On Thu, 15 Jan 2004 Jeff Trawick wrote :
> data race?  consider http://nagoya.apache.org/bugzilla/show_bug.cgi?id=25520

Thanks, your information is very useful to me. Some how when I use "data race"
as key words to search the data base, these bugs didn't show up. Now I am using
"thread safe" as key words.

> various bugs with file descriptors could lead to race conditions with 
> data going to the wrong client...  consider CAN-2003-0789, fixed in 
> Apache 2.0.48
> 
> misuse of system library functions in non-thread-safe manner?  consider 
> CAN-2003-0189, fixed a few releases ago (can't remember which)
> 
> as far as why there haven't been more (those three aren't the only ones, 
> but they haven't been prevalent in general): the Apache-specific data 
> structure use inherited from Apache 1.3 was thread-safe, so no 
> large-scale rework of core code to make it threadsafe was necessary...

Looking at these bugs, it become obvious why my system was not hitting any
of them. I am running relatively new solaris 9 and the server's execution
path is very straightforward. I just have the server fetch static files
from a directory. Even though I had 64 threads per child and 1600 threads
in total, I still didn't have any data races detected by my detector. 

I did a little more research, it seems the data race bugs show up usually
in some specific module with specific conditions. I wonder how can I setup
my server to test those unusual conditions quickly? Or, even better, are
there any known crashes caused by unknown race conditions?

Cheers,

-Min

-- 
Help them: http://www.ocef.org/newocef/en/

Reply via email to