Home
last modified time | relevance | path

Searched refs:optimizer (Results 1 – 25 of 151) sorted by relevance

1234567

/llvm-project-15.0.7/llvm/docs/
H A DLinkTimeOptimization.rst14 interface and design between the LTO optimizer and the linker.
26 bitcode files. This tight integration between the linker and LLVM optimizer
28 input allows the optimizer to avoid relying on conservative escape analysis.
97 anywhere. This information is used by the LLVM optimizer and it
102 optimizer also removes ``foo3()``.
107 linker. Here, the optimizer can not remove ``foo3()`` without the linker's
113 **Compiler driver invokes link time optimizer separately.**
114 In this model the link time optimizer is not able to take advantage of
118 optimizer from removing ``foo3()``.
142 such as a list of exported symbols. LLVM optimizer collects control flow
[all …]
H A DExtendedIntegerResults.txt79 might want the callee to extend 3) the mid-level optimizer doesn't know the
114 The one thing that is missing is the ability for the front-end and optimizer to
121 and given that it is stuck on the y function, the mid-level optimizer could
H A DHowToSubmitABug.rst42 the LLVM libraries (e.g. the optimizer or code generator) that has
46 optimizer, or backend code generator), run the ``clang`` command line as you
51 passed these options (which disable the optimizer and code generator), then
56 the code generator), you found a middle-end optimizer bug. Jump ahead to
92 If you find that a bug crashes in the optimizer, compile your test-case to a
H A DMisExpect.rst11 expected to behave at runtime to the optimizer. These annotations, however, can
16 that the optimizer can make more useful decisions about the code. MisExpect
/llvm-project-15.0.7/llvm/test/Other/
H A Dnew-pm-O0-ep-callbacks.ll2 ; RUN: opt -disable-output -debug-pass-manager -passes-ep-loop-optimizer-end=no-op-loop -passes='de…
3 ; RUN: opt -disable-output -debug-pass-manager -passes-ep-scalar-optimizer-late=no-op-function -pas…
4 ; RUN: opt -disable-output -debug-pass-manager -passes-ep-cgscc-optimizer-late=no-op-cgscc -passes=…
8 ; RUN: opt -disable-output -debug-pass-manager -passes-ep-optimizer-early=no-op-module -passes='def…
9 ; RUN: opt -disable-output -debug-pass-manager -passes-ep-optimizer-last=no-op-module -passes='defa…
/llvm-project-15.0.7/mlir/lib/Dialect/Bufferization/Transforms/
H A DBufferOptimizations.cpp368 BufferAllocationHoisting<BufferAllocationHoistingState> optimizer( in runOnOperation() local
370 optimizer.hoist(); in runOnOperation()
379 BufferAllocationHoisting<BufferAllocationLoopHoistingState> optimizer( in runOnOperation() local
381 optimizer.hoist(); in runOnOperation()
411 BufferPlacementPromotion optimizer(getOperation()); in runOnOperation() local
412 optimizer.promote(isSmallAlloc); in runOnOperation()
/llvm-project-15.0.7/llvm/docs/Frontend/
H A DPerformanceTips.rst19 As with any optimizer, LLVM has its strengths and weaknesses. In some cases,
63 SSA is the canonical form expected by much of the optimizer; if allocas can
64 not be eliminated by Mem2Reg or SROA, the optimizer is likely to be less
135 and may not be well optimized by the current optimizer. Depending on your
140 instruction. This form conveys to the optimizer that the call returns
143 desired. This is generally not required because the optimizer will convert
176 language specification to emit a particular code sequence. The optimizer
199 a long way to assisting the optimizer in reasoning about loop induction
210 generally hard for an optimizer so providing these facts from the frontend
240 known. The optimizer will try to infer these flags, but may not always be
[all …]
/llvm-project-15.0.7/clang/docs/
H A DMisExpect.rst11 expected to behave at runtime to the optimizer. These annotations, however, can
16 that the optimizer can make more useful decisions about the code.
26 of a condition is severely skewed. As a result, the optimizer can be extremely
/llvm-project-15.0.7/clang-tools-extra/docs/clang-tidy/checks/performance/
H A Dno-int-to-ptr.rst15 and are transparent to the optimizer, integer to (integral) pointer casts
16 are *NOT* transparent, and may conceal information from optimizer.
/llvm-project-15.0.7/mlir/test/IR/
H A Dcheck-help-output.mlir3 // CHECK: OVERVIEW: MLIR modular optimizer driver
/llvm-project-15.0.7/llvm/test/Transforms/LoopVectorize/
H A Dno_switch.ll6 ; CHECK: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform the requ…
9 ; NOANALYSIS: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform the…
13 ; MOREINFO: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform the r…
H A Dno_switch_disable_vectorization.ll9 ; CHECK-NOT: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform the …
12 ; NOANALYSIS-NOT: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform…
15 ; MOREINFO-NOT: warning: source.cpp:4:5: loop not vectorized: the optimizer was unable to perform t…
H A Dno_array_bounds_scalable.ll11 ; CHECK: warning: <unknown>:0:0: loop not interleaved: the optimizer was unable to perform the requ…
38 ; CHECK: warning: <unknown>:0:0: loop not vectorized: the optimizer was unable to perform the reque…
/llvm-project-15.0.7/llvm/test/CodeGen/AArch64/
H A Dstp-opt-with-renaming-undef-assert.mir3 # This test checks that aarch64 load store optimizer is not throwing an
10 # aarch64 load store optimizer should not throw an assertion if a renamable
H A Darm64-coalesce-ext.ll2 ; Check that the peephole optimizer knows about sext and zext instructions.
/llvm-project-15.0.7/llvm/test/Transforms/InstCombine/
H A D2006-10-19-SignedToUnsignedCastAndConst-2.ll1 ; The optimizer should be able to remove cast operation here.
/llvm-project-15.0.7/polly/docs/
H A DUsingPollyWithClang.rst185 Polly uses by default the isl scheduling optimizer. The isl optimizer optimizes
187 To disable the optimizer entirely use the option -polly-optimizer=none.
189 Disable tiling in the optimizer
H A DArchitecture.rst5 Polly is a loop optimizer for LLVM. Starting from LLVM-IR it detects and
51 As an early optimizer before the standard LLVM pass pipeline, as a later
52 optimizer as part of the target specialization sequence, and theoretically also
/llvm-project-15.0.7/clang/test/CodeGenObjC/
H A Dexceptions.m110 // This loop is large enough to dissuade the optimizer from just
115 // on a specific result from the optimizer is really unstable.
120 // on a specific result from the optimizer is really unstable.
/llvm-project-15.0.7/llvm/test/CodeGen/AMDGPU/
H A Dfcmp-cnd.ll4 ;registers and literal.x depending on what the optimizer does.
H A Dlower-module-lds-inactive.ll15 ; constant variables are left to the optimizer / error diagnostics
21 ; External and constant are both left to the optimizer / error diagnostics
/llvm-project-15.0.7/polly/test/ScheduleOptimizer/
H A Dstatistics.ll269 ; CHECK-DAG: 12 polly-opt-isl - Number of band members (after optimizer)
272 ; CHECK-DAG: 8 polly-opt-isl - Number of bands (after optimizer)
274 ; CHECK-DAG: 5 polly-opt-isl - Number of coincident band members (after optimizer)
276 ; CHECK-DAG: 8 polly-opt-isl - Number of permutable bands (after optimizer)
/llvm-project-15.0.7/llvm/test/ExecutionEngine/JITLink/X86/
H A DELF_x86-64_common.s13 # prevent the optimizer from relaxing the edge, we use a movl instruction.
/llvm-project-15.0.7/lld/test/ELF/lto/
H A Dltopasses-basic.ll14 ; `@ctor` doesn't do anything and so the optimizer should kill it, leaving no ctors
/llvm-project-15.0.7/clang/test/CodeGenObjCXX/
H A Dnrvo.mm25 // With the optimizer enabled, the DeadArgElim pass is able to

1234567