FailedChanges

Summary

  1. AMDGPU] gfx1010 hazard recognizer Differential Revision: https://reviews.llvm.org/D61536
  2. [AMDGPU] gfx1010: use fmac instructions Differential Revision: https://reviews.llvm.org/D61527
Revision 359961 by rampitec:
AMDGPU] gfx1010 hazard recognizer

Differential Revision: https://reviews.llvm.org/D61536
Change TypePath in RepositoryPath in Workspace
The file was modified/llvm/trunk/lib/Target/AMDGPU/GCNHazardRecognizer.cpptrunk/lib/Target/AMDGPU/GCNHazardRecognizer.cpp
The file was modified/llvm/trunk/lib/Target/AMDGPU/GCNHazardRecognizer.htrunk/lib/Target/AMDGPU/GCNHazardRecognizer.h
The file was added/llvm/trunk/test/CodeGen/AMDGPU/lds-branch-vmem-hazard.mirtrunk/test/CodeGen/AMDGPU/lds-branch-vmem-hazard.mir
The file was added/llvm/trunk/test/CodeGen/AMDGPU/nsa-vmem-hazard.mirtrunk/test/CodeGen/AMDGPU/nsa-vmem-hazard.mir
The file was added/llvm/trunk/test/CodeGen/AMDGPU/smem-war-hazard.mirtrunk/test/CodeGen/AMDGPU/smem-war-hazard.mir
The file was added/llvm/trunk/test/CodeGen/AMDGPU/vcmpx-exec-war-hazard.mirtrunk/test/CodeGen/AMDGPU/vcmpx-exec-war-hazard.mir
The file was added/llvm/trunk/test/CodeGen/AMDGPU/vmem-to-salu-hazard.mirtrunk/test/CodeGen/AMDGPU/vmem-to-salu-hazard.mir
Revision 359959 by rampitec:
[AMDGPU] gfx1010: use fmac instructions

Differential Revision: https://reviews.llvm.org/D61527
Change TypePath in RepositoryPath in Workspace
The file was modified/llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpptrunk/lib/Target/AMDGPU/SIISelLowering.cpp
The file was modified/llvm/trunk/lib/Target/AMDGPU/SIInstrInfo.cpptrunk/lib/Target/AMDGPU/SIInstrInfo.cpp
The file was modified/llvm/trunk/lib/Target/AMDGPU/SIInstructions.tdtrunk/lib/Target/AMDGPU/SIInstructions.td
The file was modified/llvm/trunk/lib/Target/AMDGPU/SIPeepholeSDWA.cpptrunk/lib/Target/AMDGPU/SIPeepholeSDWA.cpp
The file was modified/llvm/trunk/test/CodeGen/AMDGPU/fdiv.f16.lltrunk/test/CodeGen/AMDGPU/fdiv.f16.ll
The file was added/llvm/trunk/test/CodeGen/AMDGPU/fmac.sdwa.lltrunk/test/CodeGen/AMDGPU/fmac.sdwa.ll
The file was modified/llvm/trunk/test/CodeGen/AMDGPU/fmuladd.f16.lltrunk/test/CodeGen/AMDGPU/fmuladd.f16.ll
The file was modified/llvm/trunk/test/CodeGen/AMDGPU/llvm.fmuladd.f16.lltrunk/test/CodeGen/AMDGPU/llvm.fmuladd.f16.ll
The file was modified/llvm/trunk/test/CodeGen/AMDGPU/madak.lltrunk/test/CodeGen/AMDGPU/madak.ll
The file was added/llvm/trunk/test/CodeGen/AMDGPU/sdwa-peephole-instr-gfx10.mirtrunk/test/CodeGen/AMDGPU/sdwa-peephole-instr-gfx10.mir
The file was added/llvm/trunk/test/CodeGen/AMDGPU/twoaddr-fma.mirtrunk/test/CodeGen/AMDGPU/twoaddr-fma.mir

Summary

  1. Reduce amount of work ODR hashing does. When a FunctionProtoType is in the original type in a DecayedType, the decayed type is a PointerType which points back the original FunctionProtoType. The visitor for ODRHashing will attempt to process both Type's, doing double work. By chaining together multiple DecayedType's and FunctionProtoType's, this would result in 2^N Type's visited only N DecayedType's and N FunctionProtoType's exsit. Another bug where VisitDecayedType and VisitAdjustedType did redundant work doubled the work at each level, giving 4^N Type's visited. This patch removed the double work and detects when a FunctionProtoType decays to itself to only check the Type once. This lowers the exponential runtime to linear runtime. Fixes https://bugs.llvm.org/show_bug.cgi?id=41625
  2. Disallow the operand of __builtin_constant_p from modifying enclosing state when it's encountered while evaluating a constexpr function. We attempt to follow GCC trunk's behavior here, but it is somewhat inscrutible, so our behavior is only approximately the same for now. Specifically, we only permit modification of objects whose lifetime began within the operand of the __builtin_constant_p. GCC appears to have effectively the same restriction, but also some unknown restriction based on where and how the local state of the constexpr function is mentioned within the operand (see added testcases).
  3. [Driver] Create non-existent directory for -fcrash-diagnostics-dir When user specifies non-existent directory to -fcrash-diagnostics-dir, create it rather than failing with an error as would be the case before. Differential Revision: https://reviews.llvm.org/D61542
  4. [cxx_status] Don't list -fmodules / -fmodules-ts against C++ modules support; those turn on different modules modes. The real C++ modules support is behind -std=c++2a like the rest of C++20.
  5. [cxx_status] Replace "SVN" entries with Clang 8 as appropriate. Also: use the "svn" color for "explicit(bool)" rather than the "full" color.
Revision 359960 by rtrieu:
Reduce amount of work ODR hashing does.

When a FunctionProtoType is in the original type in a DecayedType, the decayed
type is a PointerType which points back the original FunctionProtoType.  The
visitor for ODRHashing will attempt to process both Type's, doing double work.
By chaining together multiple DecayedType's and FunctionProtoType's, this would
result in 2^N Type's visited only N DecayedType's and N FunctionProtoType's
exsit.  Another bug where VisitDecayedType and VisitAdjustedType did
redundant work doubled the work at each level, giving 4^N Type's visited.  This
patch removed the double work and detects when a FunctionProtoType decays to
itself to only check the Type once.  This lowers the exponential runtime to
linear runtime.  Fixes https://bugs.llvm.org/show_bug.cgi?id=41625
Change TypePath in RepositoryPath in Workspace
The file was modified/cfe/trunk/lib/AST/ODRHash.cpptrunk/lib/AST/ODRHash.cpp
The file was modified/cfe/trunk/test/Modules/odr_hash.cpptrunk/test/Modules/odr_hash.cpp
Revision 359958 by rsmith:
Disallow the operand of __builtin_constant_p from modifying enclosing
state when it's encountered while evaluating a constexpr function.

We attempt to follow GCC trunk's behavior here, but it is somewhat
inscrutible, so our behavior is only approximately the same for now.
Specifically, we only permit modification of objects whose lifetime
began within the operand of the __builtin_constant_p. GCC appears to
have effectively the same restriction, but also some unknown restriction
based on where and how the local state of the constexpr function is
mentioned within the operand (see added testcases).
Change TypePath in RepositoryPath in Workspace
The file was modified/cfe/trunk/lib/AST/ExprConstant.cpptrunk/lib/AST/ExprConstant.cpp
The file was modified/cfe/trunk/test/SemaCXX/builtin-constant-p.cpptrunk/test/SemaCXX/builtin-constant-p.cpp
Revision 359954 by phosek:
[Driver] Create non-existent directory for -fcrash-diagnostics-dir

When user specifies non-existent directory to -fcrash-diagnostics-dir,
create it rather than failing with an error as would be the case before.

Differential Revision: https://reviews.llvm.org/D61542
Change TypePath in RepositoryPath in Workspace
The file was modified/cfe/trunk/lib/Driver/Driver.cpptrunk/lib/Driver/Driver.cpp
The file was modified/cfe/trunk/test/Driver/crash-diagnostics-dir.ctrunk/test/Driver/crash-diagnostics-dir.c
Revision 359953 by rsmith:
[cxx_status] Don't list -fmodules / -fmodules-ts against C++ modules
support; those turn on different modules modes. The real C++ modules
support is behind -std=c++2a like the rest of C++20.
Change TypePath in RepositoryPath in Workspace
The file was modified/cfe/trunk/www/cxx_status.htmltrunk/www/cxx_status.html
Revision 359951 by rsmith:
[cxx_status] Replace "SVN" entries with Clang 8 as appropriate.

Also: use the "svn" color for "explicit(bool)" rather than the "full" color.
Change TypePath in RepositoryPath in Workspace
The file was modified/cfe/trunk/www/cxx_status.htmltrunk/www/cxx_status.html