ARM64: Avoid side-effects of constant rematerialization.

Thanks to Patrick Galizia.

1 files changed, 29 insertions, 17 deletions

diff --git a/src/lj_asm_arm64.h b/src/lj_asm_arm64.h
index baafa21a..ebe2e359 100644
--- a/src/lj_asm_arm64.h
+++ b/src/lj_asm_arm64.h
@@ -295,8 +295,10 @@ static void asm_fusexref(ASMState *as, A64Ins ai, Reg rd, IRRef ref,
       } else if (asm_isk32(as, ir->op1, &ofs)) {
         ref = ir->op2;
       } else {
-        Reg rn = ra_alloc1(as, ir->op1, allow);
-        IRIns *irr = IR(ir->op2);
+        Reg refk = irref_isk(ir->op1) ? ir->op1 : ir->op2;
+        Reg refv = irref_isk(ir->op1) ? ir->op2 : ir->op1;
+        Reg rn = ra_alloc1(as, refv, allow);
+        IRIns *irr = IR(refk);
         uint32_t m;
         if (irr+1 == ir && !ra_used(irr) &&
             irr->o == IR_ADD && irref_isk(irr->op2)) {
@@ -307,7 +309,7 @@ static void asm_fusexref(ASMState *as, A64Ins ai, Reg rd, IRRef ref,
             goto skipopm;
           }
         }
-        m = asm_fuseopm(as, 0, ir->op2, rset_exclude(allow, rn));
+        m = asm_fuseopm(as, 0, refk, rset_exclude(allow, rn));
         ofs = sizeof(GCstr);
       skipopm:
         emit_lso(as, ai, rd, rd, ofs);
@@ -722,6 +724,7 @@ static void asm_href(ASMState *as, IRIns *ir, IROp merge)
   Reg dest = ra_dest(as, ir, allow);
   Reg tab = ra_alloc1(as, ir->op1, rset_clear(allow, dest));
   Reg key = 0, tmp = RID_TMP;
+  Reg ftmp = RID_NONE, type = RID_NONE, scr = RID_NONE, tisnum = RID_NONE;
   IRRef refkey = ir->op2;
   IRIns *irkey = IR(refkey);
   int isk = irref_isk(ir->op2);
@@ -751,6 +754,28 @@ static void asm_href(ASMState *as, IRIns *ir, IROp merge)
     }
   }
 
+  /* Allocate constants early. */
+  if (irt_isnum(kt)) {
+    if (!isk) {
+      tisnum = ra_allock(as, LJ_TISNUM << 15, allow);
+      ftmp = ra_scratch(as, rset_exclude(RSET_FPR, key));
+      rset_clear(allow, tisnum);
+    }
+  } else if (irt_isaddr(kt)) {
+    if (isk) {
+      int64_t kk = ((int64_t)irt_toitype(irkey->t) << 47) | irkey[1].tv.u64;
+      scr = ra_allock(as, kk, allow);
+    } else {
+      scr = ra_scratch(as, allow);
+    }
+    rset_clear(allow, scr);
+  } else {
+    lua_assert(irt_ispri(kt) && !irt_isnil(kt));
+    type = ra_allock(as, ~((int64_t)~irt_toitype(ir->t) << 47), allow);
+    scr = ra_scratch(as, rset_clear(allow, type));
+    rset_clear(allow, scr);
+  }
+
   /* Key not found in chain: jump to exit (if merged) or load niltv. */
   l_end = emit_label(as);
   as->invmcp = NULL;
@@ -780,9 +805,6 @@ static void asm_href(ASMState *as, IRIns *ir, IROp merge)
         emit_nm(as, A64I_CMPx, key, tmp);
       emit_lso(as, A64I_LDRx, tmp, dest, offsetof(Node, key.u64));
     } else {
-      Reg tisnum = ra_allock(as, LJ_TISNUM << 15, allow);
-      Reg ftmp = ra_scratch(as, rset_exclude(RSET_FPR, key));
-      rset_clear(allow, tisnum);
       emit_nm(as, A64I_FCMPd, key, ftmp);
       emit_dn(as, A64I_FMOV_D_R, (ftmp & 31), (tmp & 31));
       emit_cond_branch(as, CC_LO, l_next);
@@ -790,31 +812,21 @@ static void asm_href(ASMState *as, IRIns *ir, IROp merge)
       emit_lso(as, A64I_LDRx, tmp, dest, offsetof(Node, key.n));
     }
   } else if (irt_isaddr(kt)) {
-    Reg scr;
     if (isk) {
-      int64_t kk = ((int64_t)irt_toitype(irkey->t) << 47) | irkey[1].tv.u64;
-      scr = ra_allock(as, kk, allow);
       emit_nm(as, A64I_CMPx, scr, tmp);
       emit_lso(as, A64I_LDRx, tmp, dest, offsetof(Node, key.u64));
     } else {
-      scr = ra_scratch(as, allow);
       emit_nm(as, A64I_CMPx, tmp, scr);
       emit_lso(as, A64I_LDRx, scr, dest, offsetof(Node, key.u64));
     }
-    rset_clear(allow, scr);
   } else {
-    Reg type, scr;
-    lua_assert(irt_ispri(kt) && !irt_isnil(kt));
-    type = ra_allock(as, ~((int64_t)~irt_toitype(ir->t) << 47), allow);
-    scr = ra_scratch(as, rset_clear(allow, type));
-    rset_clear(allow, scr);
     emit_nm(as, A64I_CMPw, scr, type);
     emit_lso(as, A64I_LDRx, scr, dest, offsetof(Node, key));
   }
 
   *l_loop = A64I_BCC | A64F_S19(as->mcp - l_loop) | CC_NE;
   if (!isk && irt_isaddr(kt)) {
-    Reg type = ra_allock(as, (int32_t)irt_toitype(kt), allow);
+    type = ra_allock(as, (int32_t)irt_toitype(kt), allow);
     emit_dnm(as, A64I_ADDx | A64F_SH(A64SH_LSL, 47), tmp, key, type);
     rset_clear(allow, type);
   }