OpenE2K/gcc
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

x86: Add TARGET_AVX256_[MOVE|STORE]_BY_PIECES

Browse Source 
1. Add TARGET_AVX256_MOVE_BY_PIECES to perform move by-pieces operation
with 256-bit AVX instructions.
2. Add TARGET_AVX256_STORE_BY_PIECES to perform move and store by-pieces
operations with 256-bit AVX instructions.

They are enabled only for Intel Alder Lake and Intel processors with
AVX512.

gcc/

	PR target/101935
	* config/i386/i386.h (TARGET_AVX256_MOVE_BY_PIECES): New.
	(TARGET_AVX256_STORE_BY_PIECES): Likewise.
	(MOVE_MAX): Check TARGET_AVX256_MOVE_BY_PIECES and
	TARGET_AVX256_STORE_BY_PIECES instead of
	TARGET_AVX256_SPLIT_UNALIGNED_LOAD and
	TARGET_AVX256_SPLIT_UNALIGNED_STORE.
	(STORE_MAX_PIECES): Check TARGET_AVX256_STORE_BY_PIECES instead
	of TARGET_AVX256_SPLIT_UNALIGNED_STORE.
	* config/i386/x86-tune.def (X86_TUNE_AVX256_MOVE_BY_PIECES): New.
	(X86_TUNE_AVX256_STORE_BY_PIECES): Likewise.

gcc/testsuite/

	PR target/101935
	* g++.target/i386/pr80566-1.C: Add
	-mtune-ctrl=avx256_store_by_pieces.
	* gcc.target/i386/pr100865-4a.c: Likewise.
	* gcc.target/i386/pr100865-10a.c: Likewise.
	* gcc.target/i386/pr90773-20.c: Likewise.
	* gcc.target/i386/pr90773-21.c: Likewise.
	* gcc.target/i386/pr90773-22.c: Likewise.
	* gcc.target/i386/pr90773-23.c: Likewise.
	* g++.target/i386/pr80566-2.C: Add
	-mtune-ctrl=avx256_move_by_pieces.
	* gcc.target/i386/eh_return-1.c: Likewise.
	* gcc.target/i386/pr90773-26.c: Likewise.
	* gcc.target/i386/pieces-memcpy-12.c: Replace -mtune=haswell
	with -mtune-ctrl=avx256_move_by_pieces.
	* gcc.target/i386/pieces-memcpy-15.c: Likewise.
	* gcc.target/i386/pieces-memset-2.c: Replace -mtune=haswell
	with -mtune-ctrl=avx256_store_by_pieces.
	* gcc.target/i386/pieces-memset-5.c: Likewise.
	* gcc.target/i386/pieces-memset-11.c: Likewise.
	* gcc.target/i386/pieces-memset-14.c: Likewise.
	* gcc.target/i386/pieces-memset-20.c: Likewise.
	* gcc.target/i386/pieces-memset-23.c: Likewise.
	* gcc.target/i386/pieces-memset-29.c: Likewise.
	* gcc.target/i386/pieces-memset-30.c: Likewise.
	* gcc.target/i386/pieces-memset-33.c: Likewise.
	* gcc.target/i386/pieces-memset-34.c: Likewise.
	* gcc.target/i386/pieces-memset-44.c: Likewise.
	* gcc.target/i386/pieces-memset-37.c: Replace -mtune=generic
	with -mtune-ctrl=avx256_store_by_pieces.
This commit is contained in:
H.J. Lu 2021-08-26 05:31:50 -07:00 committed by liuhongt
parent c8e4cb8adf
commit 5b01bfeb87
26 changed files with 42 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
gcc/config/i386/i386.h
View File

@ -403,6 +403,10 @@ extern unsigned char ix86_tune_features[X86_TUNE_LAST];
ix86_tune_features[X86_TUNE_AVOID_LEA_FOR_ADDR]
#define TARGET_SOFTWARE_PREFETCHING_BENEFICIAL \
ix86_tune_features[X86_TUNE_SOFTWARE_PREFETCHING_BENEFICIAL]
#define TARGET_AVX256_MOVE_BY_PIECES \
ix86_tune_features[X86_TUNE_AVX256_MOVE_BY_PIECES]
#define TARGET_AVX256_STORE_BY_PIECES \
ix86_tune_features[X86_TUNE_AVX256_STORE_BY_PIECES]
#define TARGET_AVX256_SPLIT_REGS \
ix86_tune_features[X86_TUNE_AVX256_SPLIT_REGS]
#define TARGET_GENERAL_REGS_SSE_SPILL \
@ -1793,8 +1797,8 @@ typedef struct ix86_args {
? 64 \
: ((TARGET_AVX \
&& !TARGET_PREFER_AVX128 \
&& !TARGET_AVX256_SPLIT_UNALIGNED_LOAD \
&& !TARGET_AVX256_SPLIT_UNALIGNED_STORE) \
&& (TARGET_AVX256_MOVE_BY_PIECES \
|| TARGET_AVX256_STORE_BY_PIECES)) \
? 32 \
: ((TARGET_SSE2 \
&& TARGET_SSE_UNALIGNED_LOAD_OPTIMAL \
@ -1811,7 +1815,7 @@ typedef struct ix86_args {
? 64 \
: ((TARGET_AVX \
&& !TARGET_PREFER_AVX128 \
&& !TARGET_AVX256_SPLIT_UNALIGNED_STORE) \
&& TARGET_AVX256_STORE_BY_PIECES) \
? 32 \
: ((TARGET_SSE2 \
&& TARGET_SSE_UNALIGNED_STORE_OPTIMAL) \

11
gcc/config/i386/x86-tune.def
View File

@ -484,6 +484,17 @@ DEF_TUNE (X86_TUNE_AVX128_OPTIMAL, "avx128_optimal", m_BDVER | m_BTVER2
instructions in the auto-vectorizer. */
DEF_TUNE (X86_TUNE_AVX256_OPTIMAL, "avx256_optimal", m_CORE_AVX512)
/* X86_TUNE_AVX256_MOVE_BY_PIECES: Optimize move_by_pieces with 256-bit
AVX instructions. */
DEF_TUNE (X86_TUNE_AVX256_MOVE_BY_PIECES, "avx256_move_by_pieces",
m_ALDERLAKE | m_CORE_AVX512)
/* X86_TUNE_AVX256_STORE_BY_PIECES: Optimize store_by_pieces with 256-bit
AVX instructions. */
DEF_TUNE (X86_TUNE_AVX256_STORE_BY_PIECES, "avx256_store_by_pieces",
m_ALDERLAKE | m_CORE_AVX512)
/*****************************************************************************/
/*****************************************************************************/
/* Historical relics: tuning flags that helps a specific old CPU designs */
/*****************************************************************************/

2
gcc/testsuite/g++.target/i386/pr80566-1.C
View File

@ -1,5 +1,5 @@
// { dg-do compile }
// { dg-options "-O2 -march=haswell" }
// { dg-options "-O2 -march=haswell -mtune-ctrl=avx256_store_by_pieces" }
#include <cstring>

2
gcc/testsuite/g++.target/i386/pr80566-2.C
View File

@ -1,5 +1,5 @@
// { dg-do compile }
// { dg-options "-O2 -march=haswell" }
// { dg-options "-O2 -march=haswell -mtune-ctrl=avx256_move_by_pieces" }
#include <cstring>

2
gcc/testsuite/gcc.target/i386/eh_return-1.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=haswell -mno-avx512f" } */
/* { dg-options "-O2 -march=haswell -mno-avx512f -mtune-ctrl=avx256_move_by_pieces" } */
struct _Unwind_Context
{

2
gcc/testsuite/gcc.target/i386/pieces-memcpy-12.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_move_by_pieces" } */
extern char *dst, *src;

2
gcc/testsuite/gcc.target/i386/pieces-memcpy-15.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_move_by_pieces" } */
extern char *dst, *src;

2
gcc/testsuite/gcc.target/i386/pieces-memset-11.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-14.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-2.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-20.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-23.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-29.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-30.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-33.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-34.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-37.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune=generic" } */
/* { dg-options "-O2 -mno-avx512f -mavx2 -mtune-ctrl=avx256_store_by_pieces" } */
void
foo (int a1, int a2, int a3, int a4, int a5, int a6, int x, char *dst)

2
gcc/testsuite/gcc.target/i386/pieces-memset-44.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pieces-memset-5.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune=haswell" } */
/* { dg-options "-O2 -mno-avx2 -mavx -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pr100865-10a.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile { target int128 } } */
/* { dg-options "-O3 -march=skylake" } */
/* { dg-options "-O3 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern __int128 array[16];

2
gcc/testsuite/gcc.target/i386/pr100865-4a.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern char array[64];

2
gcc/testsuite/gcc.target/i386/pr90773-20.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pr90773-21.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pr90773-22.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pr90773-23.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_store_by_pieces" } */
extern char *dst;

2
gcc/testsuite/gcc.target/i386/pr90773-26.c
View File

@ -1,5 +1,5 @@
/* { dg-do compile } */
/* { dg-options "-O2 -march=skylake" } */
/* { dg-options "-O2 -march=skylake -mtune-ctrl=avx256_move_by_pieces" } */
struct S
{