PR 1159/1412

Wed Mar 21 15:34:00 GMT 2001

Jeff Sturm wrote:
> On 2001年3月20日, Jeff Sturm wrote:
> > I began to look at 1159 and 1412. Neither is reproducible anymore.
>> Sorry folks. I'm retracting this statement. Under some conditions, these
> PR's still matter on x86. (It seems to have been corrected on Alpha, and
> reportedly PPC.)

If it can't be fixed properly before the release then we'll have to make the
front end force a "-fno-defer-pop" (which is the existing known workaround).
regards
 [ bryce ]