[cfe-dev] [PATCH] MS compatibility flag implies delayed parsing

[Francois Pichet]

On Mon, Nov 7, 2011 at 5:55 PM, Eli Friedman <eli.friedman at gmail.com> wrote:
> On Sun, Nov 6, 2011 at 3:39 PM, Aaron Ballman <aaron at aaronballman.com> wrote:
>> On Sun, Nov 6, 2011 at 3:53 PM, Aaron Ballman <aaron at aaronballman.com> wrote:
>>> On Sun, Nov 6, 2011 at 2:53 PM, Eli Friedman <eli.friedman at gmail.com> wrote:
>>>> On Sun, Nov 6, 2011 at 12:12 PM, Aaron Ballman <aaron at aaronballman.com> wrote:
>>>>> There was a small discussion on IRC about the
>>>>> -fdelayed-template-parsing flag always going hand-in-hand with the
>>>>> -fms-compatibility flag.  Since the implication is already there (you
>>>>> don't really have ms compatibility without the delayed template
>>>>> parsing, to a certain extent), this patch makes it more explicit.  So
>>>>> when you pass the fms-compatibility flag, it automatically enables the
>>>>> fdelayed-template-parsing flag as well.  If the user really does want
>>>>> one without the other, they can always pass -fms-compatibility
>>>>> -fnodelayed-template-parsing.
>>>>
>>>> Concept is fine; the patch doesn't actually implement what you say it
>>>> does, though.  Specifically, -fms-compatibility
>>>> -fnodelayed-template-parsing won't work with this patch.
>>
>> Here is a second attempt at the patch -- this time it works the way I
>> wanted it to work!  Thanks for the review.
>
> I don't see why the changes to lib/Frontend/CompilerInvocation.cpp and
> include/clang/Driver/CC1Options.td are necessary.  The changes to
> lib/Driver/Tools.cpp look fine.
>
> -Eli
>

How about we remove --fdelayed-template-parsing altogether and check
for fms-compatibility to do late template parsing?

Anybody else needs delayed template parsing without microsoft compatibility?