> Da: Bodo Eggert <7eggert@gmx.de>
> A: Hisashi Hifumi <hifumi.hisashi@oss.ntt.co.jp>; akpm@linux-foundation.org; linux-kernel@vger.kernel.org;
linux-fsdevel@vger.kernel.org
> Inviato: Martedì 16 settembre 2008, 17:32:08
> Oggetto: Re: [PATCH] VFS: make file->f_pos access atomic on 32bit arch
>
> Hisashi Hifumi wrote:
>
> > Protecting file->f_pos with a lock adds some overhead but I think we should
> > fix this.
> > I used seqlock to serialize the access to file->f_pos. This approach is
> > similar to inode->i_size synchronization.
> >
> > Comments?
>
> > + if (offset != file_pos_read(file))
> > + file_pos_write(file, offset);
>
> Maybe file_pos_write should do this check?
>