[llvm] r210641 - Create macro INITIALIZE_TM_PASS.

[Eric Christopher]

On Fri, Jun 13, 2014 at 4:25 PM, Jiangning Liu <liujiangning1 at gmail.com> wrote:
> Oops! I just committed the patch of moving GlobalMerge pass to CodeGen.
>
> Eric, sorry I didn't see your email early. If you are not happy with this
> change, I can revert both of the patch adding macro INITIALIZE_TM_PASS and
> this patch.
>
> For me, I think Candler's explanation is clear, and at least at this moment,
> there isn't pass depending on TargetMachine under lib/Transform.
>

Nope, I'm fine. Thanks.

-eric

> Thanks,
> -Jiangning
>
>
> 2014-06-14 6:46 GMT+08:00 Rafael Avila de Espindola
> <rafael.espindola at gmail.com>:
>>
>>
>>
>> Sent from my iPhone
>>
>> > On Jun 13, 2014, at 18:32, Eric Christopher <echristo at gmail.com> wrote:
>> >
>> >> On Fri, Jun 13, 2014 at 3:28 PM, Chandler Carruth
>> >> <chandlerc at google.com> wrote:
>> >>
>> >> On Fri, Jun 13, 2014 at 11:02 PM, Rafael Avila de Espindola
>> >> <rafael.espindola at gmail.com> wrote:
>> >>>
>> >>>> On Jun 13, 2014, at 17:26, Eric Christopher <echristo at gmail.com>
>> >>>> wrote:
>> >>>>
>> >>>> Ack. This is a terrible idea. (As there's a thread on it)
>> >>>>
>> >>>> But I'm not sure what the current status is here, did this patch get
>> >>>> reverted?
>> >>>
>> >>>
>> >>> I don't think so. My understanding is that Jiangning Liu is working to
>> >>> move the pass to lib/CodeGen. Would that make it ok for you or there
>> >>> are
>> >>> other issues besides the layering problem?
>> >>
>> >>
>> >> I'm not sure the nature of the problem Eric is hitting, but my problems
>> >> are
>> >> well addressed by the patch already mailed out on this thread. For me,
>> >> that
>> >> would be a step in the right directoin.
>> >
>> > Oh, I'm just going to have to rewrite it to avoid taking a TM, but
>> > this isn't making it any worse afaict.
>> >
>>
>> Because it is a module pass and you are changing each function to have a
>> different one?
>>
>>
>> > -eric
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
>