> > I'm answering that since this is plainly wrong. Bug 21809 was > > closed by yourself on 2005-05-29. This is not 1999! > > You deny that Bug 21809 is the same bug as Bug 323, which was closed in > 1999? > > Again, this is a place where you disagree that this should be considered > a "bug", but refuse to believe that reasonable people can disagree on > it.
the question then becomes whether GCC is helped or harmed by its current policy of unpredictable excess precision. it's quite unfortunate that gcc responders feel so free to vent their annoyance about this. Pinski, for instance, is absurdly unhelpful in suggesting that the bug reporter just abandon x87. again, the problem is well-understood: the lack of predictability of when excess precision is retained by x87. the problem is not EP or x87. the only constructive comment in this thread so far has been that libm might somehow rely on EP (ie can't use the _FPU_SET_CW workaround). which code is this? I'd guess it might be related to using series approximations, and the code could either set CW or accept 64b results. regards, mark hahn.
