Dorit Nuzman wrote:

I am pretty sure that adding -fwrapv will
fix the problem, but that means we are changing the test case.

gen-vect-11.c is Paolo's test, so it's his call I guess
(he already agreed to add -fwrapv -
http://gcc.gnu.org/ml/gcc-patches/2006-01/msg00579.html)
Well, gen-vect-11.c was taken from some vect-*.c test that I don't remember -- probably one that Kazu failed, and that's why I proposed adding -fwrapv. It's your test as much as mine. :-)

Paolo

Reply via email to