Home
last modified time | relevance | path

Searched refs:expensive (Results 1 – 25 of 163) sorted by relevance

1234567

/llvm-project-15.0.7/llvm/test/Transforms/Inline/
H A Dnonnull.ll10 br i1 %cmp, label %expensive, label %done
12 ; This block is designed to be too expensive to inline. We can only inline
14 expensive:
H A Dlast-callsite.ll108 ; The second call is too expensive to inline unless we update the number of
175 ; The first call is too expensive to inline unless we update the number of
253 ; The second call is too expensive to inline unless we update the number of
/llvm-project-15.0.7/llvm/test/Transforms/SimplifyCFG/
H A Dpreserve-branchweights.ll687 %expensive = icmp eq i32 %y, 0
720 %expensive = icmp eq i32 %y, 0
750 %expensive = icmp eq i32 %y, 0
782 %expensive = icmp eq i32 %y, 0
814 %expensive = icmp eq i32 %y, 0
845 %expensive = icmp eq i32 %y, 0
876 %expensive = icmp eq i32 %y, 0
907 %expensive = icmp eq i32 %y, 0
940 %expensive = icmp eq i32 %y, 0
973 %expensive = icmp eq i32 %y, 0
[all …]
/llvm-project-15.0.7/llvm/test/CodeGen/X86/
H A Dor-branch.ll2 ; RUN: llc < %s -mtriple=i386-unknown-unknown -jump-is-expensive=0 | FileCheck %s --check-prefix=JU…
3 ; RUN: llc < %s -mtriple=i386-unknown-unknown -jump-is-expensive=1 | FileCheck %s --check-prefix=JU…
45 ; regardless of whether they are expensive or not.
/llvm-project-15.0.7/llvm/test/Transforms/IndVarSimplify/
H A D2020-12-15-trunc-bug-expensive-range-inference.ll1 ; RUN: opt < %s -indvars -S -scalar-evolution-use-expensive-range-sharpening | FileCheck %s
2 ; RUN: opt < %s -passes=indvars -S -scalar-evolution-use-expensive-range-sharpening | FileCheck %s
H A Drewrite-loop-exit-values-phi.ll5 ; status. This was caught by the expensive check introduced in D86589.
/llvm-project-15.0.7/clang-tools-extra/docs/clang-tidy/checks/performance/
H A Dunnecessary-value-param.rst6 Flags value parameter declarations of expensive to copy types that are copied
9 The check is only applied to parameters of types that are expensive to copy
H A Dimplicit-conversion-in-loop.rst9 result in expensive deep copies.
H A Dfor-range-copy.rst9 The check is only applied to loop variables of types that are expensive to copy
/llvm-project-15.0.7/llvm/test/Transforms/LoopUnroll/
H A Dhigh-cost-trip-count-computation.ll5 ;; Check that we don't emit expensive instructions to compute trip
28 ;; it shouldn't be considered expensive, since the division already
/llvm-project-15.0.7/llvm/test/CodeGen/ARM/
H A Dcopy-cpsr.ll7 ; is carried in the DAG because duplicating them can be more expensive than
17 ; + We want the chains to be long enough that duplicating them is expensive.
H A Dvcvt-cost.ll2 ; instructions as expensive. If lowering is improved the cost model needs to
19 ;; We currently estimate the cost of this instruction as expensive. If lowering
66 ;; We currently estimate the cost of this instruction as expensive. If lowering
/llvm-project-15.0.7/llvm/test/Transforms/InstCombine/
H A D2006-12-05-fp-to-int-ext.ll4 ; significantly more expensive than the two conversions on some targets
/llvm-project-15.0.7/llvm/utils/gn/build/
H A Dbuildflags.gni17 # Whether to enable expensive checks.
/llvm-project-15.0.7/llvm/test/Transforms/LoopUnroll/PowerPC/
H A Da2-high-cost-trip-count-computation.ll5 ;; Check that we do emit expensive instructions to compute trip
/llvm-project-15.0.7/clang-tools-extra/docs/clang-tidy/checks/openmp/
H A Dexception-escape.rst17 WARNING! This check may be expensive on large source files.
/llvm-project-15.0.7/clang-tools-extra/docs/clang-tidy/checks/darwin/
H A Ddispatch-once-nonstatic.rst16 members of structs or classes, with the intent to lazily perform some expensive
/llvm-project-15.0.7/llvm/test/Transforms/LoopVectorize/X86/
H A Dconstant-vector-operand.ll8 ; We use to not vectorize this loop because the shift was deemed to expensive.
/llvm-project-15.0.7/llvm/test/Analysis/CostModel/ARM/
H A Dinsertelement.ll6 ; Multiple insert elements from loads into d subregisters are expensive on swift
/llvm-project-15.0.7/llvm/test/Bitcode/
H A Dmdnodes-distinct-nodes-break-cycles.ll9 ; Forward references for uniqued node operands are expensive, whereas they're
/llvm-project-15.0.7/llvm/test/DebugInfo/X86/
H A Dglobal-expression.ll39 ; This is malformed, but too expensive to detect in the verifier.
/llvm-project-15.0.7/mlir/test/Dialect/Transform/
H A Dexpensive-checks.mlir1 // RUN: mlir-opt --test-transform-dialect-interpreter='enable-expensive-checks=1' --split-input-fil…
/llvm-project-15.0.7/llvm/test/Transforms/IndVarSimplify/NVPTX/
H A Dno-widen-expensive.ll6 ; 64-bit integer is twice as expensive as that on a 32-bit integer, because the
/llvm-project-15.0.7/clang-tools-extra/docs/clang-tidy/checks/bugprone/
H A Dexception-escape.rst25 WARNING! This check may be expensive on large source files.
/llvm-project-15.0.7/llvm/test/MC/Disassembler/AArch64/
H A Darm64-basic-a64-undefined.txt1 # These spawn another process so they're rather expensive. Not many.

1234567