On 14 February 2008 22:52, Brian Ford wrote: > On Wed, 13 Feb 2008, Dave Korn wrote:
>> That diff is absolutely reliably consistently 2140 for me. I'd >> speculate that maybe the cygwin signal emulation requires a context >> switch or two, to or away from the signal processing thread, and so >> we're seeing an extra quantum or two of constant overhead for systematic >> reasons. > > Yep, two. The minimum quantum on NT based OSes is 1024. See this > thread for support of your hypothesis: > > http://cygwin.com/ml/cygwin-developers/2007-10/msg00040.html Ahh, I remember that. Hmm. I wonder if it would ever be possible to avoid context switches by doing signal processing in-thread. It needn't necessarily matter /which/ thread "is" the signal thread as long as there's only one at a time.... cheers, DaveK -- Can't think of a witty .sigline today.... -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
