[qcc] TCG

[Rob Landley]

On Sat, May 9, 2015 at 3:33 PM, Bruce Ewing <bewing0 at gmail.com> wrote:
>
>> I don't want to _fork_ tcg.
>
> It would be nicest if it could work out that way. However, the current tcg
> code uses GCC extensions. At the very least, I saw it use __func__ and
> arithmetic on void * values.

Building with llvm is explicitly of interest to the qemu guys, I
expect patches removing gcc extensions would be of interest to the
list.

> So either QCC would have to be extended to
> handle those, or the QEMU guys would have to voluntarily remove it from TCG.

It's nice to implement easy gcc extensions on general principles,
given that it's so widely used. That said, if llvm doesn't bother...

> Beyond that, there is a possibility that the logic may simply require mods
> in the TCG code (to handle floating point, maybe?) -- and then comes the
> question of whether the QEMU guys would accept non-QEMU-based mods in their
> TCG.

Indeed. In fact it probably _will_ require mods.

If you can get it working, I can try breaking up the changes and
submitting them upstream, and suggesting that tcg move to a submodule
because it's demonstrably independently useful. (I.E. I'm willing to
do some political work there, but I can't demonst in the absence of
something demonstrable.)

Rob

 1431212353.0