Date   

Re: SWAT Rotation schedule

Alexandre Belloni
 

On 23/09/2021 05:15:06+0000, Oleksiy Obitotskyy via lists.yoctoproject.org wrote:
Hi,

I will start as planned from 24/09/2021.
Thanks, there are currently 7 failures that are not triaged. The two
reproducible issues may already be fixed and we'll know that once the
nightly builds have finished.

Unfortunately, the logs for the qemuarm64 failure are already gone are
there is already a (failing) qemuarm64 build running on the same worker.
The four other ones are one off issues that didn't repeat, I'll discuss
with Richard what to do with them.

Regards,

--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: SWAT Rotation schedule

Oleksiy Obitotskyy
 

Hi,

I will start as planned from 24/09/2021.

Regards,
Oleksiy

________________________________________
From: swat@lists.yoctoproject.org <swat@lists.yoctoproject.org> on behalf of Alexandre Belloni <alexandre.belloni@bootlin.com>
Sent: Friday, September 17, 2021 00:29
To: Alejandro Hernandez Samaniego; swat@lists.yoctoproject.org
Subject: [swat] SWAT Rotation schedule

Hello,

Following the summer break, I would like to organize SWAT duty rotation.
This time, I've prepared a semi randomized schedule so it is easier for
each one of you to plan for SWAT duty.

Please check the table and let me know if you are not available for the
selected week. SWAT duty will be from Friday to Thursday and the goal is
to triage all the failures on swatbot before the weekly triage call
happening at 2:30pm UTC.

┌──────┬────────────┐
│ Week │ Start │
┌───────────────────────────────┼──────┼────────────┤
│ Alejandro Hernandez Samaniego │ 37 │ 17/09/2021 │
│ Oleksiy Obitotskyy │ 38 │ 24/09/2021 │
│ Naveen Saini │ 39 │ 01/10/2021 │
│ Thomas Perrot │ 40 │ 08/10/2021 │
│ Paul Eggleton │ 41 │ 15/10/2021 │
│ Christopher Larson │ 42 │ 22/10/2021 │
│ Jon Mason │ 43 │ 29/10/2021 │
│ Lee Chee Yang │ 44 │ 05/11/2021 │
│ Minjae Kim │ 45 │ 12/11/2021 │
│ Jaga │ 46 │ 19/11/2021 │
│ Leo Sandoval │ 47 │ 26/11/2021 │
│ Ross Burton │ 48 │ 03/12/2021 │
│ Köry Maincent │ 49 │ 10/12/2021 │
│ Anibal Limon │ 50 │ 17/12/2021 │
│ Saul Wold │ 51 │ 24/12/2021 │
└───────────────────────────────┴──────┴────────────┘

Alejandro you would be the next one on the list, starting this Friday,
can you confirm you are available?

There are currently 8 failures to triage on swatbot, I'm going to take
care of those.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: [EXTERNAL] [swat] SWAT Rotation schedule

Alejandro Hernandez Samaniego
 

Hello Alex,

Sadly I'm not available this week, I apologize for the late reply but I've been out, can you reschedule me for a different week?

Alejandro


From: swat@... <swat@...> on behalf of Alexandre Belloni via lists.yoctoproject.org <alexandre.belloni=bootlin.com@...>
Sent: Thursday, 16 September 2021, 5:16 pm
To: Alejandro Hernandez Samaniego; swat@...
Subject: [EXTERNAL] [swat] SWAT Rotation schedule

Hello,

Following the summer break, I would like to organize SWAT duty rotation.
This time, I've prepared a semi randomized schedule so it is easier for
each one of you to plan for SWAT duty.

Please check the table and let me know if you are not available for the
selected week. SWAT duty will be from Friday to Thursday and the goal is
to triage all the failures on swatbot before the weekly triage call
happening at 2:30pm UTC.

                                ┌──────┬────────────┐
                                │ Week │   Start    │
┌───────────────────────────────┼──────┼────────────┤
│ Alejandro Hernandez Samaniego │   37 │ 17/09/2021 │
│ Oleksiy Obitotskyy            │   38 │ 24/09/2021 │
│ Naveen Saini                  │   39 │ 01/10/2021 │
│ Thomas Perrot                 │   40 │ 08/10/2021 │
│ Paul Eggleton                 │   41 │ 15/10/2021 │
│ Christopher Larson            │   42 │ 22/10/2021 │
│ Jon Mason                     │   43 │ 29/10/2021 │
│ Lee Chee Yang                 │   44 │ 05/11/2021 │
│ Minjae Kim                    │   45 │ 12/11/2021 │
│ Jaga                          │   46 │ 19/11/2021 │
│ Leo Sandoval                  │   47 │ 26/11/2021 │
│ Ross Burton                   │   48 │ 03/12/2021 │
│ Köry Maincent                 │   49 │ 10/12/2021 │
│ Anibal Limon                  │   50 │ 17/12/2021 │
│ Saul Wold                     │   51 │ 24/12/2021 │
└───────────────────────────────┴──────┴────────────┘

Alejandro you would be the next one on the list, starting this Friday,
can you confirm you are available?

There are currently 8 failures to triage on swatbot, I'm going to take
care of those.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbootlin.com%2F&amp;data=04%7C01%7CAlejandro.Hernandez%40microsoft.com%7C5d12ddae8d8a43bf0e2208d9796806ed%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637674309971395320%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&amp;sdata=QTq%2FDzzYKP0TblZjXG4q3i1x2DRXUY6JwkHJGmjkylU%3D&amp;reserved=0







Re: SWAT Rotation schedule

Alexandre Belloni
 

On 17/09/2021 08:55:10-0700, Saul Wold wrote:


On 9/16/21 2:29 PM, Alexandre Belloni wrote:
Hello,

Following the summer break, I would like to organize SWAT duty rotation.
This time, I've prepared a semi randomized schedule so it is easier for
each one of you to plan for SWAT duty.

Please check the table and let me know if you are not available for the
selected week. SWAT duty will be from Friday to Thursday and the goal is
to triage all the failures on swatbot before the weekly triage call
happening at 2:30pm UTC.
Hi Alexandre,

Great work on creating a schedule that might be more predictable! Just a
heads up, the date you have me is right over the Christmas break. I am not
saying drop me as I might be working anyway, but when things get closer I
will let you know if I have other plans (going skiing!).
Sure, I realize that and I was thinking probably there will be less
builds at that time anyway!

Sau!

┌──────┬────────────┐
│ Week │ Start │
┌───────────────────────────────┼──────┼────────────┤
│ Alejandro Hernandez Samaniego │ 37 │ 17/09/2021 │
│ Oleksiy Obitotskyy │ 38 │ 24/09/2021 │
│ Naveen Saini │ 39 │ 01/10/2021 │
│ Thomas Perrot │ 40 │ 08/10/2021 │
│ Paul Eggleton │ 41 │ 15/10/2021 │
│ Christopher Larson │ 42 │ 22/10/2021 │
│ Jon Mason │ 43 │ 29/10/2021 │
│ Lee Chee Yang │ 44 │ 05/11/2021 │
│ Minjae Kim │ 45 │ 12/11/2021 │
│ Jaga │ 46 │ 19/11/2021 │
│ Leo Sandoval │ 47 │ 26/11/2021 │
│ Ross Burton │ 48 │ 03/12/2021 │
│ Köry Maincent │ 49 │ 10/12/2021 │
│ Anibal Limon │ 50 │ 17/12/2021 │
│ Saul Wold │ 51 │ 24/12/2021 │
└───────────────────────────────┴──────┴────────────┘

Alejandro you would be the next one on the list, starting this Friday,
can you confirm you are available?

There are currently 8 failures to triage on swatbot, I'm going to take
care of those.

Thanks!
--
Sau!




--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: SWAT Rotation schedule

Saul Wold
 

On 9/16/21 2:29 PM, Alexandre Belloni wrote:
Hello,
Following the summer break, I would like to organize SWAT duty rotation.
This time, I've prepared a semi randomized schedule so it is easier for
each one of you to plan for SWAT duty.
Please check the table and let me know if you are not available for the
selected week. SWAT duty will be from Friday to Thursday and the goal is
to triage all the failures on swatbot before the weekly triage call
happening at 2:30pm UTC.
Hi Alexandre,

Great work on creating a schedule that might be more predictable! Just a heads up, the date you have me is right over the Christmas break. I am not saying drop me as I might be working anyway, but when things get closer I will let you know if I have other plans (going skiing!).

Sau!

┌──────┬────────────┐
│ Week │ Start │
┌───────────────────────────────┼──────┼────────────┤
│ Alejandro Hernandez Samaniego │ 37 │ 17/09/2021 │
│ Oleksiy Obitotskyy │ 38 │ 24/09/2021 │
│ Naveen Saini │ 39 │ 01/10/2021 │
│ Thomas Perrot │ 40 │ 08/10/2021 │
│ Paul Eggleton │ 41 │ 15/10/2021 │
│ Christopher Larson │ 42 │ 22/10/2021 │
│ Jon Mason │ 43 │ 29/10/2021 │
│ Lee Chee Yang │ 44 │ 05/11/2021 │
│ Minjae Kim │ 45 │ 12/11/2021 │
│ Jaga │ 46 │ 19/11/2021 │
│ Leo Sandoval │ 47 │ 26/11/2021 │
│ Ross Burton │ 48 │ 03/12/2021 │
│ Köry Maincent │ 49 │ 10/12/2021 │
│ Anibal Limon │ 50 │ 17/12/2021 │
│ Saul Wold │ 51 │ 24/12/2021 │
└───────────────────────────────┴──────┴────────────┘
Alejandro you would be the next one on the list, starting this Friday,
can you confirm you are available?
There are currently 8 failures to triage on swatbot, I'm going to take
care of those.
Thanks!
--
Sau!


SWAT Rotation schedule

Alexandre Belloni
 

Hello,

Following the summer break, I would like to organize SWAT duty rotation.
This time, I've prepared a semi randomized schedule so it is easier for
each one of you to plan for SWAT duty.

Please check the table and let me know if you are not available for the
selected week. SWAT duty will be from Friday to Thursday and the goal is
to triage all the failures on swatbot before the weekly triage call
happening at 2:30pm UTC.

┌──────┬────────────┐
│ Week │ Start │
┌───────────────────────────────┼──────┼────────────┤
│ Alejandro Hernandez Samaniego │ 37 │ 17/09/2021 │
│ Oleksiy Obitotskyy │ 38 │ 24/09/2021 │
│ Naveen Saini │ 39 │ 01/10/2021 │
│ Thomas Perrot │ 40 │ 08/10/2021 │
│ Paul Eggleton │ 41 │ 15/10/2021 │
│ Christopher Larson │ 42 │ 22/10/2021 │
│ Jon Mason │ 43 │ 29/10/2021 │
│ Lee Chee Yang │ 44 │ 05/11/2021 │
│ Minjae Kim │ 45 │ 12/11/2021 │
│ Jaga │ 46 │ 19/11/2021 │
│ Leo Sandoval │ 47 │ 26/11/2021 │
│ Ross Burton │ 48 │ 03/12/2021 │
│ Köry Maincent │ 49 │ 10/12/2021 │
│ Anibal Limon │ 50 │ 17/12/2021 │
│ Saul Wold │ 51 │ 24/12/2021 │
└───────────────────────────────┴──────┴────────────┘

Alejandro you would be the next one on the list, starting this Friday,
can you confirm you are available?

There are currently 8 failures to triage on swatbot, I'm going to take
care of those.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: SWAT Rotation

Saul Wold
 

On 8/30/21 1:26 AM, Alexandre Belloni wrote:
Hello Saul,
As we discussed las Thursday, I count on you to take care of SWAT this
week.
ACK

I'll take care of 211545 and earlier when I'll be back so you don't
necessarily have to spend time on those.
I will look at the newer ones later this morning.

Sau!

Thanks!
--
Sau!


SWAT Rotation

Alexandre Belloni
 

Hello Saul,

As we discussed las Thursday, I count on you to take care of SWAT this
week.
I'll take care of 211545 and earlier when I'll be back so you don't
necessarily have to spend time on those.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


[swatbot] [PATCH] models/serializers: triage exception automatically

Alexandre Belloni
 

Most of the cancelled builds are actually exceptions, tirage them
automatically.

Signed-off-by: Alexandre Belloni <alexandre.belloni@bootlin.com>
---
swatapp/serializers.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/swatapp/serializers.py b/swatapp/serializers.py
index cba4fa482079..e5ca25e4442f 100644
--- a/swatapp/serializers.py
+++ b/swatapp/serializers.py
@@ -28,7 +28,7 @@ class StepFailureSerializer(serializers.ModelSerializer):
attrs['triage'] = TriageStatus.NOTFORSWAT

# Cancelled builds with only stdio (no errors/warnings) aren't interesting
- if attrs['status'] == Status.CANCELLED:
+ if attrs['status'] == Status.CANCELLED or attrs['status'] == Status.EXCEPTION:
if "/errors" not in attrs['urls'] and "/warnings" not in attrs['urls']:
attrs['triage'] = TriageStatus.CANCELLEDNOERR
return attrs
--
2.31.1


Re: SWAT Rotation

Anibal Limon
 

Hi Alexandre,

Unfortunately I will be out next two weeks for summer vacation. 

Sorry for the inconvenience.

Best regards!,
Anibal

On Fri, 13 Aug 2021 at 09:14, Alexandre Belloni <alexandre.belloni@...> wrote:
Hello Anibal,

You are the next one on the SWAT Team members list
(https://wiki.yoctoproject.org/wiki/Yocto_Build_Failure_Swat_Team#Members)

Are you available next week for SWAT duty?

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


SWAT Rotation

Alexandre Belloni
 

Hello Anibal,

You are the next one on the SWAT Team members list
(https://wiki.yoctoproject.org/wiki/Yocto_Build_Failure_Swat_Team#Members)

Are you available next week for SWAT duty?

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: glibc 2.34 tests

Randy MacLeod
 

Add  Tony who has been working on valgrind so he and Khem keep in sync.

../Randy

On 2021-07-06 5:01 a.m., Khem Raj wrote:
On Sat, Jul 3, 2021 at 2:29 PM Alexandre Belloni
<alexandre.belloni@bootlin.com> wrote:
Hello,

I've been testing multiple versions of Khem's branch for glibc2.34
Everything apart from non-gpl3 builds fine. There is a patch from Khem
to solve this issue, I didn't apply/test yet:

[yocto] [meta-gplv2] [PATCH] coreutils_6.9.bb: Fix build with glibc 2.34

On the tests side, systemd-journal was crashing, the last patch on
Khem's branch is solving the issue, thanks!

The remaining issues are on the ptests side:

https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1946/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2236/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1949/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2244/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1957/steps/12/logs/stdio

Issues happen both on qemux86-64 and qemuarm64 and they reproduced
every time:

'elfutils': ['run-backtrace-native.sh', 'run-backtrace-native-core.sh']

'strace': ['prlimit64.gen.test',
'xet_robust_list.gen.test',
'xet_robust_list--pidns-translation.gen.test'],

For valgrind, it is all over the place:

ARM has:
'valgrind': ['gdbserver_tests/hginfo',
'helgrind/tests/hg02_deadlock',
'drd/tests/boost_thread',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/swapcontext',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']

x86 has:

'valgrind': ['gdbserver_tests/hginfo',
'memcheck/tests/supp_unknown',
'helgrind/tests/annotate_smart_pointer',
'helgrind/tests/bar_bad',
'helgrind/tests/hg01_all_ok',
'helgrind/tests/hg02_deadlock',
'helgrind/tests/pth_destroy_cond',
'helgrind/tests/rwlock_test',
'helgrind/tests/tc05_simple_race',
'helgrind/tests/tc06_two_races',
'helgrind/tests/tc06_two_races_xml',
'helgrind/tests/tc14_laog_dinphils',
'helgrind/tests/tc17_sembar',
'helgrind/tests/tc19_shadowmem',
'helgrind/tests/tc21_pthonce',
'helgrind/tests/tc22_exit_w_lock',
'helgrind/tests/tc23_bogus_condwait',
'helgrind/tests/tls_threads',
'drd/tests/annotate_barrier',
'drd/tests/annotate_barrier_xml',
'drd/tests/annotate_hb_race',
'drd/tests/annotate_hbefore',
'drd/tests/annotate_ignore_read',
'drd/tests/annotate_ignore_rw',
'drd/tests/annotate_ignore_rw2',
'drd/tests/annotate_ignore_write',
'drd/tests/annotate_ignore_write2',
'drd/tests/annotate_order_1',
'drd/tests/annotate_order_2',
'drd/tests/annotate_order_3',
'drd/tests/annotate_rwlock',
'drd/tests/annotate_sem',
'drd/tests/annotate_smart_pointer',
'drd/tests/annotate_smart_pointer2',
'drd/tests/annotate_spinlock',
'drd/tests/annotate_static',
'drd/tests/atomic_var',
'drd/tests/bar_bad',
'drd/tests/bar_bad_xml',
'drd/tests/bar_trivial',
'drd/tests/boost_thread',
'drd/tests/bug-235681',
'drd/tests/bug322621',
'drd/tests/circular_buffer',
'drd/tests/concurrent_close',
'drd/tests/dlopen',
'drd/tests/fork-parallel',
'drd/tests/fork-serial',
'drd/tests/fp_race',
'drd/tests/fp_race2',
'drd/tests/fp_race_xml',
'drd/tests/free_is_write',
'drd/tests/free_is_write2',
'drd/tests/hg01_all_ok',
'drd/tests/hg02_deadlock',
'drd/tests/hg03_inherit',
'drd/tests/hg04_race',
'drd/tests/hg05_race2',
'drd/tests/hg06_readshared',
'drd/tests/linuxthreads_det',
'drd/tests/matinv',
'drd/tests/monitor_example',
'drd/tests/pth_barrier',
'drd/tests/pth_barrier2',
'drd/tests/pth_barrier3',
'drd/tests/pth_barrier_race',
'drd/tests/pth_barrier_thr_cr',
'drd/tests/pth_broadcast',
'drd/tests/pth_cancel_locked',
'drd/tests/pth_cleanup_handler',
'drd/tests/pth_cond_race',
'drd/tests/pth_cond_race2',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/pth_detached2',
'drd/tests/pth_detached3',
'drd/tests/pth_detached_sem',
'drd/tests/pth_inconsistent_cond_wait',
'drd/tests/pth_once',
'drd/tests/pth_spinlock',
'drd/tests/read_and_free_race',
'drd/tests/rwlock_race',
'drd/tests/rwlock_test',
'drd/tests/sem_as_mutex',
'drd/tests/sem_as_mutex2',
'drd/tests/sem_as_mutex3',
'drd/tests/sem_open',
'drd/tests/sem_open2',
'drd/tests/sem_open3',
'drd/tests/sem_open_traced',
'drd/tests/sigalrm',
'drd/tests/std_atomic',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/std_string',
'drd/tests/std_thread',
'drd/tests/swapcontext',
'drd/tests/tc01_simple_race',
'drd/tests/tc02_simple_tls',
'drd/tests/tc03_re_excl',
'drd/tests/tc05_simple_race',
'drd/tests/tc06_two_races',
'drd/tests/tc07_hbl1',
'drd/tests/tc08_hbl2',
'drd/tests/tc09_bad_unlock',
'drd/tests/tc11_XCHG',
'drd/tests/tc16_byterace',
'drd/tests/tc17_sembar',
'drd/tests/tc18_semabuse',
'drd/tests/tc19_shadowmem',
'drd/tests/tc21_pthonce',
'drd/tests/tc22_exit_w_lock',
'drd/tests/tc23_bogus_condwait',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']


It seems that
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233
was under more load has it had python3 fail and more strace failures.

I'll triage the remaining intermittent ptest failures

thanks, the good thing is that these packages are quite sticky to
glibc internals so usually an upgrade if available
for them will sort things out, glibc is merging libpthread/librt etc
into libc so lot of valgrind failures could be due to that
I will see what I can find.

--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
--
# Randy MacLeod
# Wind River Linux


Re: glibc 2.34 tests

Khem Raj <raj.khem@...>
 

On Sat, Jul 3, 2021 at 2:29 PM Alexandre Belloni
<alexandre.belloni@bootlin.com> wrote:

Hello,

I've been testing multiple versions of Khem's branch for glibc2.34
Everything apart from non-gpl3 builds fine. There is a patch from Khem
to solve this issue, I didn't apply/test yet:

[yocto] [meta-gplv2] [PATCH] coreutils_6.9.bb: Fix build with glibc 2.34

On the tests side, systemd-journal was crashing, the last patch on
Khem's branch is solving the issue, thanks!

The remaining issues are on the ptests side:

https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1946/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2236/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1949/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2244/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1957/steps/12/logs/stdio

Issues happen both on qemux86-64 and qemuarm64 and they reproduced
every time:

'elfutils': ['run-backtrace-native.sh', 'run-backtrace-native-core.sh']

'strace': ['prlimit64.gen.test',
'xet_robust_list.gen.test',
'xet_robust_list--pidns-translation.gen.test'],

For valgrind, it is all over the place:

ARM has:
'valgrind': ['gdbserver_tests/hginfo',
'helgrind/tests/hg02_deadlock',
'drd/tests/boost_thread',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/swapcontext',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']

x86 has:

'valgrind': ['gdbserver_tests/hginfo',
'memcheck/tests/supp_unknown',
'helgrind/tests/annotate_smart_pointer',
'helgrind/tests/bar_bad',
'helgrind/tests/hg01_all_ok',
'helgrind/tests/hg02_deadlock',
'helgrind/tests/pth_destroy_cond',
'helgrind/tests/rwlock_test',
'helgrind/tests/tc05_simple_race',
'helgrind/tests/tc06_two_races',
'helgrind/tests/tc06_two_races_xml',
'helgrind/tests/tc14_laog_dinphils',
'helgrind/tests/tc17_sembar',
'helgrind/tests/tc19_shadowmem',
'helgrind/tests/tc21_pthonce',
'helgrind/tests/tc22_exit_w_lock',
'helgrind/tests/tc23_bogus_condwait',
'helgrind/tests/tls_threads',
'drd/tests/annotate_barrier',
'drd/tests/annotate_barrier_xml',
'drd/tests/annotate_hb_race',
'drd/tests/annotate_hbefore',
'drd/tests/annotate_ignore_read',
'drd/tests/annotate_ignore_rw',
'drd/tests/annotate_ignore_rw2',
'drd/tests/annotate_ignore_write',
'drd/tests/annotate_ignore_write2',
'drd/tests/annotate_order_1',
'drd/tests/annotate_order_2',
'drd/tests/annotate_order_3',
'drd/tests/annotate_rwlock',
'drd/tests/annotate_sem',
'drd/tests/annotate_smart_pointer',
'drd/tests/annotate_smart_pointer2',
'drd/tests/annotate_spinlock',
'drd/tests/annotate_static',
'drd/tests/atomic_var',
'drd/tests/bar_bad',
'drd/tests/bar_bad_xml',
'drd/tests/bar_trivial',
'drd/tests/boost_thread',
'drd/tests/bug-235681',
'drd/tests/bug322621',
'drd/tests/circular_buffer',
'drd/tests/concurrent_close',
'drd/tests/dlopen',
'drd/tests/fork-parallel',
'drd/tests/fork-serial',
'drd/tests/fp_race',
'drd/tests/fp_race2',
'drd/tests/fp_race_xml',
'drd/tests/free_is_write',
'drd/tests/free_is_write2',
'drd/tests/hg01_all_ok',
'drd/tests/hg02_deadlock',
'drd/tests/hg03_inherit',
'drd/tests/hg04_race',
'drd/tests/hg05_race2',
'drd/tests/hg06_readshared',
'drd/tests/linuxthreads_det',
'drd/tests/matinv',
'drd/tests/monitor_example',
'drd/tests/pth_barrier',
'drd/tests/pth_barrier2',
'drd/tests/pth_barrier3',
'drd/tests/pth_barrier_race',
'drd/tests/pth_barrier_thr_cr',
'drd/tests/pth_broadcast',
'drd/tests/pth_cancel_locked',
'drd/tests/pth_cleanup_handler',
'drd/tests/pth_cond_race',
'drd/tests/pth_cond_race2',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/pth_detached2',
'drd/tests/pth_detached3',
'drd/tests/pth_detached_sem',
'drd/tests/pth_inconsistent_cond_wait',
'drd/tests/pth_once',
'drd/tests/pth_spinlock',
'drd/tests/read_and_free_race',
'drd/tests/rwlock_race',
'drd/tests/rwlock_test',
'drd/tests/sem_as_mutex',
'drd/tests/sem_as_mutex2',
'drd/tests/sem_as_mutex3',
'drd/tests/sem_open',
'drd/tests/sem_open2',
'drd/tests/sem_open3',
'drd/tests/sem_open_traced',
'drd/tests/sigalrm',
'drd/tests/std_atomic',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/std_string',
'drd/tests/std_thread',
'drd/tests/swapcontext',
'drd/tests/tc01_simple_race',
'drd/tests/tc02_simple_tls',
'drd/tests/tc03_re_excl',
'drd/tests/tc05_simple_race',
'drd/tests/tc06_two_races',
'drd/tests/tc07_hbl1',
'drd/tests/tc08_hbl2',
'drd/tests/tc09_bad_unlock',
'drd/tests/tc11_XCHG',
'drd/tests/tc16_byterace',
'drd/tests/tc17_sembar',
'drd/tests/tc18_semabuse',
'drd/tests/tc19_shadowmem',
'drd/tests/tc21_pthonce',
'drd/tests/tc22_exit_w_lock',
'drd/tests/tc23_bogus_condwait',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']


It seems that
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233
was under more load has it had python3 fail and more strace failures.

I'll triage the remaining intermittent ptest failures

thanks, the good thing is that these packages are quite sticky to
glibc internals so usually an upgrade if available
for them will sort things out, glibc is merging libpthread/librt etc
into libc so lot of valgrind failures could be due to that
I will see what I can find.

--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


glibc 2.34 tests

Alexandre Belloni
 

Hello,

I've been testing multiple versions of Khem's branch for glibc2.34
Everything apart from non-gpl3 builds fine. There is a patch from Khem
to solve this issue, I didn't apply/test yet:

[yocto] [meta-gplv2] [PATCH] coreutils_6.9.bb: Fix build with glibc 2.34

On the tests side, systemd-journal was crashing, the last patch on
Khem's branch is solving the issue, thanks!

The remaining issues are on the ptests side:

https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1946/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2236/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1949/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2244/steps/12/logs/stdio
https://autobuilder.yoctoproject.org/typhoon/#/builders/82/builds/1957/steps/12/logs/stdio

Issues happen both on qemux86-64 and qemuarm64 and they reproduced
every time:

'elfutils': ['run-backtrace-native.sh', 'run-backtrace-native-core.sh']

'strace': ['prlimit64.gen.test',
'xet_robust_list.gen.test',
'xet_robust_list--pidns-translation.gen.test'],

For valgrind, it is all over the place:

ARM has:
'valgrind': ['gdbserver_tests/hginfo',
'helgrind/tests/hg02_deadlock',
'drd/tests/boost_thread',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/swapcontext',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']

x86 has:

'valgrind': ['gdbserver_tests/hginfo',
'memcheck/tests/supp_unknown',
'helgrind/tests/annotate_smart_pointer',
'helgrind/tests/bar_bad',
'helgrind/tests/hg01_all_ok',
'helgrind/tests/hg02_deadlock',
'helgrind/tests/pth_destroy_cond',
'helgrind/tests/rwlock_test',
'helgrind/tests/tc05_simple_race',
'helgrind/tests/tc06_two_races',
'helgrind/tests/tc06_two_races_xml',
'helgrind/tests/tc14_laog_dinphils',
'helgrind/tests/tc17_sembar',
'helgrind/tests/tc19_shadowmem',
'helgrind/tests/tc21_pthonce',
'helgrind/tests/tc22_exit_w_lock',
'helgrind/tests/tc23_bogus_condwait',
'helgrind/tests/tls_threads',
'drd/tests/annotate_barrier',
'drd/tests/annotate_barrier_xml',
'drd/tests/annotate_hb_race',
'drd/tests/annotate_hbefore',
'drd/tests/annotate_ignore_read',
'drd/tests/annotate_ignore_rw',
'drd/tests/annotate_ignore_rw2',
'drd/tests/annotate_ignore_write',
'drd/tests/annotate_ignore_write2',
'drd/tests/annotate_order_1',
'drd/tests/annotate_order_2',
'drd/tests/annotate_order_3',
'drd/tests/annotate_rwlock',
'drd/tests/annotate_sem',
'drd/tests/annotate_smart_pointer',
'drd/tests/annotate_smart_pointer2',
'drd/tests/annotate_spinlock',
'drd/tests/annotate_static',
'drd/tests/atomic_var',
'drd/tests/bar_bad',
'drd/tests/bar_bad_xml',
'drd/tests/bar_trivial',
'drd/tests/boost_thread',
'drd/tests/bug-235681',
'drd/tests/bug322621',
'drd/tests/circular_buffer',
'drd/tests/concurrent_close',
'drd/tests/dlopen',
'drd/tests/fork-parallel',
'drd/tests/fork-serial',
'drd/tests/fp_race',
'drd/tests/fp_race2',
'drd/tests/fp_race_xml',
'drd/tests/free_is_write',
'drd/tests/free_is_write2',
'drd/tests/hg01_all_ok',
'drd/tests/hg02_deadlock',
'drd/tests/hg03_inherit',
'drd/tests/hg04_race',
'drd/tests/hg05_race2',
'drd/tests/hg06_readshared',
'drd/tests/linuxthreads_det',
'drd/tests/matinv',
'drd/tests/monitor_example',
'drd/tests/pth_barrier',
'drd/tests/pth_barrier2',
'drd/tests/pth_barrier3',
'drd/tests/pth_barrier_race',
'drd/tests/pth_barrier_thr_cr',
'drd/tests/pth_broadcast',
'drd/tests/pth_cancel_locked',
'drd/tests/pth_cleanup_handler',
'drd/tests/pth_cond_race',
'drd/tests/pth_cond_race2',
'drd/tests/pth_cond_race3',
'drd/tests/pth_create_chain',
'drd/tests/pth_detached',
'drd/tests/pth_detached2',
'drd/tests/pth_detached3',
'drd/tests/pth_detached_sem',
'drd/tests/pth_inconsistent_cond_wait',
'drd/tests/pth_once',
'drd/tests/pth_spinlock',
'drd/tests/read_and_free_race',
'drd/tests/rwlock_race',
'drd/tests/rwlock_test',
'drd/tests/sem_as_mutex',
'drd/tests/sem_as_mutex2',
'drd/tests/sem_as_mutex3',
'drd/tests/sem_open',
'drd/tests/sem_open2',
'drd/tests/sem_open3',
'drd/tests/sem_open_traced',
'drd/tests/sigalrm',
'drd/tests/std_atomic',
'drd/tests/std_list',
'drd/tests/std_mutex',
'drd/tests/std_string',
'drd/tests/std_thread',
'drd/tests/swapcontext',
'drd/tests/tc01_simple_race',
'drd/tests/tc02_simple_tls',
'drd/tests/tc03_re_excl',
'drd/tests/tc05_simple_race',
'drd/tests/tc06_two_races',
'drd/tests/tc07_hbl1',
'drd/tests/tc08_hbl2',
'drd/tests/tc09_bad_unlock',
'drd/tests/tc11_XCHG',
'drd/tests/tc16_byterace',
'drd/tests/tc17_sembar',
'drd/tests/tc18_semabuse',
'drd/tests/tc19_shadowmem',
'drd/tests/tc21_pthonce',
'drd/tests/tc22_exit_w_lock',
'drd/tests/tc23_bogus_condwait',
'drd/tests/tc24_nonzero_sem',
'drd/tests/thread_name',
'drd/tests/thread_name_xml',
'drd/tests/threaded-fork-vcs',
'drd/tests/threaded-fork',
'drd/tests/tls_threads',
'massif/tests/deep-D']


It seems that
https://autobuilder.yoctoproject.org/typhoon/#/builders/81/builds/2233
was under more load has it had python3 fail and more strace failures.

I'll triage the remaining intermittent ptest failures


--
Alexandre Belloni, co-owner and COO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Re: SWAT Rotation

Leonardo Sandoval
 

Hi Alexandre and team,

Sure, I can 'swat' next week.

lsg


On Thu, 1 Jul 2021 at 17:28, Alexandre Belloni <alexandre.belloni@...> wrote:
Hello Leonardo,

You are the next one on the SWAT Team members list
(https://wiki.yoctoproject.org/wiki/Yocto_Build_Failure_Swat_Team#Members)
and SWAT duty will rotate to you at EOD 2021-07-02.

Please reply to let me know whether you will be able to work on this
task.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


SWAT Rotation

Alexandre Belloni
 

Hello Leonardo,

You are the next one on the SWAT Team members list
(https://wiki.yoctoproject.org/wiki/Yocto_Build_Failure_Swat_Team#Members)
and SWAT duty will rotate to you at EOD 2021-07-02.

Please reply to let me know whether you will be able to work on this
task.

Thanks!

--
Alexandre Belloni, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com


Failed qemu bugs - notes for SWAT

Richard Purdie
 

I have some good news. Now that we have merged the rcu stall crash fix:

http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=1f802cdf857475eefb9f3ae5032a6f1ce3e01d97

I think that the cause of some of our qemu hangs under higher IO load
should be fixed. As such, I've gone through and closed out a number of qemu
AB-INT bugs that look like they're related to this.

If we do see anything that looks like some of the closed issues recur, I'd
suggest opening new bugs as the older ones are becoming unwieldy and it is
hard to tell what may have been the rcu stall and what wasn't.

For any new testimage/qemu failures, please do try and capture and include
the qemu boot log from WORKDIR/testimage/qemu_*.log if it isn't printed on the
console log. If you don't have access, ping someone who does. It is looking to
be of key importance that we get those logs. We have improved some of the code
paths but we're not quite there yet for some failure paths.

I also closed out the ltp bugs since we do have a fix there at least for the 
most common hang. I suspect there is still an arm hang lurking somewhere so
if we see that, it should have its own new bug (ping Alex or me too please as 
we'd like to poke at the hung build).

Cheers,

Richard


Re: Autobuilder "rcu stall" issue summary

Richard Purdie
 

On Fri, 2021-06-25 at 16:34 +0100, Richard Purdie via lists.yoctoproject.org wrote:
On Thu, 2021-06-24 at 17:31 +0100, Richard Purdie via lists.yoctoproject.org wrote:
I'll stop here but will follow up on the mail if I remember more info. Anyone else
feel free to add and if anyone has any insight into what is happening or how to better
debug, I'm very open to it!
I formed a new plan and set the stall detector to 3 seconds instead of 
21 seconds with this hack in master-next:

http://git.yoctoproject.org/cgit.cgi/poky/commit/?h=master-next&id=71b6bc157d39e09e8f76a15a049168eb72bbd3d9

i.e. adding CONFIG_RCU_CPU_STALL_TIMEOUT=3

I ran two builds on the autobuilder, one of which was a heavy rebuild build
so should trigger a lot of IO. I've been looking over the build results and
we have a number of builds which failed or are in the process of doing so,
I've dumped 4 below (three arm, one x86-64).

My thinking is that if we reduced the stall detector limit and only saw 
occasional hangs at the same rate as before, it was likely qemu. If we saw
an increase in hangs (which we definitely have even on the preliminary 
incomplete results), it is more likely something in the kernel RCU stall
code is taking out the system. The BUG: in occasional logs also hints at
the latter.

My own next step is probably to hack a "stall report" trigger into sysrq
and try and in some of my own images locally, see if I can break this at
will.
I realised there was a potential lock issue in the rcu stall code in the
kernel. In looking into it with Paul we found that upstream have a fix
queued for the same issue:

https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/commit/?id=406a2f008f2e

This would account for the BUG: entries in the previous email. I plan to
test this with the short rcu stall timing, see if we still see the hangs.
Adding this fix to linux-yocto would seem helpful just to rule it out
if nothing else but this fix does look like the kind of thing that would
cause the issues we've been tracking.

Cheers,

Richard


Re: Autobuilder "rcu stall" issue summary

Richard Purdie
 

On Thu, 2021-06-24 at 17:31 +0100, Richard Purdie via lists.yoctoproject.org wrote:
I'll stop here but will follow up on the mail if I remember more info. Anyone else
feel free to add and if anyone has any insight into what is happening or how to better
debug, I'm very open to it!
I formed a new plan and set the stall detector to 3 seconds instead of 
21 seconds with this hack in master-next:

http://git.yoctoproject.org/cgit.cgi/poky/commit/?h=master-next&id=71b6bc157d39e09e8f76a15a049168eb72bbd3d9

i.e. adding CONFIG_RCU_CPU_STALL_TIMEOUT=3

I ran two builds on the autobuilder, one of which was a heavy rebuild build
so should trigger a lot of IO. I've been looking over the build results and
we have a number of builds which failed or are in the process of doing so,
I've dumped 4 below (three arm, one x86-64).

My thinking is that if we reduced the stall detector limit and only saw 
occasional hangs at the same rate as before, it was likely qemu. If we saw
an increase in hangs (which we definitely have even on the preliminary 
incomplete results), it is more likely something in the kernel RCU stall
code is taking out the system. The BUG: in occasional logs also hints at
the latter.

My own next step is probably to hack a "stall report" trigger into sysrq
and try and in some of my own images locally, see if I can break this at
will.

Cheers,

Richard


qemuarm:

https://autobuilder.yoctoproject.org/typhoon/#/builders/53/builds/3608

/home/pokybuild/yocto-worker/qemuarm/build/build/tmp/work/qemuarm-poky-linux-gnueabi/core-image-sato-sdk/1.0-
r0/testimage/qemu_boot_log.20210625121601

** (matchbox-desktop:715): [1;33mWARNING[0m **: [34m12:17:35.454[0m: Error loading icon: Icon 'terminal' not
present in theme Sato
[ 2173.684752] hellomod: loading out-of-tree module taints kernel.
[ 2173.726821] Hello world!
[ 2178.269807] Cleaning up hellomod.
[ 2426.747471] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 2426.753269] rcu: 2-....: (32 ticks this GP) idle=21e/1/0x40000000 softirq=95871/95871 fqs=19
[ 2426.823824] (detected by 0, t=302 jiffies, g=214353, q=107)


qemuarm-oecore:

https://autobuilder.yoctoproject.org/typhoon/#/builders/47/builds/3601


OpenEmbedded nodistro.0 qemuarm /dev/hvc0

/home/pokybuild/yocto-worker/qemuarm-oecore/build/build/tmp-glibc/work/qemuarm-oe-linux-gnueabi/core-image-
sato/1.0-r0/testimage/qemu_boot_log.20210625114847

qemuarm login: [ 53.248661] rcu: INFO: rcu_preempt self-detected stall on CPU
[ 53.250587] rcu: 0-...!: (1 ticks this GP) idle=4aa/0/0x1 softirq=1385/1385 fqs=0
[ 53.251571] (t=1019 jiffies g=1293 q=4)
[ 53.252192] rcu: rcu_preempt kthread starved for 1019 jiffies! g1293 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x0 -
cpu=1
[ 53.253026] rcu: Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
[ 53.253697] rcu: RCU grace-period kthread stack dump:
[ 53.254348] task:rcu_preempt state:R running task stack: 0 pid: 13 ppid: 2
flags:0x00000000
[ 53.258129] [<c0a73808>] (__schedule) from [<c0a73a74>] (schedule+0xc8/0x118)
[ 53.258925] [<c0a73a74>] (schedule) from [<c0a75ef8>] (schedule_timeout+0xcc/0xfc)
[ 53.259247] [<c0a75ef8>] (schedule_timeout) from [<c028b8b0>] (rcu_gp_kthread+0x2ac/0x92c)
[ 53.259619] [<c028b8b0>] (rcu_gp_kthread) from [<c024f314>] (kthread+0x158/0x164)
[ 53.259865] [<c024f314>] (kthread) from [<c0200138>] (ret_from_fork+0x14/0x3c)
[ 53.260207] Exception stack(0xc1539fb0 to 0xc1539ff8)
[ 53.260691] 9fa0: 00000000 00000000 00000000 00000000
[ 53.260973] 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[ 53.261236] 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000
[ 53.261718] NMI backtrace for cpu 0
[ 53.262040] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.10.43-yocto-standard #1
[ 53.262324] Hardware name: Generic DT based system
[ 53.262584] [<c020fe28>] (unwind_backtrace) from [<c020ab54>] (show_stack+0x18/0x1c)
[ 53.262847] [<c020ab54>] (show_stack) from [<c0a6e304>] (dump_stack+0x88/0xa4)
[ 53.263081] [<c0a6e304>] (dump_stack) from [<c06dc0e8>] (nmi_cpu_backtrace+0xb8/0xcc)
[ 53.263340] [<c06dc0e8>] (nmi_cpu_backtrace) from [<c06dc1b8>] (nmi_trigger_cpumask_backtrace+0xbc/0x164)
[ 53.263615] [<c06dc1b8>] (nmi_trigger_cpumask_backtrace) from [<c0a69248>] (rcu_dump_cpu_stacks+0xe8/0x114)
[ 53.264190] [<c0a69248>] (rcu_dump_cpu_stacks) from [<c028d9d8>] (rcu_sched_clock_irq+0x35c/0xa78)
[ 53.264463] [<c028d9d8>] (rcu_sched_clock_irq) from [<c0296f8c>] (update_process_times+0x5c/0x80)
[ 53.264757] [<c0296f8c>] (update_process_times) from [<c02a8220>] (tick_sched_timer+0x17c/0x1fc)
[ 53.265047] [<c02a8220>] (tick_sched_timer) from [<c0297fc8>] (__hrtimer_run_queues+0x19c/0x2c4)
[ 53.265317] [<c0297fc8>] (__hrtimer_run_queues) from [<c0298528>] (hrtimer_interrupt+0xc8/0x204)
[ 53.265559] [<c0298528>] (hrtimer_interrupt) from [<c08a112c>] (arch_timer_handler_virt+0x30/0x38)
[ 53.265802] [<c08a112c>] (arch_timer_handler_virt) from [<c027f844>] (handle_percpu_devid_irq+0x60/0x118)
[ 53.266079] [<c027f844>] (handle_percpu_devid_irq) from [<c0279bf0>] (generic_handle_irq+0x30/0x40)
[ 53.266320] [<c0279bf0>] (generic_handle_irq) from [<c027a218>] (__handle_domain_irq+0x98/0xa8)
[ 53.266554] [<c027a218>] (__handle_domain_irq) from [<c06eb518>] (gic_handle_irq+0x74/0x84)
[ 53.266778] [<c06eb518>] (gic_handle_irq) from [<c0201278>] (__irq_svc+0x58/0x8c)
[ 53.266998] Exception stack(0xc1201f20 to 0xc1201f68)
[ 53.267171] 1f20: 0000a4a8 cfd6de84 00000000 c0218860 c1200000 00000000 c1200000 c1201f80
[ 53.267390] 1f40: c1204f94 00000001 c1204fd4 00000000 c1684040 c1201f70 c0207930 c0207920
[ 53.267611] 1f60: 60010113 ffffffff
[ 53.267744] [<c0201278>] (__irq_svc) from [<c0207920>] (arch_cpu_idle+0x24/0x40)
[ 53.267946] [<c0207920>] (arch_cpu_idle) from [<c0a76898>] (default_idle_call+0x40/0x5c)
[ 53.268167] [<c0a76898>] (default_idle_call) from [<c025e110>] (do_idle+0x108/0x254)
[ 53.268384] [<c025e110>] (do_idle) from [<c025e534>] (cpu_startup_entry+0x20/0x24)
[ 53.268624] [<c025e534>] (cpu_startup_entry) from [<c1000e80>] (start_kernel+0x4a0/0x57c)
[ 73.311040] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 73.313047] (detected by 2, t=428 jiffies, g=1965, q=113)
[ 73.313886] rcu: All QSes seen, last rcu_preempt kthread activity 425 (-22941--23366),
jiffies_till_next_fqs=1, root ->qsmask 0x0
[ 73.315413] rcu: rcu_preempt kthread starved for 425 jiffies! g1965 f0x2 RCU_GP_WAIT_FQS(5) ->state=0x200 -
cpu=2
[ 73.316823] rcu: Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
[ 73.317775] rcu: RCU grace-period kthread stack dump:
[ 73.318304] task:rcu_preempt state:R stack: 0 pid: 13 ppid: 2 flags:0x00000000
[ 73.318759] [<c0a73808>] (__schedule) from [<c0a73a74>] (schedule+0xc8/0x118)
[ 73.319123] [<c0a73a74>] (schedule) from [<c0a75ef8>] (schedule_timeout+0xcc/0xfc)
[ 73.319500] [<c0a75ef8>] (schedule_timeout) from [<c028b8b0>] (rcu_gp_kthread+0x2ac/0x92c)
[ 73.319886] [<c028b8b0>] (rcu_gp_kthread) from [<c024f314>] (kthread+0x158/0x164)
[ 73.320201] [<c024f314>] (kthread) from [<c0200138>] (ret_from_fork+0x14/0x3c)
[ 73.320776] Exception stack(0xc1539fb0 to 0xc1539ff8)
[ 73.321088] 9fa0: 00000000 00000000 00000000 00000000
[ 73.321403] 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
[ 73.321698] 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000
[ 73.322376] BUG: scheduling while atomic: swapper/2/0/0x00000002
[ 73.322697] Modules linked in:
[ 73.323092] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 5.10.43-yocto-standard #1
[ 73.323383] Hardware name: Generic DT based system
[ 73.323621] [<c020fe28>] (unwind_backtrace) from [<c020ab54>] (show_stack+0x18/0x1c)
[ 73.323908] [<c020ab54>] (show_stack) from [<c0a6e304>] (dump_stack+0x88/0xa4)
[ 73.324184] [<c0a6e304>] (dump_stack) from [<c02575e4>] (__schedule_bug+0x68/0x8c)
[ 73.324454] [<c02575e4>] (__schedule_bug) from [<c0a73578>] (__schedule+0x58/0x48c)
[ 73.324741] [<c0a73578>] (__schedule) from [<c0a73e00>] (schedule_idle+0x68/0x80)
[ 73.325046] [<c0a73e00>] (schedule_idle) from [<c025e534>] (cpu_startup_entry+0x20/0x24)
[ 73.325576] [<c025e534>] (cpu_startup_entry) from [<40201c2c>] (0x40201c2c)

qemuarm64-ltp (on arm64 host ubuntu1804-arm-1)

https://autobuilder.yoctoproject.org/typhoon/#/builders/96/builds/1888

/home/pokybuild/yocto-worker/qemuarm64-ltp/build/build/tmp/work/qemuarm64-poky-linux/core-image-sato/1.0-
r0/testimage/qemu_boot_log.20210625121516

qemuarm64 login: [ 21.445581] Process accounting resumed
[ 22.705496] Process accounting resumed
[ 331.583598] capability: warning: `capget01' uses 32-bit capabilities (legacy support in use)
[ 331.585764] capability: warning: `capget01' uses deprecated v2 capabilities in a way that may be insecure
[ 345.996555] Clock: inserting leap second 23:59:60 UTC
[ 713.924549] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 714.133632] rcu: Tasks blocked on level-0 rcu_node (CPUs 0-3): P9691/1:b..l
[ 714.255745] (detected by 1, t=822 jiffies, g=195937, q=3443)
[ 714.256974] task:msgstress03 state:R running task stack: 0 pid: 9589 ppid: 9543
flags:0x00000000
[ 714.258971] Call trace:
[ 714.259520] __switch_to+0x114/0x19c
[ 714.260274] __schedule+0x274/0x610
[ 714.261008] preempt_schedule_common+0x28/0x50
[ 714.261928] preempt_schedule+0x30/0x40
[ 714.262711] _raw_spin_unlock+0x38/0x50
[ 714.263507] do_msgrcv+0x284/0x4f0
[ 714.264230] __arm64_sys_msgrcv+0x38/0x44
[ 714.265042] el0_svc_common.constprop.0+0x9c/0x1c0
[ 714.266014] do_el0_svc+0x7c/0xa0
[ 714.266718] el0_svc+0x20/0x30
[ 714.267391] el0_sync_handler+0xa4/0x130
[ 714.268193] el0_sync+0x180/0x1c0
[ 714.268900] task:msgstress03 state:R running task stack: 0 pid: 9691 ppid: 32178
flags:0x00000000
[ 714.270830] Call trace:
[ 714.271369] __switch_to+0x114/0x19c
[ 714.313241] __schedule+0x274/0x610
[ 714.333040] preempt_schedule_irq+0x68/0x100
[ 714.333944] arm64_preempt_schedule_irq+0x2c/0x40
[ 714.334910] el1_irq+0xe4/0x180
[ 714.335606] wake_up_q+0xa8/0xd0
[ 714.336308] do_msgsnd+0x3f8/0x4e0
[ 714.337029] ksys_msgsnd+0x90/0xb0
[ 714.337750] __arm64_sys_msgsnd+0x2c/0x40
[ 714.338564] el0_svc_common.constprop.0+0x9c/0x1c0
[ 714.339552] do_el0_svc+0x7c/0xa0
[ 714.340266] el0_svc+0x20/0x30
[ 714.340936] el0_sync_handler+0xa4/0x130
[ 714.341735] el0_sync+0x180/0x1c0
[ 720.932547] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 721.016226] rcu: Tasks blocked on level-0 rcu_node (CPUs 0-3):
[ 721.072646] (detected by 0, t=752 jiffies, g=195941, q=7890)
[ 721.073841] rcu: All QSes seen, last rcu_preempt kthread activity 1 (4295072518-4295072517),
jiffies_till_next_fqs=1, root ->qsmask 0x0
[ 750.072043] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 750.073887] rcu: Tasks blocked on level-0 rcu_node (CPUs 0-3): P1738/1:b..l
[ 750.116802] (detected by 0, t=773 jiffies, g=196493, q=15176)
[ 750.118095] task:msgstress04 state:R running task stack: 0 pid: 1733 ppid: 1482
flags:0x00000000
[ 750.120144] Call trace:
[ 750.120730] __switch_to+0x114/0x19c
[ 750.144388] __schedule+0x274/0x610
[ 750.145215] preempt_schedule_common+0x28/0x50
[ 750.146187] preempt_schedule+0x30/0x40
[ 750.147020] _raw_spin_unlock+0x38/0x50
[ 750.147853] do_msgsnd+0x3ec/0x4e0
[ 750.148610] ksys_msgsnd+0x90/0xb0
[ 750.149365] __arm64_sys_msgsnd+0x2c/0x40
[ 750.150219] el0_svc_common.constprop.0+0x9c/0x1c0
[ 750.151244] do_el0_svc+0x7c/0xa0
[ 750.151995] el0_svc+0x20/0x30
[ 750.152698] el0_sync_handler+0xa4/0x130
[ 750.153532] el0_sync+0x180/0x1c0
[ 750.154273] task:msgstress04 state:R running task stack: 0 pid: 1738 ppid: 1482
flags:0x00000000
[ 750.156271] Call trace:
[ 750.156842] __switch_to+0x114/0x19c
[ 750.157626] __schedule+0x274/0x610
[ 750.158395] preempt_schedule_common+0x28/0x50
[ 750.159363] preempt_schedule+0x30/0x40
[ 750.160196] _raw_spin_unlock+0x38/0x50
[ 750.161020] do_msgrcv+0x284/0x4f0
[ 750.161774] __arm64_sys_msgrcv+0x38/0x44
[ 750.162620] el0_svc_common.constprop.0+0x9c/0x1c0
[ 750.163643] do_el0_svc+0x7c/0xa0
[ 750.164388] el0_svc+0x20/0x30
[ 750.165086] el0_sync_handler+0xa4/0x130
[ 750.165918] el0_sync+0x180/0x1c0
[ 763.280549] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 763.282336] (detected by 0, t=752 jiffies, g=196741, q=7494)
[ 763.283694] rcu: All QSes seen, last rcu_preempt kthread activity 703 (4295083105-4295082402),
jiffies_till_next_fqs=1, root ->qsmask 0x0
[ 763.286260] rcu: rcu_preempt kthread starved for 703 jiffies! g196741 f0x2 RCU_GP_WAIT_FQS(5) ->state=0x0 -
cpu=3
[ 763.288363] rcu: Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
[ 763.290242] rcu: RCU grace-period kthread stack dump:
[ 763.291334] task:rcu_preempt state:R running task stack: 0 pid: 13 ppid: 2
flags:0x00000028
[ 763.293387] Call trace:
[ 763.293976] __switch_to+0x114/0x19c
[ 763.294779] __schedule+0x274/0x610
[ 763.295582] schedule+0x7c/0x110
[ 763.296343] schedule_timeout+0xb0/0x1c0
[ 763.297216] rcu_gp_kthread+0x4e0/0xa90
[ 763.298073] kthread+0x168/0x16c
[ 763.298831] ret_from_fork+0x10/0x3c
[ 763.299741] BUG: scheduling while atomic: msgstress04/18860/0x00000002
[ 763.301093] Modules linked in:
[ 763.301808] CPU: 0 PID: 18860 Comm: msgstress04 Not tainted 5.10.43-yocto-standard #1
[ 763.303502] Hardware name: linux,dummy-virt (DT)
[ 763.304491] Call trace:
[ 763.305056] dump_backtrace+0x0/0x1a0
[ 763.305852] show_stack+0x24/0x30
[ 763.306595] dump_stack+0xd4/0x110
[ 763.307357] __schedule_bug+0x68/0x80
[ 763.308160] __schedule+0x454/0x610
[ 763.308925] schedule+0x7c/0x110
[ 763.309655] do_msgsnd+0x294/0x4e0
[ 763.310406] ksys_msgsnd+0x90/0xb0
[ 763.311167] __arm64_sys_msgsnd+0x2c/0x40
[ 763.312024] el0_svc_common.constprop.0+0x9c/0x1c0
[ 763.313039] do_el0_svc+0x7c/0xa0
[ 763.313780] el0_svc+0x20/0x30
[ 763.314479] el0_sync_handler+0xa4/0x130
[ 763.315326] el0_sync+0x180/0x1c0

multilib core-image-sato ubuntu1804-ty-1

https://autobuilder.yoctoproject.org/typhoon/#/builders/44/builds/3622

/home/pokybuild/yocto-worker/multilib/build/build/tmp/work/qemux86_64-poky-linux/core-image-sato/1.0-
r0/testimage/qemu_boot_log.20210625121139

libGL error: MESA-LOADER: failed to open bochs-drm: /usr/lib/dri/bochs-drm_dri.so: cannot open shared object
file: No such file or directory (search paths /usr/lib/dri)
libGL error: failed to load driver: bochs-drm
libGL error: pci id for fd 4: 1234:1111, driver (null)
libGL error: MESA-LOADER: failed to open bochs-drm: /usr/lib64/dri/bochs-drm_dri.so: cannot open shared object
file: No such file or directory (search paths /usr/lib64/dri)
libGL error: failed to load driver: bochs-drm
[settings daemon] Forking. run with -n to prevent fork
dbus-daemon[652]: Activating service name='org.a11y.atspi.Registry' requested by ':1.1' (uid=0 pid=669
comm="matchbox-panel --start-applets showdesktop,windows")
dbus-daemon[652]: Successfully activated service 'org.a11y.atspi.Registry'
SpiRegistry daemon is running with well-known name - org.a11y.atspi.Registry

** (matchbox-desktop:668): [1;33mWARNING[0m **: [34m12:12:24.699[0m: Error loading icon: Icon 'terminal' not
present in theme Sato

** (matchbox-desktop:668): [1;33mWARNING[0m **: [34m12:12:24.713[0m: Error loading icon: Icon 'net-48d24' not
present in theme Sato
[ 21.640654] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 21.642263] rcu: 3-...0: (0 ticks this GP) idle=18a/1/0x4000000000000000 softirq=1618/1618 fqs=425
[ 21.644170] (detected by 1, t=3155 jiffies, g=2653, q=16)
[ 30.951654] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:


Re: Autobuilder "rcu stall" issue summary

Denys Dmytriyenko
 

On Thu, Jun 24, 2021 at 01:57:43PM -0400, Bruce Ashfield wrote:
On Thu, Jun 24, 2021 at 12:31 PM Richard Purdie
<richard.purdie@linuxfoundation.org> wrote:

We have info all over the place relating to our "rcu stall" issue on the
autobuilder. This is an attempt to document what we know and what we don't.

We see builds fail periodically where there is an "rcu stall" mentioned
by the kernel and the image "locks up". An easier to read example is:

https://autobuilder.yoctoproject.org/typhoon/#/builders/56/builds/1488/steps/14/logs/stdio

A core-image-minimal in testimage sitting using 394% CPU having run for 97
minutes of cpu time. A minimal image doesn't execute much at all and should
be over in a minute max.

An example we were able to "catch" and debug a bit at the expense of breaking
the output is:

https://autobuilder.yoctoproject.org/typhoon/#/builders/45/builds/3576/steps/12/logs/stdio

where a musl qemux86-64 core-image-sato-sdk stopped mid "configure" on target
during testimage.

We were able to look at the kernel log, the tail end of which showed:

[settings daemon] Forking. run with -n to prevent fork
[ 102.082744] rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
[ 102.093213] rcu: 1-...!: (1 GPs behind) idle=3f6/0/0x1 softirq=1374/1375 fqs=0
[ 102.096448] rcu: 3-...!: (12 ticks this GP) idle=e34/0/0x0 softirq=1658/1658 fqs=0
[ 102.099326] (detected by 1, t=21002 jiffies, g=1937, q=278)

The kernel was:

[ 0.000000] Linux version 5.10.43-yocto-standard (oe-user@oe-host) (x86_64-poky-linux-musl-gcc (GCC) 11.1.1 20210523, GNU ld (GNU Binutils) 2.36.1.20210209) #1 SMP PREEMPT Wed Jun 16 1
6:00:47 UTC 2021
[ 0.000000] Command line: root=/dev/vda rw mem=512M ip=192.168.7.4::192.168.7.3:255.255.255.0 console=ttyS0 console=ttyS1 oprofile.timer=1 tsc=reliable no_timer_check rcupdate.rcu_exp
edited=1 printk.time=1

Full log: https://www.rpsys.net/wp/rp/qemuboot-autobuild.log

In the above state, qemu was seem with 303% CPU usage and 403% cpu usage. I do
have strace output from the hung qemu: https://www.rpsys.net/wp/rp/qemu-strace.log

It was interesting to note that pinging the image did work and you can see that in
the strace log so something was still alive. You couldn't ssh in though and existing
ssh connections failed with timeouts.

We do also have the VM core dump for this which Alexandre is looking into
(from qmp with:
{"execute": "qmp_capabilities"}
{"execute":"dump-guest-memory","arguments":{"paging":false,"protocol":"file:/tmp/vmcore.img"}}
)

I couldn't see any extra kernel log messages in the vmcore file.

We've seen this before in core-image-sato, core-image-sato-sdk and also with
qemuarm, the latter seeming to imply that KVM isn't the issue and that it
isn't arch specific, implying it is qemu related. The load itself doesn't appear
to matter, above it was a configure script but we've seen it on images without
toolchains in a variety of states/loads. We've seen on a variety of host workers,
ubuntu1804 being most common but it is a more common worker OS too.

I suspect that this is the same issue that causes images to lock up in the firmware
boot stage as in: https://bugzilla.yoctoproject.org/show_bug.cgi?id=14018
There may also be other related bugs, without looking at the kernel traces, it is hard
to say for sure https://bugzilla.yoctoproject.org/show_bug.cgi?id=14273 lists a load more.

We've not been able to replicate the issue at will. I have tried various stress-ng --all
loads locally with no success at replicating. I did compare the boot output from qemu
on my local system with the above kernel boot and it matches apart from timings and
kvm-clock msr value.

The suspicion is that some high IO load event somehow triggers qemu to fail. I
also suspect that the "rcu stall" is just a symptom of a wider issue. The stall is
suspicious as it looks incomplete.

I'll stop here but will follow up on the mail if I remember more info. Anyone else
feel free to add and if anyone has any insight into what is happening or how to better
debug, I'm very open to it!
Richard,

As mentioned earlier, we do have the kernel debug fragments available (as well
as a 'developer' kernel type). We can't globally enable the developer
ktype, since
we haven't defined BSP entry points for all of our reference boads (but I can do
that as a TODO item).

But we can get the same options enabled by adding:

KERNEL_FEATURES_append = "ktypes/developer/developer.cfg
features/debug/debug-kernel.scc"
Can this be keyed off of DEBUG_BUILD variable?


They turn on the prime suspects in the kernel to get more information. If there
are options we are missing, let me know and I'll add them to those fragments
so we can rely on them in the future.
--
Regards,
Denys Dmytriyenko <denis@denix.org>
PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
Fingerprint: 25FC E4A5 8A72 2F69 1186 6D76 4209 0272 9A92 C964

1 - 20 of 198