Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Hell of a perfectionist. #99

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Hell of a perfectionist. #99

wants to merge 1 commit into from

Conversation

ghost
Copy link

@ghost ghost commented Jun 5, 2014

No description provided.

Gnurou pushed a commit to Gnurou/linux that referenced this pull request Jun 6, 2014
Turn it into (for example):

[    0.073380] x86: Booting SMP configuration:
[    0.074005] .... node   #0, CPUs:          #1   #2   #3   #4   #5   torvalds#6   torvalds#7
[    0.603005] .... node   #1, CPUs:     torvalds#8   torvalds#9  torvalds#10  torvalds#11  torvalds#12  torvalds#13  torvalds#14  torvalds#15
[    1.200005] .... node   #2, CPUs:    torvalds#16  torvalds#17  torvalds#18  torvalds#19  torvalds#20  torvalds#21  torvalds#22  torvalds#23
[    1.796005] .... node   #3, CPUs:    torvalds#24  torvalds#25  torvalds#26  torvalds#27  torvalds#28  torvalds#29  torvalds#30  torvalds#31
[    2.393005] .... node   #4, CPUs:    torvalds#32  torvalds#33  torvalds#34  torvalds#35  torvalds#36  torvalds#37  torvalds#38  torvalds#39
[    2.996005] .... node   #5, CPUs:    torvalds#40  torvalds#41  torvalds#42  torvalds#43  torvalds#44  torvalds#45  torvalds#46  torvalds#47
[    3.600005] .... node   torvalds#6, CPUs:    torvalds#48  torvalds#49  torvalds#50  torvalds#51  #52  #53  torvalds#54  torvalds#55
[    4.202005] .... node   torvalds#7, CPUs:    torvalds#56  torvalds#57  #58  torvalds#59  torvalds#60  torvalds#61  torvalds#62  torvalds#63
[    4.811005] .... node   torvalds#8, CPUs:    torvalds#64  torvalds#65  torvalds#66  torvalds#67  torvalds#68  torvalds#69  #70  torvalds#71
[    5.421006] .... node   torvalds#9, CPUs:    torvalds#72  torvalds#73  torvalds#74  torvalds#75  torvalds#76  torvalds#77  torvalds#78  torvalds#79
[    6.032005] .... node  torvalds#10, CPUs:    torvalds#80  torvalds#81  torvalds#82  torvalds#83  torvalds#84  torvalds#85  torvalds#86  torvalds#87
[    6.648006] .... node  torvalds#11, CPUs:    torvalds#88  torvalds#89  torvalds#90  torvalds#91  torvalds#92  torvalds#93  torvalds#94  torvalds#95
[    7.262005] .... node  torvalds#12, CPUs:    torvalds#96  torvalds#97  torvalds#98  torvalds#99 torvalds#100 torvalds#101 torvalds#102 torvalds#103
[    7.865005] .... node  torvalds#13, CPUs:   torvalds#104 torvalds#105 torvalds#106 torvalds#107 torvalds#108 torvalds#109 torvalds#110 torvalds#111
[    8.466005] .... node  torvalds#14, CPUs:   torvalds#112 torvalds#113 torvalds#114 torvalds#115 torvalds#116 torvalds#117 torvalds#118 torvalds#119
[    9.073006] .... node  torvalds#15, CPUs:   torvalds#120 torvalds#121 torvalds#122 torvalds#123 torvalds#124 torvalds#125 torvalds#126 torvalds#127
[    9.679901] x86: Booted up 16 nodes, 128 CPUs

and drop useless elements.

Change num_digits() to hpa's division-avoiding, cell-phone-typed
version which he went at great lengths and pains to submit on a
Saturday evening.

Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: huawei.libin@huawei.com
Cc: wangyijing@huawei.com
Cc: fenghua.yu@intel.com
Cc: guohanjun@huawei.com
Cc: paul.gortmaker@windriver.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20130930095624.GB16383@pd.tnic
Signed-off-by: Ingo Molnar <mingo@kernel.org>
@ariccio
Copy link

ariccio commented Jun 17, 2014

Yeah, this is annoying, but _Linus does not merge pull requests from GitHub_.

martinezjavier pushed a commit to martinezjavier/linux that referenced this pull request Jul 30, 2015
WARNING: Missing a blank line after declarations
torvalds#99: FILE: mm/mlock.c:649:
+	vm_flags_t vm_flags = VM_LOCKED;
+	if (flags)

total: 0 errors, 1 warnings, 51 lines checked

./patches/mm-mlock-add-new-mlock-system-call.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: Eric B Munson <emunson@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
ddstreet pushed a commit to ddstreet/linux that referenced this pull request Jul 31, 2015
WARNING: Missing a blank line after declarations
torvalds#99: FILE: mm/mlock.c:649:
+	vm_flags_t vm_flags = VM_LOCKED;
+	if (flags)

total: 0 errors, 1 warnings, 51 lines checked

./patches/mm-mlock-add-new-mlock-system-call.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: Eric B Munson <emunson@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
ddstreet pushed a commit to ddstreet/linux that referenced this pull request Aug 6, 2015
WARNING: Missing a blank line after declarations
torvalds#99: FILE: mm/mlock.c:649:
+	vm_flags_t vm_flags = VM_LOCKED;
+	if (flags)

total: 0 errors, 1 warnings, 51 lines checked

./patches/mm-mlock-add-new-mlock-system-call.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: Eric B Munson <emunson@akamai.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Dec 18, 2015
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 1, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 6, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 13, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 14, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 15, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 21, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 22, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Jan 28, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Feb 1, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Feb 4, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Feb 22, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Feb 29, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Feb 29, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 9, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 11, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 16, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 17, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 18, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 23, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 24, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 28, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Mar 30, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
0day-ci pushed a commit to 0day-ci/linux that referenced this pull request Apr 4, 2016
Cc: David Rientjes <rientjes@google.com>

WARNING: line over 80 characters
torvalds#99: FILE: mm/page_alloc.c:2965:
+ * zone list (with a backoff mechanism which is a function of no_progress_loops).

WARNING: line over 80 characters
torvalds#129: FILE: mm/page_alloc.c:2995:
+	 * Keep reclaiming pages while there is a chance this will lead somewhere.

WARNING: line over 80 characters
torvalds#134: FILE: mm/page_alloc.c:3000:
+	for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, ac->high_zoneidx, ac->nodemask) {

WARNING: line over 80 characters
torvalds#138: FILE: mm/page_alloc.c:3004:
+		available -= DIV_ROUND_UP(no_progress_loops * available, MAX_RECLAIM_RETRIES);

WARNING: line over 80 characters
torvalds#142: FILE: mm/page_alloc.c:3008:
+		 * Would the allocation succeed if we reclaimed the whole available?

WARNING: line over 80 characters
torvalds#146: FILE: mm/page_alloc.c:3012:
+			/* Wait for some write requests to complete then retry */

total: 0 errors, 6 warnings, 202 lines checked

./patches/mm-oom-rework-oom-detection.patch has style problems, please review.

NOTE: If any of the errors are false positives, please report
      them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: David Rientjes <rientjes@google.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 21, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 22, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Feb 23, 2022
[ Upstream commit 17da2d5 ]

As reported:

[  256.104522] ======================================================
[  256.113783] WARNING: possible circular locking dependency detected
[  256.120093] 5.16.0-rc6-yocto-standard+ torvalds#99 Not tainted
[  256.125362] ------------------------------------------------------
[  256.131673] intel-speed-sel/844 is trying to acquire lock:
[  256.137290] ffffffffc036f0d0 (punit_misc_dev_lock){+.+.}-{3:3}, at: isst_if_open+0x18/0x90 [isst_if_common]
[  256.147171]
[  256.147171] but task is already holding lock:
[  256.153135] ffffffff8ee7cb50 (misc_mtx){+.+.}-{3:3}, at: misc_open+0x2a/0x170
[  256.160407]
[  256.160407] which lock already depends on the new lock.
[  256.160407]
[  256.168712]
[  256.168712] the existing dependency chain (in reverse order) is:
[  256.176327]
[  256.176327] -> #1 (misc_mtx){+.+.}-{3:3}:
[  256.181946]        lock_acquire+0x1e6/0x330
[  256.186265]        __mutex_lock+0x9b/0x9b0
[  256.190497]        mutex_lock_nested+0x1b/0x20
[  256.195075]        misc_register+0x32/0x1a0
[  256.199390]        isst_if_cdev_register+0x65/0x180 [isst_if_common]
[  256.205878]        isst_if_probe+0x144/0x16e [isst_if_mmio]
...
[  256.241976]
[  256.241976] -> #0 (punit_misc_dev_lock){+.+.}-{3:3}:
[  256.248552]        validate_chain+0xbc6/0x1750
[  256.253131]        __lock_acquire+0x88c/0xc10
[  256.257618]        lock_acquire+0x1e6/0x330
[  256.261933]        __mutex_lock+0x9b/0x9b0
[  256.266165]        mutex_lock_nested+0x1b/0x20
[  256.270739]        isst_if_open+0x18/0x90 [isst_if_common]
[  256.276356]        misc_open+0x100/0x170
[  256.280409]        chrdev_open+0xa5/0x1e0
...

The call sequence suggested that misc_device /dev file can be opened
before misc device is yet to be registered, which is done only once.

Here punit_misc_dev_lock was used as common lock, to protect the
registration by multiple ISST HW drivers, one time setup, prevent
duplicate registry of misc device and prevent load/unload when device
is open.

We can split into locks:
- One which just prevent duplicate call to misc_register() and one
time setup. Also never call again if the misc_register() failed or
required one time setup is failed. This lock is not shared with
any misc device callbacks.

- The other lock protects registry, load and unload of HW drivers.

Sequence in isst_if_cdev_register()
- Register callbacks under punit_misc_dev_open_lock
- Call isst_misc_reg() which registers misc_device on the first
registry which is under punit_misc_dev_reg_lock, which is not
shared with callbacks.

Sequence in isst_if_cdev_unregister
Just opposite of isst_if_cdev_register

Reported-and-tested-by: Liwei Song <liwei.song@windriver.com>
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://lore.kernel.org/r/20220112022521.54669-1-srinivas.pandruvada@linux.intel.com
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
akiernan pushed a commit to zuma-array/linux that referenced this pull request Nov 3, 2022
driver defect clean up:
torvalds#40
torvalds#41
torvalds#99
torvalds#100
torvalds#395
torvalds#396
torvalds#475
torvalds#614
torvalds#669

Change-Id: I581aaa8a1b950278bbf74d0c94aa647de89e07a9
Signed-off-by: Evoke Zhang <evoke.zhang@amlogic.com>
akiernan pushed a commit to zuma-array/linux that referenced this pull request Nov 4, 2022
driver defect clean up:
torvalds#40
torvalds#41
torvalds#99
torvalds#100
torvalds#395
torvalds#396
torvalds#475
torvalds#614
torvalds#669

Change-Id: I581aaa8a1b950278bbf74d0c94aa647de89e07a9
Signed-off-by: Evoke Zhang <evoke.zhang@amlogic.com>
intel-lab-lkp pushed a commit to intel-lab-lkp/linux that referenced this pull request Mar 17, 2023
Currently, test_progs outputs all stdout/stderr as it runs, and when it
is done, prints a summary.

It is non-trivial for tooling to parse that output and extract meaningful
information from it.

This change adds a new option, `--json-summary`/`-J` that let the caller
specify a file where `test_progs{,-no_alu32}` can write a summary of the
run in a json format that can later be parsed by tooling.

Currently, it creates a summary section with successes/skipped/failures
followed by a list of failed tests and subtests.

A test contains the following fields:
- name: the name of the test
- number: the number of the test
- message: the log message that was printed by the test.
- failed: A boolean indicating whether the test failed or not. Currently
we only output failed tests, but in the future, successful tests could
be added.
- subtests: A list of subtests associated with this test.

A subtest contains the following fields:
- name: same as above
- number: sanme as above
- message: the log message that was printed by the subtest.
- failed: same as above but for the subtest

An example run and json content below:
```
$ sudo ./test_progs -a $(grep -v '^#' ./DENYLIST.aarch64 | awk '{print
$1","}' | tr -d '\n') -j -J /tmp/test_progs.json
$ jq < /tmp/test_progs.json | head -n 30
{
  "success": 29,
  "success_subtest": 23,
  "skipped": 3,
  "failed": 28,
  "results": [
    {
      "name": "bpf_cookie",
      "number": 10,
      "message": "test_bpf_cookie:PASS:skel_open 0 nsec\n",
      "failed": true,
      "subtests": [
        {
          "name": "multi_kprobe_link_api",
          "number": 2,
          "message": "kprobe_multi_link_api_subtest:PASS:load_kallsyms 0
nsec\nlibbpf: extern 'bpf_testmod_fentry_test1' (strong): not
resolved\nlibbpf: failed to load object 'kprobe_multi'\nlibbpf: failed
to load BPF skeleton 'kprobe_multi':
-3\nkprobe_multi_link_api_subtest:FAIL:fentry_raw_skel_load unexpected
error: -3\n",
          "failed": true
        },
        {
          "name": "multi_kprobe_attach_api",
          "number": 3,
          "message": "libbpf: extern 'bpf_testmod_fentry_test1'
(strong): not resolved\nlibbpf: failed to load object
'kprobe_multi'\nlibbpf: failed to load BPF skeleton 'kprobe_multi':
-3\nkprobe_multi_attach_api_subtest:FAIL:fentry_raw_skel_load unexpected
error: -3\n",
          "failed": true
        },
        {
          "name": "lsm",
          "number": 8,
          "message": "lsm_subtest:PASS:lsm.link_create 0
nsec\nlsm_subtest:FAIL:stack_mprotect unexpected stack_mprotect: actual
0 != expected -1\n",
          "failed": true
        }
```

The file can then be used to print a summary of the test run and list of
failing tests/subtests:

```
$ jq -r < /tmp/test_progs.json '"Success:
\(.success)/\(.success_subtest), Skipped: \(.skipped), Failed:
\(.failed)"'

Success: 29/23, Skipped: 3, Failed: 28
$ jq -r < /tmp/test_progs.json '.results | map([
    if .failed then "#\(.number) \(.name)" else empty end,
    (
        . as {name: $tname, number: $tnum} | .subtests | map(
            if .failed then "#\($tnum)/\(.number) \($tname)/\(.name)"
else empty end
        )
    )
]) | flatten | .[]' | head -n 20
 torvalds#10 bpf_cookie
 torvalds#10/2 bpf_cookie/multi_kprobe_link_api
 torvalds#10/3 bpf_cookie/multi_kprobe_attach_api
 torvalds#10/8 bpf_cookie/lsm
 torvalds#15 bpf_mod_race
 torvalds#15/1 bpf_mod_race/ksym (used_btfs UAF)
 torvalds#15/2 bpf_mod_race/kfunc (kfunc_btf_tab UAF)
 torvalds#36 cgroup_hierarchical_stats
 torvalds#61 deny_namespace
 torvalds#61/1 deny_namespace/unpriv_userns_create_no_bpf
 torvalds#73 fexit_stress
 torvalds#83 get_func_ip_test
 torvalds#99 kfunc_dynptr_param
 torvalds#99/1 kfunc_dynptr_param/dynptr_data_null
 torvalds#99/4 kfunc_dynptr_param/dynptr_data_null
 torvalds#100 kprobe_multi_bench_attach
 torvalds#100/1 kprobe_multi_bench_attach/kernel
 torvalds#100/2 kprobe_multi_bench_attach/modules
 torvalds#101 kprobe_multi_test
 torvalds#101/1 kprobe_multi_test/skel_api
```

Signed-off-by: Manu Bretelle <chantr4@gmail.com>
ammarfaizi2 pushed a commit to ammarfaizi2/linux-fork that referenced this pull request Mar 17, 2023
Currently, test_progs outputs all stdout/stderr as it runs, and when it
is done, prints a summary.

It is non-trivial for tooling to parse that output and extract meaningful
information from it.

This change adds a new option, `--json-summary`/`-J` that let the caller
specify a file where `test_progs{,-no_alu32}` can write a summary of the
run in a json format that can later be parsed by tooling.

Currently, it creates a summary section with successes/skipped/failures
followed by a list of failed tests and subtests.

A test contains the following fields:
- name: the name of the test
- number: the number of the test
- message: the log message that was printed by the test.
- failed: A boolean indicating whether the test failed or not. Currently
we only output failed tests, but in the future, successful tests could
be added.
- subtests: A list of subtests associated with this test.

A subtest contains the following fields:
- name: same as above
- number: sanme as above
- message: the log message that was printed by the subtest.
- failed: same as above but for the subtest

An example run and json content below:
```
$ sudo ./test_progs -a $(grep -v '^#' ./DENYLIST.aarch64 | awk '{print
$1","}' | tr -d '\n') -j -J /tmp/test_progs.json
$ jq < /tmp/test_progs.json | head -n 30
{
  "success": 29,
  "success_subtest": 23,
  "skipped": 3,
  "failed": 28,
  "results": [
    {
      "name": "bpf_cookie",
      "number": 10,
      "message": "test_bpf_cookie:PASS:skel_open 0 nsec\n",
      "failed": true,
      "subtests": [
        {
          "name": "multi_kprobe_link_api",
          "number": 2,
          "message": "kprobe_multi_link_api_subtest:PASS:load_kallsyms 0 nsec\nlibbpf: extern 'bpf_testmod_fentry_test1' (strong): not resolved\nlibbpf: failed to load object 'kprobe_multi'\nlibbpf: failed to load BPF skeleton 'kprobe_multi': -3\nkprobe_multi_link_api_subtest:FAIL:fentry_raw_skel_load unexpected error: -3\n",
          "failed": true
        },
        {
          "name": "multi_kprobe_attach_api",
          "number": 3,
          "message": "libbpf: extern 'bpf_testmod_fentry_test1' (strong): not resolved\nlibbpf: failed to load object 'kprobe_multi'\nlibbpf: failed to load BPF skeleton 'kprobe_multi': -3\nkprobe_multi_attach_api_subtest:FAIL:fentry_raw_skel_load unexpected error: -3\n",
          "failed": true
        },
        {
          "name": "lsm",
          "number": 8,
          "message": "lsm_subtest:PASS:lsm.link_create 0 nsec\nlsm_subtest:FAIL:stack_mprotect unexpected stack_mprotect: actual 0 != expected -1\n",
          "failed": true
        }
```

The file can then be used to print a summary of the test run and list of
failing tests/subtests:

```
$ jq -r < /tmp/test_progs.json '"Success: \(.success)/\(.success_subtest), Skipped: \(.skipped), Failed: \(.failed)"'

Success: 29/23, Skipped: 3, Failed: 28
$ jq -r < /tmp/test_progs.json '.results | map([
    if .failed then "#\(.number) \(.name)" else empty end,
    (
        . as {name: $tname, number: $tnum} | .subtests | map(
            if .failed then "#\($tnum)/\(.number) \($tname)/\(.name)" else empty end
        )
    )
]) | flatten | .[]' | head -n 20
 torvalds#10 bpf_cookie
 torvalds#10/2 bpf_cookie/multi_kprobe_link_api
 torvalds#10/3 bpf_cookie/multi_kprobe_attach_api
 torvalds#10/8 bpf_cookie/lsm
 torvalds#15 bpf_mod_race
 torvalds#15/1 bpf_mod_race/ksym (used_btfs UAF)
 torvalds#15/2 bpf_mod_race/kfunc (kfunc_btf_tab UAF)
 torvalds#36 cgroup_hierarchical_stats
 torvalds#61 deny_namespace
 torvalds#61/1 deny_namespace/unpriv_userns_create_no_bpf
 torvalds#73 fexit_stress
 torvalds#83 get_func_ip_test
 torvalds#99 kfunc_dynptr_param
 torvalds#99/1 kfunc_dynptr_param/dynptr_data_null
 torvalds#99/4 kfunc_dynptr_param/dynptr_data_null
 torvalds#100 kprobe_multi_bench_attach
 torvalds#100/1 kprobe_multi_bench_attach/kernel
 torvalds#100/2 kprobe_multi_bench_attach/modules
 torvalds#101 kprobe_multi_test
 torvalds#101/1 kprobe_multi_test/skel_api
```

Signed-off-by: Manu Bretelle <chantr4@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230317163256.3809328-1-chantr4@gmail.com
intel-lab-lkp pushed a commit to intel-lab-lkp/linux that referenced this pull request Jul 28, 2023
The following test_verifier subtest failed due to
new encoding for BSWAP.

  $ ./test_verifier
  ...
  torvalds#99/u invalid 64-bit BPF_END FAIL
  Unexpected success to load!
  verification time 215 usec
  stack depth 0
  processed 3 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0
  torvalds#99/p invalid 64-bit BPF_END FAIL
  Unexpected success to load!
  verification time 198 usec
  stack depth 0
  processed 3 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0

Tighten the test so it still reports a failure.

Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20230728011244.3717464-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants