[llvm] r182877 - Enable FastISel on ARM for Linux and NaCl

[Rafael Espíndola]

Sorry, I should have replied to this thread too. My main concern is to
keep the bots green, so that any other failure is noticed.

If it is easy, it is probably ok to apply you patch again but keeping
mcjit with the regular instruction selection until we find why mcjit
and arm fast isel don't like each other.

On 30 May 2013 19:17, JF Bastien <jfb at google.com> wrote:
> Ah yes, I should tune my email filters some more. Or you know, read emails.
>
>
> On Thu, May 30, 2013 at 3:50 PM, Eric Christopher <echristo at gmail.com>
> wrote:
>>
>> Looks like someone already got it?
>>
>> -eric
>>
>> On Thu, May 30, 2013 at 3:44 PM, JF Bastien <jfb at google.com> wrote:
>> > I'll look into this and other failure
>> > (http://llvm.org/bugs/show_bug.cgi?id=16178), but as I'm not familiar
>> > with
>> > MCJIT I don't know how long this will take.
>> >
>> > Would you rather have me revert the ARM FastISel change?
>> >
>> >
>> > On Thu, May 30, 2013 at 2:21 PM, Renato Golin <renato.golin at linaro.org>
>> > wrote:
>> >>
>> >> Hi Rafael,
>> >>
>> >> Thanks for looking into this, I have been scratching my head to find
>> >> someone to blame for the bots... ;)
>> >>
>> >> Glad that you guys are on top of this, MCJIT has been very unreliable
>> >> on
>> >> ARM recently, but I take it there were many changes and will take a
>> >> while to
>> >> stabilize it properly.
>> >>
>> >> I have isolated the remote tests earlier to make it easier to turn them
>> >> on/off for some targets, maybe we need more divisions, at least until
>> >> this
>> >> is that unstable. I'll leave the decision to you guys sort it out.
>> >>
>> >> cheers,
>> >> --renato
>> >>
>> >>
>> >>
>> >> On 30 May 2013 22:11, Rafael Espíndola <rafael.espindola at gmail.com>
>> >> wrote:
>> >>>
>> >>> On 30 May 2013 16:52, JF Bastien <jfb at google.com> wrote:
>> >>> > Could you point me at the failure? MCJIT should work with FastISel
>> >>> > when
>> >>> > the
>> >>> > target is Darwin, so it's probably exposing another issue with
>> >>> > MCJIT+FastISel+non-Darwin.
>> >>>
>> >>> For example:
>> >>>
>> >>>
>> >>>
>> >>> http://lab.llvm.org:8011/builders/clang-native-arm-cortex-a9-self-host/builds/79/steps/check-all/logs/LLVM%3A%3Apr13727.ll
>> >>>
>> >>> Cheers,
>> >>> Rafael
>> >>
>> >>
>> >
>> >
>> > _______________________________________________
>> > llvm-commits mailing list
>> > llvm-commits at cs.uiuc.edu
>> > http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>> >
>
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>