PPC: Optimize rldicl generation for masked shifts

Masking operations (where only some number of the low bits are being kept) are
selected to rldicl(x, 0, mb). If x is a logical right shift (which would become
rldicl(y, 64-n, n)), we might be able to fold the two instructions together:

  rldicl(rldicl(x, 64-n, n), 0, mb) -> rldicl(x, 64-n, mb) for n <= mb

The right shift is really a left rotate followed by a mask, and if the explicit
mask is a more-restrictive sub-mask of the mask implied by the shift, only one
rldicl is needed.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@195185 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/PowerPC/PPCISelDAGToDAG.cpp b/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
index 6ba6af6446e5c14163a98e64d52edbf6f51fad01..324b82161800dbbf9e5ed2b823ffda68208f05ea 100644 (file)
--- a/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
+++ b/lib/Target/PowerPC/PPCISelDAGToDAG.cpp
@@ -1122,7 +1122,21 @@ SDNode *PPCDAGToDAGISel::Select(SDNode *N) {
         isMask_64(Imm64)) {
       SDValue Val = N->getOperand(0);
       MB = 64 - CountTrailingOnes_64(Imm64);
-      SDValue Ops[] = { Val, getI32Imm(0), getI32Imm(MB) };
+      SH = 0;
+
+      // If the operand is a logical right shift, we can fold it into this
+      // instruction: rldicl(rldicl(x, 64-n, n), 0, mb) -> rldicl(x, 64-n, mb)
+      // for n <= mb. The right shift is really a left rotate followed by a
+      // mask, and this mask is a more-restrictive sub-mask of the mask implied
+      // by the shift.
+      if (Val.getOpcode() == ISD::SRL &&
+          isInt32Immediate(Val.getOperand(1).getNode(), Imm) && Imm <= MB) {
+        assert(Imm < 64 && "Illegal shift amount");
+        Val = Val.getOperand(0);
+        SH = 64 - Imm;
+      }
+
+      SDValue Ops[] = { Val, getI32Imm(SH), getI32Imm(MB) };
       return CurDAG->SelectNodeTo(N, PPC::RLDICL, MVT::i64, Ops, 3);
     }
     // AND X, 0 -> 0, not "rlwinm 32".

diff --git a/test/CodeGen/PowerPC/srl-mask.ll b/test/CodeGen/PowerPC/srl-mask.ll
new file mode 100644 (file)
index 0000000..2749df9
--- /dev/null
+++ b/test/CodeGen/PowerPC/srl-mask.ll
@@ -0,0 +1,16 @@
+; RUN: llc < %s -mtriple=powerpc64-unknown-linux-gnu -mcpu=a2 | FileCheck %s
+target datalayout = "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v128:128:128-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
+
+define i64 @foo(i64 %x) #0 {
+entry:
+; CHECK-LABEL: @foo
+  %a = lshr i64 %x, 35
+  %b = and i64 %a, 65535
+; CHECK: rldicl 3, 3, 29, 48
+  ret i64 %b
+; CHECK: blr
+}
+
+attributes #0 = { nounwind }
+