Task watchdog got triggered in BLE functions (IDFGH-2271) #4413
Comments
|
Hi @ildus Could you please let me know if you have made any changes to default |
|
I get this error on default |
|
@ildus Thanks for reporting. Would you please help check whether your issue has been fixed with the latest master? Thanks. |
|
@Alvin1Zhang @prasad-alatkar sorry, will test today evening (about 20:00 UTC). |
|
Ok, just checked. The problem seems to be gone. Thank you. |
|
Hi there, I still have this same issue with stable version 4.0. and working over the bleprph example with no modifications. (I actually tested also with the master branch getting same results). At the first, I started working with no issues, but after a while it started with watchdog error just after a subscribe event. Then this error log keeps going every 5 seconds forever. I tried flashing it to a new module, and it is working properly for now, bit I think it's about time. In fact, there is no way that my first module can get it to work having the same code. Maybe I lost something? |
|
Hi @albertmbt Can you try |
|
Yep erasing memory solves the problem. I wasn't paying attention when I first got the error, probably it's a problem after flash then and not randomly on work as I thought. I mean, if it gets stuck after a while on working having to erase flash every time it's a problem. |
|
Hi @albertmbt I understand, please provide debug logs if you get this issue again. |
|
@ildus Thanks for reopening. Did you run into the issue again? Could you please help share any updates or details? Thanks. |
|
Hi, I also got into this problem. It happens when there are too many cccd's stored in NVS. I can force this behavior enabling notification for a characteristic. Here is the verbose debug log from the bleprph example: |
|
@albertmbt or @Alvin1Zhang I'm too unfamiliar with the IDF to further investigate the issue, I hope you can solve it. Thanks! |
|
@albertmbt @Unsigus, @prasad-alatkar is looking into this and will update after there is progress. Thanks! |
|
Hi @Unsigus this is actually not the same issue.
Yes, NimBLE stack makes space by deleting oldest peer Please let me know if you find any other issue. Please help to close the issue if you find the response satisfactory. |
|
Hi @prasad-alatkar, Thank you for your time, and will tweak my settings. |
|
@Unsigus Could you please update if your issue was resolved with tweaking settings from menuconfig ?
I do agree that there should be better way to handle this. However it is application developer's responsibility as well to configure minimum required settings. Having said that I am also thinking about proper handling of this particulate situation, will update you accordingly. Apart from that just to maintain history of the issue clean, the issue @Unsigus has faced is different from the original reported issue which is infrequent and |
|
@prasad-alatkar, from my point of view the issue is the same. I also got this message every x seconds, while the app stops responding to everything else: Using With that information I've reproduced the issue that I think @albertmbt has. With the DEBUG log on this was printed directly when turning on notifications:
Changing If this still not the same issue in your opinion, please inform me and I will leave you guys alone. |
|
Hi guys, i've been busy on other project those days and i haven't been able to keep looking at it. i expect to have time during the next week. In the poor test's that i've made those days i haven't been able to reproduce the issue again. i'll keep updating. |
|
@prasad-alatkar I just used old bluetooth stack for my device, and for now I'm not able to debug. But I remember in my case that the issue also was fixed after erasing the flash. But I agree with @Unsigus that is a bug, since it happens without any extra changes on example application, for example I didn't even know that there are some cccd's that are stored to flash. At least it should be logged, and there should be proper workaround. |
|
Hi @Unsigus Could you please share the entire debug log ? |
|
Here you go |
|
Hi @Unsigus Thank you for reporting this issue, I am attaching a temporary patch here to be applied at |
|
Thanks @prasad-alatkar. Sadly I was unable to reproduce the issue before and after the patch. I updated the ESP-IDF to v:latest, no issue there. @albertmbt, do you have time to look into this? I keep trying and get back to this when I know more. |
|
@Unsigus Try to reduce number of CCCDs from menuconfig to reproduce this issue, Apart from this just want to correct my earlier comment
The CCCDs configured are total CCCDs and not per device. |
mahavirj
pushed a commit
to mahavirj/esp-afr-sdk
that referenced
this issue
Jun 1, 2020
… (backport v3.3) Change list: - Reduces the size of the compiled binary, PR: espressif/esp-nimble#6 - Null pointer check, PR: apache/mynewt-nimble#701 - Pairing procedure abort on unexpected req: apache/mynewt-nimble#710 - Fix conn flags after pairing: apache/mynewt-nimble#730 - Remove notification for update process timeout (Vol 6, Part B, section 5.2 ): apache/mynewt-nimble#782 - CCCD fix : apache/mynewt-nimble#790 and apache/mynewt-nimble#804 - Host based Privacy (RPA) fix: espressif/esp-nimble#7 Closes espressif/esp-nimble#10 Closes espressif/esp-idf#4413
espressif-bot
pushed a commit
that referenced
this issue
Jun 15, 2020
… (backport v4.0) Change list: - Reduces the size of the compiled binary, PR: espressif/esp-nimble#6 - Null pointer check, PR: apache/mynewt-nimble#701 - Pairing procedure abort on unexpected req: apache/mynewt-nimble#710 - Fix conn flags after pairing: apache/mynewt-nimble#730 - Remove notification for update process timeout (Vol 6, Part B, section 5.2 ): apache/mynewt-nimble#782 - CCCD fix : apache/mynewt-nimble#790 and apache/mynewt-nimble#804 - Host based Privacy (RPA) fix: espressif/esp-nimble#7 Closes espressif/esp-nimble#10 Closes #4413
projectgus
pushed a commit
that referenced
this issue
Jul 22, 2020
… (backport v4.1) Change list: - Reduces the size of the compiled binary, PR: espressif/esp-nimble#6 - Null pointer check, PR: apache/mynewt-nimble#701 - Pairing procedure abort on unexpected req: apache/mynewt-nimble#710 - Fix conn flags after pairing: apache/mynewt-nimble#730 - Remove notification for update process timeout (Vol 6, Part B, section 5.2 ): apache/mynewt-nimble#782 - CCCD fix : apache/mynewt-nimble#790 and apache/mynewt-nimble#804 - Host based Privacy (RPA) fix: espressif/esp-nimble#7 Closes espressif/esp-nimble#10 Closes #4413
|
Hi I ma not using Nimble though, so should I open a new issue for this ? erasing the flash did not help. After getting the above error, I get this: |
3 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Recently I have started experimenting with NimBLE and in after some time I started to get these errors:
I made some research and found out that this error potentionally could be fixed by putting
vTaskDelayin these functions:nimble_port_runble_ll_taskI'm not sure that's the right way to fix the problem so please let me know if there is more information needed. The problem occures on
masterbranch, and I have not checked older branches.This happens on ESP-WROOM-32 on
bleprphexample.The text was updated successfully, but these errors were encountered: