On Fri, May 06, 2011 at 10:11:48AM -0600, Eric Blake wrote:
On 05/06/2011 07:11 AM, Daniel P. Berrange wrote:
> commit 7a2fd256cb85dd9a9d6640d9faed0b66ca617411
> Author: Daniel P. Berrange <berrange(a)redhat.com>
> Date: Fri May 6 13:36:21 2011 +0100
>
> Remove all generated files for remote protocol
>
> Stop storing the generated files for the remote protocol client
> and server in source control. The generated files will still be
> included in the result of 'make dist' to avoid end-users needing
> to generate the files
>
> * daemon/Makefile.am: Removed generated files with
> maintainer-clean target
> * src/Makefile.am: Removed generated files with
> maintainer-clean target. Always run 'rpcgen' if
> generated files are missing
>
>
> NB: hand edited diff to remove the huge deleted files
Thanks for doing that. However, it means that 'git am' didn't like this
patch, so I hope I tested it correctly. At any rate, after applying
your patch, removing the generated files, then running 'make dist', and
comparing that to a 'make dist' pre-patch, the minor differences were
all expected (in the Makefiles, not in the generated files).
> +++ b/src/remote/rpcgen_fix.pl
> @@ -1,4 +1,6 @@
> -# Fix XDR code (generated by rpcgen) so that it compiles
> +#
> +# Generate code for an XDR protocol, optionally applying
> +# fixups to the glibc rpcgen code so that it compiles
> # with warnings turned on.
> #
> # This code is evil. Arguably better would be just to compile
> @@ -17,10 +19,35 @@ use strict;
> my $in_function = 0;
> my @function = ();
>
> -while (<>) {
> +my $rpcgen = shift;
> +my $mode = shift;
> +my $xdrdef = shift;
> +my $target = shift;
> +
> +unlink $target;
> +
> +open RPCGEN, "-|", $rpcgen, $mode, $xdrdef
> + or die "cannot run $rpcgen $mode $xdrdef: $!";
> +open TARGET, ">$target"
> + or die "cannot create $target: $!";
> +
> +my $fixup = $^O eq "linux";
I guess I'll see what happens when I try this on Cygwin, but if there
are any followups needed, it should be separate patches. As is, cygwin
already can't use -Werror because of repetitive declarations in the
tirpc headers that trigger one of our default set of gcc warnings.
This shouldn't make Cygwin any worse should it ? We don't support
running bootstrap on Win32 at all, only building from a dist. So
they'd not be running rpcgen
Daniel
--
|:
http://berrange.com -o-
http://www.flickr.com/photos/dberrange/ :|
|:
http://libvirt.org -o-
http://virt-manager.org :|
|:
http://autobuild.org -o-
http://search.cpan.org/~danberr/ :|
|:
http://entangle-photo.org -o-
http://live.gnome.org/gtk-vnc :|