bofh19 / yappi

Automatically exported from code.google.com/p/yappi
MIT License
0 stars 0 forks source link

Repeating methods in statistics #39

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago
I am running yappi 0.62 on Windows and I get unexpected results. When calling 
print_stats() I get methods which repeat in several lines, in each line with 
different time and ncall statistics. Also, I get methods with tsub equal zero, 
where it is obvious that it's impossible.
How can these be explained?

Original issue reported on code.google.com by yariv.ma...@algowiz.com on 11 Feb 2013 at 4:53

GoogleCodeExporter commented 9 years ago
If there are mutiple classes using same function names, it may seem same. 
However, this is already done in the latest head by adding the class name at 
athe start of the function name. Can yiou try the latest head in the repository?

Original comment by sum...@gmail.com on 15 Feb 2013 at 9:39

GoogleCodeExporter commented 9 years ago
There are no duplicate function names in the code. Also, the line-number is the 
same. 

Original comment by yariv.ma...@algowiz.com on 15 Feb 2013 at 10:58

GoogleCodeExporter commented 9 years ago
[deleted comment]
GoogleCodeExporter commented 9 years ago
Hi,

I am not able to reproduce the issue in my setup. Is it possible you can send 
the script? Or to make the issue more speficic, can you please provide the 
following information:

-is it application is multithreaded? 

- Is the duplicate functions are class methods or module functions? 

- If the app is multithreaded, are these duplicate functions being called from 
different threads?

Thanks,

Original comment by sum...@gmail.com on 5 Mar 2013 at 12:07

GoogleCodeExporter commented 9 years ago
I can reproduce this error currently on some applications. This is due to 
following:

- get_stats_xxx() is called while profiler is running.
- the internal profile functions are still being profiled, which causes the 
results to change any maybe internal hash table grow--> which again causes the 
resultrs to be corrupted.

This is fixed by pause/resume while getting results. Fixed in 
https://bitbucket.org/sumerc/yappi/commits/8f637760be21aeaf46d0c158dc84827332014
dc6

Do you have any chance to test the latest head in the rep in your application?

Original comment by sum...@gmail.com on 8 Nov 2013 at 1:02

GoogleCodeExporter commented 9 years ago

Original comment by sum...@gmail.com on 8 Nov 2013 at 1:02

GoogleCodeExporter commented 9 years ago

Original comment by sum...@gmail.com on 8 Nov 2013 at 1:02

GoogleCodeExporter commented 9 years ago

Original comment by sum...@gmail.com on 6 Jan 2014 at 1:04

GoogleCodeExporter commented 9 years ago

Original comment by sum...@gmail.com on 30 Jan 2014 at 9:56