File: PreciseAllocation.h

package info (click to toggle)
webkit2gtk 2.44.2-1~deb11u1
  • links: PTS, VCS
  • area: main
  • in suites: bullseye
  • size: 367,728 kB
  • sloc: cpp: 2,946,520; javascript: 194,328; ansic: 137,901; python: 45,716; ruby: 18,487; asm: 16,672; perl: 16,476; xml: 4,376; yacc: 2,350; sh: 2,127; java: 1,711; lex: 1,323; pascal: 333; makefile: 323
file content (191 lines) | stat: -rw-r--r-- 7,206 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
/*
 * Copyright (C) 2016-2023 Apple Inc. All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY APPLE INC. ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE INC. OR
 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
 * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 
 */

#pragma once

#include "MarkedBlock.h"
#include "WeakSet.h"
#include <wtf/StdLibExtras.h>

namespace JSC {

class IsoSubspace;
class SlotVisitor;

// WebKit has a good malloc that already knows what to do for large allocations. The GC shouldn't
// have to think about such things. That's where PreciseAllocation comes in. We will allocate large
// objects directly using malloc, and put the PreciseAllocation header just before them. We can detect
// when a HeapCell* is a PreciseAllocation because it will have the MarkedBlock::atomSize / 2 bit set.

class PreciseAllocation : public BasicRawSentinelNode<PreciseAllocation> {
public:
    friend class LLIntOffsetsExtractor;
    friend class IsoSubspace;

    static PreciseAllocation* tryCreate(Heap&, size_t, Subspace*, unsigned indexInSpace);

    static PreciseAllocation* tryCreateForLowerTier(Heap&, size_t, Subspace*, uint8_t lowerTierIndex);
    PreciseAllocation* reuseForLowerTier();

    PreciseAllocation* tryReallocate(size_t, Subspace*);
    
    ~PreciseAllocation();
    
    static PreciseAllocation* fromCell(const void* cell)
    {
        return bitwise_cast<PreciseAllocation*>(bitwise_cast<char*>(cell) - headerSize());
    }
    
    HeapCell* cell() const
    {
        return bitwise_cast<HeapCell*>(bitwise_cast<char*>(this) + headerSize());
    }
    
    static bool isPreciseAllocation(HeapCell* cell)
    {
        return bitwise_cast<uintptr_t>(cell) & halfAlignment;
    }
    
    Subspace* subspace() const { return m_subspace; }
    
    void lastChanceToFinalize();
    
    JSC::Heap* heap() const { return m_weakSet.heap(); }
    VM& vm() const { return m_weakSet.vm(); }
    WeakSet& weakSet() { return m_weakSet; }

    static ptrdiff_t offsetOfWeakSet() { return OBJECT_OFFSETOF(PreciseAllocation, m_weakSet); }

    unsigned indexInSpace() { return m_indexInSpace; }
    void setIndexInSpace(unsigned indexInSpace) { m_indexInSpace = indexInSpace; }
    
    void clearNewlyAllocated() { m_isNewlyAllocated = false; }
    void flip();
    
    bool isNewlyAllocated() const { return m_isNewlyAllocated; }
    ALWAYS_INLINE bool isMarked() { return m_isMarked.load(std::memory_order_relaxed); }
    ALWAYS_INLINE bool isMarked(HeapCell*) { return isMarked(); }
    ALWAYS_INLINE bool isMarked(HeapCell*, Dependency) { return isMarked(); }
    ALWAYS_INLINE bool isMarked(HeapVersion, HeapCell*) { return isMarked(); }
    bool isLive() { return isMarked() || isNewlyAllocated(); }
    
    bool hasValidCell() const { return m_hasValidCell; }
    
    bool isEmpty();
    
    size_t cellSize() const { return m_cellSize; }

    uint8_t lowerTierIndex() const { return m_lowerTierIndex; }
    
    bool aboveLowerBound(const void* rawPtr)
    {
        char* ptr = bitwise_cast<char*>(rawPtr);
        char* begin = bitwise_cast<char*>(cell());
        return ptr >= begin;
    }
    
    bool belowUpperBound(const void* rawPtr)
    {
        char* ptr = bitwise_cast<char*>(rawPtr);
        char* begin = bitwise_cast<char*>(cell());
        char* end = begin + cellSize();
        // We cannot #include IndexingHeader.h because reasons. The fact that IndexingHeader is 8
        // bytes is wired deep into our engine, so this isn't so bad.
        size_t sizeOfIndexingHeader = 8;
        return ptr <= end + sizeOfIndexingHeader;
    }
    
    bool contains(const void* rawPtr)
    {
        return aboveLowerBound(rawPtr) && belowUpperBound(rawPtr);
    }

    CellAttributes attributes() const { return m_attributes; }

    Dependency aboutToMark(HeapVersion) { return Dependency(); }
    
    ALWAYS_INLINE bool testAndSetMarked()
    {
        // This method is usually called when the object is already marked. This avoids us
        // having to CAS in that case. It's profitable to reduce the total amount of CAS
        // traffic.
        if (isMarked())
            return true;
        return m_isMarked.compareExchangeStrong(false, true);
    }
    ALWAYS_INLINE bool testAndSetMarked(HeapCell*, Dependency) { return testAndSetMarked(); }
    void clearMarked() { m_isMarked.store(false); }
    
    void noteMarked() { }
    
#if ASSERT_ENABLED
    void assertValidCell(VM&, HeapCell*) const;
#else
    void assertValidCell(VM&, HeapCell*) const { }
#endif
    
    void sweep();
    
    void destroy();
    
    void dump(PrintStream&) const;

    bool isLowerTier() const { return m_lowerTierIndex != UINT8_MAX; }
    
    static constexpr unsigned alignment = MarkedBlock::atomSize;
    static constexpr unsigned halfAlignment = alignment / 2;
    static constexpr unsigned cacheLineAdjustment = 2 * halfAlignment;

    // The header size must be packed to full alignment size. Because the cell start address
    // always begins immediately after the header, this allows isAlignedForPreciseAllocation()
    // to trivially infer the alignment of the cell from the alignment of the header.
    static constexpr unsigned headerSize() { return roundUpToMultipleOf<alignment>(sizeof(PreciseAllocation)); }

private:
    PreciseAllocation(Heap&, size_t, Subspace*, unsigned indexInSpace, unsigned adjustment);
    
    void* basePointer() const;
    
    unsigned m_indexInSpace { 0 };
    size_t m_cellSize;
    bool m_isNewlyAllocated : 1;
    bool m_hasValidCell : 1;
    // Worst case adjustment needed would be halfAlignment + portionOfObjectThatMustFitInCacheLine
    // which is 8 + 16 -> 24 bytes i.e. will fit in 5 bits. If we need more bits in the future, we
    // can also encode this number of uintptr_t words to save 3 bits.
    unsigned m_adjustment : 5;
    Atomic<bool> m_isMarked;
    CellAttributes m_attributes;
    uint8_t m_lowerTierIndex { UINT8_MAX };
    Subspace* m_subspace;
    WeakSet m_weakSet;
};

inline void* PreciseAllocation::basePointer() const
{
    return bitwise_cast<char*>(this) - m_adjustment;
}

} // namespace JSC