[llvm-commits] [llvm] r60042 - in /llvm/trunk: lib/Support/APFloat.cpp test/Transforms/IndVarsSimplify/2008-11-25-APFloatAssert.ll

[Dale Johannesen]

On Nov 26, 2008, at 12:18 AMPST, Duncan Sands wrote:

>> Sure, how about adding a new opUnsupportedOp flag?  opInvalidOp means
>> something to IEEE, opUnsupportedOp could mean "we don't know how to
>> fold this" which is different.
>
> Why not teach APFloat to do arithmetic on PPCF128?

Sure, that's the most aesthetic solution.  It's not straightforward  
(the existing algorithms aren't usable) and there seems to be no  
demand for it.  It's on the list of "things we might get around to  
someday."