Threading unfairness
Peter Hansen
peter at engcorp.com
Tue May 21 00:26:55 EDT 2002
Jochen Küpper wrote:
>
> On Sun, 19 May 2002 08:50:41 -0400 Peter Hansen wrote:
>
> Peter> We had been unthinkingly using time.sleep(0.01) for a while,
> Peter> just kind of blindly using it as a free "yield timeslice".
>
> Python uses 'select' for the sleep implementation (besides special
> cases like Max, Windozer, ...).
>
> On Linux the resolution of select is depending on the HZ kernel config
> variable. On x86 it is (standard-wise) 100 (Hz), giving you 10 ms.
So does that mean that while time.sleep(0.01) really sleeps for 10 ms,
time.sleep(0.001) translates to a "yield", or does it round up to
the next tick? (Or if you'd rather teach me how to fish, where's
a good place (please don't say "the source" ;-) to find this out myself?
> You might want to look at something like userspace-RTAI or KURT if you
> do soft-realtime on Linux.
It's working fine without that step so far, but thanks for the tip.
-Peter
More information about the Python-list
mailing list