最終更新:2013-09-21 (土) 06:28:03 (3868d)
opt
Top / opt
LLVM optimizer
http://llvm.org/docs/CommandGuide/opt.html
--help
OVERVIEW: llvm .bc -> .bc modular optimizer and analysis printer
USAGE: opt [options] <input bitcode file>
OPTIONS:
-O1 - Optimization level 1. Similar to clang -O1
-O2 - Optimization level 2. Similar to clang -O2
-O3 - Optimization level 3. Similar to clang -O3
-Os - Like -O2 with extra optimizations for size. Similar to clang -Os
-Oz - Like -Os but reduces code size further. Similar to clang -Oz
-S - Write output as LLVM assembly
-analyze - Only perform analysis, no optimization
-asm-verbose - Add comments to directives.
-bounds-checking-single-trap - Use one trap block per function
-code-model - Choose code model
=default - Target default code model
=small - Small code model
=kernel - Kernel code model
=medium - Medium code model
=large - Large code model
-cppfname=<function name> - Specify the name of the generated function
-cppfor=<string> - Specify the name of the thing to generate
-cppgen - Choose what kind of output to generate
=program - Generate a complete program
=module - Generate a module definition
=contents - Generate contents of a module
=function - Generate a function definition
=functions - Generate all function definitions
=inline - Generate an inline function
=variable - Generate a variable definition
=type - Generate a type definition
-default-data-layout=<layout-string> - data layout string to use if not specified by module
-disable-fp-elim - Disable frame pointer elimination optimization
-disable-inlining - Do not run the inliner pass
-disable-internalize - Do not mark all symbols as internal
-disable-non-leaf-fp-elim - Disable frame pointer elimination optimization for non-leaf funcs
-disable-opt - Do not run any optimization passes
-disable-red-zone - Do not emit code that uses the red zone.
-disable-simplify-libcalls - Disable simplify-libcalls
-disable-spill-fusing - Disable fusing of spill code into instructions
-disable-tail-calls - Never emit tail calls
Choose driver interface:
-drvnvcl - Nvidia OpenCL driver
-drvcuda - Nvidia CUDA driver
-drvtest - Plain Test
-enable-correct-eh-support - Make the -lowerinvoke pass insert expensive, but correct, EH code
-enable-fp-mad - Enable less precise MAD instructions to be generated
-enable-load-pre -
-enable-no-infs-fp-math - Enable FP math optimizations that assume no +-Infs
-enable-no-nans-fp-math - Enable FP math optimizations that assume no NaNs
-enable-objc-arc-opts -
-enable-pie - Assume the creation of a position independent executable.
-enable-tbaa -
-enable-unsafe-fp-math - Enable optimizations that may decrease FP precision
-f - Enable binary output on terminals
-fatal-assembler-warnings - Consider warnings as error
-fdata-sections - Emit data into separate sections
-ffunction-sections - Emit functions into separate sections
-filetype - Choose a file type (not all types are supported by all targets):
=asm - Emit an assembly ('.s') file
=obj - Emit a native object ('.o') file
=null - Emit nothing, for performance testing
-float-abi - Choose float ABI type
=default - Target default float ABI type
=soft - Soft float ABI (implied by -soft-float)
=hard - Hard float ABI (uses FP registers)
-fp-contract - Enable aggresive formation of fused FP ops
=fast - Fuse FP ops whenever profitable
=on - Only fuse 'blessed' FP ops.
=off - Only fuse FP ops when the result won't be effected.
-funit-at-a-time - Enable IPO. This is same as llvm-gcc's -funit-at-a-time
-help - Display available options (-help-hidden for more)
-internalize-public-api-file=<filename> - A file containing list of symbol names to preserve
-internalize-public-api-list=<list> - A list of symbol names to preserve
-join-liveintervals - Coalesce copies (default=true)
-limit-float-precision=<uint> - Generate low-precision inline sequences for some float libcalls
-load=<pluginfilename> - Load the specified plugin
-march=<string> - Architecture to generate code for (see --version)
-mattr=<a1,+a2,-a3,...> - Target specific attributes (-mattr=help for details)
-mc-relax-all - When used with filetype=obj, relax all fixups in the emitted object file
-mc-x86-disable-arith-relaxation - Disable relaxation of arithmetic instruction for X86
-mcpu=<cpu-name> - Target a specific cpu type (-mcpu=help for details)
-msp430-hwmult-mode - Hardware multiplier use mode
=no - Do not use hardware multiplier
=interrupts - Assume hardware multiplier can be used inside interrupts
=use - Assume hardware multiplier cannot be used inside interrupts
-mtriple=<string> - Override target triple for module
Optimizations available:
-aa-eval - Exhaustive Alias Analysis Precision Evaluator
-adce - Aggressive Dead Code Elimination
-alloca-hoisting - Hoisting alloca instructions in non-entry blocks to the entry block
-always-inline - Inliner for always_inline functions
-argpromotion - Promote 'by reference' arguments to scalars
-asan - AddressSanitizer: detects use-after-free and out-of-bounds bugs.
-basicaa - Basic Alias Analysis (stateless AA impl)
-basiccg - Basic CallGraph Construction
-bb-vectorize - Basic-Block Vectorization
-block-freq - Block Frequency Analysis
-block-placement - Profile Guided Basic Block Placement
-bounds-checking - Run-time bounds checking
-branch-prob - Branch Probability Analysis
-break-crit-edges - Break critical edges in CFG
-codegenprepare - Optimize for code generation
-constmerge - Merge Duplicate Global Constants
-constprop - Simple constant propagation
-correlated-propagation - Value Propagation
-cost-model - Cost Model Analysis
-count-aa - Count Alias Analysis Query Responses
-da - Dependence Analysis
-datalayout - Data Layout
-dce - Dead Code Elimination
-deadargelim - Dead Argument Elimination
-deadarghaX0r - Dead Argument Hacking (BUGPOINT USE ONLY; DO NOT USE)
-debug-aa - AA use debugger
-die - Dead Instruction Elimination
-domfrontier - Dominance Frontier Construction
-domtree - Dominator Tree Construction
-dot-callgraph - Print Call Graph to 'dot' file
-dot-cfg - Print CFG of function to 'dot' file
-dot-cfg-only - Print CFG of function to 'dot' file (with no function bodies)
-dot-dom - Print dominance tree of function to 'dot' file
-dot-dom-only - Print dominance tree of function to 'dot' file (with no function bodies)
-dot-postdom - Print postdominance tree of function to 'dot' file
-dot-postdom-only - Print postdominance tree of function to 'dot' file (with no function bodies)
-dot-regions - Print regions of function to 'dot' file
-dot-regions-only - Print regions of function to 'dot' file (with no function bodies)
-dse - Dead Store Elimination
-early-cse - Early CSE
-extract-blocks - Extract Basic Blocks From Module (for bugpoint use)
-functionattrs - Deduce function attributes
-globaldce - Dead Global Elimination
-globalopt - Global Variable Optimizer
-globalsmodref-aa - Simple mod/ref analysis for globals
-gvn - Global Value Numbering
-indvars - Induction Variable Simplification
-inline - Function Integration/Inlining
-insert-edge-profiling - Insert instrumentation for edge profiling
-insert-gcov-profiling - Insert instrumentation for GCOV profiling
-insert-optimal-edge-profiling - Insert optimal instrumentation for edge profiling
-insert-path-profiling - Insert instrumentation for Ball-Larus path profiling
-instcombine - Combine redundant instructions
-instcount - Counts the various types of Instructions
-instnamer - Assign names to anonymous instructions
-instsimplify - Remove redundant instructions
-internalize - Internalize Global Symbols
-intervals - Interval Partition Construction
-ipconstprop - Interprocedural constant propagation
-ipsccp - Interprocedural Sparse Conditional Constant Propagation
-iv-users - Induction Variable Users
-jump-threading - Jump Threading
-lazy-value-info - Lazy Value Information Analysis
-lcssa - Loop-Closed SSA Form Pass
-libcall-aa - LibCall Alias Analysis
-licm - Loop Invariant Code Motion
-lint - Statically lint-checks LLVM IR
-loop-deletion - Delete dead loops
-loop-extract - Extract loops into new functions
-loop-extract-single - Extract at most one loop into a new function
-loop-idiom - Recognize loop idioms
-loop-instsimplify - Simplify instructions in loops
-loop-reduce - Loop Strength Reduction
-loop-rotate - Rotate Loops
-loop-simplify - Canonicalize natural loops
-loop-unroll - Unroll loops
-loop-unswitch - Unswitch loops
-loop-vectorize - Loop Vectorization
-loops - Natural Loop Information
-lower-expect - Lower 'expect' Intrinsics
-loweratomic - Lower atomic intrinsics to non-atomic form
-lowerinvoke - Lower invoke and unwind, for unwindless code generators
-lowerswitch - Lower SwitchInst's to branches
-mem2reg - Promote Memory to Register
-memcpyopt - MemCpy Optimization
-memdep - Memory Dependence Analysis
-mergefunc - Merge Functions
-mergereturn - Unify function exit nodes
-metarenamer - Assign new names to everything
-module-debuginfo - Decodes module-level debug info
-no-aa - No Alias Analysis (always returns 'may' alias)
-no-path-profile - No Path Profile Information
-no-profile - No Profile Information
-objc-arc - ObjC ARC optimization
-objc-arc-aa - ObjC-ARC-Based Alias Analysis
-objc-arc-apelim - ObjC ARC autorelease pool elimination
-objc-arc-contract - ObjC ARC contraction
-objc-arc-expand - ObjC ARC expansion
-partial-inliner - Partial Inliner
-path-profile-loader - Load path profile information from file
-path-profile-verifier - Compare the path profile derived edge profile against the edge profile.
-postdomtree - Post-Dominator Tree Construction
-preverify - Preliminary module verification
-print-alias-sets - Alias Set Printer
-print-callgraph - Print a call graph
-print-callgraph-sccs - Print SCCs of the Call Graph
-print-cfg-sccs - Print SCCs of each function CFG
-print-dbginfo - Print debug info in human readable form
-print-dom-info - Dominator Info Printer
-print-externalfnconstants - Print external fn callsites passed constants
-print-function - Print function to stderr
-print-memdeps - Print MemDeps of function
-print-module - Print module to stderr
-print-used-types - Find Used Types
-profile-estimator - Estimate profiling information
-profile-loader - Load profile information from llvmprof.out
-profile-metadata-loader - Load profile information from llvmprof.out
-profile-verifier - Verify profiling information
-prune-eh - Remove unused exception handling info
-reargs - Remove Sign and Zero Extends for Args
-reassociate - Reassociate expressions
-reg2mem - Demote all values to stack slots
-regions - Detect single entry single exit regions
-scalar-evolution - Scalar Evolution Analysis
-scalarrepl - Scalar Replacement of Aggregates (DT)
-scalarrepl-ssa - Scalar Replacement of Aggregates (SSAUp)
-sccp - Sparse Conditional Constant Propagation
-scev-aa - ScalarEvolution-based Alias Analysis
-simplify-libcalls - Simplify well-known library calls
-simplifycfg - Simplify the CFG
-sink - Code sinking
-sroa - Scalar Replacement Of Aggregates
-strip - Strip all symbols from a module
-strip-dead-debug-info - Strip debug info for unused symbols
-strip-dead-prototypes - Strip Unused Function Prototypes
-strip-debug-declare - Strip all llvm.dbg.declare intrinsics
-strip-nondebug - Strip all symbols, except dbg symbols, from a module
-tailcallelim - Tail Call Elimination
-targetlibinfo - Target Library Information
-targettransforminfo - Target Transform Info
-tbaa - Type-Based Alias Analysis
-tsan - ThreadSanitizer: detects data races.
-verify - Module Verifier
-view-cfg - View CFG of function
-view-cfg-only - View CFG of function (with no function bodies)
-view-dom - View dominance tree of function
-view-dom-only - View dominance tree of function (with no function bodies)
-view-postdom - View postdominance tree of function
-view-postdom-only - View postdominance tree of function (with no function bodies)
-view-regions - View regions of function
-view-regions-only - View regions of function (with no function bodies)
-nozero-initialized-in-bss - Don't place zero-initialized symbols into bss section
-nvptx-codegen-vectors - NVPTX Specific: Retain LLVM's vectors and generate PTX vectors
-nvptx-emit-line-numbers - NVPTX Specific: Emit Line numbers even without -G
-nvptx-emit-src - NVPTX Specific: Emit source line in ptx file
-nvptx-fma-level=<int> - NVPTX Specific: FMA contraction (0: don't do it 1: do it 2: do it aggressively
-nvptx-mad-enable - NVPTX Specific: Enable generating FMAD instructions
-nvptx-prec-divf32=<int> - NVPTX Specifies: 0 use div.approx, 1 use div.full, 2 use IEEE Compliant F32 div.rnd if avaiable.
-nvptx-remove-redundant-moves - NVPTX: Remove redundant moves introduced by vector lowering
-nvptx-sched4reg - NVPTX Specific: schedule for register pressue
-o=<filename> - Override output filename
-p - Print module after each transformation
-pre-RA-sched - Instruction schedulers available (before register allocation):
=vliw-td - VLIW scheduler
=list-ilp - Bottom-up register pressure aware list scheduling which tries to balance ILP and register pressure
=list-hybrid - Bottom-up register pressure aware list scheduling which tries to balance latency and register pressure
=source - Similar to list-burr but schedules in source order when possible
=list-burr - Bottom-up register reduction list scheduling
=linearize - Linearize DAG, no scheduling
=fast - Fast suboptimal list scheduling
=default - Best scheduler for the target
-print-after-all - Print IR after each pass
-print-before-all - Print IR before each pass
-print-breakpoints-for-testing - Print select breakpoints location for testing
-print-machineinstrs=<pass-name> - Print machine instrs
-profile-estimator-loop-weight=<loop-weight> - Number of loop executions used for profile-estimator
-profile-file=<filename> - Profile file loaded by -profile-metadata-loader
-profile-info-file=<filename> - Profile file loaded by -profile-loader
-profile-verifier-noassert - Disable assertions
-realign-stack - Realign stack if needed
-regalloc - Register allocator to use
=default - pick register allocator based on -O option
=basic - basic register allocator
=fast - fast register allocator
=greedy - greedy register allocator
=pbqp - PBQP register allocator
-relocation-model - Choose relocation model
=default - Target default relocation model
=static - Non-relocatable code
=pic - Fully relocatable, position independent code
=dynamic-no-pic - Relocatable external references, non-relocatable code
-segmented-stacks - Use segmented stacks if possible.
-shrink-wrap - Shrink wrap callee-saved register spills/restores
-soft-float - Generate software floating point library calls
-spiller - Spiller to use: (default: standard)
=trivial - trivial spiller
=inline - inline spiller
-stack-alignment=<uint> - Override default stack alignment
-stack-protector-buffer-size=<uint> - Lower bound for a buffer to be considered for stack protection
-start-after=<pass-name> - Resume compilation after a specific pass
-stats - Enable statistics output from program
-std-compile-opts - Include the standard compile time optimizations
-std-link-opts - Include the standard link time optimizations
-stop-after=<pass-name> - Stop compilation after a specific pass
-strip-debug - Strip debugger symbol info from translation unit
-tailcallopt - Turn fastcc calls into tail calls by (potentially) changing ABI.
-time-passes - Time each pass, printing elapsed time for each on exit
-use-init-array - Use .init_array instead of .ctors.
-vectorize - Run the BB vectorization passes
-vectorize-loops - Run the Loop vectorization passes
-verify-dom-info - Verify dominator info (time consuming)
-verify-each - Verify after each transform
-verify-loop-info - Verify loop info (time consuming)
-verify-regalloc - Verify during register allocation
-verify-region-info - Verify region info (time consuming)
-verify-scev - Verify ScalarEvolution's backedge taken counts (slow)
-version - Display the version of this program
-x86-asm-syntax - Choose style of code to emit from X86 backend:
=att - Emit AT&T-style assembly
=intel - Emit Intel-style assembly
-x86-early-ifcvt - Enable early if-conversion on X86
-x86-use-vzeroupper - Minimize AVX to SSE transition penalty
