> OK, if you have updated the GCC sources, would you mind updating
> the binutils-gdb.git as well? We should really keep the two as much
> in sync as we can.

Matthias asked me to push the patch for him, but the patch would not
apply. So what I did was just duplicate what we have in gcc/zlib, and
then reapplying the one patch that we had locally. Patch attached.

Tested on x86_64-linux by rebuilding GDB.

In my opinion, we should decide who has the master. Having to re-port
local patches was OK this time because there was only one, but this
is not going to be sustainable. Can we treat zlib the same way we treat
readline, and only accept patches once they have been accepted by
in the official zlib repository?

Right now, this is the change:

    commit f85997a697553fb7053de89bafe36e3d71c6f784
    Author: H.J. Lu <hjl.tools@gmail.com>
    Date:   Tue Mar 31 11:35:30 2015 -0700

        Use AM_ENABLE_MULTILIB only if with_target_subdir isn't empty

            * configure.ac (AM_ENABLE_MULTILIB): Use only if
            ${with_target_subdir} isn't empty.
            * configure: Regenerated.

At the very least, it should be pushed to GCC so both repositories
remain in sync.

-- 
Joel