In order to expose the new RPC metrics facility to all NFS operations,
we need to get rid of the rpc_call wrapper in favor of the interfaces
(like rpc_call_sync) that expose the rpc_message structure directly to
the caller.
This patch does that.
- Chuck Lever
--
corporate: <cel at netapp dot com>
personal: <chucklever at bigfoot dot com>
On Wed, 2004-03-31 at 11:34, Lever, Charles wrote:
> In order to expose the new RPC metrics facility to all NFS operations,
> we need to get rid of the rpc_call wrapper in favor of the interfaces
> (like rpc_call_sync) that expose the rpc_message structure directly to
> the caller.
>
> This patch does that.
Please could you review those "inline" conversions? I'm at all not
convinced that something like nfs3_async_handle() is particularly useful
to inline (particularly not since it is used in every function in that
file).
Cheers,
Trond
--
Trond Myklebust <[email protected]>
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
> > In order to expose the new RPC metrics facility to all NFS=20
> operations,=20
> > we need to get rid of the rpc_call wrapper in favor of the=20
> interfaces=20
> > (like rpc_call_sync) that expose the rpc_message structure=20
> directly to=20
> > the caller.
>=20
> Please could you review those "inline" conversions? I'm at=20
> all not convinced that something like nfs3_async_handle() is=20
> particularly useful to inline (particularly not since it is=20
> used in every function in that file).
i inlined nfs3_rpc_wrapper because it is used in every NFSv3
proc. it removes a calling sequence and stack frame from
the path.
the other two inlines appear on the read/write/commit path,
but this path, being synchronous, is probably not as hot as
the async read/write/commit path. so you are welcome to
remove these if you have no other complaint about this patch.
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs