[v2] AArch64: Improve immediate generation
Checks
Commit Message
v2: Use check-function-bodies in tests
Further improve immediate generation by adding support for 2-instruction
MOV/EOR bitmask immediates. This reduces the number of 3/4-instruction
immediates in SPECCPU2017 by ~2%.
Passes regress, OK for commit?
gcc/ChangeLog:
* config/aarch64/aarch64.cc (aarch64_internal_mov_immediate)
Add support for immediates using MOV/EOR bitmask.
gcc/testsuite:
* gcc.target/aarch64/imm_choice_comparison.c: Change tests.
* gcc.target/aarch64/moveor_imm.c: Add new test.
* gcc.target/aarch64/pr106583.c: Change tests.
---
Comments
On 24/10/2023 18:27, Wilco Dijkstra wrote:
> v2: Use check-function-bodies in tests
>
> Further improve immediate generation by adding support for 2-instruction
> MOV/EOR bitmask immediates. This reduces the number of 3/4-instruction
> immediates in SPECCPU2017 by ~2%.
>
> Passes regress, OK for commit?
>
> gcc/ChangeLog:
> * config/aarch64/aarch64.cc (aarch64_internal_mov_immediate)
> Add support for immediates using MOV/EOR bitmask.
>
> gcc/testsuite:
> * gcc.target/aarch64/imm_choice_comparison.c: Change tests.
> * gcc.target/aarch64/moveor_imm.c: Add new test.
> * gcc.target/aarch64/pr106583.c: Change tests.
>
> ---
>
> diff --git a/gcc/config/aarch64/aarch64.cc b/gcc/config/aarch64/aarch64.cc
> index 578a253d6e0e133e19592553fc873b3e73f9f218..ed5be2b64c9a767d74e9d78415da964c669001aa 100644
> --- a/gcc/config/aarch64/aarch64.cc
> +++ b/gcc/config/aarch64/aarch64.cc
> @@ -5748,6 +5748,26 @@ aarch64_internal_mov_immediate (rtx dest, rtx imm, bool generate,
> }
> return 2;
> }
> +
> + /* Try 2 bitmask immediates which are xor'd together. */
> + for (i = 0; i < 64; i += 16)
> + {
> + val2 = (val >> i) & mask;
> + val2 |= val2 << 16;
> + val2 |= val2 << 32;
> + if (aarch64_bitmask_imm (val2) && aarch64_bitmask_imm (val ^ val2))
> + break;
> + }
> +
> + if (i != 64)
> + {
> + if (generate)
> + {
> + emit_insn (gen_rtx_SET (dest, GEN_INT (val2)));
> + emit_insn (gen_xordi3 (dest, dest, GEN_INT (val ^ val2)));
> + }
> + return 2;
> + }
> }
>
> /* Try a bitmask plus 2 movk to generate the immediate in 3 instructions. */
> diff --git a/gcc/testsuite/gcc.target/aarch64/imm_choice_comparison.c b/gcc/testsuite/gcc.target/aarch64/imm_choice_comparison.c
> index ebc44d6dbc7287d907603d77d7b54496de177c4b..a1fc90ad73411ae8ed848fa321586afcb8d710aa 100644
> --- a/gcc/testsuite/gcc.target/aarch64/imm_choice_comparison.c
> +++ b/gcc/testsuite/gcc.target/aarch64/imm_choice_comparison.c
> @@ -1,32 +1,64 @@
> /* { dg-do compile } */
> /* { dg-options "-O2" } */
> +/* { dg-final { check-function-bodies "**" "" } } */
>
> /* Go from four moves to two. */
>
> +/*
> +** foo:
> +** mov w[0-9]+, 2576980377
> +** movk x[0-9]+, 0x9999, lsl 32
> +** ...
> +*/
> +
> int
> foo (long long x)
> {
> - return x <= 0x1999999999999998;
> + return x <= 0x0000999999999998;
> }
>
> +/*
> +** GT:
> +** mov w[0-9]+, -16777217
> +** ...
> +*/
> +
> int
> GT (unsigned int x)
> {
> return x > 0xfefffffe;
> }
>
> +/*
> +** LE:
> +** mov w[0-9]+, -16777217
> +** ...
> +*/
> +
> int
> LE (unsigned int x)
> {
> return x <= 0xfefffffe;
> }
>
> +/*
> +** GE:
> +** mov w[0-9]+, 4278190079
> +** ...
> +*/
> +
> int
> GE (long long x)
> {
> return x >= 0xff000000;
> }
>
> +/*
> +** LT:
> +** mov w[0-9]+, -16777217
> +** ...
> +*/
> +
> int
> LT (int x)
> {
> @@ -35,6 +67,13 @@ LT (int x)
>
> /* Optimize the immediate in conditionals. */
>
> +/*
> +** check:
> +** ...
> +** mov w[0-9]+, -16777217
> +** ...
> +*/
> +
> int
> check (int x, int y)
> {
> @@ -44,11 +83,15 @@ check (int x, int y)
> return x;
> }
>
> +/*
> +** tern:
> +** ...
> +** mov w[0-9]+, -16777217
> +** ...
> +*/
> +
> int
> tern (int x)
> {
> return x >= 0xff000000 ? 5 : -3;
> }
> -
> -/* baz produces one movk instruction. */
> -/* { dg-final { scan-assembler-times "movk" 1 } } */
> diff --git a/gcc/testsuite/gcc.target/aarch64/moveor_imm.c b/gcc/testsuite/gcc.target/aarch64/moveor_imm.c
> new file mode 100644
> index 0000000000000000000000000000000000000000..1c0c3f3bf8c588f9661112a8b3f9a72c5ddff95c
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/aarch64/moveor_imm.c
> @@ -0,0 +1,63 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2" } */
> +/* { dg-final { check-function-bodies "**" "" } } */
> +
> +/*
> +** f1:
> +** mov x0, -6148914691236517206
> +** eor x0, x0, -9223372036854775807
> +** ret
> +*/
Some odd white space above.
Also, I think it would be better to write the tests as
** f1:
** ...
** <pattern-to-match>
** ...
Then different prologue and epilogue options (such as BTI or pac-ret)
won't affect the tests.
> +
> +long f1 (void)
> +{
> + return 0x2aaaaaaaaaaaaaab;
> +}
> +
> +/*
> +** f2:
> +** mov x0, -1085102592571150096
> +** eor x0, x0, -2305843009213693951
> +** ret
> +*/
> +
> +long f2 (void)
> +{
> + return 0x10f0f0f0f0f0f0f1;
> +}
> +
> +/*
> +** f3:
> +** mov x0, -3689348814741910324
> +** eor x0, x0, -4611686018427387903
> +** ret
> +*/
> +
> +long f3 (void)
> +{
> + return 0xccccccccccccccd;
> +}
> +
> +/*
> +** f4:
> +** mov x0, -7378697629483820647
> +** eor x0, x0, -9223372036854775807
> +** ret
> +*/
> +
> +long f4 (void)
> +{
> + return 0x1999999999999998;
> +}
> +
> +/*
> +** f5:
> +** mov x0, 3689348814741910323
> +** eor x0, x0, 864691128656461824
> +** ret
> +*/
> +
> +long f5 (void)
> +{
> + return 0x3f3333333f333333;
> +}
> diff --git a/gcc/testsuite/gcc.target/aarch64/pr106583.c b/gcc/testsuite/gcc.target/aarch64/pr106583.c
> index 0f931580817d78dc1cc58f03b251bd21bec71f59..63df7395edf9491720e3601848e15aa773c51e6d 100644
> --- a/gcc/testsuite/gcc.target/aarch64/pr106583.c
> +++ b/gcc/testsuite/gcc.target/aarch64/pr106583.c
> @@ -1,41 +1,94 @@
> -/* { dg-do assemble } */
> -/* { dg-options "-O2 --save-temps" } */
> +/* { dg-do compile } */
> +/* { dg-options "-O2" } */
> +/* { dg-final { check-function-bodies "**" "" } } */
> +
> +/*
> +** f1:
> +** mov x0, -72340172838076674
> +** movk x0, 0xfeff, lsl 0
> +** movk x0, 0x75fe, lsl 48
> +** ret
> +*/
>
> long f1 (void)
> {
> - return 0x7efefefefefefeff;
> + return 0x75fefefefefefeff;
> }
>
> +/*
> +** f2:
> +** mov x0, -6148914691236517206
> +** movk x0, 0x5678, lsl 32
> +** movk x0, 0x1234, lsl 48
> +** ret
> +*/
> +
> long f2 (void)
> {
> return 0x12345678aaaaaaaa;
> }
>
> +/*
> +** f3:
> +** mov x0, -3689348814741910324
> +** movk x0, 0x5678, lsl 0
> +** movk x0, 0x1234, lsl 48
> +** ret
> +*/
> +
> long f3 (void)
> {
> return 0x1234cccccccc5678;
> }
>
> +/*
> +** f4:
> +** mov x0, 8608480567731124087
> +** movk x0, 0x5678, lsl 16
> +** movk x0, 0x1234, lsl 32
> +** ret
> +*/
> +
> long f4 (void)
> {
> return 0x7777123456787777;
> }
>
> +/*
> +** f5:
> +** mov x0, 6148914691236517205
> +** movk x0, 0x5678, lsl 0
> +** movk x0, 0x1234, lsl 16
> +** ret
> +*/
> +
> long f5 (void)
> {
> return 0x5555555512345678;
> }
>
> +/*
> +** f6:
> +** mov x0, -4919131752989213765
> +** movk x0, 0x5678, lsl 16
> +** movk x0, 0x1234, lsl 48
> +** ret
> +*/
> +
> long f6 (void)
> {
> return 0x1234bbbb5678bbbb;
> }
>
> +/*
> +** f7:
> +** mov x0, 4919131752989213764
> +** movk x0, 0x5678, lsl 0
> +** movk x0, 0x1234, lsl 32
> +** ret
> +*/
> +
> long f7 (void)
> {
> return 0x4444123444445678;
> }
> -
> -
> -/* { dg-final { scan-assembler-times {\tmovk\t} 14 } } */
> -/* { dg-final { scan-assembler-times {\tmov\t} 7 } } */
>
OK with that change.
R.
@@ -5748,6 +5748,26 @@ aarch64_internal_mov_immediate (rtx dest, rtx imm, bool generate,
}
return 2;
}
+
+ /* Try 2 bitmask immediates which are xor'd together. */
+ for (i = 0; i < 64; i += 16)
+ {
+ val2 = (val >> i) & mask;
+ val2 |= val2 << 16;
+ val2 |= val2 << 32;
+ if (aarch64_bitmask_imm (val2) && aarch64_bitmask_imm (val ^ val2))
+ break;
+ }
+
+ if (i != 64)
+ {
+ if (generate)
+ {
+ emit_insn (gen_rtx_SET (dest, GEN_INT (val2)));
+ emit_insn (gen_xordi3 (dest, dest, GEN_INT (val ^ val2)));
+ }
+ return 2;
+ }
}
/* Try a bitmask plus 2 movk to generate the immediate in 3 instructions. */
@@ -1,32 +1,64 @@
/* { dg-do compile } */
/* { dg-options "-O2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
/* Go from four moves to two. */
+/*
+** foo:
+** mov w[0-9]+, 2576980377
+** movk x[0-9]+, 0x9999, lsl 32
+** ...
+*/
+
int
foo (long long x)
{
- return x <= 0x1999999999999998;
+ return x <= 0x0000999999999998;
}
+/*
+** GT:
+** mov w[0-9]+, -16777217
+** ...
+*/
+
int
GT (unsigned int x)
{
return x > 0xfefffffe;
}
+/*
+** LE:
+** mov w[0-9]+, -16777217
+** ...
+*/
+
int
LE (unsigned int x)
{
return x <= 0xfefffffe;
}
+/*
+** GE:
+** mov w[0-9]+, 4278190079
+** ...
+*/
+
int
GE (long long x)
{
return x >= 0xff000000;
}
+/*
+** LT:
+** mov w[0-9]+, -16777217
+** ...
+*/
+
int
LT (int x)
{
@@ -35,6 +67,13 @@ LT (int x)
/* Optimize the immediate in conditionals. */
+/*
+** check:
+** ...
+** mov w[0-9]+, -16777217
+** ...
+*/
+
int
check (int x, int y)
{
@@ -44,11 +83,15 @@ check (int x, int y)
return x;
}
+/*
+** tern:
+** ...
+** mov w[0-9]+, -16777217
+** ...
+*/
+
int
tern (int x)
{
return x >= 0xff000000 ? 5 : -3;
}
-
-/* baz produces one movk instruction. */
-/* { dg-final { scan-assembler-times "movk" 1 } } */
new file mode 100644
@@ -0,0 +1,63 @@
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+/*
+** f1:
+** mov x0, -6148914691236517206
+** eor x0, x0, -9223372036854775807
+** ret
+*/
+
+long f1 (void)
+{
+ return 0x2aaaaaaaaaaaaaab;
+}
+
+/*
+** f2:
+** mov x0, -1085102592571150096
+** eor x0, x0, -2305843009213693951
+** ret
+*/
+
+long f2 (void)
+{
+ return 0x10f0f0f0f0f0f0f1;
+}
+
+/*
+** f3:
+** mov x0, -3689348814741910324
+** eor x0, x0, -4611686018427387903
+** ret
+*/
+
+long f3 (void)
+{
+ return 0xccccccccccccccd;
+}
+
+/*
+** f4:
+** mov x0, -7378697629483820647
+** eor x0, x0, -9223372036854775807
+** ret
+*/
+
+long f4 (void)
+{
+ return 0x1999999999999998;
+}
+
+/*
+** f5:
+** mov x0, 3689348814741910323
+** eor x0, x0, 864691128656461824
+** ret
+*/
+
+long f5 (void)
+{
+ return 0x3f3333333f333333;
+}
@@ -1,41 +1,94 @@
-/* { dg-do assemble } */
-/* { dg-options "-O2 --save-temps" } */
+/* { dg-do compile } */
+/* { dg-options "-O2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+/*
+** f1:
+** mov x0, -72340172838076674
+** movk x0, 0xfeff, lsl 0
+** movk x0, 0x75fe, lsl 48
+** ret
+*/
long f1 (void)
{
- return 0x7efefefefefefeff;
+ return 0x75fefefefefefeff;
}
+/*
+** f2:
+** mov x0, -6148914691236517206
+** movk x0, 0x5678, lsl 32
+** movk x0, 0x1234, lsl 48
+** ret
+*/
+
long f2 (void)
{
return 0x12345678aaaaaaaa;
}
+/*
+** f3:
+** mov x0, -3689348814741910324
+** movk x0, 0x5678, lsl 0
+** movk x0, 0x1234, lsl 48
+** ret
+*/
+
long f3 (void)
{
return 0x1234cccccccc5678;
}
+/*
+** f4:
+** mov x0, 8608480567731124087
+** movk x0, 0x5678, lsl 16
+** movk x0, 0x1234, lsl 32
+** ret
+*/
+
long f4 (void)
{
return 0x7777123456787777;
}
+/*
+** f5:
+** mov x0, 6148914691236517205
+** movk x0, 0x5678, lsl 0
+** movk x0, 0x1234, lsl 16
+** ret
+*/
+
long f5 (void)
{
return 0x5555555512345678;
}
+/*
+** f6:
+** mov x0, -4919131752989213765
+** movk x0, 0x5678, lsl 16
+** movk x0, 0x1234, lsl 48
+** ret
+*/
+
long f6 (void)
{
return 0x1234bbbb5678bbbb;
}
+/*
+** f7:
+** mov x0, 4919131752989213764
+** movk x0, 0x5678, lsl 0
+** movk x0, 0x1234, lsl 32
+** ret
+*/
+
long f7 (void)
{
return 0x4444123444445678;
}
-
-
-/* { dg-final { scan-assembler-times {\tmovk\t} 14 } } */
-/* { dg-final { scan-assembler-times {\tmov\t} 7 } } */