[llvm-commits] [PATCH] Instcombine: Don't optimize pow(2.0, x) -> exp2(x) if exp2 is unavailable

[Stephen Canon]

-fno-builtin-pow, I would think.

On Nov 26, 2012, at 2:37 PM, Owen Anderson <resistor at mac.com> wrote:

> The build system for the library should be passing -fno-builtin-exp2 (or something similar).
> 
> --Owen
> 
> On Nov 26, 2012, at 11:33 AM, Stephen Canon <scanon at apple.com> wrote:
> 
>> In the scenario you describe, wouldn't TLI->has(LibFunc::exp2) be true because the target has an exp2 (which calls pow(2,x))?
>> 
>> - Steve
>> 
>> On Nov 26, 2012, at 2:17 PM, Derek Schuff <dschuff at google.com> wrote:
>> 
>>> Hi,
>>> Attached is a patch to check that exp2 is available using the usual
>>> TLI mechanism before optimizing pow(2.0, x) to exp2(x). This is a
>>> problem if compiling a math library that implements exp2(x) as a call
>>> to pow(2,x).
>>> 
>>> please review, thanks
>>> <Libcalls.diff>_______________________________________________
>>> llvm-commits mailing list
>>> llvm-commits at cs.uiuc.edu
>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>> 
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>