[PowerPC] Add asm parser support for bitmask forms of rotate-and-mask instructions

Hal Finkel · Hal Finkel · commit 6e9110abe923 · 2015-03-28T19:42:41.000Z
The asm syntax for the 32-bit rotate-and-mask instructions can take a 32-bit
bitmask instead of an (mb, me) pair. This syntax is not specified in the Power
ISA manual, but is accepted by GNU as, and is documented in IBM's Assembler
Language Reference. The GNU Multiple Precision Arithmetic Library (gmp)
contains assembly that uses this syntax.

To implement this, I moved the isRunOfOnes utility function from
PPCISelDAGToDAG.cpp to PPCMCTargetDesc.h.

llvm-svn: 233483
diff --git a/llvm/lib/Target/PowerPC/AsmParser/PPCAsmParser.cpp b/llvm/lib/Target/PowerPC/AsmParser/PPCAsmParser.cpp
@@ -1071,6 +1071,58 @@ void PPCAsmParser::ProcessInstruction(MCInst &Inst,
     Inst = TmpInst;
     break;
   }
+  case PPC::RLWINMbm:
+  case PPC::RLWINMobm: {
+    unsigned MB, ME;
+    int64_t BM = Inst.getOperand(3).getImm();
+    if (!isRunOfOnes(BM, MB, ME))
+      break;
+
+    MCInst TmpInst;
+    TmpInst.setOpcode(Opcode == PPC::RLWINMbm ? PPC::RLWINM : PPC::RLWINMo);
+    TmpInst.addOperand(Inst.getOperand(0));
+    TmpInst.addOperand(Inst.getOperand(1));
+    TmpInst.addOperand(Inst.getOperand(2));
+    TmpInst.addOperand(MCOperand::CreateImm(MB));
+    TmpInst.addOperand(MCOperand::CreateImm(ME));
+    Inst = TmpInst;
+    break;
+  }
+  case PPC::RLWIMIbm:
+  case PPC::RLWIMIobm: {
+    unsigned MB, ME;
+    int64_t BM = Inst.getOperand(3).getImm();
+    if (!isRunOfOnes(BM, MB, ME))
+      break;
+
+    MCInst TmpInst;
+    TmpInst.setOpcode(Opcode == PPC::RLWIMIbm ? PPC::RLWIMI : PPC::RLWIMIo);
+    TmpInst.addOperand(Inst.getOperand(0));
+    TmpInst.addOperand(Inst.getOperand(0)); // The tied operand.
+    TmpInst.addOperand(Inst.getOperand(1));
+    TmpInst.addOperand(Inst.getOperand(2));
+    TmpInst.addOperand(MCOperand::CreateImm(MB));
+    TmpInst.addOperand(MCOperand::CreateImm(ME));
+    Inst = TmpInst;
+    break;
+  }
+  case PPC::RLWNMbm:
+  case PPC::RLWNMobm: {
+    unsigned MB, ME;
+    int64_t BM = Inst.getOperand(3).getImm();
+    if (!isRunOfOnes(BM, MB, ME))
+      break;
+
+    MCInst TmpInst;
+    TmpInst.setOpcode(Opcode == PPC::RLWNMbm ? PPC::RLWNM : PPC::RLWNMo);
+    TmpInst.addOperand(Inst.getOperand(0));
+    TmpInst.addOperand(Inst.getOperand(1));
+    TmpInst.addOperand(Inst.getOperand(2));
+    TmpInst.addOperand(MCOperand::CreateImm(MB));
+    TmpInst.addOperand(MCOperand::CreateImm(ME));
+    Inst = TmpInst;
+    break;
+  }
   }
 }
 
diff --git a/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCTargetDesc.h b/llvm/lib/Target/PowerPC/MCTargetDesc/PPCMCTargetDesc.h
@@ -18,6 +18,7 @@
 #undef PPC
 
 #include "llvm/Support/DataTypes.h"
+#include "llvm/Support/MathExtras.h"
 
 namespace llvm {
 class MCAsmBackend;
@@ -51,6 +52,35 @@ MCObjectWriter *createPPCELFObjectWriter(raw_ostream &OS,
 MCObjectWriter *createPPCMachObjectWriter(raw_ostream &OS, bool Is64Bit,
                                           uint32_t CPUType,
                                           uint32_t CPUSubtype);
+
+/// isRunOfOnes - Returns true iff Val consists of one contiguous run of 1s with
+/// any number of 0s on either side.  The 1s are allowed to wrap from LSB to
+/// MSB, so 0x000FFF0, 0x0000FFFF, and 0xFF0000FF are all runs.  0x0F0F0000 is
+/// not, since all 1s are not contiguous.
+static inline bool isRunOfOnes(unsigned Val, unsigned &MB, unsigned &ME) {
+  if (!Val)
+    return false;
+
+  if (isShiftedMask_32(Val)) {
+    // look for the first non-zero bit
+    MB = countLeadingZeros(Val);
+    // look for the first zero bit after the run of ones
+    ME = countLeadingZeros((Val - 1) ^ Val);
+    return true;
+  } else {
+    Val = ~Val; // invert mask
+    if (isShiftedMask_32(Val)) {
+      // effectively look for the first zero bit
+      ME = countLeadingZeros(Val) - 1;
+      // effectively look for the first one bit after the run of zeros
+      MB = countLeadingZeros((Val - 1) ^ Val) + 1;
+      return true;
+    }
+  }
+  // no run present
+  return false;
+}
+
 } // End llvm namespace
 
 // Generated files will use "namespace PPC". To avoid symbol clash,
diff --git a/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -105,13 +105,6 @@ namespace {
       return CurDAG->getTargetConstant(Imm, PPCLowering->getPointerTy());
     }
 
-    /// isRunOfOnes - Returns true iff Val consists of one contiguous run of 1s
-    /// with any number of 0s on either side.  The 1s are allowed to wrap from
-    /// LSB to MSB, so 0x000FFF0, 0x0000FFFF, and 0xFF0000FF are all runs.
-    /// 0x0F0F0000 is not, since all 1s are not contiguous.
-    static bool isRunOfOnes(unsigned Val, unsigned &MB, unsigned &ME);
-
-
     /// isRotateAndMask - Returns true if Mask and Shift can be folded into a
     /// rotate and mask opcode and mask operation.
     static bool isRotateAndMask(SDNode *N, unsigned Mask, bool isShiftMask,
@@ -418,30 +411,6 @@ SDNode *PPCDAGToDAGISel::getFrameIndex(SDNode *SN, SDNode *N, unsigned Offset) {
                                 getSmallIPtrImm(Offset));
 }
 
-bool PPCDAGToDAGISel::isRunOfOnes(unsigned Val, unsigned &MB, unsigned &ME) {
-  if (!Val)
-    return false;
-
-  if (isShiftedMask_32(Val)) {
-    // look for the first non-zero bit
-    MB = countLeadingZeros(Val);
-    // look for the first zero bit after the run of ones
-    ME = countLeadingZeros((Val - 1) ^ Val);
-    return true;
-  } else {
-    Val = ~Val; // invert mask
-    if (isShiftedMask_32(Val)) {
-      // effectively look for the first zero bit
-      ME = countLeadingZeros(Val) - 1;
-      // effectively look for the first one bit after the run of zeros
-      MB = countLeadingZeros((Val - 1) ^ Val) + 1;
-      return true;
-    }
-  }
-  // no run present
-  return false;
-}
-
 bool PPCDAGToDAGISel::isRotateAndMask(SDNode *N, unsigned Mask,
                                       bool isShiftMask, unsigned &SH,
                                       unsigned &MB, unsigned &ME) {
diff --git a/llvm/lib/Target/PowerPC/PPCInstrInfo.td b/llvm/lib/Target/PowerPC/PPCInstrInfo.td
@@ -3726,6 +3726,19 @@ def : InstAlias<"rotld. $rA, $rS, $rB", (RLDCLo g8rc:$rA, g8rc:$rS, gprc:$rB, 0)
 def : InstAlias<"clrldi $rA, $rS, $n", (RLDICL g8rc:$rA, g8rc:$rS, 0, u6imm:$n)>;
 def : InstAlias<"clrldi. $rA, $rS, $n", (RLDICLo g8rc:$rA, g8rc:$rS, 0, u6imm:$n)>;
 
+def RLWINMbm : PPCAsmPseudo<"rlwinm $rA, $rS, $n, $b",
+                            (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+def RLWINMobm : PPCAsmPseudo<"rlwinm. $rA, $rS, $n, $b",
+                            (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+def RLWIMIbm : PPCAsmPseudo<"rlwimi $rA, $rS, $n, $b",
+                           (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+def RLWIMIobm : PPCAsmPseudo<"rlwimi. $rA, $rS, $n, $b",
+                            (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+def RLWNMbm : PPCAsmPseudo<"rlwnm $rA, $rS, $n, $b",
+                          (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+def RLWNMobm : PPCAsmPseudo<"rlwnm. $rA, $rS, $n, $b",
+                           (ins g8rc:$rA, g8rc:$rS, u5imm:$n, i32imm:$b)>;
+
 // These generic branch instruction forms are used for the assembler parser only.
 // Defs and Uses are conservative, since we don't know the BO value.
 let PPC970_Unit = 7 in {
diff --git a/llvm/test/MC/PowerPC/ppc64-encoding.s b/llvm/test/MC/PowerPC/ppc64-encoding.s
@@ -703,6 +703,33 @@
 # CHECK-LE: rldimi. 2, 3, 4, 5              # encoding: [0x4d,0x21,0x62,0x78]
             rldimi. 2, 3, 4, 5
 
+# Aliases that take bit masks...
+
+# CHECK-BE: rlwinm  0, 0, 30, 31, 31        # encoding: [0x54,0x00,0xf7,0xfe]
+            rlwinm  0, 0, 30, 1
+# CHECK-BE: rlwinm. 0, 0, 30, 31, 31        # encoding: [0x54,0x00,0xf7,0xff]
+            rlwinm. 0, 0, 30, 1
+# CHECK-BE: rlwinm  0, 0, 30, 31, 0         # encoding: [0x54,0x00,0xf7,0xc0]
+            rlwinm  0, 0, 30, 2147483649
+# CHECK-BE: rlwinm. 0, 0, 30, 31, 0         # encoding: [0x54,0x00,0xf7,0xc1]
+            rlwinm. 0, 0, 30, 2147483649
+# CHECK-BE: rlwimi  0, 0, 30, 31, 31        # encoding: [0x50,0x00,0xf7,0xfe]
+            rlwimi  0, 0, 30, 1
+# CHECK-BE: rlwimi. 0, 0, 30, 31, 31        # encoding: [0x50,0x00,0xf7,0xff]
+            rlwimi. 0, 0, 30, 1
+# CHECK-BE: rlwimi  0, 0, 30, 31, 0         # encoding: [0x50,0x00,0xf7,0xc0]
+            rlwimi  0, 0, 30, 2147483649
+# CHECK-BE: rlwimi. 0, 0, 30, 31, 0         # encoding: [0x50,0x00,0xf7,0xc1]
+            rlwimi. 0, 0, 30, 2147483649
+# CHECK-BE: rlwnm   0, 0, 30, 31, 31        # encoding: [0x5c,0x00,0xf7,0xfe]
+            rlwnm  0, 0, 30, 1
+# CHECK-BE: rlwnm.  0, 0, 30, 31, 31        # encoding: [0x5c,0x00,0xf7,0xff]
+            rlwnm. 0, 0, 30, 1
+# CHECK-BE: rlwnm   0, 0, 30, 31, 0         # encoding: [0x5c,0x00,0xf7,0xc0]
+            rlwnm  0, 0, 30, 2147483649
+# CHECK-BE: rlwnm.  0, 0, 30, 31, 0         # encoding: [0x5c,0x00,0xf7,0xc1]
+            rlwnm. 0, 0, 30, 2147483649
+
 # CHECK-BE: slw 2, 3, 4                     # encoding: [0x7c,0x62,0x20,0x30]
 # CHECK-LE: slw 2, 3, 4                     # encoding: [0x30,0x20,0x62,0x7c]
             slw 2, 3, 4
