> From [EMAIL PROTECTED] Wed May  2 12:33:44 2007
> 
> I spent some more time on this today. I am still under the impression
> that there is a race between F_SETOWN/F_SETSIG and a counter overflow
> with notification.
> 
> Is there a way you could modifiy your program such that no thread
> call pfm_start() before all setups are done? Then I suspect the
> problem will disappear.

I hate to be the bearer of bad news, but I have done as you suggest and
the problem remains. If you'd care to see my current test code it is at
http://plg.uwaterloo.ca/~rcbilson/sigio.cc

As well, I have obtained evidence using strace that the problem can
occur after the initialization phase. There is a sample "strace -f" log
in http://plg.uwaterloo.ca/~rcbilson/perfmon-strace for an experiment
with five worker threads. At the point where the program fails each of
the five threads has been created and has successfully processed a
number of SIGIOs.

The machine in question underwent an OS upgrade yesterday, and now runs
2.6.20.3 + ubuntu patches + perfmon 070209. Now the problem occurs much
more consistently than before (the previous kernel was 2.6.20.0 +
perfmon 070209).
_______________________________________________
perfmon mailing list
[email protected]
http://www.hpl.hp.com/hosted/linux/mail-archives/perfmon/

Reply via email to