[PATCH] Move subcall decoding configuration out of the common code

[Mike Frysinger]

On Tuesday 07 May 2013 11:58:57 Dmitry V. Levin wrote:
> On Tue, May 07, 2013 at 11:42:28AM -0400, Chris Metcalf wrote:
> > On 5/7/2013 11:38 AM, Dmitry V. Levin wrote:
> > > On Tue, May 07, 2013 at 09:30:45AM -0400, Mike Frysinger wrote:
> > >> > On Tuesday 07 May 2013 08:14:46 Chris Metcalf wrote:
> > >>> > > On 5/6/2013 7:43 PM, Dmitry V. Levin wrote:
> > >>>> > > > On Mon, May 06, 2013 at 03:21:38PM -0700, Chris Zankel wrote:
> > >>>>> > > >> The Xtensa architecture also used dedicated syscalls and
> > >>>>> > > >> doesn't need to multiplex ipc and socket subcalls.
> > >>>> > > > 
> > >>>> > > > Applied, thanks.
> > >>> > > 
> > >>> > > No new architecture will ever use the subcall model, so in fact
> > >>> > > it's probably cleaner to just list the ones that do use it.
> > >> > 
> > >> > yeah, i was thinking it might be better to move those ifdefs out of
> > >> > the common code and just declare it in the relevant syscallent.h
> > >> > header
> > > 
> > > Yes, something like this?
> > 
> > Rather than all the checks to see if SYS_socket_subcall has the right
> > value, I'd just require all platforms to define it first, and then in
> > linux/subcall.h, add an #ifndef to error out if it's not defined.
> 
> I agree.

LGTM
-mike
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.strace.io/pipermail/strace-devel/attachments/20130507/f024b8a8/attachment.bin>