Viewing file:  OMPIRBuilder.h (161.05 KB) -rw-r--r--
Select action/file-type:
 (+) |  (+) |  (+) | Code (+) | Session (+) |  (+) | SDB (+) |  (+) |  (+) |  (+) |  (+) |  (+) |
//===- IR/OpenMPIRBuilder.h - OpenMP encoding builder for LLVM IR - C++ -*-===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
// This file defines the OpenMPIRBuilder class and helpers used as a convenient
// way to create LLVM instructions for OpenMP directives.
//
//===----------------------------------------------------------------------===//
#ifndef LLVM_FRONTEND_OPENMP_OMPIRBUILDER_H
#define LLVM_FRONTEND_OPENMP_OMPIRBUILDER_H
#include "llvm/Analysis/MemorySSAUpdater.h"
#include "llvm/Frontend/OpenMP/OMPConstants.h"
#include "llvm/Frontend/OpenMP/OMPGridValues.h"
#include "llvm/IR/DebugLoc.h"
#include "llvm/IR/IRBuilder.h"
#include "llvm/IR/Module.h"
#include "llvm/Support/Allocator.h"
#include "llvm/TargetParser/Triple.h"
#include <forward_list>
#include <map>
#include <optional>
namespace llvm {
class CanonicalLoopInfo;
struct TargetRegionEntryInfo;
class OffloadEntriesInfoManager;
class OpenMPIRBuilder;
/// Move the instruction after an InsertPoint to the beginning of another
/// BasicBlock.
///
/// The instructions after \p IP are moved to the beginning of \p New which must
/// not have any PHINodes. If \p CreateBranch is true, a branch instruction to
/// \p New will be added such that there is no semantic change. Otherwise, the
/// \p IP insert block remains degenerate and it is up to the caller to insert a
/// terminator.
void spliceBB(IRBuilderBase::InsertPoint IP, BasicBlock *New,
bool CreateBranch);
/// Splice a BasicBlock at an IRBuilder's current insertion point. Its new
/// insert location will stick to after the instruction before the insertion
/// point (instead of moving with the instruction the InsertPoint stores
/// internally).
void spliceBB(IRBuilder<> &Builder, BasicBlock *New, bool CreateBranch);
/// Split a BasicBlock at an InsertPoint, even if the block is degenerate
/// (missing the terminator).
///
/// llvm::SplitBasicBlock and BasicBlock::splitBasicBlock require a well-formed
/// BasicBlock. \p Name is used for the new successor block. If \p CreateBranch
/// is true, a branch to the new successor will new created such that
/// semantically there is no change; otherwise the block of the insertion point
/// remains degenerate and it is the caller's responsibility to insert a
/// terminator. Returns the new successor block.
BasicBlock *splitBB(IRBuilderBase::InsertPoint IP, bool CreateBranch,
llvm::Twine Name = {});
/// Split a BasicBlock at \p Builder's insertion point, even if the block is
/// degenerate (missing the terminator). Its new insert location will stick to
/// after the instruction before the insertion point (instead of moving with the
/// instruction the InsertPoint stores internally).
BasicBlock *splitBB(IRBuilderBase &Builder, bool CreateBranch,
llvm::Twine Name = {});
/// Split a BasicBlock at \p Builder's insertion point, even if the block is
/// degenerate (missing the terminator). Its new insert location will stick to
/// after the instruction before the insertion point (instead of moving with the
/// instruction the InsertPoint stores internally).
BasicBlock *splitBB(IRBuilder<> &Builder, bool CreateBranch, llvm::Twine Name);
/// Like splitBB, but reuses the current block's name for the new name.
BasicBlock *splitBBWithSuffix(IRBuilderBase &Builder, bool CreateBranch,
llvm::Twine Suffix = ".split");
/// Captures attributes that affect generating LLVM-IR using the
/// OpenMPIRBuilder and related classes. Note that not all attributes are
/// required for all classes or functions. In some use cases the configuration
/// is not necessary at all, because because the only functions that are called
/// are ones that are not dependent on the configuration.
class OpenMPIRBuilderConfig {
public:
/// Flag to define whether to generate code for the role of the OpenMP host
/// (if set to false) or device (if set to true) in an offloading context. It
/// is set when the -fopenmp-is-target-device compiler frontend option is
/// specified.
std::optional<bool> IsTargetDevice;
/// Flag for specifying if the compilation is done for an accelerator. It is
/// set according to the architecture of the target triple and currently only
/// true when targeting AMDGPU or NVPTX. Today, these targets can only perform
/// the role of an OpenMP target device, so `IsTargetDevice` must also be true
/// if `IsGPU` is true. This restriction might be lifted if an accelerator-
/// like target with the ability to work as the OpenMP host is added, or if
/// the capabilities of the currently supported GPU architectures are
/// expanded.
std::optional<bool> IsGPU;
/// Flag for specifying if LLVMUsed information should be emitted.
std::optional<bool> EmitLLVMUsedMetaInfo;
/// Flag for specifying if offloading is mandatory.
std::optional<bool> OpenMPOffloadMandatory;
/// First separator used between the initial two parts of a name.
std::optional<StringRef> FirstSeparator;
/// Separator used between all of the rest consecutive parts of s name
std::optional<StringRef> Separator;
// Grid Value for the GPU target
std::optional<omp::GV> GridValue;
OpenMPIRBuilderConfig();
OpenMPIRBuilderConfig(bool IsTargetDevice, bool IsGPU,
bool OpenMPOffloadMandatory,
bool HasRequiresReverseOffload,
bool HasRequiresUnifiedAddress,
bool HasRequiresUnifiedSharedMemory,
bool HasRequiresDynamicAllocators);
// Getters functions that assert if the required values are not present.
bool isTargetDevice() const {
assert(IsTargetDevice.has_value() && "IsTargetDevice is not set");
return *IsTargetDevice;
}
bool isGPU() const {
assert(IsGPU.has_value() && "IsGPU is not set");
return *IsGPU;
}
bool openMPOffloadMandatory() const {
assert(OpenMPOffloadMandatory.has_value() &&
"OpenMPOffloadMandatory is not set");
return *OpenMPOffloadMandatory;
}
omp::GV getGridValue() const {
assert(GridValue.has_value() && "GridValue is not set");
return *GridValue;
}
bool hasRequiresFlags() const { return RequiresFlags; }
bool hasRequiresReverseOffload() const;
bool hasRequiresUnifiedAddress() const;
bool hasRequiresUnifiedSharedMemory() const;
bool hasRequiresDynamicAllocators() const;
/// Returns requires directive clauses as flags compatible with those expected
/// by libomptarget.
int64_t getRequiresFlags() const;
// Returns the FirstSeparator if set, otherwise use the default separator
// depending on isGPU
StringRef firstSeparator() const {
if (FirstSeparator.has_value())
return *FirstSeparator;
if (isGPU())
return "_";
return ".";
}
// Returns the Separator if set, otherwise use the default separator depending
// on isGPU
StringRef separator() const {
if (Separator.has_value())
return *Separator;
if (isGPU())
return "$";
return ".";
}
void setIsTargetDevice(bool Value) { IsTargetDevice = Value; }
void setIsGPU(bool Value) { IsGPU = Value; }
void setEmitLLVMUsed(bool Value = true) { EmitLLVMUsedMetaInfo = Value; }
void setOpenMPOffloadMandatory(bool Value) { OpenMPOffloadMandatory = Value; }
void setFirstSeparator(StringRef FS) { FirstSeparator = FS; }
void setSeparator(StringRef S) { Separator = S; }
void setGridValue(omp::GV G) { GridValue = G; }
void setHasRequiresReverseOffload(bool Value);
void setHasRequiresUnifiedAddress(bool Value);
void setHasRequiresUnifiedSharedMemory(bool Value);
void setHasRequiresDynamicAllocators(bool Value);
private:
/// Flags for specifying which requires directive clauses are present.
int64_t RequiresFlags;
};
/// Data structure to contain the information needed to uniquely identify
/// a target entry.
struct TargetRegionEntryInfo {
std::string ParentName;
unsigned DeviceID;
unsigned FileID;
unsigned Line;
unsigned Count;
TargetRegionEntryInfo() : DeviceID(0), FileID(0), Line(0), Count(0) {}
TargetRegionEntryInfo(StringRef ParentName, unsigned DeviceID,
unsigned FileID, unsigned Line, unsigned Count = 0)
: ParentName(ParentName), DeviceID(DeviceID), FileID(FileID), Line(Line),
Count(Count) {}
static void getTargetRegionEntryFnName(SmallVectorImpl<char> &Name,
StringRef ParentName,
unsigned DeviceID, unsigned FileID,
unsigned Line, unsigned Count);
bool operator<(const TargetRegionEntryInfo &RHS) const {
return std::make_tuple(ParentName, DeviceID, FileID, Line, Count) <
std::make_tuple(RHS.ParentName, RHS.DeviceID, RHS.FileID, RHS.Line,
RHS.Count);
}
};
/// Class that manages information about offload code regions and data
class OffloadEntriesInfoManager {
/// Number of entries registered so far.
OpenMPIRBuilder *OMPBuilder;
unsigned OffloadingEntriesNum = 0;
public:
/// Base class of the entries info.
class OffloadEntryInfo {
public:
/// Kind of a given entry.
enum OffloadingEntryInfoKinds : unsigned {
/// Entry is a target region.
OffloadingEntryInfoTargetRegion = 0,
/// Entry is a declare target variable.
OffloadingEntryInfoDeviceGlobalVar = 1,
/// Invalid entry info.
OffloadingEntryInfoInvalid = ~0u
};
protected:
OffloadEntryInfo() = delete;
explicit OffloadEntryInfo(OffloadingEntryInfoKinds Kind) : Kind(Kind) {}
explicit OffloadEntryInfo(OffloadingEntryInfoKinds Kind, unsigned Order,
uint32_t Flags)
: Flags(Flags), Order(Order), Kind(Kind) {}
~OffloadEntryInfo() = default;
public:
bool isValid() const { return Order != ~0u; }
unsigned getOrder() const { return Order; }
OffloadingEntryInfoKinds getKind() const { return Kind; }
uint32_t getFlags() const { return Flags; }
void setFlags(uint32_t NewFlags) { Flags = NewFlags; }
Constant *getAddress() const { return cast_or_null<Constant>(Addr); }
void setAddress(Constant *V) {
assert(!Addr.pointsToAliveValue() && "Address has been set before!");
Addr = V;
}
static bool classof(const OffloadEntryInfo *Info) { return true; }
private:
/// Address of the entity that has to be mapped for offloading.
WeakTrackingVH Addr;
/// Flags associated with the device global.
uint32_t Flags = 0u;
/// Order this entry was emitted.
unsigned Order = ~0u;
OffloadingEntryInfoKinds Kind = OffloadingEntryInfoInvalid;
};
/// Return true if a there are no entries defined.
bool empty() const;
/// Return number of entries defined so far.
unsigned size() const { return OffloadingEntriesNum; }
OffloadEntriesInfoManager(OpenMPIRBuilder *builder) : OMPBuilder(builder) {}
//
// Target region entries related.
//
/// Kind of the target registry entry.
enum OMPTargetRegionEntryKind : uint32_t {
/// Mark the entry as target region.
OMPTargetRegionEntryTargetRegion = 0x0,
};
/// Target region entries info.
class OffloadEntryInfoTargetRegion final : public OffloadEntryInfo {
/// Address that can be used as the ID of the entry.
Constant *ID = nullptr;
public:
OffloadEntryInfoTargetRegion()
: OffloadEntryInfo(OffloadingEntryInfoTargetRegion) {}
explicit OffloadEntryInfoTargetRegion(unsigned Order, Constant *Addr,
Constant *ID,
OMPTargetRegionEntryKind Flags)
: OffloadEntryInfo(OffloadingEntryInfoTargetRegion, Order, Flags),
ID(ID) {
setAddress(Addr);
}
Constant *getID() const { return ID; }
void setID(Constant *V) {
assert(!ID && "ID has been set before!");
ID = V;
}
static bool classof(const OffloadEntryInfo *Info) {
return Info->getKind() == OffloadingEntryInfoTargetRegion;
}
};
/// Initialize target region entry.
/// This is ONLY needed for DEVICE compilation.
void initializeTargetRegionEntryInfo(const TargetRegionEntryInfo &EntryInfo,
unsigned Order);
/// Register target region entry.
void registerTargetRegionEntryInfo(TargetRegionEntryInfo EntryInfo,
Constant *Addr, Constant *ID,
OMPTargetRegionEntryKind Flags);
/// Return true if a target region entry with the provided information
/// exists.
bool hasTargetRegionEntryInfo(TargetRegionEntryInfo EntryInfo,
bool IgnoreAddressId = false) const;
// Return the Name based on \a EntryInfo using the next available Count.
void getTargetRegionEntryFnName(SmallVectorImpl<char> &Name,
const TargetRegionEntryInfo &EntryInfo);
/// brief Applies action \a Action on all registered entries.
typedef function_ref<void(const TargetRegionEntryInfo &EntryInfo,
const OffloadEntryInfoTargetRegion &)>
OffloadTargetRegionEntryInfoActTy;
void
actOnTargetRegionEntriesInfo(const OffloadTargetRegionEntryInfoActTy &Action);
//
// Device global variable entries related.
//
/// Kind of the global variable entry..
enum OMPTargetGlobalVarEntryKind : uint32_t {
/// Mark the entry as a to declare target.
OMPTargetGlobalVarEntryTo = 0x0,
/// Mark the entry as a to declare target link.
OMPTargetGlobalVarEntryLink = 0x1,
/// Mark the entry as a declare target enter.
OMPTargetGlobalVarEntryEnter = 0x2,
/// Mark the entry as having no declare target entry kind.
OMPTargetGlobalVarEntryNone = 0x3,
/// Mark the entry as a declare target indirect global.
OMPTargetGlobalVarEntryIndirect = 0x8,
/// Mark the entry as a register requires global.
OMPTargetGlobalRegisterRequires = 0x10,
};
/// Kind of device clause for declare target variables
/// and functions
/// NOTE: Currently not used as a part of a variable entry
/// used for Flang and Clang to interface with the variable
/// related registration functions
enum OMPTargetDeviceClauseKind : uint32_t {
/// The target is marked for all devices
OMPTargetDeviceClauseAny = 0x0,
/// The target is marked for non-host devices
OMPTargetDeviceClauseNoHost = 0x1,
/// The target is marked for host devices
OMPTargetDeviceClauseHost = 0x2,
/// The target is marked as having no clause
OMPTargetDeviceClauseNone = 0x3
};
/// Device global variable entries info.
class OffloadEntryInfoDeviceGlobalVar final : public OffloadEntryInfo {
/// Type of the global variable.
int64_t VarSize;
GlobalValue::LinkageTypes Linkage;
const std::string VarName;
public:
OffloadEntryInfoDeviceGlobalVar()
: OffloadEntryInfo(OffloadingEntryInfoDeviceGlobalVar) {}
explicit OffloadEntryInfoDeviceGlobalVar(unsigned Order,
OMPTargetGlobalVarEntryKind Flags)
: OffloadEntryInfo(OffloadingEntryInfoDeviceGlobalVar, Order, Flags) {}
explicit OffloadEntryInfoDeviceGlobalVar(unsigned Order, Constant *Addr,
int64_t VarSize,
OMPTargetGlobalVarEntryKind Flags,
GlobalValue::LinkageTypes Linkage,
const std::string &VarName)
: OffloadEntryInfo(OffloadingEntryInfoDeviceGlobalVar, Order, Flags),
VarSize(VarSize), Linkage(Linkage), VarName(VarName) {
setAddress(Addr);
}
int64_t getVarSize() const { return VarSize; }
StringRef getVarName() const { return VarName; }
void setVarSize(int64_t Size) { VarSize = Size; }
GlobalValue::LinkageTypes getLinkage() const { return Linkage; }
void setLinkage(GlobalValue::LinkageTypes LT) { Linkage = LT; }
static bool classof(const OffloadEntryInfo *Info) {
return Info->getKind() == OffloadingEntryInfoDeviceGlobalVar;
}
};
/// Initialize device global variable entry.
/// This is ONLY used for DEVICE compilation.
void initializeDeviceGlobalVarEntryInfo(StringRef Name,
OMPTargetGlobalVarEntryKind Flags,
unsigned Order);
/// Register device global variable entry.
void registerDeviceGlobalVarEntryInfo(StringRef VarName, Constant *Addr,
int64_t VarSize,
OMPTargetGlobalVarEntryKind Flags,
GlobalValue::LinkageTypes Linkage);
/// Checks if the variable with the given name has been registered already.
bool hasDeviceGlobalVarEntryInfo(StringRef VarName) const {
return OffloadEntriesDeviceGlobalVar.count(VarName) > 0;
}
/// Applies action \a Action on all registered entries.
typedef function_ref<void(StringRef, const OffloadEntryInfoDeviceGlobalVar &)>
OffloadDeviceGlobalVarEntryInfoActTy;
void actOnDeviceGlobalVarEntriesInfo(
const OffloadDeviceGlobalVarEntryInfoActTy &Action);
private:
/// Return the count of entries at a particular source location.
unsigned
getTargetRegionEntryInfoCount(const TargetRegionEntryInfo &EntryInfo) const;
/// Update the count of entries at a particular source location.
void
incrementTargetRegionEntryInfoCount(const TargetRegionEntryInfo &EntryInfo);
static TargetRegionEntryInfo
getTargetRegionEntryCountKey(const TargetRegionEntryInfo &EntryInfo) {
return TargetRegionEntryInfo(EntryInfo.ParentName, EntryInfo.DeviceID,
EntryInfo.FileID, EntryInfo.Line, 0);
}
// Count of entries at a location.
std::map<TargetRegionEntryInfo, unsigned> OffloadEntriesTargetRegionCount;
// Storage for target region entries kind.
typedef std::map<TargetRegionEntryInfo, OffloadEntryInfoTargetRegion>
OffloadEntriesTargetRegionTy;
OffloadEntriesTargetRegionTy OffloadEntriesTargetRegion;
/// Storage for device global variable entries kind. The storage is to be
/// indexed by mangled name.
typedef StringMap<OffloadEntryInfoDeviceGlobalVar>
OffloadEntriesDeviceGlobalVarTy;
OffloadEntriesDeviceGlobalVarTy OffloadEntriesDeviceGlobalVar;
};
/// An interface to create LLVM-IR for OpenMP directives.
///
/// Each OpenMP directive has a corresponding public generator method.
class OpenMPIRBuilder {
public:
/// Create a new OpenMPIRBuilder operating on the given module \p M. This will
/// not have an effect on \p M (see initialize)
OpenMPIRBuilder(Module &M)
: M(M), Builder(M.getContext()), OffloadInfoManager(this),
T(Triple(M.getTargetTriple())) {}
~OpenMPIRBuilder();
/// Initialize the internal state, this will put structures types and
/// potentially other helpers into the underlying module. Must be called
/// before any other method and only once! This internal state includes types
/// used in the OpenMPIRBuilder generated from OMPKinds.def.
void initialize();
void setConfig(OpenMPIRBuilderConfig C) { Config = C; }
/// Finalize the underlying module, e.g., by outlining regions.
/// \param Fn The function to be finalized. If not used,
/// all functions are finalized.
void finalize(Function *Fn = nullptr);
/// Add attributes known for \p FnID to \p Fn.
void addAttributes(omp::RuntimeFunction FnID, Function &Fn);
/// Type used throughout for insertion points.
using InsertPointTy = IRBuilder<>::InsertPoint;
/// Get the create a name using the platform specific separators.
/// \param Parts parts of the final name that needs separation
/// The created name has a first separator between the first and second part
/// and a second separator between all other parts.
/// E.g. with FirstSeparator "$" and Separator "." and
/// parts: "p1", "p2", "p3", "p4"
/// The resulting name is "p1$p2.p3.p4"
/// The separators are retrieved from the OpenMPIRBuilderConfig.
std::string createPlatformSpecificName(ArrayRef<StringRef> Parts) const;
/// Callback type for variable finalization (think destructors).
///
/// \param CodeGenIP is the insertion point at which the finalization code
/// should be placed.
///
/// A finalize callback knows about all objects that need finalization, e.g.
/// destruction, when the scope of the currently generated construct is left
/// at the time, and location, the callback is invoked.
using FinalizeCallbackTy = std::function<void(InsertPointTy CodeGenIP)>;
struct FinalizationInfo {
/// The finalization callback provided by the last in-flight invocation of
/// createXXXX for the directive of kind DK.
FinalizeCallbackTy FiniCB;
/// The directive kind of the innermost directive that has an associated
/// region which might require finalization when it is left.
omp::Directive DK;
/// Flag to indicate if the directive is cancellable.
bool IsCancellable;
};
/// Push a finalization callback on the finalization stack.
///
/// NOTE: Temporary solution until Clang CG is gone.
void pushFinalizationCB(const FinalizationInfo &FI) {
FinalizationStack.push_back(FI);
}
/// Pop the last finalization callback from the finalization stack.
///
/// NOTE: Temporary solution until Clang CG is gone.
void popFinalizationCB() { FinalizationStack.pop_back(); }
/// Callback type for body (=inner region) code generation
///
/// The callback takes code locations as arguments, each describing a
/// location where additional instructions can be inserted.
///
/// The CodeGenIP may be in the middle of a basic block or point to the end of
/// it. The basic block may have a terminator or be degenerate. The callback
/// function may just insert instructions at that position, but also split the
/// block (without the Before argument of BasicBlock::splitBasicBlock such
/// that the identify of the split predecessor block is preserved) and insert
/// additional control flow, including branches that do not lead back to what
/// follows the CodeGenIP. Note that since the callback is allowed to split
/// the block, callers must assume that InsertPoints to positions in the
/// BasicBlock after CodeGenIP including CodeGenIP itself are invalidated. If
/// such InsertPoints need to be preserved, it can split the block itself
/// before calling the callback.
///
/// AllocaIP and CodeGenIP must not point to the same position.
///
/// \param AllocaIP is the insertion point at which new alloca instructions
/// should be placed. The BasicBlock it is pointing to must
/// not be split.
/// \param CodeGenIP is the insertion point at which the body code should be
/// placed.
using BodyGenCallbackTy =
function_ref<void(InsertPointTy AllocaIP, InsertPointTy CodeGenIP)>;
// This is created primarily for sections construct as llvm::function_ref
// (BodyGenCallbackTy) is not storable (as described in the comments of
// function_ref class - function_ref contains non-ownable reference
// to the callable.
using StorableBodyGenCallbackTy =
std::function<void(InsertPointTy AllocaIP, InsertPointTy CodeGenIP)>;
/// Callback type for loop body code generation.
///
/// \param CodeGenIP is the insertion point where the loop's body code must be
/// placed. This will be a dedicated BasicBlock with a
/// conditional branch from the loop condition check and
/// terminated with an unconditional branch to the loop
/// latch.
/// \param IndVar is the induction variable usable at the insertion point.
using LoopBodyGenCallbackTy =
function_ref<void(InsertPointTy CodeGenIP, Value *IndVar)>;
/// Callback type for variable privatization (think copy & default
/// constructor).
///
/// \param AllocaIP is the insertion point at which new alloca instructions
/// should be placed.
/// \param CodeGenIP is the insertion point at which the privatization code
/// should be placed.
/// \param Original The value being copied/created, should not be used in the
/// generated IR.
/// \param Inner The equivalent of \p Original that should be used in the
/// generated IR; this is equal to \p Original if the value is
/// a pointer and can thus be passed directly, otherwise it is
/// an equivalent but different value.
/// \param ReplVal The replacement value, thus a copy or new created version
/// of \p Inner.
///
/// \returns The new insertion point where code generation continues and
/// \p ReplVal the replacement value.
using PrivatizeCallbackTy = function_ref<InsertPointTy(
InsertPointTy AllocaIP, InsertPointTy CodeGenIP, Value &Original,
Value &Inner, Value *&ReplVal)>;
/// Description of a LLVM-IR insertion point (IP) and a debug/source location
/// (filename, line, column, ...).
struct LocationDescription {
LocationDescription(const IRBuilderBase &IRB)
: IP(IRB.saveIP()), DL(IRB.getCurrentDebugLocation()) {}
LocationDescription(const InsertPointTy &IP) : IP(IP) {}
LocationDescription(const InsertPointTy &IP, const DebugLoc &DL)
: IP(IP), DL(DL) {}
InsertPointTy IP;
DebugLoc DL;
};
/// Emitter methods for OpenMP directives.
///
///{
/// Generator for '#omp barrier'
///
/// \param Loc The location where the barrier directive was encountered.
/// \param Kind The kind of directive that caused the barrier.
/// \param ForceSimpleCall Flag to force a simple (=non-cancellation) barrier.
/// \param CheckCancelFlag Flag to indicate a cancel barrier return value
/// should be checked and acted upon.
/// \param ThreadID Optional parameter to pass in any existing ThreadID value.
///
/// \returns The insertion point after the barrier.
InsertPointTy createBarrier(const LocationDescription &Loc,
omp::Directive Kind, bool ForceSimpleCall = false,
bool CheckCancelFlag = true);
/// Generator for '#omp cancel'
///
/// \param Loc The location where the directive was encountered.
/// \param IfCondition The evaluated 'if' clause expression, if any.
/// \param CanceledDirective The kind of directive that is cancled.
///
/// \returns The insertion point after the barrier.
InsertPointTy createCancel(const LocationDescription &Loc, Value *IfCondition,
omp::Directive CanceledDirective);
/// Generator for '#omp parallel'
///
/// \param Loc The insert and source location description.
/// \param AllocaIP The insertion points to be used for alloca instructions.
/// \param BodyGenCB Callback that will generate the region code.
/// \param PrivCB Callback to copy a given variable (think copy constructor).
/// \param FiniCB Callback to finalize variable copies.
/// \param IfCondition The evaluated 'if' clause expression, if any.
/// \param NumThreads The evaluated 'num_threads' clause expression, if any.
/// \param ProcBind The value of the 'proc_bind' clause (see ProcBindKind).
/// \param IsCancellable Flag to indicate a cancellable parallel region.
///
/// \returns The insertion position *after* the parallel.
IRBuilder<>::InsertPoint
createParallel(const LocationDescription &Loc, InsertPointTy AllocaIP,
BodyGenCallbackTy BodyGenCB, PrivatizeCallbackTy PrivCB,
FinalizeCallbackTy FiniCB, Value *IfCondition,
Value *NumThreads, omp::ProcBindKind ProcBind,
bool IsCancellable);
/// Generator for the control flow structure of an OpenMP canonical loop.
///
/// This generator operates on the logical iteration space of the loop, i.e.
/// the caller only has to provide a loop trip count of the loop as defined by
/// base language semantics. The trip count is interpreted as an unsigned
/// integer. The induction variable passed to \p BodyGenCB will be of the same
/// type and run from 0 to \p TripCount - 1. It is up to the callback to
/// convert the logical iteration variable to the loop counter variable in the
/// loop body.
///
/// \param Loc The insert and source location description. The insert
/// location can be between two instructions or the end of a
/// degenerate block (e.g. a BB under construction).
/// \param BodyGenCB Callback that will generate the loop body code.
/// \param TripCount Number of iterations the loop body is executed.
/// \param Name Base name used to derive BB and instruction names.
///
/// \returns An object representing the created control flow structure which
/// can be used for loop-associated directives.
CanonicalLoopInfo *createCanonicalLoop(const LocationDescription &Loc,
LoopBodyGenCallbackTy BodyGenCB,
Value *TripCount,
const Twine &Name = "loop");
/// Generator for the control flow structure of an OpenMP canonical loop.
///
/// Instead of a logical iteration space, this allows specifying user-defined
/// loop counter values using increment, upper- and lower bounds. To
/// disambiguate the terminology when counting downwards, instead of lower
/// bounds we use \p Start for the loop counter value in the first body
/// iteration.
///
/// Consider the following limitations:
///
/// * A loop counter space over all integer values of its bit-width cannot be
/// represented. E.g using uint8_t, its loop trip count of 256 cannot be
/// stored into an 8 bit integer):
///
/// DO I = 0, 255, 1
///
/// * Unsigned wrapping is only supported when wrapping only "once"; E.g.
/// effectively counting downwards:
///
/// for (uint8_t i = 100u; i > 0; i += 127u)
///
///
/// TODO: May need to add additional parameters to represent:
///
/// * Allow representing downcounting with unsigned integers.
///
/// * Sign of the step and the comparison operator might disagree:
///
/// for (int i = 0; i < 42; i -= 1u)
///
//
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the loop body code.
/// \param Start Value of the loop counter for the first iterations.
/// \param Stop Loop counter values past this will stop the loop.
/// \param Step Loop counter increment after each iteration; negative
/// means counting down.
/// \param IsSigned Whether Start, Stop and Step are signed integers.
/// \param InclusiveStop Whether \p Stop itself is a valid value for the loop
/// counter.
/// \param ComputeIP Insertion point for instructions computing the trip
/// count. Can be used to ensure the trip count is available
/// at the outermost loop of a loop nest. If not set,
/// defaults to the preheader of the generated loop.
/// \param Name Base name used to derive BB and instruction names.
///
/// \returns An object representing the created control flow structure which
/// can be used for loop-associated directives.
CanonicalLoopInfo *createCanonicalLoop(const LocationDescription &Loc,
LoopBodyGenCallbackTy BodyGenCB,
Value *Start, Value *Stop, Value *Step,
bool IsSigned, bool InclusiveStop,
InsertPointTy ComputeIP = {},
const Twine &Name = "loop");
/// Collapse a loop nest into a single loop.
///
/// Merges loops of a loop nest into a single CanonicalLoopNest representation
/// that has the same number of innermost loop iterations as the origin loop
/// nest. The induction variables of the input loops are derived from the
/// collapsed loop's induction variable. This is intended to be used to
/// implement OpenMP's collapse clause. Before applying a directive,
/// collapseLoops normalizes a loop nest to contain only a single loop and the
/// directive's implementation does not need to handle multiple loops itself.
/// This does not remove the need to handle all loop nest handling by
/// directives, such as the ordered(<n>) clause or the simd schedule-clause
/// modifier of the worksharing-loop directive.
///
/// Example:
/// \code
/// for (int i = 0; i < 7; ++i) // Canonical loop "i"
/// for (int j = 0; j < 9; ++j) // Canonical loop "j"
/// body(i, j);
/// \endcode
///
/// After collapsing with Loops={i,j}, the loop is changed to
/// \code
/// for (int ij = 0; ij < 63; ++ij) {
/// int i = ij / 9;
/// int j = ij % 9;
/// body(i, j);
/// }
/// \endcode
///
/// In the current implementation, the following limitations apply:
///
/// * All input loops have an induction variable of the same type.
///
/// * The collapsed loop will have the same trip count integer type as the
/// input loops. Therefore it is possible that the collapsed loop cannot
/// represent all iterations of the input loops. For instance, assuming a
/// 32 bit integer type, and two input loops both iterating 2^16 times, the
/// theoretical trip count of the collapsed loop would be 2^32 iteration,
/// which cannot be represented in an 32-bit integer. Behavior is undefined
/// in this case.
///
/// * The trip counts of every input loop must be available at \p ComputeIP.
/// Non-rectangular loops are not yet supported.
///
/// * At each nest level, code between a surrounding loop and its nested loop
/// is hoisted into the loop body, and such code will be executed more
/// often than before collapsing (or not at all if any inner loop iteration
/// has a trip count of 0). This is permitted by the OpenMP specification.
///
/// \param DL Debug location for instructions added for collapsing,
/// such as instructions to compute/derive the input loop's
/// induction variables.
/// \param Loops Loops in the loop nest to collapse. Loops are specified
/// from outermost-to-innermost and every control flow of a
/// loop's body must pass through its directly nested loop.
/// \param ComputeIP Where additional instruction that compute the collapsed
/// trip count. If not set, defaults to before the generated
/// loop.
///
/// \returns The CanonicalLoopInfo object representing the collapsed loop.
CanonicalLoopInfo *collapseLoops(DebugLoc DL,
ArrayRef<CanonicalLoopInfo *> Loops,
InsertPointTy ComputeIP);
/// Get the default alignment value for given target
///
/// \param TargetTriple Target triple
/// \param Features StringMap which describes extra CPU features
static unsigned getOpenMPDefaultSimdAlign(const Triple &TargetTriple,
const StringMap<bool> &Features);
/// Retrieve (or create if non-existent) the address of a declare
/// target variable, used in conjunction with registerTargetGlobalVariable
/// to create declare target global variables.
///
/// \param CaptureClause - enumerator corresponding to the OpenMP capture
/// clause used in conjunction with the variable being registered (link,
/// to, enter).
/// \param DeviceClause - enumerator corresponding to the OpenMP capture
/// clause used in conjunction with the variable being registered (nohost,
/// host, any)
/// \param IsDeclaration - boolean stating if the variable being registered
/// is a declaration-only and not a definition
/// \param IsExternallyVisible - boolean stating if the variable is externally
/// visible
/// \param EntryInfo - Unique entry information for the value generated
/// using getTargetEntryUniqueInfo, used to name generated pointer references
/// to the declare target variable
/// \param MangledName - the mangled name of the variable being registered
/// \param GeneratedRefs - references generated by invocations of
/// registerTargetGlobalVariable invoked from getAddrOfDeclareTargetVar,
/// these are required by Clang for book keeping.
/// \param OpenMPSIMD - if OpenMP SIMD mode is currently enabled
/// \param TargetTriple - The OpenMP device target triple we are compiling
/// for
/// \param LlvmPtrTy - The type of the variable we are generating or
/// retrieving an address for
/// \param GlobalInitializer - a lambda function which creates a constant
/// used for initializing a pointer reference to the variable in certain
/// cases. If a nullptr is passed, it will default to utilising the original
/// variable to initialize the pointer reference.
/// \param VariableLinkage - a lambda function which returns the variables
/// linkage type, if unspecified and a nullptr is given, it will instead
/// utilise the linkage stored on the existing global variable in the
/// LLVMModule.
Constant *getAddrOfDeclareTargetVar(
OffloadEntriesInfoManager::OMPTargetGlobalVarEntryKind CaptureClause,
OffloadEntriesInfoManager::OMPTargetDeviceClauseKind DeviceClause,
bool IsDeclaration, bool IsExternallyVisible,
TargetRegionEntryInfo EntryInfo, StringRef MangledName,
std::vector<GlobalVariable *> &GeneratedRefs, bool OpenMPSIMD,
std::vector<Triple> TargetTriple, Type *LlvmPtrTy,
std::function<Constant *()> GlobalInitializer,
std::function<GlobalValue::LinkageTypes()> VariableLinkage);
/// Registers a target variable for device or host.
///
/// \param CaptureClause - enumerator corresponding to the OpenMP capture
/// clause used in conjunction with the variable being registered (link,
/// to, enter).
/// \param DeviceClause - enumerator corresponding to the OpenMP capture
/// clause used in conjunction with the variable being registered (nohost,
/// host, any)
/// \param IsDeclaration - boolean stating if the variable being registered
/// is a declaration-only and not a definition
/// \param IsExternallyVisible - boolean stating if the variable is externally
/// visible
/// \param EntryInfo - Unique entry information for the value generated
/// using getTargetEntryUniqueInfo, used to name generated pointer references
/// to the declare target variable
/// \param MangledName - the mangled name of the variable being registered
/// \param GeneratedRefs - references generated by invocations of
/// registerTargetGlobalVariable these are required by Clang for book
/// keeping.
/// \param OpenMPSIMD - if OpenMP SIMD mode is currently enabled
/// \param TargetTriple - The OpenMP device target triple we are compiling
/// for
/// \param GlobalInitializer - a lambda function which creates a constant
/// used for initializing a pointer reference to the variable in certain
/// cases. If a nullptr is passed, it will default to utilising the original
/// variable to initialize the pointer reference.
/// \param VariableLinkage - a lambda function which returns the variables
/// linkage type, if unspecified and a nullptr is given, it will instead
/// utilise the linkage stored on the existing global variable in the
/// LLVMModule.
/// \param LlvmPtrTy - The type of the variable we are generating or
/// retrieving an address for
/// \param Addr - the original llvm value (addr) of the variable to be
/// registered
void registerTargetGlobalVariable(
OffloadEntriesInfoManager::OMPTargetGlobalVarEntryKind CaptureClause,
OffloadEntriesInfoManager::OMPTargetDeviceClauseKind DeviceClause,
bool IsDeclaration, bool IsExternallyVisible,
TargetRegionEntryInfo EntryInfo, StringRef MangledName,
std::vector<GlobalVariable *> &GeneratedRefs, bool OpenMPSIMD,
std::vector<Triple> TargetTriple,
std::function<Constant *()> GlobalInitializer,
std::function<GlobalValue::LinkageTypes()> VariableLinkage,
Type *LlvmPtrTy, Constant *Addr);
/// Get the offset of the OMP_MAP_MEMBER_OF field.
unsigned getFlagMemberOffset();
/// Get OMP_MAP_MEMBER_OF flag with extra bits reserved based on
/// the position given.
/// \param Position - A value indicating the position of the parent
/// of the member in the kernel argument structure, often retrieved
/// by the parents position in the combined information vectors used
/// to generate the structure itself. Multiple children (member's of)
/// with the same parent will use the same returned member flag.
omp::OpenMPOffloadMappingFlags getMemberOfFlag(unsigned Position);
/// Given an initial flag set, this function modifies it to contain
/// the passed in MemberOfFlag generated from the getMemberOfFlag
/// function. The results are dependent on the existing flag bits
/// set in the original flag set.
/// \param Flags - The original set of flags to be modified with the
/// passed in MemberOfFlag.
/// \param MemberOfFlag - A modified OMP_MAP_MEMBER_OF flag, adjusted
/// slightly based on the getMemberOfFlag which adjusts the flag bits
/// based on the members position in its parent.
void setCorrectMemberOfFlag(omp::OpenMPOffloadMappingFlags &Flags,
omp::OpenMPOffloadMappingFlags MemberOfFlag);
private:
/// Modifies the canonical loop to be a statically-scheduled workshare loop
/// which is executed on the device
///
/// This takes a \p CLI representing a canonical loop, such as the one
/// created by \see createCanonicalLoop and emits additional instructions to
/// turn it into a workshare loop. In particular, it calls to an OpenMP
/// runtime function in the preheader to call OpenMP device rtl function
/// which handles worksharing of loop body interations.
///
/// \param DL Debug location for instructions added for the
/// workshare-loop construct itself.
/// \param CLI A descriptor of the canonical loop to workshare.
/// \param AllocaIP An insertion point for Alloca instructions usable in the
/// preheader of the loop.
/// \param LoopType Information about type of loop worksharing.
/// It corresponds to type of loop workshare OpenMP pragma.
///
/// \returns Point where to insert code after the workshare construct.
InsertPointTy applyWorkshareLoopTarget(DebugLoc DL, CanonicalLoopInfo *CLI,
InsertPointTy AllocaIP,
omp::WorksharingLoopType LoopType);
/// Modifies the canonical loop to be a statically-scheduled workshare loop.
///
/// This takes a \p LoopInfo representing a canonical loop, such as the one
/// created by \p createCanonicalLoop and emits additional instructions to
/// turn it into a workshare loop. In particular, it calls to an OpenMP
/// runtime function in the preheader to obtain the loop bounds to be used in
/// the current thread, updates the relevant instructions in the canonical
/// loop and calls to an OpenMP runtime finalization function after the loop.
///
/// \param DL Debug location for instructions added for the
/// workshare-loop construct itself.
/// \param CLI A descriptor of the canonical loop to workshare.
/// \param AllocaIP An insertion point for Alloca instructions usable in the
/// preheader of the loop.
/// \param NeedsBarrier Indicates whether a barrier must be inserted after
/// the loop.
///
/// \returns Point where to insert code after the workshare construct.
InsertPointTy applyStaticWorkshareLoop(DebugLoc DL, CanonicalLoopInfo *CLI,
InsertPointTy AllocaIP,
bool NeedsBarrier);
/// Modifies the canonical loop a statically-scheduled workshare loop with a
/// user-specified chunk size.
///
/// \param DL Debug location for instructions added for the
/// workshare-loop construct itself.
/// \param CLI A descriptor of the canonical loop to workshare.
/// \param AllocaIP An insertion point for Alloca instructions usable in
/// the preheader of the loop.
/// \param NeedsBarrier Indicates whether a barrier must be inserted after the
/// loop.
/// \param ChunkSize The user-specified chunk size.
///
/// \returns Point where to insert code after the workshare construct.
InsertPointTy applyStaticChunkedWorkshareLoop(DebugLoc DL,
CanonicalLoopInfo *CLI,
InsertPointTy AllocaIP,
bool NeedsBarrier,
Value *ChunkSize);
/// Modifies the canonical loop to be a dynamically-scheduled workshare loop.
///
/// This takes a \p LoopInfo representing a canonical loop, such as the one
/// created by \p createCanonicalLoop and emits additional instructions to
/// turn it into a workshare loop. In particular, it calls to an OpenMP
/// runtime function in the preheader to obtain, and then in each iteration
/// to update the loop counter.
///
/// \param DL Debug location for instructions added for the
/// workshare-loop construct itself.
/// \param CLI A descriptor of the canonical loop to workshare.
/// \param AllocaIP An insertion point for Alloca instructions usable in the
/// preheader of the loop.
/// \param SchedType Type of scheduling to be passed to the init function.
/// \param NeedsBarrier Indicates whether a barrier must be insterted after
/// the loop.
/// \param Chunk The size of loop chunk considered as a unit when
/// scheduling. If \p nullptr, defaults to 1.
///
/// \returns Point where to insert code after the workshare construct.
InsertPointTy applyDynamicWorkshareLoop(DebugLoc DL, CanonicalLoopInfo *CLI,
InsertPointTy AllocaIP,
omp::OMPScheduleType SchedType,
bool NeedsBarrier,
Value *Chunk = nullptr);
/// Create alternative version of the loop to support if clause
///
/// OpenMP if clause can require to generate second loop. This loop
/// will be executed when if clause condition is not met. createIfVersion
/// adds branch instruction to the copied loop if \p ifCond is not met.
///
/// \param Loop Original loop which should be versioned.
/// \param IfCond Value which corresponds to if clause condition
/// \param VMap Value to value map to define relation between
/// original and copied loop values and loop blocks.
/// \param NamePrefix Optional name prefix for if.then if.else blocks.
void createIfVersion(CanonicalLoopInfo *Loop, Value *IfCond,
ValueToValueMapTy &VMap, const Twine &NamePrefix = "");
public:
/// Modifies the canonical loop to be a workshare loop.
///
/// This takes a \p LoopInfo representing a canonical loop, such as the one
/// created by \p createCanonicalLoop and emits additional instructions to
/// turn it into a workshare loop. In particular, it calls to an OpenMP
/// runtime function in the preheader to obtain the loop bounds to be used in
/// the current thread, updates the relevant instructions in the canonical
/// loop and calls to an OpenMP runtime finalization function after the loop.
///
/// The concrete transformation is done by applyStaticWorkshareLoop,
/// applyStaticChunkedWorkshareLoop, or applyDynamicWorkshareLoop, depending
/// on the value of \p SchedKind and \p ChunkSize.
///
/// \param DL Debug location for instructions added for the
/// workshare-loop construct itself.
/// \param CLI A descriptor of the canonical loop to workshare.
/// \param AllocaIP An insertion point for Alloca instructions usable in the
/// preheader of the loop.
/// \param NeedsBarrier Indicates whether a barrier must be insterted after
/// the loop.
/// \param SchedKind Scheduling algorithm to use.
/// \param ChunkSize The chunk size for the inner loop.
/// \param HasSimdModifier Whether the simd modifier is present in the
/// schedule clause.
/// \param HasMonotonicModifier Whether the monotonic modifier is present in
/// the schedule clause.
/// \param HasNonmonotonicModifier Whether the nonmonotonic modifier is
/// present in the schedule clause.
/// \param HasOrderedClause Whether the (parameterless) ordered clause is
/// present.
/// \param LoopType Information about type of loop worksharing.
/// It corresponds to type of loop workshare OpenMP pragma.
///
/// \returns Point where to insert code after the workshare construct.
InsertPointTy applyWorkshareLoop(
DebugLoc DL, CanonicalLoopInfo *CLI, InsertPointTy AllocaIP,
bool NeedsBarrier,
llvm::omp::ScheduleKind SchedKind = llvm::omp::OMP_SCHEDULE_Default,
Value *ChunkSize = nullptr, bool HasSimdModifier = false,
bool HasMonotonicModifier = false, bool HasNonmonotonicModifier = false,
bool HasOrderedClause = false,
omp::WorksharingLoopType LoopType =
omp::WorksharingLoopType::ForStaticLoop);
/// Tile a loop nest.
///
/// Tiles the loops of \p Loops by the tile sizes in \p TileSizes. Loops in
/// \p/ Loops must be perfectly nested, from outermost to innermost loop
/// (i.e. Loops.front() is the outermost loop). The trip count llvm::Value
/// of every loop and every tile sizes must be usable in the outermost
/// loop's preheader. This implies that the loop nest is rectangular.
///
/// Example:
/// \code
/// for (int i = 0; i < 15; ++i) // Canonical loop "i"
/// for (int j = 0; j < 14; ++j) // Canonical loop "j"
/// body(i, j);
/// \endcode
///
/// After tiling with Loops={i,j} and TileSizes={5,7}, the loop is changed to
/// \code
/// for (int i1 = 0; i1 < 3; ++i1)
/// for (int j1 = 0; j1 < 2; ++j1)
/// for (int i2 = 0; i2 < 5; ++i2)
/// for (int j2 = 0; j2 < 7; ++j2)
/// body(i1*3+i2, j1*3+j2);
/// \endcode
///
/// The returned vector are the loops {i1,j1,i2,j2}. The loops i1 and j1 are
/// referred to the floor, and the loops i2 and j2 are the tiles. Tiling also
/// handles non-constant trip counts, non-constant tile sizes and trip counts
/// that are not multiples of the tile size. In the latter case the tile loop
/// of the last floor-loop iteration will have fewer iterations than specified
/// as its tile size.
///
///
/// @param DL Debug location for instructions added by tiling, for
/// instance the floor- and tile trip count computation.
/// @param Loops Loops to tile. The CanonicalLoopInfo objects are
/// invalidated by this method, i.e. should not used after
/// tiling.
/// @param TileSizes For each loop in \p Loops, the tile size for that
/// dimensions.
///
/// \returns A list of generated loops. Contains twice as many loops as the
/// input loop nest; the first half are the floor loops and the
/// second half are the tile loops.
std::vector<CanonicalLoopInfo *>
tileLoops(DebugLoc DL, ArrayRef<CanonicalLoopInfo *> Loops,
ArrayRef<Value *> TileSizes);
/// Fully unroll a loop.
///
/// Instead of unrolling the loop immediately (and duplicating its body
/// instructions), it is deferred to LLVM's LoopUnrollPass by adding loop
/// metadata.
///
/// \param DL Debug location for instructions added by unrolling.
/// \param Loop The loop to unroll. The loop will be invalidated.
void unrollLoopFull(DebugLoc DL, CanonicalLoopInfo *Loop);
/// Fully or partially unroll a loop. How the loop is unrolled is determined
/// using LLVM's LoopUnrollPass.
///
/// \param DL Debug location for instructions added by unrolling.
/// \param Loop The loop to unroll. The loop will be invalidated.
void unrollLoopHeuristic(DebugLoc DL, CanonicalLoopInfo *Loop);
/// Partially unroll a loop.
///
/// The CanonicalLoopInfo of the unrolled loop for use with chained
/// loop-associated directive can be requested using \p UnrolledCLI. Not
/// needing the CanonicalLoopInfo allows more efficient code generation by
/// deferring the actual unrolling to the LoopUnrollPass using loop metadata.
/// A loop-associated directive applied to the unrolled loop needs to know the
/// new trip count which means that if using a heuristically determined unroll
/// factor (\p Factor == 0), that factor must be computed immediately. We are
/// using the same logic as the LoopUnrollPass to derived the unroll factor,
/// but which assumes that some canonicalization has taken place (e.g.
/// Mem2Reg, LICM, GVN, Inlining, etc.). That is, the heuristic will perform
/// better when the unrolled loop's CanonicalLoopInfo is not needed.
///
/// \param DL Debug location for instructions added by unrolling.
/// \param Loop The loop to unroll. The loop will be invalidated.
/// \param Factor The factor to unroll the loop by. A factor of 0
/// indicates that a heuristic should be used to determine
/// the unroll-factor.
/// \param UnrolledCLI If non-null, receives the CanonicalLoopInfo of the
/// partially unrolled loop. Otherwise, uses loop metadata
/// to defer unrolling to the LoopUnrollPass.
void unrollLoopPartial(DebugLoc DL, CanonicalLoopInfo *Loop, int32_t Factor,
CanonicalLoopInfo **UnrolledCLI);
/// Add metadata to simd-ize a loop. If IfCond is not nullptr, the loop
/// is cloned. The metadata which prevents vectorization is added to
/// to the cloned loop. The cloned loop is executed when ifCond is evaluated
/// to false.
///
/// \param Loop The loop to simd-ize.
/// \param AlignedVars The map which containts pairs of the pointer
/// and its corresponding alignment.
/// \param IfCond The value which corresponds to the if clause
/// condition.
/// \param Order The enum to map order clause.
/// \param Simdlen The Simdlen length to apply to the simd loop.
/// \param Safelen The Safelen length to apply to the simd loop.
void applySimd(CanonicalLoopInfo *Loop,
MapVector<Value *, Value *> AlignedVars, Value *IfCond,
omp::OrderKind Order, ConstantInt *Simdlen,
ConstantInt *Safelen);
/// Generator for '#omp flush'
///
/// \param Loc The location where the flush directive was encountered
void createFlush(const LocationDescription &Loc);
/// Generator for '#omp taskwait'
///
/// \param Loc The location where the taskwait directive was encountered.
void createTaskwait(const LocationDescription &Loc);
/// Generator for '#omp taskyield'
///
/// \param Loc The location where the taskyield directive was encountered.
void createTaskyield(const LocationDescription &Loc);
/// A struct to pack the relevant information for an OpenMP depend clause.
struct DependData {
omp::RTLDependenceKindTy DepKind = omp::RTLDependenceKindTy::DepUnknown;
Type *DepValueType;
Value *DepVal;
explicit DependData() = default;
DependData(omp::RTLDependenceKindTy DepKind, Type *DepValueType,
Value *DepVal)
: DepKind(DepKind), DepValueType(DepValueType), DepVal(DepVal) {}
};
/// Generator for `#omp task`
///
/// \param Loc The location where the task construct was encountered.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param BodyGenCB Callback that will generate the region code.
/// \param Tied True if the task is tied, false if the task is untied.
/// \param Final i1 value which is `true` if the task is final, `false` if the
/// task is not final.
/// \param IfCondition i1 value. If it evaluates to `false`, an undeferred
/// task is generated, and the encountering thread must
/// suspend the current task region, for which execution
/// cannot be resumed until execution of the structured
/// block that is associated with the generated task is
/// completed.
InsertPointTy createTask(const LocationDescription &Loc,
InsertPointTy AllocaIP, BodyGenCallbackTy BodyGenCB,
bool Tied = true, Value *Final = nullptr,
Value *IfCondition = nullptr,
SmallVector<DependData> Dependencies = {});
/// Generator for the taskgroup construct
///
/// \param Loc The location where the taskgroup construct was encountered.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param BodyGenCB Callback that will generate the region code.
InsertPointTy createTaskgroup(const LocationDescription &Loc,
InsertPointTy AllocaIP,
BodyGenCallbackTy BodyGenCB);
using FileIdentifierInfoCallbackTy =
std::function<std::tuple<std::string, uint64_t>()>;
/// Creates a unique info for a target entry when provided a filename and
/// line number from.
///
/// \param CallBack A callback function which should return filename the entry
/// resides in as well as the line number for the target entry
/// \param ParentName The name of the parent the target entry resides in, if
/// any.
static TargetRegionEntryInfo
getTargetEntryUniqueInfo(FileIdentifierInfoCallbackTy CallBack,
StringRef ParentName = "");
/// Enum class for the RedctionGen CallBack type to be used.
enum class ReductionGenCBKind { Clang, MLIR };
/// ReductionGen CallBack for Clang
///
/// \param CodeGenIP InsertPoint for CodeGen.
/// \param Index Index of the ReductionInfo to generate code for.
/// \param LHSPtr Optionally used by Clang to return the LHSPtr it used for
/// codegen, used for fixup later.
/// \param RHSPtr Optionally used by Clang to
/// return the RHSPtr it used for codegen, used for fixup later.
/// \param CurFn Optionally used by Clang to pass in the Current Function as
/// Clang context may be old.
using ReductionGenClangCBTy =
std::function<InsertPointTy(InsertPointTy CodeGenIP, unsigned Index,
Value **LHS, Value **RHS, Function *CurFn)>;
/// ReductionGen CallBack for MLIR
///
/// \param CodeGenIP InsertPoint for CodeGen.
/// \param LHS Pass in the LHS Value to be used for CodeGen.
/// \param RHS Pass in the RHS Value to be used for CodeGen.
using ReductionGenCBTy = std::function<InsertPointTy(
InsertPointTy CodeGenIP, Value *LHS, Value *RHS, Value *&Res)>;
/// Functions used to generate atomic reductions. Such functions take two
/// Values representing pointers to LHS and RHS of the reduction, as well as
/// the element type of these pointers. They are expected to atomically
/// update the LHS to the reduced value.
using ReductionGenAtomicCBTy =
std::function<InsertPointTy(InsertPointTy, Type *, Value *, Value *)>;
/// Enum class for reduction evaluation types scalar, complex and aggregate.
enum class EvalKind { Scalar, Complex, Aggregate };
/// Information about an OpenMP reduction.
struct ReductionInfo {
ReductionInfo(Type *ElementType, Value *Variable, Value *PrivateVariable,
EvalKind EvaluationKind, ReductionGenCBTy ReductionGen,
ReductionGenClangCBTy ReductionGenClang,
ReductionGenAtomicCBTy AtomicReductionGen)
: ElementType(ElementType), Variable(Variable),
PrivateVariable(PrivateVariable), EvaluationKind(EvaluationKind),
ReductionGen(ReductionGen), ReductionGenClang(ReductionGenClang),
AtomicReductionGen(AtomicReductionGen) {}
ReductionInfo(Value *PrivateVariable)
: ElementType(nullptr), Variable(nullptr),
PrivateVariable(PrivateVariable), EvaluationKind(EvalKind::Scalar),
ReductionGen(), ReductionGenClang(), AtomicReductionGen() {}
/// Reduction element type, must match pointee type of variable.
Type *ElementType;
/// Reduction variable of pointer type.
Value *Variable;
/// Thread-private partial reduction variable.
Value *PrivateVariable;
/// Reduction evaluation kind - scalar, complex or aggregate.
EvalKind EvaluationKind;
/// Callback for generating the reduction body. The IR produced by this will
/// be used to combine two values in a thread-safe context, e.g., under
/// lock or within the same thread, and therefore need not be atomic.
ReductionGenCBTy ReductionGen;
/// Clang callback for generating the reduction body. The IR produced by
/// this will be used to combine two values in a thread-safe context, e.g.,
/// under lock or within the same thread, and therefore need not be atomic.
ReductionGenClangCBTy ReductionGenClang;
/// Callback for generating the atomic reduction body, may be null. The IR
/// produced by this will be used to atomically combine two values during
/// reduction. If null, the implementation will use the non-atomic version
/// along with the appropriate synchronization mechanisms.
ReductionGenAtomicCBTy AtomicReductionGen;
};
enum class CopyAction : unsigned {
// RemoteLaneToThread: Copy over a Reduce list from a remote lane in
// the warp using shuffle instructions.
RemoteLaneToThread,
// ThreadCopy: Make a copy of a Reduce list on the thread's stack.
ThreadCopy,
};
struct CopyOptionsTy {
Value *RemoteLaneOffset = nullptr;
Value *ScratchpadIndex = nullptr;
Value *ScratchpadWidth = nullptr;
};
/// Supporting functions for Reductions CodeGen.
private:
/// Emit the llvm.used metadata.
void emitUsed(StringRef Name, std::vector<llvm::WeakTrackingVH> &List);
/// Get the id of the current thread on the GPU.
Value *getGPUThreadID();
/// Get the GPU warp size.
Value *getGPUWarpSize();
/// Get the id of the warp in the block.
/// We assume that the warp size is 32, which is always the case
/// on the NVPTX device, to generate more efficient code.
Value *getNVPTXWarpID();
/// Get the id of the current lane in the Warp.
/// We assume that the warp size is 32, which is always the case
/// on the NVPTX device, to generate more efficient code.
Value *getNVPTXLaneID();
/// Cast value to the specified type.
Value *castValueToType(InsertPointTy AllocaIP, Value *From, Type *ToType);
/// This function creates calls to one of two shuffle functions to copy
/// variables between lanes in a warp.
Value *createRuntimeShuffleFunction(InsertPointTy AllocaIP, Value *Element,
Type *ElementType, Value *Offset);
/// Function to shuffle over the value from the remote lane.
void shuffleAndStore(InsertPointTy AllocaIP, Value *SrcAddr, Value *DstAddr,
Type *ElementType, Value *Offset,
Type *ReductionArrayTy);
/// Emit instructions to copy a Reduce list, which contains partially
/// aggregated values, in the specified direction.
void emitReductionListCopy(
InsertPointTy AllocaIP, CopyAction Action, Type *ReductionArrayTy,
ArrayRef<ReductionInfo> ReductionInfos, Value *SrcBase, Value *DestBase,
CopyOptionsTy CopyOptions = {nullptr, nullptr, nullptr});
/// Emit a helper that reduces data across two OpenMP threads (lanes)
/// in the same warp. It uses shuffle instructions to copy over data from
/// a remote lane's stack. The reduction algorithm performed is specified
/// by the fourth parameter.
///
/// Algorithm Versions.
/// Full Warp Reduce (argument value 0):
/// This algorithm assumes that all 32 lanes are active and gathers
/// data from these 32 lanes, producing a single resultant value.
/// Contiguous Partial Warp Reduce (argument value 1):
/// This algorithm assumes that only a *contiguous* subset of lanes
/// are active. This happens for the last warp in a parallel region
/// when the user specified num_threads is not an integer multiple of
/// 32. This contiguous subset always starts with the zeroth lane.
/// Partial Warp Reduce (argument value 2):
/// This algorithm gathers data from any number of lanes at any position.
/// All reduced values are stored in the lowest possible lane. The set
/// of problems every algorithm addresses is a super set of those
/// addressable by algorithms with a lower version number. Overhead
/// increases as algorithm version increases.
///
/// Terminology
/// Reduce element:
/// Reduce element refers to the individual data field with primitive
/// data types to be combined and reduced across threads.
/// Reduce list:
/// Reduce list refers to a collection of local, thread-private
/// reduce elements.
/// Remote Reduce list:
/// Remote Reduce list refers to a collection of remote (relative to
/// the current thread) reduce elements.
///
/// We distinguish between three states of threads that are important to
/// the implementation of this function.
/// Alive threads:
/// Threads in a warp executing the SIMT instruction, as distinguished from
/// threads that are inactive due to divergent control flow.
/// Active threads:
/// The minimal set of threads that has to be alive upon entry to this
/// function. The computation is correct iff active threads are alive.
/// Some threads are alive but they are not active because they do not
/// contribute to the computation in any useful manner. Turning them off
/// may introduce control flow overheads without any tangible benefits.
/// Effective threads:
/// In order to comply with the argument requirements of the shuffle
/// function, we must keep all lanes holding data alive. But at most
/// half of them perform value aggregation; we refer to this half of
/// threads as effective. The other half is simply handing off their
/// data.
///
/// Procedure
/// Value shuffle:
/// In this step active threads transfer data from higher lane positions
/// in the warp to lower lane positions, creating Remote Reduce list.
/// Value aggregation:
/// In this step, effective threads combine their thread local Reduce list
/// with Remote Reduce list and store the result in the thread local
/// Reduce list.
/// Value copy:
/// In this step, we deal with the assumption made by algorithm 2
/// (i.e. contiguity assumption). When we have an odd number of lanes
/// active, say 2k+1, only k threads will be effective and therefore k
/// new values will be produced. However, the Reduce list owned by the
/// (2k+1)th thread is ignored in the value aggregation. Therefore
/// we copy the Reduce list from the (2k+1)th lane to (k+1)th lane so
/// that the contiguity assumption still holds.
///
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReduceFn The reduction function.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The ShuffleAndReduce function.
Function *emitShuffleAndReduceFunction(
ArrayRef<OpenMPIRBuilder::ReductionInfo> ReductionInfos,
Function *ReduceFn, AttributeList FuncAttrs);
/// This function emits a helper that gathers Reduce lists from the first
/// lane of every active warp to lanes in the first warp.
///
/// void inter_warp_copy_func(void* reduce_data, num_warps)
/// shared smem[warp_size];
/// For all data entries D in reduce_data:
/// sync
/// If (I am the first lane in each warp)
/// Copy my local D to smem[warp_id]
/// sync
/// if (I am the first warp)
/// Copy smem[thread_id] to my local D
///
/// \param Loc The insert and source location description.
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The InterWarpCopy function.
Function *emitInterWarpCopyFunction(const LocationDescription &Loc,
ArrayRef<ReductionInfo> ReductionInfos,
AttributeList FuncAttrs);
/// This function emits a helper that copies all the reduction variables from
/// the team into the provided global buffer for the reduction variables.
///
/// void list_to_global_copy_func(void *buffer, int Idx, void *reduce_data)
/// For all data entries D in reduce_data:
/// Copy local D to buffer.D[Idx]
///
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReductionsBufferTy The StructTy for the reductions buffer.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The ListToGlobalCopy function.
Function *emitListToGlobalCopyFunction(ArrayRef<ReductionInfo> ReductionInfos,
Type *ReductionsBufferTy,
AttributeList FuncAttrs);
/// This function emits a helper that copies all the reduction variables from
/// the team into the provided global buffer for the reduction variables.
///
/// void list_to_global_copy_func(void *buffer, int Idx, void *reduce_data)
/// For all data entries D in reduce_data:
/// Copy buffer.D[Idx] to local D;
///
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReductionsBufferTy The StructTy for the reductions buffer.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The GlobalToList function.
Function *emitGlobalToListCopyFunction(ArrayRef<ReductionInfo> ReductionInfos,
Type *ReductionsBufferTy,
AttributeList FuncAttrs);
/// This function emits a helper that reduces all the reduction variables from
/// the team into the provided global buffer for the reduction variables.
///
/// void list_to_global_reduce_func(void *buffer, int Idx, void *reduce_data)
/// void *GlobPtrs[];
/// GlobPtrs[0] = (void*)&buffer.D0[Idx];
/// ...
/// GlobPtrs[N] = (void*)&buffer.DN[Idx];
/// reduce_function(GlobPtrs, reduce_data);
///
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReduceFn The reduction function.
/// \param ReductionsBufferTy The StructTy for the reductions buffer.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The ListToGlobalReduce function.
Function *
emitListToGlobalReduceFunction(ArrayRef<ReductionInfo> ReductionInfos,
Function *ReduceFn, Type *ReductionsBufferTy,
AttributeList FuncAttrs);
/// This function emits a helper that reduces all the reduction variables from
/// the team into the provided global buffer for the reduction variables.
///
/// void global_to_list_reduce_func(void *buffer, int Idx, void *reduce_data)
/// void *GlobPtrs[];
/// GlobPtrs[0] = (void*)&buffer.D0[Idx];
/// ...
/// GlobPtrs[N] = (void*)&buffer.DN[Idx];
/// reduce_function(reduce_data, GlobPtrs);
///
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReduceFn The reduction function.
/// \param ReductionsBufferTy The StructTy for the reductions buffer.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The GlobalToListReduce function.
Function *
emitGlobalToListReduceFunction(ArrayRef<ReductionInfo> ReductionInfos,
Function *ReduceFn, Type *ReductionsBufferTy,
AttributeList FuncAttrs);
/// Get the function name of a reduction function.
std::string getReductionFuncName(StringRef Name) const;
/// Emits reduction function.
/// \param ReducerName Name of the function calling the reduction.
/// \param ReductionInfos Array type containing the ReductionOps.
/// \param ReductionGenCBKind Optional param to specify Clang or MLIR
/// CodeGenCB kind.
/// \param FuncAttrs Optional param to specify any function attributes that
/// need to be copied to the new function.
///
/// \return The reduction function.
Function *createReductionFunction(
StringRef ReducerName, ArrayRef<ReductionInfo> ReductionInfos,
ReductionGenCBKind ReductionGenCBKind = ReductionGenCBKind::MLIR,
AttributeList FuncAttrs = {});
public:
///
/// Design of OpenMP reductions on the GPU
///
/// Consider a typical OpenMP program with one or more reduction
/// clauses:
///
/// float foo;
/// double bar;
/// #pragma omp target teams distribute parallel for \
/// reduction(+:foo) reduction(*:bar)
/// for (int i = 0; i < N; i++) {
/// foo += A[i]; bar *= B[i];
/// }
///
/// where 'foo' and 'bar' are reduced across all OpenMP threads in
/// all teams. In our OpenMP implementation on the NVPTX device an
/// OpenMP team is mapped to a CUDA threadblock and OpenMP threads
/// within a team are mapped to CUDA threads within a threadblock.
/// Our goal is to efficiently aggregate values across all OpenMP
/// threads such that:
///
/// - the compiler and runtime are logically concise, and
/// - the reduction is performed efficiently in a hierarchical
/// manner as follows: within OpenMP threads in the same warp,
/// across warps in a threadblock, and finally across teams on
/// the NVPTX device.
///
/// Introduction to Decoupling
///
/// We would like to decouple the compiler and the runtime so that the
/// latter is ignorant of the reduction variables (number, data types)
/// and the reduction operators. This allows a simpler interface
/// and implementation while still attaining good performance.
///
/// Pseudocode for the aforementioned OpenMP program generated by the
/// compiler is as follows:
///
/// 1. Create private copies of reduction variables on each OpenMP
/// thread: 'foo_private', 'bar_private'
/// 2. Each OpenMP thread reduces the chunk of 'A' and 'B' assigned
/// to it and writes the result in 'foo_private' and 'bar_private'
/// respectively.
/// 3. Call the OpenMP runtime on the GPU to reduce within a team
/// and store the result on the team master:
///
/// __kmpc_nvptx_parallel_reduce_nowait_v2(...,
/// reduceData, shuffleReduceFn, interWarpCpyFn)
///
/// where:
/// struct ReduceData {
/// double *foo;
/// double *bar;
/// } reduceData
/// reduceData.foo = &foo_private
/// reduceData.bar = &bar_private
///
/// 'shuffleReduceFn' and 'interWarpCpyFn' are pointers to two
/// auxiliary functions generated by the compiler that operate on
/// variables of type 'ReduceData'. They aid the runtime perform
/// algorithmic steps in a data agnostic manner.
///
/// 'shuffleReduceFn' is a pointer to a function that reduces data
/// of type 'ReduceData' across two OpenMP threads (lanes) in the
/// same warp. It takes the following arguments as input:
///
/// a. variable of type 'ReduceData' on the calling lane,
/// b. its lane_id,
/// c. an offset relative to the current lane_id to generate a
/// remote_lane_id. The remote lane contains the second
/// variable of type 'ReduceData' that is to be reduced.
/// d. an algorithm version parameter determining which reduction
/// algorithm to use.
///
/// 'shuffleReduceFn' retrieves data from the remote lane using
/// efficient GPU shuffle intrinsics and reduces, using the
/// algorithm specified by the 4th parameter, the two operands
/// element-wise. The result is written to the first operand.
///
/// Different reduction algorithms are implemented in different
/// runtime functions, all calling 'shuffleReduceFn' to perform
/// the essential reduction step. Therefore, based on the 4th
/// parameter, this function behaves slightly differently to
/// cooperate with the runtime to ensure correctness under
/// different circumstances.
///
/// 'InterWarpCpyFn' is a pointer to a function that transfers
/// reduced variables across warps. It tunnels, through CUDA
/// shared memory, the thread-private data of type 'ReduceData'
/// from lane 0 of each warp to a lane in the first warp.
/// 4. Call the OpenMP runtime on the GPU to reduce across teams.
/// The last team writes the global reduced value to memory.
///
/// ret = __kmpc_nvptx_teams_reduce_nowait(...,
/// reduceData, shuffleReduceFn, interWarpCpyFn,
/// scratchpadCopyFn, loadAndReduceFn)
///
/// 'scratchpadCopyFn' is a helper that stores reduced
/// data from the team master to a scratchpad array in
/// global memory.
///
/// 'loadAndReduceFn' is a helper that loads data from
/// the scratchpad array and reduces it with the input
/// operand.
///
/// These compiler generated functions hide address
/// calculation and alignment information from the runtime.
/// 5. if ret == 1:
/// The team master of the last team stores the reduced
/// result to the globals in memory.
/// foo += reduceData.foo; bar *= reduceData.bar
///
///
/// Warp Reduction Algorithms
///
/// On the warp level, we have three algorithms implemented in the
/// OpenMP runtime depending on the number of active lanes:
///
/// Full Warp Reduction
///
/// The reduce algorithm within a warp where all lanes are active
/// is implemented in the runtime as follows:
///
/// full_warp_reduce(void *reduce_data,
/// kmp_ShuffleReductFctPtr ShuffleReduceFn) {
/// for (int offset = WARPSIZE/2; offset > 0; offset /= 2)
/// ShuffleReduceFn(reduce_data, 0, offset, 0);
/// }
///
/// The algorithm completes in log(2, WARPSIZE) steps.
///
/// 'ShuffleReduceFn' is used here with lane_id set to 0 because it is
/// not used therefore we save instructions by not retrieving lane_id
/// from the corresponding special registers. The 4th parameter, which
/// represents the version of the algorithm being used, is set to 0 to
/// signify full warp reduction.
///
/// In this version, 'ShuffleReduceFn' behaves, per element, as follows:
///
/// #reduce_elem refers to an element in the local lane's data structure
/// #remote_elem is retrieved from a remote lane
/// remote_elem = shuffle_down(reduce_elem, offset, WARPSIZE);
/// reduce_elem = reduce_elem REDUCE_OP remote_elem;
///
/// Contiguous Partial Warp Reduction
///
/// This reduce algorithm is used within a warp where only the first
/// 'n' (n <= WARPSIZE) lanes are active. It is typically used when the
/// number of OpenMP threads in a parallel region is not a multiple of
/// WARPSIZE. The algorithm is implemented in the runtime as follows:
///
/// void
/// contiguous_partial_reduce(void *reduce_data,
/// kmp_ShuffleReductFctPtr ShuffleReduceFn,
/// int size, int lane_id) {
/// int curr_size;
/// int offset;
/// curr_size = size;
/// mask = curr_size/2;
/// while (offset>0) {
/// ShuffleReduceFn(reduce_data, lane_id, offset, 1);
/// curr_size = (curr_size+1)/2;
/// offset = curr_size/2;
/// }
/// }
///
/// In this version, 'ShuffleReduceFn' behaves, per element, as follows:
///
/// remote_elem = shuffle_down(reduce_elem, offset, WARPSIZE);
/// if (lane_id < offset)
/// reduce_elem = reduce_elem REDUCE_OP remote_elem
/// else
/// reduce_elem = remote_elem
///
/// This algorithm assumes that the data to be reduced are located in a
/// contiguous subset of lanes starting from the first. When there is
/// an odd number of active lanes, the data in the last lane is not
/// aggregated with any other lane's dat but is instead copied over.
///
/// Dispersed Partial Warp Reduction
///
/// This algorithm is used within a warp when any discontiguous subset of
/// lanes are active. It is used to implement the reduction operation
/// across lanes in an OpenMP simd region or in a nested parallel region.
///
/// void
/// dispersed_partial_reduce(void *reduce_data,
/// kmp_ShuffleReductFctPtr ShuffleReduceFn) {
/// int size, remote_id;
/// int logical_lane_id = number_of_active_lanes_before_me() * 2;
/// do {
/// remote_id = next_active_lane_id_right_after_me();
/// # the above function returns 0 of no active lane
/// # is present right after the current lane.
/// size = number_of_active_lanes_in_this_warp();
/// logical_lane_id /= 2;
/// ShuffleReduceFn(reduce_data, logical_lane_id,
/// remote_id-1-threadIdx.x, 2);
/// } while (logical_lane_id % 2 == 0 && size > 1);
/// }
///
/// There is no assumption made about the initial state of the reduction.
/// Any number of lanes (>=1) could be active at any position. The reduction
/// result is returned in the first active lane.
///
/// In this version, 'ShuffleReduceFn' behaves, per element, as follows:
///
/// remote_elem = shuffle_down(reduce_elem, offset, WARPSIZE);
/// if (lane_id % 2 == 0 && offset > 0)
/// reduce_elem = reduce_elem REDUCE_OP remote_elem
/// else
/// reduce_elem = remote_elem
///
///
/// Intra-Team Reduction
///
/// This function, as implemented in the runtime call
/// '__kmpc_nvptx_parallel_reduce_nowait_v2', aggregates data across OpenMP
/// threads in a team. It first reduces within a warp using the
/// aforementioned algorithms. We then proceed to gather all such
/// reduced values at the first warp.
///
/// The runtime makes use of the function 'InterWarpCpyFn', which copies
/// data from each of the "warp master" (zeroth lane of each warp, where
/// warp-reduced data is held) to the zeroth warp. This step reduces (in
/// a mathematical sense) the problem of reduction across warp masters in
/// a block to the problem of warp reduction.
///
///
/// Inter-Team Reduction
///
/// Once a team has reduced its data to a single value, it is stored in
/// a global scratchpad array. Since each team has a distinct slot, this
/// can be done without locking.
///
/// The last team to write to the scratchpad array proceeds to reduce the
/// scratchpad array. One or more workers in the last team use the helper
/// 'loadAndReduceDataFn' to load and reduce values from the array, i.e.,
/// the k'th worker reduces every k'th element.
///
/// Finally, a call is made to '__kmpc_nvptx_parallel_reduce_nowait_v2' to
/// reduce across workers and compute a globally reduced value.
///
/// \param Loc The location where the reduction was
/// encountered. Must be within the associate
/// directive and after the last local access to the
/// reduction variables.
/// \param AllocaIP An insertion point suitable for allocas usable
/// in reductions.
/// \param CodeGenIP An insertion point suitable for code
/// generation. \param ReductionInfos A list of info on each reduction
/// variable. \param IsNoWait Optional flag set if the reduction is
/// marked as
/// nowait.
/// \param IsTeamsReduction Optional flag set if it is a teams
/// reduction.
/// \param HasDistribute Optional flag set if it is a
/// distribute reduction.
/// \param GridValue Optional GPU grid value.
/// \param ReductionBufNum Optional OpenMPCUDAReductionBufNumValue to be
/// used for teams reduction.
/// \param SrcLocInfo Source location information global.
InsertPointTy createReductionsGPU(
const LocationDescription &Loc, InsertPointTy AllocaIP,
InsertPointTy CodeGenIP, ArrayRef<ReductionInfo> ReductionInfos,
bool IsNoWait = false, bool IsTeamsReduction = false,
bool HasDistribute = false,
ReductionGenCBKind ReductionGenCBKind = ReductionGenCBKind::MLIR,
std::optional<omp::GV> GridValue = {}, unsigned ReductionBufNum = 1024,
Value *SrcLocInfo = nullptr);
// TODO: provide atomic and non-atomic reduction generators for reduction
// operators defined by the OpenMP specification.
/// Generator for '#omp reduction'.
///
/// Emits the IR instructing the runtime to perform the specific kind of
/// reductions. Expects reduction variables to have been privatized and
/// initialized to reduction-neutral values separately. Emits the calls to
/// runtime functions as well as the reduction function and the basic blocks
/// performing the reduction atomically and non-atomically.
///
/// The code emitted for the following:
///
/// \code
/// type var_1;
/// type var_2;
/// #pragma omp <directive> reduction(reduction-op:var_1,var_2)
/// /* body */;
/// \endcode
///
/// corresponds to the following sketch.
///
/// \code
/// void _outlined_par() {
/// // N is the number of different reductions.
/// void *red_array[] = {privatized_var_1, privatized_var_2, ...};
/// switch(__kmpc_reduce(..., N, /*size of data in red array*/, red_array,
/// _omp_reduction_func,
/// _gomp_critical_user.reduction.var)) {
/// case 1: {
/// var_1 = var_1 <reduction-op> privatized_var_1;
/// var_2 = var_2 <reduction-op> privatized_var_2;
/// // ...
/// __kmpc_end_reduce(...);
/// break;
/// }
/// case 2: {
/// _Atomic<ReductionOp>(var_1, privatized_var_1);
/// _Atomic<ReductionOp>(var_2, privatized_var_2);
/// // ...
/// break;
/// }
/// default: break;
/// }
/// }
///
/// void _omp_reduction_func(void **lhs, void **rhs) {
/// *(type *)lhs[0] = *(type *)lhs[0] <reduction-op> *(type *)rhs[0];
/// *(type *)lhs[1] = *(type *)lhs[1] <reduction-op> *(type *)rhs[1];
/// // ...
/// }
/// \endcode
///
/// \param Loc The location where the reduction was
/// encountered. Must be within the associate
/// directive and after the last local access to the
/// reduction variables.
/// \param AllocaIP An insertion point suitable for allocas usable
/// in reductions.
/// \param ReductionInfos A list of info on each reduction variable.
/// \param IsNoWait A flag set if the reduction is marked as nowait.
/// \param IsByRef A flag set if the reduction is using reference
/// or direct value.
InsertPointTy createReductions(const LocationDescription &Loc,
InsertPointTy AllocaIP,
ArrayRef<ReductionInfo> ReductionInfos,
ArrayRef<bool> IsByRef, bool IsNoWait = false);
///}
/// Return the insertion point used by the underlying IRBuilder.
InsertPointTy getInsertionPoint() { return Builder.saveIP(); }
/// Update the internal location to \p Loc.
bool updateToLocation(const LocationDescription &Loc) {
Builder.restoreIP(Loc.IP);
Builder.SetCurrentDebugLocation(Loc.DL);
return Loc.IP.getBlock() != nullptr;
}
/// Return the function declaration for the runtime function with \p FnID.
FunctionCallee getOrCreateRuntimeFunction(Module &M,
omp::RuntimeFunction FnID);
Function *getOrCreateRuntimeFunctionPtr(omp::RuntimeFunction FnID);
/// Return the (LLVM-IR) string describing the source location \p LocStr.
Constant *getOrCreateSrcLocStr(StringRef LocStr, uint32_t &SrcLocStrSize);
/// Return the (LLVM-IR) string describing the default source location.
Constant *getOrCreateDefaultSrcLocStr(uint32_t &SrcLocStrSize);
/// Return the (LLVM-IR) string describing the source location identified by
/// the arguments.
Constant *getOrCreateSrcLocStr(StringRef FunctionName, StringRef FileName,
unsigned Line, unsigned Column,
uint32_t &SrcLocStrSize);
/// Return the (LLVM-IR) string describing the DebugLoc \p DL. Use \p F as
/// fallback if \p DL does not specify the function name.
Constant *getOrCreateSrcLocStr(DebugLoc DL, uint32_t &SrcLocStrSize,
Function *F = nullptr);
/// Return the (LLVM-IR) string describing the source location \p Loc.
Constant *getOrCreateSrcLocStr(const LocationDescription &Loc,
uint32_t &SrcLocStrSize);
/// Return an ident_t* encoding the source location \p SrcLocStr and \p Flags.
/// TODO: Create a enum class for the Reserve2Flags
Constant *getOrCreateIdent(Constant *SrcLocStr, uint32_t SrcLocStrSize,
omp::IdentFlag Flags = omp::IdentFlag(0),
unsigned Reserve2Flags = 0);
/// Create a hidden global flag \p Name in the module with initial value \p
/// Value.
GlobalValue *createGlobalFlag(unsigned Value, StringRef Name);
/// Generate control flow and cleanup for cancellation.
///
/// \param CancelFlag Flag indicating if the cancellation is performed.
/// \param CanceledDirective The kind of directive that is cancled.
/// \param ExitCB Extra code to be generated in the exit block.
void emitCancelationCheckImpl(Value *CancelFlag,
omp::Directive CanceledDirective,
FinalizeCallbackTy ExitCB = {});
/// Generate a target region entry call.
///
/// \param Loc The location at which the request originated and is fulfilled.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param Return Return value of the created function returned by reference.
/// \param DeviceID Identifier for the device via the 'device' clause.
/// \param NumTeams Numer of teams for the region via the 'num_teams' clause
/// or 0 if unspecified and -1 if there is no 'teams' clause.
/// \param NumThreads Number of threads via the 'thread_limit' clause.
/// \param HostPtr Pointer to the host-side pointer of the target kernel.
/// \param KernelArgs Array of arguments to the kernel.
InsertPointTy emitTargetKernel(const LocationDescription &Loc,
InsertPointTy AllocaIP, Value *&Return,
Value *Ident, Value *DeviceID, Value *NumTeams,
Value *NumThreads, Value *HostPtr,
ArrayRef<Value *> KernelArgs);
/// Generate a flush runtime call.
///
/// \param Loc The location at which the request originated and is fulfilled.
void emitFlush(const LocationDescription &Loc);
/// The finalization stack made up of finalize callbacks currently in-flight,
/// wrapped into FinalizationInfo objects that reference also the finalization
/// target block and the kind of cancellable directive.
SmallVector<FinalizationInfo, 8> FinalizationStack;
/// Return true if the last entry in the finalization stack is of kind \p DK
/// and cancellable.
bool isLastFinalizationInfoCancellable(omp::Directive DK) {
return !FinalizationStack.empty() &&
FinalizationStack.back().IsCancellable &&
FinalizationStack.back().DK == DK;
}
/// Generate a taskwait runtime call.
///
/// \param Loc The location at which the request originated and is fulfilled.
void emitTaskwaitImpl(const LocationDescription &Loc);
/// Generate a taskyield runtime call.
///
/// \param Loc The location at which the request originated and is fulfilled.
void emitTaskyieldImpl(const LocationDescription &Loc);
/// Return the current thread ID.
///
/// \param Ident The ident (ident_t*) describing the query origin.
Value *getOrCreateThreadID(Value *Ident);
/// The OpenMPIRBuilder Configuration
OpenMPIRBuilderConfig Config;
/// The underlying LLVM-IR module
Module &M;
/// The LLVM-IR Builder used to create IR.
IRBuilder<> Builder;
/// Map to remember source location strings
StringMap<Constant *> SrcLocStrMap;
/// Map to remember existing ident_t*.
DenseMap<std::pair<Constant *, uint64_t>, Constant *> IdentMap;
/// Info manager to keep track of target regions.
OffloadEntriesInfoManager OffloadInfoManager;
/// The target triple of the underlying module.
const Triple T;
/// Helper that contains information about regions we need to outline
/// during finalization.
struct OutlineInfo {
using PostOutlineCBTy = std::function<void(Function &)>;
PostOutlineCBTy PostOutlineCB;
BasicBlock *EntryBB, *ExitBB, *OuterAllocaBB;
SmallVector<Value *, 2> ExcludeArgsFromAggregate;
/// Collect all blocks in between EntryBB and ExitBB in both the given
/// vector and set.
void collectBlocks(SmallPtrSetImpl<BasicBlock *> &BlockSet,
SmallVectorImpl<BasicBlock *> &BlockVector);
/// Return the function that contains the region to be outlined.
Function *getFunction() const { return EntryBB->getParent(); }
};
/// Collection of regions that need to be outlined during finalization.
SmallVector<OutlineInfo, 16> OutlineInfos;
/// A collection of candidate target functions that's constant allocas will
/// attempt to be raised on a call of finalize after all currently enqueued
/// outline info's have been processed.
SmallVector<llvm::Function *, 16> ConstantAllocaRaiseCandidates;
/// Collection of owned canonical loop objects that eventually need to be
/// free'd.
std::forward_list<CanonicalLoopInfo> LoopInfos;
/// Add a new region that will be outlined later.
void addOutlineInfo(OutlineInfo &&OI) { OutlineInfos.emplace_back(OI); }
/// An ordered map of auto-generated variables to their unique names.
/// It stores variables with the following names: 1) ".gomp_critical_user_" +
/// <critical_section_name> + ".var" for "omp critical" directives; 2)
/// <mangled_name_for_global_var> + ".cache." for cache for threadprivate
/// variables.
StringMap<GlobalVariable *, BumpPtrAllocator> InternalVars;
/// Computes the size of type in bytes.
Value *getSizeInBytes(Value *BasePtr);
// Emit a branch from the current block to the Target block only if
// the current block has a terminator.
void emitBranch(BasicBlock *Target);
// If BB has no use then delete it and return. Else place BB after the current
// block, if possible, or else at the end of the function. Also add a branch
// from current block to BB if current block does not have a terminator.
void emitBlock(BasicBlock *BB, Function *CurFn, bool IsFinished = false);
/// Emits code for OpenMP 'if' clause using specified \a BodyGenCallbackTy
/// Here is the logic:
/// if (Cond) {
/// ThenGen();
/// } else {
/// ElseGen();
/// }
void emitIfClause(Value *Cond, BodyGenCallbackTy ThenGen,
BodyGenCallbackTy ElseGen, InsertPointTy AllocaIP = {});
/// Create the global variable holding the offload mappings information.
GlobalVariable *createOffloadMaptypes(SmallVectorImpl<uint64_t> &Mappings,
std::string VarName);
/// Create the global variable holding the offload names information.
GlobalVariable *
createOffloadMapnames(SmallVectorImpl<llvm::Constant *> &Names,
std::string VarName);
struct MapperAllocas {
AllocaInst *ArgsBase = nullptr;
AllocaInst *Args = nullptr;
AllocaInst *ArgSizes = nullptr;
};
/// Create the allocas instruction used in call to mapper functions.
void createMapperAllocas(const LocationDescription &Loc,
InsertPointTy AllocaIP, unsigned NumOperands,
struct MapperAllocas &MapperAllocas);
/// Create the call for the target mapper function.
/// \param Loc The source location description.
/// \param MapperFunc Function to be called.
/// \param SrcLocInfo Source location information global.
/// \param MaptypesArg The argument types.
/// \param MapnamesArg The argument names.
/// \param MapperAllocas The AllocaInst used for the call.
/// \param DeviceID Device ID for the call.
/// \param NumOperands Number of operands in the call.
void emitMapperCall(const LocationDescription &Loc, Function *MapperFunc,
Value *SrcLocInfo, Value *MaptypesArg, Value *MapnamesArg,
struct MapperAllocas &MapperAllocas, int64_t DeviceID,
unsigned NumOperands);
/// Container for the arguments used to pass data to the runtime library.
struct TargetDataRTArgs {
/// The array of base pointer passed to the runtime library.
Value *BasePointersArray = nullptr;
/// The array of section pointers passed to the runtime library.
Value *PointersArray = nullptr;
/// The array of sizes passed to the runtime library.
Value *SizesArray = nullptr;
/// The array of map types passed to the runtime library for the beginning
/// of the region or for the entire region if there are no separate map
/// types for the region end.
Value *MapTypesArray = nullptr;
/// The array of map types passed to the runtime library for the end of the
/// region, or nullptr if there are no separate map types for the region
/// end.
Value *MapTypesArrayEnd = nullptr;
/// The array of user-defined mappers passed to the runtime library.
Value *MappersArray = nullptr;
/// The array of original declaration names of mapped pointers sent to the
/// runtime library for debugging
Value *MapNamesArray = nullptr;
explicit TargetDataRTArgs() {}
explicit TargetDataRTArgs(Value *BasePointersArray, Value *PointersArray,
Value *SizesArray, Value *MapTypesArray,
Value *MapTypesArrayEnd, Value *MappersArray,
Value *MapNamesArray)
: BasePointersArray(BasePointersArray), PointersArray(PointersArray),
SizesArray(SizesArray), MapTypesArray(MapTypesArray),
MapTypesArrayEnd(MapTypesArrayEnd), MappersArray(MappersArray),
MapNamesArray(MapNamesArray) {}
};
/// Data structure that contains the needed information to construct the
/// kernel args vector.
struct TargetKernelArgs {
/// Number of arguments passed to the runtime library.
unsigned NumTargetItems;
/// Arguments passed to the runtime library
TargetDataRTArgs RTArgs;
/// The number of iterations
Value *NumIterations;
/// The number of teams.
Value *NumTeams;
/// The number of threads.
Value *NumThreads;
/// The size of the dynamic shared memory.
Value *DynCGGroupMem;
/// True if the kernel has 'no wait' clause.
bool HasNoWait;
/// Constructor for TargetKernelArgs
TargetKernelArgs(unsigned NumTargetItems, TargetDataRTArgs RTArgs,
Value *NumIterations, Value *NumTeams, Value *NumThreads,
Value *DynCGGroupMem, bool HasNoWait)
: NumTargetItems(NumTargetItems), RTArgs(RTArgs),
NumIterations(NumIterations), NumTeams(NumTeams),
NumThreads(NumThreads), DynCGGroupMem(DynCGGroupMem),
HasNoWait(HasNoWait) {}
};
/// Create the kernel args vector used by emitTargetKernel. This function
/// creates various constant values that are used in the resulting args
/// vector.
static void getKernelArgsVector(TargetKernelArgs &KernelArgs,
IRBuilderBase &Builder,
SmallVector<Value *> &ArgsVector);
/// Struct that keeps the information that should be kept throughout
/// a 'target data' region.
class TargetDataInfo {
/// Set to true if device pointer information have to be obtained.
bool RequiresDevicePointerInfo = false;
/// Set to true if Clang emits separate runtime calls for the beginning and
/// end of the region. These calls might have separate map type arrays.
bool SeparateBeginEndCalls = false;
public:
TargetDataRTArgs RTArgs;
SmallMapVector<const Value *, std::pair<Value *, Value *>, 4>
DevicePtrInfoMap;
/// Indicate whether any user-defined mapper exists.
bool HasMapper = false;
/// The total number of pointers passed to the runtime library.
unsigned NumberOfPtrs = 0u;
explicit TargetDataInfo() {}
explicit TargetDataInfo(bool RequiresDevicePointerInfo,
bool SeparateBeginEndCalls)
: RequiresDevicePointerInfo(RequiresDevicePointerInfo),
SeparateBeginEndCalls(SeparateBeginEndCalls) {}
/// Clear information about the data arrays.
void clearArrayInfo() {
RTArgs = TargetDataRTArgs();
HasMapper = false;
NumberOfPtrs = 0u;
}
/// Return true if the current target data information has valid arrays.
bool isValid() {
return RTArgs.BasePointersArray && RTArgs.PointersArray &&
RTArgs.SizesArray && RTArgs.MapTypesArray &&
(!HasMapper || RTArgs.MappersArray) && NumberOfPtrs;
}
bool requiresDevicePointerInfo() { return RequiresDevicePointerInfo; }
bool separateBeginEndCalls() { return SeparateBeginEndCalls; }
};
enum class DeviceInfoTy { None, Pointer, Address };
using MapValuesArrayTy = SmallVector<Value *, 4>;
using MapDeviceInfoArrayTy = SmallVector<DeviceInfoTy, 4>;
using MapFlagsArrayTy = SmallVector<omp::OpenMPOffloadMappingFlags, 4>;
using MapNamesArrayTy = SmallVector<Constant *, 4>;
using MapDimArrayTy = SmallVector<uint64_t, 4>;
using MapNonContiguousArrayTy = SmallVector<MapValuesArrayTy, 4>;
/// This structure contains combined information generated for mappable
/// clauses, including base pointers, pointers, sizes, map types, user-defined
/// mappers, and non-contiguous information.
struct MapInfosTy {
struct StructNonContiguousInfo {
bool IsNonContiguous = false;
MapDimArrayTy Dims;
MapNonContiguousArrayTy Offsets;
MapNonContiguousArrayTy Counts;
MapNonContiguousArrayTy Strides;
};
MapValuesArrayTy BasePointers;
MapValuesArrayTy Pointers;
MapDeviceInfoArrayTy DevicePointers;
MapValuesArrayTy Sizes;
MapFlagsArrayTy Types;
MapNamesArrayTy Names;
StructNonContiguousInfo NonContigInfo;
/// Append arrays in \a CurInfo.
void append(MapInfosTy &CurInfo) {
BasePointers.append(CurInfo.BasePointers.begin(),
CurInfo.BasePointers.end());
Pointers.append(CurInfo.Pointers.begin(), CurInfo.Pointers.end());
DevicePointers.append(CurInfo.DevicePointers.begin(),
CurInfo.DevicePointers.end());
Sizes.append(CurInfo.Sizes.begin(), CurInfo.Sizes.end());
Types.append(CurInfo.Types.begin(), CurInfo.Types.end());
Names.append(CurInfo.Names.begin(), CurInfo.Names.end());
NonContigInfo.Dims.append(CurInfo.NonContigInfo.Dims.begin(),
CurInfo.NonContigInfo.Dims.end());
NonContigInfo.Offsets.append(CurInfo.NonContigInfo.Offsets.begin(),
CurInfo.NonContigInfo.Offsets.end());
NonContigInfo.Counts.append(CurInfo.NonContigInfo.Counts.begin(),
CurInfo.NonContigInfo.Counts.end());
NonContigInfo.Strides.append(CurInfo.NonContigInfo.Strides.begin(),
CurInfo.NonContigInfo.Strides.end());
}
};
/// Callback function type for functions emitting the host fallback code that
/// is executed when the kernel launch fails. It takes an insertion point as
/// parameter where the code should be emitted. It returns an insertion point
/// that points right after after the emitted code.
using EmitFallbackCallbackTy = function_ref<InsertPointTy(InsertPointTy)>;
/// Generate a target region entry call and host fallback call.
///
/// \param Loc The location at which the request originated and is fulfilled.
/// \param OutlinedFn The outlined kernel function.
/// \param OutlinedFnID The ooulined function ID.
/// \param EmitTargetCallFallbackCB Call back function to generate host
/// fallback code.
/// \param Args Data structure holding information about the kernel arguments.
/// \param DeviceID Identifier for the device via the 'device' clause.
/// \param RTLoc Source location identifier
/// \param AllocaIP The insertion point to be used for alloca instructions.
InsertPointTy emitKernelLaunch(
const LocationDescription &Loc, Function *OutlinedFn, Value *OutlinedFnID,
EmitFallbackCallbackTy EmitTargetCallFallbackCB, TargetKernelArgs &Args,
Value *DeviceID, Value *RTLoc, InsertPointTy AllocaIP);
/// Generate a target-task for the target construct
///
/// \param OutlinedFn The outlined device/target kernel function.
/// \param OutlinedFnID The ooulined function ID.
/// \param EmitTargetCallFallbackCB Call back function to generate host
/// fallback code.
/// \param Args Data structure holding information about the kernel arguments.
/// \param DeviceID Identifier for the device via the 'device' clause.
/// \param RTLoc Source location identifier
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param Dependencies Vector of DependData objects holding information of
/// dependencies as specified by the 'depend' clause.
/// \param HasNoWait True if the target construct had 'nowait' on it, false
/// otherwise
InsertPointTy emitTargetTask(
Function *OutlinedFn, Value *OutlinedFnID,
EmitFallbackCallbackTy EmitTargetCallFallbackCB, TargetKernelArgs &Args,
Value *DeviceID, Value *RTLoc, InsertPointTy AllocaIP,
SmallVector<OpenMPIRBuilder::DependData> &Dependencies, bool HasNoWait);
/// Emit the arguments to be passed to the runtime library based on the
/// arrays of base pointers, pointers, sizes, map types, and mappers. If
/// ForEndCall, emit map types to be passed for the end of the region instead
/// of the beginning.
void emitOffloadingArraysArgument(IRBuilderBase &Builder,
OpenMPIRBuilder::TargetDataRTArgs &RTArgs,
OpenMPIRBuilder::TargetDataInfo &Info,
bool EmitDebug = false,
bool ForEndCall = false);
/// Emit an array of struct descriptors to be assigned to the offload args.
void emitNonContiguousDescriptor(InsertPointTy AllocaIP,
InsertPointTy CodeGenIP,
MapInfosTy &CombinedInfo,
TargetDataInfo &Info);
/// Emit the arrays used to pass the captures and map information to the
/// offloading runtime library. If there is no map or capture information,
/// return nullptr by reference.
void emitOffloadingArrays(
InsertPointTy AllocaIP, InsertPointTy CodeGenIP, MapInfosTy &CombinedInfo,
TargetDataInfo &Info, bool IsNonContiguous = false,
function_ref<void(unsigned int, Value *)> DeviceAddrCB = nullptr,
function_ref<Value *(unsigned int)> CustomMapperCB = nullptr);
/// Creates offloading entry for the provided entry ID \a ID, address \a
/// Addr, size \a Size, and flags \a Flags.
void createOffloadEntry(Constant *ID, Constant *Addr, uint64_t Size,
int32_t Flags, GlobalValue::LinkageTypes,
StringRef Name = "");
/// The kind of errors that can occur when emitting the offload entries and
/// metadata.
enum EmitMetadataErrorKind {
EMIT_MD_TARGET_REGION_ERROR,
EMIT_MD_DECLARE_TARGET_ERROR,
EMIT_MD_GLOBAL_VAR_LINK_ERROR
};
/// Callback function type
using EmitMetadataErrorReportFunctionTy =
std::function<void(EmitMetadataErrorKind, TargetRegionEntryInfo)>;
// Emit the offloading entries and metadata so that the device codegen side
// can easily figure out what to emit. The produced metadata looks like
// this:
//
// !omp_offload.info = !{!1, ...}
//
// We only generate metadata for function that contain target regions.
void createOffloadEntriesAndInfoMetadata(
EmitMetadataErrorReportFunctionTy &ErrorReportFunction);
public:
/// Generator for __kmpc_copyprivate
///
/// \param Loc The source location description.
/// \param BufSize Number of elements in the buffer.
/// \param CpyBuf List of pointers to data to be copied.
/// \param CpyFn function to call for copying data.
/// \param DidIt flag variable; 1 for 'single' thread, 0 otherwise.
///
/// \return The insertion position *after* the CopyPrivate call.
InsertPointTy createCopyPrivate(const LocationDescription &Loc,
llvm::Value *BufSize, llvm::Value *CpyBuf,
llvm::Value *CpyFn, llvm::Value *DidIt);
/// Generator for '#omp single'
///
/// \param Loc The source location description.
/// \param BodyGenCB Callback that will generate the region code.
/// \param FiniCB Callback to finalize variable copies.
/// \param IsNowait If false, a barrier is emitted.
/// \param CPVars copyprivate variables.
/// \param CPFuncs copy functions to use for each copyprivate variable.
///
/// \returns The insertion position *after* the single call.
InsertPointTy createSingle(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB, bool IsNowait,
ArrayRef<llvm::Value *> CPVars = {},
ArrayRef<llvm::Function *> CPFuncs = {});
/// Generator for '#omp master'
///
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the region code.
/// \param FiniCB Callback to finalize variable copies.
///
/// \returns The insertion position *after* the master.
InsertPointTy createMaster(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB);
/// Generator for '#omp masked'
///
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the region code.
/// \param FiniCB Callback to finialize variable copies.
///
/// \returns The insertion position *after* the masked.
InsertPointTy createMasked(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB, Value *Filter);
/// Generator for '#omp critical'
///
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the region body code.
/// \param FiniCB Callback to finalize variable copies.
/// \param CriticalName name of the lock used by the critical directive
/// \param HintInst Hint Instruction for hint clause associated with critical
///
/// \returns The insertion position *after* the critical.
InsertPointTy createCritical(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB,
StringRef CriticalName, Value *HintInst);
/// Generator for '#omp ordered depend (source | sink)'
///
/// \param Loc The insert and source location description.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param NumLoops The number of loops in depend clause.
/// \param StoreValues The value will be stored in vector address.
/// \param Name The name of alloca instruction.
/// \param IsDependSource If true, depend source; otherwise, depend sink.
///
/// \return The insertion position *after* the ordered.
InsertPointTy createOrderedDepend(const LocationDescription &Loc,
InsertPointTy AllocaIP, unsigned NumLoops,
ArrayRef<llvm::Value *> StoreValues,
const Twine &Name, bool IsDependSource);
/// Generator for '#omp ordered [threads | simd]'
///
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the region code.
/// \param FiniCB Callback to finalize variable copies.
/// \param IsThreads If true, with threads clause or without clause;
/// otherwise, with simd clause;
///
/// \returns The insertion position *after* the ordered.
InsertPointTy createOrderedThreadsSimd(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB,
bool IsThreads);
/// Generator for '#omp sections'
///
/// \param Loc The insert and source location description.
/// \param AllocaIP The insertion points to be used for alloca instructions.
/// \param SectionCBs Callbacks that will generate body of each section.
/// \param PrivCB Callback to copy a given variable (think copy constructor).
/// \param FiniCB Callback to finalize variable copies.
/// \param IsCancellable Flag to indicate a cancellable parallel region.
/// \param IsNowait If true, barrier - to ensure all sections are executed
/// before moving forward will not be generated.
/// \returns The insertion position *after* the sections.
InsertPointTy createSections(const LocationDescription &Loc,
InsertPointTy AllocaIP,
ArrayRef<StorableBodyGenCallbackTy> SectionCBs,
PrivatizeCallbackTy PrivCB,
FinalizeCallbackTy FiniCB, bool IsCancellable,
bool IsNowait);
/// Generator for '#omp section'
///
/// \param Loc The insert and source location description.
/// \param BodyGenCB Callback that will generate the region body code.
/// \param FiniCB Callback to finalize variable copies.
/// \returns The insertion position *after* the section.
InsertPointTy createSection(const LocationDescription &Loc,
BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB);
/// Generator for `#omp teams`
///
/// \param Loc The location where the teams construct was encountered.
/// \param BodyGenCB Callback that will generate the region code.
/// \param NumTeamsLower Lower bound on number of teams. If this is nullptr,
/// it is as if lower bound is specified as equal to upperbound. If
/// this is non-null, then upperbound must also be non-null.
/// \param NumTeamsUpper Upper bound on the number of teams.
/// \param ThreadLimit on the number of threads that may participate in a
/// contention group created by each team.
/// \param IfExpr is the integer argument value of the if condition on the
/// teams clause.
InsertPointTy
createTeams(const LocationDescription &Loc, BodyGenCallbackTy BodyGenCB,
Value *NumTeamsLower = nullptr, Value *NumTeamsUpper = nullptr,
Value *ThreadLimit = nullptr, Value *IfExpr = nullptr);
/// Generate conditional branch and relevant BasicBlocks through which private
/// threads copy the 'copyin' variables from Master copy to threadprivate
/// copies.
///
/// \param IP insertion block for copyin conditional
/// \param MasterVarPtr a pointer to the master variable
/// \param PrivateVarPtr a pointer to the threadprivate variable
/// \param IntPtrTy Pointer size type
/// \param BranchtoEnd Create a branch between the copyin.not.master blocks
// and copy.in.end block
///
/// \returns The insertion point where copying operation to be emitted.
InsertPointTy createCopyinClauseBlocks(InsertPointTy IP, Value *MasterAddr,
Value *PrivateAddr,
llvm::IntegerType *IntPtrTy,
bool BranchtoEnd = true);
/// Create a runtime call for kmpc_Alloc
///
/// \param Loc The insert and source location description.
/// \param Size Size of allocated memory space
/// \param Allocator Allocator information instruction
/// \param Name Name of call Instruction for OMP_alloc
///
/// \returns CallInst to the OMP_Alloc call
CallInst *createOMPAlloc(const LocationDescription &Loc, Value *Size,
Value *Allocator, std::string Name = "");
/// Create a runtime call for kmpc_free
///
/// \param Loc The insert and source location description.
/// \param Addr Address of memory space to be freed
/// \param Allocator Allocator information instruction
/// \param Name Name of call Instruction for OMP_Free
///
/// \returns CallInst to the OMP_Free call
CallInst *createOMPFree(const LocationDescription &Loc, Value *Addr,
Value *Allocator, std::string Name = "");
/// Create a runtime call for kmpc_threadprivate_cached
///
/// \param Loc The insert and source location description.
/// \param Pointer pointer to data to be cached
/// \param Size size of data to be cached
/// \param Name Name of call Instruction for callinst
///
/// \returns CallInst to the thread private cache call.
CallInst *createCachedThreadPrivate(const LocationDescription &Loc,
llvm::Value *Pointer,
llvm::ConstantInt *Size,
const llvm::Twine &Name = Twine(""));
/// Create a runtime call for __tgt_interop_init
///
/// \param Loc The insert and source location description.
/// \param InteropVar variable to be allocated
/// \param InteropType type of interop operation
/// \param Device devide to which offloading will occur
/// \param NumDependences number of dependence variables
/// \param DependenceAddress pointer to dependence variables
/// \param HaveNowaitClause does nowait clause exist
///
/// \returns CallInst to the __tgt_interop_init call
CallInst *createOMPInteropInit(const LocationDescription &Loc,
Value *InteropVar,
omp::OMPInteropType InteropType, Value *Device,
Value *NumDependences,
Value *DependenceAddress,
bool HaveNowaitClause);
/// Create a runtime call for __tgt_interop_destroy
///
/// \param Loc The insert and source location description.
/// \param InteropVar variable to be allocated
/// \param Device devide to which offloading will occur
/// \param NumDependences number of dependence variables
/// \param DependenceAddress pointer to dependence variables
/// \param HaveNowaitClause does nowait clause exist
///
/// \returns CallInst to the __tgt_interop_destroy call
CallInst *createOMPInteropDestroy(const LocationDescription &Loc,
Value *InteropVar, Value *Device,
Value *NumDependences,
Value *DependenceAddress,
bool HaveNowaitClause);
/// Create a runtime call for __tgt_interop_use
///
/// \param Loc The insert and source location description.
/// \param InteropVar variable to be allocated
/// \param Device devide to which offloading will occur
/// \param NumDependences number of dependence variables
/// \param DependenceAddress pointer to dependence variables
/// \param HaveNowaitClause does nowait clause exist
///
/// \returns CallInst to the __tgt_interop_use call
CallInst *createOMPInteropUse(const LocationDescription &Loc,
Value *InteropVar, Value *Device,
Value *NumDependences, Value *DependenceAddress,
bool HaveNowaitClause);
/// The `omp target` interface
///
/// For more information about the usage of this interface,
/// \see openmp/libomptarget/deviceRTLs/common/include/target.h
///
///{
/// Create a runtime call for kmpc_target_init
///
/// \param Loc The insert and source location description.
/// \param IsSPMD Flag to indicate if the kernel is an SPMD kernel or not.
/// \param MinThreads Minimal number of threads, or 0.
/// \param MaxThreads Maximal number of threads, or 0.
/// \param MinTeams Minimal number of teams, or 0.
/// \param MaxTeams Maximal number of teams, or 0.
InsertPointTy createTargetInit(const LocationDescription &Loc, bool IsSPMD,
int32_t MinThreadsVal = 0,
int32_t MaxThreadsVal = 0,
int32_t MinTeamsVal = 0,
int32_t MaxTeamsVal = 0);
/// Create a runtime call for kmpc_target_deinit
///
/// \param Loc The insert and source location description.
/// \param TeamsReductionDataSize The maximal size of all the reduction data
/// for teams reduction.
/// \param TeamsReductionBufferLength The number of elements (each of up to
/// \p TeamsReductionDataSize size), in the teams reduction buffer.
void createTargetDeinit(const LocationDescription &Loc,
int32_t TeamsReductionDataSize = 0,
int32_t TeamsReductionBufferLength = 1024);
///}
/// Helpers to read/write kernel annotations from the IR.
///
///{
/// Read/write a bounds on threads for \p Kernel. Read will return 0 if none
/// is set.
static std::pair<int32_t, int32_t>
readThreadBoundsForKernel(const Triple &T, Function &Kernel);
static void writeThreadBoundsForKernel(const Triple &T, Function &Kernel,
int32_t LB, int32_t UB);
/// Read/write a bounds on teams for \p Kernel. Read will return 0 if none
/// is set.
static std::pair<int32_t, int32_t> readTeamBoundsForKernel(const Triple &T,
Function &Kernel);
static void writeTeamsForKernel(const Triple &T, Function &Kernel, int32_t LB,
int32_t UB);
///}
private:
// Sets the function attributes expected for the outlined function
void setOutlinedTargetRegionFunctionAttributes(Function *OutlinedFn);
// Creates the function ID/Address for the given outlined function.
// In the case of an embedded device function the address of the function is
// used, in the case of a non-offload function a constant is created.
Constant *createOutlinedFunctionID(Function *OutlinedFn,
StringRef EntryFnIDName);
// Creates the region entry address for the outlined function
Constant *createTargetRegionEntryAddr(Function *OutlinedFunction,
StringRef EntryFnName);
public:
/// Functions used to generate a function with the given name.
using FunctionGenCallback = std::function<Function *(StringRef FunctionName)>;
/// Create a unique name for the entry function using the source location
/// information of the current target region. The name will be something like:
///
/// __omp_offloading_DD_FFFF_PP_lBB[_CC]
///
/// where DD_FFFF is an ID unique to the file (device and file IDs), PP is the
/// mangled name of the function that encloses the target region and BB is the
/// line number of the target region. CC is a count added when more than one
/// region is located at the same location.
///
/// If this target outline function is not an offload entry, we don't need to
/// register it. This may happen if it is guarded by an if clause that is
/// false at compile time, or no target archs have been specified.
///
/// The created target region ID is used by the runtime library to identify
/// the current target region, so it only has to be unique and not
/// necessarily point to anything. It could be the pointer to the outlined
/// function that implements the target region, but we aren't using that so
/// that the compiler doesn't need to keep that, and could therefore inline
/// the host function if proven worthwhile during optimization. In the other
/// hand, if emitting code for the device, the ID has to be the function
/// address so that it can retrieved from the offloading entry and launched
/// by the runtime library. We also mark the outlined function to have
/// external linkage in case we are emitting code for the device, because
/// these functions will be entry points to the device.
///
/// \param InfoManager The info manager keeping track of the offload entries
/// \param EntryInfo The entry information about the function
/// \param GenerateFunctionCallback The callback function to generate the code
/// \param OutlinedFunction Pointer to the outlined function
/// \param EntryFnIDName Name of the ID o be created
void emitTargetRegionFunction(TargetRegionEntryInfo &EntryInfo,
FunctionGenCallback &GenerateFunctionCallback,
bool IsOffloadEntry, Function *&OutlinedFn,
Constant *&OutlinedFnID);
/// Registers the given function and sets up the attribtues of the function
/// Returns the FunctionID.
///
/// \param InfoManager The info manager keeping track of the offload entries
/// \param EntryInfo The entry information about the function
/// \param OutlinedFunction Pointer to the outlined function
/// \param EntryFnName Name of the outlined function
/// \param EntryFnIDName Name of the ID o be created
Constant *registerTargetRegionFunction(TargetRegionEntryInfo &EntryInfo,
Function *OutlinedFunction,
StringRef EntryFnName,
StringRef EntryFnIDName);
/// Type of BodyGen to use for region codegen
///
/// Priv: If device pointer privatization is required, emit the body of the
/// region here. It will have to be duplicated: with and without
/// privatization.
/// DupNoPriv: If we need device pointer privatization, we need
/// to emit the body of the region with no privatization in the 'else' branch
/// of the conditional.
/// NoPriv: If we don't require privatization of device
/// pointers, we emit the body in between the runtime calls. This avoids
/// duplicating the body code.
enum BodyGenTy { Priv, DupNoPriv, NoPriv };
/// Callback type for creating the map infos for the kernel parameters.
/// \param CodeGenIP is the insertion point where code should be generated,
/// if any.
using GenMapInfoCallbackTy =
function_ref<MapInfosTy &(InsertPointTy CodeGenIP)>;
/// Generator for '#omp target data'
///
/// \param Loc The location where the target data construct was encountered.
/// \param AllocaIP The insertion points to be used for alloca instructions.
/// \param CodeGenIP The insertion point at which the target directive code
/// should be placed.
/// \param IsBegin If true then emits begin mapper call otherwise emits
/// end mapper call.
/// \param DeviceID Stores the DeviceID from the device clause.
/// \param IfCond Value which corresponds to the if clause condition.
/// \param Info Stores all information realted to the Target Data directive.
/// \param GenMapInfoCB Callback that populates the MapInfos and returns.
/// \param BodyGenCB Optional Callback to generate the region code.
/// \param DeviceAddrCB Optional callback to generate code related to
/// use_device_ptr and use_device_addr.
/// \param CustomMapperCB Optional callback to generate code related to
/// custom mappers.
OpenMPIRBuilder::InsertPointTy createTargetData(
const LocationDescription &Loc, InsertPointTy AllocaIP,
InsertPointTy CodeGenIP, Value *DeviceID, Value *IfCond,
TargetDataInfo &Info, GenMapInfoCallbackTy GenMapInfoCB,
omp::RuntimeFunction *MapperFunc = nullptr,
function_ref<InsertPointTy(InsertPointTy CodeGenIP,
BodyGenTy BodyGenType)>
BodyGenCB = nullptr,
function_ref<void(unsigned int, Value *)> DeviceAddrCB = nullptr,
function_ref<Value *(unsigned int)> CustomMapperCB = nullptr,
Value *SrcLocInfo = nullptr);
using TargetBodyGenCallbackTy = function_ref<InsertPointTy(
InsertPointTy AllocaIP, InsertPointTy CodeGenIP)>;
using TargetGenArgAccessorsCallbackTy = function_ref<InsertPointTy(
Argument &Arg, Value *Input, Value *&RetVal, InsertPointTy AllocaIP,
InsertPointTy CodeGenIP)>;
/// Generator for '#omp target'
///
/// \param Loc where the target data construct was encountered.
/// \param CodeGenIP The insertion point where the call to the outlined
/// function should be emitted.
/// \param EntryInfo The entry information about the function.
/// \param NumTeams Number of teams specified in the num_teams clause.
/// \param NumThreads Number of teams specified in the thread_limit clause.
/// \param Inputs The input values to the region that will be passed.
/// as arguments to the outlined function.
/// \param BodyGenCB Callback that will generate the region code.
/// \param ArgAccessorFuncCB Callback that will generate accessors
/// instructions for passed in target arguments where neccessary
/// \param Dependencies A vector of DependData objects that carry
// dependency information as passed in the depend clause
InsertPointTy createTarget(const LocationDescription &Loc,
OpenMPIRBuilder::InsertPointTy AllocaIP,
OpenMPIRBuilder::InsertPointTy CodeGenIP,
TargetRegionEntryInfo &EntryInfo, int32_t NumTeams,
int32_t NumThreads,
SmallVectorImpl<Value *> &Inputs,
GenMapInfoCallbackTy GenMapInfoCB,
TargetBodyGenCallbackTy BodyGenCB,
TargetGenArgAccessorsCallbackTy ArgAccessorFuncCB,
SmallVector<DependData> Dependencies = {});
/// Returns __kmpc_for_static_init_* runtime function for the specified
/// size \a IVSize and sign \a IVSigned. Will create a distribute call
/// __kmpc_distribute_static_init* if \a IsGPUDistribute is set.
FunctionCallee createForStaticInitFunction(unsigned IVSize, bool IVSigned,
bool IsGPUDistribute);
/// Returns __kmpc_dispatch_init_* runtime function for the specified
/// size \a IVSize and sign \a IVSigned.
FunctionCallee createDispatchInitFunction(unsigned IVSize, bool IVSigned);
/// Returns __kmpc_dispatch_next_* runtime function for the specified
/// size \a IVSize and sign \a IVSigned.
FunctionCallee createDispatchNextFunction(unsigned IVSize, bool IVSigned);
/// Returns __kmpc_dispatch_fini_* runtime function for the specified
/// size \a IVSize and sign \a IVSigned.
FunctionCallee createDispatchFiniFunction(unsigned IVSize, bool IVSigned);
/// Returns __kmpc_dispatch_deinit runtime function.
FunctionCallee createDispatchDeinitFunction();
/// Declarations for LLVM-IR types (simple, array, function and structure) are
/// generated below. Their names are defined and used in OpenMPKinds.def. Here
/// we provide the declarations, the initializeTypes function will provide the
/// values.
///
///{
#define OMP_TYPE(VarName, InitValue) Type *VarName = nullptr;
#define OMP_ARRAY_TYPE(VarName, ElemTy, ArraySize) \
ArrayType *VarName##Ty = nullptr; \
PointerType *VarName##PtrTy = nullptr;
#define OMP_FUNCTION_TYPE(VarName, IsVarArg, ReturnType, ...) \
FunctionType *VarName = nullptr; \
PointerType *VarName##Ptr = nullptr;
#define OMP_STRUCT_TYPE(VarName, StrName, ...) \
StructType *VarName = nullptr; \
PointerType *VarName##Ptr = nullptr;
#include "llvm/Frontend/OpenMP/OMPKinds.def"
///}
private:
/// Create all simple and struct types exposed by the runtime and remember
/// the llvm::PointerTypes of them for easy access later.
void initializeTypes(Module &M);
/// Common interface for generating entry calls for OMP Directives.
/// if the directive has a region/body, It will set the insertion
/// point to the body
///
/// \param OMPD Directive to generate entry blocks for
/// \param EntryCall Call to the entry OMP Runtime Function
/// \param ExitBB block where the region ends.
/// \param Conditional indicate if the entry call result will be used
/// to evaluate a conditional of whether a thread will execute
/// body code or not.
///
/// \return The insertion position in exit block
InsertPointTy emitCommonDirectiveEntry(omp::Directive OMPD, Value *EntryCall,
BasicBlock *ExitBB,
bool Conditional = false);
/// Common interface to finalize the region
///
/// \param OMPD Directive to generate exiting code for
/// \param FinIP Insertion point for emitting Finalization code and exit call
/// \param ExitCall Call to the ending OMP Runtime Function
/// \param HasFinalize indicate if the directive will require finalization
/// and has a finalization callback in the stack that
/// should be called.
///
/// \return The insertion position in exit block
InsertPointTy emitCommonDirectiveExit(omp::Directive OMPD,
InsertPointTy FinIP,
Instruction *ExitCall,
bool HasFinalize = true);
/// Common Interface to generate OMP inlined regions
///
/// \param OMPD Directive to generate inlined region for
/// \param EntryCall Call to the entry OMP Runtime Function
/// \param ExitCall Call to the ending OMP Runtime Function
/// \param BodyGenCB Body code generation callback.
/// \param FiniCB Finalization Callback. Will be called when finalizing region
/// \param Conditional indicate if the entry call result will be used
/// to evaluate a conditional of whether a thread will execute
/// body code or not.
/// \param HasFinalize indicate if the directive will require finalization
/// and has a finalization callback in the stack that
/// should be called.
/// \param IsCancellable if HasFinalize is set to true, indicate if the
/// the directive should be cancellable.
/// \return The insertion point after the region
InsertPointTy
EmitOMPInlinedRegion(omp::Directive OMPD, Instruction *EntryCall,
Instruction *ExitCall, BodyGenCallbackTy BodyGenCB,
FinalizeCallbackTy FiniCB, bool Conditional = false,
bool HasFinalize = true, bool IsCancellable = false);
/// Get the platform-specific name separator.
/// \param Parts different parts of the final name that needs separation
/// \param FirstSeparator First separator used between the initial two
/// parts of the name.
/// \param Separator separator used between all of the rest consecutive
/// parts of the name
static std::string getNameWithSeparators(ArrayRef<StringRef> Parts,
StringRef FirstSeparator,
StringRef Separator);
/// Returns corresponding lock object for the specified critical region
/// name. If the lock object does not exist it is created, otherwise the
/// reference to the existing copy is returned.
/// \param CriticalName Name of the critical region.
///
Value *getOMPCriticalRegionLock(StringRef CriticalName);
/// Callback type for Atomic Expression update
/// ex:
/// \code{.cpp}
/// unsigned x = 0;
/// #pragma omp atomic update
/// x = Expr(x_old); //Expr() is any legal operation
/// \endcode
///
/// \param XOld the value of the atomic memory address to use for update
/// \param IRB reference to the IRBuilder to use
///
/// \returns Value to update X to.
using AtomicUpdateCallbackTy =
const function_ref<Value *(Value *XOld, IRBuilder<> &IRB)>;
private:
enum AtomicKind { Read, Write, Update, Capture, Compare };
/// Determine whether to emit flush or not
///
/// \param Loc The insert and source location description.
/// \param AO The required atomic ordering
/// \param AK The OpenMP atomic operation kind used.
///
/// \returns wether a flush was emitted or not
bool checkAndEmitFlushAfterAtomic(const LocationDescription &Loc,
AtomicOrdering AO, AtomicKind AK);
/// Emit atomic update for constructs: X = X BinOp Expr ,or X = Expr BinOp X
/// For complex Operations: X = UpdateOp(X) => CmpExch X, old_X, UpdateOp(X)
/// Only Scalar data types.
///
/// \param AllocaIP The insertion point to be used for alloca
/// instructions.
/// \param X The target atomic pointer to be updated
/// \param XElemTy The element type of the atomic pointer.
/// \param Expr The value to update X with.
/// \param AO Atomic ordering of the generated atomic
/// instructions.
/// \param RMWOp The binary operation used for update. If
/// operation is not supported by atomicRMW,
/// or belong to {FADD, FSUB, BAD_BINOP}.
/// Then a `cmpExch` based atomic will be generated.
/// \param UpdateOp Code generator for complex expressions that cannot be
/// expressed through atomicrmw instruction.
/// \param VolatileX true if \a X volatile?
/// \param IsXBinopExpr true if \a X is Left H.S. in Right H.S. part of the
/// update expression, false otherwise.
/// (e.g. true for X = X BinOp Expr)
///
/// \returns A pair of the old value of X before the update, and the value
/// used for the update.
std::pair<Value *, Value *>
emitAtomicUpdate(InsertPointTy AllocaIP, Value *X, Type *XElemTy, Value *Expr,
AtomicOrdering AO, AtomicRMWInst::BinOp RMWOp,
AtomicUpdateCallbackTy &UpdateOp, bool VolatileX,
bool IsXBinopExpr);
/// Emit the binary op. described by \p RMWOp, using \p Src1 and \p Src2 .
///
/// \Return The instruction
Value *emitRMWOpAsInstruction(Value *Src1, Value *Src2,
AtomicRMWInst::BinOp RMWOp);
public:
/// a struct to pack relevant information while generating atomic Ops
struct AtomicOpValue {
Value *Var = nullptr;
Type *ElemTy = nullptr;
bool IsSigned = false;
bool IsVolatile = false;
};
/// Emit atomic Read for : V = X --- Only Scalar data types.
///
/// \param Loc The insert and source location description.
/// \param X The target pointer to be atomically read
/// \param V Memory address where to store atomically read
/// value
/// \param AO Atomic ordering of the generated atomic
/// instructions.
///
/// \return Insertion point after generated atomic read IR.
InsertPointTy createAtomicRead(const LocationDescription &Loc,
AtomicOpValue &X, AtomicOpValue &V,
AtomicOrdering AO);
/// Emit atomic write for : X = Expr --- Only Scalar data types.
///
/// \param Loc The insert and source location description.
/// \param X The target pointer to be atomically written to
/// \param Expr The value to store.
/// \param AO Atomic ordering of the generated atomic
/// instructions.
///
/// \return Insertion point after generated atomic Write IR.
InsertPointTy createAtomicWrite(const LocationDescription &Loc,
AtomicOpValue &X, Value *Expr,
AtomicOrdering AO);
/// Emit atomic update for constructs: X = X BinOp Expr ,or X = Expr BinOp X
/// For complex Operations: X = UpdateOp(X) => CmpExch X, old_X, UpdateOp(X)
/// Only Scalar data types.
///
/// \param Loc The insert and source location description.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param X The target atomic pointer to be updated
/// \param Expr The value to update X with.
/// \param AO Atomic ordering of the generated atomic instructions.
/// \param RMWOp The binary operation used for update. If operation
/// is not supported by atomicRMW, or belong to
/// {FADD, FSUB, BAD_BINOP}. Then a `cmpExch` based
/// atomic will be generated.
/// \param UpdateOp Code generator for complex expressions that cannot be
/// expressed through atomicrmw instruction.
/// \param IsXBinopExpr true if \a X is Left H.S. in Right H.S. part of the
/// update expression, false otherwise.
/// (e.g. true for X = X BinOp Expr)
///
/// \return Insertion point after generated atomic update IR.
InsertPointTy createAtomicUpdate(const LocationDescription &Loc,
InsertPointTy AllocaIP, AtomicOpValue &X,
Value *Expr, AtomicOrdering AO,
AtomicRMWInst::BinOp RMWOp,
AtomicUpdateCallbackTy &UpdateOp,
bool IsXBinopExpr);
/// Emit atomic update for constructs: --- Only Scalar data types
/// V = X; X = X BinOp Expr ,
/// X = X BinOp Expr; V = X,
/// V = X; X = Expr BinOp X,
/// X = Expr BinOp X; V = X,
/// V = X; X = UpdateOp(X),
/// X = UpdateOp(X); V = X,
///
/// \param Loc The insert and source location description.
/// \param AllocaIP The insertion point to be used for alloca instructions.
/// \param X The target atomic pointer to be updated
/// \param V Memory address where to store captured value
/// \param Expr The value to update X with.
/// \param AO Atomic ordering of the generated atomic instructions
/// \param RMWOp The binary operation used for update. If
/// operation is not supported by atomicRMW, or belong to
/// {FADD, FSUB, BAD_BINOP}. Then a cmpExch based
/// atomic will be generated.
/// \param UpdateOp Code generator for complex expressions that cannot be
/// expressed through atomicrmw instruction.
/// \param UpdateExpr true if X is an in place update of the form
/// X = X BinOp Expr or X = Expr BinOp X
/// \param IsXBinopExpr true if X is Left H.S. in Right H.S. part of the
/// update expression, false otherwise.
/// (e.g. true for X = X BinOp Expr)
/// \param IsPostfixUpdate true if original value of 'x' must be stored in
/// 'v', not an updated one.
///
/// \return Insertion point after generated atomic capture IR.
InsertPointTy
createAtomicCapture(const LocationDescription &Loc, InsertPointTy AllocaIP,
AtomicOpValue &X, AtomicOpValue &V, Value *Expr,
AtomicOrdering AO, AtomicRMWInst::BinOp RMWOp,
AtomicUpdateCallbackTy &UpdateOp, bool UpdateExpr,
bool IsPostfixUpdate, bool IsXBinopExpr);
/// Emit atomic compare for constructs: --- Only scalar data types
/// cond-expr-stmt:
/// x = x ordop expr ? expr : x;
/// x = expr ordop x ? expr : x;
/// x = x == e ? d : x;
/// x = e == x ? d : x; (this one is not in the spec)
/// cond-update-stmt:
/// if (x ordop expr) { x = expr; }
/// if (expr ordop x) { x = expr; }
/// if (x == e) { x = d; }
/// if (e == x) { x = d; } (this one is not in the spec)
/// conditional-update-capture-atomic:
/// v = x; cond-update-stmt; (IsPostfixUpdate=true, IsFailOnly=false)
/// cond-update-stmt; v = x; (IsPostfixUpdate=false, IsFailOnly=false)
/// if (x == e) { x = d; } else { v = x; } (IsPostfixUpdate=false,
/// IsFailOnly=true)
/// r = x == e; if (r) { x = d; } (IsPostfixUpdate=false, IsFailOnly=false)
/// r = x == e; if (r) { x = d; } else { v = x; } (IsPostfixUpdate=false,
/// IsFailOnly=true)
///
/// \param Loc The insert and source location description.
/// \param X The target atomic pointer to be updated.
/// \param V Memory address where to store captured value (for
/// compare capture only).
/// \param R Memory address where to store comparison result
/// (for compare capture with '==' only).
/// \param E The expected value ('e') for forms that use an
/// equality comparison or an expression ('expr') for
/// forms that use 'ordop' (logically an atomic maximum or
/// minimum).
/// \param D The desired value for forms that use an equality
/// comparison. If forms that use 'ordop', it should be
/// \p nullptr.
/// \param AO Atomic ordering of the generated atomic instructions.
/// \param Op Atomic compare operation. It can only be ==, <, or >.
/// \param IsXBinopExpr True if the conditional statement is in the form where
/// x is on LHS. It only matters for < or >.
/// \param IsPostfixUpdate True if original value of 'x' must be stored in
/// 'v', not an updated one (for compare capture
/// only).
/// \param IsFailOnly True if the original value of 'x' is stored to 'v'
/// only when the comparison fails. This is only valid for
/// the case the comparison is '=='.
///
/// \return Insertion point after generated atomic capture IR.
InsertPointTy
createAtomicCompare(const LocationDescription &Loc, AtomicOpValue &X,
AtomicOpValue &V, AtomicOpValue &R, Value *E, Value *D,
AtomicOrdering AO, omp::OMPAtomicCompareOp Op,
bool IsXBinopExpr, bool IsPostfixUpdate, bool IsFailOnly);
InsertPointTy createAtomicCompare(const LocationDescription &Loc,
AtomicOpValue &X, AtomicOpValue &V,
AtomicOpValue &R, Value *E, Value *D,
AtomicOrdering AO,
omp::OMPAtomicCompareOp Op,
bool IsXBinopExpr, bool IsPostfixUpdate,
bool IsFailOnly, AtomicOrdering Failure);
/// Create the control flow structure of a canonical OpenMP loop.
///
/// The emitted loop will be disconnected, i.e. no edge to the loop's
/// preheader and no terminator in the AfterBB. The OpenMPIRBuilder's
/// IRBuilder location is not preserved.
///
/// \param DL DebugLoc used for the instructions in the skeleton.
/// \param TripCount Value to be used for the trip count.
/// \param F Function in which to insert the BasicBlocks.
/// \param PreInsertBefore Where to insert BBs that execute before the body,
/// typically the body itself.
/// \param PostInsertBefore Where to insert BBs that execute after the body.
/// \param Name Base name used to derive BB
/// and instruction names.
///
/// \returns The CanonicalLoopInfo that represents the emitted loop.
CanonicalLoopInfo *createLoopSkeleton(DebugLoc DL, Value *TripCount,
Function *F,
BasicBlock *PreInsertBefore,
BasicBlock *PostInsertBefore,
const Twine &Name = {});
/// OMP Offload Info Metadata name string
const std::string ompOffloadInfoName = "omp_offload.info";
/// Loads all the offload entries information from the host IR
/// metadata. This function is only meant to be used with device code
/// generation.
///
/// \param M Module to load Metadata info from. Module passed maybe
/// loaded from bitcode file, i.e, different from OpenMPIRBuilder::M module.
void loadOffloadInfoMetadata(Module &M);
/// Loads all the offload entries information from the host IR
/// metadata read from the file passed in as the HostFilePath argument. This
/// function is only meant to be used with device code generation.
///
/// \param HostFilePath The path to the host IR file,
/// used to load in offload metadata for the device, allowing host and device
/// to maintain the same metadata mapping.
void loadOffloadInfoMetadata(StringRef HostFilePath);
/// Gets (if variable with the given name already exist) or creates
/// internal global variable with the specified Name. The created variable has
/// linkage CommonLinkage by default and is initialized by null value.
/// \param Ty Type of the global variable. If it is exist already the type
/// must be the same.
/// \param Name Name of the variable.
GlobalVariable *getOrCreateInternalVariable(Type *Ty, const StringRef &Name,
unsigned AddressSpace = 0);
};
/// Class to represented the control flow structure of an OpenMP canonical loop.
///
/// The control-flow structure is standardized for easy consumption by
/// directives associated with loops. For instance, the worksharing-loop
/// construct may change this control flow such that each loop iteration is
/// executed on only one thread. The constraints of a canonical loop in brief
/// are:
///
/// * The number of loop iterations must have been computed before entering the
/// loop.
///
/// * Has an (unsigned) logical induction variable that starts at zero and
/// increments by one.
///
/// * The loop's CFG itself has no side-effects. The OpenMP specification
/// itself allows side-effects, but the order in which they happen, including
/// how often or whether at all, is unspecified. We expect that the frontend
/// will emit those side-effect instructions somewhere (e.g. before the loop)
/// such that the CanonicalLoopInfo itself can be side-effect free.
///
/// Keep in mind that CanonicalLoopInfo is meant to only describe a repeated
/// execution of a loop body that satifies these constraints. It does NOT
/// represent arbitrary SESE regions that happen to contain a loop. Do not use
/// CanonicalLoopInfo for such purposes.
///
/// The control flow can be described as follows:
///
/// Preheader
/// |
/// /-> Header
/// | |
/// | Cond---\
/// | | |
/// | Body |
/// | | | |
/// | <...> |
/// | | | |
/// \--Latch |
/// |
/// Exit
/// |
/// After
///
/// The loop is thought to start at PreheaderIP (at the Preheader's terminator,
/// including) and end at AfterIP (at the After's first instruction, excluding).
/// That is, instructions in the Preheader and After blocks (except the
/// Preheader's terminator) are out of CanonicalLoopInfo's control and may have
/// side-effects. Typically, the Preheader is used to compute the loop's trip
/// count. The instructions from BodyIP (at the Body block's first instruction,
/// excluding) until the Latch are also considered outside CanonicalLoopInfo's
/// control and thus can have side-effects. The body block is the single entry
/// point into the loop body, which may contain arbitrary control flow as long
/// as all control paths eventually branch to the Latch block.
///
/// TODO: Consider adding another standardized BasicBlock between Body CFG and
/// Latch to guarantee that there is only a single edge to the latch. It would
/// make loop transformations easier to not needing to consider multiple
/// predecessors of the latch (See redirectAllPredecessorsTo) and would give us
/// an equivalant to PreheaderIP, AfterIP and BodyIP for inserting code that
/// executes after each body iteration.
///
/// There must be no loop-carried dependencies through llvm::Values. This is
/// equivalant to that the Latch has no PHINode and the Header's only PHINode is
/// for the induction variable.
///
/// All code in Header, Cond, Latch and Exit (plus the terminator of the
/// Preheader) are CanonicalLoopInfo's responsibility and their build-up checked
/// by assertOK(). They are expected to not be modified unless explicitly
/// modifying the CanonicalLoopInfo through a methods that applies a OpenMP
/// loop-associated construct such as applyWorkshareLoop, tileLoops, unrollLoop,
/// etc. These methods usually invalidate the CanonicalLoopInfo and re-use its
/// basic blocks. After invalidation, the CanonicalLoopInfo must not be used
/// anymore as its underlying control flow may not exist anymore.
/// Loop-transformation methods such as tileLoops, collapseLoops and unrollLoop
/// may also return a new CanonicalLoopInfo that can be passed to other
/// loop-associated construct implementing methods. These loop-transforming
/// methods may either create a new CanonicalLoopInfo usually using
/// createLoopSkeleton and invalidate the input CanonicalLoopInfo, or reuse and
/// modify one of the input CanonicalLoopInfo and return it as representing the
/// modified loop. What is done is an implementation detail of
/// transformation-implementing method and callers should always assume that the
/// CanonicalLoopInfo passed to it is invalidated and a new object is returned.
/// Returned CanonicalLoopInfo have the same structure and guarantees as the one
/// created by createCanonicalLoop, such that transforming methods do not have
/// to special case where the CanonicalLoopInfo originated from.
///
/// Generally, methods consuming CanonicalLoopInfo do not need an
/// OpenMPIRBuilder::InsertPointTy as argument, but use the locations of the
/// CanonicalLoopInfo to insert new or modify existing instructions. Unless
/// documented otherwise, methods consuming CanonicalLoopInfo do not invalidate
/// any InsertPoint that is outside CanonicalLoopInfo's control. Specifically,
/// any InsertPoint in the Preheader, After or Block can still be used after
/// calling such a method.
///
/// TODO: Provide mechanisms for exception handling and cancellation points.
///
/// Defined outside OpenMPIRBuilder because nested classes cannot be
/// forward-declared, e.g. to avoid having to include the entire OMPIRBuilder.h.
class CanonicalLoopInfo {
friend class OpenMPIRBuilder;
private:
BasicBlock *Header = nullptr;
BasicBlock *Cond = nullptr;
BasicBlock *Latch = nullptr;
BasicBlock *Exit = nullptr;
/// Add the control blocks of this loop to \p BBs.
///
/// This does not include any block from the body, including the one returned
/// by getBody().
///
/// FIXME: This currently includes the Preheader and After blocks even though
/// their content is (mostly) not under CanonicalLoopInfo's control.
/// Re-evaluated whether this makes sense.
void collectControlBlocks(SmallVectorImpl<BasicBlock *> &BBs);
/// Sets the number of loop iterations to the given value. This value must be
/// valid in the condition block (i.e., defined in the preheader) and is
/// interpreted as an unsigned integer.
void setTripCount(Value *TripCount);
/// Replace all uses of the canonical induction variable in the loop body with
/// a new one.
///
/// The intended use case is to update the induction variable for an updated
/// iteration space such that it can stay normalized in the 0...tripcount-1
/// range.
///
/// The \p Updater is called with the (presumable updated) current normalized
/// induction variable and is expected to return the value that uses of the
/// pre-updated induction values should use instead, typically dependent on
/// the new induction variable. This is a lambda (instead of e.g. just passing
/// the new value) to be able to distinguish the uses of the pre-updated
/// induction variable and uses of the induction varible to compute the
/// updated induction variable value.
void mapIndVar(llvm::function_ref<Value *(Instruction *)> Updater);
public:
/// Returns whether this object currently represents the IR of a loop. If
/// returning false, it may have been consumed by a loop transformation or not
/// been intialized. Do not use in this case;
bool isValid() const { return Header; }
/// The preheader ensures that there is only a single edge entering the loop.
/// Code that must be execute before any loop iteration can be emitted here,
/// such as computing the loop trip count and begin lifetime markers. Code in
/// the preheader is not considered part of the canonical loop.
BasicBlock *getPreheader() const;
/// The header is the entry for each iteration. In the canonical control flow,
/// it only contains the PHINode for the induction variable.
BasicBlock *getHeader() const {
assert(isValid() && "Requires a valid canonical loop");
return Header;
}
/// The condition block computes whether there is another loop iteration. If
/// yes, branches to the body; otherwise to the exit block.
BasicBlock *getCond() const {
assert(isValid() && "Requires a valid canonical loop");
return Cond;
}
/// The body block is the single entry for a loop iteration and not controlled
/// by CanonicalLoopInfo. It can contain arbitrary control flow but must
/// eventually branch to the \p Latch block.
BasicBlock *getBody() const {
assert(isValid() && "Requires a valid canonical loop");
return cast<BranchInst>(Cond->getTerminator())->getSuccessor(0);
}
/// Reaching the latch indicates the end of the loop body code. In the
/// canonical control flow, it only contains the increment of the induction
/// variable.
BasicBlock *getLatch() const {
assert(isValid() && "Requires a valid canonical loop");
return Latch;
}
/// Reaching the exit indicates no more iterations are being executed.
BasicBlock *getExit() const {
assert(isValid() && "Requires a valid canonical loop");
return Exit;
}
/// The after block is intended for clean-up code such as lifetime end
/// markers. It is separate from the exit block to ensure, analogous to the
/// preheader, it having just a single entry edge and being free from PHI
/// nodes should there be multiple loop exits (such as from break
/// statements/cancellations).
BasicBlock *getAfter() const {
assert(isValid() && "Requires a valid canonical loop");
return Exit->getSingleSuccessor();
}
/// Returns the llvm::Value containing the number of loop iterations. It must
/// be valid in the preheader and always interpreted as an unsigned integer of
/// any bit-width.
Value *getTripCount() const {
assert(isValid() && "Requires a valid canonical loop");
Instruction *CmpI = &Cond->front();
assert(isa<CmpInst>(CmpI) && "First inst must compare IV with TripCount");
return CmpI->getOperand(1);
}
/// Returns the instruction representing the current logical induction
/// variable. Always unsigned, always starting at 0 with an increment of one.
Instruction *getIndVar() const {
assert(isValid() && "Requires a valid canonical loop");
Instruction *IndVarPHI = &Header->front();
assert(isa<PHINode>(IndVarPHI) && "First inst must be the IV PHI");
return IndVarPHI;
}
/// Return the type of the induction variable (and the trip count).
Type *getIndVarType() const {
assert(isValid() && "Requires a valid canonical loop");
return getIndVar()->getType();
}
/// Return the insertion point for user code before the loop.
OpenMPIRBuilder::InsertPointTy getPreheaderIP() const {
assert(isValid() && "Requires a valid canonical loop");
BasicBlock *Preheader = getPreheader();
return {Preheader, std::prev(Preheader->end())};
};
/// Return the insertion point for user code in the body.
OpenMPIRBuilder::InsertPointTy getBodyIP() const {
assert(isValid() && "Requires a valid canonical loop");
BasicBlock *Body = getBody();
return {Body, Body->begin()};
};
/// Return the insertion point for user code after the loop.
OpenMPIRBuilder::InsertPointTy getAfterIP() const {
assert(isValid() && "Requires a valid canonical loop");
BasicBlock *After = getAfter();
return {After, After->begin()};
};
Function *getFunction() const {
assert(isValid() && "Requires a valid canonical loop");
return Header->getParent();
}
/// Consistency self-check.
void assertOK() const;
/// Invalidate this loop. That is, the underlying IR does not fulfill the
/// requirements of an OpenMP canonical loop anymore.
void invalidate();
};
} // end namespace llvm
#endif // LLVM_FRONTEND_OPENMP_OMPIRBUILDER_H
|