/
ChangeLog
3049 lines (1839 loc) · 96.5 KB
/
ChangeLog
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
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
Fri Jun 25 17:36:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* driver.c: remove loop from the default optimizations: it seems to
interact badly with some of the other options (see bug #60613).
2004-06-25 Zoltan Varga <vargaz@freemail.hu>
* mini.c mini-x86.h mini-x86.c: Applied patch from Guenter Feldmann
(fld@informatik.uni-bremen.de): Add Solaris x86 support.
Tue Jun 22 21:29:11 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, cpu-g4.md: small updates to be able to compile
vararg-using methods.
2004-06-21 Martin Baulig <martin@ximian.com>
* mini/mini-exceptions.c
(mono_handle_exception): Added `gpointer original_ip' argument.
After calling mono_unhandled_exception(), call
mono_debugger_unhandled_exception() and if that returns true,
restore the context and return.
Mon Jun 21 19:26:40 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c: prefer the use of relative branches so
they won't need to be patched in aot code (patch from Patrick Beard).
Mon Jun 21 19:03:18 CEST 2004 Paolo Molaro <lupus@ximian.com>
* aot.c: patch from Patrick Beard to make the output assembly
more correct for the MacOSX assembler. Small tweak to
generate sane images on Linux/PPC, too.
Fri Jun 18 18:24:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c, mini.h, mini-ppc.c: handle varargs methods with a special
case until bug #59509 is fixed (shows up in #60332).
Tue Jun 15 16:36:51 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c: make sure the needed wrappers are compiled, too, with
precomp.
Mon Jun 14 18:36:08 CEST 2004 Paolo Molaro <lupus@ximian.com>
* driver.c: remove NPTL reference in --version output.
Sun Jun 13 17:25:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* aot.c: patch from Patrick Beard (pcbeard@mac.com) to
generate valid assembly for the Mach-O assembler.
Sun Jun 13 15:59:38 CEST 2004 Paolo Molaro <lupus@ximian.com>
* driver.c: don't include abcrem in the all optimization specifier
since it slows down jit compilation too much for now.
2004-06-12 Ben Maurer <bmaurer@users.sourceforge.net>
* mini.c: use BIGMUL only if both operands have the same signage.
* iltests.il: Test for bug 60056. (errors related to signage in
BIGMUL).
r=lupus.
Thu Jun 10 16:06:42 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c, aot.c: memory leak fixes.
Tue Jun 8 16:37:15 CEST 2004 Paolo Molaro <lupus@ximian.com>
* inssel-long32.brg: implemented a few missing ulong cast opcodes.
Tue Jun 8 15:36:30 CEST 2004 Paolo Molaro <lupus@ximian.com>
* Makefile.am: remove the -static hack completely, it links in
statically glib as well.
Sat Jun 5 16:32:33 CEST 2004 Paolo Molaro <lupus@ximian.com>
* iltests.il, mini.c: fixed bug#59580 in branch optimization.
* exceptions.cs: make it compile with new mcs/csc.
2004-06-03 Massimiliano Mantione <massi@ximian.com>
* cpu-pentium.md basic-float.cs Fixed bug on fpu spills (see bug 54467),
and added relevant test case.
Mon May 31 19:41:46 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c revert Zoltan's fix to bug#58863 on ppc, since it causes
regressions in gtk-sharp.
2004-05-29 Zoltan Varga <vargaz@freemail.hu>
* exceptions.cs: New regression tests.
* jit-icalls.c (mono_llmult_ovf): Fix some boundary conditions.
Sat May 29 10:45:58 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c: emit castclass/isinst in their own trees (bug #54209/59057).
2004-05-28 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.h (MONO_ARCH_NEED_DIV_CHECK): Define this.
* cpu-sparc.md mini-sparc.c: Add overflow detection to div opcodes.
2004-05-28 Patrik Torstensson <totte@hiddenpeaks.com>
* mini.c (mono_jit_runtime_invoke): Init class in this
method instead of trusting mono_jit_compile_method to
do the work (because wrappers can be in object class)
2004-05-27 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c (mono_arch_output_basic_block): Fix sub.imm.
* basic-long.cs: New regression test.
Thu May 27 15:50:52 CEST 2004 Paolo Molaro <lupus@ximian.com>
* cpu-g4.md, mini-ppc.c: fixes to long add/sub ovf opcodes
and div/rem checks.
Thu May 27 12:36:53 CEST 2004 Paolo Molaro <lupus@ximian.com>
* Makefile.am: fix miguel's change to build mono statically against
libmono (track build dependencies).
2004-05-26 Zoltan Varga <vargaz@freemail.hu>
* cfold.c: Some glib versions do not have G_MININT32.
2004-05-26 Massimiliano Mantione <massi@ximian.com>
* mini-x86.c cpu-pentium.md Makefile.am basic-math.cs: Fixed problem
with precision of tan, atan, sin and cos, and implemented related
regressions tests (fixes bug 54467, but one new problem appeared and
is not fixed yet).
2004-05-26 Zoltan Varga <vargaz@freemail.hu>
* cfold.c (FOLD_BINOPZ): Avoid division by zero.
* exceptions.cs: Add test for constant folding && division by zero.
* driver.c mini.h mini.c mini-x86.c: Revert most of the previous patch
since driver.c is in libmono too, so the optimization was useless.
* driver.c mini.h mini.c mini-x86.c: Moved the mono_lmf_addr TLS
variable to driver.c so the compiler can emit more efficient code to
access them.
2004-05-26 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* Makefile.am: don't distribute generated inssel.[ch] files.
2004-05-25 Zoltan Varga <vargaz@freemail.hu>
* mini.c (mono_jit_compile_method_with_opt): Really emit icall wrappers
into the default appdomain. Fixes #58707.
* jit-icalls.c: Remove the broken approximation for truncl, doing
no conversion is better.
* mini.c (handle_stack_args): Avoid reusing variables for stack slots.
Fixes #58863.
Tue May 25 14:33:56 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ops.h, mini-ppc.c, cpu-g4.md, inssel-ppc.brg: eliminate the use
of the mcrxr instruction which is not available on some processors
even if it's documented to be. Implement add and sub overflow correctly
(still not complete for long unsigned). Speed up icalls a bit.
2004-05-25 13:01 CET Patrik Torstenson <totte@hiddenpeaks.com>
* mini.c (mono_jit_compile_method_with_opt): Make sure that
we run .cctor in the current domain instead of target_domain.
Fixes bug #58558, .cctor not being called in -O=shared.
Tue May 25 12:46:50 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.h, jit-icalls.c: added explicit checks for divide by zero.
2004-05-24 Ben Maurer <bmaurer@users.sourceforge.net>
* mini-x86.c (EMIT_COND_BRANCH): If an OP_LABEL has an offset
which can be done with an imm8, do it that way.
(mono_arch_output_basic_block): ditto for a jmp
(mono_arch_emit_prolog): Computate maximum offset of a label.
2004-05-24 18:18 CET Patrik Torstenson <totte@hiddenpeaks.com>
* mini-x86.c (mono_arch_local_regalloc): the reg allocator
now tries to allocate prefered physical reg's for virtual
regs. This reduces the number of emited spills/loads with
20-30% on our core assemblies.
Mon May 24 18:21:51 CEST 2004 Paolo Molaro <lupus@ximian.com>
* jit-icalls.c: truncl() is not needed and trunc() is
the correct thing to do anyway (bug #58287).
2004-05-24 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c (mono_arch_flush_icache): Call sync_instruction_memory
if available.
Mon May 24 12:49:45 CEST 2004 Paolo Molaro <lupus@ximian.com>
* driver.c: enable loop optimizations by default.
Mon May 24 11:13:46 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-x86.c: fix calc of max loop size when aligning loops start.
2004-05-23 Zoltan Varga <vargaz@freemail.hu>
* ssa.c (mono_ssa_cprop): Allocate carray dynamically instead of on
the stack.
2004-05-22 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c (mono_arch_output_basic_block): ADD_IMM and SUB_IMM
should set carry.
* basic-long.cs: Add tests for add/subtract of immediates with carry.
* mini.c exceptions-x86.c: Remove MONO_USE_EXC_TABLES stuff.
* mini.c (inline_method): Allways inline some wrappers even if the cost
is too large. Fixes #58785.
* mini.c: Add support for MARSHAL_CONV_FTN_DEL.
2004-05-21 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c exceptions-sparc.c: Applied patch from Mark Crichton
(crichton@gimp.org). Beginning of support for sparc/linux.
* mini-sparc.c: Optimize retrieval of LMF address.
Fri May 21 08:00:12 EDT 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: handle alloca in methods with clauses.
Fri May 21 07:35:30 EDT 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c: cleanups, off-by-one fixes, avoid recursive thunks.
2004-05-20 Lluis Sanchez Gual <lluis@ximian.com>
* mini.c: Delegate most of the abort signal work to
mono_thread_request_interruption, which also handles Stop and Suspend
states.
2004-05-20 Zoltan Varga <vargaz@freemail.hu>
* mini.c mini.h: Allow inlining of icall wrappers if the backend
supports the save/restore lmf opcodes.
2004-05-19 Zoltan Varga <vargaz@freemail.hu>
* mini-x86.c (mono_arch_setup_jit_tls_data): Handle code generated
by gcc-3.4 as well.
* mini-x86.h mini-x86.c tramp-x86.c: Optimize lmf restoring code.
2004-05-18 Zoltan Varga <vargaz@freemail.hu>
* mini.h mini.c (mini_method_compile): Only run abc removal pass on
methods which contain array accesses.
* mini.c (CEE_LDTOKEN): Handle this instruction correctly on bb
boundaries. Fixes #58537.
* iltests.il: Add regression test for #58537.
2004-05-18 Patrik Torstensson <totte@hiddenpeaks.com>
* mini-x86.c (mono_arch_local_regalloc): Last part of
fix for bug #58633 (releasing register to early).
2004-05-18 Miguel de Icaza <miguel@ximian.com>
* basic-long.cs: Add new regression test.
2004-05-18 Patrik Torstensson <totte@hiddenpeaks.com>
* mini-x86.c (mono_arch_local_regalloc): Avoid releasing a
register too early on the chain.
2004-05-18 Zoltan Varga <vargaz@freemail.hu>
* mini.c (create_helper_signature): Use a helper function to reduce
the code which needs to be written. Also set the calling convention of
icalls on windows. Fixes #57840.
Tue May 18 11:05:18 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.h, exceptions-x86.c, exceptions-sparc.c, exceptions-s390.c,
exceptions-ppc.c: added helper function to get the instruction address
from a signal handler context.
2004-05-17 Ben Maurer <bmaurer@users.sourceforge.net>
* helpers.c: use g_get_tmp_dir. Invokes happyness
from gonzalo.
2004-05-17 Ben Maurer <bmaurer@users.sourceforge.net>
* helpers.c: Add new env variable to pass args to objdump.
Specifically for those of us who love -Mintel. r=miguel, gonzalo.
2004-05-17 Radek Doulik <rodo@ximian.com>
* Makefile.am (common_sources): added abcremoval.h so it get
disted and daily mono packages on go-mono.com will build again
2004-05-17 Massimiliano Mantione <massi@ximian.com>
* abcremoval.c: Fixed coding style, added copyright header.
* abcremoval.h: Fixed style and moved prototype to mini.h, added copyright header.
* mini.h: Added prototype for abc removal main function.
* build_relations_propagation_table.pl: Added copyright header.
2004-05-16 Patrik Torstensson <totte@hiddenpeaks.com>
* basic-long.cs: reg test for complex ceq_long bug.
2004-05-16 Patrik Torstensson <totte@hiddenpeaks.com>
* mini-x86.c (mono_arch_local_regalloc): Correctly free
reg in long and clob case (bug #58343). Fixed/added comments.
2004-05-14 Ben Maurer <bmaurer@users.sourceforge.net>
* mini.c (mono_jit_runtime_invoke): Follow new convention
of calling the invoke method with an function pointer.
2004-05-14 Zoltan Varga <vargaz@freemail.hu>
* ChangeLog: Fix author of memory leak patch.
Fri May 14 15:13:06 CEST 2004 Paolo Molaro <lupus@ximian.com>
* Makefile.am: fix make dist as well...
2004-05-14 Massimiliano Mantione <massi@ximian.com>
* cfold.c: Made so that conversions from pointer to int4 are no-ops
on archs where pointers are 4 bytes long.
* Makefile.am: Added abcremoval.c source file.
* abcremoval.c: Added abcremoval.c.
* abcremoval.h: Added abcremoval.h.
* build_relations_propagation_table.pl: Added build_relations_propagation_table.pl.
* inssel.brg: Enabled bounds check removal.
* mini.c: Added support for abcrem optimization.
* mini.h: Added abcrem optimization label.
* driver.c: Added support for abcrem optimization.
* propagated_relations_table.def: Added propagated_relations_table.def.
Fri May 14 14:30:13 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c, cfold.c: fix style.
Fri May 14 14:28:22 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c: handle issue with the low-level implementation of
some long opcodes (bug #54209).
2004-05-13 Ben Maurer <bmaurer@users.sourceforge.net>
* basic.cs: test for my new cmov stuff.
2004-05-13 Patrik Torstensson
* mini-x86.c: added OP_X86_COMPARE_MEMBASE_IMM peephole
opt and added peephole documentation.
Thu May 13 11:41:49 CEST 2004 Paolo Molaro <lupus@ximian.com>
* tramp-ppc.c: rewrote the generic trampoline code.
2004-05-11 Patrik Torstensson
* mini-x86.c: optimize long shl/shr asm code (one less branch)
2004-05-11 Zoltan Varga <vargaz@freemail.hu>
* basic.cs basic-long.cs objects.cs: Make these compile under MS csc.
* mini.h mini.c dominators.c: Applied patch from Derek Woo
(derek@eecg.toronto.edu): Fix memory leaks in loop optimizations.
* mini.c: Add new icalls for AsAny marshalling.
Tue May 11 16:00:38 CEST 2004 Paolo Molaro <lupus@ximian.com>
* tramp-ppc.c, mini-ppc.c: more cleanups.
2004-05-11 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini.c: no warnings.
Tue May 11 13:59:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, mini.c: use mono_resolve_patch_target ().
2004-05-11 Lluis Sanchez Gual <lluis@ximian.com>
* mini.c: In the thread abort signal handler, if the thread is in the
process of being stoped, don't throw the Abort exception, just stop the
thread.
Tue May 11 12:15:24 CEST 2004 Paolo Molaro <lupus@ximian.com>
* tramp-ppc.c: removed old code.
Tue May 11 12:02:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.h, mini-ppc.c, cfold.c: export mono_is_power_of_two().
do some simple speed optimizations on ppc.
Mon May 10 17:21:00 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, cpu-g4.md: fixes to handle large stack frames
and large offsets in load/store.
2004-05-07 Zoltan Varga <vargaz@freemail.hu>
* mini-x86.c (mono_arch_local_regalloc): Remove the previous fix, since
it causes regressions.
2004-05-07 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c: Fix vararg support + add beginnings of sigaltstack
support.
Fri May 7 13:25:50 CEST 2004 Paolo Molaro <lupus@ximian.com>
* jit-icalls.c: remove warnings.
* inssel-x86.brg, inssel.brg, mini-x86.c, cfold.c: some simple
speedups for unsafe code.
2004-05-07 Ben Maurer <bmaurer@users.sourceforge.net>
* inssel.brg: Optimize Stind.[ui][12]. r=zoltan.
2004-05-06 Zoltan Varga <vargaz@freemail.hu>
* basic-calls.cs: Add new regression test.
* mini.c (mono_runtime_install_handlers): Use SA_ONSTACK since it is
more portable.
* mini.c (mono_method_to_ir): Handle opcode emulation for *_OVF opcodes.
* mini-x86.c (mono_arch_local_regalloc): Free new_dest register when it
is no longer used.
2004-05-06 Patrik Torstensson
* mini-x86.[c|h], inssel-x86.brg, cpu-pentium.md, mini.c: enabled
long reg allocation in any reg (not only eax:edx) and implemented
long shl/shr ops in asm instead of helpers.
2004-05-05 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.h: Fix warnings.
* exceptions-sparc.c (mono_arch_find_jit_info): Pop unused lfm off the
stack.
* mini-exceptions.c (mono_handle_exception): Call the filter in a
separate statement for clarity.
* mini-sparc.c: Update status.
2004-05-04 Zoltan Varga <vargaz@freemail.hu>
* mini-exceptions.c (ves_icall_get_frame_info): Flush register windows
here.
Mon May 3 22:58:51 CEST 2004 Paolo Molaro <lupus@ximian.com>
* inssel-ppc.brg: another small pre-release workaround:
we don't do overflow detection for long_sub_un.
Sun May 2 20:12:22 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c, mini-ops, inssel-long32.brg: speedup ulong >> 32
(also works around a weird ppc bug: 57957).
Sat May 1 16:56:10 EDT 2004 Paolo Molaro <lupus@ximian.com>
* tramp-ppc.c: trampoline fixes.
Fri Apr 30 15:54:26 EDT 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c: fixed typos.
Thu Apr 29 20:15:41 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, exceptions-ppc.c: more code saves registers
at the top of the stack. Fixed typos. Use a frame registers
for all the methods with exception clauses.
Thu Apr 29 18:52:09 CEST 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: restore fp registers.
Thu Apr 29 18:26:56 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, exceptions-ppc.c: save the registers in reverse
order from the stack top (moved the stack room to save the
return value for trace after the param area). Fixed corruption
in restoring registers on unwind.
Thu Apr 29 16:47:15 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, cpu-g4.md: fixed unisgned -> double conversion.
Thu Apr 29 13:50:51 CEST 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c, mini-ppc.h, mini-ppc.c: fixed localloc
and prolog/epilog for methods that use it. Allow
enough param area room for varargs methods. Fix miguel's
breakage in exception handling.
Thu Apr 29 12:06:51 CEST 2004 Paolo Molaro <lupus@ximian.com>
* Makefile.am: run genmdesc only on current arch.
2004-04-29 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* exceptions-x86.c:
* mini-x86.h: fix the build on windows.
2004-04-28 Zoltan Varga <vargaz@freemail.hu>
* Makefile.am mini.h mini-exceptions.c mini-x86.h mini-sparc.h exceptions-sparc.c: Move parts of the sparc exception handling code to XP code.
* exceptions-sparc.c exceptions-ppc.c exceptions-s390.c mini-ppc.h mini-s390.h mini-sparc.h: Fix up ports after changes.
* mini-exceptions.c: New file.
* mini.c mini-exceptions.c mini-x86.h exceptions-x86.c Makefile.am:
Move some parts of the x86 exception handling code to an
arch-independent file so it can be shared with other ports.
Tue Apr 27 12:15:59 CEST 2004 Paolo Molaro <lupus@ximian.com>
* trace.c, cpu-g4.md, inssel-ppc.brg, mini-ppc.c: some fixes for FP code.
2004-04-26 David Waite <mass@akuma.org>
* driver.c: remove comma from end of enumeration declaration
2004-04-26 Jackson Harper <jackson@ximian.com>
* driver.c: parse config file before loading first assembly. This
allows the user gac to be enabled/disabled.
2004-04-23 Miguel de Icaza <miguel@ximian.com>
* mini-ppc.c (ppc_patch): Replaced the branch code patching with a
simpler mechanism: we do not care what is encoded initially
(branch absolute or relative), we care about the code and its
target. I kept the old code for reference for now.
The new code tries first to determine if the jump is anywhere in
the -/+32 absolute meg range, if it succeeds, it encodes using the
absolute branch; If not, it tried to find something in the
relative range, if not, it uses the handle_thunk code.
Fri Apr 23 14:20:40 EDT 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: use the correct ip register on macosx.
Thu Apr 22 13:23:16 EDT 2004 Paolo Molaro <lupus@ximian.com>
* exceptions.c, mini.c, mini-ppc.h: adapt code to macosx.
Thu Apr 22 18:08:37 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, cpu-g4.md: made the branch macros more flexible.
Raise exception on integer divide by zero by hand since the hw
doesn't support it. Handle NaNs in FP compares.
Thu Apr 22 16:10:31 CEST 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c, mini-ppc.h, mini.c: simplified some
code reducing duplication between the platforms and enabled
signal exception handling (on linux for now).
Wed Apr 21 12:27:48 EDT 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: more macosx support.
Wed Apr 21 16:38:28 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.h, mini-ppc.c, cpu-g4.md: enable bigmul optimization.
Wed Apr 21 16:20:32 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.h, mini-ppc.c, cpu-g4.md: support for implicit exceptions.
2004-04-19 Ben Maurer <bmaurer@users.sourceforge.net>
* iltests.il: more tests.
2004-04-19 Zoltan Varga <vargaz@freemail.hu>
* mini-*.c (mono_arch_get_allocatable_int_vars): Skip written-only
vars as well.
Mon Apr 19 19:39:47 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c: some fixes to bootstrap mcs/corlib/etc.
2004-04-19 Zoltan Varga <vargaz@freemail.hu>
* liveness.c: Mark variables as volatile in all basic blocks reachable
from exception clauses.
2004-04-18 Zoltan Varga <vargaz@freemail.hu>
* exceptions.cs (test_0_rethow_stacktrace): Make this work with
inlining.
2004-04-18 Ben Maurer <bmaurer@users.sourceforge.net>
* iltests.il, basic.cs: more tests for regalloc.
2004-04-17 Ben Maurer <bmaurer@users.sourceforge.net>
* iltests.il: Some tests for register allocation modifications
I have locally.
2004-04-16 Zoltan Varga <vargaz@freemail.hu>
* exceptions.cs: Add regression test for bug #56782.
* exceptions-*.c (mono_arch_handle_exception): Do not overwrite the
original stack trace if an exception is rethrown. Fixes #56782. Oh,
the beauty of fixing the same thing in 5 different files...
2004-04-15 Zoltan Varga <vargaz@freemail.hu>
* mini.c (mono_method_to_ir): Do not compute coverage for inlined
methods.
2004-04-14 Zoltan Varga <vargaz@freemail.hu>
* mini.c: Add support for STRWLPARRAY marshalling convention.
Wed Apr 14 18:15:55 CEST 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: missing fixes in mono_jit_walk_stack (need
to init the context to setup the regs pointer).
Wed Apr 14 17:59:09 CEST 2004 Paolo Molaro <lupus@ximian.com>
* exceptions-ppc.c: more exceptions work.
Wed Apr 14 17:46:22 CEST 2004 Paolo Molaro <lupus@ximian.com>
* mini.c: avoid reusing the same MonoInst on multiple trees: this is
not allowed.
2004-04-13 Miguel de Icaza <miguel@ximian.com>
* inssel-x86.brg (reg): Add new rules for add, sub and mul that
can use the memory directly.
* cpu-pentium.md: Update documentation from a post from Zoltan.
add x86_add_membase, x86_sub_membase, x86_mul_membase
2004-04-13 Miguel de Icaza <miguel@ximian.com>
* mini-ppc.c: Remove unused definitions FLOAT_REGS and
GENERAL_REGS they were also hardcoded for all PPC ports.
(add_general): Use PPC_NUM_REG_ARGS instead of GENERAL_REGS.
Remove hard-coded limit for floating point registers, use
PPC_LAST_FPARG_REG instead in MONO_TYPE_R4 and MONO_TYPE_R8.
Notice that in MacOS X calling conventions you can fit a lot more
floating point values in registers, so I should update the PInvoke
test to excercise the passing of floating point values on the
stack (currently broken).
2004-04-06 Miguel de Icaza <miguel@ximian.com>
* tramp-ppc.c (create_trampoline_code): Added
JUMP_TRAMPOLINE_SIZE.
(ppc_magic_trampoline): Follow the pattern from
x86_magic_trampoline: if code is set to zero, return.
(create_trampoline_code): Always pass MonoMethod to the jump
trampoline, before it was passing a null.
(mono_arch_create_jump_trampoline): Implement the jump stub, could
share the code with mono_arch_create_jit_trampoline.
* mini-ppc.c (mono_arch_output_basic_block): CEE_JMP opcode
implemented.
(mono_arch_patch_code): MONO_PATCH_INFO_METHOD_JUMP patch type
implemented.
* cpu-g4.md: Added length for jmp instruction, the worst case
scenario is 92 bytes (4 mandatory bytes, potential 19 registers
for save_lmf).
2004-04-08 Zoltan Varga <vargaz@freemail.hu>
* aot.c (mono_compile_assembly): Add back unlink removed by mistake.
2004-04-07 Zoltan Varga <vargaz@freemail.hu>
* mini.c: Only set bblock->real_offset when adding a new bblock, and
before each IL instruction.
* mini.c (CEE_BOX): Fix warnings.
2004-04-07 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini.c: removed a few unused vars and extra whitespace.
2004-04-05 Ben Maurer <bmaurer@users.sourceforge.net>
* inssel.brg (MONO_EMIT_BOUNDS_CHECK): a new macro to emit bounds
checks.
(MONO_EMIT_BOUNDS_CHECK_IMM): the above, but when you know the
index.
(OP_GETCHR): use the above
(CEE_LDELEMA): use the above.
* inseel-x86.brg (MONO_EMIT_BOUNDS_CHECK): a faster and smaller
version of the generic impl.
(MONO_EMIT_BOUNDS_CHECK_IMM): the same
(CEE_LDELEMA): use the above.
2004-04-05 Zoltan Varga <vargaz@freemail.hu>
* inssel-long32.brg (CEE_CONV_OVF_I8): Sign extend the i4 value to i8.
Fixes #56317.
* iltests.il: Added new regression test for #56317.
2004-04-05 Zoltan Varga <vargaz@freemail.hu>
* mini-x86.c (mono_arch_setup_jit_tls_data): Use pthread_attr_get_np
under NetBSD. Fixes #56450.
* liveness.c (update_gen_kill_set): Fix previous patch.
2004-04-04 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini-x86.h: SA_STACK defined as SA_ONSTACK. Fixed build under NetBSD.
2004-04-02 Zoltan Varga <vargaz@freemail.hu>
* mini.c (mono_method_to_ir): Avoid handle_loaded_temps in
ldsfld and ldsflda.
* inssel-sparc.brg: Add more optimizations.
* mini-sparc.c: Replace multiply/divide with shifts if possible.
2004-04-01 Martin Baulig <martin@ximian.com>
* mini.c (handle_box): New static function; moved the
implementation of CEE_BOX here.
(mono_method_to_ir): Added `constrained_call' variable.
(mono_method_to_ir:CEE_CONSTRAINED_): Set it.
(mono_method_to_ir:CEE_CALL): If `constrained_call' is set, use
mono_method_get_constrained() to get the method.
2004-04-01 Martin Baulig <martin@ximian.com>
* mini.c (TYPE_PARAM_TO_TYPE, TYPE_PARAM_TO_CLASS): Removed.
(MTYPE_PARAM_TO_TYPE, MTYPE_PARAM_TO_CLASS): Removed.
(mono_method_to_ir): We don't need these macros anymore since
mono_class_get_full() already takes care of it.
2004-03-31 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* aot.c: set aot_verbose to 0, fixed format string that caused sigsegv,
use @function (as doesn't accept #function here) and check the return
value of system and stop if fails.
2004-03-31 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini.c: set the timeout to 2s when calling mono_domain_finalize.
2004-03-31 Zoltan Varga <vargaz@freemail.hu>
* mini-ppc.c (mono_arch_patch_code): Fix ppc build.
* inssel-sparc.brg mini-sparc.c aot.c: Implement AOT support.
* inssel-long32.brg (OP_LNEG): Use ADC instead of ADD here. Fixes
#56223.
* basic-long.cs: Add test for negation of Int64.MinValue.
2004-03-30 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c: Update status.
* mini-sparc.c tramp-sparc.c: Save lmf in trampolines.
* exceptions-sparc.c: Fix return value in filters.
* inssel-sparc.brg: Fix register allocation in some rules.
2004-03-28 Martin Baulig <martin@ximian.com>
* mini.c (mmono_method_to_ir): In CEE_STELEM, do a handle_stobj()
if neccessary.
2004-03-28 Zoltan Varga <vargaz@freemail.hu>
* mini-x86.c (mono_arch_patch_code): Fix warnings.
* mini-x86.c (mono_arch_output_basic_block): Fix CEE_MUL_OVF_UN if
dreg is not EAX. Thanks to Willibard Krenn for spotting this. Also
remove unused conv_u4 opcode.
* mini-x86.c: Remove valgrind workaround since it slows down things
even when mono is not run under valgrind.
2004-03-26 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c: Update status.
* inssel-sparc.brg: Add some optimizations.
* inssel-sparc.brg mini-sparc.c: Rework branch instructions to allow
future delay slot filling. Add support for varargs, tail calls and JMP.
* inssel.brg mini-ops.h mini.c: Use OP_REFANYTYPE instead of
CEE_REFANYTYPE, since CEE_REFANYTYPE needs a prefix to be unique.
* inssel.brg: Fix register allocation in OP_ARGLIST.
* inssel.brg: Fix warnings.
2004-03-25 Martin Baulig <martin@ximian.com>
* mini.c (inflate_generic_field): Removed.
(mini_get_method): Removed, use mono_get_method_full(),
(mini_get_class): Removed, use mono_class_get_full().
(mono_method_to_ir): Pass our generic context to
mono_field_from_token().
2004-03-25 Martin Baulig <martin@ximian.com>
* mini.c (mini_get_class): Take a `MonoGenericContext *' instead
of a `MonoMethod *'.
(mini_get_method): Take a `MonoGenericContext *' instead
of a `MonoMethod *'.
(TYPE_PARAM_TO_TYPE, MTYPE_PARAM_TO_TYPE): mono_method_to_ir() has
a new local variable called `generic_context' which holds the
current `MonoGenericContext *'; use it to lookup things.
2004-03-24 Martin Baulig <martin@ximian.com>
* mini.c (mini_get_class): New static method; if we're inside a
generic instance, inflate the class if neccessary.
(mono_method_to_ir): Use mini_get_class() instead of mono_class_get().
2004-03-24 Zoltan Varga <vargaz@freemail.hu>
* iltests.il: New regression test for #55976.
* mini.c (mono_method_to_ir): Empty the stack in endfinally. Fixes
#55976.
2004-03-23 Zoltan Varga <vargaz@freemail.hu>
* exceptions-sparc.c (mono_sparc_handle_exception): Remove debugging
output.
2004-03-23 Zoltan Varga <vargaz@freemail.hu>
* liveness.c: Consider SSA stores as well as loads when making vars
volatile.
* exceptions.cs: New regression tests for register allocation.
2004-03-22 Lluis Sanchez Gual <lluis@ximian.com>
* mini-ppc.c, tramp-ppc.c: Added lock for accessing the domain code manager.
* mini.c: Removed domain lock from mono_jit_compile_method_with_opt. Use
domain lock only to protect puntual access to data structures.
Added access lock for sighash, jit_icall_hash_name,
jit_icall_hash_addr and domain->code_mp.
2004-03-20 Zoltan Varga <vargaz@freemail.hu>
* driver.c: Print SIGSEGV handling method.
* mini-x86.c (mono_arch_free_jit_tls_data): Add missing ifdef.
* mini.c (setup_jit_tls_data): Handle case when this is called
multiple times for a thread.
* mini-x86.c cpu-pentium.md: Fix floating point branch opcodes so fbxx
is different from fbxx_un. Fixes #54303. Also use constants instead of
magic numbers in a lot of places.
2004-03-19 Zoltan Varga <vargaz@freemail.hu>
* exceptions.cs: Fix cctor test when --regression is used.
Thu Mar 18 19:57:56 CET 2004 Paolo Molaro <lupus@ximian.com>
* mini-ppc.c, exceptions-ppc.c: basic exceptions support
for Linux/ppc.
Thu Mar 18 19:56:19 CET 2004 Paolo Molaro <lupus@ximian.com>
* inssel-ppc.brg: fixed register assignments for some rules.
2004-03-17 Zoltan Varga <vargaz@freemail.hu>
* exceptions.cs: Add test for exceptions in static constructors.
* mini.c (mono_jit_compile_method_with_out): Move the calling of
static constructors outside the domain lock. Fixes #55720.
2004-03-17 Martin Baulig <martin@ximian.com>
* mini.c (get_generic_field_inst): Removed, this'll never happen.
(inflate_generic_field): Take the `MonoMethod *' instead of the
`MonoClass *' and added support for generic method.
(mono_method_to_ir): In CEE_LDSFLD and CEE_STSFLD, assert we never
have a `field->parent->gen_params', only inflate the field if it's
an open constructed type.
2004-03-17 Zoltan Varga <vargaz@freemail.hu>
* exceptions-x86.c (mono_arch_handle_exception): Allocate a new
exception object instead of the preconstructed ones.
2004-03-17 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini.c: reverted changed to sigsegv_signal_handler commited
accidentally in the previous patch.
2004-03-17 Gonzalo Paniagua Javier <gonzalo@ximian.com>
* mini.c:
(mono_method_to_ir): CEE_CALLVIRT, abort if no method. It hanged when
running --aot with an old assembly.
2004-03-16 Zoltan Varga <vargaz@freemail.hu>
* mini-sparc.c (mono_arch_instrument_epilog): Fix handling of floating
point values.
* mini-sparc.c: Add support for v9 branches with prediction.
2004-03-15 Bernie Solomon <bernard@ugsolutions.com>
* mini.c (mini_init): #warning is GNUC only
* mini-sparc.h: implement __builtin_frame_address
and __builtin_return_address for Sun C compiler
2004-03-15 Zoltan Varga <vargaz@freemail.hu>
* exceptions-sparc.c (mono_arch_has_unwind_info): Add missing function.
2004-03-14 Zoltan Varga <vargaz@freemail.hu>
* basic-calls.cs: Add test for unaligned byref long argument passing.
* mini-ops.h: Add sparcv9 compare and branch instructions.