llvm dev - Mar 2017 - [GlobalISel][AArch64] Toward flipping the switch for O0: Please give it a try!

Hi Renato,

If Kristof is busy I can make runs on AArch64 Linux (Cortex-A53 and Cortex-57).

Thanks,
Evgeny Astigeevich
Senior Compiler Engineer
Compilation Tools
ARM
> -----Original Message-----
> From: llvm-dev [mailto:llvm-dev-bounces at lists.llvm.org] On Behalf Of
> Renato Golin via llvm-dev
> Sent: Thursday, March 30, 2017 9:54 AM
> To: Quentin Colombet
> Cc: llvm-dev; Justin Bogner; Ahmed Bougacha; Aditya Nandakumar
> Subject: Re: [llvm-dev] [GlobalISel][AArch64] Toward flipping the switch
for
> O0: Please give it a try!
> 
> On 30 March 2017 at 00:27, Quentin Colombet <qcolombet at apple.com>
> wrote:
> > On iOS we are at 100% pass rate in 00 g for the LLVM test suite,
> > standard benchmarks and unit tests. In about 5% of all functions
> > GlobalIsel falls back to SDIsel.
> > (Kristof Beyls would have the linux numbers.) The self host compiler
> > correctly builds and runs the LLVM test suite in O0.
> 
> Having done no tests at all on my side, I think we need to have similar
> numbers on Linux to be able to flip across the board.
> 
> I don't want to flip it only for Darwin and not Linux, as that will
fragment the
> effort too much.
> 
> I'll check with Diana and Kristof to know what's the best way
forward, but it
> should be reasonably quick.
> 
> cheers,
> --renato
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

Hi,

Just a heads-up for anyone building things with CMake: if you use the
default build type (Release), it will put CMAKE_CXX_FLAGS_RELEASE
after the CMAKE_CXX_FLAGS, so you'll end up running -O3 instead of
-O0. You might want to check your command lines.

Kristof has also found an issue in the test-suite where it actually
runs -O2 due to some weirdness in the build system, we'll try to
commit a fix for that.

Cheers,
Diana

On 30 March 2017 at 16:13, Evgeny Astigeevich via llvm-dev
<llvm-dev at lists.llvm.org> wrote:
> Hi Renato,
>
> If Kristof is busy I can make runs on AArch64 Linux (Cortex-A53 and
Cortex-57).
>
> Thanks,
> Evgeny Astigeevich
> Senior Compiler Engineer
> Compilation Tools
> ARM
>
>> -----Original Message-----
>> From: llvm-dev [mailto:llvm-dev-bounces at lists.llvm.org] On Behalf Of
>> Renato Golin via llvm-dev
>> Sent: Thursday, March 30, 2017 9:54 AM
>> To: Quentin Colombet
>> Cc: llvm-dev; Justin Bogner; Ahmed Bougacha; Aditya Nandakumar
>> Subject: Re: [llvm-dev] [GlobalISel][AArch64] Toward flipping the
switch for
>> O0: Please give it a try!
>>
>> On 30 March 2017 at 00:27, Quentin Colombet <qcolombet at
apple.com>
>> wrote:
>> > On iOS we are at 100% pass rate in 00 g for the LLVM test suite,
>> > standard benchmarks and unit tests. In about 5% of all functions
>> > GlobalIsel falls back to SDIsel.
>> > (Kristof Beyls would have the linux numbers.) The self host
compiler
>> > correctly builds and runs the LLVM test suite in O0.
>>
>> Having done no tests at all on my side, I think we need to have similar
>> numbers on Linux to be able to flip across the board.
>>
>> I don't want to flip it only for Darwin and not Linux, as that will
fragment the
>> effort too much.
>>
>> I'll check with Diana and Kristof to know what's the best way
forward, but it
>> should be reasonably quick.
>>
>> cheers,
>> --renato
>> _______________________________________________
>> LLVM Developers mailing list
>> llvm-dev at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

Hi Diana,

Thank you for information. Yes, I know about the problem with O options.
I've been hit by this whilst I was testing my changes for Oz. 

Thanks,
Evgeny
> -----Original Message-----
> From: Diana Picus [mailto:diana.picus at linaro.org]
> Sent: Thursday, March 30, 2017 4:44 PM
> To: Evgeny Astigeevich
> Cc: Renato Golin; llvm-dev; nd
> Subject: Re: [llvm-dev] [GlobalISel][AArch64] Toward flipping the switch
for
> O0: Please give it a try!
> 
> Hi,
> 
> Just a heads-up for anyone building things with CMake: if you use the
default
> build type (Release), it will put CMAKE_CXX_FLAGS_RELEASE after the
> CMAKE_CXX_FLAGS, so you'll end up running -O3 instead of -O0. You might
> want to check your command lines.
> 
> Kristof has also found an issue in the test-suite where it actually runs
-O2 due
> to some weirdness in the build system, we'll try to commit a fix for
that.
> 
> Cheers,
> Diana
> 
> On 30 March 2017 at 16:13, Evgeny Astigeevich via llvm-dev <llvm-
> dev at lists.llvm.org> wrote:
> > Hi Renato,
> >
> > If Kristof is busy I can make runs on AArch64 Linux (Cortex-A53 and
Cortex-
> 57).
> >
> > Thanks,
> > Evgeny Astigeevich
> > Senior Compiler Engineer
> > Compilation Tools
> > ARM
> >
> >> -----Original Message-----
> >> From: llvm-dev [mailto:llvm-dev-bounces at lists.llvm.org] On
Behalf Of
> >> Renato Golin via llvm-dev
> >> Sent: Thursday, March 30, 2017 9:54 AM
> >> To: Quentin Colombet
> >> Cc: llvm-dev; Justin Bogner; Ahmed Bougacha; Aditya Nandakumar
> >> Subject: Re: [llvm-dev] [GlobalISel][AArch64] Toward flipping the
> >> switch for
> >> O0: Please give it a try!
> >>
> >> On 30 March 2017 at 00:27, Quentin Colombet <qcolombet at
apple.com>
> >> wrote:
> >> > On iOS we are at 100% pass rate in 00 g for the LLVM test
suite,
> >> > standard benchmarks and unit tests. In about 5% of all
functions
> >> > GlobalIsel falls back to SDIsel.
> >> > (Kristof Beyls would have the linux numbers.) The self host
> >> > compiler correctly builds and runs the LLVM test suite in O0.
> >>
> >> Having done no tests at all on my side, I think we need to have
> >> similar numbers on Linux to be able to flip across the board.
> >>
> >> I don't want to flip it only for Darwin and not Linux, as that
will
> >> fragment the effort too much.
> >>
> >> I'll check with Diana and Kristof to know what's the best
way
> >> forward, but it should be reasonably quick.
> >>
> >> cheers,
> >> --renato
> >> _______________________________________________
> >> LLVM Developers mailing list
> >> llvm-dev at lists.llvm.org
> >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
> > _______________________________________________
> > LLVM Developers mailing list
> > llvm-dev at lists.llvm.org
> > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev

On 30 March 2017 at 08:44, Diana Picus via llvm-dev
<llvm-dev at lists.llvm.org> wrote:
> Just a heads-up for anyone building things with CMake: if you use the
> default build type (Release), it will put CMAKE_CXX_FLAGS_RELEASE
> after the CMAKE_CXX_FLAGS, so you'll end up running -O3 instead of
> -O0. You might want to check your command lines.
It does actually work if you add "-fno-vectorize -fno-slp-vectorize"
to CMAKE_CXX_FLAGS_RELASE (and C). Not a standard configuration, but
probably throwing more varied IR at it than just doing -O0.

Cheers.

Tim.