Porting pretty printers to Python 2+3

[Samuel Bronson]

TL;DR: I started the copyright assignment process after jwakely sent me
the first form off-list, and now I'm wondering if it's gone through yet.

I'm also responding to some of his points, which I may or may not have
responded to already on IRC.

Jonathan Wakely <jwakely@redhat.com> writes:

> On 21/05/14 15:45 -0400, Samuel Bronson wrote:
>>I have two patches:
>
> N.B. all GCC patches need to be sent to gcc-patches@gcc.gnu.org,
> libstdc++ ones get sent to this list and that one.

Oops.  I guess missed the reference to the global GCC contribution
rules...

[...]
>>This is just to make it easy to test my changes, since my computer is
>>evidently too pathetic to build GCC.
>
> Testing an installed compiler would usually be done to test the same
> version as the testsuite belongs to, but I suppose this change doesn't
> hurt.

Well, my understanding was that these pretty printers were supposed to
be compatible with at least any *past* version of libstdc++.so.6, so I
figured it should work for at least a rough idea of whether I broke the
world or not.  (Also, libstdc++ is newer than gcc here anyway; Debian
shares a single libstdc++ among all the available g++-X.Y packages.)

[...]
>>My changes give me the exact same set of failures in GDB 7.7 when
>>linked with python2.7 and with python3.3, for what seem to be the
>>exact same reasons; I'm expecting them to go away in GDB 7.7.1 once I
>>get that built.
>
> Which failures? There shouldn't be any, although consistency is a good
> start!

It doesn't particularly matter; they went away once genattrtab decided
not to OOM on me and I could run the testsuite in the usual way, with
fresh-built gcc, libgcc, libstdc++, etc.

>>Besides review, I believe I also need an Assignment Form -- I'm
>>assuming I can't get away with claiming that this code is effectively
>>part of GDB, and that my GDB assignment papers are thus all I need?
>
> I don't think you can get away with that, no, sorry.
> The first patch is probably trivial enough to take without an
> assignment though.

Okay, so, um, I've gotten as far as signing the papers and mailing them
back, but I don't seem to have gotten the fully signed ones returned to
me yet.  Should I go ahead and send again anyway -- this time to both
lists -- or keep waiting?

-- 
Hi! I'm a .signature virus! Copy me into your ~/.signature to help me spread!