This is the mail archive of the
mailing list for the Cygwin project.
Re: Cygwin signal bandwidth
- From: Rolf Campbell <rcampbell at tropicnetworks dot com>
- To: cygwin at cygwin dot com
- Date: Sun, 09 Feb 2003 14:05:18 -0500
- Subject: Re: Cygwin signal bandwidth
- References: <email@example.com>
Sorry about the multiple posts, I was trying to use gmane and it delayed
delivery of my messages for 90 minutes, so I did the naive thing and
sent it two more times.
Rolf Campbell wrote:
While working on a project, I had the need to run a function every
millisecond. So, my first attempt was to fork() and then loop sending a
signal then delaying 1ms. What I found was that a cygwin process can
only recieve about 100 signals/sec before it uses up 100% of the
processor (on a PIII/850MHz).
Is it reasonable for me to conclude that cygwin uses about 8,000,000
cycles to process a single signal, or have I done something wrong?
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Bug reporting: http://cygwin.com/bugs.html