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

[David Edelsohn]

>>>>> "Kevin B Hendricks" writes:


Kevin> Has this bug always existed in gcc 2.95.X or was it recently changed for the 
Kevin> upcoming 3.0?

	The STACK_BOUNDARY 64 existed in gcc-2.95 branch.

Kevin> I suppose I could rewrite his code in libffi to remove this constraint but I 
Kevin> wouldn't want to do that without finding out what Geoff wants to do here.

Kevin> Should I e-mail Geoff directly?

	It probably would not hurt to bring this to his attention
directly.

David