[gcc r13-5710] LoongArch: Generate bytepick.[wd] for suitable bit operation pattern

[gcc r13-5710] LoongArch: Generate bytepick.[wd] for suitable bit operation pattern

[Xi Ruoyao]

https://gcc.gnu.org/g:5ebfd7ba2f612283656de4e7b8082bf82ef2b957

commit r13-5710-g5ebfd7ba2f612283656de4e7b8082bf82ef2b957
Author: Xi Ruoyao <xry111@xry111.site>
Date:   Fri Feb 3 17:06:06 2023 +0800

    LoongArch: Generate bytepick.[wd] for suitable bit operation pattern
    
    We can use bytepick.[wd] for
    
        a << (8 * x) | b >> (8 * (sizeof(a) - x))
    
    while a and b are uint32_t or uint64_t.  This is useful for some cases,
    for example:
    https://sourceware.org/pipermail/libc-alpha/2023-February/145203.html
    
    gcc/ChangeLog:
    
            * config/loongarch/loongarch.md (bytepick_w_ashift_amount):
            New define_int_iterator.
            (bytepick_d_ashift_amount): Likewise.
            (bytepick_imm): New define_int_attr.
            (bytepick_w_lshiftrt_amount): Likewise.
            (bytepick_d_lshiftrt_amount): Likewise.
            (bytepick_w_<bytepick_imm>): New define_insn template.
            (bytepick_w_<bytepick_imm>_extend): Likewise.
            (bytepick_d_<bytepick_imm>): Likewise.
            (bytepick_w): Remove unused define_insn.
            (bytepick_d): Likewise.
            (UNSPEC_BYTEPICK_W): Remove unused unspec.
            (UNSPEC_BYTEPICK_D): Likewise.
            * config/loongarch/predicates.md (const_0_to_3_operand):
            Remove unused define_predicate.
            (const_0_to_7_operand): Likewise.
    
    gcc/testsuite/ChangeLog:
    
            * g++.target/loongarch/bytepick.C: New test.

Diff:
---
 gcc/config/loongarch/loongarch.md             | 60 ++++++++++++++++++++-------
 gcc/config/loongarch/predicates.md            |  8 ----
 gcc/testsuite/g++.target/loongarch/bytepick.C | 32 ++++++++++++++
 3 files changed, 77 insertions(+), 23 deletions(-)

diff --git a/gcc/config/loongarch/loongarch.md b/gcc/config/loongarch/loongarch.md
index 0884ec09dfb..3509c3c21c1 100644
--- a/gcc/config/loongarch/loongarch.md
+++ b/gcc/config/loongarch/loongarch.md
@@ -48,8 +48,6 @@
   UNSPEC_EH_RETURN
 
   ;; Bit operation
-  UNSPEC_BYTEPICK_W
-  UNSPEC_BYTEPICK_D
   UNSPEC_BITREV_4B
   UNSPEC_BITREV_8B
 
@@ -544,6 +542,27 @@
 				      (UNSPEC_FTINTRM "0")
 				      (UNSPEC_FTINTRP "0")])
 
+;; Iterator and attributes for bytepick.d
+(define_int_iterator bytepick_w_ashift_amount [8 16 24])
+(define_int_attr bytepick_w_lshiftrt_amount [(8 "24")
+					     (16 "16")
+					     (24 "8")])
+(define_int_iterator bytepick_d_ashift_amount [8 16 24 32 40 48 56])
+(define_int_attr bytepick_d_lshiftrt_amount [(8 "56")
+					     (16 "48")
+					     (24 "40")
+					     (32 "32")
+					     (40 "24")
+					     (48 "16")
+					     (56 "8")])
+(define_int_attr bytepick_imm [(8 "1")
+				 (16 "2")
+				 (24 "3")
+				 (32 "4")
+				 (40 "5")
+				 (48 "6")
+				 (56 "7")])
+
 ;;
 ;;  ....................
 ;;
@@ -3364,24 +3383,35 @@
   [(set_attr "type" "unknown")
    (set_attr "mode" "<MODE>")])
 
-(define_insn "bytepick_w"
+(define_insn "bytepick_w_<bytepick_imm>"
   [(set (match_operand:SI 0 "register_operand" "=r")
-	(unspec:SI [(match_operand:SI 1 "register_operand" "r")
-		    (match_operand:SI 2 "register_operand" "r")
-		    (match_operand:SI 3 "const_0_to_3_operand" "n")]
-		    UNSPEC_BYTEPICK_W))]
+	(ior:SI (lshiftrt (match_operand:SI 1 "register_operand" "r")
+			  (const_int <bytepick_w_lshiftrt_amount>))
+		(ashift (match_operand:SI 2 "register_operand" "r")
+			(const_int bytepick_w_ashift_amount))))]
   ""
-  "bytepick.w\t%0,%1,%2,%z3"
+  "bytepick.w\t%0,%1,%2,<bytepick_imm>"
   [(set_attr "mode" "SI")])
 
-(define_insn "bytepick_d"
+(define_insn "bytepick_w_<bytepick_imm>_extend"
   [(set (match_operand:DI 0 "register_operand" "=r")
-	(unspec:DI [(match_operand:DI 1 "register_operand" "r")
-		    (match_operand:DI 2 "register_operand" "r")
-		    (match_operand:DI 3 "const_0_to_7_operand" "n")]
-		    UNSPEC_BYTEPICK_D))]
-  ""
-  "bytepick.d\t%0,%1,%2,%z3"
+	(sign_extend:DI
+	  (ior:SI (lshiftrt (match_operand:SI 1 "register_operand" "r")
+			    (const_int <bytepick_w_lshiftrt_amount>))
+		  (ashift (match_operand:SI 2 "register_operand" "r")
+			  (const_int bytepick_w_ashift_amount)))))]
+  "TARGET_64BIT"
+  "bytepick.w\t%0,%1,%2,<bytepick_imm>"
+  [(set_attr "mode" "SI")])
+
+(define_insn "bytepick_d_<bytepick_imm>"
+  [(set (match_operand:DI 0 "register_operand" "=r")
+	(ior:DI (lshiftrt (match_operand:DI 1 "register_operand" "r")
+			  (const_int <bytepick_d_lshiftrt_amount>))
+		(ashift (match_operand:DI 2 "register_operand" "r")
+			(const_int bytepick_d_ashift_amount))))]
+  "TARGET_64BIT"
+  "bytepick.d\t%0,%1,%2,<bytepick_imm>"
   [(set_attr "mode" "DI")])
 
 (define_insn "bitrev_4b"
diff --git a/gcc/config/loongarch/predicates.md b/gcc/config/loongarch/predicates.md
index de0a363e2e4..95140280f1e 100644
--- a/gcc/config/loongarch/predicates.md
+++ b/gcc/config/loongarch/predicates.md
@@ -91,14 +91,6 @@
   (ior (match_operand 0 "const_1_operand")
        (match_operand 0 "register_operand")))
 
-(define_predicate "const_0_to_3_operand"
-  (and (match_code "const_int")
-       (match_test "IN_RANGE (INTVAL (op), 0, 3)")))
-
-(define_predicate "const_0_to_7_operand"
-  (and (match_code "const_int")
-       (match_test "IN_RANGE (INTVAL (op), 0, 7)")))
-
 (define_predicate "lu52i_mask_operand"
   (and (match_code "const_int")
        (match_test "UINTVAL (op) == 0xfffffffffffff")))
diff --git a/gcc/testsuite/g++.target/loongarch/bytepick.C b/gcc/testsuite/g++.target/loongarch/bytepick.C
new file mode 100644
index 00000000000..a39e2fa65b7
--- /dev/null
+++ b/gcc/testsuite/g++.target/loongarch/bytepick.C
@@ -0,0 +1,32 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -mabi=lp64d" } */
+/* { dg-final { scan-assembler-times "bytepick.w\t\\\$r4,\\\$r5,\\\$r4" 3 } } */
+/* { dg-final { scan-assembler-times "bytepick.d\t\\\$r4,\\\$r5,\\\$r4" 7 } } */
+/* { dg-final { scan-assembler-not "slli.w" } } */
+
+template <class T, int offs>
+T
+merge (T a, T b)
+{
+  return a << offs | b >> (8 * sizeof (T) - offs);
+}
+
+using u32 = __UINT32_TYPE__;
+using u64 = __UINT64_TYPE__;
+using i64 = __INT64_TYPE__;
+
+template u32 merge<u32, 8> (u32, u32);
+template u32 merge<u32, 16> (u32, u32);
+template u32 merge<u32, 24> (u32, u32);
+
+template u64 merge<u64, 8> (u64, u64);
+template u64 merge<u64, 16> (u64, u64);
+template u64 merge<u64, 24> (u64, u64);
+template u64 merge<u64, 32> (u64, u64);
+template u64 merge<u64, 40> (u64, u64);
+template u64 merge<u64, 48> (u64, u64);
+template u64 merge<u64, 56> (u64, u64);
+
+/* we cannot use bytepick for the following cases */
+template i64 merge<i64, 8> (i64, i64);
+template u64 merge<u64, 42> (u64, u64);