[clang] [flang] [compiler-rt] [openmp] [libcxxabi] [libc] [clang-tools-extra] [libcxx] [llvm] [mlir] [AArch64] Add custom lowering for load <3 x i8>. (PR #78632)

Eli Friedman via cfe-commits cfe-commits at lists.llvm.org
Tue Jan 23 14:04:19 PST 2024


================
@@ -21248,6 +21248,51 @@ static SDValue foldTruncStoreOfExt(SelectionDAG &DAG, SDNode *N) {
   return SDValue();
 }
 
+// A custom combine to lower load <3 x i8> as the more efficient sequence
+// below:
+//    ldrb wX, [x0, #2]
+//    ldrh wY, [x0]
+//    orr wX, wY, wX, lsl #16
+//    fmov s0, wX
+//
+static SDValue combineV3I8LoadExt(LoadSDNode *LD, SelectionDAG &DAG) {
+  EVT MemVT = LD->getMemoryVT();
+  if (MemVT != EVT::getVectorVT(*DAG.getContext(), MVT::i8, 3) ||
+      LD->getOriginalAlign() >= 4)
+    return SDValue();
+
+  SDLoc DL(LD);
+  SDValue Chain = LD->getChain();
+  SDValue BasePtr = LD->getBasePtr();
+  assert(LD->getOffset().isUndef() && "undef offset expected");
+
+  // Load 2 x i8, then 1 x i8.
+  SDValue L16 = DAG.getLoad(MVT::i16, DL, Chain, BasePtr, LD->getPointerInfo(),
+                            LD->getOriginalAlign());
+  TypeSize Offset2 = TypeSize::getFixed(2);
+  SDValue L8 = DAG.getLoad(
+      MVT::i8, DL, Chain, DAG.getMemBasePlusOffset(BasePtr, Offset2, DL),
+      LD->getPointerInfo(), commonAlignment(LD->getOriginalAlign(), Offset2));
----------------
efriedma-quic wrote:

MachineFunction::getMachineMemOperand has an overload that takes an existing MachineMemOperand and adds an offset; that will produce a more accurate result here.

https://github.com/llvm/llvm-project/pull/78632


More information about the cfe-commits mailing list