[PATCH] D17444: PR26672: [MSVC] Clang does not recognize "static_assert" keyword in C mode
Aaron Ballman via cfe-commits
cfe-commits at lists.llvm.org
Thu Mar 3 10:40:22 PST 2016
On Thu, Mar 3, 2016 at 12:06 PM, David Majnemer
<david.majnemer at gmail.com> wrote:
> On Thu, Mar 3, 2016 at 7:52 AM, Aaron Ballman <aaron at aaronballman.com>
>> On Thu, Mar 3, 2016 at 10:43 AM, Andrey Bokhanko
>> <andreybokhanko at gmail.com> wrote:
>> > Now I'm completely confused... :-)
>> > Can we rely that this MS engineer has enough authority to declare this
>> > to be a bug?
>> He's on the compiler team, so yes.
>> > And more importantly -- is MS willing to fix the [supposed] bug in
>> > future MSVC compilers? I frankly don't think so...
>> We can only guess what Microsoft is going to do and when. However, we
>> do not aim to be bug for bug compatible with MSVC unless there's very
>> good reason to do so, and I've yet to see any justification to warrant
>> that for this bug. Further, there's additional burden. If Microsoft
>> fixes this bug, are we going to then update our code to only support
>> it in a certain range of -fms-compatibility-version values? Or are we
>> going to have a compatibility hack that Microsoft doesn't have? etc.
> Here is my thinking:
> If there is code out there in a wild that depends on this, then we should
> take this patch.
That was what I meant by "justification". I would say it has to be
reasonably compelling code (win32 headers, boost, some other major
library) as that's our usual bar for these sort of bug-for-bug
compatible things, as I understand it.
> Once Microsoft releases a compiler which supports the conforming behavior,
> we should limit it's scope to a specific range of versions.
> We've done this in the past for things like _Atomic.
Agreed, we have a way forward if we need it. I mostly just want to
avoid the burden of supporting that because this is sufficiently weird
(being a non-conforming keyword).
>> > Either way, it's up to Reid to decide what to do. Reid?
>> It is not solely up to Reid to decide what to do. It's up to us as a
>> community as to what we want to support and maintain long-term and
>> that requires some amount of consensus. Also, Reid is not the only
>> Windows maintainer. ;-)
More information about the cfe-commits