Hi all,I'd like to fix or replace the current logging code in Arducopter. It works but can fail in some circumstances which I can't repeat or predict.We need to keep the current messages and but can replace the logic for log management.If you are interested, please have a look at the code and let me know.Thanks!Jason
You need to be a member of diydrones to add comments!
There is a problem in start_new_log function in Log.pde.
When num_existing_logs equals 0, after erasing log for example, the second if is end_pages[num_existing_logs - 1] which is end_pages[- 1] which is impossible or in C somewhere in the memory out of the end_pages array and then the value is unexpected.
I looked at the code briefly last night and I think I can help - at least look for what is happening. When you say "fail" do you mean a corrupted log/no log recorded, or are there dropped samples or corrupted samples. I haven't noticed any log issues to speak of in my short testing over the last week, but I'm interested in looking at the timing of the eeprom writes with respect to the high rate loop and any other timing issues with my logic analyzer and adding some bit blips as outputs to the code to watch while the log is writing.
Replies
Jason,
There is a problem in start_new_log function in Log.pde.
When num_existing_logs equals 0, after erasing log for example, the second if is end_pages[num_existing_logs - 1] which is end_pages[- 1] which is impossible or in C somewhere in the memory out of the end_pages array and then the value is unexpected.
I just experiment this
erasing logs
reboot in flying mode
num_existing_logs = 0
end_pages[num_existing_logs - 1] = 6146
no log created !
At least one should modify the if with :
if(num_existing_logs==0 || ((end_pages[num_existing_logs - 1] < 4095) && (num_existing_logs < MAX_NUM_LOGS /*50*/))) {Don't know if this will fix all the log problems but should contribute to.
Olivier.
Jason,
I looked at the code briefly last night and I think I can help - at least look for what is happening. When you say "fail" do you mean a corrupted log/no log recorded, or are there dropped samples or corrupted samples. I haven't noticed any log issues to speak of in my short testing over the last week, but I'm interested in looking at the timing of the eeprom writes with respect to the high rate loop and any other timing issues with my logic analyzer and adding some bit blips as outputs to the code to watch while the log is writing.
Heino