[Bug target/64821] New: [AArch64] Improve target folding for vsqrt_f64 intrinsic

[Bug target/64821] New: [AArch64] Improve target folding for vsqrt_f64 intrinsic

[ktkachov at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

            Bug ID: 64821
           Summary: [AArch64] Improve target folding for vsqrt_f64
                    intrinsic
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Keywords: missed-optimization
          Severity: minor
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ktkachov at gcc dot gnu.org

Following the implementation of vsqrt_f64 with a target builtin:
https://gcc.gnu.org/ml/gcc-patches/2015-01/msg00699.html

it was suggested to add some target folding code into a sqrt when
-fno-math-errno is given:
https://gcc.gnu.org/ml/gcc-patches/2015-01/msg00710.html

This issue tracks that work that should be done for GCC 6

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-02-16
     Ever confirmed|0                           |1
           Severity|minor                       |enhancement

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|2015-02-16 00:00:00         |2022-1-9

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Here is a testcase:
  #include <arm_neon.h>
  float64x1_t f(float64x1_t a)
  {
      return vsqrt_f64 (a);
  }

We should fold the __builtin_aarch64_sqrtdf into the internal function .SQRT
which is indepdent of -fno-math-errno .

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |pinskia at gcc dot gnu.org

--- Comment #4 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I am going to fix this but maybe not until stage 1.

I think this is enough too:
diff --git a/gcc/config/aarch64/aarch64-builtins.c
b/gcc/config/aarch64/aarch64-builtins.c
index 58bcbd9875f..1bf487477eb 100644
--- a/gcc/config/aarch64/aarch64-builtins.c
+++ b/gcc/config/aarch64/aarch64-builtins.c
@@ -2820,6 +2820,13 @@ aarch64_general_gimple_fold_builtin (unsigned int fcode,
gcall *stmt,
        gimple_call_set_lhs (new_stmt, gimple_call_lhs (stmt));
        break;

+      /* Lower sqrt builtins to gimple/internal function sqrt. */
+      BUILTIN_VHSDF_DF (UNOP, sqrt, 2, FP)
+       new_stmt = gimple_build_call_internal (IFN_SQRT,
+                                              1, args[0]);
+       gimple_call_set_lhs (new_stmt, gimple_call_lhs (stmt));
+       break;
+
      /*lower store and load neon builtins to gimple.  */
      BUILTIN_VALL_F16 (LOAD1, ld1, 0, LOAD)
      BUILTIN_VDQ_I (LOAD1_U, ld1, 0, LOAD)

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

--- Comment #5 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #4)
> I am going to fix this but maybe not until stage 1.
> 
> I think this is enough too:

It is.

A testcase too:
/* { dg-do compile } */
/* { dg-options "-fdump-tree-optimized" } */
#include <arm_neon.h>
/* { dg-final { scan-tree-dump-times "__builtin_aarch64" 0 "optimized" } } */
/* { dg-final { scan-tree-dump-times ".SQRT" 4 "optimized" } } */

float64x1_t f64(float64x1_t a)
{
      return vsqrt_f64 (a);
}

float64x2_t f64q(float64x2_t a)
{
      return vsqrtq_f64 (a);
}

float32x2_t f32(float32x2_t a)
{
      return vsqrt_f32 (a);
}

float32x4_t f32q(float32x4_t a)
{
      return vsqrtq_f32 (a);
}

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

--- Comment #6 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
One more testcase where we get folding now:
/* { dg-do compile } */
/* { dg-options "-O2 -fdump-tree-optimized" } */
/* { dg-final { scan-tree-dump-times ".SQRT" 0 "optimized" } } */
/* { dg-final { scan-tree-dump-times "2.0" 1 "optimized" } } */
/* { dg-final { scan-assembler-times "fsqrt" 0 } } */
/* We should produce a fmov to d0 with 2.0 but currently don't, see PR 103959.
*/
/* { dg-final { scan-assembler-times "\n\tfmov\td0, 2.0e.0" 1 { xfail *-*-* } }
} */

#include <arm_neon.h>

float64x1_t f64(void)
{
   float64x1_t a = (float64x1_t){4.0};
   return vsqrt_f64 (a);
}

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Created attachment 52151
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=52151&action=edit
Patch which is in testing

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch
                URL|                            |https://gcc.gnu.org/piperma
                   |                            |il/gcc-patches/2022-January
                   |                            |/588012.html

--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Patch submitted:
https://gcc.gnu.org/pipermail/gcc-patches/2022-January/588012.html

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

--- Comment #9 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Andrew Pinski <pinskia@gcc.gnu.org>:

https://gcc.gnu.org/g:cbcf4a50fa21abd7a4a50a7ce47ada80b115febc

commit r12-6824-gcbcf4a50fa21abd7a4a50a7ce47ada80b115febc
Author: Andrew Pinski <apinski@marvell.com>
Date:   Sun Jan 9 23:39:31 2022 -0800

    [aarch64/64821]: Simplify __builtin_aarch64_sqrt* into internal function
.SQRT.

    This is a simple patch which simplifies the __builtin_aarch64_sqrt*
builtins
    into the internal function SQRT which allows for constant folding and other
    optimizations at the gimple level. It was originally suggested we do to
    __builtin_sqrt just for __builtin_aarch64_sqrtdf when -fno-math-errno
    but since r6-4969-g686ee9719a4 we have the internal function SQRT which
does
    the same so it makes we don't need to check -fno-math-errno either now.

    Applied as approved after bootstrapped and tested on aarch64-linux-gnu with
no regressions.

            PR target/64821

    gcc/ChangeLog:

            * config/aarch64/aarch64-builtins.cc
            (aarch64_general_gimple_fold_builtin): Handle
            __builtin_aarch64_sqrt* and simplify into SQRT internal
            function.

    gcc/testsuite/ChangeLog:

            * gcc.target/aarch64/vsqrt-1.c: New test.
            * gcc.target/aarch64/vsqrt-2.c: New test.

[Bug target/64821] [AArch64] Improve target folding for vsqrt_f64 intrinsic

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64821

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED
   Target Milestone|---                         |12.0

--- Comment #10 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed.