 
[lkml]   [2004]   [Apr]   [7]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [Patch] BME, noatime and nodiratime
    On Wed, Apr 07, 2004 at 12:19:12PM +0200, Herbert Poetzl wrote:

    > > Yes, but that's S_NOATIME in inode->i_flags, not MS_NOATIME in sb->s_flags.
    > > Actually, I've been wrong here - we *do* need that check, since there are
    > > filesystems that force noatime or nodiratime.
    >
    > so we keep them?

    yes.

    > > it on remount. Most of those are my fault - I've missed the remount side of
    > > things in readdir patch. They should have nodiratime always on to match
    > > the original behaviour. IMO, both (a) and (b) should be handled by a new
    > > field - sb->s_forced_flags. do_remount_sb() would set those in flags before
    > > doing anything else. Note that assignment to ->s_flags in do_remount_sb()
    > > is not safe - e.g. ext[23] can get forced r/o by fs error and if that
    > > happens after return from ->remount_sb() but before assignement, we are
    > > screwed. IOW, do_remount_sb() will need more work.
    >
    > are you going to do this?

    Yes - for now I'll do a fix that doesn't change API (IOW, add/update
    ->remount_fs() to filesystems with forced flags, forced r/o has the
    same problems); we'll see what should be done in the long run when
    the things clean up a bit. I'd rather avoid struct super_block changes
    that would have to be reverted soon.

    > >> simple, to match the MS_* counterparts, something which
    > >> actually confused me in the first place (in the code)
    >
    > so is this okay? actually I'd prefer to use the same
    > values for the MNT_NOATIME and MNT_NODIRATIME too ...
    > (as in the previous version I did)

    Hmm... Potentially we are breaking ABI for no good reason, since these
    defines are visible to out-of-tree code. I don't think that we should
    care about matching MS_... stuff, simply because MS_... encoding is ugly
    as hell and there's no reason to use MNT_... and MS_... in the same
    context.
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

    \
     
     \ /
      Last update: 2005-03-22 14:02    [W:0.032 / U:59.856 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site