procyberian/gcc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

OMP SIMD inbranch call vectorization for AVX512 style masks

Browse source 
The following teaches vectorizable_simd_clone_call to handle
integer mode masks.  The tricky bit is to second-guess the
number of lanes represented by a single mask argument - the following
uses simdlen and the number of mask arguments to calculate that,
assuming ABIs have them uniform.

Similar to the VOIDmode handling there's a restriction on not
supporting splitting/merging of incoming vector masks to
more/less SIMD call arguments.

	PR tree-optimization/111795
	* tree-vect-stmts.cc (vectorizable_simd_clone_call): Handle
	integer mode mask arguments.

	* gcc.target/i386/vect-simd-clone-avx512-1.c: New testcase.
	* gcc.target/i386/vect-simd-clone-avx512-2.c: Likewise.
	* gcc.target/i386/vect-simd-clone-avx512-3.c: Likewise.
This commit is contained in:
Richard Biener 2023-10-13 12:32:51 +02:00
parent 63eaccd114
commit 3179ad72f6
4 changed files with 175 additions and 30 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

43
gcc/testsuite/gcc.target/i386/vect-simd-clone-avx512-1.c Normal file
View file

@ -0,0 +1,43 @@
/* { dg-do run } */
/* { dg-require-effective-target avx512vl } */
/* { dg-options "-O2 -fopenmp-simd -mavx512vl" } */
#include "avx512vl-check.h"
#ifndef SIMDLEN
#define SIMDLEN 4
#endif
int x[1024];
#pragma omp declare simd simdlen(SIMDLEN)
__attribute__((noinline)) int
foo (int a, int b)
{
return a + b;
}
void __attribute__((noipa))
bar (void)
{
#pragma omp simd
for (int i = 0; i < 1024; i++)
if (x[i] < 20)
x[i] = foo (x[i], x[i]);
}
void avx512vl_test ()
{
int i;
#pragma GCC novector
for (i = 0; i < 1024; i++)
x[i] = i;
bar ();
#pragma GCC novector
for (i = 0; i < 1024; i++)
if ((i < 20 && x[i] != i + i)
|| (i >= 20 && x[i] != i))
abort ();
}

6
gcc/testsuite/gcc.target/i386/vect-simd-clone-avx512-2.c Normal file
View file

@ -0,0 +1,6 @@
/* { dg-do run } */
/* { dg-require-effective-target avx512vl } */
/* { dg-options "-O2 -fopenmp-simd -mavx512vl" } */
#define SIMDLEN 8
#include "vect-simd-clone-avx512-1.c"

6
gcc/testsuite/gcc.target/i386/vect-simd-clone-avx512-3.c Normal file
View file

@ -0,0 +1,6 @@
/* { dg-do run } */
/* { dg-require-effective-target avx512vl } */
/* { dg-options "-O2 -fopenmp-simd -mavx512vl" } */
#define SIMDLEN 16
#include "vect-simd-clone-avx512-1.c"

150
gcc/tree-vect-stmts.cc
View file

@ -4385,6 +4385,9 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
i = -1;
break;
case SIMD_CLONE_ARG_TYPE_MASK:
if (SCALAR_INT_MODE_P (n->simdclone->mask_mode)
!= SCALAR_INT_MODE_P (TYPE_MODE (arginfo[i].vectype)))
i = -1;
break;
}
if (i == (size_t) -1)
@ -4410,6 +4413,12 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
if (bestn == NULL)
return false;
unsigned int num_mask_args = 0;
if (SCALAR_INT_MODE_P (bestn->simdclone->mask_mode))
for (i = 0; i < nargs; i++)
if (bestn->simdclone->args[i].arg_type == SIMD_CLONE_ARG_TYPE_MASK)
num_mask_args++;
for (i = 0; i < nargs; i++)
{
if ((arginfo[i].dt == vect_constant_def
@ -4434,30 +4443,50 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
return false;
}
if (bestn->simdclone->args[i].arg_type == SIMD_CLONE_ARG_TYPE_MASK
&& bestn->simdclone->mask_mode == VOIDmode
&& (simd_clone_subparts (bestn->simdclone->args[i].vector_type)
!= simd_clone_subparts (arginfo[i].vectype)))
if (bestn->simdclone->args[i].arg_type == SIMD_CLONE_ARG_TYPE_MASK)
{
/* FORNOW we only have partial support for vector-type masks that
can't hold all of simdlen. */
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION,
vect_location,
"in-branch vector clones are not yet"
" supported for mismatched vector sizes.\n");
return false;
}
if (bestn->simdclone->args[i].arg_type == SIMD_CLONE_ARG_TYPE_MASK
&& bestn->simdclone->mask_mode != VOIDmode)
{
/* FORNOW don't support integer-type masks. */
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION,
vect_location,
"in-branch vector clones are not yet"
" supported for integer mask modes.\n");
return false;
if (bestn->simdclone->mask_mode == VOIDmode)
{
if (simd_clone_subparts (bestn->simdclone->args[i].vector_type)
!= simd_clone_subparts (arginfo[i].vectype))
{
/* FORNOW we only have partial support for vector-type masks
that can't hold all of simdlen. */
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION,
vect_location,
"in-branch vector clones are not yet"
" supported for mismatched vector sizes.\n");
return false;
}
}
else if (SCALAR_INT_MODE_P (bestn->simdclone->mask_mode))
{
if (!SCALAR_INT_MODE_P (TYPE_MODE (arginfo[i].vectype))
|| maybe_ne (exact_div (bestn->simdclone->simdlen,
num_mask_args),
simd_clone_subparts (arginfo[i].vectype)))
{
/* FORNOW we only have partial support for integer-type masks
that represent the same number of lanes as the
vectorized mask inputs. */
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION,
vect_location,
"in-branch vector clones are not yet "
"supported for mismatched vector sizes.\n");
return false;
}
}
else
{
if (dump_enabled_p ())
dump_printf_loc (MSG_MISSED_OPTIMIZATION,
vect_location,
"in-branch vector clones not supported"
" on this target.\n");
return false;
}
}
}
@ -4674,14 +4703,9 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
}
break;
case SIMD_CLONE_ARG_TYPE_MASK:
atype = bestn->simdclone->args[i].vector_type;
if (bestn->simdclone->mask_mode != VOIDmode)
{
/* FORNOW: this is disabled above. */
gcc_unreachable ();
}
else
if (bestn->simdclone->mask_mode == VOIDmode)
{
atype = bestn->simdclone->args[i].vector_type;
tree elt_type = TREE_TYPE (atype);
tree one = fold_convert (elt_type, integer_one_node);
tree zero = fold_convert (elt_type, integer_zero_node);
@ -4732,6 +4756,72 @@ vectorizable_simd_clone_call (vec_info *vinfo, stmt_vec_info stmt_info,
}
}
}
else if (SCALAR_INT_MODE_P (bestn->simdclone->mask_mode))
{
atype = bestn->simdclone->args[i].vector_type;
/* Guess the number of lanes represented by atype. */
unsigned HOST_WIDE_INT atype_subparts
= exact_div (bestn->simdclone->simdlen,
num_mask_args).to_constant ();
o = vector_unroll_factor (nunits, atype_subparts);
for (m = j * o; m < (j + 1) * o; m++)
{
if (m == 0)
{
if (!slp_node)
vect_get_vec_defs_for_operand (vinfo, stmt_info,
o * ncopies,
op,
&vec_oprnds[i]);
vec_oprnds_i[i] = 0;
}
if (atype_subparts
< simd_clone_subparts (arginfo[i].vectype))
{
/* The mask argument has fewer elements than the
input vector. */
/* FORNOW */
gcc_unreachable ();
}
else if (atype_subparts
== simd_clone_subparts (arginfo[i].vectype))
{
/* The vector mask argument matches the input
in the number of lanes, but not necessarily
in the mode. */
vec_oprnd0 = vec_oprnds[i][vec_oprnds_i[i]++];
tree st = lang_hooks.types.type_for_mode
(TYPE_MODE (TREE_TYPE (vec_oprnd0)), 1);
vec_oprnd0 = build1 (VIEW_CONVERT_EXPR, st,
vec_oprnd0);
gassign *new_stmt
= gimple_build_assign (make_ssa_name (st),
vec_oprnd0);
vect_finish_stmt_generation (vinfo, stmt_info,
new_stmt, gsi);
if (!types_compatible_p (atype, st))
{
new_stmt
= gimple_build_assign (make_ssa_name (atype),
NOP_EXPR,
gimple_assign_lhs
(new_stmt));
vect_finish_stmt_generation (vinfo, stmt_info,
new_stmt, gsi);
}
vargs.safe_push (gimple_assign_lhs (new_stmt));
}
else
{
/* The mask argument has more elements than the
input vector. */
/* FORNOW */
gcc_unreachable ();
}
}
}
else
gcc_unreachable ();
break;
case SIMD_CLONE_ARG_TYPE_UNIFORM:
vargs.safe_push (op);