Also...
--- lib/Sema/SemaTemplateInstantiateDecl.cpp (revision 181010)
+++ lib/Sema/SemaTemplateInstantiateDecl.cpp (working copy)
@@ -1152,12 +1152,12 @@
// this declaration.
FunctionTemplateDecl *FunctionTemplate =
D->getDescribedFunctionTemplate();
if (FunctionTemplate && !TemplateParams) {
- std::pair<const TemplateArgument *, unsigned> Innermost
+ ArrayRef<TemplateArgument> Innermost
= TemplateArgs.getInnermost();
May as well fold this to one line now.
On Fri, May 3, 2013 at 2:36 PM, Richard Smith <[email protected]> wrote:
> The data() + size() changes shouldn't be necessary; there's an
> ArrayRef(const T *data, size_t length) constructor. Otherwise, looks great.
> Do you have commit access yet?
>
>
> On Fri, May 3, 2013 at 2:17 PM, Faisal Vali <[email protected]> wrote:
>
>> As requested - a patch that replaces the pair kludge with ArrayRef.
>>
>> Sorry about that earlier non-sense comment - I just assumed I remembered
>> how that member was used correctly when i fiddled with it last (months
>> ago). Should have looked at the code.
>>
>> Let me know what you think of this patch.
>>
>> Faisal Vali
>>
>>
>>
>> On Fri, May 3, 2013 at 1:19 PM, Richard Smith <[email protected]>wrote:
>>
>>> The Template.h comment is not right -- an ArgList is really a
>>> second-rate facsimile of an ArrayRef. A patch making it be a typedef for
>>> ArrayRef<TemplateArgument> would be great, though...
>>>
>>> getTemplateInstantiationArgs already has a doc comment in
>>> SemaTemplateInstantiate.cpp, which seems pretty good.
>>>
>>>
>>> On Fri, May 3, 2013 at 8:00 AM, Faisal Vali <[email protected]> wrote:
>>>
>>>> What do you think Richard? Are the comments accurate - and should we
>>>> add them?
>>>>
>>>> Faisal Vali
>>>>
>>>>
>>>
>>
>
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
