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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
|
; Licensed to the .NET Foundation under one or more agreements.
; The .NET Foundation licenses this file to you under the MIT license.
; See the LICENSE file in the project root for more information.
; ==++==
;
;
; ==--==
; ***********************************************************************
; File: JitHelpers_InlineGetThread.asm, see history in jithelp.asm
;
; Notes: These routinues will be patched at runtime with the location in
; the TLS to find the Thread* and are the fastest implementation
; of their specific functionality.
; ***********************************************************************
include AsmMacros.inc
include asmconstants.inc
; Min amount of stack space that a nested function should allocate.
MIN_SIZE equ 28h
; Macro to create a patchable inline GetAppdomain, if we decide to create patchable
; high TLS inline versions then just change this macro to make sure to create enough
; space in the asm to patch the high TLS getter instructions.
PATCHABLE_INLINE_GETTHREAD macro Reg, PatchLabel
PATCH_LABEL PatchLabel
mov Reg, gs:[OFFSET__TEB__TlsSlots]
endm
JIT_NEW equ ?JIT_New@@YAPEAVObject@@PEAUCORINFO_CLASS_STRUCT_@@@Z
Object__DEBUG_SetAppDomain equ ?DEBUG_SetAppDomain@Object@@QEAAXPEAVAppDomain@@@Z
CopyValueClassUnchecked equ ?CopyValueClassUnchecked@@YAXPEAX0PEAVMethodTable@@@Z
JIT_Box equ ?JIT_Box@@YAPEAVObject@@PEAUCORINFO_CLASS_STRUCT_@@PEAX@Z
g_pStringClass equ ?g_pStringClass@@3PEAVMethodTable@@EA
FramedAllocateString equ ?FramedAllocateString@@YAPEAVStringObject@@K@Z
JIT_NewArr1 equ ?JIT_NewArr1@@YAPEAVObject@@PEAUCORINFO_CLASS_STRUCT_@@_J@Z
INVALIDGCVALUE equ 0CCCCCCCDh
extern JIT_NEW:proc
extern CopyValueClassUnchecked:proc
extern JIT_Box:proc
extern g_pStringClass:QWORD
extern FramedAllocateString:proc
extern JIT_NewArr1:proc
extern JIT_InternalThrow:proc
ifdef _DEBUG
extern DEBUG_TrialAllocSetAppDomain:proc
extern DEBUG_TrialAllocSetAppDomain_NoScratchArea:proc
endif
; IN: rcx: MethodTable*
; OUT: rax: new object
LEAF_ENTRY JIT_TrialAllocSFastMP_InlineGetThread, _TEXT
mov edx, [rcx + OFFSET__MethodTable__m_BaseSize]
; m_BaseSize is guaranteed to be a multiple of 8.
PATCHABLE_INLINE_GETTHREAD r11, JIT_TrialAllocSFastMP_InlineGetThread__PatchTLSOffset
mov r10, [r11 + OFFSET__Thread__m_alloc_context__alloc_limit]
mov rax, [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr]
add rdx, rax
cmp rdx, r10
ja AllocFailed
mov [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr], rdx
mov [rax], rcx
ifdef _DEBUG
call DEBUG_TrialAllocSetAppDomain_NoScratchArea
endif ; _DEBUG
ret
AllocFailed:
jmp JIT_NEW
LEAF_END JIT_TrialAllocSFastMP_InlineGetThread, _TEXT
; HCIMPL2(Object*, JIT_Box, CORINFO_CLASS_HANDLE type, void* unboxedData)
NESTED_ENTRY JIT_BoxFastMP_InlineGetThread, _TEXT
mov rax, [rcx + OFFSETOF__MethodTable__m_pWriteableData]
; Check whether the class has not been initialized
test dword ptr [rax + OFFSETOF__MethodTableWriteableData__m_dwFlags], MethodTableWriteableData__enum_flag_Unrestored
jnz ClassNotInited
mov r8d, [rcx + OFFSET__MethodTable__m_BaseSize]
; m_BaseSize is guaranteed to be a multiple of 8.
PATCHABLE_INLINE_GETTHREAD r11, JIT_BoxFastMPIGT__PatchTLSLabel
mov r10, [r11 + OFFSET__Thread__m_alloc_context__alloc_limit]
mov rax, [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr]
add r8, rax
cmp r8, r10
ja AllocFailed
mov [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr], r8
mov [rax], rcx
ifdef _DEBUG
call DEBUG_TrialAllocSetAppDomain_NoScratchArea
endif ; _DEBUG
; Check whether the object contains pointers
test dword ptr [rcx + OFFSETOF__MethodTable__m_dwFlags], MethodTable__enum_flag_ContainsPointers
jnz ContainsPointers
; We have no pointers - emit a simple inline copy loop
; Copy the contents from the end
mov ecx, [rcx + OFFSET__MethodTable__m_BaseSize]
sub ecx, 18h ; sizeof(ObjHeader) + sizeof(Object) + last slot
align 16
CopyLoop:
mov r8, [rdx+rcx]
mov [rax+rcx+8], r8
sub ecx, 8
jge CopyLoop
REPRET
ContainsPointers:
; Do call to CopyValueClassUnchecked(object, data, pMT)
push_vol_reg rax
alloc_stack 20h
END_PROLOGUE
mov r8, rcx
lea rcx, [rax + 8]
call CopyValueClassUnchecked
add rsp, 20h
pop rax
ret
ClassNotInited:
AllocFailed:
jmp JIT_Box
NESTED_END JIT_BoxFastMP_InlineGetThread, _TEXT
LEAF_ENTRY AllocateStringFastMP_InlineGetThread, _TEXT
; We were passed the number of characters in ECX
; we need to load the method table for string from the global
mov r9, [g_pStringClass]
; Instead of doing elaborate overflow checks, we just limit the number of elements
; to (LARGE_OBJECT_SIZE - 256)/sizeof(WCHAR) or less.
; This will avoid all overflow problems, as well as making sure
; big string objects are correctly allocated in the big object heap.
cmp ecx, (ASM_LARGE_OBJECT_SIZE - 256)/2
jae OversizedString
mov edx, [r9 + OFFSET__MethodTable__m_BaseSize]
; Calculate the final size to allocate.
; We need to calculate baseSize + cnt*2, then round that up by adding 7 and anding ~7.
lea edx, [edx + ecx*2 + 7]
and edx, -8
PATCHABLE_INLINE_GETTHREAD r11, AllocateStringFastMP_InlineGetThread__PatchTLSOffset
mov r10, [r11 + OFFSET__Thread__m_alloc_context__alloc_limit]
mov rax, [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr]
add rdx, rax
cmp rdx, r10
ja AllocFailed
mov [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr], rdx
mov [rax], r9
mov [rax + OFFSETOF__StringObject__m_StringLength], ecx
ifdef _DEBUG
call DEBUG_TrialAllocSetAppDomain_NoScratchArea
endif ; _DEBUG
ret
OversizedString:
AllocFailed:
jmp FramedAllocateString
LEAF_END AllocateStringFastMP_InlineGetThread, _TEXT
; HCIMPL2(Object*, JIT_NewArr1VC_MP_InlineGetThread, CORINFO_CLASS_HANDLE arrayMT, INT_PTR size)
LEAF_ENTRY JIT_NewArr1VC_MP_InlineGetThread, _TEXT
; We were passed a (shared) method table in RCX, which contains the element type.
; The element count is in RDX
; NOTE: if this code is ported for CORINFO_HELP_NEWSFAST_ALIGN8, it will need
; to emulate the double-specific behavior of JIT_TrialAlloc::GenAllocArray.
; Do a conservative check here. This is to avoid overflow while doing the calculations. We don't
; have to worry about "large" objects, since the allocation quantum is never big enough for
; LARGE_OBJECT_SIZE.
; For Value Classes, this needs to be 2^16 - slack (2^32 / max component size),
; The slack includes the size for the array header and round-up ; for alignment. Use 256 for the
; slack value out of laziness.
; In both cases we do a final overflow check after adding to the alloc_ptr.
cmp rdx, (65535 - 256)
jae OversizedArray
movzx r8d, word ptr [rcx + OFFSETOF__MethodTable__m_dwFlags] ; component size is low 16 bits
imul r8d, edx
add r8d, dword ptr [rcx + OFFSET__MethodTable__m_BaseSize]
; round the size to a multiple of 8
add r8d, 7
and r8d, -8
PATCHABLE_INLINE_GETTHREAD r11, JIT_NewArr1VC_MP_InlineGetThread__PatchTLSOffset
mov r10, [r11 + OFFSET__Thread__m_alloc_context__alloc_limit]
mov rax, [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr]
add r8, rax
jc AllocFailed
cmp r8, r10
ja AllocFailed
mov [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr], r8
mov [rax], rcx
mov dword ptr [rax + OFFSETOF__ArrayBase__m_NumComponents], edx
ifdef _DEBUG
call DEBUG_TrialAllocSetAppDomain_NoScratchArea
endif ; _DEBUG
ret
OversizedArray:
AllocFailed:
jmp JIT_NewArr1
LEAF_END JIT_NewArr1VC_MP_InlineGetThread, _TEXT
; HCIMPL2(Object*, JIT_NewArr1OBJ_MP_InlineGetThread, CORINFO_CLASS_HANDLE arrayMT, INT_PTR size)
LEAF_ENTRY JIT_NewArr1OBJ_MP_InlineGetThread, _TEXT
; We were passed a (shared) method table in RCX, which contains the element type.
; The element count is in RDX
; NOTE: if this code is ported for CORINFO_HELP_NEWSFAST_ALIGN8, it will need
; to emulate the double-specific behavior of JIT_TrialAlloc::GenAllocArray.
; Verifies that LARGE_OBJECT_SIZE fits in 32-bit. This allows us to do array size
; arithmetic using 32-bit registers.
.erre ASM_LARGE_OBJECT_SIZE lt 100000000h
cmp rdx, (ASM_LARGE_OBJECT_SIZE - 256)/8 ; sizeof(void*)
jae OversizedArray
; In this case we know the element size is sizeof(void *), or 8 for x64
; This helps us in two ways - we can shift instead of multiplying, and
; there's no need to align the size either
mov r8d, dword ptr [rcx + OFFSET__MethodTable__m_BaseSize]
lea r8d, [r8d + edx * 8]
; No need for rounding in this case - element size is 8, and m_BaseSize is guaranteed
; to be a multiple of 8.
PATCHABLE_INLINE_GETTHREAD r11, JIT_NewArr1OBJ_MP_InlineGetThread__PatchTLSOffset
mov r10, [r11 + OFFSET__Thread__m_alloc_context__alloc_limit]
mov rax, [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr]
add r8, rax
cmp r8, r10
ja AllocFailed
mov [r11 + OFFSET__Thread__m_alloc_context__alloc_ptr], r8
mov [rax], rcx
mov dword ptr [rax + OFFSETOF__ArrayBase__m_NumComponents], edx
ifdef _DEBUG
call DEBUG_TrialAllocSetAppDomain_NoScratchArea
endif ; _DEBUG
ret
OversizedArray:
AllocFailed:
jmp JIT_NewArr1
LEAF_END JIT_NewArr1OBJ_MP_InlineGetThread, _TEXT
end
|
