[PATCH] D118352: [clang][ABI] New c++20 modules mangling scheme

Iain Sandoe via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Mon Mar 7 08:12:33 PST 2022


iains added a comment.

In D118352#3363909 <https://reviews.llvm.org/D118352#3363909>, @urnathan wrote:

> In D118352#3362694 <https://reviews.llvm.org/D118352#3362694>, @ChuanqiXu wrote:
>
>> In D118352#3359626 <https://reviews.llvm.org/D118352#3359626>, @urnathan wrote:
>>
>>> 
>
>
>
>>> Correct, it is not called as the global initializer pieces are not yet implemented.  Let's not hold up this patch for that nor remove piece that will become necessary at that point.
>>
>> It's odd to see unused functions. I just worry about that other people might delete these functions as cleaning up (maybe there wouldn't be).
>
> That's a risk I'm ok with.

I agree - I think we already a fair amount of infrastructure in the compiler to handle module initialisers, ahead of a full implementation - so I do not see a reason for waiting longer to apply this because we cannot yet exercise it full.

>> Although the general declarations in partitions wouldn't be mangled specially, I think it would be better to add test case to show that.
>
> good idea, added.




CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D118352/new/

https://reviews.llvm.org/D118352



More information about the cfe-commits mailing list