CalledValuePropagation.cpp File Reference

#include "llvm/Transforms/IPO/CalledValuePropagation.h"
#include "llvm/Analysis/SparsePropagation.h"
#include "llvm/Analysis/ValueLatticeUtils.h"
#include "llvm/IR/InstVisitor.h"
#include "llvm/IR/MDBuilder.h"
#include "llvm/Transforms/IPO.h"

Include dependency graph for CalledValuePropagation.cpp:

Go to the source code of this file.

Classes
struct	llvm::LatticeKeyInfo< CVPLatticeKey >
	A specialization of LatticeKeyInfo for CVPLatticeKeys. More...

Namespaces
	llvm
	This class represents lattice values for constants.

Macros
#define	DEBUG_TYPE   "called-value-propagation"

Enumerations
enum	IPOGrouping
	To enable interprocedural analysis, we assign LLVM values to the following groups. More...

Functions
static bool	runCVP (Module &M)
	INITIALIZE_PASS (CalledValuePropagationLegacyPass, "called-value-propagation", "Called Value Propagation", false, false) ModulePass *llvm

Variables
static cl::opt< unsigned >	MaxFunctionsPerValue ("cvp-max-functions-per-value", cl::Hidden, cl::init(4), cl::desc("The maximum number of functions to track per lattice value"))
	The maximum number of functions to track per lattice value. More...

Macro Definition Documentation

DEBUG_TYPE

#define DEBUG_TYPE   "called-value-propagation"

Definition at line 28 of file CalledValuePropagation.cpp.

Enumeration Type Documentation

IPOGrouping

enum IPOGrouping

strong

To enable interprocedural analysis, we assign LLVM values to the following groups.

The register group represents SSA registers, the return group represents the return values of functions, and the memory group represents in-memory values. An LLVM Value can technically be in more than one group. It's necessary to distinguish these groups so we can, for example, track a global variable separately from the value stored at its location.

Definition at line 48 of file CalledValuePropagation.cpp.

Function Documentation

INITIALIZE_PASS()

INITIALIZE_PASS	(	CalledValuePropagationLegacyPass	,
		"called-value-propagation"	,
		"Called Value Propagation"	,
		false	,
		false
	)

Definition at line 433 of file CalledValuePropagation.cpp.

runCVP()

static bool runCVP ( Module &  M )

static

Definition at line 371 of file CalledValuePropagation.cpp.

References C, llvm::canTrackArgumentsInterprocedurally(), F(), llvm::CallSiteBase< FunTy, BBTy, ValTy, UserTy, UseTy, InstrTy, CallTy, InvokeTy, IterTy >::getCalledValue(), llvm::SparseSolver< LatticeKey, LatticeVal, KeyInfo >::getExistingValueState(), llvm::SparseSolver< LatticeKey, LatticeVal, KeyInfo >::MarkBlockExecutable(), llvm::LLVMContext::MD_callees, Register, llvm::Instruction::setMetadata(), and llvm::SparseSolver< LatticeKey, LatticeVal, KeyInfo >::Solve().

Referenced by llvm::CalledValuePropagationPass::run().

Variable Documentation

MaxFunctionsPerValue

cl::opt<unsigned> MaxFunctionsPerValue("cvp-max-functions-per-value", cl::Hidden, cl::init(4), cl::desc("The maximum number of functions to track per lattice value"))

static

The maximum number of functions to track per lattice value.

Once the number of functions a call site can possibly target exceeds this threshold, it's lattice value becomes overdefined. The number of possible lattice values is bounded by Ch(F, M), where F is the number of functions in the module and M is MaxFunctionsPerValue. As such, this value should be kept very small. We likely can't do anything useful for call sites with a large number of possible targets, anyway.