[llvm-dev] Speculative FP operation.
Umesh Kalappa via llvm-dev
We have a situation ,where the floating operations were hoisted
speculative from a conditional block on ARM64.
Was using the option "-mllvm --speculate-fp-insts=false" but doesn't
have any effect from LLVM 12 .
and "-ftrapping-math" is not enabled for aarch64 .
So question to the expert ,while investigating the passes we found
that simplifyCFG does this transformation and before we investigate
and prevent this transformation .
Do the group recommend any other options here ?
Thank you
~Umesh
_______________________________________________
LLVM Developers mailing list
llvm...@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
Umesh Kalappa via llvm-dev
Did someone tried enabling the trapping math for the aarch64 arch ?
So that we can surpass the transformation like below.
If not ,we would like to take it up for ARM arch.
Thank you
~Umesh
Florian Hahn via llvm-dev
> On Jan 12, 2022, at 04:50, Umesh Kalappa via llvm-dev <llvm...@lists.llvm.org> wrote:
>
> Hi All,
>
> Did someone tried enabling the trapping math for the aarch64 arch ?
> So that we can surpass the transformation like below.
>
As far as I know, initial trapping math support only existing for X86 at the moment.
> If not ,we would like to take it up for ARM arch.
I am not aware of anybody working on trapping math for AArch64, so any work in this area could be very valuable. I am CC’ing a couple of people who might know more about the current state of trapping math in Clang.
Cheers,
Florian
John Brawn via llvm-dev
Sent: 12 January 2022 15:16
To: Umesh Kalappa <umesh.k...@gmail.com>; cfe-dev <cfe...@lists.llvm.org>
Cc: Kaylor, Andrew via llvm-dev <llvm...@lists.llvm.org>; Blower, Melanie I via cfe-dev <cfe...@lists.llvm.org>
Subject: Re: [cfe-dev] [llvm-dev] Speculative FP operation.
> On Jan 12, 2022, at 04:50, Umesh Kalappa via llvm-dev <llvm...@lists.llvm.org> wrote:
>
> Hi All,
>
> Did someone tried enabling the trapping math for the aarch64 arch ?
> So that we can surpass the transformation like below.
>
As far as I know, initial trapping math support only existing for X86 at the moment.
> If not ,we would like to take it up for ARM arch.
I am not aware of anybody working on trapping math for AArch64, so any work in this area could be very valuable. I am CC’ing a couple of people who might know more about the current state of trapping math in Clang.
Cheers,
Florian
_______________________________________________
Kevin P. Neal via llvm-dev
> Hi,
>
> > On Jan 12, 2022, at 04:50, Umesh Kalappa via llvm-dev <llvm...@lists.llvm.org> wrote:
> >
> > Hi All,
> >
> > Did someone tried enabling the trapping math for the aarch64 arch ?
> > So that we can surpass the transformation like below.
> >
>
> As far as I know, initial trapping math support only existing for X86 at the moment.
X86, PowerPC, SystemZ.
I believe PowerPC is the only target that properly handles dynamic rounding
mode, where instructions don't get moved past a call instruction since we
don't know if the rounding mode will be changed. I might have missed the
fix for the other targets, though.
--
Kevin P. Neal http://www.pobox.com/~kpn/
Seen on bottom of IBM part number 1887724:
DO NOT EXPOSE MOUSE PAD TO DIRECT SUNLIGHT FOR EXTENDED PERIODS OF TIME.
Florian Hahn via llvm-dev
On Jan 12, 2022, at 16:23, John Brawn <John....@arm.com> wrote:I'm currently working on AArch64 strict fp support (which is what's needed to enable -ftrapping-math).There's a few patches currently in review:https://reviews.llvm.org/D115352 (though I need to rebase this)https://reviews.llvm.org/D115620 (needs some extra work to enable a currently xfailed test)And I'm working on some other patches:Miscellaneous fixes for vector instruction loweringModel FPCR use in AArch64 instructionsFixes related to GlobalISel (not strictly needed as GlobalISel falls back to default isel when it can't handle things)Once all of the above is done I think that I can then mark AArch64 as supporting strict fp in clang, so that -ftrapping-math isn't ignored.