[PATCH] AArch64: Enable subregister liveness tracking.

[Eric Christopher]

On Thu, Mar 19, 2015 at 4:33 PM Matthias Braun <matze at braunis.de> wrote:

> Hi t.p.northover, qcolombet,
>
> Splitting this into an own commit to get some visibility.
>
> After doing some extensive benchmarking wiht llvm test-suite, I see no
> correctness or compiletime changes. I see some very slight benchmark
> runtime improvements (which may or may not be noise). The only noticeable
> regression is in a benchmark were literally 6 lines of assembly changed for
> the better, so that must be cache issues.
>
> REPOSITORY
>   rL LLVM
>
> http://reviews.llvm.org/D8474
>
> Files:
>   lib/Target/AArch64/AArch64Subtarget.h
>
> Index: lib/Target/AArch64/AArch64Subtarget.h
> ===================================================================
> --- lib/Target/AArch64/AArch64Subtarget.h
> +++ lib/Target/AArch64/AArch64Subtarget.h
> @@ -91,6 +91,7 @@
>    bool enablePostMachineScheduler() const override {
>      return isCortexA53() || isCortexA57();
>    }
> +  bool enableSubRegLiveness() const override { return true; }
>
>
Drive by feedback:

Most of AArch64 is option controlled for this sort of thing. Did you want
to add an option with init(true) for this?

Personally no preference, but thought I'd bring it up.

-eric


>    bool hasZeroCycleRegMove() const { return HasZeroCycleRegMove; }
>
> EMAIL PREFERENCES
>   http://reviews.llvm.org/settings/panel/emailpreferences/
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150319/b5b6ca72/attachment.html>