[libcxx-commits] [PATCH] D141208: libc++: bring back the unsigned in the return type in wcstoull_l

[Nikolas Klauser via Phabricator via libcxx-commits]

philnik added a comment.

In D141208#4035450 <https://reviews.llvm.org/D141208#4035450>, @sylvestre.ledru wrote:

> @philnik Sorry but we don't really have a way to know who is the "libc++ team". I don't know if @brad is part of it or not
>
>> provide a CI bot, which is the case here currently.
>
> I am running it for https://apt.llvm.org/ and happy to open bugs in case of regression

They are the people who are part of the #libc <https://reviews.llvm.org/tag/libc/> review group, which gets automatically added as a blocking reviewer on libc++ patches.

In D141208#4035239 <https://reviews.llvm.org/D141208#4035239>, @thesamesam wrote:

> In D141208#4034019 <https://reviews.llvm.org/D141208#4034019>, @philnik wrote:
>
>> In D141208#4034013 <https://reviews.llvm.org/D141208#4034013>, @sylvestre.ledru wrote:
>>
>>> In D141208#4033997 <https://reviews.llvm.org/D141208#4033997>, @brad wrote:
>>>
>>>> Sorry about that.
>>>
>>> no worries ;)
>>
>> Please don't land libc++ patches unless approved by the libc++ team. I won't revert, but this code will be removed in the future unless we find someone who is willing to support the musl configuration in libc++. Our policy is to not accept patches where nobody is willing to provide a CI bot, which is the case here currently.
>
> Could you show me where that policy is written? I wasn't aware of it.

I'm not sure it's written down explicitly anywhere, but we should definitely update our documentation. I guess part of the problem is that we still have multiple configurations in our code base which we have neither removed nor got a CI bot.

> This issue wasn't raised in D124227 <https://reviews.llvm.org/D124227> either.

Yeah, I' not sure why.

In D141208#4035242 <https://reviews.llvm.org/D141208#4035242>, @mgorny wrote:

>> Our policy is to not accept patches where nobody is willing to provide a CI bot, which is the case here currently.
>
> That "policy" is really against the principles of open source. Corporations have money to maintain CI bots. Small open source projects can't afford the huge cost of running CI for libc++.

I'm not sure that this is against the principles of open source, though I don't know what these principles are in your opinion. Yes, this makes it a lot harder for small open source projects, but it's a lot worse to get aware of an issue literally half a year after it has been introduced, while it could have quite easily been detected before the patch has landed.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141208/new/

https://reviews.llvm.org/D141208