-
Notifications
You must be signed in to change notification settings - Fork 6.5k
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
frdm_k64f:running testcase tests/subsys/debug/coredump/ and tests/subsys/debug/coredump_backends/ failed #32514
Comments
This is not a bug, but expected bahaviour. The test case just tests that Zephyr catches an illegal Null pointer assignment and prints a core dump. So although the core dump looks like an error, this test checks that it is indeed generated. Testcase is here So I think this bug report can be closed. |
As @laodzu said, this is expected behavior as the test is to make sure a coredump can be generated. So closing... |
@Zhaoningx , my understanding is that twister takes this as a failure of the test case? If that is the case, maybe there is something wrong with the test case. |
Hi @LeiW000 , yes, so we need to spend some time to debug this I think, by the way, this issue has also met on reel board platform, so I'll talk with Kelvin later, maybe he has good idea for this. |
I don't think we should close the issue now. The dump info is expected behavior. But twister takes it as a failure. It is still a bug. |
I think there is now some confusion in this ticket. Twister does not take this is a failure. As the original report already states, Twister shows "successful". Here is the log from my system running the test:
So again, the test passes - as expected. This is all expected behaviour so the bug can indeed be closed. |
@Zhaoningx Was your initial report not correct that twister shows "successful" running the test? As explained, the test correctly test the ability of Zephyr to do a crash dump. For this an illegal access is done on purpose and the crash is observed correctly by twister - just as you state right in your report. This is also what I confirmed by running the test on my own frdm_k64f board. So can anyone explain to me, why this bug has been re-opened? |
Hi @laodzu , yep, I totally agree with you, but there are still some issues about this bug, so I re-open this, see log below: It shows that the testcase is successful to test, but marked as failure by twister. |
@Zhaoningx Ah, slowly I understand what you are talking about. So there is a discrepancy between what twister reports in its console output and between what it writes into the XML report files.
And this is the XML report it produced:
So the problems seems to be in tests/subsys/debug/coredump/ - the |
Running the two tests individually works as expected - both tests pass and the XML files are correct. The bug only appears when running both tests in a single twister invocation. |
@nashif Have you seen the same issue with twister? |
works for me on frdm_k64f. |
@nashif Thanks for also testing. Can you clarify if running both tests at once yield correct XML output even though the console output signaled success? |
confirm there is an issue with xml output, looking. |
@nashif @andyross Thanks for fixing one race. Unfortunately the bug in this issue is not fixed on my system. I still have the same failure mode. Running the tests individually again work, running them from one twister invocation fails. I tried this 5 times and it failed 5 times, so if it is another race, I can hit it in a pretty deterministic fashion:
Results:
|
@Zhaoningx Can you please verify if the bug is really fixed for you? As I wrote, I still see the problem but I cannot reopen the bug. |
Hi @laodzu , after double check, It is not really fixed, I saw the report of yesterday, the failures of coredump is still met. See log as below: |
Do not set results before we finished parsing output. Fixes zephyrproject-rtos#32514 Signed-off-by: Anas Nashif <[email protected]>
Do not set results before we finished parsing output. Fixes #32514 Signed-off-by: Anas Nashif <[email protected]>
Do not set results before we finished parsing output. Fixes zephyrproject-rtos#32514 Signed-off-by: Anas Nashif <[email protected]>
To Reproduce
When running coredumptestcases, it showed me 'successful', but it is always showed some error log in handler.log.
Steps to reproduce the behavior:
twister -p frdm_k64f --device-testing --device-serial /dev/ttyACM0 -T tests/subsys/debug/coredump/ -j 1
twister -p frdm_k64f --device-testing --device-serial /dev/ttyACM0 -T tests/subsys/debug/coredump_backends/ -j 1
see error
handler.log
X*** Booting Zephyr OS build zephyr-v2.5.0 ***
Coredump: frdm_k64f
E: ***** USAGE FAULT *****
E: Attempt to execute undefined instruction
E: r0/a1: 0x00000014 r1/a2: 0x00000000 r2/a3: 0x00000080
E: r3/a4: 0x00000000 r12/ip: 0x0000000a r14/lr: 0x00000727
E: xpsr: 0x41000000
E: Faulting instruction address (r15/pc): 0x0000072a
E: >>> ZEPHYR FATAL ERROR 0: CPU exception on CPU 0
E: Current thread: 0x200001b0 (unknown)
E: #CD:BEGIN#
E: #CD:5a4501000300050000000000
E: #CD:4101002400
E: #CD:140000000000000080000000000000000a000000270700002a07000000000041
E: #CD:48070020
E: #CD:4d0100b001002030020020
E: #CD:cc020020cc020020000000000180000000000000000000000000000000000000
E: #CD:00000000000000000000000000000000e0010020e00100203d2600003d260000
E: #CD:800700203001002080070020001100203d260000010000005007002000000000
E: #CD:00000000000000008003002000040000000000000000000000000000f5ffffff
E: #CD:4d01008003002080070020
E: #CD:0068046001200e49c8710198000ac8720198887205984872fff7b6fe8346bbf1
E: #CD:000f01d05846d1e7a6eb080644443d4400bf002edad10020c8e700000e010000
E: #CD:0000024070b504460d460cb9042070bd40200349c87108468571fff795fef6e7
E: #CD:000002402de9ff4d0c4617461d46dde90cb8ddf838a01db9042004b0bde8f08d
E: #CD:04233a4621460098fff71eff06460eb13046f2e722e014487844006804600220
E: #CD:1249c871084680f80bb00f4979441439286809688860fff767fe06465eb1b8f1
E: #CD:000f01d0c8f80040baf1000f02d00020caf8000004e03f1f2d1d241d002fdad1
E: #CD:00bf3046c9e70000740000000000024000000000000008000000100000002000
E: #CD:0000400000008000000000010000000104000240000000000000000000001000
E: #CD:0200000000100000000000000000000000000000c0ec1b3629221c3660ccff00
E: #CD:0000000000000000c00d00202000000008822000700700000000000000000000
E: #CD:0000000000000000000000000000000000000000000000000000000000000000
E: #CD:0000000000000000000000000000000000000000000000000000000080f0fa02
E: #CD:0000000000000000401700205803002058030020000000000000000000000000
E: #CD:4005002040050020580300204c0500204c050020000000000000000000000000
E: #CD:0000000000000000000000000000000040170020401700200000000005010000
E: #CD:50030020b8040020000000000200000051000000b40000204edc1a10ff869303
E: #CD:5003002000000000000000000000000000000000000000000000000000000000
E: #CD:0000000000000000000000000000000000000000000000000000000000000000
E: #CD:0000000080f0fa0200000000b8040020b8040020000000008838002018040020
E: #CD:180400200000000000000000000000001006002010060020180400201c060020
E: #CD:1c06002000000000180400200000000000000000500300200000000018040020
E: #CD:0000000000000000000000000000000000000000000000000000000000000000
E: #CD:0000000000000000000000000000000000000000000000000000000000000000
E: #CD:0000000000000000000000000000000000000000000000001337000066000000
E: #CD:38020020cb0f0000ad0f0000053f0000133700000a00000038020020cb0f0000
E: #CD:ad0f00006007002013000000950700002e000000950700003507000000000000
E: #CD:350700000000000000000000000000000000000000000000fc3e000000000000
E: #CD:80000373000000000000000000000000000000003d2600003d26000080070020
E: #CD:3001002080070020001100203d2600000100000063070000063f000014000000
E: #CD:a402002017300000140000000000000080000000000000000a00000027070000
E: #CD:2a07000000000041b00100205d260000000000002d3000000000000041130000
E: #CD:END#
Environment (please complete the following information):
OS: Fedora33
Toolchain: zephyr-sdk-0.12.2
Commit ID: fe7c2ef
The text was updated successfully, but these errors were encountered: