AArch64: Add more accurate constraint [PR117292]

As shown in the PR, reload may only check the constraint in some cases and
and not check the predicate is still valid for the resulting instruction.
To fix the issue, add a new constraint which matches the predicate exactly.

gcc/ChangeLog:
	PR target/117292
	* config/aarch64/aarch64-simd.md (xor<mode>3<vczle><vczbe>): Use
	'De' constraint.
	* config/aarch64/constraints.md (De): Add new constraint.

gcc/testsuite/ChangeLog:
	PR target/117292
	* gcc.target/aarch64/sve/single_5.c: Remove xfails.
	* gcc.target/aarch64/pr117292.c: New test.
This commit is contained in:
Wilco Dijkstra 2024-10-25 14:53:58 +00:00
parent 6cb1da72ca
commit 7c17058eac
4 changed files with 50 additions and 3 deletions

View file

@ -1151,7 +1151,7 @@
"TARGET_SIMD"
{@ [ cons: =0 , 1 , 2 ]
[ w , w , w ] eor\t%0.<Vbtype>, %1.<Vbtype>, %2.<Vbtype>
[ w , 0 , Do ] << aarch64_output_simd_xor_imm (operands[2], <bitsize>);
[ w , 0 , De ] << aarch64_output_simd_xor_imm (operands[2], <bitsize>);
}
[(set_attr "type" "neon_logic<q>")]
)

View file

@ -472,6 +472,12 @@
(and (match_code "const_vector")
(match_test "aarch64_simd_valid_and_imm (op)")))
(define_constraint "De"
"@internal
A constraint that matches vector of immediates for xor."
(and (match_code "const_vector")
(match_test "aarch64_simd_valid_xor_imm (op)")))
(define_constraint "Dn"
"@internal
A constraint that matches vector of immediates."

View file

@ -0,0 +1,41 @@
/* { dg-do compile } */
/* { dg-options "-Os" } */
#pragma GCC target "+nosve"
typedef char v8u8;
typedef __attribute__((__vector_size__ (2))) char v16u8;
typedef __attribute__((__vector_size__ (4))) char v32u8;
typedef __attribute__((__vector_size__ (8))) char v64u8;
typedef short v128u8;
typedef __attribute__((__vector_size__ (32))) char v256u8;
typedef __attribute__((__vector_size__ (64))) char v512u8;
v16u8 foo0_v16s16_0;
__attribute__((__vector_size__ (16))) int foo0_v512u32_0;
v8u8 foo0_ret;
static __attribute__((__noinline__)) __attribute__((__noclone__)) void
foo0 (signed char s8_0, v512u8 v512u16_0, v512u8 v512s16_0,
v512u8 v512s32_0, v512u8 v512u64_0, v512u8 v512s64_0,
v512u8 v512u128_0, v512u8 v512s128_0)
{
char v8s8_0;
v8s8_0 ^= s8_0;
foo0_v512u32_0 /= foo0_v512u32_0 ^ s8_0;
v512u8 v512u8_r = v512u16_0 + v512s16_0 + v512s32_0 +
v512u64_0 + v512s64_0 + v512u128_0 + v512s128_0;
v16u8 v16u8_r = ((union { v64u8 a; v16u8 b;})
((union { v128u8 a; v64u8 b;})
((union { v256u8 a; v128u8 b;})
((union { v512u8 a; v256u8 b;}) v512u8_r).b).b).b).b +
foo0_v16s16_0;
v8u8 v8u8_r = ((union { v16u8 a; v8u8 b;}) v16u8_r).b + v8s8_0;
foo0_ret = v8u8_r;
}
void
main ()
{
foo0 (3, (v512u8){}, (v512u8){}, (v512u8){}, (v512u8){},
(v512u8){}, (v512u8){}, (v512u8){});
}

View file

@ -11,8 +11,8 @@
/* { dg-final { scan-assembler-times {\tmovi\tv[0-9]+\.8h, 0x4\n} 1 } } */
/* { dg-final { scan-assembler-times {\tmovi\tv[0-9]+\.4s, 0x5\n} 1 } } */
/* { dg-final { scan-assembler-times {\tmovi\tv[0-9]+\.4s, 0x6\n} 1 } } */
/* { dg-final { scan-assembler-times {\tmov\tz[0-9]+\.d, #7\n} 1 { xfail *-*-* } } } */
/* { dg-final { scan-assembler-times {\tmov\tz[0-9]+\.d, #8\n} 1 { xfail *-*-* } } } */
/* { dg-final { scan-assembler-times {\tmov\tz[0-9]+\.d, #7\n} 1 } } */
/* { dg-final { scan-assembler-times {\tmov\tz[0-9]+\.d, #8\n} 1 } } */
/* { dg-final { scan-assembler-times {\tfmov\tv[0-9]+\.8h, 1\.0e\+0\n} 1 } } */
/* { dg-final { scan-assembler-times {\tfmov\tv[0-9]+\.4s, 2\.0e\+0\n} 1 } } */
/* { dg-final { scan-assembler-times {\tfmov\tv[0-9]+\.2d, 3\.0e\+0\n} 1 } } */