RZ_A1H Tests Failing #3273
Assignees
Labels
Comments
|
Thank you for giving me the comment. And sorry for this late reply. I don't understand well the occurs. cc @toyowata |
|
@TomoYamanaka Thanks for your reply. I was just doing a general check up on this platform and saw there were a few test failures and I thought I'd notify you. The The The |
|
@c1728p9 Russ can you comment please? |
|
I took a look at the test tests-mbedmicro-rtos-mbed-malloc in depth and found a couple of problems that are causing these failures:
|
|
Hi @TomoYamanaka, the old version of RTX is 1 of the 3 problems. To pass testing all 3 need to be fixed. The old version of RTX is causing problems because rt_tsk_delete() has a bug in it - it does not release mutexes held by the terminated task. |
|
@c1728p9 |
I recognize this THREAD_STACK_SIZE macro defined to the following file. It is right ? This main.cpp is a test file for Is it possible to modify it by an ARM member who is a test case creator? |
|
Hi @TomoYamanaka I created this test so feel free to modify it. I would recommend an "#if defined(__CORTEX_A9)" and inside it set the stack to a larger value. I think cortex-A in general will require more stack space. If you would like I can make this change. |
|
@c1728p9 I will be very helpful if you change it. Please make this change. |
|
Hi @TomoYamanaka, I created #3362 to increase the stack size. |
|
@c1728p9 |
|
Now Renesas is resolving this issue. If anything, please let me ask a question again. Best regards |
|
I pushed PR for this issue. |
|
Hi @TomoYamanaka, thanks for all the work! The tests look good when testing with your PR, however in CI we also enable heap and stack stats by running the following commands: I'm receiving an mbed assertion from the memory metrics library for each test. I'm leaving for the holidays so I can't debug this at the moment, but perhaps you could look into it with @c1728p9 ? |
|
I excuted the above two commands. But I did not receive an mbed assertion, did not find like the message in test log. If you can reproduce in your environment, I would like to let me share test log. |
|
@TomoYamanaka Thanks for trying this locally. Could you tell me the version of GCC_ARM you're using? I'd like to reproduce your results locally as well. Here's my environment: |
|
Hi @bridadan Thank you for your reply. It seems the same environment as yours. |
|
Thanks @TomoYamanaka, unfortunately I'm out of office for the holidays and don't have access to a RZ_A1H, so I can't test at this time. I will do so when I return! I will try in other environments as well, it may just be something wrong with my machine. |
|
Hi @TomoYamanaka, it turns out it wasn't the HEAP_STATS but it was the STACK_STATS that was causing the issue. Could you please run the following commands your machine and see if you get a similar error? In a fresh workspace: This was the output from the test command: This is the error line causing the TIMEOUT: Can you confirm you see something similar? If you don't see it with GCC_ARM, try using the ARM or IAR toolchains. Thanks for your patience! |
sg-
pushed a commit
that referenced
this issue
Jan 9, 2017
* [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @russ already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702 * Revert "[RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API" This reverts commit e71f79a. * [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @c1728p9 already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702
|
Thank you for your commnets. As a result of executing in your method, I could reproduce the similar phenomenon. Does this TIMEOUT message mean that there is not enough stack size to set in the test ? |
|
@TomoYamanaka The TIMEOUT is actually because the device halted with an This points to this line: https://github.com/ARMmbed/mbed-os/blob/master/features/frameworks/greentea-client/source/greentea_metrics.cpp#L150 @c1728p9 Do you have suggestions on where to begin debugging this assert? |
|
I found the cause of this assert. As |
adbridge
pushed a commit
that referenced
this issue
Jan 13, 2017
PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @russ already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702
adbridge
pushed a commit
that referenced
this issue
Jan 13, 2017
PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @c1728p9 already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702
|
Hi @TomoYamanaka, it looks like the changes here added stack init, so if you want you should be able implement osThreadInfoStackMax (just copy the cortex-m version). If that is a problem or you don't want to implment it, you could wrap this in a check to "if (queue->empty())" to fix the problem. |
|
Thanks you for your comments, and sorry for this late reply. I am trying with the method that you suggested now. |
As a result of trying with the method that you suggested, I could get SUCCESS in the test case of The Code that I changed is here.
In addition, test result after code changed is here. If there is no problem in this changes, I will PR this. |
|
@TomoYamanak I think those changes should be fine, thanks for looking at this! Please submit them as a PR and we will review further. |
|
Thank you for your quick reply. |
TomoYamanaka
added a commit
to TomoYamanaka/mbed-os
that referenced
this issue
Jan 23, 2017
…NABLED=1" In CI, I receive an mbed assert of stack stats by running the following commands. Its assert is TIMEOUT. mbed test -m RZ_A1H -t GCC_ARM -DMBED_STACK_STATS_ENABLED=1 --clean --compile mbed test -m RZ_A1H -t GCC_ARM --run -n tests-mbed_drivers-stl_features -v So, I fixed the process of related to stack stats (osThreadInfoStackMax).The Detail contents is here. ARMmbed#3273 (comment)
simonqhughes
pushed a commit
to simonqhughes/mbed-os-1
that referenced
this issue
Jan 31, 2017
…3470) * [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @russ already supported in PR ARMmbed#3362. Reference for issue of test error is here. ARMmbed#3273 Reference for PR to update RTX for Cortex-M. ARMmbed#1702 * Revert "[RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API" This reverts commit e71f79a. * [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @c1728p9 already supported in PR ARMmbed#3362. Reference for issue of test error is here. ARMmbed#3273 Reference for PR to update RTX for Cortex-M. ARMmbed#1702
simonqhughes
pushed a commit
that referenced
this issue
Jan 31, 2017
* [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @russ already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702 * Revert "[RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API" This reverts commit e71f79a. * [RZ/A1H]Support RTX v4.80 for Cortex-A and a few Malloc API PR to update RTX to v4.80 for Cortex-A. In the latest CI test, three tests result in an error. The cause is mainly three points - Unsupport RTX v4.80 for Cortex-A - Insufficient stack size of thread defined in test - A few Lack of Malloc API in GCC Therefore, I supported to these. For the stack size for test, @c1728p9 already supported in PR #3362. Reference for issue of test error is here. #3273 Reference for PR to update RTX for Cortex-M. #1702
|
Fixed by #3624, thanks for all of your work @TomoYamanaka! |
0xc0170
pushed a commit
that referenced
this issue
Feb 13, 2017
…NABLED=1" In CI, I receive an mbed assert of stack stats by running the following commands. Its assert is TIMEOUT. mbed test -m RZ_A1H -t GCC_ARM -DMBED_STACK_STATS_ENABLED=1 --clean --compile mbed test -m RZ_A1H -t GCC_ARM --run -n tests-mbed_drivers-stl_features -v So, I fixed the process of related to stack stats (osThreadInfoStackMax).The Detail contents is here. #3273 (comment)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Description
Bug
Target
RZ_A1H
Toolchain:
GCC_ARM, ARM, and IAR
Toolchain version:
GCC_ARM - 4.9.3 20141119
ARM - ARM Compiler 5.06 update 1
IAR - V7.80.1.11864/W32 for ARM
mbed-cli version:
0.9.10meed-os sha:
9d8ec61df5e7442bbde12beabe744ed3a3e13486DAPLink version:
0221 according to DETAILS.TXT, but I think this is running CMSIS-DAP
Expected behavior
All tests pass!
Actual behavior
Not all tests pass!
Steps to reproduce
mbed test -m RZ_A1H -t GCC_ARM -v -n tests-mbed_drivers-race_test,tests-mbedmicro-rtos-mbed-malloc,tests-mbedmicro-rtos-mbed-threadsFailing tests
NOTE: These failures were consistent across all toolchains
tests-mbed_drivers-race_test
tests-mbedmicro-rtos-mbed-malloc
tests-mbedmicro-rtos-mbed-threads
cc @toyowata, @TomoYamanaka
The text was updated successfully, but these errors were encountered: