forked from ibm-power-utilities/powerpc-utils
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathChangelog
3507 lines (2438 loc) · 133 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
powerpc-utils-1.3.0
=====================================================================
commit ad4cf85f4bc80f6f5cbf4ce0b71eada696848081
Author: Nathan Fontenot <[email protected]>
Date: Thu Jan 7 15:17:41 2016 -0500
Move the powerpc-utils package from the IBM license to the GNU GPL license.
Signed-off-by: Nathan Fontenot <[email protected]>
commit f037afe3bcece02ca3a3b1711a57c66db088cfa1
Author: Nathan Fontenot <[email protected]>
Date: Tue Dec 1 16:36:11 2015 -0500
drmgr: Correct checking for display adapters when adding PHB
A previous commit (c2091227) added formalization for checking for display
adapters. Unfortunately the update tries to verify if there is a display
adapter for a NULL phb pointer when trying to add a PHB resulting in a
segfault on phb add requests.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 85a5037cb01c1aaa5ad3dbd91f74dbcdd823dbb1
Author: Nathan Fontenot <[email protected]>
Date: Tue Dec 1 16:35:04 2015 -0500
drmgr: Remove extraneous debug output
Remove extra debug statements for allocating and free'ing drc information.
These were mistakenly left in during debug of a previous patch.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 33aa48a0c4ebc4e660080514939232ad2912cb6d
Author: Nathan Fontenot <[email protected]>
Date: Tue Dec 1 16:33:45 2015 -0500
drmgr: Fix node add buffer overflow
When adding a node to the device tree, the drmgr command writes a terminatin
null in the first character past the end of the allocated buffer. This
results in a SIGABRT being generated by libc when the buffer is free'ed.
The fix is to move the writing of the space between properties to the
beginning of the loop instead of at the end of the loop.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 55544a975f87e9d76366e6fefcf571ae01f3e7eb
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 23 21:10:27 2015 -0500
drmgr: Remove extraneous "d" in buffer
Remove an erroneous character left in a print statement used
to update the drconf property.
Signed-off-by: Nathan Fontenot <[email protected]>
commit f4006893d84c85bbf5b86919c28ba56047001204
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 23 21:09:47 2015 -0500
drmgr: Correct return code from set_mem_scn_state()
The set_mem_scn_state() routine incorrectly returns the number of chars
written in the case of success instead of returning 0. This results in the
caller assuming a failure occurred when setting the memroy section state
and fails the memory online/offline operation.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 4f515ea179cb36f26adb53c93b79f834e32c2abe
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 23 21:08:54 2015 -0500
drmgr: Correct LMB counting when discovering LMBs
Recent updates to how drmgr counted the number of LMBs found did
not save the count value correctly and in the case of LMBs found
in the ibm,dynamic-reconfiguration-memory property only counted
LMBs that were present. This caused havoc when trying to shuffle
the LMBs.
Save the LMB count in the lmb_list_head struct so we have it across
function calls. Also update the shuffle_lmbs to just take a pointer to
the lmb_list_head and make changes locally in the routine.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 9a1957588b040db0dddfa284dbb96a817e699792
Author: Cédric Le Goater <[email protected]>
Date: Mon Nov 23 21:07:17 2015 -0500
update_flash_nv: improve check_ipmitool
This patch adds a quick test in check_ipmitool to make sure the
BMC is alive
Signed-off-by: Cédric Le Goater <[email protected]>
commit 4e4e5f9f8d16a2ad196d1fd5388de23aa2e1fdbc
Author: Cédric Le Goater <[email protected]>
Date: Mon Nov 23 21:06:52 2015 -0500
update_flash_nv: fix firmware update sequence for powernv
This patch modifies the way the firmware update is performed from
the host. It adds a cold reset of the BMC before and after the
update.
Signed-off-by: Cédric Le Goater <[email protected]>
commit 1f25105c1c7f1384be4aef03f2aa6d5bd065eb11
Author: Cédric Le Goater <[email protected]>
Date: Mon Nov 23 21:06:22 2015 -0500
update_flash_nv: force the use of the ipmitool usb backend
Signed-off-by: Cédric Le Goater <[email protected]>
commit b0e5ba4f55dcf1a5d013eddbb95b660c65504b3c
Author: Nishanth Aravamudan <[email protected]>
Date: Mon Nov 23 21:04:45 2015 -0500
ppc64_cpu: output to user if the requested number of cores are not brought o
Now that we check the return code from set_one_core, we should ensure
that the requested number of cores in --cores-on is achieved, or return
-1 intead. Also add a slightly verbose message in that case.
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 1a1bb7167a3946712a5b59b5566bac6ced46cdc9
Author: Nishanth Aravamudan <[email protected]>
Date: Mon Nov 23 21:03:37 2015 -0500
ppc64_cpu: use rc instead of boolean flag
We can overload the rc variable, reducing the number of local variables
in do_online_cores.
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 37eeba5583d48558f2ca1de6096df83494005e6f
Author: Nishanth Aravamudan <[email protected]>
Date: Mon Nov 23 21:00:14 2015 -0500
ppc64_cpu: check return code from set_one_core
In both --cores-on and --online-cores, the return code from the helper
function set_one_core is not checked. This can lead to ppc64_cpu
returning 0 even though the command technically failed to do what was
requested. After this patch, --cores-on still doesn't return a proper
error code (fixing in a follow-on).
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 9711ee00b604a39d47dfced7f0c49cf40600131f
Author: Nishanth Aravamudan <[email protected]>
Date: Tue Nov 17 16:24:37 2015 -0500
ppc64_cpu: add --{on,off}line-cores to {on,off}line specific cores
We currently have support for a mechanism to ensure an exact number of
cores (or all cores after a recent patch submission) are online.
However, there is no control over which cores are selected. Add two new
options to ppc64_cpu, --online-cores and --offline-cores to support
specifying which cores to bring online or offline, respectively. Both
options take either no parameters, in which case they output a
comma-separated list of cores that are online/offline, or a
comma-separated list of cores, in which case the command sets those
cores' states to online/offline.
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 2239dd005e00d63c6e68fe962928b5f0b08ef6cb
Author: Nishanth Aravamudan <[email protected]>
Date: Tue Nov 17 09:33:55 2015 -0500
ppc64_cpu: rename do_cores_online to do_cores_on
This matches the other function names more closely.
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit bb8472f0658565e5188150a141fa8856b55000ca
Author: Nishanth Aravamudan <[email protected]>
Date: Tue Nov 17 09:31:57 2015 -0500
ppc64_cpu: add all parameter to --cores-on to online all cores
Rather than finding out how many cores are in a system, and then passing
that value to --cores-on, it can be handy to have a shortcut to just
online all cores.
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 17cd4bbf7c20d8096523c765c1157e4a58d31363
Author: Nishanth Aravamudan <[email protected]>
Date: Tue Nov 17 09:29:06 2015 -0500
ppc64_cpu: handle errors from strtol in --cores-on parsing
Currently, on a 24-core system, passing certain values to `ppc64_cpu
--cores-on` leads to some surprising results:
ppc64_cpu --cores-on=all
ppc64_cpu --cores-on=adfd
offlines all but one core
This is because we are not using the endptr parameter of strtol to
determine if any of the string is consumed (and if none of the string is
consumed, strtol still returns a valid long value, namely 0).
ppc64_cpu --cores-on=12df
onlines 12 cores
This is because we are not using the endptr parameter to strtol to
determine if the entire string is consumed (checking for trailing
characters).
Add both these checks.
After this patch:
ppc64_cpu --cores-on=all
Invalid number of cores to online: all
ppc64_cpu --cores-on=adfd
Invalid number of cores to online: adfd
ppc64_cpu --cores-on=12fd
Invalid number of cores to online: 12fd
Signed-off-by: Nishanth Aravamudan <[email protected]>
commit 475919d4294819e5f890f000ccadf5b13280a0a9
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:50:11 2015 -0500
powerpc-utils.spec: Fix typo in spec-file
In line:
%{__rm} -rf $RPM_BULD_ROOT
The directory should be RPM_BUILD_ROOT not RPM_BULD_ROOT
Signed-off-by: Chandni Verma <[email protected]>
commit b1d909378cf8a9502c7207f4f51a2cf3931d5263
Author: Michael Bringmann <[email protected]>
Date: Mon Nov 16 16:48:30 2015 -0500
errinjct/rtas_debug: Update RPM packaging specs
This patch updates the powerpc-utils.spec.in file to install the binaries
associated with errinjct and rtas_dbg.
Signed-off-by: Michael Bringmann <[email protected]>
commit 03704771a36c025eb8636a90f67eaa77acd5880d
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 16 16:40:47 2015 -0500
lsslot/drmgr: Performance enhancement for parsing LMBs
Signed-off-by: Nathan Fontenot <[email protected]>
commit 268116acde0d5f98efa150044b68d977d578ccb4
Author: Vasant Hegde <[email protected]>
Date: Mon Nov 16 16:39:15 2015 -0500
errinjct: Add format specificer
We hit below build warning with "-Werror=format-security" compilation option
CC src/errinjct/icache.o
src/errinjct/errinjct.c: In function 'perr':
src/errinjct/errinjct.c:240:18: error: format not a string literal and no f
fprintf(stderr, buf);
^
Signed-off-by: Vasant Hegde <[email protected]>
commit c2091227bdf40192a37456445a6dedf3f3da8cda
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 16 16:27:08 2015 -0500
drmgr: Do not allow DLPAR of display adapters
The DLPAR of display adapters has never been supported but that support
has not ever been enforced in drmgr. Recent tests show that trying to DLPAR
remove a display adapter causes the drmngr command to appear hung as it
waits for display adapters to power down.
Add the ability to check for a drc-type of 'display' when trying to add
or remove a pci or slot, and verify that none of the children under a
PHB are display adapters when trying to add or remove a PHB.
Signed-off-by: Nathan Fontenot <[email protected]>
commit aef0c5daa20f38b66e5fa0255ecf5a53ba202138
Author: Nathan Fontenot <[email protected]>
Date: Mon Nov 16 16:26:01 2015 -0500
drmgr: Move check for non-hotplug children to common routine
In more than one place we check to see if a PHB has any children that
are not hotplug capable. Move this code to common routine.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 2d8ad5ef33948f07e3022c601ee99cd8efc85a76
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:22:19 2015 -0500
activate_firmware: Have man page reflect the changes made in this branch
-Edited description
-added Option section
-Added 2 new return codes 7 and 8
Signed-off-by: Chandni Verma <[email protected]>
commit 11041ab9e8f0e75b58d2c1a593c9c410d7a82f00
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:21:02 2015 -0500
activate_fw: Return code 8, to report a parameter error when activating firm
This is a new return code mapped from rtas return code -3 which is returned
case of parameter error, when the utility is used without options.
Signed-off-by: Chandni Verma <[email protected]>
commit d1217de2906f22dbeed54f52c683f28ffcfe90c9
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:19:56 2015 -0500
activate_fw: get update access key expiry date, set update access key
This patch provides two feature extensions:
1. activate_firmware -e
(with only -e, fetches the update access key expiry date)
2. activate_firmware -e /path/to/keyfile
(with an argument to -e applies the provided update access key-file to
extend the update access key expiry date)
These are required to provide an interface to check validity of and
update the Update Access Key. ESA will be one of the users of this tool.
Signed-off-by: Chandni Verma <[email protected]>
commit 097c16987a65aceb0be6d66befaa0104b518abbd
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:18:09 2015 -0500
activate_fw.c: Make use of errx() to print error messages to stderr and exit
This is a more compact way to do it compared to using fprintf and explicitly
exiting in main()
Signed-off-by: Chandni Verma <[email protected]>
commit c2dc2191bb5963b09e781b78cbf2a152520c8e87
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:17:21 2015 -0500
activate_fw.c: Factor out activate_firmware() from main()
This is done to reserve main solely for option processing and to
abstract out rtas calling function.
Signed-off-by: Chandni Verma <[email protected]>
commit 72bc893a5f4b59d0e7742726e85f3cffab65ebd3
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 16:14:46 2015 -0500
activate_fw: fix typo
The default case in error handler should return code 6, not 4.
Signed-off-by: Chandni Verma <[email protected]>
commit 782445384c165ea28f00c9cb233e41bdeef7a962
Author: Chandni Verma <[email protected]>
Date: Mon Nov 16 15:53:42 2015 -0500
activate_fw: Coding style fixes
This patch fixes trailing white spaces and indentation issues
and in general makes to code confirm to linux's checkpatch.pl
Signed-off-by: Chandni Verma <[email protected]>
commit ad93027170afe3b7bc760fed0ab3e078390ba088
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:58:44 2015 -0400
serv_config: Compile warning cleanups
Correct several compile warnings.
Remove unchecked return code from execv().
Remove unused variables.
Correct de-reference of type punned pointers.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 8ad7081b5f1d98593c8192d4d46c10050e5b8d88
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:57:34 2015 -0400
rtas_ibm_get_vpd: Compile warning cleanups
Clean up compile warning by removing the unused lflag variable.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 28aeacd59e0e35753bae243cf63150f6ecc2a10b
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:56:32 2015 -0400
rtas_dbg: Compile warning cleanups
Correct compile errors to check the return value from a fread() call and
add an appropriate error message, and ensuring we do not use an
uninitialized 'rc' variable.
Signed-off-by: Nathan Fontenot <[email protected]>
commit dfab2e6f11a11e25da342b703bd054aaf906d818
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:55:33 2015 -0400
ppc64_cpu: Compile warning cleanups
Correct compile warning for dereference of type punned pointer.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 92fd02382fbec464562e2a84841391f09cf1e2ea
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:54:34 2015 -0400
nvram: Compile warning cleanups
Correct compile warning by removing the unused variable memctrl_data.
Signed-off-by: Nathan Fontenot <[email protected]>
commit f069a159755b70bf4596b2bcce5b3fb4f8c3831f
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:53:35 2015 -0400
lparstat: Compile warning cleanups
Clean up compile warnings by adding checks for the return codes from
fgets() calls and adding appropriate error messages.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 2cc2e6590a43a367ee6ca6474bc98cc0a7245caf
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:51:59 2015 -0400
drmgr: Compile warning cleanups
Fix multiple warnings generated when compiling drmgr code. These include
not checking return codes of fread() and write() calls where a check is
added and appropriate error messages.
Updating print formats to use the appropriate type of value being printed.
Correcting the use of rc values to ensure the rc variable is set.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 0e8d1534f5bf57798b74584b6f2b42bad57ddf89
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:49:57 2015 -0400
drmgr: Validate the status of a CPU before adding or removing it
When asked to perform a CPU DLPAR operation we should verify the current
status of the CPU before performing the operation. Failure to verify the
status has caused the kernel to oops in cases where we try to add a CPU
that is already owned, namely in a kvm guest the following is seen.
Signed-off-by: Nathan Fontenot <[email protected]>
commit dabc0264a9f3b64226da63421378ed02f2f1b201
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:48:25 2015 -0400
drmgr: Add missing free's of allocated memory
Using valgrind to do analysis of the memory usage of the drmgr command
revealed multiple instances of memory not being free'ed. Correct this
by adding appropriate calls to free commands.
Note that the analysis done, and this patch, does not cover all of the code
paths in drmgr. There may be additional patches as testing continues.
Signed-off-by: Nathan Fontenot <[email protected]>
commit d5a1c1bfae839e159ba049736a9719835e2a42e6
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:45:26 2015 -0400
drmgr: Add EXTRA_DEBUG output level
Introduce a new internal developer debug output level. In some cases the
drmgr command dumps all the information about slots/adapters it gathers
from the device tree, this is really meant for internal debugging purposes.
Additionally, one large systems this information can easily overwhelm
the output seen from drmgr.
This patches introduces a debug level for the say() command, EXTRA_DEBUG.
Developers can get this output by using the option '-d 5'. Several
instances of debug output are also updated to use this new option.
Signed-off-by: Nathan Fontenot <[email protected]>
commit a7ebbd4bbc61cd0c7d6c3c3eea9f7ea9724558d9
Author: Nathan Fontenot <[email protected]>
Date: Tue Oct 27 16:43:25 2015 -0400
drmgr: free slot nodes before adding slots
When handling a request for slot dlpar add the device tree is parsed twice,
once prior to the add and again after adding the slot. On systems with a
large number of slot devices, 200+, we are seeing that the drmgr command
segfaults when trying to parse the device tree the second time.
To avoid crashing here we need to free the parsed information from the
device tree prior to adding the new slot. While this isn't really a true
root cause fix it does alloow drmgr to run without segfaulting.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 767e3dd780f1a18b2c832b5d8db82a5aac9a9818
Author: Vasant Hegde <[email protected]>
Date: Fri Oct 16 10:01:50 2015 -0400
update_flash: Add support for OpenPower system
On OpenPower system we use IPMI interface to pass image to service processor
(BMC) and it will take care of updating FW.
This patch adds wrapper around ipmitool so that we can use update_flash
command to update FW.
Validation:
BMC take care of validating image. We just pass the image to BMC.
Commit/Reject operation:
Like FSP system, BMC has two sides. Golden and boot side. Golden side
is permanent and is never updated in the field. So commit/reject
operations are not allowed on BMC based system.
FW update:
OpenPower system has two components (BMC FW and host FW). We send all
the component to BMC in one command. BMC will take care of updating
all the components.
Note that we are not using "force" option. So BMC will skip the component
which are not changed.
FW version display:
We display host FW version using ipmitool fru command. like below:
Product Name : OpenPOWER Firmware
Product Version : IBM-firestone-ibm-OP8_PFD_v1.6_0.29
Product Extra : hostboot-6847d73-b8d7c0a
Product Extra : occ-0726e69
Product Extra : skiboot-5.1.3-6221bd2
Product Extra : hostboot-binaries-43d5a59
Product Extra : firestone-xml-db1a93e-4ae8032
Product Extra : capp-ucode-105cb8f
Signed-off-by: Vasant Hegde <[email protected]>
commit 1422abed03f78866339a7b29e3eba4b2d29cb468
Author: Vasant Hegde <[email protected]>
Date: Fri Oct 16 10:01:04 2015 -0400
update_flash_nv: Add platform check
Presently we assume this script is running on FSP based PowerNV platform.
Lets add platform check to validate platform before running. That way
adding new platform support (like OpenPower) becomes easy.
Signed-off-by: Vasant Hegde <[email protected]>
Reviewed-by: Kamalesh Babulal <[email protected]>
commit b7d4678e0963ff4956d02188fae33cb8d5a7fe11
Author: Vasant Hegde <[email protected]>
Date: Fri Oct 16 10:00:28 2015 -0400
update_flash_nv: Rename function name to reflect platform
Presently update_flash_nv supports code update on FSP based PowerNV
(Non-Virtualized) platform. We want to support code update on OpenPower
(BMC based) platform. Code update procedure on OpenPower system is
different than FSP based machine. Hence rename FSP specific function
in this script as fsp_*.
Signed-off-by: Vasant Hegde <[email protected]>
Reviewed-by: Kamalesh Babulal <[email protected]>
commit a3e1c3eddb6e3486ae48dfd23f0c95c4b714322d
Author: Kamalesh Babulal <[email protected]>
Date: Fri Oct 16 09:59:06 2015 -0400
snap: Warn users about deprecated support for SLES 12 onwards
This patch introduce deprecated warning on SLES 12 onwards
to use supportconfig. supportconfig captures all the information
collected by snap, so it safe to deprecate snap.
It also rearranges check for distribution (RHEL/SLES/Ubuntu) into
common function check_distro_support().
Signed-off-by: Kamalesh Babulal <[email protected]>
Cc: Vasant Hegde <[email protected]>
Cc: Nathan Fontenot <[email protected]>
powerpc-utils-1.2.27
=====================================================================
commit 8e22e19dab80a1d17ca5ad5cdf112148282223eb
Author: John Allen <[email protected]>
Date: Tue Sep 29 14:41:14 2015 -0400
errinjct/rtas_dbg: Update Makefile
This patch updates the Makefile to include errinjct and rtas_dbg.
Signed-off-by: John Allen <[email protected]>
commit ef5f204ca2d5de67ef47de1ac29169e7e39b54a6
Author: John Allen <[email protected]>
Date: Tue Sep 29 14:37:12 2015 -0400
rtas_dbg: Initial contribution of the rtas_dbg tool
This patch contributes the previously IBM internal rtas_dbg tool. The rtas_d
tool enables rtas debug output to the system console for a given rtas call.
This version of the code is functionally identical to the old internal versi
but I have made some superficial changes to remove IBM confidential informat
and improve coding style.
Signed-off-by: John Allen <[email protected]>
commit 784885efc522d49e29c3f1fb679a31e7ba9b413d
Author: John Allen <[email protected]>
Date: Tue Sep 29 14:35:14 2015 -0400
errinjct: Initial contribution of errinjct tool
This patch contributes the previously IBM internal errinjct tool. The errinj
tool is used to inject artificial hardware errors into a system. This versio
of the code is functionally identical to the old internal version, but I hav
made some superficial changes to remove IBM confidential information and
improve coding style.
Changes in v2: Now checks for platform. If a user is not on a PowerVM system
the tool exits and prints a descriptive error message. Updated the man page
document lack of support for non-PowerVM systems.
Signed-off-by: John Allen <[email protected]>
commit 16e49f436470d597bb29656ed02b40fcdf47fcc5
Author: Nathan Fontenot <[email protected]>
Date: Wed Sep 9 10:30:53 2015 -0400
lparstat: print "Shared Physical CPUS in system" data
The information for "Shared Physical CPUS in system" is not printed when
the lparstat command is run. This patch adds this information to the list
of items to print and modifies the name to match the output of the lparstat
command on other OS'es.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 4153cccdc21a0c8810aab169fb5b04a611c7b154
Author: Bharata B Rao <[email protected]>
Date: Tue Sep 8 11:38:35 2015 -0400
drmgr: Don't return unsuable CPU from get_available_cpu()
From: Bharata B Rao <[email protected]>
When querying CPU by name or index (and not count), let
get_available_cpu() return NULL for CPU that is marked as unusable.
This will prevent the callers (eg. add_cpus()) from infinitely
querying for the same CPU.
Signed-off-by: Bharata B Rao <[email protected]>
commit d291c43e4ad1141591008e39e77549a6fd33d48a
Author: Nathan Fontenot <[email protected]>
Date: Tue Sep 8 11:35:09 2015 -0400
drmgr: Generate error message when PHB is not found
An error message should be generated when drmgr is asked to remove a PHB
that does not exist. The current behavior is to just exit without any
message.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 8ff5144da07985cd84508465007f9031840d4ecd
Author: Vasant Hegde <[email protected]>
Date: Tue Sep 8 11:33:24 2015 -0400
update_flash: Stop kexec service before starting firmware update on PowerVM
We run 'reboot' call before initiating firmware update. If kexec service
is enabled, then we endup booting into new kernel instead of starting
firmware update.
Commit 62dc6d38 fix this issue on PowerNV platform.. but missed to fix
PowerVM LPAR. This patch fixes above described issue on PowerVM LPAR.
Signed-off-by: Vasant Hegde <[email protected]>
commit 7a94345ff14fbb6d18a885616a5a7f347daa95f1
Author: John Allen <[email protected]>
Date: Wed Sep 2 12:43:15 2015 -0400
drmgr: Fix file leak in dr_lock/dr_unlock interaction
This patch fixes a file leak in dr_lock(). The file descriptor
is meant to be saved to a global variable dr_lock_fd, but we
have an extraneous local variable named dr_lock_fd as well. So
When we get to dr_unlock, the global file descriptor that it
tries to close was never set and quietly fails.
Signed-off-by: John Allen <[email protected]>
commit 29b8f317985249912b4025f704aae16811ea1775
Author: Boqun Feng <[email protected]>
Date: Wed Sep 2 12:19:54 2015 -0400
drmgr: handle read failures in get_att_prop()
In get_att_prop(), though rarely, reading from /proc or /sysfs may
fail. This patch handles failures of these readings based on the fact,
if we successfully read the property, both fread() and fscanf() will
return 1. This patch also updates the fread() path so that we only
try to read the size of the file and not the size of the buffer
passed in.
This can also fix the compiler WARNING saying that 'rc' is unused
variable.
Signed-off-by: Boqun Feng <[email protected]>
Signed-off-by: Nathan Fontenot <[email protected]>
commit 9b8a3b449de4508adca248631e7ffda62a9a4994
Author: John Allen <[email protected]>
Date: Wed Sep 2 12:14:09 2015 -0400
drmgr, ppc64_cpu: Fix several memory/file leaks
This patch fixes a few resource leak problems discoverd through
static analysis.
Signed-off-by: John Allen <[email protected]>
commit 5fee10ac758e16d49e42d5cf0855df44b1f90a7a
Author: Nathan Fontenot <[email protected]>
Date: Wed Sep 2 11:03:28 2015 -0400
drmgr: Combine the init_node() and examine_child() routines
The complexity of the init_node() and examine_child() routines that call
each other recursively is causing drmgr to segfault in odd and sometimes
slightly differing ways. Each instance does involve drmgr segfaulting in
libc malloc code.
This patch re-writes these two routines to combine them into a single
routine that is called recursively. The result is cleaner code that no
longer segfaults.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 96a83ab8b84576517a28bc8f4d8e684e6c86befd
Author: Nathan Fontenot <[email protected]>
Date: Wed Sep 2 10:36:25 2015 -0400
drmgr: Allocate workarea memory instead of using stack
The configure_connector() routine declares a 4K buffer on the stack, this
is causing odd behaviors in drmgr such as segfaults in libc malloc code.
The workarea should be malloc()'ed instead of being on the stack.
Signed-off-by: Nathan Fontenot <[email protected]>
commit 62dc6d38de177588f205e01cec15b93b3e2b9a30
Author: Vasant Hegde <[email protected]>
Date: Wed Sep 2 10:26:23 2015 -0400
update_flash: Stop kexec service before starting firmware update
We run 'reboot' call before initiating firmware update. If kexec service
is enabled, then we endup booting into new kernel instead of starting
firmware update.
This patch validates kexec service and if required stops this service
before starting code update.
Sample output:
update_flash -v -f /home/hegdevasant/01SV840_032_032.img
info: Current fixpack level is unknown.
Projected Flash Update Results:
Current T Image: UNKNOWN
Current P Image: SV840_031
New T Image: SV840_032
New P Image: SV840_031
info: kexec service is running. It will be stopped before flashing.
Signed-off-by: Vasant Hegde <[email protected]>
commit 546174be41c74894126414515644888ed8b31584
Author: Vasant Hegde <[email protected]>
Date: Wed Sep 2 10:23:21 2015 -0400
update_flash: Fix bad variable issue
Shells like dash throws error if we try to assign string with space
without double quote. ex:
local bug=ML FW830.00 FW830.00
Sample output without this patch:
update_flash -f 01SV840_016_016.img
/usr/sbin/update_flash: 185: local: FW830.00: bad variable name
This patch fixes above issue by covering all string assignment with
double quote.
sample output with patch:
info: Current Temporary side will be committed to
Permanent side before being replaced with the new
image.
Projected Flash Update Results:
Current T Image: SV840_031
Current P Image: SV840_027
New T Image: SV840_032
New P Image: SV840_031
Signed-off-by: Vasant Hegde <[email protected]>
commit 2d9c67ba3a0763b152be184f81841a6b8ea5580d
Author: Boqun Feng <[email protected]>
Date: Wed Aug 26 12:13:41 2015 -0400
In cpu_index_to_path(), sometimes 'rc' may not be assigned to a certain
value, so we need to give 'rc' an initial value, as we may only success
if we find a path, so 'rc' is initialized to -1, which indicates
failure.
Signed-off-by: Boqun Feng <[email protected]>
commit b31215f78806bf0b511e39b3c76fc1b12234310a
Author: Boqun Feng <[email protected]>
Date: Wed Aug 26 10:50:30 2015 -0400
In say(), we use variable 'rc' for the return value of write(), however
that return value is never used, though rarely, writting to log_fd may not
results in writting exact number of bytes we want. To be aware of the
return value of write() and also to remove unused variables, we update
'len' to the return value of write() if log_fd is not NULL.
Signed-off-by: Boqun Feng <[email protected]>
commit 8b654c544f5e6243df1a3694162dc466023b39c4
Author: Kamalesh Babulal <[email protected]>
Date: Wed Aug 26 10:48:30 2015 -0400
Fix trivial spelling mistakes and trailing white spaces.
'avaiable' -> 'available'
'parmaeter' -> 'parameter'
'intialize' -> 'initialize'
Signed-off-by: Kamalesh Babulal <[email protected]>
commit f768e8c6465185255142cdfb0577bd3ed729494a
Author: Joel Stanley <[email protected]>
Date: Mon Jun 29 11:07:28 2015 -0400
lsslot: Fix missing header for __be32 types
On my system (Ubuntu 15.04) lsslot does not build as it's missing the
__be32 types.
src/drmgr/lsslot.c: In function ‘print_drconf_mem’:
src/drmgr/lsslot.c:712:2: error: unknown type name ‘__be32’
__be32 *aa;
Signed-off-by: Joel Stanley <[email protected]>
commit b15b0a7f9ffcd1bc099a166723475bb53bc945cf
Author: Joel Stanley <[email protected]>
Date: Mon Jun 29 11:04:24 2015 -0400
nvram: don't allow long names
The maximum partition name length is 12 characters but nvram accepts
strings of any length. This change will enforce the maximum length
passed to options that take a partition name to avoid confusion.
Signed-off-by: Joel Stanley <[email protected]>
commit 9adb99b0490ca249ef1f2edfefe58d1369e2a278
Author: Joel Stanley <[email protected]>
Date: Mon Jun 29 11:03:19 2015 -0400
nvram: fix stack overflow in logging function
Signed-off-by: Joel Stanley <[email protected]>
commit cfce8ea751cd7d87554e59197e8e0c8ab78a8a70
Author: Nathan Fontenot <[email protected]>
Date: Mon Jun 29 11:00:20 2015 -0400
ppc64_cpu: Use calloc for array allocation
Use calloc instead of malloc + memset for array allocation.
Signed-off-by: Nathan Fontenot <[email protected]>
Reviewed-by: Vasant Hegde <[email protected]>
commit b9af10826189fceaa0dd714e793c4753f8241eda
Author: Nathan Fontenot <[email protected]>
Date: Mon Jun 29 10:58:52 2015 -0400
ppc64_cpu: Correct pthread handling for exiting threads
The pthread handling for checking cpu frequency in the ppc64_cpu
command does not properly handle 'soak' threads exiting from a failed
sched_setaffinity call.
We need to update the number of threads handled to be dynamically set
instead of defining MAX_NR_CPUS, which is out of date. We also need to
check for 'soak' threads that have exited before reading the perf counters
Signed-off-by: Nathan Fontenot <[email protected]>
powerpc-utils-1.2.26
=====================================================================
commit 7653d5814de2abb209c198a8b763c99139bf750d
Author: Dinar Valeev <[email protected]>
Date: Wed Jun 17 16:46:48 2015 -0400
PowerKVM is a distro, not a platform
Name OPAL platform correctly.. Bare metal systems are not
PowerKVM (which is IBM distro), but PowerNV.
Signed-off-by: Dinar Valeev <[email protected]>
Reviewed-by: Vasant Hegde <[email protected]>
commit 09630ae59954de064f379f085e710ba34b999ae4
Author: Nathan Fontenot <[email protected]>
Date: Wed Jun 3 08:21:09 2015 -0400
drmgr: Correct memory affinity when adding memory
When adding a LMB we need to update the affinity as part of the add process.
The affinity for the LMB we are adding is conveyed in the
ibm,associativity property for the device tree node of the LMB we are adding
To clarify, a little history. Previously LMB information in the device tree
was done as a node per LMB in the form memory@XXX, and all of the informatio
about the LMB as properties under this node. When we moved to using the
ibm,dynamic-reconfiguration-memory node to convey LMB information the
ibm,dynamic-memory property provides an index for each LMB into the
ibm,associativity-lookup-arrays property. This is where we determine the
affinity for each LMB.
In order to update the affinity for a LMB we are adding we have to call
configure connector to get the memory@XXX node and its properties for the
LMB we are adding (yes, rtas did not change this part), find the
ibm,associativity property, then find the entry in the associativity
lookup array that matches it. We then update the ibm,dynamic-memory
property with the proper index into the associativity lookup array.
Signed-off-by: Nathan Fontenot <[email protected]>
Tested-by: Bharata B Rao <[email protected]>
commit f95a18deaaaa9c67ca1453e6f12784ecebf45e19
Author: Michael Roth <[email protected]>
Date: Wed Jun 3 08:19:04 2015 -0400
drmgr: allow PHBs to be specified using numerical DRC index
For each invocation involving '-c phb', if a numerical value is given
for -s do a lookup through PHB list to find the corresponding DRC