bugpoint / cbe Problems

Hello, David.

After hacking around in the CBE output I managed to compile it with
gcc, only to discover that gcc 4.1.2 has the SAME bug LLVM does with
respect to alignment.

Automatic stack realignment was atted to X86 backend ~3 months ago. Everything should
work with LLVM. If not - please fill out a PR. AFAIR, automatic stack realignment still does
not land into gcc mainline (it's on maybe 8th or 9th iteration or so).

I don't see a way to run opt directly with bugpoint. It seems to want to go
through the CBE, which has the problems I mentioned before (generating
syntactically-incorrect C).

Please file a PR.

Hello, David.

> After hacking around in the CBE output I managed to compile it with
> gcc, only to discover that gcc 4.1.2 has the SAME bug LLVM does with
> respect to alignment.

Automatic stack realignment was atted to X86 backend ~3 months ago.
Everything should work with LLVM. If not - please fill out a PR. AFAIR,
automatic stack realignment still does not land into gcc mainline (it's on
maybe 8th or 9th iteration or so).

Right. The object is on the stack. This is a fortran code where all sorts of
interesting things with pointers can be done behind your back. I'll try to
write a simple C test that has the equivalent problem.

> I don't see a way to run opt directly with bugpoint. It seems to want to
> go through the CBE, which has the problems I mentioned before (generating
> syntactically-incorrect C).

Please file a PR.

It's on my list. :slight_smile:

                                                     -Dave

Testcases in LLVM IL are fine, if that's easier.

-Eli