Bug in gcc CVS HEAD from a few days ago with alloca and stackalignment

[jeff.sturm@appnet.com]

On Sat, 10 Feb 2001, Kevin B. Hendricks wrote:
> I don't know whether this is a an implementation issue in alloca but either 
> way the the code is missing to round up r9 to the nearest power of 16.

Do you suppose Anthony Green's patch to use __builtin_alloca() might help:

http://gcc.gnu.org/ml/java-patches/2001-q1/msg00225.html

(This is not yet in the tree.)  Since it replaces the alloca
implementation I'm wondering if it wouldn't affect the outcome of your
libffi testing.

Jeff