Re: [PATCH 0/5] Allow more than PAGESIZE data read in configfs

[Andrew Morton <akpm@osdl.org>]

On Sat, 14 Oct 2006 01:01:07 -0700
Greg KH <greg@kroah.com> wrote:

> On Wed, Oct 11, 2006 at 04:27:13PM -0700, Chandra Seetharaman wrote:
> > On Wed, 2006-10-11 at 15:48 -0700, Andrew Morton wrote:
> > > On Wed, 11 Oct 2006 15:18:22 -0700
> > > Joel Becker <Joel.Becker@oracle.com> wrote:
> > > 
> > > > On Wed, Oct 11, 2006 at 01:19:35PM -0700, Andrew Morton wrote:
> > > > > The patch deletes a pile of custom code from configfs and replaces it with
> > > > > calls to standard kernel infrastructure and fixes a shortcoming/bug in the
> > > > > process.  Migration over to the new interface is trivial and almost
> > > > > scriptable.
> > > > 
> > > > 	The configfs stuff is based on the sysfs code too.  Should we
> > > > migrate sysfs/file.c to the same seq_file code?  Serious question, if
> > > > the cleanup is considered better.
> > > > 
> > > 
> > > I don't see why not.  I don't know if anyone has though of/proposed it
> > > before.
> > 
> > I can generate a patch for that too.
> 
> Argh!!!!
> 
> Are you going to honestly tell me you have a single attribute in sysfs
> that is larger than PAGE_SIZE?

He does not.  It's a matter of reusing existing facilities rather than
impementing similar things in multiple places.  The equivalent patch in
configfs removed a decent amount of code:

fs/configfs/file.c       |  130 ++++++++++-------------------------------------
include/linux/configfs.h |    3 -
2 files changed, 31 insertions(+), 102 deletions(-)