[linux-cifs-client] Re: [PATCH 3/3] cifs: have cifs_create use POSIX create if it's available

[Steve French]

It saves a round trip on query path info (a huge reduction in the number of
smbs sent)

On Wed, Feb 11, 2009 at 5:51 PM, Jeff Layton <jlayton at redhat.com> wrote:

> On Wed, 11 Feb 2009 13:05:32 -0600
> Shirish Pargaonkar <shirishpargaonkar at gmail.com> wrote:
>
> > On Wed, Feb 11, 2009 at 10:41 AM, Steve French <smfrench at gmail.com>
> wrote:
> > >
> > >
> > > On Wed, Feb 11, 2009 at 7:08 AM, Jeff Layton <jlayton at redhat.com>
> wrote:
> > >>
> > >> This reduces the number of calls to the server when creating files,
> and
> > >> also makes the force_create_mode option in samba work correctly.
> > >>
> > >> It also refactors some of the error handling in this function to
> > >> reduce layers of indentation.
> > >>
> > >> This is the second attempt at this patch. The first patch had a
> > >> bug in the open flag conversion routines and didn't try to enable
> > >> O_DIRECT on direct I/O mounts.
> > >
> > > This is great, but wondering if the posix create code should be split
> out to
> > > a different function - calling out early from cifs_create to a new
> > > posix_create (when the server supports posix extensions) to reduce the
> size
> > > of this (now) 250+ line cifs_create function (as I had done in my
> earlier
> > > start on posix_create last month).
> > >
> > >
> > >
> > > --
> > > Thanks,
> > >
> > > Steve
> > >
> > > _______________________________________________
> > > linux-cifs-client mailing list
> > > linux-cifs-client at lists.samba.org
> > > https://lists.samba.org/mailman/listinfo/linux-cifs-client
> > >
> > >
> >
> >
> > Nitpicking here, I think it would be a good idea to carve out
> > posix_create function that cifs_create calls.
> > Then this posix_create function could also be called from cifs_lookup
> > when lookup intents are implemented.
> > And may be it should be called posix_open instead of posix_create, it
> > is all open anyway.
> >
>
> What's the advantage of doing the actual create on lookup rather than
> in the create operation? It seems like doing the actual file creation
> with the ->create op is a more logical scheme...
>
> --
> Jeff Layton <jlayton at redhat.com>
>



-- 
Thanks,

Steve
-------------- next part --------------
HTML attachment scrubbed and removed