Alias Analysis Infrastructure in LLVM |
- Introduction
- AliasAnalysis Overview
- Writing a new AliasAnalysis Implementation
- Using AliasAnalysis results
- Helpful alias analysis related tools
Written by Chris Lattner
Alias Analysis (or Pointer Analysis) is a technique which attempts to determine
whether or not two pointers ever can point to the same object in memory.
Traditionally, Alias Analyses respond to a query with either a Must, May, or No alias response, indicating that two
pointers do point to the same object, might point to the same object, or are
known not to point to the same object.
The AliasAnalysis class is the
centerpiece of the LLVM Alias Analysis related infrastructure. This class is
the common interface between clients of alias analysis information and the
implementations providing it. In addition to simple alias analysis information,
this class exposes Mod/Ref information from those implementations which can
provide it, allowing for powerful analyses and transformations to work well
together.
This document contains information necessary to successfully implement this
interface, use it, and to test both sides. It also explains some of the finer
points about what exactly results mean. If you feel that something is unclear
or should be added, please let me know.
The AliasAnalysis class defines
the interface that Alias Analysis implementations should support. This class
exports two important enums: AliasResult and ModRefResult
which represent the result of an alias query or a mod/ref query,
respectively.
The AliasAnalysis interface exposes information about memory, represented in
several different ways. In particular, memory objects are represented as a
starting address and size, and function calls are represented as the actual
call or invoke instructions that performs the call. The
AliasAnalysis interface also exposes some helper methods which allow you to get
mod/ref information for arbitrary instructions.
Most importantly, the AliasAnalysis class provides several methods which are
used to query whether or not pointers alias, whether function calls can modify
or read memory, etc.
Representing memory objects as a starting address and a size is critically
important for precise Alias Analyses. For example, consider this (silly) C
code:
int i;
char C[2];
char A[10];
/* ... */
for (i = 0; i != 10; ++i) {
C[0] = A[i]; /* One byte store */
C[1] = A[9-i]; /* One byte store */
}
In this case, the basicaa pass will disambiguate the stores to
C[0] and C[1] because they are accesses to two distinct
locations one byte apart, and the accesses are each one byte. In this case, the
LICM pass can use store motion to remove the stores from the loop. In
constrast, the following code:
int i;
char C[2];
char A[10];
/* ... */
for (i = 0; i != 10; ++i) {
((short*)C)[0] = A[i]; /* Two byte store! */
C[1] = A[9-i]; /* One byte store */
}
In this case, the two stores to C do alias each other, because the access to the
&C[0] element is a two byte access. If size information wasn't
available in the query, even the first case would have to conservatively assume
that the accesses alias.
An Alias Analysis implementation can return one of three responses: MustAlias,
MayAlias, and NoAlias. The No and May alias results are obvious: if the two
pointers may never equal each other, return NoAlias, if they might, return
MayAlias.
The Must Alias response is trickier though. In LLVM, the Must Alias response
may only be returned if the two memory objects are guaranteed to always start at
exactly the same location. If two memory objects overlap, but do not start at
the same location, MayAlias must be returned.
The getModRefInfo methods return information about whether the
execution of an instruction can read or modify a memory location. Mod/Ref
information is always conservative: if an action may read a location, Ref
is returned.
Writing a new alias analysis implementation for LLVM is quite straight-forward.
There are already several implementations that you can use for examples, and the
following information should help fill in any details. For a minimal example,
take a look at the no-aa
implementation.
The first step to determining what type of LLVM
pass you need to use for your Alias Analysis. As is the case with most
other analyses and transformations, the answer should be fairly obvious from
what type of problem you are trying to solve:
- If you require interprocedural analysis, it should be a Pass.
- If you are a global analysis, subclass FunctionPass.
- If you are a local pass, subclass BasicBlockPass.
- If you don't need to look at the program at all, subclass
ImmutablePass.
In addition to the pass that you subclass, you should also inherit from the
AliasAnalysis interface, of course, and use the
RegisterAnalysisGroup template to register as an implementation of
AliasAnalysis.
Your subclass of AliasAnalysis is required to invoke two methods on the
AliasAnalysis base class: getAnalysisUsage and
InitializeAliasAnalysis. In particular, your implementation of
getAnalysisUsage should explicitly call into the
AliasAnalysis::getAnalysisUsage method in addition to doing any
declaring any pass dependencies your pass has. Thus you should have something
like this:
void getAnalysisUsage(AnalysisUsage &AU) const {
AliasAnalysis::getAnalysisUsage(AU);
// declare your dependencies here.
}
Additionally, your must invoke the InitializeAliasAnalysis method from
your analysis run method (run for a Pass,
runOnFunction for a FunctionPass, runOnBasicBlock for
a BasicBlockPass, or InitializeAliasAnalysis for an
ImmutablePass). For example (as part of a Pass):
bool run(Module &M) {
InitializeAliasAnalysis(this);
// Perform analysis here...
return false;
}
All of the AliasAnalysis virtual
methods default to providing conservatively correct information (returning "May"
Alias and "Mod/Ref" for alias and mod/ref queries respectively). Depending on
the capabilities of the analysis you are implementing, you just override the
interfaces you can improve.
With only two special exceptions (the basicaa and no-aa passes) every alias analysis pass should chain
to another alias analysis implementation (for example, you could specify
"-basic-aa -ds-aa -andersens-aa -licm" to get the maximum benefit from
the three alias analyses). To do this, simply "Require" AliasAnalysis in your
getAnalysisUsage method, and if you need to return a conservative
MayAlias or Mod/Ref result, simply chain to a lower analysis.
From the LLVM perspective, the only thing you need to do to provide an efficient
alias analysis is to make sure that alias analysis queries are serviced
quickly. The actual calculation of the alias analysis results (the "run"
method) is only performed once, but many (perhaps duplicate) queries may be
performed. Because of this, try to move as much computation to the run method
as possible (within reason).
There are several different ways to use alias analysis results. In order of
preference, these are...
The load-vn pass uses alias analysis to provide value numbering
information for load instructions. If your analysis or transformation
can be modelled in a form that uses value numbering information, you don't have
to do anything special to handle load instructions: just use the
load-vn pass, which uses alias analysis.
Many transformations need information about alias sets that are active in
some scope, rather than information about pairwise aliasing. The AliasSetTracker class is used
to efficiently build these Alias Sets from the pairwise alias analysis
information provided by the AliasAnalysis interface.
First you initialize the AliasSetTracker by use the "add" methods to
add information about various potentially aliasing instructions in the scope you
are interested in. Once all of the alias sets are completed, your pass should
simply iterate through the constructed alias sets, using the AliasSetTracker
begin()/end() methods.
The AliasSets formed by the AliasSetTracker are guaranteed to
be disjoint, calculate mod/ref information for the set, and keep track of
whether or not all of the pointers in the set are Must aliases. The
AliasSetTracker also makes sure that sets are properly folded due to call
instructions, and can provide a list of pointers in each set.
As an example user of this, the Loop
Invariant Code Motion pass uses AliasSetTrackers to build alias information
about each loop nest. If an AliasSet in a loop is not modified, then all load
instructions from that set may be hoisted out of the loop. If any alias sets
are stored and are must alias sets, then the stores may be sunk to
outside of the loop. Both of these transformations obviously only apply if the
pointer argument is loop-invariant.
As a last resort, your pass could use the AliasAnalysis interface directly to
service your pass. If you find the need to do this, please let me know so I can see if something new
needs to be added to LLVM.
If you're going to be working with the AliasAnalysis infrastructure, there are
several nice tools that may be useful for you and are worth knowing about...
The -no-aa analysis is just like what it sounds: an alias analysis that
never returns any useful information. This pass can be useful if you think that
alias analysis is doing something wrong and are trying to narrow down a problem.
If you don't specify an alias analysis, the default will be to use the
basicaa pass which does quite a bit of disambiguation on its own.
The -print-alias-sets pass is exposed as part of the analyze
tool to print out the Alias Sets formed by the AliasSetTracker class. This is useful if you're using
the AliasSetTracker.
The -count-aa pass is useful to see how many queries a particular pass
is making and what kinds of responses are returned by the alias analysis. An
example usage is:
$ opt -basicaa -count-aa -ds-aa -count-aa -licm
Which will print out how many queries (and what responses are returned) by the
-licm pass (of the -ds-aa pass) and how many queries are made
of the -basicaa pass by the -ds-aa pass. This can be useful
when evaluating an alias analysis for precision.
The -aa-eval pass simply iterates through all pairs of pointers in a
function and asks an alias analysis whether or not the pointers alias. This
gives an indication of the precision of the alias analysis. Statistics are
printed.
Chris Lattner
Last modified: Tue Mar 4 13:36:53 CST 2003