LLVM 4.0.0 Final Release

This commit is contained in:
Tom Stellard 2017-03-23 14:27:53 +00:00
parent 05920e6f29
commit 1bbe5adca1
27 changed files with 203 additions and 4139 deletions

View File

@ -0,0 +1,193 @@
From a61fc423f3c043314efd4c0cdb1367de2077ac36 Mon Sep 17 00:00:00 2001
From: Eric Fiselier <eric@efcs.ca>
Date: Fri, 10 Feb 2017 01:59:20 +0000
Subject: [PATCH] [CMake] Fix pthread handling for out-of-tree builds
LLVM defines `PTHREAD_LIB` which is used by AddLLVM.cmake and various projects
to correctly link the threading library when needed. Unfortunately
`PTHREAD_LIB` is defined by LLVM's `config-ix.cmake` file which isn't installed
and therefore can't be used when configuring out-of-tree builds. This causes
such builds to fail since `pthread` isn't being correctly linked.
This patch attempts to fix that problem by renaming and exporting
`LLVM_PTHREAD_LIB` as part of`LLVMConfig.cmake`. I renamed `PTHREAD_LIB`
because It seemed likely to cause collisions with downstream users of
`LLVMConfig.cmake`.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294690 91177308-0d34-0410-b5e6-96231b3b80d8
---
cmake/config-ix.cmake | 2 +-
cmake/modules/AddLLVM.cmake | 6 +++---
cmake/modules/LLVMConfig.cmake.in | 4 ++++
examples/ParallelJIT/CMakeLists.txt | 2 +-
lib/CodeGen/CMakeLists.txt | 2 +-
lib/ExecutionEngine/IntelJITEvents/CMakeLists.txt | 2 +-
lib/Fuzzer/CMakeLists.txt | 4 ++--
lib/Support/CMakeLists.txt | 2 +-
unittests/ExecutionEngine/Orc/CMakeLists.txt | 2 +-
unittests/Support/CMakeLists.txt | 2 +-
utils/unittest/CMakeLists.txt | 4 ++--
11 files changed, 18 insertions(+), 14 deletions(-)
diff --git a/cmake/config-ix.cmake b/cmake/config-ix.cmake
index 50bcc50..6bd2b53 100755
--- a/cmake/config-ix.cmake
+++ b/cmake/config-ix.cmake
@@ -115,7 +115,7 @@ if(HAVE_LIBPTHREAD)
set(CMAKE_THREAD_PREFER_PTHREAD TRUE)
set(THREADS_HAVE_PTHREAD_ARG Off)
find_package(Threads REQUIRED)
- set(PTHREAD_LIB ${CMAKE_THREAD_LIBS_INIT})
+ set(LLVM_PTHREAD_LIB ${CMAKE_THREAD_LIBS_INIT})
endif()
# Don't look for these libraries on Windows. Also don't look for them if we're
diff --git a/cmake/modules/AddLLVM.cmake b/cmake/modules/AddLLVM.cmake
index b3c7746..cb4171c 100644
--- a/cmake/modules/AddLLVM.cmake
+++ b/cmake/modules/AddLLVM.cmake
@@ -718,11 +718,11 @@ macro(add_llvm_executable name)
if(NOT ARG_IGNORE_EXTERNALIZE_DEBUGINFO)
llvm_externalize_debuginfo(${name})
endif()
- if (PTHREAD_LIB)
+ if (LLVM_PTHREAD_LIB)
# libpthreads overrides some standard library symbols, so main
# executable must be linked with it in order to provide consistent
# API for all shared libaries loaded by this executable.
- target_link_libraries(${name} ${PTHREAD_LIB})
+ target_link_libraries(${name} ${LLVM_PTHREAD_LIB})
endif()
endmacro(add_llvm_executable name)
@@ -1027,7 +1027,7 @@ function(add_unittest test_suite test_name)
# libpthreads overrides some standard library symbols, so main
# executable must be linked with it in order to provide consistent
# API for all shared libaries loaded by this executable.
- target_link_libraries(${test_name} gtest_main gtest ${PTHREAD_LIB})
+ target_link_libraries(${test_name} gtest_main gtest ${LLVM_PTHREAD_LIB})
add_dependencies(${test_suite} ${test_name})
get_target_property(test_suite_folder ${test_suite} FOLDER)
diff --git a/cmake/modules/LLVMConfig.cmake.in b/cmake/modules/LLVMConfig.cmake.in
index 2aea2dc..7a8eb36 100644
--- a/cmake/modules/LLVMConfig.cmake.in
+++ b/cmake/modules/LLVMConfig.cmake.in
@@ -45,6 +45,10 @@ set(LLVM_ENABLE_PIC @LLVM_ENABLE_PIC@)
set(LLVM_BUILD_32_BITS @LLVM_BUILD_32_BITS@)
+if (NOT "@LLVM_PTHREAD_LIB@" STREQUAL "")
+ set(LLVM_PTHREAD_LIB "@LLVM_PTHREAD_LIB@")
+endif()
+
set(LLVM_ENABLE_PLUGINS @LLVM_ENABLE_PLUGINS@)
set(LLVM_EXPORT_SYMBOLS_FOR_PLUGINS @LLVM_EXPORT_SYMBOLS_FOR_PLUGINS@)
set(LLVM_PLUGIN_EXT @LLVM_PLUGIN_EXT@)
diff --git a/examples/ParallelJIT/CMakeLists.txt b/examples/ParallelJIT/CMakeLists.txt
index e85b470..deeee07 100644
--- a/examples/ParallelJIT/CMakeLists.txt
+++ b/examples/ParallelJIT/CMakeLists.txt
@@ -11,4 +11,4 @@ add_llvm_example(ParallelJIT
ParallelJIT.cpp
)
-target_link_libraries(ParallelJIT ${PTHREAD_LIB})
+target_link_libraries(ParallelJIT ${LLVM_PTHREAD_LIB})
diff --git a/lib/CodeGen/CMakeLists.txt b/lib/CodeGen/CMakeLists.txt
index a1e5fd4..a9a3d85 100644
--- a/lib/CodeGen/CMakeLists.txt
+++ b/lib/CodeGen/CMakeLists.txt
@@ -150,7 +150,7 @@ add_llvm_library(LLVMCodeGen
${LLVM_MAIN_INCLUDE_DIR}/llvm/CodeGen
${LLVM_MAIN_INCLUDE_DIR}/llvm/CodeGen/PBQP
- LINK_LIBS ${PTHREAD_LIB}
+ LINK_LIBS ${LLVM_PTHREAD_LIB}
DEPENDS
intrinsics_gen
diff --git a/lib/ExecutionEngine/IntelJITEvents/CMakeLists.txt b/lib/ExecutionEngine/IntelJITEvents/CMakeLists.txt
index 3b8c4b9..e6c33b2 100644
--- a/lib/ExecutionEngine/IntelJITEvents/CMakeLists.txt
+++ b/lib/ExecutionEngine/IntelJITEvents/CMakeLists.txt
@@ -4,7 +4,7 @@ if( HAVE_LIBDL )
set(LLVM_INTEL_JIT_LIBS ${CMAKE_DL_LIBS})
endif()
-set(LLVM_INTEL_JIT_LIBS ${PTHREAD_LIB} ${LLVM_INTEL_JIT_LIBS})
+set(LLVM_INTEL_JIT_LIBS ${LLVM_PTHREAD_LIB} ${LLVM_INTEL_JIT_LIBS})
add_llvm_library(LLVMIntelJITEvents
diff --git a/lib/Fuzzer/CMakeLists.txt b/lib/Fuzzer/CMakeLists.txt
index 5ba126e..f490b36 100644
--- a/lib/Fuzzer/CMakeLists.txt
+++ b/lib/Fuzzer/CMakeLists.txt
@@ -34,12 +34,12 @@ if( LLVM_USE_SANITIZE_COVERAGE )
add_library(LLVMFuzzerNoMain STATIC
$<TARGET_OBJECTS:LLVMFuzzerNoMainObjects>
)
- target_link_libraries(LLVMFuzzerNoMain ${PTHREAD_LIB})
+ target_link_libraries(LLVMFuzzerNoMain ${LLVM_PTHREAD_LIB})
add_library(LLVMFuzzer STATIC
FuzzerMain.cpp
$<TARGET_OBJECTS:LLVMFuzzerNoMainObjects>
)
- target_link_libraries(LLVMFuzzer ${PTHREAD_LIB})
+ target_link_libraries(LLVMFuzzer ${LLVM_PTHREAD_LIB})
if( LLVM_INCLUDE_TESTS )
add_subdirectory(test)
diff --git a/lib/Support/CMakeLists.txt b/lib/Support/CMakeLists.txt
index 3301364..f7cfa76 100644
--- a/lib/Support/CMakeLists.txt
+++ b/lib/Support/CMakeLists.txt
@@ -17,7 +17,7 @@ elseif( CMAKE_HOST_UNIX )
if( LLVM_ENABLE_THREADS AND HAVE_LIBATOMIC )
set(system_libs ${system_libs} atomic)
endif()
- set(system_libs ${system_libs} ${PTHREAD_LIB})
+ set(system_libs ${system_libs} ${LLVM_PTHREAD_LIB})
if ( LLVM_ENABLE_ZLIB AND HAVE_LIBZ )
set(system_libs ${system_libs} z)
endif()
diff --git a/unittests/ExecutionEngine/Orc/CMakeLists.txt b/unittests/ExecutionEngine/Orc/CMakeLists.txt
index 68f6d0c..80c344e 100644
--- a/unittests/ExecutionEngine/Orc/CMakeLists.txt
+++ b/unittests/ExecutionEngine/Orc/CMakeLists.txt
@@ -21,4 +21,4 @@ add_llvm_unittest(OrcJITTests
RPCUtilsTest.cpp
)
-target_link_libraries(OrcJITTests ${PTHREAD_LIB})
+target_link_libraries(OrcJITTests ${LLVM_PTHREAD_LIB})
diff --git a/unittests/Support/CMakeLists.txt b/unittests/Support/CMakeLists.txt
index 4c9bb5e..ea26079 100644
--- a/unittests/Support/CMakeLists.txt
+++ b/unittests/Support/CMakeLists.txt
@@ -64,4 +64,4 @@ add_llvm_unittest(SupportTests
)
# ManagedStatic.cpp uses <pthread>.
-target_link_libraries(SupportTests ${PTHREAD_LIB})
+target_link_libraries(SupportTests ${LLVM_PTHREAD_LIB})
diff --git a/utils/unittest/CMakeLists.txt b/utils/unittest/CMakeLists.txt
index a50733a..b42ac83 100644
--- a/utils/unittest/CMakeLists.txt
+++ b/utils/unittest/CMakeLists.txt
@@ -40,8 +40,8 @@ if (NOT LLVM_ENABLE_THREADS)
add_definitions( -DGTEST_HAS_PTHREAD=0 )
endif()
-find_library(PTHREAD_LIBRARY_PATH pthread)
-if (PTHREAD_LIBRARY_PATH)
+find_library(LLVM_PTHREAD_LIBRARY_PATH pthread)
+if (LLVM_PTHREAD_LIBRARY_PATH)
list(APPEND LIBS pthread)
endif()
--
1.8.3.1

View File

@ -1,435 +0,0 @@
From 29cf3bd00fe84ddab138c9311fe288bb9da8a273 Mon Sep 17 00:00:00 2001
From: root <root@mammon-seattle-raw.austin.arm.com>
Date: Thu, 9 Mar 2017 12:22:48 -0600
Subject: [PATCH] Fix R_AARCH64_MOVW_UABS_G3 relocation
Summary: The relocation is missing mask so an address that
has non-zero bits in 47:43 may overwrite the register
number. (Frequently shows up as target register changed
to xzr....)
Reviewers: t.p.northover, lhames
Subscribers: davide, aemerson, rengolin, llvm-commits
Differential Revision: https://reviews.llvm.org/D27609
---
llvm-3.9.1.src/include/llvm/Object/ELFObjectFile.h | 2 +-
llvm-3.9.1.src/include/llvm/Object/RelocVisitor.h | 1 +
.../ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp | 67 +++++++++-----
.../RuntimeDyld/AArch64/ELF_ARM64_BE-relocations.s | 102 +++++++++++++++++++++
.../RuntimeDyld/AArch64/ELF_ARM64_relocations.s | 99 ++++++++++++++++++++
5 files changed, 249 insertions(+), 22 deletions(-)
create mode 100644 llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_BE-relocations.s
create mode 100644 llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_relocations.s
diff --git a/llvm-3.9.1.src/include/llvm/Object/ELFObjectFile.h b/llvm-3.9.1.src/include/llvm/Object/ELFObjectFile.h
index 07c6364..d3b83f9 100644
--- a/llvm-3.9.1.src/include/llvm/Object/ELFObjectFile.h
+++ b/llvm-3.9.1.src/include/llvm/Object/ELFObjectFile.h
@@ -907,7 +907,7 @@ unsigned ELFObjectFile<ELFT>::getArch() const {
case ELF::EM_X86_64:
return Triple::x86_64;
case ELF::EM_AARCH64:
- return Triple::aarch64;
+ return IsLittleEndian ? Triple::aarch64 : Triple::aarch64_be;
case ELF::EM_ARM:
return Triple::arm;
case ELF::EM_AVR:
diff --git a/llvm-3.9.1.src/include/llvm/Object/RelocVisitor.h b/llvm-3.9.1.src/include/llvm/Object/RelocVisitor.h
index 5e0df98..b59e8ec 100644
--- a/llvm-3.9.1.src/include/llvm/Object/RelocVisitor.h
+++ b/llvm-3.9.1.src/include/llvm/Object/RelocVisitor.h
@@ -86,6 +86,7 @@ private:
return RelocToApply();
}
case Triple::aarch64:
+ case Triple::aarch64_be:
switch (RelocType) {
case llvm::ELF::R_AARCH64_ABS32:
return visitELF_AARCH64_ABS32(R, Value);
diff --git a/llvm-3.9.1.src/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp b/llvm-3.9.1.src/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp
index 9cbdb13..9e04b5d 100644
--- a/llvm-3.9.1.src/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp
+++ b/llvm-3.9.1.src/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp
@@ -309,6 +309,8 @@ void RuntimeDyldELF::resolveAArch64Relocation(const SectionEntry &Section,
uint32_t *TargetPtr =
reinterpret_cast<uint32_t *>(Section.getAddressWithOffset(Offset));
uint64_t FinalAddress = Section.getLoadAddressWithOffset(Offset);
+ // Data should use target endian. Code should always use little endian.
+ bool isBE = Arch == Triple::aarch64_be;
DEBUG(dbgs() << "resolveAArch64Relocation, LocalAddress: 0x"
<< format("%llx", Section.getAddressWithOffset(Offset))
@@ -324,14 +326,22 @@ void RuntimeDyldELF::resolveAArch64Relocation(const SectionEntry &Section,
case ELF::R_AARCH64_ABS64: {
uint64_t *TargetPtr =
reinterpret_cast<uint64_t *>(Section.getAddressWithOffset(Offset));
- *TargetPtr = Value + Addend;
+ if (isBE)
+ support::ubig64_t::ref{TargetPtr} = Value + Addend;
+ else
+ support::ulittle64_t::ref{TargetPtr} = Value + Addend;
break;
}
case ELF::R_AARCH64_PREL32: {
uint64_t Result = Value + Addend - FinalAddress;
assert(static_cast<int64_t>(Result) >= INT32_MIN &&
static_cast<int64_t>(Result) <= UINT32_MAX);
- *TargetPtr = static_cast<uint32_t>(Result & 0xffffffffU);
+ if (isBE)
+ support::ubig32_t::ref{TargetPtr} =
+ static_cast<uint32_t>(Result & 0xffffffffU);
+ else
+ support::ulittle32_t::ref{TargetPtr} =
+ static_cast<uint32_t>(Result & 0xffffffffU);
break;
}
case ELF::R_AARCH64_CALL26: // fallthrough
@@ -339,6 +349,7 @@ void RuntimeDyldELF::resolveAArch64Relocation(const SectionEntry &Section,
// Operation: S+A-P. Set Call or B immediate value to bits fff_fffc of the
// calculation.
uint64_t BranchImm = Value + Addend - FinalAddress;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// "Check that -2^27 <= result < 2^27".
assert(isInt<28>(BranchImm));
@@ -352,91 +363,105 @@ void RuntimeDyldELF::resolveAArch64Relocation(const SectionEntry &Section,
}
case ELF::R_AARCH64_MOVW_UABS_G3: {
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffe0001fU;
+ TargetValue &= 0xffe0001fU;
// Immediate goes in bits 20:5 of MOVZ/MOVK instruction
- *TargetPtr |= Result >> (48 - 5);
+ TargetValue |= ((Result & 0xffff000000000000ULL) >> (48 - 5));
// Shift must be "lsl #48", in bits 22:21
- assert((*TargetPtr >> 21 & 0x3) == 3 && "invalid shift for relocation");
+ assert((TargetValue >> 21 & 0x3) == 3 && "invalid shift for relocation");
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_MOVW_UABS_G2_NC: {
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffe0001fU;
+ TargetValue &= 0xffe0001fU;
// Immediate goes in bits 20:5 of MOVZ/MOVK instruction
- *TargetPtr |= ((Result & 0xffff00000000ULL) >> (32 - 5));
+ TargetValue |= ((Result & 0xffff00000000ULL) >> (32 - 5));
// Shift must be "lsl #32", in bits 22:21
- assert((*TargetPtr >> 21 & 0x3) == 2 && "invalid shift for relocation");
+ assert((TargetValue >> 21 & 0x3) == 2 && "invalid shift for relocation");
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_MOVW_UABS_G1_NC: {
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffe0001fU;
+ TargetValue &= 0xffe0001fU;
// Immediate goes in bits 20:5 of MOVZ/MOVK instruction
- *TargetPtr |= ((Result & 0xffff0000U) >> (16 - 5));
+ TargetValue |= ((Result & 0xffff0000U) >> (16 - 5));
// Shift must be "lsl #16", in bits 22:2
- assert((*TargetPtr >> 21 & 0x3) == 1 && "invalid shift for relocation");
+ assert((TargetValue >> 21 & 0x3) == 1 && "invalid shift for relocation");
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_MOVW_UABS_G0_NC: {
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffe0001fU;
+ TargetValue &= 0xffe0001fU;
// Immediate goes in bits 20:5 of MOVZ/MOVK instruction
- *TargetPtr |= ((Result & 0xffffU) << 5);
+ TargetValue |= ((Result & 0xffffU) << 5);
// Shift must be "lsl #0", in bits 22:21.
- assert((*TargetPtr >> 21 & 0x3) == 0 && "invalid shift for relocation");
+ assert((TargetValue >> 21 & 0x3) == 0 && "invalid shift for relocation");
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_ADR_PREL_PG_HI21: {
// Operation: Page(S+A) - Page(P)
uint64_t Result =
((Value + Addend) & ~0xfffULL) - (FinalAddress & ~0xfffULL);
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// Check that -2^32 <= X < 2^32
assert(isInt<33>(Result) && "overflow check failed for relocation");
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0x9f00001fU;
+ TargetValue &= 0x9f00001fU;
// Immediate goes in bits 30:29 + 5:23 of ADRP instruction, taken
// from bits 32:12 of X.
- *TargetPtr |= ((Result & 0x3000U) << (29 - 12));
- *TargetPtr |= ((Result & 0x1ffffc000ULL) >> (14 - 5));
+ TargetValue |= ((Result & 0x3000U) << (29 - 12));
+ TargetValue |= ((Result & 0x1ffffc000ULL) >> (14 - 5));
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_LDST32_ABS_LO12_NC: {
// Operation: S + A
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffc003ffU;
+ TargetValue &= 0xffc003ffU;
// Immediate goes in bits 21:10 of LD/ST instruction, taken
// from bits 11:2 of X
- *TargetPtr |= ((Result & 0xffc) << (10 - 2));
+ TargetValue |= ((Result & 0xffc) << (10 - 2));
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
case ELF::R_AARCH64_LDST64_ABS_LO12_NC: {
// Operation: S + A
uint64_t Result = Value + Addend;
+ uint32_t TargetValue = support::ulittle32_t::ref{TargetPtr};
// AArch64 code is emitted with .rela relocations. The data already in any
// bits affected by the relocation on entry is garbage.
- *TargetPtr &= 0xffc003ffU;
+ TargetValue &= 0xffc003ffU;
// Immediate goes in bits 21:10 of LD/ST instruction, taken
// from bits 11:3 of X
- *TargetPtr |= ((Result & 0xff8) << (10 - 3));
+ TargetValue |= ((Result & 0xff8) << (10 - 3));
+ support::ulittle32_t::ref{TargetPtr} = TargetValue;
break;
}
}
diff --git a/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_BE-relocations.s b/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_BE-relocations.s
new file mode 100644
index 0000000..01d01e5
--- /dev/null
+++ b/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_BE-relocations.s
@@ -0,0 +1,102 @@
+# RUN: llvm-mc -triple=aarch64_be-none-linux-gnu -filetype=obj -o %T/be-reloc.o %s
+# RUN: llvm-rtdyld -triple=aarch64_be-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/be-reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# LE instructions read as BE
+# rtdyld-check: *{4}(g) = 0x6024e0d2
+# rtdyld-check: *{4}(g + 4) = 0xe0acc8f2
+# rtdyld-check: *{4}(g + 8) = 0x6035b1f2
+# rtdyld-check: *{4}(g + 12) = 0xe0bd99f2
+# rtdyld-check: *{8}k = f
+# RUN: llvm-mc -triple=aarch64_be-none-linux-gnu -filetype=obj -o %T/be-reloc.o %s
+# RUN: llvm-rtdyld -triple=aarch64_be-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/be-reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# LE instructions read as BE
+# rtdyld-check: *{4}(g) = 0x6024e0d2
+# rtdyld-check: *{4}(g + 4) = 0xe0acc8f2
+# rtdyld-check: *{4}(g + 8) = 0x6035b1f2
+# rtdyld-check: *{4}(g + 12) = 0xe0bd99f2
+# rtdyld-check: *{8}k = f
+# RUN: llvm-mc -triple=aarch64_be-none-linux-gnu -filetype=obj -o %T/be-reloc.o %s
+# RUN: llvm-rtdyld -triple=aarch64_be-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/be-reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# LE instructions read as BE
+# rtdyld-check: *{4}(g) = 0x6024e0d2
+# rtdyld-check: *{4}(g + 4) = 0xe0acc8f2
+# rtdyld-check: *{4}(g + 8) = 0x6035b1f2
+# rtdyld-check: *{4}(g + 12) = 0xe0bd99f2
+# rtdyld-check: *{8}k = f
diff --git a/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_relocations.s b/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_relocations.s
new file mode 100644
index 0000000..e07fa97
--- /dev/null
+++ b/llvm-3.9.1.src/test/ExecutionEngine/RuntimeDyld/AArch64/ELF_ARM64_relocations.s
@@ -0,0 +1,99 @@
+# RUN: llvm-mc -triple=arm64-none-linux-gnu -filetype=obj -o %T/reloc.o %s
+# RUN: llvm-rtdyld -triple=arm64-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# rtdyld-check: *{4}(g) = 0xd2e02460
+# rtdyld-check: *{4}(g + 4) = 0xf2c8ace0
+# rtdyld-check: *{4}(g + 8) = 0xf2b13560
+# rtdyld-check: *{4}(g + 12) = 0xf299bde0
+# rtdyld-check: *{8}k = f
+# RUN: llvm-mc -triple=arm64-none-linux-gnu -filetype=obj -o %T/reloc.o %s
+# RUN: llvm-rtdyld -triple=arm64-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# rtdyld-check: *{4}(g) = 0xd2e02460
+# rtdyld-check: *{4}(g + 4) = 0xf2c8ace0
+# rtdyld-check: *{4}(g + 8) = 0xf2b13560
+# rtdyld-check: *{4}(g + 12) = 0xf299bde0
+# rtdyld-check: *{8}k = f
+# RUN: llvm-mc -triple=arm64-none-linux-gnu -filetype=obj -o %T/reloc.o %s
+# RUN: llvm-rtdyld -triple=arm64-none-linux-gnu -verify -dummy-extern f=0x0123456789abcdef -check=%s %T/reloc.o
+
+ .text
+ .globl g
+ .p2align 2
+ .type g,@function
+g:
+# R_AARCH64_MOVW_UABS_G3
+ movz x0, #:abs_g3:f
+# R_AARCH64_MOVW_UABS_G2_NC
+ movk x0, #:abs_g2_nc:f
+# R_AARCH64_MOVW_UABS_G1_NC
+ movk x0, #:abs_g1_nc:f
+# R_AARCH64_MOVW_UABS_G0_NC
+ movk x0, #:abs_g0_nc:f
+ ret
+ .Lfunc_end0:
+ .size g, .Lfunc_end0-g
+
+ .type k,@object
+ .data
+ .globl k
+ .p2align 3
+k:
+ .xword f
+ .size k, 8
+
+# rtdyld-check: *{4}(g) = 0xd2e02460
+# rtdyld-check: *{4}(g + 4) = 0xf2c8ace0
+# rtdyld-check: *{4}(g + 8) = 0xf2b13560
+# rtdyld-check: *{4}(g + 12) = 0xf299bde0
+# rtdyld-check: *{8}k = f
--
2.12.0

View File

@ -1,53 +0,0 @@
From 04c252cc93fe5905a625773cd174ed6ca7651463 Mon Sep 17 00:00:00 2001
From: Will Schmidt <will_schmidt@vnet.ibm.com>
Date: Mon, 24 Mar 2014 16:04:15 +0000
Subject: [PATCH] [PPC64LE] ELFv2 ABI updates for the .opd section
[PPC64LE] ELFv2 ABI updates for the .opd section
The PPC64 Little Endian (PPC64LE) target supports the ELFv2 ABI, and as
such, does not have a ".opd" section. This is keyed off a _CALL_ELF=2
macro check.
The CALL_ELF check is not clearly documented at this time. The basis
for usage in this patch is from the gcc thread here:
http://gcc.gnu.org/ml/gcc-patches/2013-11/msg01144.html
> Adding comment from Uli:
Looks good to me. I think the old-style JIT doesn't really work
anyway for 64-bit, but at least with this patch LLVM will compile
and link again on a ppc64le host ...
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@204614 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Target/PowerPC/PPCJITInfo.cpp | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/lib/Target/PowerPC/PPCJITInfo.cpp b/lib/Target/PowerPC/PPCJITInfo.cpp
index 5e3a48d..227919c 100644
--- a/lib/Target/PowerPC/PPCJITInfo.cpp
+++ b/lib/Target/PowerPC/PPCJITInfo.cpp
@@ -214,6 +214,10 @@ asm(
".text\n"
".align 2\n"
".globl PPC64CompilationCallback\n"
+#if _CALL_ELF == 2
+ ".type PPC64CompilationCallback,@function\n"
+"PPC64CompilationCallback:\n"
+#else
".section \".opd\",\"aw\",@progbits\n"
".align 3\n"
"PPC64CompilationCallback:\n"
@@ -223,6 +227,7 @@ asm(
".align 4\n"
".type PPC64CompilationCallback,@function\n"
".L.PPC64CompilationCallback:\n"
+#endif
# else
asm(
".text\n"
--
1.9.3

View File

@ -1,119 +0,0 @@
From 95b15b3d2f180b15267032e16c947c0f9b8a112d Mon Sep 17 00:00:00 2001
From: Dave Airlie <airlied@redhat.com>
Date: Wed, 1 Mar 2017 13:02:38 +1000
Subject: [PATCH] Revert "Merging r280589:"
This reverts commit 25e2616626caafb896517e18cd8aa724fba2b200.
---
lib/Target/AMDGPU/SIInstructions.td | 1 -
lib/Target/AMDGPU/SIWholeQuadMode.cpp | 7 +++++
test/CodeGen/AMDGPU/wqm.ll | 49 +++--------------------------------
3 files changed, 11 insertions(+), 46 deletions(-)
diff --git a/lib/Target/AMDGPU/SIInstructions.td b/lib/Target/AMDGPU/SIInstructions.td
index dde5f2f..18b7d5d 100644
--- a/lib/Target/AMDGPU/SIInstructions.td
+++ b/lib/Target/AMDGPU/SIInstructions.td
@@ -2029,7 +2029,6 @@ def SI_RETURN : PseudoInstSI <
let hasSideEffects = 1;
let SALU = 1;
let hasNoSchedulingInfo = 1;
- let DisableWQM = 1;
}
let Uses = [EXEC], Defs = [EXEC, VCC, M0],
diff --git a/lib/Target/AMDGPU/SIWholeQuadMode.cpp b/lib/Target/AMDGPU/SIWholeQuadMode.cpp
index 1534d58..b200c15 100644
--- a/lib/Target/AMDGPU/SIWholeQuadMode.cpp
+++ b/lib/Target/AMDGPU/SIWholeQuadMode.cpp
@@ -219,6 +219,13 @@ char SIWholeQuadMode::scanInstructions(MachineFunction &MF,
markInstruction(MI, Flags, Worklist);
GlobalFlags |= Flags;
}
+
+ if (WQMOutputs && MBB.succ_empty()) {
+ // This is a prolog shader. Make sure we go back to exact mode at the end.
+ Blocks[&MBB].OutNeeds = StateExact;
+ Worklist.push_back(&MBB);
+ GlobalFlags |= StateExact;
+ }
}
return GlobalFlags;
diff --git a/test/CodeGen/AMDGPU/wqm.ll b/test/CodeGen/AMDGPU/wqm.ll
index 41e4264..809a7ba 100644
--- a/test/CodeGen/AMDGPU/wqm.ll
+++ b/test/CodeGen/AMDGPU/wqm.ll
@@ -17,18 +17,17 @@ main_body:
;CHECK-LABEL: {{^}}test2:
;CHECK-NEXT: ; %main_body
;CHECK-NEXT: s_wqm_b64 exec, exec
+;CHECK: image_sample
;CHECK-NOT: exec
-define amdgpu_ps void @test2(<8 x i32> inreg %rsrc, <4 x i32> inreg %sampler, float addrspace(1)* inreg %ptr, <4 x i32> %c) {
+;CHECK: _load_dword v0,
+define amdgpu_ps float @test2(<8 x i32> inreg %rsrc, <4 x i32> inreg %sampler, float addrspace(1)* inreg %ptr, <4 x i32> %c) {
main_body:
%c.1 = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %c, <8 x i32> %rsrc, <4 x i32> %sampler, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
%c.2 = bitcast <4 x float> %c.1 to <4 x i32>
%c.3 = extractelement <4 x i32> %c.2, i32 0
%gep = getelementptr float, float addrspace(1)* %ptr, i32 %c.3
%data = load float, float addrspace(1)* %gep
-
- call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %data, float undef, float undef, float undef)
-
- ret void
+ ret float %data
}
; ... but disabled for stores (and, in this simple case, not re-enabled).
@@ -415,46 +414,6 @@ entry:
ret void
}
-; Must return to exact at the end of a non-void returning shader,
-; otherwise the EXEC mask exported by the epilog will be wrong. This is true
-; even if the shader has no kills, because a kill could have happened in a
-; previous shader fragment.
-;
-; CHECK-LABEL: {{^}}test_nonvoid_return:
-; CHECK: s_mov_b64 [[LIVE:s\[[0-9]+:[0-9]+\]]], exec
-; CHECK: s_wqm_b64 exec, exec
-;
-; CHECK: s_and_b64 exec, exec, [[LIVE]]
-; CHECK-NOT: exec
-define amdgpu_ps <4 x float> @test_nonvoid_return() nounwind {
- %tex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> undef, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
- %tex.i = bitcast <4 x float> %tex to <4 x i32>
- %dtex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %tex.i, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
- ret <4 x float> %dtex
-}
-
-; CHECK-LABEL: {{^}}test_nonvoid_return_unreachable:
-; CHECK: s_mov_b64 [[LIVE:s\[[0-9]+:[0-9]+\]]], exec
-; CHECK: s_wqm_b64 exec, exec
-;
-; CHECK: s_and_b64 exec, exec, [[LIVE]]
-; CHECK-NOT: exec
-define amdgpu_ps <4 x float> @test_nonvoid_return_unreachable(i32 inreg %c) nounwind {
-entry:
- %tex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> undef, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
- %tex.i = bitcast <4 x float> %tex to <4 x i32>
- %dtex = call <4 x float> @llvm.SI.image.sample.v4i32(<4 x i32> %tex.i, <8 x i32> undef, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0)
-
- %cc = icmp sgt i32 %c, 0
- br i1 %cc, label %if, label %else
-
-if:
- store volatile <4 x float> %dtex, <4 x float>* undef
- unreachable
-
-else:
- ret <4 x float> %dtex
-}
declare void @llvm.amdgcn.image.store.v4i32(<4 x float>, <4 x i32>, <8 x i32>, i32, i1, i1, i1, i1) #1
declare void @llvm.amdgcn.buffer.store.f32(float, <4 x i32>, i32, i32, i1, i1) #1
--
2.9.3

View File

@ -1,30 +0,0 @@
From ce04fe5f8eb9f3a27504db75672083c8aaf80ddd Mon Sep 17 00:00:00 2001
From: Aaron Ballman <aaron@aaronballman.com>
Date: Tue, 19 Jul 2016 17:46:55 +0000
Subject: [PATCH] This code block breaks the docs build
(http://lab.llvm.org:8011/builders/llvm-sphinx-docs/builds/11920/steps/docs-llvm-html/logs/stdio),
but I cannot see anything immediately wrong with it and cannot reproduce the
diagnostic locally. Setting the code highlighting to none instead of nasm to
hopefully get the bot stumbling back towards green.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@275998 91177308-0d34-0410-b5e6-96231b3b80d8
---
docs/AMDGPUUsage.rst | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/docs/AMDGPUUsage.rst b/docs/AMDGPUUsage.rst
index 34a9b60..7d1ef11 100644
--- a/docs/AMDGPUUsage.rst
+++ b/docs/AMDGPUUsage.rst
@@ -171,7 +171,7 @@ keys, see the comments in lib/Target/AMDGPU/AmdKernelCodeT.h
Here is an example of a minimal amd_kernel_code_t specification:
-.. code-block:: nasm
+.. code-block:: none
.hsa_code_object_version 1,0
.hsa_code_object_isa
--
2.5.5

File diff suppressed because it is too large Load Diff

View File

@ -1,39 +0,0 @@
From fdda55bb968b2c39da76baa85a29114f53154944 Mon Sep 17 00:00:00 2001
From: Chris Bieneman <beanz@apple.com>
Date: Thu, 25 Aug 2016 20:53:00 +0000
Subject: [PATCH] cmake: Install CheckAtomic.cmake (needed by lldb)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Summary:
Install CheckAtomic.cmake along with other LLVM modules, therefore making it possible for other projects to use it. This file is needed for LLDB to be built standalone, and installing it was suggested in https://reviews.llvm.org/D23881.
Patch by: Michał Górny
Reviewers: krytarowski, zturner, eugenis, jyknight, labath, beanz
Subscribers: beanz, llvm-commits
Differential Revision: https://reviews.llvm.org/D23887
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@279777 91177308-0d34-0410-b5e6-96231b3b80d8
---
cmake/modules/CMakeLists.txt | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/cmake/modules/CMakeLists.txt b/cmake/modules/CMakeLists.txt
index 826dd36..d2510b8 100644
--- a/cmake/modules/CMakeLists.txt
+++ b/cmake/modules/CMakeLists.txt
@@ -91,6 +91,5 @@ if (NOT LLVM_INSTALL_TOOLCHAIN_ONLY)
PATTERN LLVMConfig.cmake EXCLUDE
PATTERN LLVMConfigVersion.cmake EXCLUDE
PATTERN LLVM-Config.cmake EXCLUDE
- PATTERN GetHostTriple.cmake EXCLUDE
- PATTERN CheckAtomic.cmake EXCLUDE)
+ PATTERN GetHostTriple.cmake EXCLUDE)
endif()
--
2.5.5

View File

@ -1,22 +0,0 @@
Preserve timestamps when installing data files
---
Makefile.rules | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Makefile.rules b/Makefile.rules
index f0c542b..0ff92bb 100644
--- a/Makefile.rules
+++ b/Makefile.rules
@@ -773,7 +773,7 @@ BCCompile.CXX = $(LLVMCXX) $(CPP.Flags) $(CXX.Flags) $(CXXFLAGS) $(CPPFLAGS) \
ProgInstall = $(INSTALL) $(Install.StripFlag) -m 0755
ScriptInstall = $(INSTALL) -m 0755
-DataInstall = $(INSTALL) -m 0644
+DataInstall = $(INSTALL) -p -m 0644
# When compiling under Mingw/Cygwin, the tblgen tool expects Windows
# paths. In this case, the SYSPATH function (defined in
--
1.8.3.1

View File

@ -1,44 +0,0 @@
From f12c36b2bc2e1db86098c181b88b8003c595e63c Mon Sep 17 00:00:00 2001
From: Renato Golin <renato.golin@linaro.org>
Date: Wed, 20 Jul 2016 09:47:09 +0000
Subject: [PATCH] [docs] fix cmake code-block warning
This will unblock the llvm-sphinx-buildbot, which is currently failing due
to a warning being treated as error.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@276100 91177308-0d34-0410-b5e6-96231b3b80d8
---
docs/CMakePrimer.rst | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/docs/CMakePrimer.rst b/docs/CMakePrimer.rst
index 0347790..1e3a09e 100644
--- a/docs/CMakePrimer.rst
+++ b/docs/CMakePrimer.rst
@@ -246,11 +246,11 @@ In general CMake if blocks work the way you'd expect:
.. code-block:: cmake
if(<condition>)
- .. do stuff
+ message("do stuff")
elseif(<condition>)
- .. do other stuff
+ message("do other stuff")
else()
- .. do other other stuff
+ message("do other other stuff")
endif()
The single most important thing to know about CMake's if blocks coming from a C
@@ -265,7 +265,7 @@ The most common form of the CMake ``foreach`` block is:
.. code-block:: cmake
foreach(var ...)
- .. do stuff
+ message("do stuff")
endforeach()
The variable argument portion of the ``foreach`` block can contain dereferenced
--
2.5.5

View File

@ -1,38 +0,0 @@
From de4fbfe93560c78f29c8b92cafab0793f5d26bc6 Mon Sep 17 00:00:00 2001
From: Aaron Ballman <aaron@aaronballman.com>
Date: Tue, 19 Jul 2016 20:20:03 +0000
Subject: [PATCH] This code block breaks the docs build
(http://lab.llvm.org:8011/builders/llvm-sphinx-docs/builds/11921/steps/docs-llvm-html/logs/stdio).
Setting the code highlighting to none instead of llvm to hopefully get the
bot stumbling back towards green.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@276018 91177308-0d34-0410-b5e6-96231b3b80d8
---
docs/BitCodeFormat.rst | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/docs/BitCodeFormat.rst b/docs/BitCodeFormat.rst
index ffa2176..89c7c1b 100644
--- a/docs/BitCodeFormat.rst
+++ b/docs/BitCodeFormat.rst
@@ -596,7 +596,7 @@ will be encoded as 1.
For example, instead of
-.. code-block:: llvm
+.. code-block:: none
#n = load #n-1
#n+1 = icmp eq #n, #const0
@@ -604,7 +604,7 @@ For example, instead of
version 1 will encode the instructions as
-.. code-block:: llvm
+.. code-block:: none
#n = load #1
#n+1 = icmp eq #1, (#n+1)-#const0
--
2.5.5

View File

@ -1,46 +0,0 @@
From 9871423412faa2ed8380445a26ed1b0991a18502 Mon Sep 17 00:00:00 2001
From: Aaron Ballman <aaron@aaronballman.com>
Date: Tue, 19 Jul 2016 23:50:11 +0000
Subject: [PATCH] This code block breaks the docs build
(http://lab.llvm.org:8011/builders/llvm-sphinx-docs/builds/11925/steps/docs-llvm-html/logs/stdio).
Setting the code highlighting to none instead of llvm.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@276060 91177308-0d34-0410-b5e6-96231b3b80d8
---
docs/BranchWeightMetadata.rst | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/docs/BranchWeightMetadata.rst b/docs/BranchWeightMetadata.rst
index 6cbcb0f..9e61d23 100644
--- a/docs/BranchWeightMetadata.rst
+++ b/docs/BranchWeightMetadata.rst
@@ -29,7 +29,7 @@ Supported Instructions
Metadata is only assigned to the conditional branches. There are two extra
operands for the true and the false branch.
-.. code-block:: llvm
+.. code-block:: none
!0 = metadata !{
metadata !"branch_weights",
@@ -43,7 +43,7 @@ operands for the true and the false branch.
Branch weights are assigned to every case (including the ``default`` case which
is always case #0).
-.. code-block:: llvm
+.. code-block:: none
!0 = metadata !{
metadata !"branch_weights",
@@ -56,7 +56,7 @@ is always case #0).
Branch weights are assigned to every destination.
-.. code-block:: llvm
+.. code-block:: none
!0 = metadata !{
metadata !"branch_weights",
--
2.5.5

View File

@ -1,24 +0,0 @@
From e746af90e6aa52af6e6bb0b88ffe20dbd97a1e4d Mon Sep 17 00:00:00 2001
From: Stephan Bergmann <sbergman@redhat.com>
Date: Thu, 3 Dec 2015 14:53:39 +0100
Subject: [PATCH 2/3] Adapt previous (Clang trunk) patch to Clang 3.7
---
lib/AST/ItaniumMangle.cpp | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/AST/ItaniumMangle.cpp b/lib/AST/ItaniumMangle.cpp
index e32b659..19a09dc 100644
--- a/tools/clang/lib/AST/ItaniumMangle.cpp
+++ b/tools/clang/lib/AST/ItaniumMangle.cpp
@@ -31,6 +31,7 @@
#include "llvm/ADT/StringExtras.h"
#include "llvm/Support/ErrorHandling.h"
#include "llvm/Support/raw_ostream.h"
+#include <set>
#define MANGLE_CHECKER 0
--
2.4.3

View File

@ -1,12 +0,0 @@
diff -up llvm-3.4/tools/llvm-shlib/Makefile.orig llvm-3.4/tools/llvm-shlib/Makefile
--- llvm-3.4/tools/llvm-shlib/Makefile.orig 2013-11-01 00:35:00.000000000 +1000
+++ llvm-3.4/tools/llvm-shlib/Makefile 2014-01-14 10:13:20.069858909 +1000
@@ -75,7 +75,7 @@ endif
ifeq ($(HOST_OS), $(filter $(HOST_OS), Linux GNU GNU/kFreeBSD))
# Don't allow unresolved symbols.
- LLVMLibsOptions += -Wl,--no-undefined
+ LLVMLibsOptions += -Wl,--no-undefined -Wl,-Bsymbolic
endif
ifeq ($(HOST_OS),SunOS)

View File

@ -1,16 +0,0 @@
diff -up llvm-3.4.1.src/tools/clang/lib/Driver/Tools.cpp.jx llvm-3.4.1.src/tools/clang/lib/Driver/Tools.cpp
--- llvm-3.4.1.src/tools/clang/lib/Driver/Tools.cpp.jx 2013-12-08 21:59:27.000000000 -0500
+++ llvm-3.4.1.src/tools/clang/lib/Driver/Tools.cpp 2014-06-02 11:24:07.628292753 -0400
@@ -763,9 +763,9 @@ static StringRef getARMFloatABI(const Dr
break;
}
default:
- // Assume "soft", but warn the user we are guessing.
- FloatABI = "soft";
- D.Diag(diag::warn_drv_assuming_mfloat_abi_is) << "soft";
+ // Assume "hard", but warn the user we are guessing.
+ FloatABI = "hard";
+ D.Diag(diag::warn_drv_assuming_mfloat_abi_is) << "hard";
break;
}
}

View File

@ -1,12 +0,0 @@
diff -up llvm-3.7.0.src/include/llvm/CodeGen/CommandFlags.h.dma llvm-3.7.0.src/include/llvm/CodeGen/CommandFlags.h
--- llvm-3.7.0.src/include/llvm/CodeGen/CommandFlags.h.dma 2015-09-15 15:01:41.570871134 +1000
+++ llvm-3.7.0.src/include/llvm/CodeGen/CommandFlags.h 2015-09-15 15:01:53.642191476 +1000
@@ -21,7 +21,7 @@
#include "llvm/IR/Intrinsics.h"
#include "llvm/IR/Module.h"
#include "llvm/MC/MCTargetOptionsCommandFlags.h"
-#include "llvm//MC/SubtargetFeature.h"
+#include "llvm/MC/SubtargetFeature.h"
#include "llvm/Support/CodeGen.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/Host.h"

View File

@ -1,16 +0,0 @@
Fix broken symlink to Python's lldb module
Resolves: #1177143
diff -rupN lldb-3.5.0.src/scripts/Python/finish-swig-Python-LLDB.sh lldb-3.5.0.src-new/scripts/Python/finish-swig-Python-LLDB.sh
--- lldb-3.5.0.src/scripts/Python/finish-swig-Python-LLDB.sh 2014-07-01 19:57:19.000000000 +0200
+++ lldb-3.5.0.src-new/scripts/Python/finish-swig-Python-LLDB.sh 2014-12-24 14:18:10.068604693 +0100
@@ -158,7 +158,7 @@ then
then
ln -s "../../../LLDB" _lldb.so
else
- ln -s "../../../liblldb${SOEXT}" _lldb.so
+ ln -s "../../../llvm/liblldb${SOEXT}" _lldb.so
fi
else
if [ $Debug -eq 1 ]

View File

@ -1,16 +0,0 @@
diff -up lldb/scripts/Python/modules/readline/Makefile.jx lldb/scripts/Python/modules/readline/Makefile
--- lldb/scripts/Python/modules/readline/Makefile.jx 2014-02-26 10:05:48.000000000 -0500
+++ lldb/scripts/Python/modules/readline/Makefile 2014-10-14 11:55:05.112566400 -0400
@@ -91,9 +91,8 @@ $(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)
# Target to move the shared library from the build python lib dir to
# the install python lib dir.
install-local:: $(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)$(SHLIBEXT)
- $(Echo) Installing $(BuildMode) $(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)$(SHLIBEXT) to $(DESTDIR)$(prefix)/lib/$(LLDB_PYTHON_MODULE_REL_DIR)
- $(Verb) $(MKDIR) "$(DESTDIR)$(prefix)/lib/$(LLDB_PYTHON_MODULE_REL_DIR)"
- $(Verb) $(ProgInstall) "$(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)$(SHLIBEXT)" "$(DESTDIR)$(prefix)/lib/$(LLDB_PYTHON_MODULE_REL_DIR)"
- $(Verb) $(RM) "$(DESTDIR)$(prefix)/lib/$(LIBRARYNAME)$(SHLIBEXT)"
+ $(Echo) Installing $(BuildMode) $(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)$(SHLIBEXT) to $(DESTDIR)$(prefix)/@lib@/$(LLDB_PYTHON_MODULE_REL_DIR)
+ $(Verb) $(MKDIR) "$(DESTDIR)$(prefix)/@lib@/$(LLDB_PYTHON_MODULE_REL_DIR)"
+ $(Verb) $(ProgInstall) "$(LLDB_PYTHON_MODULE_DIR)/$(LIBRARYNAME)$(SHLIBEXT)" "$(DESTDIR)$(prefix)/@lib@/$(LLDB_PYTHON_MODULE_REL_DIR)"
endif # if !defined(LLDB_DISABLE_PYTHON)

View File

@ -1,459 +0,0 @@
diff --git a/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.cpp b/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.cpp
index 99e1377..7105879 100644
--- a/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.cpp
+++ b/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.cpp
@@ -316,6 +316,37 @@ void AMDGPUInstPrinter::printKCache(const MCInst *MI, unsigned OpNo,
}
}
+void AMDGPUInstPrinter::printSendMsg(const MCInst *MI, unsigned OpNo,
+ raw_ostream &O) {
+ unsigned SImm16 = MI->getOperand(OpNo).getImm();
+ unsigned Msg = SImm16 & 0xF;
+ if (Msg == 2 || Msg == 3) {
+ unsigned Op = (SImm16 >> 4) & 0xF;
+ if (Msg == 3)
+ O << "Gs_done(";
+ else
+ O << "Gs(";
+ if (Op == 0) {
+ O << "nop";
+ } else {
+ unsigned Stream = (SImm16 >> 8) & 0x3;
+ if (Op == 1)
+ O << "cut";
+ else if (Op == 2)
+ O << "emit";
+ else if (Op == 3)
+ O << "emit-cut";
+ O << " stream " << Stream;
+ }
+ O << "), [m0] ";
+ } else if (Msg == 1)
+ O << "interrupt ";
+ else if (Msg == 15)
+ O << "system ";
+ else
+ O << "unknown(" << Msg << ") ";
+}
+
void AMDGPUInstPrinter::printWaitFlag(const MCInst *MI, unsigned OpNo,
raw_ostream &O) {
// Note: Mask values are taken from SIInsertWaits.cpp and not from ISA docs
diff --git a/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.h b/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.h
index 77af942..2876dd2 100644
--- a/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.h
+++ b/lib/Target/R600/InstPrinter/AMDGPUInstPrinter.h
@@ -53,6 +53,7 @@ private:
void printRSel(const MCInst *MI, unsigned OpNo, raw_ostream &O);
void printCT(const MCInst *MI, unsigned OpNo, raw_ostream &O);
void printKCache(const MCInst *MI, unsigned OpNo, raw_ostream &O);
+ void printSendMsg(const MCInst *MI, unsigned OpNo, raw_ostream &O);
void printWaitFlag(const MCInst *MI, unsigned OpNo, raw_ostream &O);
};
diff --git a/lib/Target/R600/SIInsertWaits.cpp b/lib/Target/R600/SIInsertWaits.cpp
index 7ef662e..695ec40 100644
--- a/lib/Target/R600/SIInsertWaits.cpp
+++ b/lib/Target/R600/SIInsertWaits.cpp
@@ -314,6 +314,12 @@ Counters SIInsertWaits::handleOperands(MachineInstr &MI) {
Counters Result = ZeroCounts;
+ // S_SENDMSG implicitly waits for all outstanding LGKM transfers to finish,
+ // but we also want to wait for any other outstanding transfers before
+ // signalling other hardware blocks
+ if (MI.getOpcode() == AMDGPU::S_SENDMSG)
+ return LastIssued;
+
// For each register affected by this
// instruction increase the result sequence
for (unsigned i = 0, e = MI.getNumOperands(); i != e; ++i) {
diff --git a/lib/Target/R600/SIInstrInfo.td b/lib/Target/R600/SIInstrInfo.td
index 4cd0daa..19d2171 100644
--- a/lib/Target/R600/SIInstrInfo.td
+++ b/lib/Target/R600/SIInstrInfo.td
@@ -425,26 +425,48 @@ class MTBUF_Store_Helper <bits<3> op, string asm, RegisterClass regClass> : MTBU
multiclass MUBUF_Load_Helper <bits<7> op, string asm, RegisterClass regClass> {
- let glc = 0, lds = 0, slc = 0, tfe = 0, soffset = 128 /* ZERO */,
- mayLoad = 1 in {
-
- let offen = 1, idxen = 0, addr64 = 0, offset = 0 in {
- def _OFFEN : MUBUF <op, (outs regClass:$vdata),
- (ins SReg_128:$srsrc, VReg_32:$vaddr),
- asm#" $vdata, $srsrc + $vaddr", []>;
- }
-
- let offen = 0, idxen = 1, addr64 = 0 in {
- def _IDXEN : MUBUF <op, (outs regClass:$vdata),
- (ins SReg_128:$srsrc, VReg_32:$vaddr, i16imm:$offset),
- asm#" $vdata, $srsrc[$vaddr] + $offset", []>;
- }
+ let lds = 0, mayLoad = 1 in {
+
+ let addr64 = 0 in {
+
+ let offen = 0, idxen = 0 in {
+ def _OFFSET : MUBUF <op, (outs regClass:$vdata),
+ (ins SReg_128:$srsrc, VReg_32:$vaddr,
+ i16imm:$offset, SSrc_32:$soffset, i1imm:$glc,
+ i1imm:$slc, i1imm:$tfe),
+ asm#" $vdata, $srsrc + $offset + $soffset, glc=$glc, slc=$slc, tfe=$tfe", []>;
+ }
+
+ let offen = 1, idxen = 0, offset = 0 in {
+ def _OFFEN : MUBUF <op, (outs regClass:$vdata),
+ (ins SReg_128:$srsrc, VReg_32:$vaddr,
+ SSrc_32:$soffset, i1imm:$glc, i1imm:$slc,
+ i1imm:$tfe),
+ asm#" $vdata, $srsrc + $vaddr + $soffset, glc=$glc, slc=$slc, tfe=$tfe", []>;
+ }
+
+ let offen = 0, idxen = 1 in {
+ def _IDXEN : MUBUF <op, (outs regClass:$vdata),
+ (ins SReg_128:$srsrc, VReg_32:$vaddr,
+ i16imm:$offset, SSrc_32:$soffset, i1imm:$glc,
+ i1imm:$slc, i1imm:$tfe),
+ asm#" $vdata, $srsrc[$vaddr] + $offset + $soffset, glc=$glc, slc=$slc, tfe=$tfe", []>;
+ }
+
+ let offen = 1, idxen = 1 in {
+ def _BOTHEN : MUBUF <op, (outs regClass:$vdata),
+ (ins SReg_128:$srsrc, VReg_64:$vaddr,
+ SSrc_32:$soffset, i1imm:$glc,
+ i1imm:$slc, i1imm:$tfe),
+ asm#" $vdata, $srsrc[$vaddr[0]] + $vaddr[1] + $soffset, glc=$glc, slc=$slc, tfe=$tfe", []>;
+ }
+ }
- let offen = 0, idxen = 0, addr64 = 1 in {
- def _ADDR64 : MUBUF <op, (outs regClass:$vdata),
- (ins SReg_128:$srsrc, VReg_64:$vaddr, i16imm:$offset),
- asm#" $vdata, $srsrc + $vaddr + $offset", []>;
- }
+ let offen = 0, idxen = 0, addr64 = 1, glc = 0, slc = 0, tfe = 0, soffset = 128 /* ZERO */ in {
+ def _ADDR64 : MUBUF <op, (outs regClass:$vdata),
+ (ins SReg_128:$srsrc, VReg_64:$vaddr, i16imm:$offset),
+ asm#" $vdata, $srsrc + $vaddr + $offset", []>;
+ }
}
}
diff --git a/lib/Target/R600/SIInstructions.td b/lib/Target/R600/SIInstructions.td
index 76f05eb..9acb9b6 100644
--- a/lib/Target/R600/SIInstructions.td
+++ b/lib/Target/R600/SIInstructions.td
@@ -22,6 +22,10 @@ def InterpSlot : Operand<i32> {
let PrintMethod = "printInterpSlot";
}
+def SendMsgImm : Operand<i32> {
+ let PrintMethod = "printSendMsg";
+}
+
def isSI : Predicate<"Subtarget.getGeneration() "
">= AMDGPUSubtarget::SOUTHERN_ISLANDS">;
@@ -826,17 +830,25 @@ def S_BARRIER : SOPP <0x0000000a, (ins), "S_BARRIER",
def S_WAITCNT : SOPP <0x0000000c, (ins WAIT_FLAG:$simm16), "S_WAITCNT $simm16",
[]
>;
-} // End hasSideEffects
//def S_SETHALT : SOPP_ <0x0000000d, "S_SETHALT", []>;
//def S_SLEEP : SOPP_ <0x0000000e, "S_SLEEP", []>;
//def S_SETPRIO : SOPP_ <0x0000000f, "S_SETPRIO", []>;
-//def S_SENDMSG : SOPP_ <0x00000010, "S_SENDMSG", []>;
+
+let Uses = [EXEC] in {
+ def S_SENDMSG : SOPP <0x00000010, (ins SendMsgImm:$simm16, M0Reg:$m0), "S_SENDMSG $simm16",
+ [(int_SI_sendmsg imm:$simm16, M0Reg:$m0)]
+ > {
+ let DisableEncoding = "$m0";
+ }
+} // End Uses = [EXEC]
+
//def S_SENDMSGHALT : SOPP_ <0x00000011, "S_SENDMSGHALT", []>;
//def S_TRAP : SOPP_ <0x00000012, "S_TRAP", []>;
//def S_ICACHE_INV : SOPP_ <0x00000013, "S_ICACHE_INV", []>;
//def S_INCPERFLEVEL : SOPP_ <0x00000014, "S_INCPERFLEVEL", []>;
//def S_DECPERFLEVEL : SOPP_ <0x00000015, "S_DECPERFLEVEL", []>;
//def S_TTRACEDATA : SOPP_ <0x00000016, "S_TTRACEDATA", []>;
+} // End hasSideEffects
def V_CNDMASK_B32_e32 : VOP2 <0x00000000, (outs VReg_32:$dst),
(ins VSrc_32:$src0, VReg_32:$src1, VCCReg:$vcc),
@@ -1305,8 +1317,8 @@ def SI_END_CF : InstSI <
def SI_KILL : InstSI <
(outs),
- (ins VReg_32:$src),
- "SI_KIL $src",
+ (ins VSrc_32:$src),
+ "SI_KILL $src",
[(int_AMDGPU_kill f32:$src)]
>;
@@ -1397,13 +1409,13 @@ def : Pat<
def : Pat <
(int_AMDGPU_kilp),
- (SI_KILL (V_MOV_B32_e32 0xbf800000))
+ (SI_KILL 0xbf800000)
>;
/* int_SI_vs_load_input */
def : Pat<
(SIload_input i128:$tlst, IMM12bit:$attr_offset, i32:$buf_idx_vgpr),
- (BUFFER_LOAD_FORMAT_XYZW_IDXEN $tlst, $buf_idx_vgpr, imm:$attr_offset)
+ (BUFFER_LOAD_FORMAT_XYZW_IDXEN $tlst, $buf_idx_vgpr, imm:$attr_offset, 0, 0, 0, 0)
>;
/* int_SI_export */
@@ -1809,7 +1821,7 @@ def : Pat <
// 3. Offset in an 32Bit VGPR
def : Pat <
(SIload_constant i128:$sbase, i32:$voff),
- (BUFFER_LOAD_DWORD_OFFEN $sbase, $voff)
+ (BUFFER_LOAD_DWORD_OFFEN $sbase, $voff, 0, 0, 0, 0)
>;
// The multiplication scales from [0,1] to the unsigned integer range
@@ -1970,6 +1982,50 @@ defm : MUBUFStore_Pattern <BUFFER_STORE_DWORDX2, i64, global_store>;
defm : MUBUFStore_Pattern <BUFFER_STORE_DWORDX2, v2i32, global_store>;
defm : MUBUFStore_Pattern <BUFFER_STORE_DWORDX4, v4i32, global_store>;
+// BUFFER_LOAD_DWORD*, addr64=0
+multiclass MUBUF_Load_Dword <ValueType vt, MUBUF offset, MUBUF offen, MUBUF idxen,
+ MUBUF bothen> {
+
+ def : Pat <
+ (vt (int_SI_buffer_load_dword i128:$rsrc, i32:$vaddr, i32:$soffset,
+ imm:$offset, 0, 0, imm:$glc, imm:$slc,
+ imm:$tfe)),
+ (offset $rsrc, $vaddr, (as_i16imm $offset), $soffset, (as_i1imm $glc),
+ (as_i1imm $slc), (as_i1imm $tfe))
+ >;
+
+ def : Pat <
+ (vt (int_SI_buffer_load_dword i128:$rsrc, i32:$vaddr, i32:$soffset,
+ imm, 1, 0, imm:$glc, imm:$slc,
+ imm:$tfe)),
+ (offen $rsrc, $vaddr, $soffset, (as_i1imm $glc), (as_i1imm $slc),
+ (as_i1imm $tfe))
+ >;
+
+ def : Pat <
+ (vt (int_SI_buffer_load_dword i128:$rsrc, i32:$vaddr, i32:$soffset,
+ imm:$offset, 0, 1, imm:$glc, imm:$slc,
+ imm:$tfe)),
+ (idxen $rsrc, $vaddr, (as_i16imm $offset), $soffset, (as_i1imm $glc),
+ (as_i1imm $slc), (as_i1imm $tfe))
+ >;
+
+ def : Pat <
+ (vt (int_SI_buffer_load_dword i128:$rsrc, v2i32:$vaddr, i32:$soffset,
+ imm, 1, 1, imm:$glc, imm:$slc,
+ imm:$tfe)),
+ (bothen $rsrc, $vaddr, $soffset, (as_i1imm $glc), (as_i1imm $slc),
+ (as_i1imm $tfe))
+ >;
+}
+
+defm : MUBUF_Load_Dword <i32, BUFFER_LOAD_DWORD_OFFSET, BUFFER_LOAD_DWORD_OFFEN,
+ BUFFER_LOAD_DWORD_IDXEN, BUFFER_LOAD_DWORD_BOTHEN>;
+defm : MUBUF_Load_Dword <v2i32, BUFFER_LOAD_DWORDX2_OFFSET, BUFFER_LOAD_DWORDX2_OFFEN,
+ BUFFER_LOAD_DWORDX2_IDXEN, BUFFER_LOAD_DWORDX2_BOTHEN>;
+defm : MUBUF_Load_Dword <v4i32, BUFFER_LOAD_DWORDX4_OFFSET, BUFFER_LOAD_DWORDX4_OFFEN,
+ BUFFER_LOAD_DWORDX4_IDXEN, BUFFER_LOAD_DWORDX4_BOTHEN>;
+
//===----------------------------------------------------------------------===//
// MTBUF Patterns
//===----------------------------------------------------------------------===//
diff --git a/lib/Target/R600/SIIntrinsics.td b/lib/Target/R600/SIIntrinsics.td
index 7fcc964..00e32c0 100644
--- a/lib/Target/R600/SIIntrinsics.td
+++ b/lib/Target/R600/SIIntrinsics.td
@@ -38,6 +38,22 @@ let TargetPrefix = "SI", isTarget = 1 in {
llvm_i32_ty], // tfe(imm)
[]>;
+ // Fully-flexible BUFFER_LOAD_DWORD_* except for the ADDR64 bit, which is not exposed
+ def int_SI_buffer_load_dword : Intrinsic <
+ [llvm_anyint_ty], // vdata(VGPR), overloaded for types i32, v2i32, v4i32
+ [llvm_anyint_ty, // rsrc(SGPR)
+ llvm_anyint_ty, // vaddr(VGPR)
+ llvm_i32_ty, // soffset(SGPR)
+ llvm_i32_ty, // inst_offset(imm)
+ llvm_i32_ty, // offen(imm)
+ llvm_i32_ty, // idxen(imm)
+ llvm_i32_ty, // glc(imm)
+ llvm_i32_ty, // slc(imm)
+ llvm_i32_ty], // tfe(imm)
+ [IntrReadArgMem]>;
+
+ def int_SI_sendmsg : Intrinsic <[], [llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
+
class Sample : Intrinsic <[llvm_v4f32_ty], [llvm_anyvector_ty, llvm_v32i8_ty, llvm_anyint_ty, llvm_i32_ty], [IntrNoMem]>;
def int_SI_sample : Sample;
diff --git a/lib/Target/R600/SILowerControlFlow.cpp b/lib/Target/R600/SILowerControlFlow.cpp
index 958763d..254f3a6 100644
--- a/lib/Target/R600/SILowerControlFlow.cpp
+++ b/lib/Target/R600/SILowerControlFlow.cpp
@@ -55,6 +55,7 @@
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
#include "llvm/CodeGen/MachineRegisterInfo.h"
+#include "llvm/IR/Constants.h"
using namespace llvm;
@@ -145,7 +146,9 @@ void SILowerControlFlowPass::SkipIfDead(MachineInstr &MI) {
MachineBasicBlock &MBB = *MI.getParent();
DebugLoc DL = MI.getDebugLoc();
- if (!shouldSkip(&MBB, &MBB.getParent()->back()))
+ if (MBB.getParent()->getInfo<SIMachineFunctionInfo>()->ShaderType !=
+ ShaderType::PIXEL ||
+ !shouldSkip(&MBB, &MBB.getParent()->back()))
return;
MachineBasicBlock::iterator Insert = &MI;
@@ -295,15 +298,27 @@ void SILowerControlFlowPass::Kill(MachineInstr &MI) {
MachineBasicBlock &MBB = *MI.getParent();
DebugLoc DL = MI.getDebugLoc();
+ const MachineOperand &Op = MI.getOperand(0);
- // Kill is only allowed in pixel shaders
+ // Kill is only allowed in pixel / geometry shaders
assert(MBB.getParent()->getInfo<SIMachineFunctionInfo>()->ShaderType ==
- ShaderType::PIXEL);
-
- // Clear this pixel from the exec mask if the operand is negative
- BuildMI(MBB, &MI, DL, TII->get(AMDGPU::V_CMPX_LE_F32_e32), AMDGPU::VCC)
- .addImm(0)
- .addOperand(MI.getOperand(0));
+ ShaderType::PIXEL ||
+ MBB.getParent()->getInfo<SIMachineFunctionInfo>()->ShaderType ==
+ ShaderType::GEOMETRY);
+
+ // Clear this thread from the exec mask if the operand is negative
+ if ((Op.isImm() || Op.isFPImm())) {
+ // Constant operand: Set exec mask to 0 or do nothing
+ if (Op.isImm() ? (Op.getImm() & 0x80000000) :
+ Op.getFPImm()->isNegative()) {
+ BuildMI(MBB, &MI, DL, TII->get(AMDGPU::S_MOV_B64), AMDGPU::EXEC)
+ .addImm(0);
+ }
+ } else {
+ BuildMI(MBB, &MI, DL, TII->get(AMDGPU::V_CMPX_LE_F32_e32), AMDGPU::VCC)
+ .addImm(0)
+ .addOperand(Op);
+ }
MI.eraseFromParent();
}
diff --git a/test/CodeGen/R600/llvm.AMDGPU.kill.ll b/test/CodeGen/R600/llvm.AMDGPU.kill.ll
new file mode 100644
index 0000000..4ab6a8a
--- /dev/null
+++ b/test/CodeGen/R600/llvm.AMDGPU.kill.ll
@@ -0,0 +1,22 @@
+; RUN: llc < %s -march=r600 -mcpu=verde -verify-machineinstrs | FileCheck --check-prefix=SI %s
+
+; SI-LABEL: @kill_gs_const
+; SI-NOT: V_CMPX_LE_F32
+; SI: S_MOV_B64 exec, 0
+
+define void @kill_gs_const() #0 {
+main_body:
+ %0 = icmp ule i32 0, 3
+ %1 = select i1 %0, float 1.000000e+00, float -1.000000e+00
+ call void @llvm.AMDGPU.kill(float %1)
+ %2 = icmp ule i32 3, 0
+ %3 = select i1 %2, float 1.000000e+00, float -1.000000e+00
+ call void @llvm.AMDGPU.kill(float %3)
+ ret void
+}
+
+declare void @llvm.AMDGPU.kill(float)
+
+attributes #0 = { "ShaderType"="2" }
+
+!0 = metadata !{metadata !"const", null, i32 1}
diff --git a/test/CodeGen/R600/llvm.SI.load.dword.ll b/test/CodeGen/R600/llvm.SI.load.dword.ll
new file mode 100644
index 0000000..a622775
--- /dev/null
+++ b/test/CodeGen/R600/llvm.SI.load.dword.ll
@@ -0,0 +1,40 @@
+;RUN: llc < %s -march=r600 -mcpu=verde -verify-machineinstrs | FileCheck %s
+
+; Example of a simple geometry shader loading vertex attributes from the
+; ESGS ring buffer
+
+; CHECK-LABEL: @main
+; CHECK: BUFFER_LOAD_DWORD
+; CHECK: BUFFER_LOAD_DWORD
+; CHECK: BUFFER_LOAD_DWORD
+; CHECK: BUFFER_LOAD_DWORD
+
+define void @main([17 x <16 x i8>] addrspace(2)* byval, [32 x <16 x i8>] addrspace(2)* byval, [16 x <32 x i8>] addrspace(2)* byval, [2 x <16 x i8>] addrspace(2)* byval, [17 x <16 x i8>] addrspace(2)* inreg, [17 x <16 x i8>] addrspace(2)* inreg, i32, i32, i32, i32) #0 {
+main_body:
+ %10 = getelementptr [2 x <16 x i8>] addrspace(2)* %3, i64 0, i32 1
+ %11 = load <16 x i8> addrspace(2)* %10, !tbaa !0
+ %12 = shl i32 %6, 2
+ %13 = call i32 @llvm.SI.buffer.load.dword.i32.i32(<16 x i8> %11, i32 0, i32 0, i32 0, i32 0, i32 0, i32 1, i32 1, i32 0)
+ %14 = bitcast i32 %13 to float
+ %15 = call i32 @llvm.SI.buffer.load.dword.i32.i32(<16 x i8> %11, i32 %12, i32 0, i32 0, i32 1, i32 0, i32 1, i32 1, i32 0)
+ %16 = bitcast i32 %15 to float
+ %17 = call i32 @llvm.SI.buffer.load.dword.i32.i32(<16 x i8> %11, i32 %12, i32 0, i32 0, i32 0, i32 1, i32 1, i32 1, i32 0)
+ %18 = bitcast i32 %17 to float
+ %19 = call i32 @llvm.SI.buffer.load.dword.i32.v2i32(<16 x i8> %11, <2 x i32> <i32 0, i32 0>, i32 0, i32 0, i32 1, i32 1, i32 1, i32 1, i32 0)
+ %20 = bitcast i32 %19 to float
+ call void @llvm.SI.export(i32 15, i32 0, i32 1, i32 12, i32 0, float %14, float %16, float %18, float %20)
+ ret void
+}
+
+; Function Attrs: nounwind readonly
+declare i32 @llvm.SI.buffer.load.dword.i32.i32(<16 x i8>, i32, i32, i32, i32, i32, i32, i32, i32) #1
+
+; Function Attrs: nounwind readonly
+declare i32 @llvm.SI.buffer.load.dword.i32.v2i32(<16 x i8>, <2 x i32>, i32, i32, i32, i32, i32, i32, i32) #1
+
+declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)
+
+attributes #0 = { "ShaderType"="1" }
+attributes #1 = { nounwind readonly }
+
+!0 = metadata !{metadata !"const", null, i32 1}
diff --git a/test/CodeGen/R600/llvm.SI.sendmsg.ll b/test/CodeGen/R600/llvm.SI.sendmsg.ll
new file mode 100644
index 0000000..581d422
--- /dev/null
+++ b/test/CodeGen/R600/llvm.SI.sendmsg.ll
@@ -0,0 +1,21 @@
+;RUN: llc < %s -march=r600 -mcpu=verde -verify-machineinstrs | FileCheck %s
+
+; CHECK-LABEL: @main
+; CHECK: S_SENDMSG Gs(emit stream 0)
+; CHECK: S_SENDMSG Gs(cut stream 1)
+; CHECK: S_SENDMSG Gs(emit-cut stream 2)
+; CHECK: S_SENDMSG Gs_done(nop)
+
+define void @main() {
+main_body:
+ call void @llvm.SI.sendmsg(i32 34, i32 0);
+ call void @llvm.SI.sendmsg(i32 274, i32 0);
+ call void @llvm.SI.sendmsg(i32 562, i32 0);
+ call void @llvm.SI.sendmsg(i32 3, i32 0);
+ ret void
+}
+
+; Function Attrs: nounwind
+declare void @llvm.SI.sendmsg(i32, i32) #0
+
+attributes #0 = { nounwind }

View File

@ -1,13 +0,0 @@
Index: docs/CommandGuide/lit.rst
===================================================================
--- docs/CommandGuide/lit.rst (revision 294645)
+++ docs/CommandGuide/lit.rst (revision 294646)
@@ -56,7 +56,7 @@
Search for :file:`{NAME}.cfg` and :file:`{NAME}.site.cfg` when searching for
test suites, instead of :file:`lit.cfg` and :file:`lit.site.cfg`.
-.. option:: -D NAME, -D NAME=VALUE, --param NAME, --param NAME=VALUE
+.. option:: -D NAME[=VALUE], --param NAME[=VALUE]
Add a user defined parameter ``NAME`` with the given ``VALUE`` (or the empty
string if not given). The meaning and use of these parameters is test suite

View File

@ -6,8 +6,8 @@
%endif
Name: llvm
Version: 3.9.1
Release: 6%{?dist}
Version: 4.0.0
Release: 1%{?dist}
Summary: The Low Level Virtual Machine
License: NCSA
@ -18,27 +18,7 @@ Source100: llvm-config.h
# recognize s390 as SystemZ when configuring build
Patch0: llvm-3.7.1-cmake-s390.patch
Patch1: 0001-This-code-block-breaks-the-docs-build-http-lab.llvm..patch
Patch2: 0001-fix-docs-2.patch
Patch3: 0001-fix-docs-3.patch
Patch4: 0001-docs-fix-cmake-code-block-warning.patch
# backport from upstream to fix lldb out of tree
Patch5: 0001-cmake-Install-CheckAtomic.cmake-needed-by-lldb.patch
# Upstream patch to fix doc build
# http://llvm.org/viewvc/llvm-project?view=revision&revision=294646
Patch6: llvm-r294646.patch
# This fix caused regressions
Patch7: 0001-Revert-Merging-r280589.patch
# https://reviews.llvm.org/D27609
Patch8: 0001-Fix-R_AARCH64_MOVW_UABS_G3-relocation.patch
# backports cribbed from https://github.com/rust-lang/llvm/
Patch47: rust-lang-llvm-pr47.patch
Patch53: rust-lang-llvm-pr53.patch
Patch54: rust-lang-llvm-pr54.patch
Patch55: rust-lang-llvm-pr55.patch
Patch57: rust-lang-llvm-pr57.patch
Patch1: 0001-CMake-Fix-pthread-handling-for-out-of-tree-builds.patch
BuildRequires: cmake
BuildRequires: zlib-devel
@ -91,19 +71,7 @@ Static libraries for the LLVM compiler infrastructure.
%prep
%setup -q -n %{name}-%{version}.src
%patch0 -p1 -b .s390
%patch1 -p1 -b .sphinx
%patch2 -p1 -b .docs2
%patch3 -p1 -b .docs3
%patch4 -p1 -b .docs4
%patch5 -p1 -b .lldbfix
%patch6 -p0 -b .doc-lit
%patch7 -p1 -b .amdfix
%patch8 -p2 -b .arm64
%patch47 -p1 -b .rust47
%patch53 -p1 -b .rust53
%patch54 -p1 -b .rust54
%patch55 -p1 -b .rust55
%patch57 -p1 -b .rust57
%patch1 -p1 -b .pthread-fix
%ifarch armv7hl
@ -211,8 +179,8 @@ fi
%if %{with gold}
%{_libdir}/LLVMgold.so
%endif
%{_libdir}/libLLVM-3.9*.so
%{_libdir}/libLTO.so
%{_libdir}/libLLVM-4.0*.so
%{_libdir}/libLTO.so*
%files devel
%{_bindir}/llvm-config-%{__isa_bits}
@ -229,6 +197,9 @@ fi
%{_libdir}/*.a
%changelog
* Thu Mar 23 2017 Tom Stellard <tstellar@redhat.com> - 4.0.0-1
- LLVM 4.0.0 Final Release
* Wed Mar 22 2017 tstellar@redhat.com - 3.9.1-6
- Fix %postun sep for -devel package.

View File

@ -1,13 +0,0 @@
diff --git a/tools/clang/lib/Driver/Tools.cpp b/tools/clang/lib/Driver/Tools.cpp
index 29713ed..0d23694 100644
--- a/tools/clang/lib/Driver/Tools.cpp
+++ b/tools/clang/lib/Driver/Tools.cpp
@@ -747,7 +747,7 @@ static StringRef getARMFloatABI(const Driver &D,
FloatABI = "hard";
break;
case llvm::Triple::GNUEABI:
- FloatABI = "softfp";
+ FloatABI = Triple.getVendorName() == "hardfloat" ? "hard" : "softfp";
break;
case llvm::Triple::EABI:
// EABI is always AAPCS, and if it was not marked 'hard', it's softfp

View File

@ -1,272 +0,0 @@
From ae32815f9281a5a8d48014e180901fcdb658285a Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Sun, 7 Aug 2016 07:58:00 +0000
Subject: [rust-lang/llvm#47 1/4] [InstCombine] Infer inbounds on geps of
allocas
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@277950 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Transforms/InstCombine/InstructionCombining.cpp | 19 +++++++++++++++++++
test/Transforms/InstCombine/getelementptr.ll | 6 +++---
2 files changed, 22 insertions(+), 3 deletions(-)
diff --git a/lib/Transforms/InstCombine/InstructionCombining.cpp b/lib/Transforms/InstCombine/InstructionCombining.cpp
index 377ccb9c37f7..31b5ad6ae8af 100644
--- a/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -1898,6 +1898,25 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
}
}
+ if (!GEP.isInBounds()) {
+ unsigned PtrWidth =
+ DL.getPointerSizeInBits(PtrOp->getType()->getPointerAddressSpace());
+ APInt BasePtrOffset(PtrWidth, 0);
+ Value *UnderlyingPtrOp =
+ PtrOp->stripAndAccumulateInBoundsConstantOffsets(DL,
+ BasePtrOffset);
+ if (auto *AI = dyn_cast<AllocaInst>(UnderlyingPtrOp)) {
+ if (GEP.accumulateConstantOffset(DL, BasePtrOffset) &&
+ BasePtrOffset.isNonNegative()) {
+ APInt AllocSize(PtrWidth, DL.getTypeAllocSize(AI->getAllocatedType()));
+ if (BasePtrOffset.ule(AllocSize)) {
+ return GetElementPtrInst::CreateInBounds(
+ PtrOp, makeArrayRef(Ops).slice(1), GEP.getName());
+ }
+ }
+ }
+ }
+
return nullptr;
}
diff --git a/test/Transforms/InstCombine/getelementptr.ll b/test/Transforms/InstCombine/getelementptr.ll
index 7446734e210c..14abd84fd18e 100644
--- a/test/Transforms/InstCombine/getelementptr.ll
+++ b/test/Transforms/InstCombine/getelementptr.ll
@@ -366,7 +366,7 @@ define i32 @test21() {
%rval = load i32, i32* %pbobel
ret i32 %rval
; CHECK-LABEL: @test21(
-; CHECK: getelementptr %intstruct, %intstruct* %pbob1, i64 0, i32 0
+; CHECK: getelementptr inbounds %intstruct, %intstruct* %pbob1, i64 0, i32 0
}
@@ -540,8 +540,8 @@ define i8* @test32(i8* %v) {
%G = load i8*, i8** %F
ret i8* %G
; CHECK-LABEL: @test32(
-; CHECK: %D = getelementptr [4 x i8*], [4 x i8*]* %A, i64 0, i64 1
-; CHECK: %F = getelementptr [4 x i8*], [4 x i8*]* %A, i64 0, i64 2
+; CHECK: %D = getelementptr inbounds [4 x i8*], [4 x i8*]* %A, i64 0, i64 1
+; CHECK: %F = getelementptr inbounds [4 x i8*], [4 x i8*]* %A, i64 0, i64 2
}
; PR3290
--
2.7.4
From d31c987130ff1bf9cea9a287195ecceda91c37d1 Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Sun, 7 Aug 2016 07:58:10 +0000
Subject: [rust-lang/llvm#47 2/4] [InstSimplify] Try hard to simplify pointer
comparisons
Simplify ptrtoint comparisons involving operands with different source
types.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@277951 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Analysis/InstructionSimplify.cpp | 10 ++++++++++
test/Transforms/InstSimplify/compare.ll | 13 +++++++++++++
2 files changed, 23 insertions(+)
diff --git a/lib/Analysis/InstructionSimplify.cpp b/lib/Analysis/InstructionSimplify.cpp
index 7c6edbfca270..8b70d89d62cb 100644
--- a/lib/Analysis/InstructionSimplify.cpp
+++ b/lib/Analysis/InstructionSimplify.cpp
@@ -3092,6 +3092,16 @@ static Value *SimplifyICmpInst(unsigned Predicate, Value *LHS, Value *RHS,
if (LHS->getType()->isPointerTy())
if (auto *C = computePointerICmp(Q.DL, Q.TLI, Q.DT, Pred, Q.CxtI, LHS, RHS))
return C;
+ if (auto *CLHS = dyn_cast<PtrToIntOperator>(LHS))
+ if (auto *CRHS = dyn_cast<PtrToIntOperator>(RHS))
+ if (Q.DL.getTypeSizeInBits(CLHS->getPointerOperandType()) ==
+ Q.DL.getTypeSizeInBits(CLHS->getType()) &&
+ Q.DL.getTypeSizeInBits(CRHS->getPointerOperandType()) ==
+ Q.DL.getTypeSizeInBits(CRHS->getType()))
+ if (auto *C = computePointerICmp(Q.DL, Q.TLI, Q.DT, Pred, Q.CxtI,
+ CLHS->getPointerOperand(),
+ CRHS->getPointerOperand()))
+ return C;
if (GetElementPtrInst *GLHS = dyn_cast<GetElementPtrInst>(LHS)) {
if (GEPOperator *GRHS = dyn_cast<GEPOperator>(RHS)) {
diff --git a/test/Transforms/InstSimplify/compare.ll b/test/Transforms/InstSimplify/compare.ll
index 9d6fd74ae56f..3e7316ec6b48 100644
--- a/test/Transforms/InstSimplify/compare.ll
+++ b/test/Transforms/InstSimplify/compare.ll
@@ -205,6 +205,19 @@ define i1 @gep16(i8* %ptr, i32 %a) {
; CHECK-NEXT: ret i1 false
}
+define i1 @gep17() {
+; CHECK-LABEL: @gep17(
+ %alloca = alloca i32, align 4
+ %bc = bitcast i32* %alloca to [4 x i8]*
+ %gep1 = getelementptr inbounds i32, i32* %alloca, i32 1
+ %pti1 = ptrtoint i32* %gep1 to i32
+ %gep2 = getelementptr inbounds [4 x i8], [4 x i8]* %bc, i32 0, i32 1
+ %pti2 = ptrtoint i8* %gep2 to i32
+ %cmp = icmp ugt i32 %pti1, %pti2
+ ret i1 %cmp
+; CHECK-NEXT: ret i1 true
+}
+
define i1 @zext(i32 %x) {
; CHECK-LABEL: @zext(
%e1 = zext i32 %x to i64
--
2.7.4
From bd3e05cb1f5293635edff14fcf23cfc73985c977 Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Sun, 7 Aug 2016 07:58:12 +0000
Subject: [rust-lang/llvm#47 3/4] [InstSimplify] Fold gep (gep V, C), (sub 0,
V) to C
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@277952 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Analysis/InstructionSimplify.cpp | 20 ++++++++++++++++++++
test/Transforms/InstSimplify/compare.ll | 13 +++++++++++++
2 files changed, 33 insertions(+)
diff --git a/lib/Analysis/InstructionSimplify.cpp b/lib/Analysis/InstructionSimplify.cpp
index 8b70d89d62cb..9d2a47957125 100644
--- a/lib/Analysis/InstructionSimplify.cpp
+++ b/lib/Analysis/InstructionSimplify.cpp
@@ -3597,6 +3597,26 @@ static Value *SimplifyGEPInst(Type *SrcTy, ArrayRef<Value *> Ops,
}
}
+ // gep (gep V, C), (sub 0, V) -> C
+ if (Q.DL.getTypeAllocSize(LastType) == 1 &&
+ all_of(Ops.slice(1).drop_back(1),
+ [](Value *Idx) { return match(Idx, m_Zero()); })) {
+ unsigned PtrWidth =
+ Q.DL.getPointerSizeInBits(Ops[0]->getType()->getPointerAddressSpace());
+ if (Q.DL.getTypeSizeInBits(Ops.back()->getType()) == PtrWidth) {
+ APInt BasePtrOffset(PtrWidth, 0);
+ Value *StrippedBasePtr =
+ Ops[0]->stripAndAccumulateInBoundsConstantOffsets(Q.DL,
+ BasePtrOffset);
+
+ if (match(Ops.back(),
+ m_Sub(m_Zero(), m_PtrToInt(m_Specific(StrippedBasePtr))))) {
+ auto *CI = ConstantInt::get(GEPTy->getContext(), BasePtrOffset);
+ return ConstantExpr::getIntToPtr(CI, GEPTy);
+ }
+ }
+ }
+
// Check to see if this is constant foldable.
for (unsigned i = 0, e = Ops.size(); i != e; ++i)
if (!isa<Constant>(Ops[i]))
diff --git a/test/Transforms/InstSimplify/compare.ll b/test/Transforms/InstSimplify/compare.ll
index 3e7316ec6b48..addb63c57222 100644
--- a/test/Transforms/InstSimplify/compare.ll
+++ b/test/Transforms/InstSimplify/compare.ll
@@ -218,6 +218,19 @@ define i1 @gep17() {
; CHECK-NEXT: ret i1 true
}
+define i32 @gep18() {
+; CHECK-LABEL: @gep18(
+ %alloca = alloca i32, align 4 ; alloca + 0
+ %gep = getelementptr inbounds i32, i32* %alloca, i32 1 ; alloca + 4
+ %bc = bitcast i32* %gep to [4 x i8]* ; alloca + 4
+ %pti = ptrtoint i32* %alloca to i32 ; alloca
+ %sub = sub i32 0, %pti ; -alloca
+ %add = getelementptr [4 x i8], [4 x i8]* %bc, i32 0, i32 %sub ; alloca + 4 - alloca == 4
+ %add_to_int = ptrtoint i8* %add to i32 ; 4
+ ret i32 %add_to_int ; 4
+; CHECK-NEXT: ret i32 4
+}
+
define i1 @zext(i32 %x) {
; CHECK-LABEL: @zext(
%e1 = zext i32 %x to i64
--
2.7.4
From c3eb3c7608f439231d0c1340af6b720f113b4bf4 Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Tue, 16 Aug 2016 06:13:46 +0000
Subject: [rust-lang/llvm#47 4/4] [InstSimplify] Fold gep (gep V, C), (xor V,
-1) to C-1
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@278779 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Analysis/InstructionSimplify.cpp | 8 +++++++-
test/Transforms/InstSimplify/compare.ll | 13 -------------
2 files changed, 7 insertions(+), 14 deletions(-)
diff --git a/lib/Analysis/InstructionSimplify.cpp b/lib/Analysis/InstructionSimplify.cpp
index 9d2a47957125..f7a435d1ad46 100644
--- a/lib/Analysis/InstructionSimplify.cpp
+++ b/lib/Analysis/InstructionSimplify.cpp
@@ -3597,7 +3597,6 @@ static Value *SimplifyGEPInst(Type *SrcTy, ArrayRef<Value *> Ops,
}
}
- // gep (gep V, C), (sub 0, V) -> C
if (Q.DL.getTypeAllocSize(LastType) == 1 &&
all_of(Ops.slice(1).drop_back(1),
[](Value *Idx) { return match(Idx, m_Zero()); })) {
@@ -3609,11 +3608,18 @@ static Value *SimplifyGEPInst(Type *SrcTy, ArrayRef<Value *> Ops,
Ops[0]->stripAndAccumulateInBoundsConstantOffsets(Q.DL,
BasePtrOffset);
+ // gep (gep V, C), (sub 0, V) -> C
if (match(Ops.back(),
m_Sub(m_Zero(), m_PtrToInt(m_Specific(StrippedBasePtr))))) {
auto *CI = ConstantInt::get(GEPTy->getContext(), BasePtrOffset);
return ConstantExpr::getIntToPtr(CI, GEPTy);
}
+ // gep (gep V, C), (xor V, -1) -> C-1
+ if (match(Ops.back(),
+ m_Xor(m_PtrToInt(m_Specific(StrippedBasePtr)), m_AllOnes()))) {
+ auto *CI = ConstantInt::get(GEPTy->getContext(), BasePtrOffset - 1);
+ return ConstantExpr::getIntToPtr(CI, GEPTy);
+ }
}
}
diff --git a/test/Transforms/InstSimplify/compare.ll b/test/Transforms/InstSimplify/compare.ll
index addb63c57222..3e7316ec6b48 100644
--- a/test/Transforms/InstSimplify/compare.ll
+++ b/test/Transforms/InstSimplify/compare.ll
@@ -218,19 +218,6 @@ define i1 @gep17() {
; CHECK-NEXT: ret i1 true
}
-define i32 @gep18() {
-; CHECK-LABEL: @gep18(
- %alloca = alloca i32, align 4 ; alloca + 0
- %gep = getelementptr inbounds i32, i32* %alloca, i32 1 ; alloca + 4
- %bc = bitcast i32* %gep to [4 x i8]* ; alloca + 4
- %pti = ptrtoint i32* %alloca to i32 ; alloca
- %sub = sub i32 0, %pti ; -alloca
- %add = getelementptr [4 x i8], [4 x i8]* %bc, i32 0, i32 %sub ; alloca + 4 - alloca == 4
- %add_to_int = ptrtoint i8* %add to i32 ; 4
- ret i32 %add_to_int ; 4
-; CHECK-NEXT: ret i32 4
-}
-
define i1 @zext(i32 %x) {
; CHECK-LABEL: @zext(
%e1 = zext i32 %x to i64
--
2.7.4

View File

@ -1,883 +0,0 @@
From 0d4331af6e01235479e44f9775b3fde9e19200a3 Mon Sep 17 00:00:00 2001
From: Keith Walker <kwalker@arm.com>
Date: Tue, 27 Sep 2016 16:46:07 +0000
Subject: [rust-lang/llvm#53 1/2] Propagate DBG_VALUE entries when there are
unvisited predecessors
Variables are sometimes missing their debug location information in
blocks in which the variables should be available. This would occur
when one or more predecessor blocks had not yet been visited by the
routine which propagated the information from predecessor blocks.
This is addressed by only considering predecessor blocks which have
already been visited.
The solution to this problem was suggested by Daniel Berlin on the
LLVM developer mailing list.
Differential Revision: https://reviews.llvm.org/D24927
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@282506 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/CodeGen/LiveDebugValues.cpp | 34 ++--
test/CodeGen/ARM/dbg-range-extension.mir | 282 +++++++++++++++++++++++++++++++
2 files changed, 306 insertions(+), 10 deletions(-)
create mode 100644 test/CodeGen/ARM/dbg-range-extension.mir
diff --git a/lib/CodeGen/LiveDebugValues.cpp b/lib/CodeGen/LiveDebugValues.cpp
index 4ff88d528108..4cadd5855ed5 100644
--- a/lib/CodeGen/LiveDebugValues.cpp
+++ b/lib/CodeGen/LiveDebugValues.cpp
@@ -201,7 +201,8 @@ private:
VarLocInMBB &OutLocs, VarLocMap &VarLocIDs);
bool join(MachineBasicBlock &MBB, VarLocInMBB &OutLocs, VarLocInMBB &InLocs,
- const VarLocMap &VarLocIDs);
+ const VarLocMap &VarLocIDs,
+ SmallPtrSet<const MachineBasicBlock *, 16> &Visited);
bool ExtendRanges(MachineFunction &MF);
@@ -368,7 +369,8 @@ bool LiveDebugValues::transfer(MachineInstr &MI, OpenRangesSet &OpenRanges,
/// inserting a new DBG_VALUE instruction at the start of the @MBB - if the same
/// source variable in all the predecessors of @MBB reside in the same location.
bool LiveDebugValues::join(MachineBasicBlock &MBB, VarLocInMBB &OutLocs,
- VarLocInMBB &InLocs, const VarLocMap &VarLocIDs) {
+ VarLocInMBB &InLocs, const VarLocMap &VarLocIDs,
+ SmallPtrSet<const MachineBasicBlock *, 16> &Visited) {
DEBUG(dbgs() << "join MBB: " << MBB.getName() << "\n");
bool Changed = false;
@@ -376,21 +378,32 @@ bool LiveDebugValues::join(MachineBasicBlock &MBB, VarLocInMBB &OutLocs,
// For all predecessors of this MBB, find the set of VarLocs that
// can be joined.
+ int NumVisited = 0;
for (auto p : MBB.predecessors()) {
+ // Ignore unvisited predecessor blocks. As we are processing
+ // the blocks in reverse post-order any unvisited block can
+ // be considered to not remove any incoming values.
+ if (!Visited.count(p))
+ continue;
auto OL = OutLocs.find(p);
// Join is null in case of empty OutLocs from any of the pred.
if (OL == OutLocs.end())
return false;
- // Just copy over the Out locs to incoming locs for the first predecessor.
- if (p == *MBB.pred_begin()) {
+ // Just copy over the Out locs to incoming locs for the first visited
+ // predecessor, and for all other predecessors join the Out locs.
+ if (!NumVisited)
InLocsT = OL->second;
- continue;
- }
- // Join with this predecessor.
- InLocsT &= OL->second;
+ else
+ InLocsT &= OL->second;
+ NumVisited++;
}
+ // As we are processing blocks in reverse post-order we
+ // should have processed at least one predecessor, unless it
+ // is the entry block which has no predecessor.
+ assert((NumVisited || MBB.pred_empty()) &&
+ "Should have processed at least one predecessor");
if (InLocsT.empty())
return false;
@@ -463,6 +476,7 @@ bool LiveDebugValues::ExtendRanges(MachineFunction &MF) {
// To solve it, we perform join() and transfer() using the two worklist method
// until the ranges converge.
// Ranges have converged when both worklists are empty.
+ SmallPtrSet<const MachineBasicBlock *, 16> Visited;
while (!Worklist.empty() || !Pending.empty()) {
// We track what is on the pending worklist to avoid inserting the same
// thing twice. We could avoid this with a custom priority queue, but this
@@ -471,8 +485,8 @@ bool LiveDebugValues::ExtendRanges(MachineFunction &MF) {
while (!Worklist.empty()) {
MachineBasicBlock *MBB = OrderToBB[Worklist.top()];
Worklist.pop();
- MBBJoined = join(*MBB, OutLocs, InLocs, VarLocIDs);
-
+ MBBJoined = join(*MBB, OutLocs, InLocs, VarLocIDs, Visited);
+ Visited.insert(MBB);
if (MBBJoined) {
MBBJoined = false;
Changed = true;
diff --git a/test/CodeGen/ARM/dbg-range-extension.mir b/test/CodeGen/ARM/dbg-range-extension.mir
new file mode 100644
index 000000000000..b18f56197948
--- /dev/null
+++ b/test/CodeGen/ARM/dbg-range-extension.mir
@@ -0,0 +1,282 @@
+# RUN: llc -mtriple=arm-eabi -run-pass=livedebugvalues %s -o - | FileCheck %s
+#
+# Check that the debug information for variables are propagated into the correct blocks.
+#
+# Generated from the C source:
+#
+# int func2(int, int);
+# void func(int a) {
+# int b = func2(10, 11);
+# if (a) {
+# int c = func2(12, 13);
+# for(int i = 1; i < a; i++) {
+# func2(i, i+b);
+# }
+# func2(b,c);
+# }
+# func2(b,a);
+# }
+
+# CHECK: [[VAR_A:![0-9]+]] = !DILocalVariable(name: "a",
+# CHECK: [[VAR_B:![0-9]+]] = !DILocalVariable(name: "b",
+# CHECK: [[VAR_C:![0-9]+]] = !DILocalVariable(name: "c",
+# CHECK: [[VAR_I:![0-9]+]] = !DILocalVariable(name: "i",
+
+# CHECK: bb.0.entry
+# CHECK: DBG_VALUE debug-use %r0, debug-use _, [[VAR_A]]
+# CHECK: DBG_VALUE debug-use [[REG_A:%r[0-9]+]], debug-use _, [[VAR_A]]
+# CHECK: DBG_VALUE debug-use [[REG_B:%r[0-9]+]], debug-use _, [[VAR_B]]
+
+# CHECK: bb.1.if.then
+# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
+# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
+# CHECK: DBG_VALUE debug-use [[REG_C:%r[0-9]+]], debug-use _, [[VAR_C]]
+# CHECK: DBG_VALUE 1, 0, [[VAR_I]]
+
+# CHECK: bb.2.for.body
+# CHECK: DBG_VALUE debug-use [[REG_I:%r[0-9]+]], debug-use _, [[VAR_I]]
+# CHECK: DBG_VALUE debug-use [[REG_C]], debug-use _, [[VAR_C]]
+# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
+# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
+# CHECK: DBG_VALUE debug-use [[REG_I]], debug-use _, [[VAR_I]]
+
+# CHECK: bb.3.for.cond
+# CHECK: DBG_VALUE debug-use [[REG_C]], debug-use _, [[VAR_C]]
+# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
+# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
+# CHECK: DBG_VALUE debug-use [[REG_I]], debug-use _, [[VAR_I]]
+
+# CHECK: bb.4.for.cond.cleanup
+# CHECK: DBG_VALUE debug-use [[REG_I]], debug-use _, [[VAR_I]]
+# CHECK: DBG_VALUE debug-use [[REG_C]], debug-use _, [[VAR_C]]
+# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
+# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
+
+# CHECK: bb.5.if.end
+# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
+# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
+--- |
+ ; ModuleID = '/data/kwalker/work/OpenSource-llvm/llvm/test/CodeGen/ARM/dbg-range-extension.ll'
+ source_filename = "/data/kwalker/work/OpenSource-llvm/llvm/test/CodeGen/ARM/dbg-range-extension.ll"
+ target datalayout = "e-m:e-p:32:32-i64:64-v128:64:128-a:0:32-n32-S64"
+ target triple = "arm---eabi"
+
+ ; Function Attrs: minsize nounwind optsize
+ define void @func(i32 %a) local_unnamed_addr #0 !dbg !8 {
+ entry:
+ tail call void @llvm.dbg.value(metadata i32 %a, i64 0, metadata !13, metadata !20), !dbg !21
+ %call = tail call i32 @func2(i32 10, i32 11) #0, !dbg !22
+ tail call void @llvm.dbg.value(metadata i32 %call, i64 0, metadata !14, metadata !20), !dbg !23
+ %tobool = icmp eq i32 %a, 0, !dbg !24
+ br i1 %tobool, label %if.end, label %if.then, !dbg !25
+
+ if.then: ; preds = %entry
+ %call1 = tail call i32 @func2(i32 12, i32 13) #0, !dbg !26
+ tail call void @llvm.dbg.value(metadata i32 %call1, i64 0, metadata !15, metadata !20), !dbg !27
+ tail call void @llvm.dbg.value(metadata i32 1, i64 0, metadata !18, metadata !20), !dbg !28
+ br label %for.cond, !dbg !29
+
+ for.cond: ; preds = %for.body, %if.then
+ %i.0 = phi i32 [ 1, %if.then ], [ %inc, %for.body ]
+ tail call void @llvm.dbg.value(metadata i32 %i.0, i64 0, metadata !18, metadata !20), !dbg !28
+ %cmp = icmp slt i32 %i.0, %a, !dbg !30
+ br i1 %cmp, label %for.body, label %for.cond.cleanup, !dbg !33
+
+ for.cond.cleanup: ; preds = %for.cond
+ %call3 = tail call i32 @func2(i32 %call, i32 %call1) #0, !dbg !34
+ br label %if.end, !dbg !35
+
+ for.body: ; preds = %for.cond
+ %0 = add i32 %call, %i.0, !dbg !36
+ %call2 = tail call i32 @func2(i32 %i.0, i32 %0) #0, !dbg !36
+ %inc = add nuw nsw i32 %i.0, 1, !dbg !38
+ tail call void @llvm.dbg.value(metadata i32 %inc, i64 0, metadata !18, metadata !20), !dbg !28
+ br label %for.cond, !dbg !40, !llvm.loop !41
+
+ if.end: ; preds = %for.cond.cleanup, %entry
+ %call4 = tail call i32 @func2(i32 %call, i32 %a) #0, !dbg !43
+ ret void, !dbg !44
+ }
+
+ ; Function Attrs: minsize optsize
+ declare i32 @func2(i32, i32) local_unnamed_addr #1
+
+ ; Function Attrs: nounwind readnone
+ declare void @llvm.dbg.value(metadata, i64, metadata, metadata) #2
+
+ ; Function Attrs: nounwind
+ declare void @llvm.stackprotector(i8*, i8**) #3
+
+ attributes #0 = { minsize nounwind optsize }
+ attributes #1 = { minsize optsize }
+ attributes #2 = { nounwind readnone }
+ attributes #3 = { nounwind }
+
+ !llvm.dbg.cu = !{!0}
+ !llvm.module.flags = !{!3, !4, !5, !6}
+ !llvm.ident = !{!7}
+
+ !0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2)
+ !1 = !DIFile(filename: "loop.c", directory: "/tmp")
+ !2 = !{}
+ !3 = !{i32 2, !"Dwarf Version", i32 4}
+ !4 = !{i32 2, !"Debug Info Version", i32 3}
+ !5 = !{i32 1, !"wchar_size", i32 4}
+ !6 = !{i32 1, !"min_enum_size", i32 4}
+ !7 = !{!"clang version 4.0.0 (http://llvm.org/git/clang.git b8f10df3679b36f51e1de7c4351b82d297825089) (http://llvm.org/git/llvm.git c2a5d16d1e3b8c49f5bbb1ff87a76ac4f88edb89)"}
+ !8 = distinct !DISubprogram(name: "func", scope: !1, file: !1, line: 2, type: !9, isLocal: false, isDefinition: true, scopeLine: 2, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !12)
+ !9 = !DISubroutineType(types: !10)
+ !10 = !{null, !11}
+ !11 = !DIBasicType(name: "int", size: 32, align: 32, encoding: DW_ATE_signed)
+ !12 = !{!13, !14, !15, !18}
+ !13 = !DILocalVariable(name: "a", arg: 1, scope: !8, file: !1, line: 2, type: !11)
+ !14 = !DILocalVariable(name: "b", scope: !8, file: !1, line: 3, type: !11)
+ !15 = !DILocalVariable(name: "c", scope: !16, file: !1, line: 5, type: !11)
+ !16 = distinct !DILexicalBlock(scope: !17, file: !1, line: 4, column: 9)
+ !17 = distinct !DILexicalBlock(scope: !8, file: !1, line: 4, column: 6)
+ !18 = !DILocalVariable(name: "i", scope: !19, file: !1, line: 6, type: !11)
+ !19 = distinct !DILexicalBlock(scope: !16, file: !1, line: 6, column: 3)
+ !20 = !DIExpression()
+ !21 = !DILocation(line: 2, column: 15, scope: !8)
+ !22 = !DILocation(line: 3, column: 17, scope: !8)
+ !23 = !DILocation(line: 3, column: 13, scope: !8)
+ !24 = !DILocation(line: 4, column: 6, scope: !17)
+ !25 = !DILocation(line: 4, column: 6, scope: !8)
+ !26 = !DILocation(line: 5, column: 11, scope: !16)
+ !27 = !DILocation(line: 5, column: 7, scope: !16)
+ !28 = !DILocation(line: 6, column: 11, scope: !19)
+ !29 = !DILocation(line: 6, column: 7, scope: !19)
+ !30 = !DILocation(line: 6, column: 20, scope: !31)
+ !31 = !DILexicalBlockFile(scope: !32, file: !1, discriminator: 1)
+ !32 = distinct !DILexicalBlock(scope: !19, file: !1, line: 6, column: 3)
+ !33 = !DILocation(line: 6, column: 3, scope: !31)
+ !34 = !DILocation(line: 9, column: 3, scope: !16)
+ !35 = !DILocation(line: 10, column: 2, scope: !16)
+ !36 = !DILocation(line: 7, column: 4, scope: !37)
+ !37 = distinct !DILexicalBlock(scope: !32, file: !1, line: 6, column: 30)
+ !38 = !DILocation(line: 6, column: 26, scope: !39)
+ !39 = !DILexicalBlockFile(scope: !32, file: !1, discriminator: 3)
+ !40 = !DILocation(line: 6, column: 3, scope: !39)
+ !41 = distinct !{!41, !42}
+ !42 = !DILocation(line: 6, column: 3, scope: !16)
+ !43 = !DILocation(line: 11, column: 2, scope: !8)
+ !44 = !DILocation(line: 12, column: 1, scope: !8)
+
+...
+---
+name: func
+alignment: 2
+exposesReturnsTwice: false
+legalized: false
+regBankSelected: false
+selected: false
+tracksRegLiveness: false
+liveins:
+ - { reg: '%r0' }
+calleeSavedRegisters: [ '%lr', '%d8', '%d9', '%d10', '%d11', '%d12', '%d13',
+ '%d14', '%d15', '%q4', '%q5', '%q6', '%q7', '%r4',
+ '%r5', '%r6', '%r7', '%r8', '%r9', '%r10', '%r11',
+ '%s16', '%s17', '%s18', '%s19', '%s20', '%s21',
+ '%s22', '%s23', '%s24', '%s25', '%s26', '%s27',
+ '%s28', '%s29', '%s30', '%s31', '%d8_d10', '%d9_d11',
+ '%d10_d12', '%d11_d13', '%d12_d14', '%d13_d15',
+ '%q4_q5', '%q5_q6', '%q6_q7', '%q4_q5_q6_q7', '%r4_r5',
+ '%r6_r7', '%r8_r9', '%r10_r11', '%d8_d9_d10', '%d9_d10_d11',
+ '%d10_d11_d12', '%d11_d12_d13', '%d12_d13_d14',
+ '%d13_d14_d15', '%d8_d10_d12', '%d9_d11_d13', '%d10_d12_d14',
+ '%d11_d13_d15', '%d8_d10_d12_d14', '%d9_d11_d13_d15',
+ '%d9_d10', '%d11_d12', '%d13_d14', '%d9_d10_d11_d12',
+ '%d11_d12_d13_d14' ]
+frameInfo:
+ isFrameAddressTaken: false
+ isReturnAddressTaken: false
+ hasStackMap: false
+ hasPatchPoint: false
+ stackSize: 24
+ offsetAdjustment: 0
+ maxAlignment: 4
+ adjustsStack: true
+ hasCalls: true
+ maxCallFrameSize: 0
+ hasOpaqueSPAdjustment: false
+ hasVAStart: false
+ hasMustTailInVarArgFunc: false
+stack:
+ - { id: 0, type: spill-slot, offset: -4, size: 4, alignment: 4, callee-saved-register: '%lr' }
+ - { id: 1, type: spill-slot, offset: -8, size: 4, alignment: 4, callee-saved-register: '%r11' }
+ - { id: 2, type: spill-slot, offset: -12, size: 4, alignment: 4, callee-saved-register: '%r7' }
+ - { id: 3, type: spill-slot, offset: -16, size: 4, alignment: 4, callee-saved-register: '%r6' }
+ - { id: 4, type: spill-slot, offset: -20, size: 4, alignment: 4, callee-saved-register: '%r5' }
+ - { id: 5, type: spill-slot, offset: -24, size: 4, alignment: 4, callee-saved-register: '%r4' }
+body: |
+ bb.0.entry:
+ successors: %bb.5.if.end, %bb.1.if.then
+ liveins: %r0, %r4, %r5, %r6, %r7, %r11, %lr
+
+ %sp = frame-setup STMDB_UPD %sp, 14, _, killed %r4, killed %r5, killed %r6, killed %r7, killed %r11, killed %lr
+ frame-setup CFI_INSTRUCTION def_cfa_offset 24
+ frame-setup CFI_INSTRUCTION offset %lr, -4
+ frame-setup CFI_INSTRUCTION offset %r11, -8
+ frame-setup CFI_INSTRUCTION offset %r7, -12
+ frame-setup CFI_INSTRUCTION offset %r6, -16
+ frame-setup CFI_INSTRUCTION offset %r5, -20
+ frame-setup CFI_INSTRUCTION offset %r4, -24
+ DBG_VALUE debug-use %r0, debug-use _, !13, !20, debug-location !21
+ %r4 = MOVr killed %r0, 14, _, _
+ DBG_VALUE debug-use %r4, debug-use _, !13, !20, debug-location !21
+ %r0 = MOVi 10, 14, _, _, debug-location !22
+ %r1 = MOVi 11, 14, _, _, debug-location !22
+ BL @func2, csr_aapcs, implicit-def dead %lr, implicit %sp, implicit killed %r0, implicit killed %r1, implicit-def %sp, implicit-def %r0, debug-location !22
+ %r5 = MOVr killed %r0, 14, _, _, debug-location !22
+ DBG_VALUE debug-use %r5, debug-use _, !14, !20, debug-location !23
+ CMPri %r4, 0, 14, _, implicit-def %cpsr, debug-location !25
+ Bcc %bb.5.if.end, 0, killed %cpsr
+
+ bb.1.if.then:
+ successors: %bb.3.for.cond
+ liveins: %r4, %r5
+
+ %r0 = MOVi 12, 14, _, _, debug-location !26
+ %r1 = MOVi 13, 14, _, _, debug-location !26
+ BL @func2, csr_aapcs, implicit-def dead %lr, implicit %sp, implicit killed %r0, implicit killed %r1, implicit-def %sp, implicit-def %r0, debug-location !26
+ %r6 = MOVr killed %r0, 14, _, _, debug-location !26
+ DBG_VALUE debug-use %r6, debug-use _, !15, !20, debug-location !27
+ %r7 = MOVi 1, 14, _, _
+ DBG_VALUE 1, 0, !18, !20, debug-location !28
+ B %bb.3.for.cond
+
+ bb.2.for.body:
+ successors: %bb.3.for.cond
+ liveins: %r4, %r5, %r6, %r7
+
+ %r1 = ADDrr %r5, %r7, 14, _, _, debug-location !36
+ %r0 = MOVr %r7, 14, _, _, debug-location !36
+ BL @func2, csr_aapcs, implicit-def dead %lr, implicit %sp, implicit killed %r0, implicit killed %r1, implicit-def %sp, implicit-def dead %r0, debug-location !36
+ %r7 = ADDri killed %r7, 1, 14, _, _, debug-location !38
+ DBG_VALUE debug-use %r7, debug-use _, !18, !20, debug-location !28
+
+ bb.3.for.cond:
+ successors: %bb.2.for.body, %bb.4.for.cond.cleanup
+ liveins: %r4, %r5, %r6, %r7
+
+ DBG_VALUE debug-use %r7, debug-use _, !18, !20, debug-location !28
+ CMPrr %r7, %r4, 14, _, implicit-def %cpsr, debug-location !33
+ Bcc %bb.2.for.body, 11, killed %cpsr, debug-location !33
+
+ bb.4.for.cond.cleanup:
+ successors: %bb.5.if.end
+ liveins: %r4, %r5, %r6
+
+ %r0 = MOVr %r5, 14, _, _, debug-location !34
+ %r1 = MOVr killed %r6, 14, _, _, debug-location !34
+ BL @func2, csr_aapcs, implicit-def dead %lr, implicit %sp, implicit killed %r0, implicit killed %r1, implicit-def %sp, implicit-def dead %r0, debug-location !34
+
+ bb.5.if.end:
+ liveins: %r4, %r5
+
+ %r0 = MOVr killed %r5, 14, _, _, debug-location !43
+ %r1 = MOVr killed %r4, 14, _, _, debug-location !43
+ %sp = LDMIA_UPD %sp, 14, _, def %r4, def %r5, def %r6, def %r7, def %r11, def %lr, debug-location !43
+ TAILJMPd @func2, implicit %sp, implicit %sp, implicit killed %r0, implicit killed %r1, debug-location !43
+
+...
--
2.7.4
From 8a0fc26559123bb6eab3ceae93d5a2c94943614b Mon Sep 17 00:00:00 2001
From: Adrian Prantl <aprantl@apple.com>
Date: Wed, 28 Sep 2016 17:51:14 +0000
Subject: [rust-lang/llvm#53 2/2] Teach LiveDebugValues about lexical scopes.
This addresses PR26055 LiveDebugValues is very slow.
Contrary to the old LiveDebugVariables pass LiveDebugValues currently
doesn't look at the lexical scopes before inserting a DBG_VALUE
intrinsic. This means that we often propagate DBG_VALUEs much further
down than necessary. This is especially noticeable in large C++
functions with many inlined method calls that all use the same
"this"-pointer.
For example, in the following code it makes no sense to propagate the
inlined variable a from the first inlined call to f() into any of the
subsequent basic blocks, because the variable will always be out of
scope:
void sink(int a);
void __attribute((always_inline)) f(int a) { sink(a); }
void foo(int i) {
f(i);
if (i)
f(i);
f(i);
}
This patch reuses the LexicalScopes infrastructure we have for
LiveDebugVariables to take this into account.
The effect on compile time and memory consumption is quite noticeable:
I tested a benchmark that is a large C++ source with an enormous
amount of inlined "this"-pointers that would previously eat >24GiB
(most of them for DBG_VALUE intrinsics) and whose compile time was
dominated by LiveDebugValues. With this patch applied the memory
consumption is 1GiB and 1.7% of the time is spent in LiveDebugValues.
https://reviews.llvm.org/D24994
Thanks to Daniel Berlin and Keith Walker for reviewing!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@282611 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/CodeGen/LiveDebugValues.cpp | 51 ++++-
test/CodeGen/ARM/dbg-range-extension.mir | 1 -
test/DebugInfo/COFF/register-variables.ll | 8 +-
test/DebugInfo/MIR/X86/livedebugvalues-limit.mir | 228 +++++++++++++++++++++++
test/DebugInfo/X86/fission-ranges.ll | 6 +-
5 files changed, 278 insertions(+), 16 deletions(-)
create mode 100644 test/DebugInfo/MIR/X86/livedebugvalues-limit.mir
diff --git a/lib/CodeGen/LiveDebugValues.cpp b/lib/CodeGen/LiveDebugValues.cpp
index 4cadd5855ed5..969944eb24a2 100644
--- a/lib/CodeGen/LiveDebugValues.cpp
+++ b/lib/CodeGen/LiveDebugValues.cpp
@@ -23,6 +23,7 @@
#include "llvm/ADT/SparseBitVector.h"
#include "llvm/ADT/Statistic.h"
#include "llvm/ADT/UniqueVector.h"
+#include "llvm/CodeGen/LexicalScopes.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
#include "llvm/CodeGen/MachineInstrBuilder.h"
@@ -60,6 +61,26 @@ class LiveDebugValues : public MachineFunctionPass {
private:
const TargetRegisterInfo *TRI;
const TargetInstrInfo *TII;
+ LexicalScopes LS;
+
+ /// Keeps track of lexical scopes associated with a user value's source
+ /// location.
+ class UserValueScopes {
+ DebugLoc DL;
+ LexicalScopes &LS;
+ SmallPtrSet<const MachineBasicBlock *, 4> LBlocks;
+
+ public:
+ UserValueScopes(DebugLoc D, LexicalScopes &L) : DL(std::move(D)), LS(L) {}
+
+ /// Return true if current scope dominates at least one machine
+ /// instruction in a given machine basic block.
+ bool dominates(MachineBasicBlock *MBB) {
+ if (LBlocks.empty())
+ LS.getMachineBasicBlocks(DL, LBlocks);
+ return LBlocks.count(MBB) != 0 || LS.dominates(DL, MBB);
+ }
+ };
/// Based on std::pair so it can be used as an index into a DenseMap.
typedef std::pair<const DILocalVariable *, const DILocation *>
@@ -83,7 +104,7 @@ private:
struct VarLoc {
const DebugVariable Var;
const MachineInstr &MI; ///< Only used for cloning a new DBG_VALUE.
-
+ mutable UserValueScopes UVS;
enum { InvalidKind = 0, RegisterKind } Kind;
/// The value location. Stored separately to avoid repeatedly
@@ -96,9 +117,9 @@ private:
uint64_t Hash;
} Loc;
- VarLoc(const MachineInstr &MI)
+ VarLoc(const MachineInstr &MI, LexicalScopes &LS)
: Var(MI.getDebugVariable(), MI.getDebugLoc()->getInlinedAt()), MI(MI),
- Kind(InvalidKind) {
+ UVS(MI.getDebugLoc(), LS), Kind(InvalidKind) {
static_assert((sizeof(Loc) == sizeof(uint64_t)),
"hash does not cover all members of Loc");
assert(MI.isDebugValue() && "not a DBG_VALUE");
@@ -125,6 +146,10 @@ private:
return 0;
}
+ /// Determine whether the lexical scope of this value's debug location
+ /// dominates MBB.
+ bool dominates(MachineBasicBlock &MBB) const { return UVS.dominates(&MBB); }
+
void dump() const { MI.dump(); }
bool operator==(const VarLoc &Other) const {
@@ -229,6 +254,7 @@ public:
/// Calculate the liveness information for the given machine function.
bool runOnMachineFunction(MachineFunction &MF) override;
};
+
} // namespace
//===----------------------------------------------------------------------===//
@@ -295,7 +321,7 @@ void LiveDebugValues::transferDebugValue(const MachineInstr &MI,
// Add the VarLoc to OpenRanges from this DBG_VALUE.
// TODO: Currently handles DBG_VALUE which has only reg as location.
if (isDbgValueDescribedByReg(MI)) {
- VarLoc VL(MI);
+ VarLoc VL(MI, LS);
unsigned ID = VarLocIDs.insert(VL);
OpenRanges.insert(ID, VL.Var);
}
@@ -399,6 +425,13 @@ bool LiveDebugValues::join(MachineBasicBlock &MBB, VarLocInMBB &OutLocs,
NumVisited++;
}
+ // Filter out DBG_VALUES that are out of scope.
+ VarLocSet KillSet;
+ for (auto ID : InLocsT)
+ if (!VarLocIDs[ID].dominates(MBB))
+ KillSet.set(ID);
+ InLocsT.intersectWithComplement(KillSet);
+
// As we are processing blocks in reverse post-order we
// should have processed at least one predecessor, unless it
// is the entry block which has no predecessor.
@@ -519,12 +552,14 @@ bool LiveDebugValues::ExtendRanges(MachineFunction &MF) {
}
bool LiveDebugValues::runOnMachineFunction(MachineFunction &MF) {
+ if (!MF.getFunction()->getSubprogram())
+ // LiveDebugValues will already have removed all DBG_VALUEs.
+ return false;
+
TRI = MF.getSubtarget().getRegisterInfo();
TII = MF.getSubtarget().getInstrInfo();
+ LS.initialize(MF);
- bool Changed = false;
-
- Changed |= ExtendRanges(MF);
-
+ bool Changed = ExtendRanges(MF);
return Changed;
}
diff --git a/test/CodeGen/ARM/dbg-range-extension.mir b/test/CodeGen/ARM/dbg-range-extension.mir
index b18f56197948..466f69396948 100644
--- a/test/CodeGen/ARM/dbg-range-extension.mir
+++ b/test/CodeGen/ARM/dbg-range-extension.mir
@@ -47,7 +47,6 @@
# CHECK: DBG_VALUE debug-use [[REG_I]], debug-use _, [[VAR_I]]
# CHECK: bb.4.for.cond.cleanup
-# CHECK: DBG_VALUE debug-use [[REG_I]], debug-use _, [[VAR_I]]
# CHECK: DBG_VALUE debug-use [[REG_C]], debug-use _, [[VAR_C]]
# CHECK: DBG_VALUE debug-use [[REG_B]], debug-use _, [[VAR_B]]
# CHECK: DBG_VALUE debug-use [[REG_A]], debug-use _, [[VAR_A]]
diff --git a/test/DebugInfo/COFF/register-variables.ll b/test/DebugInfo/COFF/register-variables.ll
index 9bb782853a3d..08246fef9603 100644
--- a/test/DebugInfo/COFF/register-variables.ll
+++ b/test/DebugInfo/COFF/register-variables.ll
@@ -37,8 +37,8 @@
; ASM: #DEBUG_VALUE: c <- %EAX
; ASM: testl %esi, %esi
; ASM: je .LBB0_2
+; ASM: [[after_je:\.Ltmp.*]]:
; ASM: # BB#1: # %if.then
-; ASM-DAG: #DEBUG_VALUE: c <- %EAX
; ASM-DAG: #DEBUG_VALUE: inlineinc:a <- %EAX
; ASM-DAG: #DEBUG_VALUE: a <- %EAX
; ASM-DAG: #DEBUG_VALUE: f:p <- %ESI
@@ -65,7 +65,7 @@
; ASM: .cv_def_range [[after_getint]] [[after_inc_eax]], "A\021\021\000\000\000"
; ASM: .short 4414 # Record kind: S_LOCAL
; ASM: .asciz "c"
-; ASM: .cv_def_range [[after_getint]] [[after_inc_eax]], "A\021\021\000\000\000"
+; ASM: .cv_def_range [[after_getint]] [[after_je]], "A\021\021\000\000\000"
; ASM: .short 4414 # Record kind: S_LOCAL
; ASM: .asciz "b"
; ASM: .cv_def_range [[after_inc_eax]] [[after_if]], "A\021\021\000\000\000"
@@ -132,7 +132,7 @@
; OBJ: LocalVariableAddrRange {
; OBJ: OffsetStart: .text+0xC
; OBJ: ISectStart: 0x0
-; OBJ: Range: 0x6
+; OBJ: Range: 0x4
; OBJ: }
; OBJ: }
; OBJ: Local {
@@ -143,7 +143,7 @@
; OBJ: }
; OBJ: DefRangeRegister {
; OBJ: Register: 17
-; OBJ: LocalVariableAddrRange {
+; OBJ: MayHaveNoName: 0
; OBJ: OffsetStart: .text+0x12
; OBJ: ISectStart: 0x0
; OBJ: Range: 0x6
diff --git a/test/DebugInfo/MIR/X86/livedebugvalues-limit.mir b/test/DebugInfo/MIR/X86/livedebugvalues-limit.mir
new file mode 100644
index 000000000000..4c87543636d5
--- /dev/null
+++ b/test/DebugInfo/MIR/X86/livedebugvalues-limit.mir
@@ -0,0 +1,228 @@
+--- |
+ ; RUN: llc -run-pass=livedebugvalues -march=x86-64 -o - %s | FileCheck %s
+ ; Created from:
+ ; void sink(int a);
+ ; void __attribute((always_inline)) f(int a) { sink(a); }
+ ; void foo(int i) {
+ ; f(i);
+ ; if (i)
+ ; f(i);
+ ; f(i);
+ ; }
+ ;
+ ; This test verifies that LiveDebugValues doesn't propagate DBG_VALUEs into
+ ; basic blocks that are beyond the scope of the source variable.
+ ;
+ ; CHECK: bb.1.if.then:
+ ; CHECK: DBG_VALUE debug-use %ebx, debug-use _, !19, !13, debug-location !20
+ ; CHECK-NOT: DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !21
+ ; CHECK: DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !27
+ ; CHECK: bb.2.if.end:
+ ; CHECK: DBG_VALUE debug-use %ebx, debug-use _, !19, !13, debug-location !20
+ ; CHECK-NOT: DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !21
+ ; CHECK: DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !31
+ ;
+ ; ModuleID = 'livedebugvalues-limit.ll'
+ source_filename = "livedebugvalues-limit.c"
+ target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
+ target triple = "x86_64-apple-macosx"
+
+ ; Function Attrs: alwaysinline nounwind ssp uwtable
+ define void @f(i32 %a) local_unnamed_addr #0 !dbg !7 {
+ entry:
+ tail call void @llvm.dbg.value(metadata i32 %a, i64 0, metadata !12, metadata !13), !dbg !14
+ tail call void @sink(i32 %a) #4, !dbg !15
+ ret void, !dbg !16
+ }
+
+ declare void @sink(i32) local_unnamed_addr
+
+ ; Function Attrs: nounwind ssp uwtable
+ define void @foo(i32 %i) local_unnamed_addr #2 !dbg !17 {
+ entry:
+ tail call void @llvm.dbg.value(metadata i32 %i, i64 0, metadata !19, metadata !13), !dbg !20
+ tail call void @llvm.dbg.value(metadata i32 %i, i64 0, metadata !12, metadata !13) #4, !dbg !21
+ tail call void @sink(i32 %i) #4, !dbg !23
+ %tobool = icmp eq i32 %i, 0, !dbg !24
+ br i1 %tobool, label %if.end, label %if.then, !dbg !26
+
+ if.then: ; preds = %entry
+ tail call void @llvm.dbg.value(metadata i32 %i, i64 0, metadata !12, metadata !13) #4, !dbg !27
+ tail call void @sink(i32 %i) #4, !dbg !29
+ br label %if.end, !dbg !30
+
+ if.end: ; preds = %if.then, %entry
+ tail call void @llvm.dbg.value(metadata i32 %i, i64 0, metadata !12, metadata !13) #4, !dbg !31
+ tail call void @sink(i32 %i) #4, !dbg !33
+ ret void, !dbg !34
+ }
+
+ ; Function Attrs: nounwind readnone
+ declare void @llvm.dbg.value(metadata, i64, metadata, metadata) #3
+
+ ; Function Attrs: nounwind
+ declare void @llvm.stackprotector(i8*, i8**) #4
+
+ attributes #0 = { alwaysinline nounwind ssp uwtable }
+ attributes #2 = { nounwind ssp uwtable }
+ attributes #3 = { nounwind readnone }
+ attributes #4 = { nounwind }
+
+ !llvm.dbg.cu = !{!0}
+ !llvm.module.flags = !{!3, !4, !5}
+ !llvm.ident = !{!6}
+
+ !0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 4.0.0 (trunk 281923) (llvm/trunk 281916)", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2)
+ !1 = !DIFile(filename: "livedebugvalues-limit.c", directory: "/Volumes/Fusion/Data/llvm")
+ !2 = !{}
+ !3 = !{i32 2, !"Dwarf Version", i32 4}
+ !4 = !{i32 2, !"Debug Info Version", i32 3}
+ !5 = !{i32 1, !"PIC Level", i32 2}
+ !6 = !{!"clang version 4.0.0 (trunk 281923) (llvm/trunk 281916)"}
+ !7 = distinct !DISubprogram(name: "f", scope: !1, file: !1, line: 3, type: !8, isLocal: false, isDefinition: true, scopeLine: 3, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !11)
+ !8 = !DISubroutineType(types: !9)
+ !9 = !{null, !10}
+ !10 = !DIBasicType(name: "int", size: 32, align: 32, encoding: DW_ATE_signed)
+ !11 = !{!12}
+ !12 = !DILocalVariable(name: "a", arg: 1, scope: !7, file: !1, line: 3, type: !10)
+ !13 = !DIExpression()
+ !14 = !DILocation(line: 3, column: 41, scope: !7)
+ !15 = !DILocation(line: 3, column: 46, scope: !7)
+ !16 = !DILocation(line: 3, column: 55, scope: !7)
+ !17 = distinct !DISubprogram(name: "foo", scope: !1, file: !1, line: 4, type: !8, isLocal: false, isDefinition: true, scopeLine: 4, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !18)
+ !18 = !{!19}
+ !19 = !DILocalVariable(name: "i", arg: 1, scope: !17, file: !1, line: 4, type: !10)
+ !20 = !DILocation(line: 4, column: 14, scope: !17)
+ !21 = !DILocation(line: 3, column: 41, scope: !7, inlinedAt: !22)
+ !22 = distinct !DILocation(line: 5, column: 3, scope: !17)
+ !23 = !DILocation(line: 3, column: 46, scope: !7, inlinedAt: !22)
+ !24 = !DILocation(line: 6, column: 7, scope: !25)
+ !25 = distinct !DILexicalBlock(scope: !17, file: !1, line: 6, column: 7)
+ !26 = !DILocation(line: 6, column: 7, scope: !17)
+ !27 = !DILocation(line: 3, column: 41, scope: !7, inlinedAt: !28)
+ !28 = distinct !DILocation(line: 7, column: 5, scope: !25)
+ !29 = !DILocation(line: 3, column: 46, scope: !7, inlinedAt: !28)
+ !30 = !DILocation(line: 7, column: 5, scope: !25)
+ !31 = !DILocation(line: 3, column: 41, scope: !7, inlinedAt: !32)
+ !32 = distinct !DILocation(line: 8, column: 3, scope: !17)
+ !33 = !DILocation(line: 3, column: 46, scope: !7, inlinedAt: !32)
+ !34 = !DILocation(line: 9, column: 1, scope: !17)
+
+...
+---
+name: f
+alignment: 4
+exposesReturnsTwice: false
+legalized: false
+regBankSelected: false
+selected: false
+tracksRegLiveness: true
+liveins:
+ - { reg: '%edi' }
+calleeSavedRegisters: [ '%bh', '%bl', '%bp', '%bpl', '%bx', '%ebp', '%ebx',
+ '%rbp', '%rbx', '%r12', '%r13', '%r14', '%r15',
+ '%r12b', '%r13b', '%r14b', '%r15b', '%r12d', '%r13d',
+ '%r14d', '%r15d', '%r12w', '%r13w', '%r14w', '%r15w' ]
+frameInfo:
+ isFrameAddressTaken: false
+ isReturnAddressTaken: false
+ hasStackMap: false
+ hasPatchPoint: false
+ stackSize: 8
+ offsetAdjustment: 0
+ maxAlignment: 0
+ adjustsStack: false
+ hasCalls: false
+ maxCallFrameSize: 0
+ hasOpaqueSPAdjustment: false
+ hasVAStart: false
+ hasMustTailInVarArgFunc: false
+fixedStack:
+ - { id: 0, type: spill-slot, offset: -16, size: 8, alignment: 16 }
+body: |
+ bb.0.entry:
+ liveins: %edi, %rbp
+
+ frame-setup PUSH64r killed %rbp, implicit-def %rsp, implicit %rsp
+ CFI_INSTRUCTION def_cfa_offset 16
+ CFI_INSTRUCTION offset %rbp, -16
+ %rbp = frame-setup MOV64rr %rsp
+ CFI_INSTRUCTION def_cfa_register %rbp
+ DBG_VALUE debug-use %edi, debug-use _, !12, !13, debug-location !14
+ %rbp = POP64r implicit-def %rsp, implicit %rsp, debug-location !15
+ TAILJMPd64 @sink, csr_64, implicit %rsp, implicit %rsp, implicit %edi, debug-location !15
+
+...
+---
+name: foo
+alignment: 4
+exposesReturnsTwice: false
+legalized: false
+regBankSelected: false
+selected: false
+tracksRegLiveness: true
+liveins:
+ - { reg: '%edi' }
+calleeSavedRegisters: [ '%bh', '%bl', '%bp', '%bpl', '%bx', '%ebp', '%ebx',
+ '%rbp', '%rbx', '%r12', '%r13', '%r14', '%r15',
+ '%r12b', '%r13b', '%r14b', '%r15b', '%r12d', '%r13d',
+ '%r14d', '%r15d', '%r12w', '%r13w', '%r14w', '%r15w' ]
+frameInfo:
+ isFrameAddressTaken: false
+ isReturnAddressTaken: false
+ hasStackMap: false
+ hasPatchPoint: false
+ stackSize: 24
+ offsetAdjustment: -8
+ maxAlignment: 0
+ adjustsStack: true
+ hasCalls: true
+ maxCallFrameSize: 0
+ hasOpaqueSPAdjustment: false
+ hasVAStart: false
+ hasMustTailInVarArgFunc: false
+fixedStack:
+ - { id: 0, type: spill-slot, offset: -24, size: 8, alignment: 8, callee-saved-register: '%rbx' }
+ - { id: 1, type: spill-slot, offset: -16, size: 8, alignment: 16 }
+body: |
+ bb.0.entry:
+ successors: %bb.2.if.end, %bb.1.if.then
+ liveins: %edi, %rbx, %rbp
+
+ frame-setup PUSH64r killed %rbp, implicit-def %rsp, implicit %rsp
+ CFI_INSTRUCTION def_cfa_offset 16
+ CFI_INSTRUCTION offset %rbp, -16
+ %rbp = frame-setup MOV64rr %rsp
+ CFI_INSTRUCTION def_cfa_register %rbp
+ frame-setup PUSH64r killed %rbx, implicit-def %rsp, implicit %rsp
+ frame-setup PUSH64r undef %rax, implicit-def %rsp, implicit %rsp
+ CFI_INSTRUCTION offset %rbx, -24
+ DBG_VALUE debug-use %edi, debug-use _, !19, !13, debug-location !20
+ %ebx = MOV32rr %edi
+ DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !21
+ DBG_VALUE debug-use %ebx, debug-use _, !19, !13, debug-location !20
+ CALL64pcrel32 @sink, csr_64, implicit %rsp, implicit %edi, implicit-def %rsp, debug-location !23
+ TEST32rr %ebx, %ebx, implicit-def %eflags, debug-location !24
+ JE_1 %bb.2.if.end, implicit %eflags
+
+ bb.1.if.then:
+ successors: %bb.2.if.end
+ liveins: %ebx, %rbp
+
+ DBG_VALUE debug-use %ebx, debug-use _, !19, !13, debug-location !20
+ DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !27
+ %edi = MOV32rr %ebx, debug-location !29
+ CALL64pcrel32 @sink, csr_64, implicit %rsp, implicit %edi, implicit-def %rsp, debug-location !29
+
+ bb.2.if.end:
+ liveins: %ebx, %rbp
+
+ DBG_VALUE debug-use %ebx, debug-use _, !19, !13, debug-location !20
+ %edi = MOV32rr killed %ebx, debug-location !33
+ %rsp = ADD64ri8 %rsp, 8, implicit-def dead %eflags, debug-location !33
+ DBG_VALUE debug-use %ebx, debug-use _, !12, !13, debug-location !31
+ %rbx = POP64r implicit-def %rsp, implicit %rsp, debug-location !33
+ %rbp = POP64r implicit-def %rsp, implicit %rsp, debug-location !33
+ TAILJMPd64 @sink, csr_64, implicit %rsp, implicit %rsp, implicit %edi, debug-location !33
+
+...
diff --git a/test/DebugInfo/X86/fission-ranges.ll b/test/DebugInfo/X86/fission-ranges.ll
index 3c05f223ee79..0dfb13ab66b7 100644
--- a/test/DebugInfo/X86/fission-ranges.ll
+++ b/test/DebugInfo/X86/fission-ranges.ll
@@ -32,13 +32,13 @@
; CHECK-NEXT: Length: 25
; CHECK-NEXT: Location description: 50 93 04
; CHECK: [[E]]: Beginning address index: 4
-; CHECK-NEXT: Length: 23
+; CHECK-NEXT: Length: 19
; CHECK-NEXT: Location description: 50 93 04
; CHECK: [[B]]: Beginning address index: 5
-; CHECK-NEXT: Length: 21
+; CHECK-NEXT: Length: 17
; CHECK-NEXT: Location description: 50 93 04
; CHECK: [[D]]: Beginning address index: 6
-; CHECK-NEXT: Length: 21
+; CHECK-NEXT: Length: 17
; CHECK-NEXT: Location description: 50 93 04
; Make sure we don't produce any relocations in any .dwo section (though in particular, debug_info.dwo)
--
2.7.4

View File

@ -1,103 +0,0 @@
From 9b74379db3f9e30516f053138fd3697d42b23164 Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Fri, 7 Oct 2016 01:38:35 +0000
Subject: [rust-lang/llvm#54] [SimplifyCFG] Correctly test for unconditional
branches in GetCaseResults
GetCaseResults assumed that a terminator with one successor was an
unconditional branch. This is not necessarily the case, it could be a
cleanupret.
Strengthen the check by querying whether or not the terminator is
exceptional.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@283517 91177308-0d34-0410-b5e6-96231b3b80d8
---
lib/Transforms/Utils/SimplifyCFG.cpp | 2 +-
.../SimplifyCFG/X86/switch_to_lookup_table.ll | 60 ++++++++++++++++++++++
2 files changed, 61 insertions(+), 1 deletion(-)
diff --git a/lib/Transforms/Utils/SimplifyCFG.cpp b/lib/Transforms/Utils/SimplifyCFG.cpp
index c197317ac771..d274f97296dc 100644
--- a/lib/Transforms/Utils/SimplifyCFG.cpp
+++ b/lib/Transforms/Utils/SimplifyCFG.cpp
@@ -4228,7 +4228,7 @@ GetCaseResults(SwitchInst *SI, ConstantInt *CaseVal, BasicBlock *CaseDest,
++I) {
if (TerminatorInst *T = dyn_cast<TerminatorInst>(I)) {
// If the terminator is a simple branch, continue to the next block.
- if (T->getNumSuccessors() != 1)
+ if (T->getNumSuccessors() != 1 || T->isExceptional())
return false;
Pred = CaseDest;
CaseDest = T->getSuccessor(0);
diff --git a/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll b/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
index bae8c1dc5a4b..77e355a00528 100644
--- a/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
+++ b/test/Transforms/SimplifyCFG/X86/switch_to_lookup_table.ll
@@ -1334,3 +1334,63 @@ cleanup4:
br label %while.body
}
+declare void @throw(i1)
+
+define void @wineh_test(i64 %val) personality i32 (...)* @__CxxFrameHandler3 {
+entry:
+ invoke void @throw(i1 false)
+ to label %unreachable unwind label %cleanup1
+
+unreachable:
+ unreachable
+
+cleanup1:
+ %cleanuppad1 = cleanuppad within none []
+ switch i64 %val, label %cleanupdone2 [
+ i64 0, label %cleanupdone1
+ i64 1, label %cleanupdone1
+ i64 6, label %cleanupdone1
+ ]
+
+cleanupdone1:
+ cleanupret from %cleanuppad1 unwind label %cleanup2
+
+cleanupdone2:
+ cleanupret from %cleanuppad1 unwind label %cleanup2
+
+cleanup2:
+ %phi = phi i1 [ true, %cleanupdone1 ], [ false, %cleanupdone2 ]
+ %cleanuppad2 = cleanuppad within none []
+ call void @throw(i1 %phi) [ "funclet"(token %cleanuppad2) ]
+ unreachable
+}
+
+; CHECK-LABEL: @wineh_test(
+; CHECK: entry:
+; CHECK: invoke void @throw(i1 false)
+; CHECK: to label %[[unreachable:.*]] unwind label %[[cleanup1:.*]]
+
+; CHECK: [[unreachable]]:
+; CHECK: unreachable
+
+; CHECK: [[cleanup1]]:
+; CHECK: %[[cleanuppad1:.*]] = cleanuppad within none []
+; CHECK: switch i64 %val, label %[[cleanupdone2:.*]] [
+; CHECK: i64 0, label %[[cleanupdone1:.*]]
+; CHECK: i64 1, label %[[cleanupdone1]]
+; CHECK: i64 6, label %[[cleanupdone1]]
+; CHECK: ]
+
+; CHECK: [[cleanupdone1]]:
+; CHECK: cleanupret from %[[cleanuppad1]] unwind label %[[cleanup2:.*]]
+
+; CHECK: [[cleanupdone2]]:
+; CHECK: cleanupret from %[[cleanuppad1]] unwind label %[[cleanup2]]
+
+; CHECK: [[cleanup2]]:
+; CHECK: %[[phi:.*]] = phi i1 [ true, %[[cleanupdone1]] ], [ false, %[[cleanupdone2]] ]
+; CHECK: %[[cleanuppad2:.*]] = cleanuppad within none []
+; CHECK: call void @throw(i1 %[[phi]]) [ "funclet"(token %[[cleanuppad2]]) ]
+; CHECK: unreachable
+
+declare i32 @__CxxFrameHandler3(...)
--
2.7.4

View File

@ -1,85 +0,0 @@
From eff5dc809ed54701f2bb3e15c58d01881299cedf Mon Sep 17 00:00:00 2001
From: David Majnemer <david.majnemer@gmail.com>
Date: Tue, 11 Oct 2016 01:00:45 +0000
Subject: [rust-lang/llvm#55] [InstCombine] Transform !range metadata to
!nonnull when combining loads
When combining an integer load with !range metadata that does not include 0 to a pointer load, make sure emit !nonnull metadata on the newly-created pointer load. This prevents the !nonnull metadata from being dropped during a ptrtoint/inttoptr pair.
This fixes PR30597.
Patch by Ariel Ben-Yehuda!
Differential Revision: https://reviews.llvm.org/D25215
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@283836 91177308-0d34-0410-b5e6-96231b3b80d8
---
.../InstCombine/InstCombineLoadStoreAlloca.cpp | 12 ++++++--
test/Transforms/InstCombine/PR30597.ll | 32 ++++++++++++++++++++++
2 files changed, 42 insertions(+), 2 deletions(-)
create mode 100644 test/Transforms/InstCombine/PR30597.ll
diff --git a/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp b/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
index d312983ed51b..26f4e764501a 100644
--- a/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
+++ b/lib/Transforms/InstCombine/InstCombineLoadStoreAlloca.cpp
@@ -380,8 +380,16 @@ static LoadInst *combineLoadToNewType(InstCombiner &IC, LoadInst &LI, Type *NewT
break;
case LLVMContext::MD_range:
// FIXME: It would be nice to propagate this in some way, but the type
- // conversions make it hard. If the new type is a pointer, we could
- // translate it to !nonnull metadata.
+ // conversions make it hard.
+
+ // If it's a pointer now and the range does not contain 0, make it !nonnull.
+ if (NewTy->isPointerTy()) {
+ unsigned BitWidth = IC.getDataLayout().getTypeSizeInBits(NewTy);
+ if (!getConstantRangeFromMetadata(*N).contains(APInt(BitWidth, 0))) {
+ MDNode *NN = MDNode::get(LI.getContext(), None);
+ NewLoad->setMetadata(LLVMContext::MD_nonnull, NN);
+ }
+ }
break;
}
}
diff --git a/test/Transforms/InstCombine/PR30597.ll b/test/Transforms/InstCombine/PR30597.ll
new file mode 100644
index 000000000000..c0803ed71204
--- /dev/null
+++ b/test/Transforms/InstCombine/PR30597.ll
@@ -0,0 +1,32 @@
+; RUN: opt < %s -instcombine -S | FileCheck %s
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+; Function Attrs: readonly uwtable
+define i1 @dot_ref_s(i32** noalias nocapture readonly dereferenceable(8)) {
+entry-block:
+ %loadedptr = load i32*, i32** %0, align 8, !nonnull !0
+ %ptrtoint = ptrtoint i32* %loadedptr to i64
+ %inttoptr = inttoptr i64 %ptrtoint to i32*
+ %switchtmp = icmp eq i32* %inttoptr, null
+ ret i1 %switchtmp
+
+; CHECK-LABEL: @dot_ref_s
+; CHECK-NEXT: entry-block:
+; CHECK-NEXT: ret i1 false
+}
+
+; Function Attrs: readonly uwtable
+define i64* @function(i64* noalias nocapture readonly dereferenceable(8)) {
+entry-block:
+ %loaded = load i64, i64* %0, align 8, !range !1
+ %inttoptr = inttoptr i64 %loaded to i64*
+ ret i64* %inttoptr
+; CHECK-LABEL: @function
+; CHECK: %{{.+}} = load i64*, i64** %{{.+}}, align 8, !nonnull
+}
+
+
+!0 = !{}
+!1 = !{i64 1, i64 140737488355327}
--
2.9.3

View File

@ -1,118 +0,0 @@
From 5ac4f80be3e8b5d42475aeaba246455e0016c7ef Mon Sep 17 00:00:00 2001
From: Anthony Ramine <n.oxyde@gmail.com>
Date: Sun, 27 Nov 2016 16:28:12 +0100
Subject: [rust-lang/llvm#57] Backport rL277331
---
lib/Target/AArch64/AArch64InstrInfo.cpp | 3 +
.../MIR/AArch64/inst-size-tlsdesc-callseq.mir | 84 ++++++++++++++++++++++
2 files changed, 87 insertions(+)
create mode 100644 test/CodeGen/MIR/AArch64/inst-size-tlsdesc-callseq.mir
diff --git a/lib/Target/AArch64/AArch64InstrInfo.cpp b/lib/Target/AArch64/AArch64InstrInfo.cpp
index 0aa4708f35ac..d39542a8e4eb 100644
--- a/lib/Target/AArch64/AArch64InstrInfo.cpp
+++ b/lib/Target/AArch64/AArch64InstrInfo.cpp
@@ -56,6 +56,9 @@ unsigned AArch64InstrInfo::GetInstSizeInBytes(const MachineInstr &MI) const {
case TargetOpcode::IMPLICIT_DEF:
case TargetOpcode::KILL:
return 0;
+ case AArch64::TLSDESC_CALLSEQ:
+ // This gets lowered to an instruction sequence which takes 16 bytes
+ return 16;
}
llvm_unreachable("GetInstSizeInBytes()- Unable to determin insn size");
diff --git a/test/CodeGen/MIR/AArch64/inst-size-tlsdesc-callseq.mir b/test/CodeGen/MIR/AArch64/inst-size-tlsdesc-callseq.mir
new file mode 100644
index 000000000000..2d966ece768e
--- /dev/null
+++ b/test/CodeGen/MIR/AArch64/inst-size-tlsdesc-callseq.mir
@@ -0,0 +1,84 @@
+# RUN: llc -mtriple=aarch64-unknown -run-pass aarch64-branch-relax -aarch64-tbz-offset-bits=4 %s -o - | FileCheck %s
+--- |
+ ; ModuleID = 'test.ll'
+ source_filename = "test.ll"
+ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128"
+ target triple = "aarch64-unknown"
+
+ @ThreadLocalGlobal = external thread_local local_unnamed_addr global i32, align 8
+
+ define i32 @test_tlsdesc_callseq_length(i32 %in) {
+ %val = and i32 %in, 1
+ %tst = icmp eq i32 %val, 0
+ br i1 %tst, label %true, label %false
+
+ true: ; preds = %0
+ %1 = load i32, i32* @ThreadLocalGlobal, align 8
+ ret i32 %1
+
+ false: ; preds = %0
+ ret i32 0
+ }
+
+...
+---
+# CHECK-LABEL: name:{{.*}}test_tlsdesc_callseq_length
+# If the size of TLSDESC_CALLSEQ is computed correctly, that will push
+# the bb.2.false block too far away from the TBNZW, so the branch will
+# have to be relaxed (note that we're using -aarch64-tbz-offset-bits to
+# constrain the range that can be reached with the TBNZW to something smaller
+# than what TLSDESC_CALLSEQ is lowered to).
+# CHECK: TBZW killed %w0, 0, %bb.1.true
+# CHECK: B %bb.2.false
+name: test_tlsdesc_callseq_length
+alignment: 2
+exposesReturnsTwice: false
+hasInlineAsm: false
+allVRegsAllocated: true
+isSSA: false
+tracksRegLiveness: false
+tracksSubRegLiveness: false
+liveins:
+ - { reg: '%w0' }
+frameInfo:
+ isFrameAddressTaken: false
+ isReturnAddressTaken: false
+ hasStackMap: false
+ hasPatchPoint: false
+ stackSize: 16
+ offsetAdjustment: 0
+ maxAlignment: 16
+ adjustsStack: false
+ hasCalls: true
+ maxCallFrameSize: 0
+ hasOpaqueSPAdjustment: false
+ hasVAStart: false
+ hasMustTailInVarArgFunc: false
+stack:
+ - { id: 0, type: spill-slot, offset: -16, size: 8, alignment: 16, callee-saved-register: '%lr' }
+body: |
+ bb.0 (%ir-block.0):
+ successors: %bb.1.true, %bb.2.false
+ liveins: %w0, %lr
+
+ TBNZW killed %w0, 0, %bb.2.false
+
+ bb.1.true:
+ liveins: %lr
+
+ early-clobber %sp = frame-setup STRXpre killed %lr, %sp, -16 :: (store 8 into %stack.0)
+ frame-setup CFI_INSTRUCTION def_cfa_offset 16
+ frame-setup CFI_INSTRUCTION offset %w30, -16
+ TLSDESC_CALLSEQ target-flags(aarch64-tls) @ThreadLocalGlobal, implicit-def dead %lr, implicit-def %x0, implicit-def dead %x1
+ %x8 = MRS 56962
+ %w0 = LDRWroX killed %x8, killed %x0, 0, 0 :: (load 4 from @ThreadLocalGlobal, align 8)
+ early-clobber %sp, %lr = LDRXpost %sp, 16 :: (load 8 from %stack.0)
+ RET killed %lr, implicit killed %w0
+
+ bb.2.false:
+ liveins: %lr
+
+ %w0 = ORRWrs %wzr, %wzr, 0
+ RET killed %lr, implicit killed %w0
+
+...
--
2.9.3

View File

@ -1 +1 @@
SHA512 (llvm-3.9.1.src.tar.xz) = 50cbe8ee911080f586e77861c442348701bd02e2de0c090c54c34f82ac275ecfcd712af0f41e387c33b4a6057778a4258a27554292fe68ab4af3fd9dd6d90683
SHA512 (llvm-4.0.0.src.tar.xz) = cf681f0626ef6d568d951cdc3e143471a1d7715a0ba11e52aa273cf5d8d421e1357ef2645cc85879eaefcd577e99e74d07b01566825b3d0461171ef2cbfc7704