1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-22 12:33:33 +02:00
llvm-mirror/include/llvm/IR/GlobalObject.h
Oliver Stannard 6aaf81e821 Reland: Dead Virtual Function Elimination
Remove dead virtual functions from vtables with
replaceNonMetadataUsesWith, so that CGProfile metadata gets cleaned up
correctly.

Original commit message:

Currently, it is hard for the compiler to remove unused C++ virtual
functions, because they are all referenced from vtables, which are referenced
by constructors. This means that if the constructor is called from any live
code, then we keep every virtual function in the final link, even if there
are no call sites which can use it.

This patch allows unused virtual functions to be removed during LTO (and
regular compilation in limited circumstances) by using type metadata to match
virtual function call sites to the vtable slots they might load from. This
information can then be used in the global dead code elimination pass instead
of the references from vtables to virtual functions, to more accurately
determine which functions are reachable.

To make this transformation safe, I have changed clang's code-generation to
always load virtual function pointers using the llvm.type.checked.load
intrinsic, instead of regular load instructions. I originally tried writing
this using clang's existing code-generation, which uses the llvm.type.test
and llvm.assume intrinsics after doing a normal load. However, it is possible
for optimisations to obscure the relationship between the GEP, load and
llvm.type.test, causing GlobalDCE to fail to find virtual function call
sites.

The existing linkage and visibility types don't accurately describe the scope
in which a virtual call could be made which uses a given vtable. This is
wider than the visibility of the type itself, because a virtual function call
could be made using a more-visible base class. I've added a new
!vcall_visibility metadata type to represent this, described in
TypeMetadata.rst. The internalization pass and libLTO have been updated to
change this metadata when linking is performed.

This doesn't currently work with ThinLTO, because it needs to see every call
to llvm.type.checked.load in the linkage unit. It might be possible to
extend this optimisation to be able to use the ThinLTO summary, as was done
for devirtualization, but until then that combination is rejected in the
clang driver.

To test this, I've written a fuzzer which generates random C++ programs with
complex class inheritance graphs, and virtual functions called through object
and function pointers of different types. The programs are spread across
multiple translation units and DSOs to test the different visibility
restrictions.

I've also tried doing bootstrap builds of LLVM to test this. This isn't
ideal, because only classes in anonymous namespaces can be optimised with
-fvisibility=default, and some parts of LLVM (plugins and bugpoint) do not
work correctly with -fvisibility=hidden. However, there are only 12 test
failures when building with -fvisibility=hidden (and an unmodified compiler),
and this change does not cause any new failures for either value of
-fvisibility.

On the 7 C++ sub-benchmarks of SPEC2006, this gives a geomean code-size
reduction of ~6%, over a baseline compiled with "-O2 -flto
-fvisibility=hidden -fwhole-program-vtables". The best cases are reductions
of ~14% in 450.soplex and 483.xalancbmk, and there are no code size
increases.

I've also run this on a set of 8 mbed-os examples compiled for Armv7M, which
show a geomean size reduction of ~3%, again with no size increases.

I had hoped that this would have no effect on performance, which would allow
it to awlays be enabled (when using -fwhole-program-vtables). However, the
changes in clang to use the llvm.type.checked.load intrinsic are causing ~1%
performance regression in the C++ parts of SPEC2006. It should be possible to
recover some of this perf loss by teaching optimisations about the
llvm.type.checked.load intrinsic, which would make it worth turning this on
by default (though it's still dependent on -fwhole-program-vtables).

Differential revision: https://reviews.llvm.org/D63932

llvm-svn: 375094
2019-10-17 09:58:57 +00:00

216 lines
7.1 KiB
C++

//===-- llvm/GlobalObject.h - Class to represent global objects -*- 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 represents an independent object. That is, a function or a global
// variable, but not an alias.
//
//===----------------------------------------------------------------------===//
#ifndef LLVM_IR_GLOBALOBJECT_H
#define LLVM_IR_GLOBALOBJECT_H
#include "llvm/ADT/StringRef.h"
#include "llvm/IR/GlobalValue.h"
#include "llvm/IR/Value.h"
#include "llvm/Support/Alignment.h"
#include <string>
#include <utility>
namespace llvm {
class Comdat;
class MDNode;
class Metadata;
class GlobalObject : public GlobalValue {
public:
// VCallVisibility - values for visibility metadata attached to vtables. This
// describes the scope in which a virtual call could end up being dispatched
// through this vtable.
enum VCallVisibility {
// Type is potentially visible to external code.
VCallVisibilityPublic = 0,
// Type is only visible to code which will be in the current Module after
// LTO internalization.
VCallVisibilityLinkageUnit = 1,
// Type is only visible to code in the current Module.
VCallVisibilityTranslationUnit = 2,
};
protected:
GlobalObject(Type *Ty, ValueTy VTy, Use *Ops, unsigned NumOps,
LinkageTypes Linkage, const Twine &Name,
unsigned AddressSpace = 0)
: GlobalValue(Ty, VTy, Ops, NumOps, Linkage, Name, AddressSpace),
ObjComdat(nullptr) {
setGlobalValueSubClassData(0);
}
Comdat *ObjComdat;
enum {
LastAlignmentBit = 4,
HasMetadataHashEntryBit,
HasSectionHashEntryBit,
GlobalObjectBits,
};
static const unsigned GlobalObjectSubClassDataBits =
GlobalValueSubClassDataBits - GlobalObjectBits;
private:
static const unsigned AlignmentBits = LastAlignmentBit + 1;
static const unsigned AlignmentMask = (1 << AlignmentBits) - 1;
static const unsigned GlobalObjectMask = (1 << GlobalObjectBits) - 1;
public:
GlobalObject(const GlobalObject &) = delete;
unsigned getAlignment() const {
unsigned Data = getGlobalValueSubClassData();
unsigned AlignmentData = Data & AlignmentMask;
MaybeAlign Align = decodeMaybeAlign(AlignmentData);
return Align ? Align->value() : 0;
}
/// FIXME: Remove this setter once the migration to MaybeAlign is over.
LLVM_ATTRIBUTE_DEPRECATED(void setAlignment(unsigned Align),
"Please use `void setAlignment(MaybeAlign Align)`");
void setAlignment(MaybeAlign Align);
unsigned getGlobalObjectSubClassData() const {
unsigned ValueData = getGlobalValueSubClassData();
return ValueData >> GlobalObjectBits;
}
void setGlobalObjectSubClassData(unsigned Val) {
unsigned OldData = getGlobalValueSubClassData();
setGlobalValueSubClassData((OldData & GlobalObjectMask) |
(Val << GlobalObjectBits));
assert(getGlobalObjectSubClassData() == Val && "representation error");
}
/// Check if this global has a custom object file section.
///
/// This is more efficient than calling getSection() and checking for an empty
/// string.
bool hasSection() const {
return getGlobalValueSubClassData() & (1 << HasSectionHashEntryBit);
}
/// Get the custom section of this global if it has one.
///
/// If this global does not have a custom section, this will be empty and the
/// default object file section (.text, .data, etc) will be used.
StringRef getSection() const {
return hasSection() ? getSectionImpl() : StringRef();
}
/// Change the section for this global.
///
/// Setting the section to the empty string tells LLVM to choose an
/// appropriate default object file section.
void setSection(StringRef S);
bool hasComdat() const { return getComdat() != nullptr; }
const Comdat *getComdat() const { return ObjComdat; }
Comdat *getComdat() { return ObjComdat; }
void setComdat(Comdat *C) { ObjComdat = C; }
/// Check if this has any metadata.
bool hasMetadata() const { return hasMetadataHashEntry(); }
/// Check if this has any metadata of the given kind.
bool hasMetadata(unsigned KindID) const {
return getMetadata(KindID) != nullptr;
}
bool hasMetadata(StringRef Kind) const {
return getMetadata(Kind) != nullptr;
}
/// Get the current metadata attachments for the given kind, if any.
///
/// These functions require that the function have at most a single attachment
/// of the given kind, and return \c nullptr if such an attachment is missing.
/// @{
MDNode *getMetadata(unsigned KindID) const;
MDNode *getMetadata(StringRef Kind) const;
/// @}
/// Appends all attachments with the given ID to \c MDs in insertion order.
/// If the global has no attachments with the given ID, or if ID is invalid,
/// leaves MDs unchanged.
/// @{
void getMetadata(unsigned KindID, SmallVectorImpl<MDNode *> &MDs) const;
void getMetadata(StringRef Kind, SmallVectorImpl<MDNode *> &MDs) const;
/// @}
/// Set a particular kind of metadata attachment.
///
/// Sets the given attachment to \c MD, erasing it if \c MD is \c nullptr or
/// replacing it if it already exists.
/// @{
void setMetadata(unsigned KindID, MDNode *MD);
void setMetadata(StringRef Kind, MDNode *MD);
/// @}
/// Add a metadata attachment.
/// @{
void addMetadata(unsigned KindID, MDNode &MD);
void addMetadata(StringRef Kind, MDNode &MD);
/// @}
/// Appends all attachments for the global to \c MDs, sorting by attachment
/// ID. Attachments with the same ID appear in insertion order.
void
getAllMetadata(SmallVectorImpl<std::pair<unsigned, MDNode *>> &MDs) const;
/// Erase all metadata attachments with the given kind.
///
/// \returns true if any metadata was removed.
bool eraseMetadata(unsigned KindID);
/// Copy metadata from Src, adjusting offsets by Offset.
void copyMetadata(const GlobalObject *Src, unsigned Offset);
void addTypeMetadata(unsigned Offset, Metadata *TypeID);
void addVCallVisibilityMetadata(VCallVisibility Visibility);
VCallVisibility getVCallVisibility() const;
protected:
void copyAttributesFrom(const GlobalObject *Src);
public:
// Methods for support type inquiry through isa, cast, and dyn_cast:
static bool classof(const Value *V) {
return V->getValueID() == Value::FunctionVal ||
V->getValueID() == Value::GlobalVariableVal;
}
void clearMetadata();
private:
void setGlobalObjectFlag(unsigned Bit, bool Val) {
unsigned Mask = 1 << Bit;
setGlobalValueSubClassData((~Mask & getGlobalValueSubClassData()) |
(Val ? Mask : 0u));
}
bool hasMetadataHashEntry() const {
return getGlobalValueSubClassData() & (1 << HasMetadataHashEntryBit);
}
void setHasMetadataHashEntry(bool HasEntry) {
setGlobalObjectFlag(HasMetadataHashEntryBit, HasEntry);
}
StringRef getSectionImpl() const;
};
} // end namespace llvm
#endif // LLVM_IR_GLOBALOBJECT_H