|
 | mlir/lib/Conversion/AMDGPUToROCDL/AMDGPUToROCDL.cpp (diff) |
 | mlir/test/Dialect/AMDGPU/ops.mlir (diff) |
 | mlir/include/mlir/Dialect/AMDGPU/AMDGPU.td (diff) |
 | mlir/test/Conversion/AMDGPUToROCDL/amdgpu-to-rocdl.mlir (diff) |
|
 | llvm/test/CodeGen/PowerPC/pr61315.ll |
Commit
eecb8c5f06149baf970fa0943e9fb9a6afe00207
by aeubanks[SampleProfile] Use LazyCallGraph instead of CallGraph
The function order in some tests had to be changed because they relied on ordering of functions returned in an SCC which is consistent but unspecified.
|
 | llvm/test/Other/new-pm-thinlto-prelink-samplepgo-defaults.ll (diff) |
 | llvm/test/Transforms/SampleProfile/profile-context-order.ll (diff) |
 | llvm/lib/Transforms/IPO/SampleProfile.cpp (diff) |
 | llvm/test/Other/new-pm-thinlto-postlink-samplepgo-defaults.ll (diff) |
 | llvm/test/Transforms/SampleProfile/profile-topdown-order.ll (diff) |
Commit
091422adc1d7478b126a967c795414840c5c0c97
by listmail[LSR] Fix wrapping bug in lsr-term-fold logic
The existing logic was unsound, in two ways.
First, due to wrapping on the trip count computation, it could compute a value which convert a loop which exiting on iteration 256, to one which exited at 255. (With i8 trip counts.)
Second, it allowed rewriting when the trip count implies wrapping around the alternate IV. As a trivial example, it allowed rewriting an i128 exit test in terms of an i64 IV. This is obviously wrong.
Note that the test change is fairly minimal - i.e. only the targeted test - but that's only because I precommitted a change which switched the test from 32 to 64 bit pointers. For 32 bit point architectures with 32 bit primary inductions, this transform is almost always unsound to perform.
Differential Revision: https://reviews.llvm.org/D146429
|
 | llvm/test/Transforms/LoopStrengthReduce/lsr-term-fold.ll (diff) |
 | llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp (diff) |
Commit
3af1c48c66a3cf906bd35c2a6e9d96b28ad19438
by ravishankarmChanges to `SCFFuseProducerOfSliceResult` to also return the operations created during fusion.
This is follow up to https://reviews.llvm.org/D145133 that allows propogating information about ops that are fused back to the caller.
Reviewed By: hanchung
Differential Revision: https://reviews.llvm.org/D146254
|
 | mlir/include/mlir/Dialect/SCF/Transforms/TileUsingInterface.h (diff) |
 | mlir/lib/Dialect/SCF/Transforms/TileUsingInterface.cpp (diff) |
Commit
fb1b9945be7657a3a25b727eaf0eeb3f74525210
by Lang Hames[JITLink][ELF][AArch64] Add support for ELF R_AARCH64_ABS32 relocation.
This relocation is commonly used in debug sections. Failure to handle it caused the test failure in https://lab.llvm.org/buildbot/#/builders/197/builds/4272, which forced the reversion, in f721fcb6ed0, of 57aeb305460 ("[JITLink][ELF] Don't skip debug info sections by default"). This fix should allow us to re-land 57aeb305460.
|
 | llvm/lib/ExecutionEngine/JITLink/ELF_aarch64.cpp (diff) |
 | llvm/test/ExecutionEngine/JITLink/AArch64/ELF_R_AARCH64_ABS32.yaml |
|
 | mlir/include/mlir/Dialect/Tensor/IR/TensorOps.td (diff) |
 | mlir/lib/Dialect/Tensor/IR/TensorOps.cpp (diff) |
 | mlir/lib/Dialect/Utils/StructuredOpsUtils.cpp (diff) |
 | mlir/include/mlir/Dialect/Utils/StructuredOpsUtils.h (diff) |
 | mlir/include/mlir/Dialect/Linalg/Utils/Utils.h (diff) |
|
 | utils/bazel/llvm-project-overlay/mlir/BUILD.bazel (diff) |
|
 | llvm/lib/Option/Option.cpp (diff) |
|
 | llvm/lib/Option/Option.cpp (diff) |
|
 | mlir/lib/Dialect/Utils/StructuredOpsUtils.cpp (diff) |
Commit
411b1d8f079533860a990ee615abae3b0e6dbd8b
by ravishankarm[mlir][Tensor] Make `TilingInterface` implementation only return handle to the created `pad` operation.
Pad tiling implementation only needs to return the tiled pad operation. The rest of the generated code is related to handling boundary conditions.
Reviewed By: hanchung
Differential Revision: https://reviews.llvm.org/D146439
|
 | mlir/lib/Dialect/Tensor/IR/TensorTilingInterfaceImpl.cpp (diff) |
|
 | mlir/lib/Dialect/SparseTensor/Transforms/Sparsification.cpp (diff) |
 | mlir/include/mlir/Dialect/SparseTensor/Utils/Merger.h (diff) |
 | mlir/lib/Dialect/SparseTensor/Utils/Merger.cpp (diff) |
 | mlir/lib/Dialect/SparseTensor/Transforms/CodegenEnv.h (diff) |
|
 | llvm/test/CodeGen/PowerPC/pr61315.ll (diff) |
Commit
919a3f1c751bf27c3c28018d8ff6cb55a81a5164
by Vitaly BukaAdd declaration for `__tsan_default_options` to tsan_interface.h
`__tsan_default_options` is part of the tsan interface so should be exposed in tsan_interface.h.
Differential Revision: https://reviews.llvm.org/D146259
|
 | compiler-rt/lib/tsan/rtl/tsan_flags.cpp (diff) |
 | compiler-rt/lib/tsan/rtl/tsan_interface.h (diff) |
Commit
f2c3a9cbd229791b63d58f7499fe9ebf74c3303b
by michaelrj[libc][obvious] fix missing memory_check_utils
memmove and bcopy tests were missing a dependancy on memory_check_utils in the bazel build, causing build failures.
Differential Revision: https://reviews.llvm.org/D146462
|
 | utils/bazel/llvm-project-overlay/libc/test/src/string/BUILD.bazel (diff) |
Commit
99047c0501e0fe5c60bb583185f3b45bdc112199
by michaelrj[libc][bazel] add targets for sprintf
The bazel build is currently overlay mode only, so the FILE functions are still out of reach for it, but sprintf only uses strings. This adds targets for sprintf, snprintf, and all the interal printf pieces, as well as tests.
Reviewed By: sivachandra, lntue
Differential Revision: https://reviews.llvm.org/D146100
|
 | utils/bazel/llvm-project-overlay/libc/test/UnitTest/BUILD.bazel (diff) |
 | libc/src/stdio/printf_core/CMakeLists.txt (diff) |
 | utils/bazel/llvm-project-overlay/libc/BUILD.bazel (diff) |
 | utils/bazel/llvm-project-overlay/libc/test/src/stdio/BUILD.bazel |
Commit
1187d8a62ba288e2221731f1795fa184571cd854
by Ian Anderson[libunwind][Modules] Add unwind_arm_ehabi.h and unwind_itanium.h to the unwind module)
Add unwind_arm_ehabi.h and unwind_itanium.h to the unwind module and use angle includes to include them.
Reviewed By: ldionne, #libunwind
Differential Revision: https://reviews.llvm.org/D144323
|
 | libunwind/include/unwind.h (diff) |
 | libunwind/include/libunwind.modulemap (diff) |
|
 | llvm/lib/CodeGen/ValueTypes.cpp (diff) |
 | llvm/include/llvm/Support/MachineValueType.h (diff) |
 | llvm/include/llvm/CodeGen/ValueTypes.td (diff) |
Commit
8d024a79ea783ed3fbb5691aeaf186ad3f0a4ae9
by jinghamFix a problem with "watchpoint triggers before" watchpoint handling.
We need to step the watchpoint instruction in these cases, but the when we queued the ThreadPlanStepOverWatchpoint to do this, we didn't make it a Controlling plan. So if you are stepping, this plan returns as though it were a utility plan, and the stepping plan keeps going.
This only partially fixes the problem on Darwin; there's another bug with reporting a watchpoint when we're instruction single stepping over an instruction that triggers a watchpoint. The kernel reports the "single step completed" but not the watchpoint hit. So this commit also refactors the test into a part that works (at least on Darwin) and a part that still fails.
We may have to adjust the test result expectations for other systems after this fix.
Differential Revision: https://reviews.llvm.org/D146337
|
 | lldb/source/Target/Thread.cpp (diff) |
 | lldb/source/Target/StopInfo.cpp (diff) |
 | lldb/test/API/commands/watchpoints/step_over_watchpoint/TestStepOverWatchpoint.py (diff) |
 | lldb/source/Plugins/Process/Utility/StopInfoMachException.cpp (diff) |
 | lldb/test/API/commands/watchpoints/step_over_watchpoint/main.c (diff) |
 | lldb/include/lldb/Target/Process.h (diff) |