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
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
|
#
# Copyright (c) 2004, 2007, Oracle and/or its affiliates. All rights reserved.
# DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
#
# This code is free software; you can redistribute it and/or modify it
# under the terms of the GNU General Public License version 2 only, as
# published by the Free Software Foundation.
#
# This code is distributed in the hope that it will be useful, but WITHOUT
# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
# FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
# version 2 for more details (a copy is included in the LICENSE file that
# accompanied this code).
#
# You should have received a copy of the GNU General Public License version
# 2 along with this work; if not, write to the Free Software Foundation,
# Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
#
# Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
# or visit www.oracle.com if you need additional information or have any
# questions.
#
#ifdef __APPLE__
# Darwin uses _ prefixed global symbols
#define SYMBOL(s) _ ## s
#define ELF_TYPE(name, description)
#else
#define SYMBOL(s) s
#define ELF_TYPE(name, description) .type name,description
#endif
# NOTE WELL! The _Copy functions are called directly
# from server-compiler-generated code via CallLeafNoFP,
# which means that they *must* either not use floating
# point or use it in the same manner as does the server
# compiler.
.globl SYMBOL(_Copy_arrayof_conjoint_bytes)
.globl SYMBOL(_Copy_arrayof_conjoint_jshorts)
.globl SYMBOL(_Copy_conjoint_jshorts_atomic)
.globl SYMBOL(_Copy_arrayof_conjoint_jints)
.globl SYMBOL(_Copy_conjoint_jints_atomic)
.globl SYMBOL(_Copy_arrayof_conjoint_jlongs)
.globl SYMBOL(_Copy_conjoint_jlongs_atomic)
.text
.globl SYMBOL(SpinPause)
.p2align 4,,15
ELF_TYPE(SpinPause,@function)
SYMBOL(SpinPause):
rep
nop
movq $1, %rax
ret
# Support for void Copy::arrayof_conjoint_bytes(void* from,
# void* to,
# size_t count)
# rdi - from
# rsi - to
# rdx - count, treated as ssize_t
#
.p2align 4,,15
ELF_TYPE(_Copy_arrayof_conjoint_bytes,@function)
SYMBOL(_Copy_arrayof_conjoint_bytes):
movq %rdx,%r8 # byte count
shrq $3,%rdx # qword count
cmpq %rdi,%rsi
leaq -1(%rdi,%r8,1),%rax # from + bcount*1 - 1
jbe acb_CopyRight
cmpq %rax,%rsi
jbe acb_CopyLeft
acb_CopyRight:
leaq -8(%rdi,%rdx,8),%rax # from + qcount*8 - 8
leaq -8(%rsi,%rdx,8),%rcx # to + qcount*8 - 8
negq %rdx
jmp 7f
.p2align 4,,15
1: movq 8(%rax,%rdx,8),%rsi
movq %rsi,8(%rcx,%rdx,8)
addq $1,%rdx
jnz 1b
2: testq $4,%r8 # check for trailing dword
jz 3f
movl 8(%rax),%esi # copy trailing dword
movl %esi,8(%rcx)
addq $4,%rax
addq $4,%rcx # original %rsi is trashed, so we
# can't use it as a base register
3: testq $2,%r8 # check for trailing word
jz 4f
movw 8(%rax),%si # copy trailing word
movw %si,8(%rcx)
addq $2,%rcx
4: testq $1,%r8 # check for trailing byte
jz 5f
movb -1(%rdi,%r8,1),%al # copy trailing byte
movb %al,8(%rcx)
5: ret
.p2align 4,,15
6: movq -24(%rax,%rdx,8),%rsi
movq %rsi,-24(%rcx,%rdx,8)
movq -16(%rax,%rdx,8),%rsi
movq %rsi,-16(%rcx,%rdx,8)
movq -8(%rax,%rdx,8),%rsi
movq %rsi,-8(%rcx,%rdx,8)
movq (%rax,%rdx,8),%rsi
movq %rsi,(%rcx,%rdx,8)
7: addq $4,%rdx
jle 6b
subq $4,%rdx
jl 1b
jmp 2b
acb_CopyLeft:
testq $1,%r8 # check for trailing byte
jz 1f
movb -1(%rdi,%r8,1),%cl # copy trailing byte
movb %cl,-1(%rsi,%r8,1)
subq $1,%r8 # adjust for possible trailing word
1: testq $2,%r8 # check for trailing word
jz 2f
movw -2(%rdi,%r8,1),%cx # copy trailing word
movw %cx,-2(%rsi,%r8,1)
2: testq $4,%r8 # check for trailing dword
jz 5f
movl (%rdi,%rdx,8),%ecx # copy trailing dword
movl %ecx,(%rsi,%rdx,8)
jmp 5f
.p2align 4,,15
3: movq -8(%rdi,%rdx,8),%rcx
movq %rcx,-8(%rsi,%rdx,8)
subq $1,%rdx
jnz 3b
ret
.p2align 4,,15
4: movq 24(%rdi,%rdx,8),%rcx
movq %rcx,24(%rsi,%rdx,8)
movq 16(%rdi,%rdx,8),%rcx
movq %rcx,16(%rsi,%rdx,8)
movq 8(%rdi,%rdx,8),%rcx
movq %rcx,8(%rsi,%rdx,8)
movq (%rdi,%rdx,8),%rcx
movq %rcx,(%rsi,%rdx,8)
5: subq $4,%rdx
jge 4b
addq $4,%rdx
jg 3b
ret
# Support for void Copy::arrayof_conjoint_jshorts(void* from,
# void* to,
# size_t count)
# Equivalent to
# conjoint_jshorts_atomic
#
# If 'from' and/or 'to' are aligned on 4- or 2-byte boundaries, we
# let the hardware handle it. The tow or four words within dwords
# or qwords that span cache line boundaries will still be loaded
# and stored atomically.
#
# rdi - from
# rsi - to
# rdx - count, treated as ssize_t
#
.p2align 4,,15
ELF_TYPE(_Copy_arrayof_conjoint_jshorts,@function)
ELF_TYPE(_Copy_conjoint_jshorts_atomic,@function)
SYMBOL(_Copy_arrayof_conjoint_jshorts):
SYMBOL(_Copy_conjoint_jshorts_atomic):
movq %rdx,%r8 # word count
shrq $2,%rdx # qword count
cmpq %rdi,%rsi
leaq -2(%rdi,%r8,2),%rax # from + wcount*2 - 2
jbe acs_CopyRight
cmpq %rax,%rsi
jbe acs_CopyLeft
acs_CopyRight:
leaq -8(%rdi,%rdx,8),%rax # from + qcount*8 - 8
leaq -8(%rsi,%rdx,8),%rcx # to + qcount*8 - 8
negq %rdx
jmp 6f
1: movq 8(%rax,%rdx,8),%rsi
movq %rsi,8(%rcx,%rdx,8)
addq $1,%rdx
jnz 1b
2: testq $2,%r8 # check for trailing dword
jz 3f
movl 8(%rax),%esi # copy trailing dword
movl %esi,8(%rcx)
addq $4,%rcx # original %rsi is trashed, so we
# can't use it as a base register
3: testq $1,%r8 # check for trailing word
jz 4f
movw -2(%rdi,%r8,2),%si # copy trailing word
movw %si,8(%rcx)
4: ret
.p2align 4,,15
5: movq -24(%rax,%rdx,8),%rsi
movq %rsi,-24(%rcx,%rdx,8)
movq -16(%rax,%rdx,8),%rsi
movq %rsi,-16(%rcx,%rdx,8)
movq -8(%rax,%rdx,8),%rsi
movq %rsi,-8(%rcx,%rdx,8)
movq (%rax,%rdx,8),%rsi
movq %rsi,(%rcx,%rdx,8)
6: addq $4,%rdx
jle 5b
subq $4,%rdx
jl 1b
jmp 2b
acs_CopyLeft:
testq $1,%r8 # check for trailing word
jz 1f
movw -2(%rdi,%r8,2),%cx # copy trailing word
movw %cx,-2(%rsi,%r8,2)
1: testq $2,%r8 # check for trailing dword
jz 4f
movl (%rdi,%rdx,8),%ecx # copy trailing dword
movl %ecx,(%rsi,%rdx,8)
jmp 4f
2: movq -8(%rdi,%rdx,8),%rcx
movq %rcx,-8(%rsi,%rdx,8)
subq $1,%rdx
jnz 2b
ret
.p2align 4,,15
3: movq 24(%rdi,%rdx,8),%rcx
movq %rcx,24(%rsi,%rdx,8)
movq 16(%rdi,%rdx,8),%rcx
movq %rcx,16(%rsi,%rdx,8)
movq 8(%rdi,%rdx,8),%rcx
movq %rcx,8(%rsi,%rdx,8)
movq (%rdi,%rdx,8),%rcx
movq %rcx,(%rsi,%rdx,8)
4: subq $4,%rdx
jge 3b
addq $4,%rdx
jg 2b
ret
# Support for void Copy::arrayof_conjoint_jints(jint* from,
# jint* to,
# size_t count)
# Equivalent to
# conjoint_jints_atomic
#
# If 'from' and/or 'to' are aligned on 4-byte boundaries, we let
# the hardware handle it. The two dwords within qwords that span
# cache line boundaries will still be loaded and stored atomically.
#
# rdi - from
# rsi - to
# rdx - count, treated as ssize_t
#
.p2align 4,,15
ELF_TYPE(_Copy_arrayof_conjoint_jints,@function)
ELF_TYPE(_Copy_conjoint_jints_atomic,@function)
SYMBOL(_Copy_arrayof_conjoint_jints):
SYMBOL(_Copy_conjoint_jints_atomic):
movq %rdx,%r8 # dword count
shrq %rdx # qword count
cmpq %rdi,%rsi
leaq -4(%rdi,%r8,4),%rax # from + dcount*4 - 4
jbe aci_CopyRight
cmpq %rax,%rsi
jbe aci_CopyLeft
aci_CopyRight:
leaq -8(%rdi,%rdx,8),%rax # from + qcount*8 - 8
leaq -8(%rsi,%rdx,8),%rcx # to + qcount*8 - 8
negq %rdx
jmp 5f
.p2align 4,,15
1: movq 8(%rax,%rdx,8),%rsi
movq %rsi,8(%rcx,%rdx,8)
addq $1,%rdx
jnz 1b
2: testq $1,%r8 # check for trailing dword
jz 3f
movl 8(%rax),%esi # copy trailing dword
movl %esi,8(%rcx)
3: ret
.p2align 4,,15
4: movq -24(%rax,%rdx,8),%rsi
movq %rsi,-24(%rcx,%rdx,8)
movq -16(%rax,%rdx,8),%rsi
movq %rsi,-16(%rcx,%rdx,8)
movq -8(%rax,%rdx,8),%rsi
movq %rsi,-8(%rcx,%rdx,8)
movq (%rax,%rdx,8),%rsi
movq %rsi,(%rcx,%rdx,8)
5: addq $4,%rdx
jle 4b
subq $4,%rdx
jl 1b
jmp 2b
aci_CopyLeft:
testq $1,%r8 # check for trailing dword
jz 3f
movl -4(%rdi,%r8,4),%ecx # copy trailing dword
movl %ecx,-4(%rsi,%r8,4)
jmp 3f
1: movq -8(%rdi,%rdx,8),%rcx
movq %rcx,-8(%rsi,%rdx,8)
subq $1,%rdx
jnz 1b
ret
.p2align 4,,15
2: movq 24(%rdi,%rdx,8),%rcx
movq %rcx,24(%rsi,%rdx,8)
movq 16(%rdi,%rdx,8),%rcx
movq %rcx,16(%rsi,%rdx,8)
movq 8(%rdi,%rdx,8),%rcx
movq %rcx,8(%rsi,%rdx,8)
movq (%rdi,%rdx,8),%rcx
movq %rcx,(%rsi,%rdx,8)
3: subq $4,%rdx
jge 2b
addq $4,%rdx
jg 1b
ret
# Support for void Copy::arrayof_conjoint_jlongs(jlong* from,
# jlong* to,
# size_t count)
# Equivalent to
# conjoint_jlongs_atomic
# arrayof_conjoint_oops
# conjoint_oops_atomic
#
# rdi - from
# rsi - to
# rdx - count, treated as ssize_t
#
.p2align 4,,15
ELF_TYPE(_Copy_arrayof_conjoint_jlongs,@function)
ELF_TYPE(_Copy_conjoint_jlongs_atomic,@function)
SYMBOL(_Copy_arrayof_conjoint_jlongs):
SYMBOL(_Copy_conjoint_jlongs_atomic):
cmpq %rdi,%rsi
leaq -8(%rdi,%rdx,8),%rax # from + count*8 - 8
jbe acl_CopyRight
cmpq %rax,%rsi
jbe acl_CopyLeft
acl_CopyRight:
leaq -8(%rsi,%rdx,8),%rcx # to + count*8 - 8
negq %rdx
jmp 3f
1: movq 8(%rax,%rdx,8),%rsi
movq %rsi,8(%rcx,%rdx,8)
addq $1,%rdx
jnz 1b
ret
.p2align 4,,15
2: movq -24(%rax,%rdx,8),%rsi
movq %rsi,-24(%rcx,%rdx,8)
movq -16(%rax,%rdx,8),%rsi
movq %rsi,-16(%rcx,%rdx,8)
movq -8(%rax,%rdx,8),%rsi
movq %rsi,-8(%rcx,%rdx,8)
movq (%rax,%rdx,8),%rsi
movq %rsi,(%rcx,%rdx,8)
3: addq $4,%rdx
jle 2b
subq $4,%rdx
jl 1b
ret
4: movq -8(%rdi,%rdx,8),%rcx
movq %rcx,-8(%rsi,%rdx,8)
subq $1,%rdx
jnz 4b
ret
.p2align 4,,15
5: movq 24(%rdi,%rdx,8),%rcx
movq %rcx,24(%rsi,%rdx,8)
movq 16(%rdi,%rdx,8),%rcx
movq %rcx,16(%rsi,%rdx,8)
movq 8(%rdi,%rdx,8),%rcx
movq %rcx,8(%rsi,%rdx,8)
movq (%rdi,%rdx,8),%rcx
movq %rcx,(%rsi,%rdx,8)
acl_CopyLeft:
subq $4,%rdx
jge 5b
addq $4,%rdx
jg 4b
ret
|
