On Fri, Jun 7, 2013 at 4:46 PM, John McCall <[email protected]> wrote:
> On Jun 7, 2013, at 4:42 PM, Arthur O'Dwyer <[email protected]> wrote:
>> On Fri, Jun 7, 2013 at 3:22 PM, John McCall <[email protected]> wrote:
>>> On Jun 7, 2013, at 12:07 PM, Arthur O'Dwyer <[email protected]> 
>>> wrote:
>>>>
>>>> +VALUE_CODEGENOPT(PassSmallStructsConvention, 2, 0) /// If
>>>> -fpcc-struct-return (==1) or -freg-struct-return (==2) is specified.
>>>
>>> Use ENUM_CODEGENOPT instead of magic numbers, please.
>>
>> Fixed. I renamed this variable to SmallStructsConvention and the enum to
>> SmallStructsConventionKind to save a bit of horizontal space.
>> I removed the "default: assert(false);" case in my helper function, as it was
>> now triggering a warning from -Wcovered-switch-default.
>
> How about "StructReturnConventionKind"?

Fixed, although I realized afterward that this changes "SSC_" to "SRC_",
which is slightly unfortunate ("SRC" being short for "source" in other
contexts).

> Please add tests and then attach the final patch instead of including it 
> inline.

Attached.

-Arthur

Attachment: fpcc-struct-return-v4.diff
Description: Binary data

_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to