[DAG] getNode() - remove oneuse limit from (zext (trunc (assertzext x))) -> (assertzext x) fold

Noticed on D159533 and I've finally deal with the x86 regressions - MatchingStackOffset wasn't peeking through AssertZext nodes while trying to find CopyFromReg/Load sources, it was only removing them if they were part of a (trunc (assertzext x)) pattern.

Author: RKSimon

llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp

@@ -5700,7 +5700,7 @@ SDValue SelectionDAG::getNode(unsigned Opcode, const SDLoc &DL, EVT VT,
     if (OpOpcode == ISD::TRUNCATE) {
       SDValue OpOp = N1.getOperand(0);
       if (OpOp.getValueType() == VT) {
-        if (OpOp.getOpcode() == ISD::AssertZext && N1->hasOneUse()) {
+        if (OpOp.getOpcode() == ISD::AssertZext) {
           APInt HiBits = APInt::getBitsSetFrom(VT.getScalarSizeInBits(),
                                                N1.getScalarValueSizeInBits());
           if (MaskedValueIsZero(OpOp, HiBits)) {

llvm/lib/Target/X86/X86ISelLoweringCall.cpp

@@ -2645,7 +2645,8 @@ bool MatchingStackOffset(SDValue Arg, unsigned Offset, ISD::ArgFlagsTy Flags,
   for (;;) {
     // Look through nodes that don't alter the bits of the incoming value.
     unsigned Op = Arg.getOpcode();
-    if (Op == ISD::ZERO_EXTEND || Op == ISD::ANY_EXTEND || Op == ISD::BITCAST) {
+    if (Op == ISD::ZERO_EXTEND || Op == ISD::ANY_EXTEND || Op == ISD::BITCAST ||
+        Op == ISD::AssertZext) {
       Arg = Arg.getOperand(0);
       continue;
     }

llvm/test/CodeGen/AArch64/setcc_knownbits.ll

@@ -4,8 +4,6 @@
 define i1 @load_bv_v4i8(i1 zeroext %a) {
 ; CHECK-LABEL: load_bv_v4i8:
 ; CHECK:       // %bb.0:
-; CHECK-NEXT:    cmp w0, #0
-; CHECK-NEXT:    cset w0, ne
 ; CHECK-NEXT:    ret
   %b = zext i1 %a to i32
   %c = icmp eq i32 %b, 1

llvm/test/CodeGen/RISCV/rvv/fold-vp-fadd-and-vp-fmul.ll

@@ -62,9 +62,9 @@ define <vscale x 1 x double> @fma_reassociate(<vscale x 1 x double> %a, <vscale
 ; CHECK-LABEL: fma_reassociate:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetvli zero, a0, e64, m1, ta, ma
-; CHECK-NEXT:    vfmadd.vv v9, v8, v12, v0.t
-; CHECK-NEXT:    vfmadd.vv v11, v10, v9, v0.t
-; CHECK-NEXT:    vmv.v.v v8, v11
+; CHECK-NEXT:    vfmadd.vv v11, v10, v12, v0.t
+; CHECK-NEXT:    vfmadd.vv v9, v8, v11, v0.t
+; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %1 = call fast <vscale x 1 x double> @llvm.vp.fmul.nxv1f64(<vscale x 1 x double> %a, <vscale x 1 x double> %b, <vscale x 1 x i1> %m, i32 %vl)
   %2 = call fast <vscale x 1 x double> @llvm.vp.fmul.nxv1f64(<vscale x 1 x double> %c, <vscale x 1 x double> %d, <vscale x 1 x i1> %m, i32 %vl)