Merge branch 'main' into hjagasiaAMD-offload-tunnel-cmake

Author: hjagasiaAMD

clang/lib/CIR/CodeGen/CIRGenRecordLayoutBuilder.cpp

@@ -615,20 +615,20 @@ void CIRRecordLowering::determinePacked(bool nvBaseType) {
       continue;
     // If any member falls at an offset that it not a multiple of its alignment,
     // then the entire record must be packed.
-    if (member.offset % getAlignment(member.data))
+    if (!member.offset.isMultipleOf(getAlignment(member.data)))
       packed = true;
     if (member.offset < nvSize)
       nvAlignment = std::max(nvAlignment, getAlignment(member.data));
     alignment = std::max(alignment, getAlignment(member.data));
   }
   // If the size of the record (the capstone's offset) is not a multiple of the
   // record's alignment, it must be packed.
-  if (members.back().offset % alignment)
+  if (!members.back().offset.isMultipleOf(alignment))
     packed = true;
   // If the non-virtual sub-object is not a multiple of the non-virtual
   // sub-object's alignment, it must be packed.  We cannot have a packed
   // non-virtual sub-object and an unpacked complete object or vise versa.
-  if (nvSize % nvAlignment)
+  if (!nvSize.isMultipleOf(nvAlignment))
     packed = true;
   // Update the alignment of the sentinel.
   if (!packed)
@@ -824,7 +824,7 @@ void CIRRecordLowering::lowerUnion() {
     appendPaddingBytes(layoutSize - getSize(storageType));
 
   // Set packed if we need it.
-  if (layoutSize % getAlignment(storageType))
+  if (!layoutSize.isMultipleOf(getAlignment(storageType)))
     packed = true;
 }
 

clang/test/CodeGenHLSL/RootSignature.hlsl

@@ -82,8 +82,8 @@ void RootDescriptorsEntry() {}
 // checking minLOD, maxLOD
 // CHECK-SAME: float -1.280000e+02, float 1.280000e+02,
 
-// checking register, space and visibility
-// CHECK-SAME: i32 42, i32 0, i32 0}
+// checking register, space, visibility and flags
+// CHECK-SAME: i32 42, i32 0, i32 0, i32 0}
 
 #define SampleStaticSampler \
   "StaticSampler(s42, " \

llvm/include/llvm/Frontend/HLSL/HLSLRootSignature.h

@@ -131,6 +131,7 @@ struct StaticSampler {
   float MaxLOD = std::numeric_limits<float>::max();
   uint32_t Space = 0;
   dxbc::ShaderVisibility Visibility = dxbc::ShaderVisibility::All;
+  dxbc::StaticSamplerFlags Flags = dxbc::StaticSamplerFlags::None;
 };
 
 /// Models RootElement : RootFlags | RootConstants | RootParam

llvm/lib/Frontend/HLSL/RootSignatureMetadata.cpp

@@ -218,6 +218,7 @@ MDNode *MetadataBuilder::BuildStaticSampler(const StaticSampler &Sampler) {
       ConstantAsMetadata::get(Builder.getInt32(Sampler.Space)),
       ConstantAsMetadata::get(
           Builder.getInt32(to_underlying(Sampler.Visibility))),
+      ConstantAsMetadata::get(Builder.getInt32(to_underlying(Sampler.Flags))),
   };
   return MDNode::get(Ctx, Operands);
 }
@@ -417,7 +418,7 @@ Error MetadataParser::parseDescriptorTable(mcdxbc::RootSignatureDesc &RSD,
 
 Error MetadataParser::parseStaticSampler(mcdxbc::RootSignatureDesc &RSD,
                                          MDNode *StaticSamplerNode) {
-  if (StaticSamplerNode->getNumOperands() != 14)
+  if (StaticSamplerNode->getNumOperands() != 15)
     return make_error<InvalidRSMetadataFormat>("Static Sampler");
 
   mcdxbc::StaticSampler Sampler;
@@ -501,6 +502,17 @@ Error MetadataParser::parseStaticSampler(mcdxbc::RootSignatureDesc &RSD,
     return Error(std::move(E));
   Sampler.ShaderVisibility = *Visibility;
 
+  if (RSD.Version < 3) {
+    RSD.StaticSamplers.push_back(Sampler);
+    return Error::success();
+  }
+  assert(RSD.Version >= 3);
+
+  if (std::optional<uint32_t> Val = extractMdIntValue(StaticSamplerNode, 14))
+    Sampler.Flags = *Val;
+  else
+    return make_error<InvalidRSMetadataValue>("Static Sampler Flags");
+
   RSD.StaticSamplers.push_back(Sampler);
   return Error::success();
 }

llvm/lib/Support/ScopedPrinter.cpp

@@ -1,12 +1,17 @@
-#include "llvm/Support/ScopedPrinter.h"
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
 
+#include "llvm/Support/ScopedPrinter.h"
 #include "llvm/Support/Format.h"
 
-using namespace llvm::support;
+using namespace llvm;
 
-namespace llvm {
-
-raw_ostream &operator<<(raw_ostream &OS, const HexNumber &Value) {
+raw_ostream &llvm::operator<<(raw_ostream &OS, const HexNumber &Value) {
   OS << "0x" << utohexstr(Value.Value);
   return OS;
 }
@@ -45,5 +50,3 @@ JSONScopedPrinter::JSONScopedPrinter(
   if (this->OuterScope)
     this->OuterScope->setPrinter(*this);
 }
-
-} // namespace llvm

llvm/lib/Target/RISCV/RISCVTargetTransformInfo.cpp

@@ -3139,8 +3139,8 @@ bool RISCVTTIImpl::isProfitableToSinkOperands(
     bool IsVPSplat = match(Op, m_Intrinsic<Intrinsic::experimental_vp_splat>(
                                    m_Value(), m_Value(), m_Value()));
     if (!IsVPSplat &&
-        !match(Op, m_Shuffle(m_InsertElt(m_Undef(), m_Value(), m_ZeroInt()),
-                             m_Undef(), m_ZeroMask())))
+        !match(Op, m_Shuffle(m_InsertElt(m_Value(), m_Value(), m_ZeroInt()),
+                             m_Value(), m_ZeroMask())))
       continue;
 
     // Don't sink i1 splats.

llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

@@ -2241,10 +2241,9 @@ class BoUpSLP {
   /// TODO: If load combining is allowed in the IR optimizer, this analysis
   ///       may not be necessary.
   bool isLoadCombineCandidate(ArrayRef<Value *> Stores) const;
-  bool isStridedLoad(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
-                     ArrayRef<unsigned> Order, const TargetTransformInfo &TTI,
-                     const DataLayout &DL, ScalarEvolution &SE,
-                     const int64_t Diff, StridedPtrInfo &SPtrInfo) const;
+  bool isStridedLoad(ArrayRef<Value *> PointerOps, Type *ScalarTy,
+                     Align Alignment, const int64_t Diff, Value *Ptr0,
+                     Value *PtrN, StridedPtrInfo &SPtrInfo) const;
 
   /// Checks if the given array of loads can be represented as a vectorized,
   /// scatter or just simple gather.
@@ -6824,13 +6823,10 @@ isMaskedLoadCompress(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
 /// 4. Any pointer operand is an instruction with the users outside of the
 /// current graph (for masked gathers extra extractelement instructions
 /// might be required).
-bool BoUpSLP::isStridedLoad(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
-                            ArrayRef<unsigned> Order,
-                            const TargetTransformInfo &TTI,
-                            const DataLayout &DL, ScalarEvolution &SE,
-                            const int64_t Diff,
-                            StridedPtrInfo &SPtrInfo) const {
-  const size_t Sz = VL.size();
+bool BoUpSLP::isStridedLoad(ArrayRef<Value *> PointerOps, Type *ScalarTy,
+                            Align Alignment, const int64_t Diff, Value *Ptr0,
+                            Value *PtrN, StridedPtrInfo &SPtrInfo) const {
+  const size_t Sz = PointerOps.size();
   if (Diff % (Sz - 1) != 0)
     return false;
 
@@ -6842,7 +6838,6 @@ bool BoUpSLP::isStridedLoad(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
   });
 
   const uint64_t AbsoluteDiff = std::abs(Diff);
-  Type *ScalarTy = VL.front()->getType();
   auto *VecTy = getWidenedType(ScalarTy, Sz);
   if (IsAnyPointerUsedOutGraph ||
       (AbsoluteDiff > Sz &&
@@ -6853,20 +6848,9 @@ bool BoUpSLP::isStridedLoad(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
     int64_t Stride = Diff / static_cast<int64_t>(Sz - 1);
     if (Diff != Stride * static_cast<int64_t>(Sz - 1))
       return false;
-    Align Alignment =
-        cast<LoadInst>(Order.empty() ? VL.front() : VL[Order.front()])
-            ->getAlign();
-    if (!TTI.isLegalStridedLoadStore(VecTy, Alignment))
+    if (!TTI->isLegalStridedLoadStore(VecTy, Alignment))
       return false;
-    Value *Ptr0;
-    Value *PtrN;
-    if (Order.empty()) {
-      Ptr0 = PointerOps.front();
-      PtrN = PointerOps.back();
-    } else {
-      Ptr0 = PointerOps[Order.front()];
-      PtrN = PointerOps[Order.back()];
-    }
+
     // Iterate through all pointers and check if all distances are
     // unique multiple of Dist.
     SmallSet<int64_t, 4> Dists;
@@ -6875,14 +6859,14 @@ bool BoUpSLP::isStridedLoad(ArrayRef<Value *> VL, ArrayRef<Value *> PointerOps,
       if (Ptr == PtrN)
         Dist = Diff;
       else if (Ptr != Ptr0)
-        Dist = *getPointersDiff(ScalarTy, Ptr0, ScalarTy, Ptr, DL, SE);
+        Dist = *getPointersDiff(ScalarTy, Ptr0, ScalarTy, Ptr, *DL, *SE);
       // If the strides are not the same or repeated, we can't
       // vectorize.
       if (((Dist / Stride) * Stride) != Dist || !Dists.insert(Dist).second)
         break;
     }
     if (Dists.size() == Sz) {
-      Type *StrideTy = DL.getIndexType(Ptr0->getType());
+      Type *StrideTy = DL->getIndexType(Ptr0->getType());
       SPtrInfo.StrideVal = ConstantInt::get(StrideTy, Stride);
       SPtrInfo.Ty = getWidenedType(ScalarTy, Sz);
       return true;
@@ -6971,7 +6955,11 @@ BoUpSLP::LoadsState BoUpSLP::canVectorizeLoads(
                                    cast<Instruction>(V), UserIgnoreList);
                              }))
       return LoadsState::CompressVectorize;
-    if (isStridedLoad(VL, PointerOps, Order, *TTI, *DL, *SE, *Diff, SPtrInfo))
+    Align Alignment =
+        cast<LoadInst>(Order.empty() ? VL.front() : VL[Order.front()])
+            ->getAlign();
+    if (isStridedLoad(PointerOps, ScalarTy, Alignment, *Diff, Ptr0, PtrN,
+                      SPtrInfo))
       return LoadsState::StridedVectorize;
   }
   if (!TTI->isLegalMaskedGather(VecTy, CommonAlignment) ||

llvm/test/CMakeLists.txt

@@ -71,7 +71,6 @@ set(LLVM_TEST_DEPENDS
   ${LLVM_TEST_DEPENDS_COMMON}
   BugpointPasses
   LLVMWindowsDriver
-  UnitTests
   bugpoint
   llc
   lli
@@ -270,10 +269,11 @@ add_lit_testsuites(LLVM ${CMAKE_CURRENT_SOURCE_DIR}
   ${exclude_from_check_all}
   DEPENDS ${LLVM_TEST_DEPENDS}
   FOLDER "Tests/Subdirectories"
-  SKIP "^FileCheck" "^TableGen"
+  SKIP "^FileCheck" "^TableGen" "^Unit"
   )
 add_subdirectory(FileCheck)
 add_subdirectory(TableGen)
+add_subdirectory(Unit)
 
 # Setup an alias for 'check-all'.
 add_custom_target(check)

llvm/test/CodeGen/DirectX/ContainerData/RootSignature-StaticSamplers-Invalid-AddressU.ll

@@ -16,4 +16,4 @@ attributes #0 = { "hlsl.numthreads"="1,1,1" "hlsl.shader"="compute" }
 !dx.rootsignatures = !{!2} ; list of function/root signature pairs
 !2 = !{ ptr @main, !3, i32 2 } ; function, root signature
 !3 = !{ !5 } ; list of root signature elements
-!5 = !{ !"StaticSampler", i32 4, i32 666, i32 3, i32 5, float 0x3FF6CCCCC0000000, i32 9, i32 3, i32 2, float -1.280000e+02, float 1.280000e+02, i32 42, i32 0, i32 0 }
+!5 = !{ !"StaticSampler", i32 4, i32 666, i32 3, i32 5, float 0x3FF6CCCCC0000000, i32 9, i32 3, i32 2, float -1.280000e+02, float 1.280000e+02, i32 42, i32 0, i32 0, i32 0 }

llvm/test/CodeGen/DirectX/ContainerData/RootSignature-StaticSamplers-Invalid-AddressV.ll

@@ -16,4 +16,4 @@ attributes #0 = { "hlsl.numthreads"="1,1,1" "hlsl.shader"="compute" }
 !dx.rootsignatures = !{!2} ; list of function/root signature pairs
 !2 = !{ ptr @main, !3, i32 2 } ; function, root signature
 !3 = !{ !5 } ; list of root signature elements
-!5 = !{ !"StaticSampler", i32 4, i32 2, i32 666, i32 5, float 0x3FF6CCCCC0000000, i32 9, i32 3, i32 2, float -1.280000e+02, float 1.280000e+02, i32 42, i32 0, i32 0 }
+!5 = !{ !"StaticSampler", i32 4, i32 2, i32 666, i32 5, float 0x3FF6CCCCC0000000, i32 9, i32 3, i32 2, float -1.280000e+02, float 1.280000e+02, i32 42, i32 0, i32 0, i32 0 }