[Ocfs2-devel] [PATCH]2.6 fixes in super.c

Rusty Lynch rusty at linux.co.intel.com
Tue Feb 24 14:38:48 CST 2004


On Tue, Feb 24, 2004 at 02:12:12PM -0800, Mark Fasheh wrote:
> On Mon, Feb 23, 2004 at 08:25:01PM -0800, Rusty Lynch wrote:
> > Here is a redone version of the patch that applies cleanly to the current 
> > svn repository.
> > 
> >     --rusty
> <snip>
> 
> > @@ -721,15 +733,27 @@
> >          buf->f_type = OCFS_MAGIC;
> >          buf->f_bsize = sb->s_blocksize;
> >          buf->f_namelen = OCFS_MAX_FILENAME_LENGTH;
> > +        buf->f_bavail = buf->f_bfree;
> > +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,0)
> >          buf->f_blocks =
> > -            (unsigned long) ((unsigned long) (numbits) *
> > -                             (unsigned long) (osb->vol_layout.cluster_size >> 9) -
> > -			     (8 * ONE_MEGA_BYTE / osb->vol_layout.cluster_size)); 
> > +		(unsigned long) ((unsigned long) (numbits) *
> > +				 (unsigned long) (osb->vol_layout.cluster_size >> 9) -
> > +				 (8 * ONE_MEGA_BYTE / osb->vol_layout.cluster_size)); 
> >          buf->f_bfree =
> >              (unsigned long) (freebits * (osb->vol_layout.cluster_size >> 9));
> >          buf->f_bavail = buf->f_bfree;
> >          buf->f_files = (unsigned long) (numbits);
> >          buf->f_ffree = (unsigned long) (numbits) - freebits;
> > +#else
> > +        buf->f_blocks =
> > +		(sector_t) ((unsigned long) (numbits) *
> > +			    (unsigned long) (osb->vol_layout.cluster_size >> 9) -
> > +			    (8 * ONE_MEGA_BYTE / osb->vol_layout.cluster_size)); 
> > +        buf->f_bfree =
> > +            (sector_t) (freebits * (osb->vol_layout.cluster_size >> 9));
> > +        buf->f_files = (sector_t) (numbits);
> > +        buf->f_ffree = (sector_t) (numbits) - freebits;
> > +#endif
> 
> In this preprocessor conditional, don't we mean: 
> #if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,0)
> 
> instead of 
> #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,0)
> 
> If so, I'll just change it in my tree and commit that (you should prolly fix
> it in yours too).
> 	--Mark

Yeap, good catch.

    --rusty


More information about the Ocfs2-devel mailing list