bsd patch regression?

[Pedro Giffuni]

Hi Sean;

El 27/08/2013 5:58 p. m., Sean Bruno escribió:
> Colin generated a patch for xen things that does some pretty typical
> behavior.  bsdpatch really didn't handle it well and rejected some
> things and flat out refused to create sys/modules/xenhvm/Makefile for
> me.
>
> http://lists.freebsd.org/pipermail/freebsd-xen/2013-August/001697.html
>
> When applying this patch with gnupatch I get:
> http://people.freebsd.org/~sbruno/gnupatch.txt
>
>
> When applying this patch with bsdpatch I get:
> http://people.freebsd.org/~sbruno/bsdpatch.txt
>

I note that you are using GNU patch from the ports tree.

It would appear that both utilities reject the patches for:
sys/amd64/conf/XENHVM
sys/i386/conf/XENHVM

bsdpatch is uglily more verbose and handles the rejected patches much 
less gracefully but it seems like gpatch also has issues with the same 
patches.

I am not sure we can call this a regression: please note that bsd patch
is meant to replace the ancient GNU patch that we had in the tree (it's
still there under the name "gnupatch"). We ran an exp-run on ports and
there was only a small regression related to the patch level which is a
little stricter in BSD patch.

>
> Any ideas here?
>

Not very helpful but I suggest using "svn patch" when possible. :(

Pedro.