[patch libffi]: Add thiscall-closure support for fixing regression PR/51500

[Uros Bizjak]

On Tue, Jan 31, 2012 at 10:51 AM, Kai Tietz <ktietz70@googlemail.com> wrote:
> Ok, this variant keeps 16-byte alignment for thiscall-closure.

It doesn't. You have to account for the fact that caller pushed return
address to the stack, so %esp was already decreased by 4 behind your
back. To maintain alignment, you have to decrease %esp by $12, not by
$16.

Uros.