[ARM][MVE] Select VQNEG

[anwel]

Summary

Adds a pattern to ARMInstrMVE.td to use a VQNEG instruction if an equivalent multi-instruction construct is found.

[anwel]

Remove trailing whitespace from test

llvm/lib/Target/ARM/ARMInstrMVE.td

 def MVE_VQABSs32 : MVE_VQABSNEG<"vqabs", "s32", 0b10, 0b0>;
 
 def MVE_VQNEGs8  : MVE_VQABSNEG<"vqneg", "s8",  0b00, 0b1>;
 def MVE_VQNEGs16 : MVE_VQABSNEG<"vqneg", "s16", 0b01, 0b1>;
 def MVE_VQNEGs32 : MVE_VQABSNEG<"vqneg", "s32", 0b10, 0b1>;
 
 // int_min/int_max: vector containing INT_MIN/INT_MAX VTI.Size times
 // zero_vec: v4i32-initialized zero vector, potentially wrapped in a bitconvert
-multiclass vqabs_pattern<MVEVectorVTInfo VTI, dag int_min, dag int_max,
-                         dag zero_vec, MVE_VQABSNEG vqabs_instruction> {
+multiclass vqabsneg_pattern<MVEVectorVTInfo VTI, dag int_min, dag int_max,
+                         dag zero_vec,  MVE_VQABSNEG vqabs_instruction,
+                         MVE_VQABSNEG vqneg_instruction> {
+  let Predicates = [HasMVEInt] in {
     // The below tree can be replaced by a vqabs instruction, as it represents
     // the following vectorized expression (r being the value in $reg):
     // r > 0 ? r : (r == INT_MIN ? INT_MAX : -r)
-  let Predicates = [HasMVEInt] in {
     def : Pat<(VTI.Vec (vselect
                       (VTI.Pred (ARMvcmpz (VTI.Vec MQPR:$reg), (i32 12))),
                       (VTI.Vec MQPR:$reg),
                       (VTI.Vec (vselect
                                 (VTI.Pred (ARMvcmp (VTI.Vec MQPR:$reg), int_min, (i32 0))),
                                 int_max,
                                 (sub (VTI.Vec zero_vec), (VTI.Vec MQPR:$reg)))))),
             (VTI.Vec (vqabs_instruction (VTI.Vec MQPR:$reg)))>;
+    // Similarly, this tree represents vqneg, i.e. the following vectorized expression:
+    // r == INT_MIN ? INT_MAX : -r
+    def : Pat<(VTI.Vec (vselect
+                        (VTI.Pred (ARMvcmp (VTI.Vec MQPR:$reg), int_min, (i32 0))),
+                        int_max,
+                        (sub (VTI.Vec zero_vec), (VTI.Vec MQPR:$reg)))),
+               (VTI.Vec (vqneg_instruction (VTI.Vec MQPR:$reg)))>;
   }
 }
 
-defm MVE_VQABS_Ps8  : vqabs_pattern<MVE_v16i8,
+defm MVE_VQABSNEG_Ps8  : vqabsneg_pattern<MVE_v16i8,
                                     (v16i8 (ARMvmovImm (i32 3712))),
                                     (v16i8 (ARMvmovImm (i32 3711))),
                                     (bitconvert (v4i32 (ARMvmovImm (i32 0)))),
-                                    MVE_VQABSs8>;
-defm MVE_VQABS_Ps16 : vqabs_pattern<MVE_v8i16,
+                                    MVE_VQABSs8, MVE_VQNEGs8>;
+defm MVE_VQABSNEG_Ps16 : vqabsneg_pattern<MVE_v8i16,
                                     (v8i16 (ARMvmovImm (i32 2688))),
                                     (v8i16 (ARMvmvnImm (i32 2688))),
                                     (bitconvert (v4i32 (ARMvmovImm (i32 0)))),
-                                    MVE_VQABSs16>;
-defm MVE_VQABS_Ps32 : vqabs_pattern<MVE_v4i32,
+                                    MVE_VQABSs16, MVE_VQNEGs16>;
+defm MVE_VQABSNEG_Ps32 : vqabsneg_pattern<MVE_v4i32,
                                     (v4i32 (ARMvmovImm (i32 1664))),
                                     (v4i32 (ARMvmvnImm (i32 1664))),
                                     (ARMvmovImm (i32 0)),
-                                    MVE_VQABSs32>;
+                                    MVE_VQABSs32, MVE_VQNEGs32>;
 
 class MVE_mod_imm<string iname, string suffix, bits<4> cmode, bit op,
                   dag iops, list<dag> pattern=[]>
   : MVE_p<(outs MQPR:$Qd), iops, NoItinerary, iname, suffix, "$Qd, $imm",
           vpred_r, "", pattern> {
   bits<13> imm;
   bits<4> Qd;
 

llvm/test/CodeGen/Thumb2/vqneg.ll

+; RUN: llc -mtriple=thumbv8.1m.main-arm-none-eabi -mattr=+mve %s -o - | FileCheck %s
+
+define arm_aapcs_vfpcc <16 x i8> @vqneg_test16(<16 x i8> %A) nounwind {
+; CHECK-LABEL: vqneg_test16:
+; CHECK:       @ %bb.0: @ %entry
+; CHECK-NEXT:    vqneg.s8 q0, q0
+; CHECK-NEXT:    bx lr
+entry:
+
+  %0 = icmp eq <16 x i8> %A, <i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128, i8 -128>
+  %1 = sub nsw <16 x i8> zeroinitializer, %A
+  %2 = select <16 x i1> %0, <16 x i8> <i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127, i8 127>, <16 x i8> %1
+
+  ret <16 x i8> %2
+}
+
+define arm_aapcs_vfpcc <8 x i16> @vqneg_test8(<8 x i16> %A) nounwind {
+; CHECK-LABEL: vqneg_test8:
+; CHECK:       @ %bb.0: @ %entry
+; CHECK-NEXT:    vqneg.s16 q0, q0
+; CHECK-NEXT:    bx lr
+entry:
+
+  %0 = icmp eq <8 x i16> %A, <i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768, i16 -32768>
+  %1 = sub nsw <8 x i16> zeroinitializer, %A
+  %2 = select <8 x i1> %0, <8 x i16> <i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767, i16 32767>, <8 x i16> %1
+
+  ret <8 x i16> %2
+}
+
+define arm_aapcs_vfpcc <4 x i32> @vqneg_test4(<4 x i32> %A) nounwind {
+; CHECK-LABEL: vqneg_test4:
+; CHECK:       @ %bb.0: @ %entry
+; CHECK-NEXT:    vqneg.s32 q0, q0
+; CHECK-NEXT:    bx lr
+entry:
+
+  %0 = icmp eq <4 x i32> %A, <i32 -2147483648, i32 -2147483648, i32 -2147483648, i32 -2147483648>
+  %1 = sub nsw <4 x i32> zeroinitializer, %A
+  %2 = select <4 x i1> %0, <4 x i32> <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>, <4 x i32> %1
+
+  ret <4 x i32> %2
+}
+