| /openbsd-src/gnu/llvm/llvm/lib/Transforms/Scalar/ |
| H A D | CMakeLists.txt | 27 LICM.cpp
|
| H A D | LICM.cpp | 213 : LoopPass(ID), LICM(LicmMssaOptCap, LicmMssaNoAccForPromotionCap, in LegacyLICMPass()
233 return LICM.runOnLoop( in runOnLoop()
261 LoopInvariantCodeMotion LICM; member
276 LoopInvariantCodeMotion LICM(Opts.MssaOptCap, Opts.MssaNoAccForPromotionCap, in run() local
278 if (!LICM.runOnLoop(&L, &AR.AA, &AR.LI, &AR.DT, &AR.AC, &AR.TLI, &AR.TTI, in run()
313 LoopInvariantCodeMotion LICM(Opts.MssaOptCap, Opts.MssaNoAccForPromotionCap, in run() local
317 bool Changed = LICM.runOnLoop(&OutermostLoop, &AR.AA, &AR.LI, &AR.DT, &AR.AC, in run()
|
| /openbsd-src/gnu/usr.bin/clang/libLLVMScalarOpts/ |
| H A D | Makefile | 34 LICM.cpp \
|
| /openbsd-src/gnu/llvm/llvm/utils/gn/secondary/llvm/lib/Transforms/Scalar/ |
| H A D | BUILD.gn | 38 "LICM.cpp",
|
| /openbsd-src/gnu/usr.bin/clang/include/llvm/Config/ |
| H A D | llvm-config.h.in | 24 * lib/Transforms/Scalar/LICM.cpp
|
| /openbsd-src/gnu/llvm/llvm/docs/HistoricalNotes/ |
| H A D | 2001-06-01-GCCOptimizations.txt | 24 11. LICM
|
| /openbsd-src/gnu/llvm/llvm/lib/Target/WebAssembly/ |
| H A D | README.txt | 54 AArch64 runs SeparateConstOffsetFromGEPPass, followed by EarlyCSE and LICM.
|
| /openbsd-src/gnu/llvm/llvm/docs/Frontend/ |
| H A D | PerformanceTips.rst | 152 block is a loop exiting conditional branch, the effectiveness of LICM will
263 two of LoopUnswitch and LICM to your pass order. The standard pass order,
|
| /openbsd-src/gnu/llvm/llvm/docs/ |
| H A D | Passes.rst | 728 also handles other optimizations than LICM that increase live-ranges.
804 analyses and transformations, such as :ref:`LICM <passes-licm>`.
810 into LICM.
870 This pass expects :ref:`LICM <passes-licm>` to be run before it to hoist
975 promote better constant propagation, GCSE, :ref:`LICM <passes-licm>`, PRE, etc.
|
| H A D | HowToUpdateDebugInfo.rst | 59 * LICM. E.g., if an instruction is moved from the loop body to the preheader,
97 * Merging identical loop-invariant stores (see the LICM utility
|
| H A D | Lexicon.rst | 189 **LICM**
|
| H A D | WritingAnLLVMPass.rst | 806 void LICM::getAnalysisUsage(AnalysisUsage &AU) const {
824 bool LICM::runOnFunction(Function &F) {
1067 Here we see that GVN uses dominator tree information to do its job. The LICM pass
1070 After the LICM pass, the module verifier runs (which is automatically added by
|
| H A D | AliasAnalysis.rst | 83 (LICM) pass can use store motion to remove the stores from the loop. In
415 The ``AliasSetTracker`` class (which is used by ``LICM``) makes a
|
| H A D | Atomics.rst | 365 Unordered. This would be checked, for example, by LICM before hoisting an
392 * LICM: Unordered loads/stores can be moved out of a loop. It just treats
|
| H A D | TransformMetadata.rst | 349 Versioning LICM
|
| H A D | MemorySSA.rst | 340 If you'd like examples, ``GVNHoist`` and ``LICM`` are users of ``MemorySSA``\ s
|
| H A D | Coroutines.rst | 1732 along that path (e.g. LICM), which can lead to use-after-free. At the moment
1733 we disabled LICM for loops that have coro.suspend, but the general problem still
|
| /openbsd-src/gnu/llvm/llvm/include/llvm/IR/ |
| H A D | Intrinsics.td | 591 [IntrInaccessibleMemOnly]>; // blocks LICM and some more
|
| H A D | IntrinsicsX86.td | 2741 // LICM.
|
| /openbsd-src/gnu/llvm/llvm/lib/Target/ |
| H A D | README.txt | 279 interchange, memory promotion (which LICM already does), vectorization and
|
| /openbsd-src/gnu/llvm/llvm/lib/Target/PowerPC/ |
| H A D | PPCInstr64Bit.td | 1437 // oppotunities to some opts(CSE, LICM and etc.) for the result of adding
|
| /openbsd-src/distrib/sets/lists/comp/ |
| H A D | clang.amd64 | 2708 ./usr/include/llvm/Transforms/Scalar/LICM.h
|
| H A D | clang.macppc | 2707 ./usr/include/llvm/Transforms/Scalar/LICM.h
|
| H A D | clang.loongson | 2707 ./usr/include/llvm/Transforms/Scalar/LICM.h
|
| H A D | clang.octeon | 2707 ./usr/include/llvm/Transforms/Scalar/LICM.h
|