[llvm] r233067 - Silencing some MSVC warnings "C4805: '^' : unsafe mix of type 'bool' and type 'unsigned int' in operation"; NFC.

[Aaron Ballman]

On Tue, Mar 24, 2015 at 12:13 PM, David Blaikie <dblaikie at gmail.com> wrote:
>
>
> On Tue, Mar 24, 2015 at 5:47 AM, Aaron Ballman <aaron at aaronballman.com>
> wrote:
>>
>> Author: aaronballman
>> Date: Tue Mar 24 07:47:51 2015
>> New Revision: 233067
>>
>> URL: http://llvm.org/viewvc/llvm-project?rev=233067&view=rev
>> Log:
>> Silencing some MSVC warnings "C4805: '^' : unsafe mix of type 'bool' and
>> type 'unsigned int' in operation"; NFC.
>
>
> FWIW, Andy Kaylor's started a review/discussion on disabling some MSVC
> warnings and cleaning up others to get a baseline.
>
> This is one I'm sort of inclined to disable (as mentioned in the review
> thread & some other recent discussions - the basic bar that's certainly been
> applied to GCC and I think is a reasonable one is: If the warning doesn't
> meet the quality bar to be added to Clang, then it probably doesn't meet the
> quality bar to be enabled for building LLVM projects and we should just
> disable the warning rather than fixing it even once)

I've not looked at what code is generated by MSVC for this construct,
but when the warning says "unsafe" in it, I get a bit worried. I'd be
opposed to silencing this warning if it turns out this really is an
unsafe operation with MSVC (though I cannot fathom why it would be
since bool can be implicitly converted to int with well-defined
semantics). If the operation is safe in MSVC and the warning is more
stylistic in nature, then I agree this would be one we can silence
globally.

In the meantime, I wanted to get the build back to warning-free.

~Aaron

>
>>
>>
>> Modified:
>>     llvm/trunk/lib/Support/APFloat.cpp
>>
>> Modified: llvm/trunk/lib/Support/APFloat.cpp
>> URL:
>> http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Support/APFloat.cpp?rev=233067&r1=233066&r2=233067&view=diff
>>
>> ==============================================================================
>> --- llvm/trunk/lib/Support/APFloat.cpp (original)
>> +++ llvm/trunk/lib/Support/APFloat.cpp Tue Mar 24 07:47:51 2015
>> @@ -1430,7 +1430,7 @@ APFloat::addOrSubtractSignificand(const
>>
>>    /* Determine if the operation on the absolute values is effectively
>>       an addition or subtraction.  */
>> -  subtract ^= sign ^ rhs.sign;
>> +  subtract ^= static_cast<bool>(sign ^ rhs.sign);
>>
>>    /* Are we bigger exponent-wise than the RHS?  */
>>    bits = exponent - rhs.exponent;
>>
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>