[PATCH] Introduce llvm/ADT/OptionSet.h utility class

[Chandler Carruth via llvm-commits]

On Fri, Feb 12, 2016 at 3:58 PM Chris Lattner <clattner at apple.com> wrote:

> On Feb 12, 2016, at 2:29 PM, Philip Reames <listmail at philipreames.com>
> wrote:
>
> If so, that would resolve the licensing concern.  In the future, let's
> make sure that gets mentioned in the review/commit thread to avoid
> confusion.
>
>
> I can understand your concern, but for better or worse, we don’t ask llvm
> contributors to state the provenance of their code that they are posting.
> If you’re asking for some new rule to be put in place, please specify what
> the rule is and what the rationale for that rule is.
>

I don't want to speak for Philip, but I think the thing that made this a
bit different was the explicit statement that the code came from some
particular source (a different open source project in this case) and that
triggered a concern about whether it was reasonable to contribute it. That
doesn't seem unreasonable.

For example, when someone contributed a patch from the GCC fork of the
sanitizer runtimes, we asked similar questions to what Philip has asked
here because the statement that the patch came from somewhere else seemed
directly in conflict  with the contributor being able to correctly
contribute it to LLVM.

That pattern might be a reasonable basis for new guidelines, or might not.
I'm not really trying to have an opinion about that, just giving some other
context.

-Chandler
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20160213/b6244809/attachment.html>