Hi
I updated the patch to correct the length of insn adddi3_neon.
thanks
Carrot
2012-06-04 Wei Guozhi <car...@google.com>
PR target/53447
* gcc.target/arm/pr53447-1.c: New testcase.
2012-06-04 Wei Guozhi <car...@google.com>
PR target/53447
* config/arm/arm-protos.h (const_ok_for_adddi): New prototype.
* config/arm/arm.c (const_ok_for_adddi): New function.
* config/arm/constraints.md (Dd): New constraint.
* config/arm/arm.md (adddi3): Extend it to handle constants.
(arm_adddi3): Likewise.
* config/arm/neon.md (adddi3_neon): Likewise.
Index: testsuite/gcc.target/arm/pr53447-1.c
===================================================================
--- testsuite/gcc.target/arm/pr53447-1.c (revision 0)
+++ testsuite/gcc.target/arm/pr53447-1.c (revision 0)
@@ -0,0 +1,8 @@
+/* { dg-options "-O2" } */
+/* { dg-require-effective-target arm32 } */
+/* { dg-final { scan-assembler-not "mov" } } */
+
+void t0p(long long * p)
+{
+ *p += 0x100000001;
+}
Index: config/arm/arm.c
===================================================================
--- config/arm/arm.c (revision 187751)
+++ config/arm/arm.c (working copy)
@@ -2497,6 +2497,17 @@
}
}
+/* Return TRUE if int I is a valid immediate constant used by pattern
+ arm_adddi3. */
+int
+const_ok_for_adddi (HOST_WIDE_INT i)
+{
+ HOST_WIDE_INT high = (i >> 32) & 0xFFFFFFFF;
+ HOST_WIDE_INT low = i & 0xFFFFFFFF;
+ return (const_ok_for_arm (high)
+ && (const_ok_for_arm (low) || const_ok_for_arm (-low)));
+}
+
/* Emit a sequence of insns to handle a large constant.
CODE is the code of the operation required, it can be any of SET, PLUS,
IOR, AND, XOR, MINUS;
Index: config/arm/arm-protos.h
===================================================================
--- config/arm/arm-protos.h (revision 187751)
+++ config/arm/arm-protos.h (working copy)
@@ -47,6 +47,7 @@
extern bool arm_small_register_classes_for_mode_p (enum machine_mode);
extern int arm_hard_regno_mode_ok (unsigned int, enum machine_mode);
extern bool arm_modes_tieable_p (enum machine_mode, enum machine_mode);
+extern int const_ok_for_adddi (HOST_WIDE_INT);
extern int const_ok_for_arm (HOST_WIDE_INT);
extern int const_ok_for_op (HOST_WIDE_INT, enum rtx_code);
extern int arm_split_constant (RTX_CODE, enum machine_mode, rtx,
Index: config/arm/neon.md
===================================================================
--- config/arm/neon.md (revision 187751)
+++ config/arm/neon.md (working copy)
@@ -588,9 +588,9 @@
)
(define_insn "adddi3_neon"
- [(set (match_operand:DI 0 "s_register_operand" "=w,?&r,?&r,?w")
- (plus:DI (match_operand:DI 1 "s_register_operand" "%w,0,0,w")
- (match_operand:DI 2 "s_register_operand" "w,r,0,w")))
+ [(set (match_operand:DI 0 "s_register_operand" "=w,?&r,?&r,?w,?&r,?&r,?&r")
+ (plus:DI (match_operand:DI 1 "s_register_operand" "%w,0,0,w,r,0,r")
+ (match_operand:DI 2 "reg_or_int_operand" "w,r,0,w,r,Dd,Dd")))
(clobber (reg:CC CC_REGNUM))]
"TARGET_NEON"
{
@@ -600,13 +600,16 @@
case 3: return "vadd.i64\t%P0, %P1, %P2";
case 1: return "#";
case 2: return "#";
+ case 4: return "#";
+ case 5: return "#";
+ case 6: return "#";
default: gcc_unreachable ();
}
}
- [(set_attr "neon_type" "neon_int_1,*,*,neon_int_1")
- (set_attr "conds" "*,clob,clob,*")
- (set_attr "length" "*,8,8,*")
- (set_attr "arch" "nota8,*,*,onlya8")]
+ [(set_attr "neon_type" "neon_int_1,*,*,neon_int_1,*,*,*")
+ (set_attr "conds" "*,clob,clob,*,clob,clob,clob")
+ (set_attr "length" "*,8,8,*,8,8,8")
+ (set_attr "arch" "nota8,*,*,onlya8,*,*,*")]
)
(define_insn "*sub<mode>3_neon"
Index: config/arm/constraints.md
===================================================================
--- config/arm/constraints.md (revision 187751)
+++ config/arm/constraints.md (working copy)
@@ -29,7 +29,7 @@
;; in Thumb-1 state: I, J, K, L, M, N, O
;; The following multi-letter normal constraints have been used:
-;; in ARM/Thumb-2 state: Da, Db, Dc, Dn, Dl, DL, Dv, Dy, Di, Dt, Dz
+;; in ARM/Thumb-2 state: Da, Db, Dc, Dd, Dn, Dl, DL, Dv, Dy, Di, Dt, Dz
;; in Thumb-1 state: Pa, Pb, Pc, Pd, Pe
;; in Thumb-2 state: Pj, PJ, Ps, Pt, Pu, Pv, Pw, Px, Py
@@ -251,6 +251,12 @@
(match_test "TARGET_32BIT && arm_const_double_inline_cost (op) == 4
&& !(optimize_size || arm_ld_sched)")))
+(define_constraint "Dd"
+ "@internal
+ In ARM/Thumb-2 state a const_int that can be used by insn adddi."
+ (and (match_code "const_int")
+ (match_test "TARGET_32BIT && const_ok_for_adddi (ival)")))
+
(define_constraint "Di"
"@internal
In ARM/Thumb-2 state a const_int or const_double where both the high
Index: config/arm/arm.md
===================================================================
--- config/arm/arm.md (revision 187751)
+++ config/arm/arm.md (working copy)
@@ -574,10 +574,21 @@
[(parallel
[(set (match_operand:DI 0 "s_register_operand" "")
(plus:DI (match_operand:DI 1 "s_register_operand" "")
- (match_operand:DI 2 "s_register_operand" "")))
+ (match_operand:DI 2 "reg_or_int_operand" "")))
(clobber (reg:CC CC_REGNUM))])]
"TARGET_EITHER"
"
+ if (GET_CODE (operands[2]) == CONST_INT)
+ {
+ if (TARGET_32BIT && const_ok_for_adddi (INTVAL (operands[2])))
+ {
+ emit_insn (gen_arm_adddi3 (operands[0], operands[1], operands[2]));
+ DONE;
+ }
+ else
+ operands[2] = force_reg (DImode, operands[2]);
+ }
+
if (TARGET_HARD_FLOAT && TARGET_MAVERICK)
{
if (!cirrus_fp_register (operands[0], DImode))
@@ -609,10 +620,10 @@
[(set_attr "length" "4")]
)
-(define_insn_and_split "*arm_adddi3"
- [(set (match_operand:DI 0 "s_register_operand" "=&r,&r")
- (plus:DI (match_operand:DI 1 "s_register_operand" "%0, 0")
- (match_operand:DI 2 "s_register_operand" "r, 0")))
+(define_insn_and_split "arm_adddi3"
+ [(set (match_operand:DI 0 "s_register_operand" "=&r,&r,&r,&r,&r")
+ (plus:DI (match_operand:DI 1 "s_register_operand" "%0, 0, r, 0, r")
+ (match_operand:DI 2 "reg_or_int_operand" "r, 0, r, Dd,Dd")))
(clobber (reg:CC CC_REGNUM))]
"TARGET_32BIT && !(TARGET_HARD_FLOAT && TARGET_MAVERICK) && !TARGET_NEON"
"#"
@@ -630,8 +641,17 @@
operands[0] = gen_lowpart (SImode, operands[0]);
operands[4] = gen_highpart (SImode, operands[1]);
operands[1] = gen_lowpart (SImode, operands[1]);
- operands[5] = gen_highpart (SImode, operands[2]);
- operands[2] = gen_lowpart (SImode, operands[2]);
+ if (GET_CODE (operands[2]) == CONST_INT)
+ {
+ HOST_WIDE_INT v = INTVAL (operands[2]);
+ operands[5] = GEN_INT (ARM_SIGN_EXTEND ((v >> 32) & 0xFFFFFFFF));
+ operands[2] = GEN_INT (ARM_SIGN_EXTEND (v & 0xFFFFFFFF));
+ }
+ else
+ {
+ operands[5] = gen_highpart (SImode, operands[2]);
+ operands[2] = gen_lowpart (SImode, operands[2]);
+ }
}"
[(set_attr "conds" "clob")
(set_attr "length" "8")]
On Sat, May 26, 2012 at 9:42 PM, Carrot Wei <car...@google.com> wrote:
> Hi,
>
> As described in PR53447, many 64bit ALU operations with constant can be
> optimized to use corresponding 32bit instructions with immediate operands.
>
> This is the first part of the patches that deals with 64bit add. It directly
> extends the patterns adddi3, arm_adddi3 and adddi3_neon to handle constant
> operands.
>
> Tested on arm qemu without regression.
>
> OK for trunk?
>
> thanks
> Carrot
>
> 2012-05-26 Wei Guozhi <car...@google.com>
>
> PR target/53447
> * gcc.target/arm/pr53447-1.c: New testcase.
>
>
> 2012-05-26 Wei Guozhi <car...@google.com>
>
> PR target/53447
> * config/arm/arm-protos.h (const_ok_for_adddi): New prototype.
> * config/arm/arm.c (const_ok_for_adddi): New function.
> * config/arm/constraints.md (Dd): New constraint.
> * config/arm/arm.md (adddi3): Extend it to handle constants.
> (arm_adddi3): Likewise.
> * config/arm/neon.md (adddi3_neon): Likewise.
>
>
> Index: testsuite/gcc.target/arm/pr53447-1.c
> ===================================================================
> --- testsuite/gcc.target/arm/pr53447-1.c (revision 0)
> +++ testsuite/gcc.target/arm/pr53447-1.c (revision 0)
> @@ -0,0 +1,8 @@
> +/* { dg-options "-O2" } */
> +/* { dg-require-effective-target arm32 } */
> +/* { dg-final { scan-assembler-not "mov" } } */
> +
> +void t0p(long long * p)
> +{
> + *p += 0x100000001;
> +}
> Index: config/arm/arm.c
> ===================================================================
> --- config/arm/arm.c (revision 187751)
> +++ config/arm/arm.c (working copy)
> @@ -2497,6 +2497,17 @@
> }
> }
>
> +/* Return TRUE if int I is a valid immediate constant used by pattern
> + arm_adddi3. */
> +int
> +const_ok_for_adddi (HOST_WIDE_INT i)
> +{
> + HOST_WIDE_INT high = (i >> 32) & 0xFFFFFFFF;
> + HOST_WIDE_INT low = i & 0xFFFFFFFF;
> + return (const_ok_for_arm (high)
> + && (const_ok_for_arm (low) || const_ok_for_arm (-low)));
> +}
> +
> /* Emit a sequence of insns to handle a large constant.
> CODE is the code of the operation required, it can be any of SET, PLUS,
> IOR, AND, XOR, MINUS;
> Index: config/arm/arm-protos.h
> ===================================================================
> --- config/arm/arm-protos.h (revision 187751)
> +++ config/arm/arm-protos.h (working copy)
> @@ -47,6 +47,7 @@
> extern bool arm_small_register_classes_for_mode_p (enum machine_mode);
> extern int arm_hard_regno_mode_ok (unsigned int, enum machine_mode);
> extern bool arm_modes_tieable_p (enum machine_mode, enum machine_mode);
> +extern int const_ok_for_adddi (HOST_WIDE_INT);
> extern int const_ok_for_arm (HOST_WIDE_INT);
> extern int const_ok_for_op (HOST_WIDE_INT, enum rtx_code);
> extern int arm_split_constant (RTX_CODE, enum machine_mode, rtx,
> Index: config/arm/neon.md
> ===================================================================
> --- config/arm/neon.md (revision 187751)
> +++ config/arm/neon.md (working copy)
> @@ -588,9 +588,9 @@
> )
>
> (define_insn "adddi3_neon"
> - [(set (match_operand:DI 0 "s_register_operand" "=w,?&r,?&r,?w")
> - (plus:DI (match_operand:DI 1 "s_register_operand" "%w,0,0,w")
> - (match_operand:DI 2 "s_register_operand" "w,r,0,w")))
> + [(set (match_operand:DI 0 "s_register_operand" "=w,?&r,?&r,?w,?&r,?&r,?&r")
> + (plus:DI (match_operand:DI 1 "s_register_operand" "%w,0,0,w,r,0,r")
> + (match_operand:DI 2 "reg_or_int_operand"
> "w,r,0,w,r,Dd,Dd")))
> (clobber (reg:CC CC_REGNUM))]
> "TARGET_NEON"
> {
> @@ -600,13 +600,16 @@
> case 3: return "vadd.i64\t%P0, %P1, %P2";
> case 1: return "#";
> case 2: return "#";
> + case 4: return "#";
> + case 5: return "#";
> + case 6: return "#";
> default: gcc_unreachable ();
> }
> }
> - [(set_attr "neon_type" "neon_int_1,*,*,neon_int_1")
> - (set_attr "conds" "*,clob,clob,*")
> - (set_attr "length" "*,8,8,*")
> - (set_attr "arch" "nota8,*,*,onlya8")]
> + [(set_attr "neon_type" "neon_int_1,*,*,neon_int_1,*,*,*")
> + (set_attr "conds" "*,clob,clob,*,clob,clob,clob")
> + (set_attr "length" "*,8,8,*,*,*,*")
> + (set_attr "arch" "nota8,*,*,onlya8,*,*,*")]
> )
>
> (define_insn "*sub<mode>3_neon"
> Index: config/arm/constraints.md
> ===================================================================
> --- config/arm/constraints.md (revision 187751)
> +++ config/arm/constraints.md (working copy)
> @@ -29,7 +29,7 @@
> ;; in Thumb-1 state: I, J, K, L, M, N, O
>
> ;; The following multi-letter normal constraints have been used:
> -;; in ARM/Thumb-2 state: Da, Db, Dc, Dn, Dl, DL, Dv, Dy, Di, Dt, Dz
> +;; in ARM/Thumb-2 state: Da, Db, Dc, Dd, Dn, Dl, DL, Dv, Dy, Di, Dt, Dz
> ;; in Thumb-1 state: Pa, Pb, Pc, Pd, Pe
> ;; in Thumb-2 state: Pj, PJ, Ps, Pt, Pu, Pv, Pw, Px, Py
>
> @@ -251,6 +251,12 @@
> (match_test "TARGET_32BIT && arm_const_double_inline_cost (op) == 4
> && !(optimize_size || arm_ld_sched)")))
>
> +(define_constraint "Dd"
> + "@internal
> + In ARM/Thumb-2 state a const_int that can be used by insn adddi."
> + (and (match_code "const_int")
> + (match_test "TARGET_32BIT && const_ok_for_adddi (ival)")))
> +
> (define_constraint "Di"
> "@internal
> In ARM/Thumb-2 state a const_int or const_double where both the high
> Index: config/arm/arm.md
> ===================================================================
> --- config/arm/arm.md (revision 187751)
> +++ config/arm/arm.md (working copy)
> @@ -574,10 +574,21 @@
> [(parallel
> [(set (match_operand:DI 0 "s_register_operand" "")
> (plus:DI (match_operand:DI 1 "s_register_operand" "")
> - (match_operand:DI 2 "s_register_operand" "")))
> + (match_operand:DI 2 "reg_or_int_operand" "")))
> (clobber (reg:CC CC_REGNUM))])]
> "TARGET_EITHER"
> "
> + if (GET_CODE (operands[2]) == CONST_INT)
> + {
> + if (TARGET_32BIT && const_ok_for_adddi (INTVAL (operands[2])))
> + {
> + emit_insn (gen_arm_adddi3 (operands[0], operands[1], operands[2]));
> + DONE;
> + }
> + else
> + operands[2] = force_reg (DImode, operands[2]);
> + }
> +
> if (TARGET_HARD_FLOAT && TARGET_MAVERICK)
> {
> if (!cirrus_fp_register (operands[0], DImode))
> @@ -609,10 +620,10 @@
> [(set_attr "length" "4")]
> )
>
> -(define_insn_and_split "*arm_adddi3"
> - [(set (match_operand:DI 0 "s_register_operand" "=&r,&r")
> - (plus:DI (match_operand:DI 1 "s_register_operand" "%0, 0")
> - (match_operand:DI 2 "s_register_operand" "r, 0")))
> +(define_insn_and_split "arm_adddi3"
> + [(set (match_operand:DI 0 "s_register_operand" "=&r,&r,&r,&r,&r")
> + (plus:DI (match_operand:DI 1 "s_register_operand" "%0, 0, r, 0, r")
> + (match_operand:DI 2 "reg_or_int_operand" "r, 0, r, Dd,Dd")))
> (clobber (reg:CC CC_REGNUM))]
> "TARGET_32BIT && !(TARGET_HARD_FLOAT && TARGET_MAVERICK) && !TARGET_NEON"
> "#"
> @@ -630,8 +641,17 @@
> operands[0] = gen_lowpart (SImode, operands[0]);
> operands[4] = gen_highpart (SImode, operands[1]);
> operands[1] = gen_lowpart (SImode, operands[1]);
> - operands[5] = gen_highpart (SImode, operands[2]);
> - operands[2] = gen_lowpart (SImode, operands[2]);
> + if (GET_CODE (operands[2]) == CONST_INT)
> + {
> + HOST_WIDE_INT v = INTVAL (operands[2]);
> + operands[5] = GEN_INT (ARM_SIGN_EXTEND ((v >> 32) & 0xFFFFFFFF));
> + operands[2] = GEN_INT (ARM_SIGN_EXTEND (v & 0xFFFFFFFF));
> + }
> + else
> + {
> + operands[5] = gen_highpart (SImode, operands[2]);
> + operands[2] = gen_lowpart (SImode, operands[2]);
> + }
> }"
> [(set_attr "conds" "clob")
> (set_attr "length" "8")]
