Nicolas Williams wrote: > On Tue, Jan 29, 2008 at 01:35:01PM -0800, Dan Price wrote: >> Finally, it seems to me that this might be more easily fixed by masking >> the offending signal in all threads except for the dedicated >> signal catcher-- in which case the signal will be forced to go to >> that thread and only that thread; this could be done with some signal >> manipulation in startd_thread_create(). > > Simpler still: have the signal handler port_send() a user-defined event > that causes wait_thread() to exit(). > _______________________________________________ > smf-discuss mailing list > smf-discuss at opensolaris.org
Or better yet, use the port_alert() interface; this is exactly the sort of thing it was designed for.... - Bart -- Bart Smaalders Solaris Kernel Performance barts at cyber.eng.sun.com http://blogs.sun.com/barts
