Sproadic LoadProhibited (CON-907) #748
Comments
github-actions
bot
changed the title
Does this happen everytime? With the default examples in esp-matter? |
|
I have this issue as well. With the default examples in esp-matter but with PSRAM enabled (NimBLE on top of PSRAM as well) |
|
Crash log: |
|
Here my sdkconfig.zip |
|
Any updates? I'm having this issue too. |
|
@shubhamdp It's been quite a while and this issue still exists at release/v1.2 branch, how can this be fixed? I am on ESP32-PICO-MINI-02 |
|
I'm having this issue. How can this be fixed? |
|
As I opened this issue and am currently not suffering from this issue, I think it was very likely caused by a memory leak/misuse in the main application. So I would say, the Kernel structures are likely to be damaged by the tasks. Please be sure:
|
I don't think this is true, as it also happens on a blank light project example. |
|
@PhLuReh @law-ko We've found the problem and we'll release the fix soon. Till then, can you please check with this patch if your crash disappears? Edit: This patch is on release/v1.2 branch |
|
I'm not able to test this, as I'm still tied to v1.1 due to my test lab, which has not shifted to v1.2 so far. |
|
@PhLuReh did you see this problem on v1.1 as well? |
|
As I wrote in #748 (comment) - I saw it, that's why I raised this issue. But p to now, it did not come up again. |
|
@shubhamdp Tested and randomly got this error: |
|
@law-ko can you please share the esp-matter or connectedhomeip commit id? Also, please specify if you are using specific branch of esp-matter. This crash is similar to project-chip/connectedhomeip#28142 and was fixed in project-chip/connectedhomeip#29180. |
|
@shubhamdp Upon further testing, the error still exists even after applying the patch: Branch: Does this patch need to be on the latest commit ID on the |
|
I did use the latest release/v1.2 for reproducing the crash, and fixing it. Please give a try there, its an intermittent crash (I guess). If you can, please help test two more minor changes separately. These are workarounds I tried to figure out the root cause of crash.
|
|
@law-ko any updates here? |
Testing with the first option adding delay with 1000 ms works now, will continue to observe as this is a intermittent crash. |
|
@law-ko I have put up some solid fix here, PTAL: project-chip/connectedhomeip#33109 |
Does this mean the first option of adding 1000ms is not necessary? |
|
Yes, the above suggested work arounds won't be necessary, I'll post once we merge the changes |
|
@shubhamdp Is the merge completed yet? |
|
@shubhamdp I don't think it is fixed, this is the crash log: |
|
@law-ko Thanks for trying it out, this is a fresh one and the one due to ble deinit race is fixed. For this one, can you tell me please tell me the espressif/mdns version that you are using. Open up the If its |
@shubhamdp The version I found in Under Note this is an intermittent issue as well. |
|
Please let me know if this occurs on |
|
@shubhamdp I think this occurs after merging as well: |
|
@law-ko please share the sdkconfig, esp-idf and esp-matter commits. Do you have any patches on top of esp-matter, esp-idf? In which example this reproducible? Anything different that you are doing in your application? Any more details that would help debug this issue. As per time stamp it looks like its getting hit very late, around 20s after bootup, anything different going on before starting matter? |
@shubhamdp We do not have additional patches on top of it, but we have other programs that we run in other tasks. We are using PSRAM as well. This can occur once after the boot up, after Wi-Fi is connected. The timestamp late is due to copying the latter crash log. |
|
Keep in mind we are using ESP-PICO-MINI-02 with BLE on top of DRAM and Matter stack on top of PSRAM. A full log from boot: SDKCONFIG: |
|
@shubhamdp Another crash log: I've changed BLE to CPU1.. |
|
@shubhamdp Another crash.. |
|
@shubhamdp Please see this issue for related-crash: #1079 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the bug
While starting up, the FreeRTOS spradically get's killed by a LoadProhibited Exception. I already adressed this issue to the esp-idf, because I thought it may be idf related or even related to the linked issue. But they told me to report the error here.
Backtrace (full log attached below):
Maybe this issue #722 is also related.
Environment
Linux version 6.1.29-1-MANJARO (builduser@fv-az292-908) (gcc (GCC) 12.2.1 20230201, GNU ld (GNU Binutils) 2.40) #1 SMP PREEMPT_DYNAMIC Wed May 17 14:00:55 UTC 2023Any additional details
I modified my code to figure out what was the source of this exception. My list.c has this diff
I think it could be a race condition or some illegal memory access. As posted on the esp-idf repo i could drill it down to this location:
if (pxItemToRemove < (ListItem_t* const)0xff) abort();So I would assume, pxItemToRemove get's corrupted, is not NULL but according to the BT pxItemToRemove was 0x0000000a. Did not get any clue how up to now.
The text was updated successfully, but these errors were encountered: