[RFC PATCH] tcg/ppc: implement rem[u]_i{32,64} with mod[su][wd]

[Matheus Kowalczuk Ferst]

Power ISA v3.0 introduced mod[su][wd] insns that can be used to
implement rem[u]_i{32,64}.

Signed-off-by: Matheus Ferst <matheus.ferst@eldorado.org.br>
---
 tcg/ppc/tcg-target.c.inc | 22 ++++++++++++++++++++++
 tcg/ppc/tcg-target.h     |  4 ++--
 2 files changed, 24 insertions(+), 2 deletions(-)

diff --git a/tcg/ppc/tcg-target.c.inc b/tcg/ppc/tcg-target.c.inc
index de4483e43b..1cbd047ab3 100644
--- a/tcg/ppc/tcg-target.c.inc
+++ b/tcg/ppc/tcg-target.c.inc
@@ -371,6 +371,8 @@ static bool tcg_target_const_match(int64_t val, TCGType 
type, int ct)
 #define MULHWU XO31( 11)
 #define DIVW   XO31(491)
 #define DIVWU  XO31(459)
+#define MODSW  XO31(779)
+#define MODUW  XO31(267)
 #define CMP    XO31(  0)
 #define CMPL   XO31( 32)
 #define LHBRX  XO31(790)
@@ -403,6 +405,8 @@ static bool tcg_target_const_match(int64_t val, TCGType 
type, int ct)
 #define MULHDU XO31(  9)
 #define DIVD   XO31(489)
 #define DIVDU  XO31(457)
+#define MODSD  XO31(777)
+#define MODUD  XO31(265)
 
 #define LBZX   XO31( 87)
 #define LHZX   XO31(279)
@@ -2806,6 +2810,14 @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc,
         tcg_out32(s, DIVWU | TAB(args[0], args[1], args[2]));
         break;
 
+    case INDEX_op_rem_i32:
+        tcg_out32(s, MODSW | TAB(args[0], args[1], args[2]));
+        break;
+
+    case INDEX_op_remu_i32:
+        tcg_out32(s, MODUW | TAB(args[0], args[1], args[2]));
+        break;
+
     case INDEX_op_shl_i32:
         if (const_args[2]) {
             /* Limit immediate shift count lest we create an illegal insn.  */
@@ -2947,6 +2959,12 @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc,
     case INDEX_op_divu_i64:
         tcg_out32(s, DIVDU | TAB(args[0], args[1], args[2]));
         break;
+    case INDEX_op_rem_i64:
+        tcg_out32(s, MODSD | TAB(args[0], args[1], args[2]));
+        break;
+    case INDEX_op_remu_i64:
+        tcg_out32(s, MODUD | TAB(args[0], args[1], args[2]));
+        break;
 
     case INDEX_op_qemu_ld_i32:
         tcg_out_qemu_ld(s, args, false);
@@ -3722,6 +3740,8 @@ static TCGConstraintSetIndex tcg_target_op_def(TCGOpcode 
op)
 
     case INDEX_op_div_i32:
     case INDEX_op_divu_i32:
+    case INDEX_op_rem_i32:
+    case INDEX_op_remu_i32:
     case INDEX_op_nand_i32:
     case INDEX_op_nor_i32:
     case INDEX_op_muluh_i32:
@@ -3732,6 +3752,8 @@ static TCGConstraintSetIndex tcg_target_op_def(TCGOpcode 
op)
     case INDEX_op_nor_i64:
     case INDEX_op_div_i64:
     case INDEX_op_divu_i64:
+    case INDEX_op_rem_i64:
+    case INDEX_op_remu_i64:
     case INDEX_op_mulsh_i64:
     case INDEX_op_muluh_i64:
         return C_O1_I2(r, r, r);
diff --git a/tcg/ppc/tcg-target.h b/tcg/ppc/tcg-target.h
index e6cf72503f..b5cd225cfa 100644
--- a/tcg/ppc/tcg-target.h
+++ b/tcg/ppc/tcg-target.h
@@ -83,7 +83,7 @@ extern bool have_vsx;
 
 /* optional instructions */
 #define TCG_TARGET_HAS_div_i32          1
-#define TCG_TARGET_HAS_rem_i32          0
+#define TCG_TARGET_HAS_rem_i32          have_isa_3_00
 #define TCG_TARGET_HAS_rot_i32          1
 #define TCG_TARGET_HAS_ext8s_i32        1
 #define TCG_TARGET_HAS_ext16s_i32       1
@@ -117,7 +117,7 @@ extern bool have_vsx;
 #define TCG_TARGET_HAS_extrl_i64_i32    0
 #define TCG_TARGET_HAS_extrh_i64_i32    0
 #define TCG_TARGET_HAS_div_i64          1
-#define TCG_TARGET_HAS_rem_i64          0
+#define TCG_TARGET_HAS_rem_i64          have_isa_3_00
 #define TCG_TARGET_HAS_rot_i64          1
 #define TCG_TARGET_HAS_ext8s_i64        1
 #define TCG_TARGET_HAS_ext16s_i64       1
-- 
2.25.1