LLVM’s Analysis and Transform Passes

Introduction

This document serves as a high level summary of the optimization features thatLLVM provides. Optimizations are implemented as Passes that traverse someportion of a program to either collect information or transform the program.The table below divides the passes that LLVM provides into three categories.Analysis passes compute information that other passes can use or for debuggingor program visualization purposes. Transform passes can use (or invalidate)the analysis passes. Transform passes all mutate the program in some way.Utility passes provides some utility but don’t otherwise fit categorization.For example passes to extract functions to bitcode or write a module to bitcodeare neither analysis nor transform passes. The table of contents aboveprovides a quick summary of each pass and links to the more complete passdescription later in the document.

Analysis Passes

This section describes the LLVM Analysis Passes.

-aa-eval: Exhaustive Alias Analysis Precision Evaluator

This is a simple N^2 alias analysis accuracy evaluator. Basically, for eachfunction in the program, it simply queries to see how the alias analysisimplementation answers alias queries between each pair of pointers in thefunction.

This is inspired and adapted from code by: Naveen Neelakantam, FrancescoSpadini, and Wojciech Stryjewski.

-basicaa: Basic Alias Analysis (stateless AA impl)

A basic alias analysis pass that implements identities (two different globalscannot alias, etc), but does no stateful analysis.

-basiccg: Basic CallGraph Construction

Yet to be written.

-count-aa: Count Alias Analysis Query Responses

A pass which can be used to count how many alias queries are being made and howthe alias analysis implementation being used responds.

-da: Dependence Analysis

Dependence analysis framework, which is used to detect dependences in memoryaccesses.

-debug-aa: AA use debugger

This simple pass checks alias analysis users to ensure that if they create anew value, they do not query AA without informing it of the value. It acts asa shim over any other AA pass you want.

Yes keeping track of every value in the program is expensive, but this is adebugging pass.

-domfrontier: Dominance Frontier Construction

This pass is a simple dominator construction algorithm for finding forwarddominator frontiers.

-domtree: Dominator Tree Construction

This pass is a simple dominator construction algorithm for finding forwarddominators.

-dot-callgraph: Print Call Graph to “dot” file

This pass, only available in opt, prints the call graph into a .dotgraph. This graph can then be processed with the “dot” tool to convert it topostscript or some other suitable format.

-dot-cfg: Print CFG of function to “dot” file

This pass, only available in opt, prints the control flow graph into a.dot graph. This graph can then be processed with the dot toolto convert it to postscript or some other suitable format.

-dot-cfg-only: Print CFG of function to “dot” file (with no function bodies)

This pass, only available in opt, prints the control flow graph into a.dot graph, omitting the function bodies. This graph can then be processedwith the dot tool to convert it to postscript or some other suitableformat.

-dot-dom: Print dominance tree of function to “dot” file

This pass, only available in opt, prints the dominator tree into a .dotgraph. This graph can then be processed with the dot tool toconvert it to postscript or some other suitable format.

-dot-dom-only: Print dominance tree of function to “dot” file (with no function bodies)

This pass, only available in opt, prints the dominator tree into a .dotgraph, omitting the function bodies. This graph can then be processed with thedot tool to convert it to postscript or some other suitable format.

-dot-postdom: Print postdominance tree of function to “dot” file

This pass, only available in opt, prints the post dominator tree into a.dot graph. This graph can then be processed with the dot toolto convert it to postscript or some other suitable format.

-dot-postdom-only: Print postdominance tree of function to “dot” file (with no function bodies)

This pass, only available in opt, prints the post dominator tree into a.dot graph, omitting the function bodies. This graph can then be processedwith the dot tool to convert it to postscript or some other suitableformat.

-globalsmodref-aa: Simple mod/ref analysis for globals

This simple pass provides alias and mod/ref information for global values thatdo not have their address taken, and keeps track of whether functions read orwrite memory (are “pure”). For this simple (but very common) case, we canprovide pretty accurate and useful information.

-instcount: Counts the various types of Instructions

This pass collects the count of all instructions and reports them.

-intervals: Interval Partition Construction

This analysis calculates and represents the interval partition of a function,or a preexisting interval partition.

In this way, the interval partition may be used to reduce a flow graph down toits degenerate single node interval partition (unless it is irreducible).

-iv-users: Induction Variable Users

Bookkeeping for “interesting” users of expressions computed from inductionvariables.

-lazy-value-info: Lazy Value Information Analysis

Interface for lazy computation of value constraint information.

-libcall-aa: LibCall Alias Analysis

LibCall Alias Analysis.

-lint: Statically lint-checks LLVM IR

This pass statically checks for common and easily-identified constructs whichproduce undefined or likely unintended behavior in LLVM IR.

It is not a guarantee of correctness, in two ways. First, it isn’tcomprehensive. There are checks which could be done statically which are notyet implemented. Some of these are indicated by TODO comments, but thosearen’t comprehensive either. Second, many conditions cannot be checkedstatically. This pass does no dynamic instrumentation, so it can’t check forall possible problems.

Another limitation is that it assumes all code will be executed. A storethrough a null pointer in a basic block which is never reached is harmless, butthis pass will warn about it anyway.

Optimization passes may make conditions that this pass checks for more or lessobvious. If an optimization pass appears to be introducing a warning, it maybe that the optimization pass is merely exposing an existing condition in thecode.

This code may be run before instcombine. In manycases, instcombine checks for the same kinds of things and turns instructionswith undefined behavior into unreachable (or equivalent). Because of this,this pass makes some effort to look through bitcasts and so on.

-loops: Natural Loop Information

This analysis is used to identify natural loops and determine the loop depth ofvarious nodes of the CFG. Note that the loops identified may actually beseveral natural loops that share the same header node… not just a singlenatural loop.

-memdep: Memory Dependence Analysis

An analysis that determines, for a given memory operation, what precedingmemory operations it depends on. It builds on alias analysis information, andtries to provide a lazy, caching interface to a common kind of aliasinformation query.

-module-debuginfo: Decodes module-level debug info

This pass decodes the debug info metadata in a module and prints in a(sufficiently-prepared-) human-readable form.

For example, run this pass from opt along with the -analyze option, andit’ll print to standard output.

-postdomfrontier: Post-Dominance Frontier Construction

This pass is a simple post-dominator construction algorithm for findingpost-dominator frontiers.

-postdomtree: Post-Dominator Tree Construction

This pass is a simple post-dominator construction algorithm for findingpost-dominators.

-print-alias-sets: Alias Set Printer

Yet to be written.

-print-callgraph: Print a call graph

This pass, only available in opt, prints the call graph to standard errorin a human-readable form.

-print-callgraph-sccs: Print SCCs of the Call Graph

This pass, only available in opt, prints the SCCs of the call graph tostandard error in a human-readable form.

-print-cfg-sccs: Print SCCs of each function CFG

This pass, only available in opt, printsthe SCCs of each function CFG tostandard error in a human-readable fom.

-print-dom-info: Dominator Info Printer

Dominator Info Printer.

-print-externalfnconstants: Print external fn callsites passed constants

This pass, only available in opt, prints out call sites to externalfunctions that are called with constant arguments. This can be useful whenlooking for standard library functions we should constant fold or handle inalias analyses.

-print-function: Print function to stderr

The PrintFunctionPass class is designed to be pipelined with otherFunctionPasses, and prints out the functions of the module as they areprocessed.

-print-module: Print module to stderr

This pass simply prints out the entire module when it is executed.

-print-used-types: Find Used Types

This pass is used to seek out all of the types in use by the program. Notethat this analysis explicitly does not include types only used by the symboltable.

-regions: Detect single entry single exit regions

The RegionInfo pass detects single entry single exit regions in a function,where a region is defined as any subgraph that is connected to the remaininggraph at only two spots. Furthermore, a hierarchical region tree is built.

-scalar-evolution: Scalar Evolution Analysis

The ScalarEvolution analysis can be used to analyze and catagorize scalarexpressions in loops. It specializes in recognizing general inductionvariables, representing them with the abstract and opaque SCEV class.Given this analysis, trip counts of loops and other important properties can beobtained.

This analysis is primarily useful for induction variable substitution andstrength reduction.

-scev-aa: ScalarEvolution-based Alias Analysis

Simple alias analysis implemented in terms of ScalarEvolution queries.

This differs from traditional loop dependence analysis in that it tests fordependencies within a single iteration of a loop, rather than dependenciesbetween different iterations.

ScalarEvolution has a more complete understanding of pointer arithmeticthan BasicAliasAnalysis’ collection of ad-hoc analyses.

-stack-safety: Stack Safety Analysis

The StackSafety analysis can be used to determine if stack allocatedvariables can be considered safe from memory access bugs.

This analysis’ primary purpose is to be used by sanitizers to avoid unnecessaryinstrumentation of safe variables.

-targetdata: Target Data Layout

Provides other passes access to information on how the size and alignmentrequired by the target ABI for various data types.

Transform Passes

This section describes the LLVM Transform Passes.

-adce: Aggressive Dead Code Elimination

ADCE aggressively tries to eliminate code. This pass is similar to DCE but it assumes that values are dead until proven otherwise. Thisis similar to SCCP, except applied to the liveness ofvalues.

-always-inline: Inliner for always_inline functions

A custom inliner that handles only functions that are marked as “alwaysinline”.

-argpromotion: Promote ‘by reference’ arguments to scalars

This pass promotes “by reference” arguments to be “by value” arguments. Inpractice, this means looking for internal functions that have pointerarguments. If it can prove, through the use of alias analysis, that anargument is only loaded, then it can pass the value into the function insteadof the address of the value. This can cause recursive simplification of codeand lead to the elimination of allocas (especially in C++ template code likethe STL).

This pass also handles aggregate arguments that are passed into a function,scalarizing them if the elements of the aggregate are only loaded. Note thatit refuses to scalarize aggregates which would require passing in more thanthree operands to the function, because passing thousands of operands for alarge array or structure is unprofitable!

Note that this transformation could also be done for arguments that are onlystored to (returning the value instead), but does not currently. This casewould be best handled when and if LLVM starts supporting multiple return valuesfrom functions.

-bb-vectorize: Basic-Block Vectorization

This pass combines instructions inside basic blocks to form vectorinstructions. It iterates over each basic block, attempting to pair compatibleinstructions, repeating this process until no additional pairs are selected forvectorization. When the outputs of some pair of compatible instructions areused as inputs by some other pair of compatible instructions, those pairs arepart of a potential vectorization chain. Instruction pairs are only fused intovector instructions when they are part of a chain longer than some thresholdlength. Moreover, the pass attempts to find the best possible chain for eachpair of compatible instructions. These heuristics are intended to preventvectorization in cases where it would not yield a performance increase of theresulting code.

-block-placement: Profile Guided Basic Block Placement

This pass is a very simple profile guided basic block placement algorithm. Theidea is to put frequently executed blocks together at the start of the functionand hopefully increase the number of fall-through conditional branches. Ifthere is no profile information for a particular function, this pass basicallyorders blocks in depth-first order.

-break-crit-edges: Break critical edges in CFG

Break all of the critical edges in the CFG by inserting a dummy basic block.It may be “required” by passes that cannot deal with critical edges. Thistransformation obviously invalidates the CFG, but can update forward dominator(set, immediate dominators, tree, and frontier) information.

-codegenprepare: Optimize for code generation

This pass munges the code in the input function to better prepare it forSelectionDAG-based code generation. This works around limitations in itsbasic-block-at-a-time approach. It should eventually be removed.

-constmerge: Merge Duplicate Global Constants

Merges duplicate global constants together into a single constant that isshared. This is useful because some passes (i.e., TraceValues) insert a lot ofstring constants into the program, regardless of whether or not an existingstring is available.

-constprop: Simple constant propagation

This pass implements constant propagation and merging. It looks forinstructions involving only constant operands and replaces them with a constantvalue instead of an instruction. For example:

add i32 1, 2

becomes

i32 3

NOTE: this pass has a habit of making definitions be dead. It is a good ideato run a Dead Instruction Elimination pass sometime afterrunning this pass.

-dce: Dead Code Elimination

Dead code elimination is similar to dead instruction elimination, but it rechecks instructions that were used by removedinstructions to see if they are newly dead.

-deadargelim: Dead Argument Elimination

This pass deletes dead arguments from internal functions. Dead argumentelimination removes arguments which are directly dead, as well as argumentsonly passed into function calls as dead arguments of other functions. Thispass also deletes dead arguments in a similar way.

This pass is often useful as a cleanup pass to run after aggressiveinterprocedural passes, which add possibly-dead arguments.

-deadtypeelim: Dead Type Elimination

This pass is used to cleanup the output of GCC. It eliminate names for typesthat are unused in the entire translation unit, using the find used types pass.

-die: Dead Instruction Elimination

Dead instruction elimination performs a single pass over the function, removinginstructions that are obviously dead.

-dse: Dead Store Elimination

A trivial dead store elimination that only considers basic-block localredundant stores.

-functionattrs: Deduce function attributes

A simple interprocedural pass which walks the call-graph, looking for functionswhich do not access or only read non-local memory, and marking themreadnone/readonly. In addition, it marks function arguments (ofpointer type) “nocapture” if a call to the function does not create anycopies of the pointer value that outlive the call. This more or less meansthat the pointer is only dereferenced, and not returned from the function orstored in a global. This pass is implemented as a bottom-up traversal of thecall-graph.

-globaldce: Dead Global Elimination

This transform is designed to eliminate unreachable internal globals from theprogram. It uses an aggressive algorithm, searching out globals that are knownto be alive. After it finds all of the globals which are needed, it deleteswhatever is left over. This allows it to delete recursive chunks of theprogram which are unreachable.

-globalopt: Global Variable Optimizer

This pass transforms simple global variables that never have their addresstaken. If obviously true, it marks read/write globals as constant, deletesvariables only stored to, etc.

-gvn: Global Value Numbering

This pass performs global value numbering to eliminate fully and partiallyredundant instructions. It also performs redundant load elimination.

-indvars: Canonicalize Induction Variables

This transformation analyzes and transforms the induction variables (andcomputations derived from them) into simpler forms suitable for subsequentanalysis and transformation.

This transformation makes the following changes to each loop with anidentifiable induction variable:

• All loops are transformed to have a single canonical induction variablewhich starts at zero and steps by one.
• The canonical induction variable is guaranteed to be the first PHI node inthe loop header block.
• Any pointer arithmetic recurrences are raised to use array subscripts.

If the trip count of a loop is computable, this pass also makes the followingchanges:

• The exit condition for the loop is canonicalized to compare the inductionvalue against the exit value. This turns loops like:

for (i = 7; i*i < 1000; ++i)
 
into

for (i = 0; i != 25; ++i)

• Any use outside of the loop of an expression derived from the indvar ischanged to compute the derived value outside of the loop, eliminating thedependence on the exit value of the induction variable. If the only purposeof the loop is to compute the exit value of some derived expression, thistransformation will make the loop dead.

This transformation should be followed by strength reduction after all of thedesired loop transformations have been performed. Additionally, on targetswhere it is profitable, the loop could be transformed to count down to zero(the “do loop” optimization).

-inline: Function Integration/Inlining

Bottom-up inlining of functions into callees.

-instcombine: Combine redundant instructions

Combine instructions to form fewer, simple instructions. This pass does notmodify the CFG. This pass is where algebraic simplification happens.

This pass combines things like:

%Y = add i32 %X, 1
%Z = add i32 %Y, 1

into:

%Z = add i32 %X, 2

This is a simple worklist driven algorithm.

This pass guarantees that the following canonicalizations are performed on theprogram:

• If a binary operator has a constant operand, it is moved to the right-handside.
• Bitwise operators with constant operands are always grouped so that shiftsare performed first, then ors, then ands, then xors.
• Compare instructions are converted from <, >, ≤, or ≥ to= or ≠ if possible.
• All cmp instructions on boolean values are replaced with logicaloperations.
• add X, X is represented as mul X, 2 ⇒ shl X, 1
• Multiplies with a constant power-of-two argument are transformed intoshifts.
• … etc.This pass can also simplify calls to specific well-known function calls (e.g.runtime library functions). For example, a call exit(3) that occurs withinthe main() function can be transformed into simply return 3. Whether ornot library calls are simplified is controlled by the-functionattrs pass and LLVM’s knowledge oflibrary calls on different targets.

-aggressive-instcombine: Combine expression patterns

Combine expression patterns to form expressions with fewer, simple instructions.This pass does not modify the CFG.

For example, this pass reduce width of expressions post-dominated by TruncInstinto smaller width when applicable.

It differs from instcombine pass in that it contains pattern optimization thatrequires higher complexity than the O(1), thus, it should run fewer times thaninstcombine pass.

-internalize: Internalize Global Symbols

This pass loops over all of the functions in the input module, looking for amain function. If a main function is found, all other functions and all globalvariables with initializers are marked as internal.

-ipconstprop: Interprocedural constant propagation

This pass implements an extremely simple interprocedural constant propagationpass. It could certainly be improved in many different ways, like using aworklist. This pass makes arguments dead, but does not remove them. Theexisting dead argument elimination pass should be run after this to clean upthe mess.

-ipsccp: Interprocedural Sparse Conditional Constant Propagation

An interprocedural variant of Sparse Conditional Constant Propagation.

-jump-threading: Jump Threading

Jump threading tries to find distinct threads of control flow running through abasic block. This pass looks at blocks that have multiple predecessors andmultiple successors. If one or more of the predecessors of the block can beproven to always cause a jump to one of the successors, we forward the edgefrom the predecessor to the successor by duplicating the contents of thisblock.

An example of when this can occur is code like this:

if () { ...
  X = 4;
}
if (X < 3) {

In this case, the unconditional branch at the end of the first if can berevectored to the false side of the second if.

-lcssa: Loop-Closed SSA Form Pass

This pass transforms loops by placing phi nodes at the end of the loops for allvalues that are live across the loop boundary. For example, it turns the leftinto the right code:

for (...)                for (...)
    if (c)                   if (c)
        X1 = ...                 X1 = ...
    else                     else
        X2 = ...                 X2 = ...
    X3 = phi(X1, X2)         X3 = phi(X1, X2)
... = X3 + 4              X4 = phi(X3)
                            ... = X4 + 4

This is still valid LLVM; the extra phi nodes are purely redundant, and will betrivially eliminated by InstCombine. The major benefit of thistransformation is that it makes many other loop optimizations, such asLoopUnswitching, simpler.

-licm: Loop Invariant Code Motion

This pass performs loop invariant code motion, attempting to remove as muchcode from the body of a loop as possible. It does this by either hoisting codeinto the preheader block, or by sinking code to the exit blocks if it is safe.This pass also promotes must-aliased memory locations in the loop to live inregisters, thus hoisting and sinking “invariant” loads and stores.

This pass uses alias analysis for two purposes:

• Moving loop invariant loads and calls out of loops. If we can determinethat a load or call inside of a loop never aliases anything stored to, wecan hoist it or sink it like any other instruction.

• Scalar Promotion of Memory. If there is a store instruction inside of theloop, we try to move the store to happen AFTER the loop instead of inside ofthe loop. This can only happen if a few conditions are true:

  • The pointer stored through is loop invariant.
  • There are no stores or loads in the loop which may alias the pointer.There are no calls in the loop which mod/ref the pointer.If these conditions are true, we can promote the loads and stores in theloop of the pointer to use a temporary alloca’d variable. We then use themem2reg functionality to construct the appropriateSSA form for the variable.

-loop-deletion: Delete dead loops

This file implements the Dead Loop Deletion Pass. This pass is responsible foreliminating loops with non-infinite computable trip counts that have no sideeffects or volatile instructions, and do not contribute to the computation ofthe function’s return value.

-loop-extract: Extract loops into new functions

A pass wrapper around the ExtractLoop() scalar transformation to extracteach top-level loop into its own new function. If the loop is the only loopin a given function, it is not touched. This is a pass most useful fordebugging via bugpoint.

-loop-extract-single: Extract at most one loop into a new function

Similar to Extract loops into new functions, thispass extracts one natural loop from the program into a function if it can.This is used by bugpoint.

-loop-reduce: Loop Strength Reduction

This pass performs a strength reduction on array references inside loops thathave as one or more of their components the loop induction variable. This isaccomplished by creating a new value to hold the initial value of the arrayaccess for the first iteration, and then creating a new GEP instruction in theloop to increment the value by the appropriate amount.

-loop-rotate: Rotate Loops

A simple loop rotation transformation. A summary of it can be found inLoop Terminology for Rotated Loops.

-loop-simplify: Canonicalize natural loops

This pass performs several transformations to transform natural loops into asimpler form, which makes subsequent analyses and transformations simpler andmore effective. A summary of it can be found inLoop Terminology, Loop Simplify Form.

Loop pre-header insertion guarantees that there is a single, non-critical entryedge from outside of the loop to the loop header. This simplifies a number ofanalyses and transformations, such as LICM.

Loop exit-block insertion guarantees that all exit blocks from the loop (blockswhich are outside of the loop that have predecessors inside of the loop) onlyhave predecessors from inside of the loop (and are thus dominated by the loopheader). This simplifies transformations such as store-sinking that are builtinto LICM.

This pass also guarantees that loops will have exactly one backedge.

Note that the simplifycfg pass will clean up blockswhich are split out but end up being unnecessary, so usage of this pass shouldnot pessimize generated code.

This pass obviously modifies the CFG, but updates loop information anddominator information.

-loop-unroll: Unroll loops

This pass implements a simple loop unroller. It works best when loops havebeen canonicalized by the indvars pass, allowing it todetermine the trip counts of loops easily.

-loop-unroll-and-jam: Unroll and Jam loops

This pass implements a simple unroll and jam classical loop optimisation pass.It transforms loop from:

for i.. i+= 1              for i.. i+= 4
  for j..                    for j..
    code(i, j)                 code(i, j)
                               code(i+1, j)
                               code(i+2, j)
                               code(i+3, j)
                           remainder loop

Which can be seen as unrolling the outer loop and “jamming” (fusing) the innerloops into one. When variables or loads can be shared in the new inner loop, thiscan lead to significant performance improvements. It usesDependence Analysis for proving the transformations are safe.

-loop-unswitch: Unswitch loops

This pass transforms loops that contain branches on loop-invariant conditionsto have multiple loops. For example, it turns the left into the right code:

for (...)                  if (lic)
    A                          for (...)
    if (lic)                       A; B; C
        B                  else
    C                          for (...)
                                   A; C

This can increase the size of the code exponentially (doubling it every time aloop is unswitched) so we only unswitch if the resultant code will be smallerthan a threshold.

This pass expects LICM to be run before it to hoistinvariant conditions out of the loop, to make the unswitching opportunityobvious.

-loweratomic: Lower atomic intrinsics to non-atomic form

This pass lowers atomic intrinsics to non-atomic form for use in a knownnon-preemptible environment.

The pass does not verify that the environment is non-preemptible (in generalthis would require knowledge of the entire call graph of the program includingany libraries which may not be available in bitcode form); it simply lowersevery atomic intrinsic.

-lowerinvoke: Lower invokes to calls, for unwindless code generators

This transformation is designed for use by code generators which do not yetsupport stack unwinding. This pass converts invoke instructions tocall instructions, so that any exception-handling landingpad blocksbecome dead code (which can be removed by running the -simplifycfg passafterwards).

-lowerswitch: Lower SwitchInsts to branches

Rewrites switch instructions with a sequence of branches, which allows targetsto get away with not implementing the switch instruction until it isconvenient.

-mem2reg: Promote Memory to Register

This file promotes memory references to be register references. It promotesalloca instructions which only have loads and stores as uses. An alloca istransformed by using dominator frontiers to place phi nodes, then traversingthe function in depth-first order to rewrite loads and stores as appropriate.This is just the standard SSA construction algorithm to construct “pruned” SSAform.

-memcpyopt: MemCpy Optimization

This pass performs various transformations related to eliminating memcpycalls, or transforming sets of stores into memsets.

-mergefunc: Merge Functions

This pass looks for equivalent functions that are mergable and folds them.

Total-ordering is introduced among the functions set: we define comparisonthat answers for every two functions which of them is greater. It allows toarrange functions into the binary tree.

For every new function we check for equivalent in tree.

If equivalent exists we fold such functions. If both functions are overridable,we move the functionality into a new internal function and leave twooverridable thunks to it.

If there is no equivalent, then we add this function to tree.

Lookup routine has O(log(n)) complexity, while whole merging process hascomplexity of O(n*log(n)).

Readthisarticle for more details.

-mergereturn: Unify function exit nodes

Ensure that functions have at most one ret instruction in them.Additionally, it keeps track of which node is the new exit node of the CFG.

-partial-inliner: Partial Inliner

This pass performs partial inlining, typically by inlining an if statementthat surrounds the body of the function.

-prune-eh: Remove unused exception handling info

This file implements a simple interprocedural pass which walks the call-graph,turning invoke instructions into call instructions if and only if the calleecannot throw an exception. It implements this as a bottom-up traversal of thecall-graph.

-reassociate: Reassociate expressions

This pass reassociates commutative expressions in an order that is designed topromote better constant propagation, GCSE, LICM, PRE, etc.

For example: 4 + (x + 5) ⇒ x + (4 + 5)

In the implementation of this algorithm, constants are assigned rank = 0,function arguments are rank = 1, and other values are assigned rankscorresponding to the reverse post order traversal of current function (startingat 2), which effectively gives values in deep loops higher rank than values notin loops.

-reg2mem: Demote all values to stack slots

This file demotes all registers to memory references. It is intended to be theinverse of mem2reg. By converting to loadinstructions, the only values live across basic blocks are allocainstructions and load instructions before phi nodes. It is intendedthat this should make CFG hacking much easier. To make later hacking easier,the entry block is split into two, such that all introduced allocainstructions (and nothing else) are in the entry block.

-sroa: Scalar Replacement of Aggregates

The well-known scalar replacement of aggregates transformation. This transformbreaks up alloca instructions of aggregate type (structure or array) intoindividual alloca instructions for each member if possible. Then, ifpossible, it transforms the individual alloca instructions into nice cleanscalar SSA form.

-sccp: Sparse Conditional Constant Propagation

Sparse conditional constant propagation and merging, which can be summarizedas:

• Assumes values are constant unless proven otherwise
• Assumes BasicBlocks are dead unless proven otherwise
• Proves values to be constant, and replaces them with constants
• Proves conditional branches to be unconditional

Note that this pass has a habit of making definitions be dead. It is a goodidea to run a DCE pass sometime after running this pass.

-simplifycfg: Simplify the CFG

Performs dead code elimination and basic block merging. Specifically:

• Removes basic blocks with no predecessors.
• Merges a basic block into its predecessor if there is only one and thepredecessor only has one successor.
• Eliminates PHI nodes for basic blocks with a single predecessor.
• Eliminates a basic block that only contains an unconditional branch.

-sink: Code sinking

This pass moves instructions into successor blocks, when possible, so that theyaren’t executed on paths where their results aren’t needed.

-strip: Strip all symbols from a module

Performs code stripping. This transformation can delete:

• names for virtual registers
• symbols for internal globals and functions
• debug information

Note that this transformation makes code much less readable, so it should onlybe used in situations where the strip utility would be used, such as reducingcode size or making it harder to reverse engineer code.

-strip-dead-debug-info: Strip debug info for unused symbols

performs code stripping. this transformation can delete:

• names for virtual registers
• symbols for internal globals and functions
• debug information

note that this transformation makes code much less readable, so it should onlybe used in situations where the strip utility would be used, such as reducingcode size or making it harder to reverse engineer code.

-strip-dead-prototypes: Strip Unused Function Prototypes

This pass loops over all of the functions in the input module, looking for deaddeclarations and removes them. Dead declarations are declarations of functionsfor which no implementation is available (i.e., declarations for unused libraryfunctions).

-strip-debug-declare: Strip all llvm.dbg.declare intrinsics

This pass implements code stripping. Specifically, it can delete:

• names for virtual registers
• symbols for internal globals and functions
• debug informationNote that this transformation makes code much less readable, so it should onlybe used in situations where the ‘strip’ utility would be used, such as reducingcode size or making it harder to reverse engineer code.

-strip-nondebug: Strip all symbols, except dbg symbols, from a module

This pass implements code stripping. Specifically, it can delete:

• names for virtual registers
• symbols for internal globals and functions
• debug informationNote that this transformation makes code much less readable, so it should onlybe used in situations where the ‘strip’ utility would be used, such as reducingcode size or making it harder to reverse engineer code.

-tailcallelim: Tail Call Elimination

This file transforms calls of the current function (self recursion) followed bya return instruction with a branch to the entry of the function, creating aloop. This pass also implements the following extensions to the basicalgorithm:

• Trivial instructions between the call and return do not prevent thetransformation from taking place, though currently the analysis cannotsupport moving any really useful instructions (only dead ones).
• This pass transforms functions that are prevented from being tail recursiveby an associative expression to use an accumulator variable, thus compilingthe typical naive factorial or fib implementation into efficient code.
• TRE is performed if the function returns void, if the return returns theresult returned by the call, or if the function returns a run-time constanton all exits from the function. It is possible, though unlikely, that thereturn returns something else (like constant 0), and can still be TRE’d. Itcan be TRE’d if all other return instructions in the function return theexact same value.
• If it can prove that callees do not access theier caller stack frame, theyare marked as eligible for tail call elimination (by the code generator).

Utility Passes

This section describes the LLVM Utility Passes.

-deadarghaX0r: Dead Argument Hacking (BUGPOINT USE ONLY; DO NOT USE)

Same as dead argument elimination, but deletes arguments to functions which areexternal. This is only for use by bugpoint.

-extract-blocks: Extract Basic Blocks From Module (for bugpoint use)

This pass is used by bugpoint to extract all blocks from the module into theirown functions.

-instnamer: Assign names to anonymous instructions

This is a little utility pass that gives instructions names, this is mostlyuseful when diffing the effect of an optimization because deleting an unnamedinstruction can change all other instruction numbering, making the diff verynoisy.

-verify: Module Verifier

Verifies an LLVM IR code. This is useful to run after an optimization which isundergoing testing. Note that llvm-as verifies its input before emittingbitcode, and also that malformed bitcode is likely to make LLVM crash. Alllanguage front-ends are therefore encouraged to verify their output beforeperforming optimizing transformations.

• Both of a binary operator’s parameters are of the same type.
• Verify that the indices of mem access instructions match other operands.
• Verify that arithmetic and other things are only performed on first-classtypes. Verify that shifts and logicals only happen on integrals f.e.
• All of the constants in a switch statement are of the correct type.
• The code is in valid SSA form.
• It is illegal to put a label into any other type (like a structure) or toreturn one.
• Only phi nodes can be self referential: %x = add i32 %x, %x isinvalid.
• PHI nodes must have an entry for each predecessor, with no extras.
• PHI nodes must be the first thing in a basic block, all grouped together.
• PHI nodes must have at least one entry.
• All basic blocks should only end with terminator insts, not contain them.
• The entry node to a function must not have predecessors.
• All Instructions must be embedded into a basic block.
• Functions cannot take a void-typed parameter.
• Verify that a function’s argument list agrees with its declared type.
• It is illegal to specify a name for a void value.
• It is illegal to have an internal global value with no initializer.
• It is illegal to have a ret instruction that returns a value that doesnot agree with the function return value type.
• Function call argument types match the function prototype.
• All other things that are tested by asserts spread about the code.Note that this does not provide full security verification (like Java), butinstead just tries to ensure that code is well-formed.

-view-cfg: View CFG of function

Displays the control flow graph using the GraphViz tool.

-view-cfg-only: View CFG of function (with no function bodies)

Displays the control flow graph using the GraphViz tool, but omitting functionbodies.

-view-dom: View dominance tree of function

Displays the dominator tree using the GraphViz tool.

-view-dom-only: View dominance tree of function (with no function bodies)

Displays the dominator tree using the GraphViz tool, but omitting functionbodies.

-view-postdom: View postdominance tree of function

Displays the post dominator tree using the GraphViz tool.

-view-postdom-only: View postdominance tree of function (with no function bodies)

Displays the post dominator tree using the GraphViz tool, but omitting functionbodies.

-transform-warning: Report missed forced transformations

Emits warnings about not yet applied forced transformations (e.g. from#pragma omp simd).