-
Notifications
You must be signed in to change notification settings - Fork 16
Expand file tree
/
Copy pathChangeLog
More file actions
4693 lines (3328 loc) · 180 KB
/
ChangeLog
File metadata and controls
4693 lines (3328 loc) · 180 KB
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
2015-10-28 Pierangelo Masarati <pierangelo.masarati@polimi.it>
* RTAI 5.0
2015-09-24 Pierangelo Masarati <pierangelo.masarati@polimi.it>
* RTAI 4.1.1 (4.1 closure release)
2015-05-09 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/addons/rtdm/rtdm_driver.h, xn.h: support changing, deprecated
and not, smp_mb__before_atomic versions.
2015-04-18 Paul Corner <paul_c@tuxcnc.org>
* base/scripts/rtai-config.in: added datarootdir; applied and checked
by Pierangelo (Masarati).
2015-04-13 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: if CONFIG_SMP is not defined set up an
internal declaration for it.
2015-04-07 Pierangelo Masarati <pierangelo.masarati@polimi.it>
* base/include/GNUmakefile.am: make selfcalibration work also
for VPATH builds
2015-04-03 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai.h: insert dummy rt_mount_rtai and rt_umount_rtai
to make comedi making happy.
2015-03-25 Shahbaz Youssefi <shabbyx@gmail.com>
* rtai-py/rtai_lxrt.py: fixed nam2num and num2nam typos, argstype
used in place argtypes.
2015-03-22 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sys.c: revised lxrt_invoke, argument list and catching
Linux signal at entering moved to syscall interception in sched.c.
* base/sched/sched.c: revised linux signal wakeup and RTAI-Linux
syscalls intercpting.
2015-03-19 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: omit argument in macros
"...fast_flush_pipeline", not needed any more and return though linux
after intercepting syscalls, theu can never be in hard real time
without the scheduler.
* base/include/rtai_schedcore.h: make TASK_HARDREALTIME and
TASK_RTAISRVSLEEP as TASK_INTERRUPTIBLE.
* base/include/asm-x86/rtai_lxrt.h: experiment with the possibility of
* base/sched/api.c: split the function rt_task_masked_unblock in a
basic unprotected function, to be used in specifically protected
forms, both in api.c and wherever it is needed.
2015-03-17 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/msg/msg.c: improve the check of the sending task in
rt_Receive.
2015-03-06 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: type rtai_hirq_dispatcher as void, no
return value used by the function calling it.
* base/sched/sched.c: removed the unused code related to the old way
of using a list of tasks to transition to hardmode
(lxrt_intercept_schedule_tail).
2015-03-04 Pierangelo Masarati <pierangelo.masarati@polimi.it>
* configure.in, configure: revised to support self calibration and
revised configuration paraeters.
* base/sched/selfcalibrate.sh: added to support self calibration.
* base/arch/x86/Kconfig: restructured calibration parameters structure
names and help, both to clean no more used names and to support self
calibration.
* base/arch/x86/calibration/calibrate.c: modified to be usable both
for self and user guided calibrations.
2015-02-27 Pierangelo Masarati <pierangelo.masarati@polimi.it>
* configure.in, configure, testsuite/kthreads/*/GNUmakefile.*,
testsuite/kthreads/*/Makefile.kbuild: add prefix kthreads_ to
modules testsuite/kthreads, showing the use of hardened Linux
kernel threads in the same way as user space threads.
* base/arch/x86/calibration/GNUmakefile.*: module rtai_calibrate.ko
no longer built; build calibrate instead.
2015-02-22 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sys.c: avoid waiting on a Linux sem when a kthread
creation is requested from hard real; the Linux task structure
address is returned through a call argument.
2015-02-21 Paolo Mantegazza <mantegazza@aero.polimi.it>
* testsuite/kthreads/latency/display.c and latency-module.c,
testsuite/kthreads/preempt/display.c and preempt-module.c,
testsuite/kthreads/switches/switches-module.c: set of examples
showing the usage of RTAI enabled soft-hardened Linux kthreads.
* base/sched/sys.c: added a Linux kthread server, to allow creating
hadr-soft RTAI Linux kernel threads from RTAI own kernel tasks and
Linux real time hardened kernel thread, without loosing hard real
time execution; it implies a revision of what prviously committed.
* base/sched/sched.c: run and termininate, at init-exiting the Linux
kthread server.
2015-02-19 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/calibrate/calibrate.c: do not self calibrate if
SCHED_LATENCY is set already; update also .rtai-config.
* base/sched/sched.c: do not convert negative kern-user busy waits.
* base/arch/x86/Kconfig: display kern busy wait again.
2015-02-18 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c and sys.c: busy timed sync waits only if there
is a sched_latency.
* base/sched/sys.c: set sched_latency also when sched_latency is zero.
* base/include/rtai_schedcore.h: enable timed sync busy waits at wake
up only if there is a sched_latency.
* base/arch/x86/calibrate/calibrate.c: recoded for use within the
make self tuning.
2015-02-16 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_fpu_64.h: cleaned legacy oldnames.
2015-02-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sys.c: rt_thread_create returns an error when called from
an RTAI own kernel task; use a wrapper function and a LINUX kernel
sem to grant the new kthread is running before retunring,
substitutes the previous polled wait for an RTAI enabled new khread.
2015-02-14 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/drivers/serial/serial.c and serialP.h: cleaned legacy
oldnames.
* addons/rtdm/drvlib.c, module.c, rtdm_driver.h and xn.h: cleaned
legacy oldnames, aligned with renewed variable and structure names.
* base/arch/x86/Kconfig: prepared for a few next config changes.
* base/arch/x86/hal/hal/c: cleaned legacy oldnames and macros, aligned
with renewed variable and structure names.
* base/include/rtai_hal_names.h: reduced to what is still needed with
the latest patches and a more direct use of its raw names.
* base/include/asm-x86/rtai_hal.h: redefined struct calibration_data,
cleaned legacy oldnames, messy redundant macros and unused parts of
code; defined new macros, aligned with renewed variable and structure
names.
* base/include/asm-x86/rtai_lxrt.h: cleaned legacy oldnames, messy
redundant macros and unused parts of code; defined new macros,
aligned with renewed variable and structure names.
* base/include/asm-x86/rtai_vectors.h: cleaned of vectors and ipis
macros definition, introduced a reminder of what in use now.
* base/sched/api.c: cleaned legacy oldnames and cared of the no more
used rt_times tunables, now rt_smp_times[0].
* base/sched/sched.c: cleaned legacy oldnames, messy redundant macros;
aligned with renewed variable, macros and structure names; cared of
the no more used rt_times tunables, now rt_smp_times[0].
* base/sched/sys.c: cleaned legacy oldnames, aligned with renewed
variable and structure names.
* base/tasklets/tasklets.c: cleaned legacy oldnames, aligned with
renewed variable and structure names.
* base/wd/wd.c: cleaned legacy oldnames, aligned with renewed variable
and structure names.
* testsuite/kern/latency/latency-module.c,
testsuite/kern/preempt/preempt-module.c,
testsuite/kern/switches/switches-module.c: cleaned legacy oldnames,
aligned with renewed variable and structure names.
2015-02-11 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: got rid of rt_times variable, beware
just the variable, not of the structure with the same name.
2015-02-08 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sys.c: rt_thread_create poll waits till the created
thread is rtai enabled.
2015-02-07 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: added the possibility of testing different ways
to set active_mm for kernel threads.
2015-02-06 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: avoid loosing the active_mm stolen for kernel
threads working in hard real time, by incrementing its mm_count
when it is acquired in fast_schedule, then mmdrop active_mm any
time they are returned to soft mode in give_back_to_linux.
2015-02-05 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_sched.h: added the new declaration for
rt_kthread_init, hiding the old one. Still to be cleaned a bit.
* base/include/rtai_schedcore.h: hided the old declaration of
rt_kthread_init. Awaiting some cleaning.
* base/sched/sched.c: made fast_schedule support hard transitions
for kernel threads, without touching the context switch in use.
* base/sched/sys.c: reinstated support for hard-soft kernel threads,
emulating the scheme used in kernel space. Beware of the way of
calling RTAI services in the case of soft threads. See
soft calling functions. The legacy RESERVOIR option awaits a new
helper module. TBD: the releated testsuite.
2015-02-04 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: got rid of all old timer request/free.
* base/sched/sched.c: got rid of all old start/stop_rt_timer and
likes. Added fastcall hook.
2015-02-02 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_hal.h: macro rtai_tskext modified so to
include the related LINUX task, and to use either ptd in task_struct
and thread_info.
* base/ipc/msg/msg.c;
base/ipc/shm/shm.c;
addons/rtdm/xn.h;
base/sched/api.c, sched.c, sys.c: use the new macro rtai_tskext,
modified so to include the related LINUX task.
2015-02-01 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: force oneshot mode ALWAYS, through a bad, but
temporary, hack.
2015-01-31 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_lxrt.h: gather the common parts of 32-64
bits archs, unify the definition of union rtai_lxrt_t.
* base/include/asm-x86/rtai_lxrt_32.h: keap only the 32 bits specific
definitions and services, required copying __do_strncpy_from_user
from LINUX, as for 64 bits.
* base/include/asm-x86/rtai_lxrt_64.h: keap only the 64 bits specific
definitions and services.
* base/include/asm-x86/rtai_fpu.h: gather the common parts of 32-64
* base/include/asm-x86/rtai_fpu_32.h and rtai_fpu_64.h: keap only the
32-64 bits specific bits archs.
2015-01-31 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: significantly refurbished, x86-32-64
unified, patch's hard timers services; from now on only the oneshot
mode must be used.
2015-01-27 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm/rtai_hal.h: unify 32-64 bits independent code here
and revise real time spin locks and global lock.
* base/include/asm/rtai_hal_32.h and rtai_hal_64.h: left with only
32-64 bits arch specific code.
* base/sched/sched.c: use non rtai_cli protected global lock get and
release from rtai_hal.h.
* base/arch/x86/hal/hal_32.c: redefine rtai_cpu_lock according to the
global lock structure.
* addons/rtdm/xn.h: make xnlock_t comply with the new global_lock
structure declaration.
2015-01-24 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: exploit oneshot hardware timing setting
from the modified patch, working but still to be finished along with
some cleaning more.
* base/include/asm-x86/rtai_hal_32.h: optionally set the oneshot delay
by optionally using ipipe_timer_set if X86_FEATURE_TSC_DEADLINE_TIMER.
* base/sched/sched.c: do not provide rt_linux_hrt_set_mode any more,
use the one found of rtai_hal.c.
2015-01-22 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: serialize symultaneous uses of steal_from_linux
on the same cpu, directly pass fast_schedule as a migration_hook for
the patch; removed rt_daemonize stuff.
* base/sched/sys.c: removed rt_daemonize stuff.
2015-01-18 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/rtdm_driver.h and xn.h: a few changes to work also
without legacy constraints.
* base/arch/x86/hal/hal_32.c and hal_64.c: a series of changes to work
also without legacy constraints and some cleaning.
* base/include/rtai_hal_names.h: cleaned up, also in relation to a few
changes to work also without legacy constraints.
* base/include/rtai_hal_schecore.h: a few changes to work also without
legacy constraints.
* base/include/asm-x86/rtai_hal_32.h: a few changes to work also without
legacy constraints.
* base/sched/sys.c: grant zeroing of ptd and use the patch support
* base/sched/sched.c: some cleaning and changes to work also without
legacy constraints.
2015-01-09 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal, hal_32.c and hal_64.c : if IsolCpusMask is
null use that of Linux, so that there is nomore the need to set it
in RTAI.
2015-01-08 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c and hal_64.c : avoid reprogramming the
apic timer mode when tsc-deadline is available.
2015-01-07 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_hal_32.h and rtai_hal_64.h: reworked
rt_set_timer_delay.
* base/include/asm-x86/rtai_hal.h: reorder a set of declarations.
* base/sched/sched.c: avoid multiple oneshot delay units conversion,
unified in rt_set_timer_delay.
2015-01-05 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_lxrt_32.h and rtai_lxrt_64.h: update
caring of the Linux version when manupulating the fpu_counter in
contet switches.
* base/sched/sched.c: care of Linux version in setting need_resched.
2014-05-13 Pierangelo Masarati <masarati@aero.polimi.it>
* base/sched/{GNUmakefile.am,sched.c}: use MODULE_ALIAS()
instead of manual symlinks for legacy schedulers,
thanks to Lukas Anzinger <lukas@lukasanzinger.at>
* base/sched/GNUmakefile.in: regenerate
2014-05-06 Pierangelo Masarati <masarati@aero.polimi.it>
* configure.in: fix linux source dir extraction when further args
follow -C, thanks to Lukas Anzinger <lukas@lukasanzinger.at>
* configure: rebuild
2014-02-19 Pierangelo Masarati <masarati@aero.polimi.it>
* addons/rdtm/GNUmakefile.*: updated for Linux >= 3.10.0.
* addons/rtdm/module.c: removed provisional quick and dirty
inclusion of vfile.c.
2014-02-19 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/vfile.h, vfile.c and proc.c: updated for Linux >= 3.10.0,
lifted from native nucleus and rtdm.
* addons/rtdm/rtdm_driver.h: added declaration to support the new proc
file system.
* addons/rtdm/module.c: provisional quick and dirty inclusion of
vfile.c for a preliminary test.
2014-02-16 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_proc_fs.h: added support for changes of proc files
usage for Linux >= 3.10.
* base/arch/x86/hal/hal_32.c and hal_64.c: modified support of proc
files for Linux >= 3.10.
* base/sched/sched.c and api.c: modified support of proc files for
Linux >= 3.10.
* base/ipi/fifos/fifos.c: modified support of proc files for
Linux >= 3.10.
* base/ipi/mq/mq.c: modified support of proc files for Linux >= 3.10.
* base/wd/wd.c: modified support of proc files for Linux >= 3.10.
* testsuite/kern/latency/latency-module.c: modified support of proc
files for Linux >= 3.10.
2013-11-23 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/math/libm.c: added function cd2str, i.e. compex double to
string.
2013-11-19 Paolo Mantegazza <mantegazza@aero.polimi.it> and
Marco Morandini <morandini@aero.polimi.it>
* base/math/libm.c: updated to use complex functions and glibc-libm.a,
i386 only, newlib-libm.a and uClibi-libm.a
* base/math/export_glibc, export_newlib.h and export_uclibc.h: private
headers exporting what is specific to each of the supported libm.a.
* base/math/README.KLIBM: explanation of what the new way math support
in kernel space is.
* base/include/rtai_math.h: new self contained header.
* base/arch/x86/Kconfig: add parameters to configure the refurbished
math support in kernel space.
2013-12-06 Pierangelo Masarati <masarati@aero.polimi.it>
* RTAI 4.0
2013-11-14 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c: do not make available rtai_sched_affinity
and free_isolcpus_from_linux under UP.
2013-11-11 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/math/libm.c: added a trivial d2str (double to string).
* base/include/rtai_kerrno.h: added the declaration of d2str.
2013-11-09 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/math/libm.c: uses just the standard GPLed glibc, added the
possibility of calling both the float and double version of
glibc-libm functions.
* base/include/rtai_sched.h: a added kerrno to RT_TASK, to support
a per kernel task errno management with the refurbished libm.c
* base/include/rtai_kerrno.h: new header to support the management
of per kernel tasks errno.
2013-11-05 Pierangelo Masarati <masarati@aero.polimi.it>
* RTAI 4.0-test2
2013-11-03 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: added rt_get_cpu_freq and static
inlines for an easier converion of tsc counts from/to nanos,
directly in user space; simple inlined conversion functions
provided in rtai_lxrt.h.
* base/sched/sys.c: implemented the above rt_get_cpu_freq.
* base/arch/x86/hal/hal.c: added a function to be called for ensuring
that no Linux task will be on the CPUs isolated for RTAI.
* base/arch/x86/hal/hal_32.c and hal_64.c: when initting RTAI hal call
the function removing Linux tasks from the CPUs isolated for RTAI.
* README.ISOLCPUS: add a short explanation of what newly added above.
2013-11-02 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c and hal_64.c: check irq_set_affinity
in desc, to avoid an unharmful but annoying Linux WARN_ONCE.
2013-10-28 Pierangelo Masarati <masarati@aero.polimi.it>
* RTAI 4.0-test1
2013-10-27 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/hal/hal_32.c and hal_64.c: let hal proc file
display IsolCpusMask.
2013-10-26 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_hal.h: make available an inlined support
for reading the tsc directly from user space.
* base/include/asm-x86/hal/hal_32.c and hal_64.c: echo more proper
informations about system frequencies and timings.
* base/include/rtai_lxrt.h: rt_get_time can read the tsc directly.
* base/include/rtai_posix.h: read the tsc directly from user space.
* testsuite/user/ latency/latency.c and latency/latency.c: use tsc
reading from user space, so to be more consistent with the
calibration-helper results.
2013-10-25 Pierangelo Masarati <masarati@aero.polimi.it>
* configure.in, configure: propagate TSC detection to user space
2013-10-25 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_schedcore.h: assign and use only
wake_up_srq[0].srq.
* base/sched/sched.c: pend hard to soft task transitions to
wake_up_srq[0].srq.
2013-10-15 Pierangelo Masarati <masarati@aero.polimi.it>
* RTAI 3.9.2
2013-10-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: make rt_smp_half_tick available to the whole
module files, have wake_up_srq_handler loop over all pending
cpu task lists.
* base/sched/sys.c: avoid stopping/starting the hard oneshot timer by
setting rt_smp_haf_tick accordingly to the just update latency.
2013-10-05 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/calibration/calibration_helper.c: restructured in a,
hopefully, better way, with respect to its first commit.
* base/arch/x86/calibration/README: rewritten according to the new
calibration way.
* base/arch/x86/Kconfig: added BUSY_ALIGN_RET_DELAY parameters for
kernel and user space.
* base/include/rtai_posix.h: fixed the return values of user space
to better comply with the related specs.
* base/include/asm-x86/rtai_hal.h: added various definition related to
the setting and declaration of busy aligns in the calibrate struct.
* base/sched/sched.c: reordered the one shot firing macros and enabled
the newest busy wait align stuff.
* base/arch/i386/hal/hal.immed and base/arch/x86_646/hal/hal.immed:
check irq desc and chip addr to avoid invalid pointers when used
in a brute mode to divert all Linux interrupts belonging to
IsolCpuMask; return a null cpumask in case of failure an irq
affinity assignment.
2013-10-04 Pierangelo Masarati <masarati@aero.polimi.it>
* configure.in, base/arch/x86/Kconfig: add support for
configurable user/kernel busy align delay
* base/arch/x86/calibration/calibration_helper.c, GNUmakefile.am:
insert/remove modules directly from within helper
2013-09-30 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: take into account the newly settable return
delay in busy wait alignment.
* base/sched/sys.c: added floating point operations to the kernel
calibrator.
* base/include/asm-x86/rtai_hal.h, rtai_hal_32.h, rtai_hal_64.h:
remove struct calibration_data the 32/64 bits hal parts and define
a unique one in the common hal.
* base/include/rtai_lxrt.h: add the call to the kernel calibrator.
* base/arch/x86/Kconfig: fixed a few spelling errors and added a new
configuration option.
* base/arch/x86/calibration/calibration_helper.c: added floating point
operations to the user calibrator.
* base/arch/x86/calibration/runinfo: added the command line to launch
the new calibration way.
2013-09-29 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: calibrate the hard timers at the scheduler
insmoding, care of avoiding null timer delays programming.
* base/sched/sys.c: add the support for an easier determination of
the scheduling latencies of RTAI own kernel tasks.
* base/arch/x86/hal/hal.c: put the determination of the hardtimers
programming delay in the common hal part.
* base/arch/x86/calibration/calibration_helper.c: simplify the
calibration of user and kernel space scheduling latencies.
2013-09-24 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_schedcore.h: use just one wake up srq handler,
keeping a per cpu list of tasks to be woken up.
* base/sched/sched.c: omit the CONFIG_SMP ifdef in wake up srq
handler.
* base/arch/x86/hal/hal_32.c, hal_64.c: use cpu_active_mask in
selecting hard timers.
2013-09-13 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: provide a do nothing rt_daemonize for
linux > 3.0.
2013-08-31 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/module.c: aligned the intr part to its latest sibling.
2013-08-30 Pierangelo Masarati <masarati@aero.polimi.it>
Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/comedi/GNUmakefile.am: no kernel headers in user space
* addons/comedi/rtai_comedi.h: do not include linux/compiler.h
in user space
2013-08-28 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_64.c: aligned to work with Linux-3.x,
it is much as its hal_32.c counterpart.
2013-08-27 Pierangelo Masarati <masarati@aero.polimi.it>
Paolo Mantegazza <mantegazza@aero.polimi.it>
* configure.in, base/arch/x86/Kconfig: enable hard-soft toggling,
by automatically installing a SIGUSR1 Linux signal & handler,
using standard Linux KILL support
2013-08-26 Ian Abbott <abbotti@mev.co.uk>
* configure.in: Some AC_CONFIG_LINKS() go wrong when building
outside the source tree.
2013-08-25 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: return to requesting the 8254-PIT
interrupt in rt_timer_request, we care of the related issue,
alrady solved in the way of the explanation below, by modifying
the patch.
* base/arch/x86/hal/hal_64.c: same as above, but, once more, since
X86_64 has an APIC always, it is likely non an issue in this case.
2013-08-22 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: do not request the 8254-PIT interrupt
in rt_timer_request, it is done by the patch already; so if the
related timer handler is installed in advance the following
rtai_request_tickdev will get an error and the related 8254 stuff
will not be initialized.
* base/arch/x86/hal/hal_64.c: same as above, since X86_64 has an APIC
always, it is likely non an issue in this case.
2013-08-20 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: use the irq desc eoi in rt_eoi_irq.
* base/arch/x86/hal/hal_64.c: same as above.
2013-08-19 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/Kconfig: ask for RTNet installation directory if
NETRPC has to use it.
2013-08-19 Roberto Bucher <roberto.bucher@supsi.ch>
* rtai-lab/xrtailab.cpp: few blank lines for better read code
* rtai-lab/scicoslab/devices/scicos_block4.h: new version from
scicoslab-4.4.1
* rtai-lab/scicoslab/macros/Makefile: fitted to scicoslab-4.4.1
* rtai-lab/scicoslab/macros/RTAICodeGen_.sci: new version for
scicoslab-4.4.1
* rtai-lab/scicoslab/macros/SetTarget_.sci: new version for
scicoslab-4.4.1
* rtai-lab/scicoslab/macros/RT_templates/rtai.mak: added blas,
lapack and gfortran libs for the linker
2013-08-18 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_lxrt.h: set appropriate use of the
32-64 specific versions; GNUmakefile.am GNUmakefile.in also
modified accordingly.
2013-08-17 Mark Fortescue <mark@mtfhpc.demon.co.uk>
* base/addons/rtdm/module.c: split the loop initializing the timers
into two parts, fixes possible SMP preemption issues.
* base/addons/rtdm/xn.h: add missing list pointers at timers
initialization.
(N.B. a forgotten due recovery of a year old, 2012-08-03, fix
existing in VULCANO CVS already.)
2013-08-07 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: modified the support and declarations for
externally forcing a task to soft, so to make it possible toggling
hard_soft by enforcing a signal handler for SIGUSR1, at task init,
providing also a support function to get the pid of the RTAI task
to be toggled using Linux kill.
* base/sched/sys.c: provided the toggling support handler, adding
a function to return the pid of any RTAI task; notice that the new
way wholly substitute the old forcing of a task to soft only. Yet
to be made optionally configurable at RTAI making.
2013-08-03 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: check if irq disable-enable are NULL
and divert to mask_unmask if so, add rt_irq_mask.
* base/include/rtai_usi.h: #if 0 ... #endif the request of an srq for
USI, see comment in the file.
2013-08-02 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: fixed wrong arg list struct setup in all
USI support functions.
2013-08-01 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86/rtai_sched_32.h, rtai_sched_64.h and
rtai_sched.h: removed and gathered code common to 64 and 32 into
rtai_sched.h.
2013-07-30 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/msg/msg.c: tightened task and message copying pointer
checks.
2013-07-29 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/netrpc/netrpc.c added a missing kclose for x86_64.
2013-07-26 Pierangelo Masarati <masarati@aero.polimi.it>
* change the making of the schedulers, finally we got rid of a useless
double creation of the same scheduler; now the only real scheduler
is rtai_sched, legacy is saved by linking lxrt, ksched, up, mup and
smp to it.
2013-07-26 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal.c and hal_32.c: fixed TSC sync support, broken
because of both the new x86 unified making and the need to avoid
using kernel_thread(...) directly.
2013-07-25 Pierangelo Masarati <masarati@aero.polimi.it>
* change making configuration so to not include Linux headers for user
space code; it made GCC >= 4.7.2 yell for conflicting definitions
within the same headers; it does not break under older GCCs, which
did not yell.
2013-07-25 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/module.c: fix the changed name of the isr scheduler lock.
* base/sched/api.c: opt for a randomization in generating names for
NULL adr requests (pevious nameseed way kept under #if #else #endif).
* base/sched/sched.c: make rtai_isr_sched_handle available and
exported always.
* base/arch/x86/hal/hal.c: moved random number generator to
base/include/rtai.h, so to have it available anywhere.
* base/include/rtai.h: put the random number generator here, so to
have it available anywhere.
* base/include/rtai_nam2num.h: echo a "|null|" for names that are
either 0 or generated through rt_get_name(NULL).
2013-07-24 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/api.c: make getting a unique name from registry safer
against duplications.
* base/include/rtai_nam2num.h: embellished the coding and added a
couple of usable characters more.
2013-07-21 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/mq/mq.c: trivial changes to avoid annoying compiler
warnings.
* base/include/rtai_mq.h: fixed a few wrong working space indicators.
* base/sched/sys.c: added test to avoid copying from null user space
pointers into the kernel space task specific buffer.
2013-07-20 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/netrpc/netrpc.c: aligned to the latest kthreads Linux
support.
* base/sched/sched.c: rt_signal_wake_up reverted to wake up on
Linux signals hanyhow.
2013-07-18 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: rt_signal_wake_up will allow hard real time
tasks to receive signals only if the user has indicated to be
willing to accept them by installing a signal handler for SIGTERM.
2013-07-17 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c and sys.c: use RTAI own task pid for Pierre's
intertask messaging stuff.
* base/include/rtai_lxrt.h: return pid<->task stuff back here, it is
used also elsewhere than in Pierre's intertask messaging stuff.
* base/ipc/msg/msg.c: use unified rtai own task pids in Pierre's
intertask messaging stuff.
2013-07-16 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: removed call to rt_kthread_init and moved
pid<->rttask stuff into msg.c
2013-07-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: keep formal kthread_init for compatibility, with
task_init behind it, used by msg.c.
* base/ipc/msg/msg.c: brounght here pid<->rttask inlines from rtai_lxrt
and reworked proxy support.
2013-07-10 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: rtai_print_to_screen is rt_printk now,
added call rt_kthread_init.
2013-07-03 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/liblxrt/init.c: used nomore, just an empty code.
* base/sched/liblxrt/services.c: include RTAI services to be libbed
from the LXRT user space part of header files directly and not
through the inclusion of rtai_schedcore.h.
2013-07-02 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_lxrt.h: remove unused rt_clone and lower
RT_THREAD_STACK_MIN.
2013-06-09 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/xn.h and base/ipc/shm/kvmem.c: define VM_RESERVED if it
does not exist.
* base/arch/x86/calibration/smi-module.c: discard __devinit in function
definitions, found no more in Linux.
* base/include/rtai_schedcore.h: add e new Linux version dependenc for
OOM_DISABLE.
* base/sched/sys.c:just a first fix to have it compile against the
gone for ever Linux kernel_thread, to be fixed, it is just for
having it compile at the moment.
2013-03-14 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/arch/x86/hal/hal_32.c: use CPU_MASK_ALL to fix a quick and
dirty, but very very bad hack, in selecting Linux timers for RTAI.
2013-02-28 John Morris <john@zultron.com>
* base/sched/sched.c: changed set_cpu_allowed to set_cpu_allowed_ptr
and to cpumask_of_cpu to cpumask_of, to allow compilation against
kernels that have CONFIG_CPUMASK_OFFSTACK=y
2013-02-28 Shahbaz Youssefi <ShabbyX@gmail.com>
* base/include/rtai_rwl.h: fixed a missing nam2num in rt_named_rwl_init.
* base/include/rtai_spl.h: fixed a missing nam2num in rt_named_spl_init.
2013-02-28 Sebastian Kuzminsky <seb@highlab.com\>
* base/arch/x86/patches/hal-linux-3.5.7-x86-3.patch: fix to help
DEBIAN packaging.
2013-02-28 Paolo Mantegazza <mantegazza@aero.polimi.it>
Pierangelo Masarati <masarati@aero.polimi.it>
* Completed the initial tructuring for a unified X86 arch, making
it work for X86_32-i386. X86_64 still in progress.
2013-02-10 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c, api.c, sys.c: made compatible with Linux-3.x.
* base/include/rtai_lxrt.h: specialized NARG macro into LXRT_NARG, to
avoid warning with the same name appearing in Linux.
* base/include/rtai_wrappers.h: modified to be made compatible with
Linux-3.x.
* base/include/asm/rtai_fpu.h: modified to be made compatible with
Linux-3.x.
2012-01-03 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/x86_64/hal/hal.immed: updated to work with Linux up to
2.6.38.
2011-12-31 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c and sys.c: if PF_EVNOTIFY is defined use the
related Linux patch service, instead of or-ring the related flag
directly.
* addons/rtdm/device.c: avoid DECLARE_MUTEX, use the related struct
declaration and initialize it explicitly at module init.
2011-12-30 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c and sys.c: update PF_EVNOTIFY flag assignement
according to Linux version.
* base/arch/i386/hal/hal.immed: upgraded to work up to Linux 2.6.38.8.
2011-05-09 Pierangelo Masarati <masarati@aero.polimi.it>
* configure.in, configure: avoid potential shell syntax error,
allow separate src and build dirs for Linux, from Ian Abbott
* configure.in, configure: create/install setsmi for x86_64,
from Ian Abbott
2011-03-04 Pierangelo Masarati <masarati@aero.polimi.it>
* makefile: remove erroneous '%' in 'all ::' rule
2011-02-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_wrappers.h: define init_MUTEX_LOCKED if not
defined in Linux.
* base/ipc/fifos/fifos.c: define iopl according to
HAVE_UNLOCKED_IOCTL.
* base/ipc/shm/shm.c: define iopl according to HAVE_UNLOCKED_IOCTL.
2011-01-11 Holger Nahrstaedt <nahrstaedt@control.tu-berlin.de>
* rtai-lab/scilab5/RTAI/macros/gif_icons: add icons for each block
* rtai-lab/scilab5/RTAI/macros/svg_icons: add svg-icons for each block
* rtai-lab/scilab5/RTAI/macros/h5: add directory for xcos-block generation
* rtai-lab/scilab5/RTAI/macros/do_compile_superblock_rt.sci: code generation (same function as for scilab 4.1.2)
* rtai-lab/scilab5/RTAI/macros/RTCodeGen.sci: code generation for scilab 5.3 (replaced RTAICodeGen_.sci)
* rtai-lab/scilab5/RTAI/macros/buildmacros.sce
* rtai-lab/scilab5/RTAI/macros/*.sci: replace getvalue by scicos_getvalue and change exprs from a column vector to a row vector.
* rtai-lab/scilab5/RTAI/license.txt
* rtai-lab/scilab5/RTAI/builder.sce
* rtai-lab/scilab5/RTAI/loader.sce
* rtai-lab/scilab5/RTAI/readme.txt
* rtai-lab/scilab5/RTAI/src/c/builder_c.sce
* rtai-lab/scilab5/RTAI/src/c/rtai_sinus.sce
* rtai-lab/scilab5/RTAI/src/c/rtai_square.sce
* rtai-lab/scilab5/RTAI/src/c/rtai_step.sce
* rtai-lab/scilab5/RTAI/RT_templates/rtai.mak
* rtai-lab/scilab5/RTAI/RT_templates/rtai_sh.mak
* rtai-lab/scilab5/RTAI/etc/RTAI.start
* rtai-lab/scilab5/libs/scicoslibs.tgz
* rtai-lab/scilab5/examples/test.xcos:
* rtai-lab/scilab5/Makefile:
* rtai-lab/scilab5/scilab
* rtai-lab/scilab5/devices/Makefile:
* rtai-lab/scila5/devices/scicos_block4.h: modifications for scilab 5.3.0
2010-11-27 Fernando Augusto <fernando_aug@yahoo.com.
* base/ipc/sem/sem.c: fixed a possible nasty locking loop in
rt_sem_broadcast.
2010-11-16 Holger Nahrstaedt <nahrstaedt@control.tu-berlin.de>
* rtai-lab/scilab/rtmain.c:
* rtai-lab/scilab5/devices/rtmain.h:
* rtai-lab/scilab/devices/rtmain.h: rtRegisterLogData should not be static (otherwise there is a problem with shared libraries)
2010-11-11 Holger Nahrstaedt <nahrstaedt@control.tu-berlin.de>
* rtai-lab/scilab/loader.sce:
* rtai-lab/scilab/macros/RTAI/loadmacros.sce:
* rtai-lab/scilab/macros/Makefile:
* rtai-lab/scilab/macros/scilab: New approach in including RTAI files
in scilab (same approach as for scicoslab).
2010-10-12 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/msg/msg.c: make rt_sendx_if usable form interrupt handlers.
2010-10-08 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-x86_64/rtai_fpu.h: upgrade to new FP link in thread
struct definition for > 2.6.34.
2010-10-06 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-i386/rtai_fpu.h: upgrade to new FP link in thread
struct definition for > 2.6.34.
2010-09-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/drivers/serial/rtai_serial.h: make all declared prototypes
compilible in a lib.
* rtai-py/rtai_ser.py: add RTAI serial support for python.
2010-09-09 Pierangelo Masarati <masarati@aero.polimi.it>
* base/arch/x86_64/calibration/GNUmakefile.am,
base/arch/x86_64/calibration/GNUmakefile.in,
base/arch/x86_64/calibration/Makefile.kbuild:
sync'ed with i386
* rtai-lab/scicoslab/GNUmakefile.am, rtai-lab/scicoslab/GNUmakefile.in:
sync'ed with vulcano
* base/arch/i386/calibration/smi-module.c,
base/arch/x86_64/calibration/smi-module.c,::
#ifndef PCI_DEVICE_ID_INTEL_ICH10_1, for older kernel without it.
2010-08-23 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/asm-i386/rtai_hal.h, base/include/asm-x86_64/rtai_hal.h:
define a blank IPIPE_IRQ_DOALL when not defined yet.
2010-08-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
suggested by Tommaso Falchi Delitala <volalto86@gmail.com>
* addons/rtdm/rtdm.h: change RTAI_LXRT into RTDM_RTAI_LXRT to avoid
naming conflicts
2010-06-29 Roberto Bucher <roberto.bucher@supsi.ch>
* rtai-lab/scicoslab/macros/RTAI/RTAICodeGen_.sci:
* rtai-lab/scicoslab/macros/RTAI/gif_icons: Patches
from Rafael Campos (thanks!!!)+modification from INRIA.
2010-06-09 Rafael Campos <rafael@freedom.ind.br>
* addons/comedi/rtai_comedi.h: fix a macro to keep it compatible for
both COMEDI site and Linux staging drivers versions.
2010-05-17 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/* and addons/drivers/16559A/*: upgraded to the latest
rtdm changes.
2010-05-16 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/sched/sched.c: use RT_TASK "running" variable to mark a truly
owned resource semaphores when scheduling.
* base/ipc/sem/sem.c: modify "ubi maior minor cessat" basing it on
the "running" variable of RT_TASK; add the feature also to mutex
signal specific to conds.
* base/include/rtai_sem.h: the use of RT_TASK "running" in "ubi maior
minor cessat" allows reverting to previous sem structure declaration.
2010-05-15 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/include/rtai_posix.h: fixed return value of pthread_cond_wait.
2010-05-14 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/sem/sem.c: a higher prio task can grab a resem that is
owned by a task that has not been scheduled yet "ubi maior minor
cessat"; mimicking Linux rt-mutexes.
* base/include/rtai_sem.h: add a variable to the sem struct to allow
checking if a resem is truly owned already.
2010-05-07 Paolo Mantegazza <mantegazza@aero.polimi.it>
* rtai-py/rtai_lxrt.py: fix rt_task_delete and rt_thread_delete, to
match changes in rtai_lxrt.h (thanks to Chris Cole).
2010-05-01 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/drvlib.c: fix wrong return value in sems support
functions _sem_wait and _sem_wait_timed.
2010-04-28 Paolo Mantegazza <mantegazza@aero.polimi.it>
* addons/rtdm/module.c: allocate selector on the real time heap as
its deletion is now delayed to an srq service.
2010-04-25 Paolo Mantegazza <mantegazza@aero.polimi.it>
* base/ipc/msg/msg.c: strengthened correctness check of task args,
used a more appropriate unsigned long as return arg of rpcs, revised
sendx, previous implementation could overwrite previous sends,