-
Notifications
You must be signed in to change notification settings - Fork 0
/
mutate.jl
602 lines (537 loc) · 41 KB
/
mutate.jl
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
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
using MemoryMutate
# runs on my machine: archlinux x86_64
include("testutils.jl")
#= we have
- Julep: setfield! for mutable references to immutables
https://github.com/JuliaLang/julia/issues/17115
- "we propose making it possible to have setfield! modify fields inside of immutable objects that are wrapped in mutable objects"
- "To support this proposal, the setfield! function will get a multi-arg form, with the following behaviors:"
"setfield!(x, a, b, c, value) mutates the right most mutable object to change the value of its fields to be"
"equivalent to copying the immutable objects and updating the referenced field."
- "tl;dr The syntax:"
"x.a.b.c = 3"
"would now be valid, as long as at least one of the referenced fields is mutable."
- WIP: Make mutating immutables easier
https://github.com/JuliaLang/julia/pull/21912
- proposes an @ operator as in `x@a = 2` for an immutable `x`
- "The ways this works is that under the hood, it creates a new immutable object with the specified field modified and then assigns it back to the appropriate place."
"Syntax wise, everything to the left of the @ is what's being assigned to, everything to the right of the @ is what is to be modified."
- Make immutable mutable again - or make arrays of structures stack-allocated
https://discourse.julialang.org/t/make-immutable-mutable-again-or-make-arrays-of-structures-stack-allocated/20800
- colleced a few related links
=#
################################################################################
struct G1; x :: Float32; end # struct G2 { const float x; };
struct F1; x :: Float32; g :: G1; end # struct F2 { const float x; const G2 g; };
struct E1; x :: Float32; f :: F1; end # struct E2 { const float x; const F2 g; };
struct D1; x :: Float32; e :: E1; end # struct D2 { const float x; const E2 g; };
struct C1; x :: Float32; d :: D1; end # struct C2 { const float x; const D2 g; };
struct B1; x :: Float32; c :: C1; end # struct B2 { const float x; const C2 g; };
mutable struct A1; x :: Float32; b :: B1; end # struct A2 { float x; B2 g; };
g1 = G1(0.0f0)
f1 = F1(0.0f0,g1)
e1 = E1(0.0f0,f1)
d1 = D1(0.0f0,e1)
c1 = C1(0.0f0,d1)
b1 = B1(0.0f0,c1)
a1 = A1(0.0f0,b1)
test101(a::A1,v::Float32) = @mem a.b.c.d.e.f.g.x = v
test102(a::A1,v::G1) = @mem a.b.c.d.e.f.g = v
test103(a::A1,v::Float32) = @mem a.b.c.d.e.f.x = v
test104(a::A1,v::F1) = @mem a.b.c.d.e.f = v
test105(a::A1,v::Float32) = @mem a.b.c.d.e.x = v
test106(a::A1,v::E1) = @mem a.b.c.d.e = v
test107(a::A1,v::Float32) = @mem a.b.c.d.x = v
test108(a::A1,v::D1) = @mem a.b.c.d = v
test109(a::A1,v::Float32) = @mem a.b.c.x = v
test110(a::A1,v::C1) = @mem a.b.c = v
test111(a::A1,v::Float32) = @mem a.b.x = v
test112(a::A1,v::B1) = @mem a.b = v
test113(a::A1,v::Float32) = @mem a.x = v
test114(a::A1,v::A1) = @mem a = v
v = 0f0
v += 1f0; @assert a1.b.c.d.e.f.g.x != v; test101(a1,v); @assert a1.b.c.d.e.f.g.x == v
v += 1f0; @assert a1.b.c.d.e.f.g != G1(v); test102(a1,G1(v)); @assert a1.b.c.d.e.f.g == G1(v)
v += 1f0; @assert a1.b.c.d.e.f.x != v; test103(a1,v); @assert a1.b.c.d.e.f.x == v
v += 1f0; @assert a1.b.c.d.e.f != F1(v,g1); test104(a1,F1(v,g1)); @assert a1.b.c.d.e.f == F1(v,g1)
v += 1f0; @assert a1.b.c.d.e.x != v; test105(a1,v); @assert a1.b.c.d.e.x == v
v += 1f0; @assert a1.b.c.d.e != E1(v,f1); test106(a1,E1(v,f1)); @assert a1.b.c.d.e == E1(v,f1)
v += 1f0; @assert a1.b.c.d.x != v; test107(a1,v); @assert a1.b.c.d.x == v
v += 1f0; @assert a1.b.c.d != D1(v,e1); test108(a1,D1(v,e1)); @assert a1.b.c.d == D1(v,e1)
v += 1f0; @assert a1.b.c.x != v; test109(a1,v); @assert a1.b.c.x == v
v += 1f0; @assert a1.b.c != C1(v,d1); test110(a1,C1(v,d1)); @assert a1.b.c == C1(v,d1)
v += 1f0; @assert a1.b.x != v; test111(a1,v); @assert a1.b.x == v
v += 1f0; @assert a1.b != B1(v,c1); test112(a1,B1(v,c1)); @assert a1.b == B1(v,c1)
v += 1f0; @assert a1.x != v; test113(a1,v); @assert a1.x == v
v += 1f0; @assert a1 != A1(v,b1); test114(a1,A1(v,b1)); # @assert a1 == A1(v,b1) # call by "sharing"
code_native(io,test101,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test102,(A1,G1)) ; display_asm_stat_io(io) # (total = 4, movs = 3, mov = 2, vmov = 1)
code_native(io,test103,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test104,(A1,F1)) ; display_asm_stat_io(io) # (total = 5, movs = 4, mov = 2, vmov = 2)
code_native(io,test105,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test106,(A1,E1)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 9, vmov = 0)
code_native(io,test107,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test108,(A1,D1)) ; display_asm_stat_io(io) # (total = 6, movs = 5, mov = 1, vmov = 4)
code_native(io,test109,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test110,(A1,C1)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test111,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test112,(A1,B1)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test113,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test114,(A1,A1)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
a1 = A1(0.0f0,b1)
test101b(a::A1,v::Float32) = @mem getfield(a.b.c,:d).e.f.g.x = v
test102b(a::A1,v::G1) = @mem getfield(a.b.c,:d).e.f.g = v
test103b(a::A1,v::Float32) = @mem getfield(a.b.c,:d).e.f.x = v
test104b(a::A1,v::F1) = @mem getfield(a.b.c,:d).e.f = v
test105b(a::A1,v::Float32) = @mem getfield(a.b.c,:d).e.x = v
test106b(a::A1,v::E1) = @mem getfield(a.b.c,:d).e = v
test107b(a::A1,v::Float32) = @mem getfield(a.b.c,:d).x = v
test108b(a::A1,v::D1) = @mem getfield(a.b.c,:d) = v
test109b(a::A1,v::Float32) = @mem a.b.c.x = v
test110b(a::A1,v::C1) = @mem a.b.c = v
test111b(a::A1,v::Float32) = @mem a.b.x = v
test112b(a::A1,v::B1) = @mem a.b = v
test113b(a::A1,v::Float32) = @mem a.x = v
test114b(a::A1,v::A1) = @mem a = v
v = 0f0
v += 1f0; @assert a1.b.c.d.e.f.g.x != v; test101b(a1,v); @assert a1.b.c.d.e.f.g.x == v
v += 1f0; @assert a1.b.c.d.e.f.g != G1(v); test102b(a1,G1(v)); @assert a1.b.c.d.e.f.g == G1(v)
v += 1f0; @assert a1.b.c.d.e.f.x != v; test103b(a1,v); @assert a1.b.c.d.e.f.x == v
v += 1f0; @assert a1.b.c.d.e.f != F1(v,g1); test104b(a1,F1(v,g1)); @assert a1.b.c.d.e.f == F1(v,g1)
v += 1f0; @assert a1.b.c.d.e.x != v; test105b(a1,v); @assert a1.b.c.d.e.x == v
v += 1f0; @assert a1.b.c.d.e != E1(v,f1); test106b(a1,E1(v,f1)); @assert a1.b.c.d.e == E1(v,f1)
v += 1f0; @assert a1.b.c.d.x != v; test107b(a1,v); @assert a1.b.c.d.x == v
v += 1f0; @assert a1.b.c.d != D1(v,e1); test108b(a1,D1(v,e1)); @assert a1.b.c.d == D1(v,e1)
v += 1f0; @assert a1.b.c.x != v; test109b(a1,v); @assert a1.b.c.x == v
v += 1f0; @assert a1.b.c != C1(v,d1); test110b(a1,C1(v,d1)); @assert a1.b.c == C1(v,d1)
v += 1f0; @assert a1.b.x != v; test111b(a1,v); @assert a1.b.x == v
v += 1f0; @assert a1.b != B1(v,c1); test112b(a1,B1(v,c1)); @assert a1.b == B1(v,c1)
v += 1f0; @assert a1.x != v; test113b(a1,v); @assert a1.x == v
v += 1f0; @assert a1 != A1(v,b1); test114b(a1,A1(v,b1)); # @assert a1 == A1(v,b1) # call by "sharing"
code_native(io,test101b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test102b,(A1,G1)) ; display_asm_stat_io(io) # (total = 4, movs = 3, mov = 2, vmov = 1)
code_native(io,test103b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test104b,(A1,F1)) ; display_asm_stat_io(io) # (total = 5, movs = 4, mov = 2, vmov = 2)
code_native(io,test105b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test106b,(A1,E1)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 9, vmov = 0)
code_native(io,test107b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test108b,(A1,D1)) ; display_asm_stat_io(io) # (total = 6, movs = 5, mov = 1, vmov = 4)
code_native(io,test109b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test110b,(A1,C1)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test111b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test112b,(A1,B1)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test113b,(A1,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test114b,(A1,A1)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
struct G2; x :: Float32; end # struct G2 { const float x; };
struct F2; x :: Float32; g :: G2; end # struct F2 { const float x; const G2 g; };
struct E2; x :: Float32; f :: F2; end # struct E2 { const float x; const F2 g; };
mutable struct D2; x :: Float32; e :: E2; end # struct D2 { float x; E2 g; };
struct C2; x :: Float32; d :: D2; end # struct C2 { const float x; const D2& g; };
struct B2; x :: Float32; c :: C2; end # struct B2 { const float x; const C2& g; };
mutable struct A2; x :: Float32; b :: B2; end # struct A2 { float x; B2& g; };
g2 = G2(6.0f0)
f2 = F2(5.0f0,g2)
e2 = E2(4.0f0,f2)
d2 = D2(3.0f0,e2)
c2 = C2(2.0f0,d2)
b2 = B2(1.0f0,c2)
a2 = A2(0.0f0,b2)
test201(a::A2,v::Float32) = @mem a.b.c.d.e.f.g.x = v
test202(a::A2,v::G2) = @mem a.b.c.d.e.f.g = v
test203(a::A2,v::Float32) = @mem a.b.c.d.e.f.x = v
test204(a::A2,v::F2) = @mem a.b.c.d.e.f = v
test205(a::A2,v::Float32) = @mem a.b.c.d.e.x = v
test206(a::A2,v::E2) = @mem a.b.c.d.e = v
test207(a::A2,v::Float32) = @mem a.b.c.d.x = v
test208(a::A2,v::D2) = @yolo a.b.c.d = v
test209(a::A2,v::Float32) = @yolo a.b.c.x = v
test210(a::A2,v::C2) = @yolo a.b.c = v
test211(a::A2,v::Float32) = @yolo a.b.x = v
test212(a::A2,v::B2) = @mem a.b = v
test213(a::A2,v::Float32) = @mem a.x = v
test214(a::A2,v::A2) = @mem a = v
v = 0f0
v += 1f0; @assert a2.b.c.d.e.f.g.x != v; test201(a2,v); @assert a2.b.c.d.e.f.g.x == v
v += 1f0; @assert a2.b.c.d.e.f.g != G2(v); test202(a2,G2(v)); @assert a2.b.c.d.e.f.g == G2(v)
v += 1f0; @assert a2.b.c.d.e.f.x != v; test203(a2,v); @assert a2.b.c.d.e.f.x == v
v += 1f0; @assert a2.b.c.d.e.f != F2(v,g2); test204(a2,F2(v,g2)); @assert a2.b.c.d.e.f == F2(v,g2)
v += 1f0; @assert a2.b.c.d.e.x != v; test205(a2,v); @assert a2.b.c.d.e.x == v
v += 1f0; @assert a2.b.c.d.e != E2(v,f2); test206(a2,E2(v,f2)); @assert a2.b.c.d.e == E2(v,f2)
v += 1f0; @assert a2.b.c.d.x != v; test207(a2,v); @assert a2.b.c.d.x == v
try; global v += 1f0; @assert a2.b.c.d != D2(v,e2); test208(a2,D2(v,e2)); @assert a2.b.c.d.x == D2(v,e2).x # mutables with equal field values are not equal
@assert a2.b.c.d.e == D2(v,e2).e; catch e; print(e); end
try; global v += 1f0; @assert a2.b.c.x != v; test209(a2,v); @assert a2.b.c.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a2.b.c != C2(v,d2); test210(a2,C2(v,d2)); @assert a2.b.c == C2(v,d2) ; catch e; print(e); end
try; global v += 1f0; @assert a2.b.x != v; test211(a2,v); @assert a2.b.x == v ; catch e; print(e); end
v += 1f0; @assert a2.b != B2(v,c2); test212(a2,B2(v,c2)); @assert a2.b == B2(v,c2)
v += 1f0; @assert a2.x != v; test213(a2,v); @assert a2.x == v
v += 1f0; @assert a2 != A2(v,b2); test214(a2,A2(v,b2)); # @assert a1 == A2(v,b2) # call by "sharing"
code_native(io,test201,(A2,Float32)); display_asm_stat_io(io) # (total = 5, movs = 4, mov = 3, vmov = 1)
code_native(io,test202,(A2,G2)) ; display_asm_stat_io(io) # (total = 7, movs = 6, mov = 5, vmov = 1)
code_native(io,test203,(A2,Float32)); display_asm_stat_io(io) # (total = 5, movs = 4, mov = 3, vmov = 1)
code_native(io,test204,(A2,F2)) ; display_asm_stat_io(io) # (total = 9, movs = 7, mov = 5, vmov = 2)
code_native(io,test205,(A2,Float32)); display_asm_stat_io(io) # (total = 5, movs = 4, mov = 3, vmov = 1)
code_native(io,test206,(A2,E2)) ; display_asm_stat_io(io) # (total = 14, movs = 12, mov = 12, vmov = 0)
code_native(io,test207,(A2,Float32)); display_asm_stat_io(io) # (total = 5, movs = 4, mov = 3, vmov = 1)
code_native(io,test208,(A2,D2)) ; display_asm_stat_io(io) # (total = 7, movs = 6, mov = 6, vmov = 0)
code_native(io,test209,(A2,Float32)); display_asm_stat_io(io) # (total = 28, movs = 18, mov = 15, vmov = 3)
code_native(io,test210,(A2,C2)) ; display_asm_stat_io(io) # (total = 18, movs = 15, mov = 14, vmov = 1)
code_native(io,test211,(A2,Float32)); display_asm_stat_io(io) # (total = 26, movs = 17, mov = 14, vmov = 3)
code_native(io,test212,(A2,B2)) ; display_asm_stat_io(io) # (total = 24, movs = 8, mov = 8, vmov = 0)
code_native(io,test213,(A2,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test214,(A2,A2)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
struct G3; x :: Float32; end # struct G3 { const float x; };
struct F3; x :: Float32; g :: G3; end # struct F3 { const float x; const G3 g; };
struct E3; x :: Float32; f :: F3; end # struct E3 { const float x; const F3 g; };
struct D3; x :: Float32; e :: E3; end # struct D3 { const float x; const E3 g; };
struct C3; x :: Float32; d :: D3; end # struct C3 { const float x; const D3 g; };
struct B3; x :: Float32; c :: C3; end # struct B3 { const float x; const C3 g; };
struct A3; x :: Float32; b :: B3; end # struct A3 { const float x; const B3 g; };
g3 = G3(6.0f0)
f3 = F3(5.0f0,g3)
e3 = E3(4.0f0,f3)
d3 = D3(3.0f0,e3)
c3 = C3(2.0f0,d3)
b3 = B3(1.0f0,c3)
a3 = A3(0.0f0,b3)
test301(a::Base.RefValue{A3},v::Float32) = @mem a[].b.c.d.e.f.g.x = v
test302(a::Base.RefValue{A3},v::G3) = @mem a[].b.c.d.e.f.g = v
test303(a::Base.RefValue{A3},v::Float32) = @mem a[].b.c.d.e.f.x = v
test304(a::Base.RefValue{A3},v::F3) = @mem a[].b.c.d.e.f = v
test305(a::Base.RefValue{A3},v::Float32) = @mem a[].b.c.d.e.x = v
test306(a::Base.RefValue{A3},v::E3) = @mem a[].b.c.d.e = v
test307(a::Base.RefValue{A3},v::Float32) = @mem a[].b.c.d.x = v
test308(a::Base.RefValue{A3},v::D3) = @mem a[].b.c.d = v
test309(a::Base.RefValue{A3},v::Float32) = @mem a[].b.c.x = v
test310(a::Base.RefValue{A3},v::C3) = @mem a[].b.c = v
test311(a::Base.RefValue{A3},v::Float32) = @mem a[].b.x = v
test312(a::Base.RefValue{A3},v::B3) = @mem a[].b = v
test313(a::Base.RefValue{A3},v::Float32) = @mem a[].x = v
test314(a::Base.RefValue{A3},v::A3) = @mem a[] = v
a3r = Ref(a3)
v = 0f0
v += 1f0; @assert a3r[].b.c.d.e.f.g.x != v; test301(a3r,v); @assert a3r[].b.c.d.e.f.g.x == v
v += 1f0; @assert a3r[].b.c.d.e.f.g != G3(v); test302(a3r,G3(v)); @assert a3r[].b.c.d.e.f.g == G3(v)
v += 1f0; @assert a3r[].b.c.d.e.f.x != v; test303(a3r,v); @assert a3r[].b.c.d.e.f.x == v
v += 1f0; @assert a3r[].b.c.d.e.f != F3(v,g3); test304(a3r,F3(v,g3)); @assert a3r[].b.c.d.e.f == F3(v,g3)
v += 1f0; @assert a3r[].b.c.d.e.x != v; test305(a3r,v); @assert a3r[].b.c.d.e.x == v
v += 1f0; @assert a3r[].b.c.d.e != E3(v,f3); test306(a3r,E3(v,f3)); @assert a3r[].b.c.d.e == E3(v,f3)
v += 1f0; @assert a3r[].b.c.d.x != v; test307(a3r,v); @assert a3r[].b.c.d.x == v
v += 1f0; @assert a3r[].b.c.d != D3(v,e3); test308(a3r,D3(v,e3)); @assert a3r[].b.c.d == D3(v,e3)
v += 1f0; @assert a3r[].b.c.x != v; test309(a3r,v); @assert a3r[].b.c.x == v
v += 1f0; @assert a3r[].b.c != C3(v,d3); test310(a3r,C3(v,d3)); @assert a3r[].b.c == C3(v,d3)
v += 1f0; @assert a3r[].b.x != v; test311(a3r,v); @assert a3r[].b.x == v
v += 1f0; @assert a3r[].b != B3(v,c3); test312(a3r,B3(v,c3)); @assert a3r[].b == B3(v,c3)
v += 1f0; @assert a3r[].x != v; test313(a3r,v); @assert a3r[].x == v
v += 1f0; @assert a3r[] != A3(v,b3); test314(a3r,A3(v,b3)); @assert a3r[] == A3(v,b3)
code_native(io,test301,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test302,(Base.RefValue{A3},G3)) ; display_asm_stat_io(io) # (total = 4, movs = 3, mov = 2, vmov = 1)
code_native(io,test303,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test304,(Base.RefValue{A3},F3)) ; display_asm_stat_io(io) # (total = 5, movs = 4, mov = 2, vmov = 2)
code_native(io,test305,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test306,(Base.RefValue{A3},E3)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 9, vmov = 0)
code_native(io,test307,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test308,(Base.RefValue{A3},D3)) ; display_asm_stat_io(io) # (total = 6, movs = 5, mov = 1, vmov = 4)
code_native(io,test309,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test310,(Base.RefValue{A3},C3)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test311,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test312,(Base.RefValue{A3},B3)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test313,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test314,(Base.RefValue{A3},A3)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 1, vmov = 8)
################################################################################
test301b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.c.d.e.f.g.x = v
test302b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::G3) = @mem a[][][].b.c.d.e.f.g = v
test303b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.c.d.e.f.x = v
test304b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::F3) = @mem a[][][].b.c.d.e.f = v
test305b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.c.d.e.x = v
test306b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::E3) = @mem a[][][].b.c.d.e = v
test307b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.c.d.x = v
test308b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::D3) = @mem a[][][].b.c.d = v
test309b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.c.x = v
test310b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::C3) = @mem a[][][].b.c = v
test311b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].b.x = v
test312b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::B3) = @mem a[][][].b = v
test313b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::Float32) = @mem a[][][].x = v
test314b(a::Base.RefValue{Base.RefValue{Base.RefValue{A3}}},v::A3) = @mem a[][][] = v
a3rrr = Ref(Ref(Ref(a3)))
# @assert false
v = 0f0
v += 1f0; @assert a3rrr[][][].b.c.d.e.f.g.x != v; test301b(a3rrr,v); @assert a3rrr[][][].b.c.d.e.f.g.x == v
v += 1f0; @assert a3rrr[][][].b.c.d.e.f.g != G3(v); test302b(a3rrr,G3(v)); @assert a3rrr[][][].b.c.d.e.f.g == G3(v)
v += 1f0; @assert a3rrr[][][].b.c.d.e.f.x != v; test303b(a3rrr,v); @assert a3rrr[][][].b.c.d.e.f.x == v
v += 1f0; @assert a3rrr[][][].b.c.d.e.f != F3(v,g3); test304b(a3rrr,F3(v,g3)); @assert a3rrr[][][].b.c.d.e.f == F3(v,g3)
v += 1f0; @assert a3rrr[][][].b.c.d.e.x != v; test305b(a3rrr,v); @assert a3rrr[][][].b.c.d.e.x == v
v += 1f0; @assert a3rrr[][][].b.c.d.e != E3(v,f3); test306b(a3rrr,E3(v,f3)); @assert a3rrr[][][].b.c.d.e == E3(v,f3)
v += 1f0; @assert a3rrr[][][].b.c.d.x != v; test307b(a3rrr,v); @assert a3rrr[][][].b.c.d.x == v
v += 1f0; @assert a3rrr[][][].b.c.d != D3(v,e3); test308b(a3rrr,D3(v,e3)); @assert a3rrr[][][].b.c.d == D3(v,e3)
v += 1f0; @assert a3rrr[][][].b.c.x != v; test309b(a3rrr,v); @assert a3rrr[][][].b.c.x == v
v += 1f0; @assert a3rrr[][][].b.c != C3(v,d3); test310b(a3rrr,C3(v,d3)); @assert a3rrr[][][].b.c == C3(v,d3)
v += 1f0; @assert a3rrr[][][].b.x != v; test311b(a3rrr,v); @assert a3rrr[][][].b.x == v
v += 1f0; @assert a3rrr[][][].b != B3(v,c3); test312b(a3rrr,B3(v,c3)); @assert a3rrr[][][].b == B3(v,c3)
v += 1f0; @assert a3rrr[][][].x != v; test313b(a3rrr,v); @assert a3rrr[][][].x == v
v += 1f0; @assert a3rrr[][][] != A3(v,b3); test314b(a3rrr,A3(v,b3)); @assert a3rrr[][][] == A3(v,b3)
code_native(io,test301,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test302,(Base.RefValue{A3},G3)) ; display_asm_stat_io(io) # (total = 4, movs = 3, mov = 2, vmov = 1)
code_native(io,test303,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test304,(Base.RefValue{A3},F3)) ; display_asm_stat_io(io) # (total = 5, movs = 4, mov = 2, vmov = 2)
code_native(io,test305,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test306,(Base.RefValue{A3},E3)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 9, vmov = 0)
code_native(io,test307,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test308,(Base.RefValue{A3},D3)) ; display_asm_stat_io(io) # (total = 6, movs = 5, mov = 1, vmov = 4)
code_native(io,test309,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test310,(Base.RefValue{A3},C3)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test311,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test312,(Base.RefValue{A3},B3)) ; display_asm_stat_io(io) # (total = 10, movs = 9, mov = 5, vmov = 4)
code_native(io,test313,(Base.RefValue{A3},Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test314,(Base.RefValue{A3},A3)) ; display_asm_stat_io(io) # (total = 11, movs = 9, mov = 1, vmov = 8)
################################################################################
struct G4; x :: Float32; end # struct G4 { const float x; };
struct F4; x :: Float32; g :: G4; end # struct F4 { const float x; const G4 g; };
struct E4; x :: Float32; f :: F4; end # struct E4 { const float x; const F4 g; };
struct D4; x :: Float32; e :: Base.RefValue{E4}; end # struct D4 { const float x; const E4& g; };
struct C4; x :: Float32; d :: D4; end # struct C4 { const float x; const D4& g; };
struct B4; x :: Float32; c :: C4; end # struct B4 { const float x; const C4& g; };
mutable struct A4; x :: Float32; b :: B4; end # struct A4 { float x; B4& g; };
g4 = G4(0.0f0)
f4 = F4(0.0f0,g4)
e4 = E4(0.0f0,f4)
d4 = D4(0.0f0,Ref(e4))
c4 = C4(0.0f0,d4)
b4 = B4(0.0f0,c4)
a4 = A4(0.0f0,b4)
test401(a::A4,v::Float32) = @mem a.b.c.d.e[].f.g.x = v
test402(a::A4,v::G4) = @mem a.b.c.d.e[].f.g = v
test403(a::A4,v::Float32) = @mem a.b.c.d.e[].f.x = v
test404(a::A4,v::F4) = @mem a.b.c.d.e[].f = v
test405(a::A4,v::Float32) = @mem a.b.c.d.e[].x = v
test406(a::A4,v::E4) = @mem a.b.c.d.e[] = v
test407(a::A4,v::Float32) = @yolo a.b.c.d.x = v
test408(a::A4,v::D4) = @yolo a.b.c.d = v
test409(a::A4,v::Float32) = @yolo a.b.c.x = v
test410(a::A4,v::C4) = @yolo a.b.c = v
test411(a::A4,v::Float32) = @yolo a.b.x = v
test412(a::A4,v::B4) = @mem a.b = v
test413(a::A4,v::Float32) = @mem a.x = v
test414(a::A4,v::A4) = @mem a = v
v = 0f0
v += 1f0; @assert a4.b.c.d.e[].f.g.x != v; test401(a4,v); @assert a4.b.c.d.e[].f.g.x == v
v += 1f0; @assert a4.b.c.d.e[].f.g != G4(v); test402(a4,G4(v)); @assert a4.b.c.d.e[].f.g == G4(v)
v += 1f0; @assert a4.b.c.d.e[].f.x != v; test403(a4,v); @assert a4.b.c.d.e[].f.x == v
v += 1f0; @assert a4.b.c.d.e[].f != F4(v,g4); test404(a4,F4(v,g4)); @assert a4.b.c.d.e[].f == F4(v,g4)
v += 1f0; @assert a4.b.c.d.e[].x != v; test405(a4,v); @assert a4.b.c.d.e[].x == v
v += 1f0; @assert a4.b.c.d.e[] != E4(v,f4); test406(a4,E4(v,f4)); @assert a4.b.c.d.e[] == E4(v,f4)
try; global v += 1f0; @assert a4.b.c.d.x != v; test407(a4,v); @assert a4.b.c.d.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c.d != D4(v,Ref(e4)); test408(a4,D4(v,Ref(e4))); @assert a4.b.c.d.e[] == e4 # references to equal immutables are not equal
@assert a4.b.c.d.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c.x != v; test409(a4,v); @assert a4.b.c.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c != C4(v,d4); test410(a4,C4(v,d4)); @assert a4.b.c == C4(v,d4) ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.x != v; test411(a4,v); @assert a4.b.x == v ; catch e; print(e); end
v += 1f0; @assert a4.b != B4(v,c4); test412(a4,B4(v,c4)); @assert a4.b == B4(v,c4)
v += 1f0; @assert a4.x != v; test413(a4,v); @assert a4.x == v
v += 1f0; @assert a4 != A4(v,b4); test414(a4,A4(v,b4)); # @assert a4 == A4(v,b4) # call by "sharing"
code_native(io,test401,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test402,(A4,G4)) ; display_asm_stat_io(io) # (total = 8, movs = 7, mov = 6, vmov = 1)
code_native(io,test403,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test404,(A4,F4)) ; display_asm_stat_io(io) # (total = 9, movs = 8, mov = 6, vmov = 2)
code_native(io,test405,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test406,(A4,E4)) ; display_asm_stat_io(io) # (total = 15, movs = 13, mov = 13, vmov = 0)
code_native(io,test407,(A4,Float32)); display_asm_stat_io(io) # (total = 29, movs = 19, mov = 16, vmov = 3)
code_native(io,test408,(A4,D4)) ; display_asm_stat_io(io) # (total = 19, movs = 16, mov = 15, vmov = 1)
code_native(io,test409,(A4,Float32)); display_asm_stat_io(io) # (total = 28, movs = 18, mov = 15, vmov = 3)
code_native(io,test410,(A4,C4)) ; display_asm_stat_io(io) # (total = 18, movs = 15, mov = 14, vmov = 1)
code_native(io,test411,(A4,Float32)); display_asm_stat_io(io) # (total = 26, movs = 17, mov = 14, vmov = 3)
code_native(io,test412,(A4,B4)) ; display_asm_stat_io(io) # (total = 24, movs = 8, mov = 8, vmov = 0)
code_native(io,test413,(A4,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test414,(A4,A4)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
g4 = G4(0.0f0)
f4 = F4(0.0f0,g4)
e4 = E4(0.0f0,f4)
d4 = D4(0.0f0,Ref(e4))
c4 = C4(0.0f0,d4)
b4 = B4(0.0f0,c4)
a4 = A4(0.0f0,b4)
test401b(a::A4,v::Float32) = @mem a.b.c.d.e.x.f.g.x = v
test402b(a::A4,v::G4) = @mem a.b.c.d.e.x.f.g = v
test403b(a::A4,v::Float32) = @mem a.b.c.d.e.x.f.x = v
test404b(a::A4,v::F4) = @mem a.b.c.d.e.x.f = v
test405b(a::A4,v::Float32) = @mem a.b.c.d.e.x.x = v
test406b(a::A4,v::E4) = @mem a.b.c.d.e.x = v
test407b(a::A4,v::Float32) = @yolo a.b.c.d.x = v
test408b(a::A4,v::D4) = @yolo a.b.c.d = v
test409b(a::A4,v::Float32) = @yolo a.b.c.x = v
test410b(a::A4,v::C4) = @yolo a.b.c = v
test411b(a::A4,v::Float32) = @yolo a.b.x = v
test412b(a::A4,v::B4) = @mem a.b = v
test413b(a::A4,v::Float32) = @mem a.x = v
test414b(a::A4,v::A4) = @mem a = v
v = 0f0
v += 1f0; @assert a4.b.c.d.e.x.f.g.x != v; test401b(a4,v); @assert a4.b.c.d.e.x.f.g.x == v
v += 1f0; @assert a4.b.c.d.e.x.f.g != G4(v); test402b(a4,G4(v)); @assert a4.b.c.d.e.x.f.g == G4(v)
v += 1f0; @assert a4.b.c.d.e.x.f.x != v; test403b(a4,v); @assert a4.b.c.d.e.x.f.x == v
v += 1f0; @assert a4.b.c.d.e.x.f != F4(v,g4); test404b(a4,F4(v,g4)); @assert a4.b.c.d.e.x.f == F4(v,g4)
v += 1f0; @assert a4.b.c.d.e.x.x != v; test405b(a4,v); @assert a4.b.c.d.e.x.x == v
v += 1f0; @assert a4.b.c.d.e.x != E4(v,f4); test406b(a4,E4(v,f4)); @assert a4.b.c.d.e.x == E4(v,f4)
try; global v += 1f0; @assert a4.b.c.d.x != v; test407b(a4,v); @assert a4.b.c.d.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c.d != D4(v,Ref(e4)); test408b(a4,D4(v,Ref(e4))); @assert a4.b.c.d.e[] == e4 # references to equal immutables are not equal
@assert a4.b.c.d.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c.x != v; test409b(a4,v); @assert a4.b.c.x == v ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.c != C4(v,d4); test410b(a4,C4(v,d4)); @assert a4.b.c == C4(v,d4) ; catch e; print(e); end
try; global v += 1f0; @assert a4.b.x != v; test411b(a4,v); @assert a4.b.x == v ; catch e; print(e); end
v += 1f0; @assert a4.b != B4(v,c4); test412b(a4,B4(v,c4)); @assert a4.b == B4(v,c4)
v += 1f0; @assert a4.x != v; test413b(a4,v); @assert a4.x == v
v += 1f0; @assert a4 != A4(v,b4); test414b(a4,A4(v,b4)); # @assert a4 == A4(v,b4) # call by "sharing"
code_native(io,test401b,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test402b,(A4,G4)) ; display_asm_stat_io(io) # (total = 8, movs = 7, mov = 6, vmov = 1)
code_native(io,test403b,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test404b,(A4,F4)) ; display_asm_stat_io(io) # (total = 9, movs = 8, mov = 6, vmov = 2)
code_native(io,test405b,(A4,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 4, vmov = 1)
code_native(io,test406b,(A4,E4)) ; display_asm_stat_io(io) # (total = 15, movs = 13, mov = 13, vmov = 0)
code_native(io,test407b,(A4,Float32)); display_asm_stat_io(io) # (total = 29, movs = 19, mov = 16, vmov = 3)
code_native(io,test408b,(A4,D4)) ; display_asm_stat_io(io) # (total = 19, movs = 16, mov = 15, vmov = 1)
code_native(io,test409b,(A4,Float32)); display_asm_stat_io(io) # (total = 28, movs = 18, mov = 15, vmov = 3)
code_native(io,test410b,(A4,C4)) ; display_asm_stat_io(io) # (total = 18, movs = 15, mov = 14, vmov = 1)
code_native(io,test411b,(A4,Float32)); display_asm_stat_io(io) # (total = 26, movs = 17, mov = 14, vmov = 3)
code_native(io,test412b,(A4,B4)) ; display_asm_stat_io(io) # (total = 24, movs = 8, mov = 8, vmov = 0)
code_native(io,test413b,(A4,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test414b,(A4,A4)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
# template for the next try, currently the same as for case 4
# struct G5; x :: Float32; end # struct G5 { const float x; };
# struct F5; x :: Float32; g :: G5; end # struct F5 { const float x; const G5 g; };
# struct E5; x :: Float32; f :: F5; end # struct E5 { const float x; const F5 g; };
# mutable struct D5; x :: Float32; e :: Base.RefValue{E5}; end # struct D5 { float x; E5& g; };
# struct C5; x :: Float32; d :: D5; end # struct C5 { const float x; const D5& g; };
# struct B5; x :: Float32; c :: C5; end # struct B5 { const float x; const C5& g; };
# mutable struct A5; x :: Float32; b :: B5; end # struct A5 { float x; B5& g; };
#
# g5 = G5(0.0f0)
# f5 = F5(0.0f0,g5)
# e5 = E5(0.0f0,f5)
# d5 = D5(0.0f0,Ref(e5))
# c5 = C5(0.0f0,d5)
# b5 = B5(0.0f0,c5)
# a5 = A5(0.0f0,b5)
#
# test501(a::A5,v::Float32) = @mem a.b.c.d.e[].f.g.x = v
# test502(a::A5,v::G5) = @mem a.b.c.d.e[].f.g = v
# test503(a::A5,v::Float32) = @mem a.b.c.d.e[].f.x = v
# test504(a::A5,v::F5) = @mem a.b.c.d.e[].f = v
# test505(a::A5,v::Float32) = @mem a.b.c.d.e[].x = v
# test506(a::A5,v::E5) = @mem a.b.c.d.e[] = v
# test507(a::A5,v::Float32) = @mem a.b.c.d.x = v
# test508(a::A5,v::D5) = @mem a.b.c.d = v
# test509(a::A5,v::Float32) = @mem a.b.c.x = v
# test510(a::A5,v::C5) = @mem a.b.c = v
# test511(a::A5,v::Float32) = @mem a.b.x = v
# test512(a::A5,v::B5) = @mem a.b = v
# test513(a::A5,v::Float32) = @mem a.x = v
# test514(a::A5,v::A5) = @mem a = v
#
# v = 0f0
# v += 1f0; @assert a5.b.c.d.e[].f.g.x != v; test501(a5,v); @assert a5.b.c.d.e[].f.g.x == v
# v += 1f0; @assert a5.b.c.d.e[].f.g != G5(v); test502(a5,G5(v)); @assert a5.b.c.d.e[].f.g == G5(v)
# v += 1f0; @assert a5.b.c.d.e[].f.x != v; test503(a5,v); @assert a5.b.c.d.e[].f.x == v
# v += 1f0; @assert a5.b.c.d.e[].f != F5(v,g5); test504(a5,F5(v,g5)); @assert a5.b.c.d.e[].f == F5(v,g5)
# v += 1f0; @assert a5.b.c.d.e[].x != v; test505(a5,v); @assert a5.b.c.d.e[].x == v
# v += 1f0; @assert a5.b.c.d.e[] != E5(v,f5); test506(a5,E5(v,f5)); @assert a5.b.c.d.e[] == E5(v,f5)
# v += 1f0; @assert a5.b.c.d.x != v; test507(a5,v); @assert a5.b.c.d.x == v
# try; global v += 1f0; @assert a5.b.c.d != D5(v,Ref(e5)); test508(a5,D5(v,Ref(e5))); @assert a5.b.c.d == D5(v,Ref(e5)); catch e; print(e); end
# try; global v += 1f0; @assert a5.b.c.x != v; test509(a5,v); @assert a5.b.c.x == v ; catch e; print(e); end
# try; global v += 1f0; @assert a5.b.c != C5(v,d5); test510(a5,C5(v,d5)); @assert a5.b.c == C5(v,d5) ; catch e; print(e); end
# try; global v += 1f0; @assert a5.b.x != v; test511(a5,v); @assert a5.b.x == v ; catch e; print(e); end
# v += 1f0; @assert a5.b != B5(v,c5); test512(a5,B5(v,c5)); @assert a5.b == B5(v,c5)
# v += 1f0; @assert a5.x != v; test513(a5,v); @assert a5.x == v
# v += 1f0; @assert a5 != A5(v,b5); test514(a5,A5(v,b5)); # @assert a5 == A5(v,b5) # call by "sharing"
#
# code_native(io,test501,(A5,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 5, vmov = 1)
# code_native(io,test502,(A5,G5)) ; display_asm_stat_io(io) # (total = 8, movs = 7, mov = 6, vmov = 1)
# code_native(io,test503,(A5,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 5, vmov = 1)
# code_native(io,test504,(A5,F5)) ; display_asm_stat_io(io) # (total = 9, movs = 8, mov = 6, vmov = 2)
# code_native(io,test505,(A5,Float32)); display_asm_stat_io(io) # (total = 6, movs = 5, mov = 5, vmov = 1)
# code_native(io,test506,(A5,E5)) ; display_asm_stat_io(io) # (total = 15, movs = 13, mov = 13, vmov = 0)
# # code_native(io,test507,(A5,Float32)); display_asm_stat_io(io)
# # code_native(io,test508,(A5,D5)) ; display_asm_stat_io(io)
# # code_native(io,test509,(A5,Float32)); display_asm_stat_io(io)
# # code_native(io,test510,(A5,C5)) ; display_asm_stat_io(io)
# code_native(io,test511,(A5,Float32)); display_asm_stat_io(io) # (total = 25, movs = 16, mov = 13, vmov = 3)
# code_native(io,test512,(A5,B5)) ; display_asm_stat_io(io) # (total = 25, movs = 8, mov = 8, vmov = 0)
# code_native(io,test513,(A5,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
# code_native(io,test514,(A5,A5)) ; display_asm_stat_io(io) # (total = 3, movs = 2, mov = 2, vmov = 0)
################################################################################
# @macroexpand @mem a.b.c.d.e.f.g.x = v
# a,v = (a1,v)
#
# @macroexpand @mem a.b.c.d = v
# a,v = (a2,D2(v,e2))
#
# @macroexpand @mem a.b.c.d.e.f.g.x = v
# a,v = (a2,v)
#
# @macroexpand @mem a[].b.c.d.e.f.g.x = v
# a,v = (a3r,v)
#
# @macroexpand @mem a[] = v
# a,v = (a3r,A3(v,b3))
#
# @macroexpand @mem a.b = v
# a,v = (a2,B2(v,c2))
#
# @macroexpand @mem a[][][].b.c.d.e.f.g.x = v
# a,v = (a3rrr,v)
#
# @macroexpand @mem a.b.c.d.e[].f.g.x = v
# a,v = (a4,v)
#
# @macroexpand @mem a.b.c.d = v
# a,v = (a2,D2(v,e2))
#
# @macroexpand @mem a.b.c.d = v
# a,v = (a4,D4(v,Ref(e4)))
#
# @macroexpand @mem a[].b.c.d.e.f.g.x = v
# a,v = (a3r,v)
using StaticArrays
m = SArray{Tuple{4,3,2},Int64,3,4*3*2}(1:4*3*2)
r = Ref(m)
@assert r[][1,1,2] == 13
@mem r[][1,1,2] = 77
@assert r[][1,1,2] == 77
@assert r[][1,1,2] == unsafe_load(@ptr r[][1,1,2])
m = Ref(SVector{8,UInt8}(reverse([0xde, 0xad, 0xbe, 0xef, 0xba, 0xad, 0xf0, 0x0d])))
unsafe_load(@ptr m[][8]) # 0xde
unsafe_load(@ptr m[][7]) # 0xad
unsafe_load(@typedptr UInt32 m[][3]) # 0xbeefbaad
@assert pointer_from_objref(r) == @voidptr r[][1,1,1]
@voidptr r[][1,1,1]
@ptr r[][1,1,2]
const B6 = SArray{Tuple{4,3,2},Float32,3,4*3*2}
const A6 = Base.RefValue{B6}
a = Ref(B6(1:4*3*2))
test601(a :: A6, v :: Float32) = @mem a[][1,1,2] = v
v = 77f0
@assert a[][1,1,2] != v; test601(a,v); @assert a[][1,1,2] == v;
code_native(io,test601,(A6,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
const B6b = SArray{Tuple{4,3,2},Float32,3,4*3*2}
mutable struct A6b
b :: B6b
end
b = B6b(1:4*3*2)
a = A6b(b)
test601b(a :: A6b, v :: Float32) = @mem a.b[1,1,2] = v
test602b(a :: A6b, x :: Int64, y :: Int64, z :: Int64, v :: Float32) = @mem a.b[x,y,z] = v
v = 77f0
x, y, z = 2, 1, 2
@assert a.b[1,1,2] != v; test601b(a,v); @assert a.b[1,1,2] == v;
@assert a.b[x,y,z] != v; test602b(a,x,y,z,v); @assert a.b[x,y,z] == v;
code_native(io,test601b,(A6b,Float32)); display_asm_stat_io(io) # (total = 2, movs = 1, mov = 0, vmov = 1)
code_native(io,test602b,(A6b,Int64,Int64,Int64,Float32)); display_asm_stat_io(io) # (total = 5, movs = 1, mov = 0, vmov = 1)
# leaq (%rcx,%rcx,2), %rax
# leaq (%rsi,%rdx,4), %rcx
# leaq (%rcx,%rax,4), %rax
# vmovss %xmm0, -68(%rdi,%rax,4)
# TODO: a test of a non-isbitstype SArray; since it allocates an intermediate array and needs double dereferencing
# same for MArray