llvm-project/clang/lib/CodeGen/CGRecordLayout.h
Ties Stuij 208987844f [ARM] Follow AACPS standard for volatile bit-fields access width
This patch resumes the work of D16586.
According to the AAPCS, volatile bit-fields should
be accessed using containers of the widht of their
declarative type. In such case:
```
struct S1 {
  short a : 1;
}
```
should be accessed using load and stores of the width
(sizeof(short)), where now the compiler does only load
the minimum required width (char in this case).
However, as discussed in D16586,
that could overwrite non-volatile bit-fields, which
conflicted with C and C++ object models by creating
data race conditions that are not part of the bit-field,
e.g.
```
struct S2 {
  short a;
  int  b : 16;
}
```
Accessing `S2.b` would also access `S2.a`.

The AAPCS Release 2020Q2
(https://documentation-service.arm.com/static/5efb7fbedbdee951c1ccf186?token=)
section 8.1 Data Types, page 36, "Volatile bit-fields -
preserving number and width of container accesses" has been
updated to avoid conflict with the C++ Memory Model.
Now it reads in the note:
```
This ABI does not place any restrictions on the access widths of bit-fields where the container
overlaps with a non-bit-field member or where the container overlaps with any zero length bit-field
placed between two other bit-fields. This is because the C/C++ memory model defines these as being
separate memory locations, which can be accessed by two threads simultaneously. For this reason,
compilers must be permitted to use a narrower memory access width (including splitting the access into
multiple instructions) to avoid writing to a different memory location. For example, in
struct S { int a:24; char b; }; a write to a must not also write to the location occupied by b, this requires at least two
memory accesses in all current Arm architectures. In the same way, in struct S { int a:24; int:0; int b:8; };,
writes to a or b must not overwrite each other.
```

I've updated the patch D16586 to follow such behavior by verifying that we
only change volatile bit-field access when:
 - it won't overlap with any other non-bit-field member
 - we only access memory inside the bounds of the record
 - avoid overlapping zero-length bit-fields.

Regarding the number of memory accesses, that should be preserved, that will
be implemented by D67399.

Reviewed By: ostannard

Differential Revision: https://reviews.llvm.org/D72932
2020-10-13 10:31:48 +01:00

233 lines
8.4 KiB
C++

//===--- CGRecordLayout.h - LLVM Record Layout Information ------*- 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
//
//===----------------------------------------------------------------------===//
#ifndef LLVM_CLANG_LIB_CODEGEN_CGRECORDLAYOUT_H
#define LLVM_CLANG_LIB_CODEGEN_CGRECORDLAYOUT_H
#include "clang/AST/CharUnits.h"
#include "clang/AST/DeclCXX.h"
#include "clang/Basic/LLVM.h"
#include "llvm/ADT/DenseMap.h"
#include "llvm/IR/DerivedTypes.h"
namespace llvm {
class StructType;
}
namespace clang {
namespace CodeGen {
/// Structure with information about how a bitfield should be accessed.
///
/// Often we layout a sequence of bitfields as a contiguous sequence of bits.
/// When the AST record layout does this, we represent it in the LLVM IR's type
/// as either a sequence of i8 members or a byte array to reserve the number of
/// bytes touched without forcing any particular alignment beyond the basic
/// character alignment.
///
/// Then accessing a particular bitfield involves converting this byte array
/// into a single integer of that size (i24 or i40 -- may not be power-of-two
/// size), loading it, and shifting and masking to extract the particular
/// subsequence of bits which make up that particular bitfield. This structure
/// encodes the information used to construct the extraction code sequences.
/// The CGRecordLayout also has a field index which encodes which byte-sequence
/// this bitfield falls within. Let's assume the following C struct:
///
/// struct S {
/// char a, b, c;
/// unsigned bits : 3;
/// unsigned more_bits : 4;
/// unsigned still_more_bits : 7;
/// };
///
/// This will end up as the following LLVM type. The first array is the
/// bitfield, and the second is the padding out to a 4-byte alignment.
///
/// %t = type { i8, i8, i8, i8, i8, [3 x i8] }
///
/// When generating code to access more_bits, we'll generate something
/// essentially like this:
///
/// define i32 @foo(%t* %base) {
/// %0 = gep %t* %base, i32 0, i32 3
/// %2 = load i8* %1
/// %3 = lshr i8 %2, 3
/// %4 = and i8 %3, 15
/// %5 = zext i8 %4 to i32
/// ret i32 %i
/// }
///
struct CGBitFieldInfo {
/// The offset within a contiguous run of bitfields that are represented as
/// a single "field" within the LLVM struct type. This offset is in bits.
unsigned Offset : 16;
/// The total size of the bit-field, in bits.
unsigned Size : 15;
/// Whether the bit-field is signed.
unsigned IsSigned : 1;
/// The storage size in bits which should be used when accessing this
/// bitfield.
unsigned StorageSize;
/// The offset of the bitfield storage from the start of the struct.
CharUnits StorageOffset;
/// The offset within a contiguous run of bitfields that are represented as a
/// single "field" within the LLVM struct type, taking into account the AAPCS
/// rules for volatile bitfields. This offset is in bits.
unsigned VolatileOffset : 16;
/// The storage size in bits which should be used when accessing this
/// bitfield.
unsigned VolatileStorageSize;
/// The offset of the bitfield storage from the start of the struct.
CharUnits VolatileStorageOffset;
CGBitFieldInfo()
: Offset(), Size(), IsSigned(), StorageSize(), StorageOffset(),
VolatileOffset(), VolatileStorageSize(), VolatileStorageOffset() {}
CGBitFieldInfo(unsigned Offset, unsigned Size, bool IsSigned,
unsigned StorageSize, CharUnits StorageOffset)
: Offset(Offset), Size(Size), IsSigned(IsSigned),
StorageSize(StorageSize), StorageOffset(StorageOffset) {}
void print(raw_ostream &OS) const;
void dump() const;
/// Given a bit-field decl, build an appropriate helper object for
/// accessing that field (which is expected to have the given offset and
/// size).
static CGBitFieldInfo MakeInfo(class CodeGenTypes &Types,
const FieldDecl *FD,
uint64_t Offset, uint64_t Size,
uint64_t StorageSize,
CharUnits StorageOffset);
};
/// CGRecordLayout - This class handles struct and union layout info while
/// lowering AST types to LLVM types.
///
/// These layout objects are only created on demand as IR generation requires.
class CGRecordLayout {
friend class CodeGenTypes;
CGRecordLayout(const CGRecordLayout &) = delete;
void operator=(const CGRecordLayout &) = delete;
private:
/// The LLVM type corresponding to this record layout; used when
/// laying it out as a complete object.
llvm::StructType *CompleteObjectType;
/// The LLVM type for the non-virtual part of this record layout;
/// used when laying it out as a base subobject.
llvm::StructType *BaseSubobjectType;
/// Map from (non-bit-field) struct field to the corresponding llvm struct
/// type field no. This info is populated by record builder.
llvm::DenseMap<const FieldDecl *, unsigned> FieldInfo;
/// Map from (bit-field) struct field to the corresponding llvm struct type
/// field no. This info is populated by record builder.
llvm::DenseMap<const FieldDecl *, CGBitFieldInfo> BitFields;
// FIXME: Maybe we could use a CXXBaseSpecifier as the key and use a single
// map for both virtual and non-virtual bases.
llvm::DenseMap<const CXXRecordDecl *, unsigned> NonVirtualBases;
/// Map from virtual bases to their field index in the complete object.
llvm::DenseMap<const CXXRecordDecl *, unsigned> CompleteObjectVirtualBases;
/// False if any direct or indirect subobject of this class, when
/// considered as a complete object, requires a non-zero bitpattern
/// when zero-initialized.
bool IsZeroInitializable : 1;
/// False if any direct or indirect subobject of this class, when
/// considered as a base subobject, requires a non-zero bitpattern
/// when zero-initialized.
bool IsZeroInitializableAsBase : 1;
public:
CGRecordLayout(llvm::StructType *CompleteObjectType,
llvm::StructType *BaseSubobjectType,
bool IsZeroInitializable,
bool IsZeroInitializableAsBase)
: CompleteObjectType(CompleteObjectType),
BaseSubobjectType(BaseSubobjectType),
IsZeroInitializable(IsZeroInitializable),
IsZeroInitializableAsBase(IsZeroInitializableAsBase) {}
/// Return the "complete object" LLVM type associated with
/// this record.
llvm::StructType *getLLVMType() const {
return CompleteObjectType;
}
/// Return the "base subobject" LLVM type associated with
/// this record.
llvm::StructType *getBaseSubobjectLLVMType() const {
return BaseSubobjectType;
}
/// Check whether this struct can be C++ zero-initialized
/// with a zeroinitializer.
bool isZeroInitializable() const {
return IsZeroInitializable;
}
/// Check whether this struct can be C++ zero-initialized
/// with a zeroinitializer when considered as a base subobject.
bool isZeroInitializableAsBase() const {
return IsZeroInitializableAsBase;
}
/// Return llvm::StructType element number that corresponds to the
/// field FD.
unsigned getLLVMFieldNo(const FieldDecl *FD) const {
FD = FD->getCanonicalDecl();
assert(FieldInfo.count(FD) && "Invalid field for record!");
return FieldInfo.lookup(FD);
}
unsigned getNonVirtualBaseLLVMFieldNo(const CXXRecordDecl *RD) const {
assert(NonVirtualBases.count(RD) && "Invalid non-virtual base!");
return NonVirtualBases.lookup(RD);
}
/// Return the LLVM field index corresponding to the given
/// virtual base. Only valid when operating on the complete object.
unsigned getVirtualBaseIndex(const CXXRecordDecl *base) const {
assert(CompleteObjectVirtualBases.count(base) && "Invalid virtual base!");
return CompleteObjectVirtualBases.lookup(base);
}
/// Return the BitFieldInfo that corresponds to the field FD.
const CGBitFieldInfo &getBitFieldInfo(const FieldDecl *FD) const {
FD = FD->getCanonicalDecl();
assert(FD->isBitField() && "Invalid call for non-bit-field decl!");
llvm::DenseMap<const FieldDecl *, CGBitFieldInfo>::const_iterator
it = BitFields.find(FD);
assert(it != BitFields.end() && "Unable to find bitfield info");
return it->second;
}
void print(raw_ostream &OS) const;
void dump() const;
};
} // end namespace CodeGen
} // end namespace clang
#endif