[AVX512] adding PROLQ and PROLD Intrinsics

[AsafBadouh]

LGTM

lib/Target/X86/X86ISelLowering.h

319

typo: Roted -> Rotate

Bit rotation left

[delena]

llvm/trunk/lib/Target/X86/X86ISelLowering.h

319 ↗

(On Diff #44665)

try to use ISD::ROTL, please

include/llvm/IR/IntrinsicsX86.td

def int_x86_avx512_mask_prorv_q_128 : GCCBuiltin<"__builtin_ia32_prorvq128_mask">,
                          llvm_v2i64_ty, llvm_v2i64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_prorv_q_256 : GCCBuiltin<"__builtin_ia32_prorvq256_mask">,
               Intrinsic<[llvm_v4i64_ty], [llvm_v4i64_ty,
                          llvm_v4i64_ty, llvm_v4i64_ty, llvm_i8_ty], [IntrNoMem]>;
   def int_x86_avx512_mask_prorv_q_512 : GCCBuiltin<"__builtin_ia32_prorvq512_mask">,
               Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty,
                          llvm_v8i64_ty, llvm_v8i64_ty, llvm_i8_ty], [IntrNoMem]>;
 
+   def int_x86_avx512_mask_prol_d_128 : GCCBuiltin<"__builtin_ia32_prold128_mask">,
+              Intrinsic<[llvm_v4i32_ty] , [llvm_v4i32_ty,
+                         llvm_i8_ty, llvm_v4i32_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_prol_d_256 : GCCBuiltin<"__builtin_ia32_prold256_mask">,
+              Intrinsic<[llvm_v8i32_ty] , [llvm_v8i32_ty,
+                         llvm_i8_ty, llvm_v8i32_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_prol_d_512 : GCCBuiltin<"__builtin_ia32_prold512_mask">,
+              Intrinsic<[llvm_v16i32_ty] , [llvm_v16i32_ty,
+                         llvm_i8_ty, llvm_v16i32_ty, llvm_i16_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_prol_q_128 : GCCBuiltin<"__builtin_ia32_prolq128_mask">,
+              Intrinsic<[llvm_v2i64_ty] , [llvm_v2i64_ty,
+                         llvm_i8_ty, llvm_v2i64_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_prol_q_256 : GCCBuiltin<"__builtin_ia32_prolq256_mask">,
+              Intrinsic<[llvm_v4i64_ty] , [llvm_v4i64_ty,
+                         llvm_i8_ty, llvm_v4i64_ty, llvm_i8_ty], [IntrNoMem]>;
+  def int_x86_avx512_mask_prol_q_512 : GCCBuiltin<"__builtin_ia32_prolq512_mask">,
+              Intrinsic<[llvm_v8i64_ty] , [llvm_v8i64_ty,
+                         llvm_i8_ty, llvm_v8i64_ty, llvm_i8_ty], [IntrNoMem]>;
+
+
 }
 
 // Gather ops
 let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
   def int_x86_avx2_gather_d_pd : GCCBuiltin<"__builtin_ia32_gatherd_pd">,
       Intrinsic<[llvm_v2f64_ty],
         [llvm_v2f64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v2f64_ty, llvm_i8_ty],
         [IntrReadArgMem]>;

lib/Target/X86/X86ISelLowering.h

enum NodeType : unsigned {
       VSHLDQ, VSRLDQ,
 
       // Vector shift elements
       VSHL, VSRL, VSRA,
 
       // Vector shift elements by immediate
       VSHLI, VSRLI, VSRAI,
 
+      // Bit Roted By immediate
+
+      VROTLI,
+      
       // Vector packed double/float comparison.
       CMPP,
 
       // Vector integer comparisons.
       PCMPEQ, PCMPGT,
       // Vector integer comparisons, the result is in a mask vector.
       PCMPEQM, PCMPGTM,
 

lib/Target/X86/X86ISelLowering.cpp

const char *X86TargetLowering::getTargetNodeName(unsigned Opcode) const {
   case X86ISD::VSHLDQ:             return "X86ISD::VSHLDQ";
   case X86ISD::VSRLDQ:             return "X86ISD::VSRLDQ";
   case X86ISD::VSHL:               return "X86ISD::VSHL";
   case X86ISD::VSRL:               return "X86ISD::VSRL";
   case X86ISD::VSRA:               return "X86ISD::VSRA";
   case X86ISD::VSHLI:              return "X86ISD::VSHLI";
   case X86ISD::VSRLI:              return "X86ISD::VSRLI";
   case X86ISD::VSRAI:              return "X86ISD::VSRAI";
+  case X86ISD::VROTLI:             return "X86ISD::VROTLI";
   case X86ISD::CMPP:               return "X86ISD::CMPP";
   case X86ISD::PCMPEQ:             return "X86ISD::PCMPEQ";
   case X86ISD::PCMPGT:             return "X86ISD::PCMPGT";
   case X86ISD::PCMPEQM:            return "X86ISD::PCMPEQM";
   case X86ISD::PCMPGTM:            return "X86ISD::PCMPGTM";
   case X86ISD::ADD:                return "X86ISD::ADD";
   case X86ISD::SUB:                return "X86ISD::SUB";
   case X86ISD::ADC:                return "X86ISD::ADC";

lib/Target/X86/X86InstrAVX512.td

 
 defm VPSLL : avx512_shift_rmi_dq<0x72, 0x73, MRM6r, MRM6m, "vpsll", X86vshli>,
              avx512_shift_rmi_w<0x71, MRM6r, MRM6m, "vpsllw", X86vshli>, AVX512BIi8Base, EVEX_4V;
 
 defm VPSRA : avx512_shift_rmi_dq<0x72, 0x72, MRM4r, MRM4m, "vpsra", X86vsrai>,
              avx512_shift_rmi_w<0x71, MRM4r, MRM4m, "vpsraw", X86vsrai>, AVX512BIi8Base, EVEX_4V;
 
 defm VPROR : avx512_shift_rmi_dq<0x72, 0x72, MRM0r, MRM0m, "vpror", rotr>, AVX512BIi8Base, EVEX_4V;
-defm VPROL : avx512_shift_rmi_dq<0x72, 0x72, MRM1r, MRM1m, "vprol", rotl>, AVX512BIi8Base, EVEX_4V;
+defm VPROL : avx512_shift_rmi_dq<0x72, 0x72, MRM1r, MRM1m, "vprol", X86vrotli>, AVX512BIi8Base, EVEX_4V;
 
 defm VPSLL : avx512_shift_types<0xF2, 0xF3, 0xF1, "vpsll", X86vshl>;
 defm VPSRA : avx512_shift_types<0xE2, 0xE2, 0xE1, "vpsra", X86vsra>;
 defm VPSRL : avx512_shift_types<0xD2, 0xD3, 0xD1, "vpsrl", X86vsrl>;
 
 //===-------------------------------------------------------------------===//
 // Variable Bit Shifts
 //===-------------------------------------------------------------------===//

lib/Target/X86/X86InstrFragmentsSIMD.td

 def X86vsra    : SDNode<"X86ISD::VSRA",
                         SDTypeProfile<1, 2, [SDTCisVec<0>, SDTCisSameAs<0,1>,
                                       SDTCisVec<2>]>>;
 
 def X86vshli   : SDNode<"X86ISD::VSHLI", SDTIntShiftOp>;
 def X86vsrli   : SDNode<"X86ISD::VSRLI", SDTIntShiftOp>;
 def X86vsrai   : SDNode<"X86ISD::VSRAI", SDTIntShiftOp>;
 
+def X86vrotli  : SDNode<"X86ISD::VROTLI", SDTIntShiftOp>;
+
 def X86vprot   : SDNode<"X86ISD::VPROT",
                         SDTypeProfile<1, 2, [SDTCisVec<0>, SDTCisSameAs<0,1>,
                                              SDTCisSameAs<0,2>]>>;
 def X86vproti  : SDNode<"X86ISD::VPROTI",
                         SDTypeProfile<1, 2, [SDTCisVec<0>, SDTCisSameAs<0,1>,
                                              SDTCisVT<2, i8>]>>;
 
 def X86vpshl   : SDNode<"X86ISD::VPSHL",

lib/Target/X86/X86IntrinsicsInfo.h

static const IntrinsicData  IntrinsicsWithoutChain[] = {
   X86_INTRINSIC_DATA(avx512_mask_pmulu_dq_512, INTR_TYPE_2OP_MASK,
                      X86ISD::PMULUDQ, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_d_128, INTR_TYPE_2OP_MASK, ISD::OR, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_d_256, INTR_TYPE_2OP_MASK, ISD::OR, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_d_512, INTR_TYPE_2OP_MASK, ISD::OR, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_q_128, INTR_TYPE_2OP_MASK, ISD::OR, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_q_256, INTR_TYPE_2OP_MASK, ISD::OR, 0),
   X86_INTRINSIC_DATA(avx512_mask_por_q_512, INTR_TYPE_2OP_MASK, ISD::OR, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_d_128,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_d_256,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_d_512,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_q_128,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_q_256,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
+  X86_INTRINSIC_DATA(avx512_mask_prol_q_512,  INTR_TYPE_2OP_MASK, X86ISD::VROTLI, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_d_128, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_d_256, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_d_512, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_q_128, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_q_256, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_prorv_q_512, INTR_TYPE_2OP_MASK, ISD::ROTR, 0),
   X86_INTRINSIC_DATA(avx512_mask_pshuf_b_128, INTR_TYPE_2OP_MASK,
                      X86ISD::PSHUFB, 0),

test/CodeGen/X86/avx512-intrinsics.ll

 ; CHECK-NEXT:    retq 
   %res = call <8 x i64> @llvm.x86.avx512.mask.prorv.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 %x3)
   %res1 = call <8 x i64> @llvm.x86.avx512.mask.prorv.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> zeroinitializer, i8 %x3)
   %res2 = call <8 x i64> @llvm.x86.avx512.mask.prorv.q.512(<8 x i64> %x0, <8 x i64> %x1, <8 x i64> %x2, i8 -1)
   %res3 = add <8 x i64> %res, %res1
   %res4 = add <8 x i64> %res3, %res2
   ret <8 x i64> %res4
 }
+
+declare <16 x i32> @llvm.x86.avx512.mask.prol.d.512(<16 x i32>, i8, <16 x i32>, i16)
+
+define <16 x i32>@test_int_x86_avx512_mask_prol_d_512(<16 x i32> %x0, i8 %x1, <16 x i32> %x2, i16 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_d_512:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovw %esi, %k1 
+; CHECK-NEXT:    vprold $3, %zmm0, %zmm1 {%k1} 
+; CHECK-NEXT:    vprold $3, %zmm0, %zmm2 {%k1} {z} 
+; CHECK-NEXT:    vprold $3, %zmm0, %zmm0 
+; CHECK-NEXT:    vpaddd %zmm2, %zmm1, %zmm1 
+; CHECK-NEXT:    vpaddd %zmm0, %zmm1, %zmm0 
+; CHECK-NEXT:    retq 
+  %res = call <16 x i32> @llvm.x86.avx512.mask.prol.d.512(<16 x i32> %x0, i8 3, <16 x i32> %x2, i16 %x3)
+  %res1 = call <16 x i32> @llvm.x86.avx512.mask.prol.d.512(<16 x i32> %x0, i8 3, <16 x i32> zeroinitializer, i16 %x3)
+  %res2 = call <16 x i32> @llvm.x86.avx512.mask.prol.d.512(<16 x i32> %x0, i8 3, <16 x i32> %x2, i16 -1)
+  %res3 = add <16 x i32> %res, %res1
+  %res4 = add <16 x i32> %res3, %res2
+  ret <16 x i32> %res4
+}
+
+declare <8 x i64> @llvm.x86.avx512.mask.prol.q.512(<8 x i64>, i8, <8 x i64>, i8)
+
+define <8 x i64>@test_int_x86_avx512_mask_prol_q_512(<8 x i64> %x0, i8 %x1, <8 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_q_512:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movzbl %sil, %eax 
+; CHECK-NEXT:    kmovw %eax, %k1 
+; CHECK-NEXT:    vprolq $3, %zmm0, %zmm1 {%k1} 
+; CHECK-NEXT:    vprolq $3, %zmm0, %zmm2 {%k1} {z} 
+; CHECK-NEXT:    vprolq $3, %zmm0, %zmm0 
+; CHECK-NEXT:    vpaddq %zmm2, %zmm1, %zmm1 
+; CHECK-NEXT:    vpaddq %zmm0, %zmm1, %zmm0 
+; CHECK-NEXT:    retq 
+  %res = call <8 x i64> @llvm.x86.avx512.mask.prol.q.512(<8 x i64> %x0, i8 3, <8 x i64> %x2, i8 %x3)
+  %res1 = call <8 x i64> @llvm.x86.avx512.mask.prol.q.512(<8 x i64> %x0, i8 3, <8 x i64> zeroinitializer, i8 %x3)
+  %res2 = call <8 x i64> @llvm.x86.avx512.mask.prol.q.512(<8 x i64> %x0, i8 3, <8 x i64> %x2, i8 -1)
+  %res3 = add <8 x i64> %res, %res1
+  %res4 = add <8 x i64> %res3, %res2
+  ret <8 x i64> %res4
+}
+
+

test/CodeGen/X86/avx512vl-intrinsics.ll

 ; CHECK-NEXT:    retq 
   %res = call <4 x i64> @llvm.x86.avx512.mask.prorv.q.256(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> %x2, i8 %x3)
   %res1 = call <4 x i64> @llvm.x86.avx512.mask.prorv.q.256(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> zeroinitializer, i8 %x3)
   %res2 = call <4 x i64> @llvm.x86.avx512.mask.prorv.q.256(<4 x i64> %x0, <4 x i64> %x1, <4 x i64> %x2, i8 -1)
   %res3 = add <4 x i64> %res, %res1
   %res4 = add <4 x i64> %res3, %res2
   ret <4 x i64> %res4
 }
+declare <4 x i32> @llvm.x86.avx512.mask.prol.d.128(<4 x i32>, i8, <4 x i32>, i8)
+
+define <4 x i32>@test_int_x86_avx512_mask_prol_d_128(<4 x i32> %x0, i8 %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_d_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movzbl %sil, %eax 
+; CHECK-NEXT:    kmovw %eax, %k1 
+; CHECK-NEXT:    vprold $3, %xmm0, %xmm1 {%k1} 
+; CHECK-NEXT:    vprold $3, %xmm0, %xmm2 {%k1} {z} 
+; CHECK-NEXT:    vprold $3, %xmm0, %xmm0 
+; CHECK-NEXT:    vpaddd %xmm2, %xmm1, %xmm1 
+; CHECK-NEXT:    vpaddd %xmm0, %xmm1, %xmm0 
+; CHECK-NEXT:    retq 
+  %res = call <4 x i32> @llvm.x86.avx512.mask.prol.d.128(<4 x i32> %x0, i8 3, <4 x i32> %x2, i8 %x3)
+  %res1 = call <4 x i32> @llvm.x86.avx512.mask.prol.d.128(<4 x i32> %x0, i8 3, <4 x i32> zeroinitializer, i8 %x3)
+  %res2 = call <4 x i32> @llvm.x86.avx512.mask.prol.d.128(<4 x i32> %x0, i8 3, <4 x i32> %x2, i8 -1)
+  %res3 = add <4 x i32> %res, %res1
+  %res4 = add <4 x i32> %res3, %res2
+  ret <4 x i32> %res4
+}
+
+declare <8 x i32> @llvm.x86.avx512.mask.prol.d.256(<8 x i32>, i8, <8 x i32>, i8)
+
+define <8 x i32>@test_int_x86_avx512_mask_prol_d_256(<8 x i32> %x0, i8 %x1, <8 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_d_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movzbl %sil, %eax 
+; CHECK-NEXT:    kmovw %eax, %k1 
+; CHECK-NEXT:    vprold $3, %ymm0, %ymm1 {%k1} 
+; CHECK-NEXT:    vprold $3, %ymm0, %ymm2 {%k1} {z} 
+; CHECK-NEXT:    vprold $3, %ymm0, %ymm0 
+; CHECK-NEXT:    vpaddd %ymm2, %ymm1, %ymm1 
+; CHECK-NEXT:    vpaddd %ymm0, %ymm1, %ymm0 
+; CHECK-NEXT:    retq 
+  %res = call <8 x i32> @llvm.x86.avx512.mask.prol.d.256(<8 x i32> %x0, i8 3, <8 x i32> %x2, i8 %x3)
+  %res1 = call <8 x i32> @llvm.x86.avx512.mask.prol.d.256(<8 x i32> %x0, i8 3, <8 x i32> zeroinitializer, i8 %x3)
+  %res2 = call <8 x i32> @llvm.x86.avx512.mask.prol.d.256(<8 x i32> %x0, i8 3, <8 x i32> %x2, i8 -1)
+  %res3 = add <8 x i32> %res, %res1
+  %res4 = add <8 x i32> %res3, %res2
+  ret <8 x i32> %res4
+}
+
+declare <2 x i64> @llvm.x86.avx512.mask.prol.q.128(<2 x i64>, i8, <2 x i64>, i8)
+
+define <2 x i64>@test_int_x86_avx512_mask_prol_q_128(<2 x i64> %x0, i8 %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_q_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movzbl %sil, %eax 
+; CHECK-NEXT:    kmovw %eax, %k1 
+; CHECK-NEXT:    vprolq $3, %xmm0, %xmm1 {%k1} 
+; CHECK-NEXT:    vprolq $3, %xmm0, %xmm2 {%k1} {z} 
+; CHECK-NEXT:    vprolq $3, %xmm0, %xmm0 
+; CHECK-NEXT:    vpaddq %xmm2, %xmm1, %xmm1 
+; CHECK-NEXT:    vpaddq %xmm0, %xmm1, %xmm0 
+; CHECK-NEXT:    retq 
+  %res = call <2 x i64> @llvm.x86.avx512.mask.prol.q.128(<2 x i64> %x0, i8 3, <2 x i64> %x2, i8 %x3)
+  %res1 = call <2 x i64> @llvm.x86.avx512.mask.prol.q.128(<2 x i64> %x0, i8 3, <2 x i64> zeroinitializer, i8 %x3)
+  %res2 = call <2 x i64> @llvm.x86.avx512.mask.prol.q.128(<2 x i64> %x0, i8 3, <2 x i64> %x2, i8 -1)
+  %res3 = add <2 x i64> %res, %res1
+  %res4 = add <2 x i64> %res3, %res2
+  ret <2 x i64> %res4
+}
+
+declare <4 x i64> @llvm.x86.avx512.mask.prol.q.256(<4 x i64>, i8, <4 x i64>, i8)
+
+define <4 x i64>@test_int_x86_avx512_mask_prol_q_256(<4 x i64> %x0, i8 %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_prol_q_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    movzbl %sil, %eax 
+; CHECK-NEXT:    kmovw %eax, %k1 
+; CHECK-NEXT:    vprolq $3, %ymm0, %ymm1 {%k1} 
+; CHECK-NEXT:    vprolq $3, %ymm0, %ymm2 {%k1} {z} 
+; CHECK-NEXT:    vprolq $3, %ymm0, %ymm0 
+; CHECK-NEXT:    vpaddq %ymm2, %ymm1, %ymm1 
+; CHECK-NEXT:    vpaddq %ymm0, %ymm1, %ymm0 
+; CHECK-NEXT:    retq 
+  %res = call <4 x i64> @llvm.x86.avx512.mask.prol.q.256(<4 x i64> %x0, i8 3, <4 x i64> %x2, i8 %x3)
+  %res1 = call <4 x i64> @llvm.x86.avx512.mask.prol.q.256(<4 x i64> %x0, i8 3, <4 x i64> zeroinitializer, i8 %x3)
+  %res2 = call <4 x i64> @llvm.x86.avx512.mask.prol.q.256(<4 x i64> %x0, i8 3, <4 x i64> %x2, i8 -1)
+  %res3 = add <4 x i64> %res, %res1
+  %res4 = add <4 x i64> %res3, %res2
+  ret <4 x i64> %res4
+}