20 Mar
2003
20 Mar
'03
12:30 a.m.
I don't recall the exact situation now. It might have been a Linux poll issue, or that Pike loops over every socket that made poll exit until it calls poll again (which is a good idea, generally?).
Is there a dumped infokom archive somewhere?
/ Mirar
Previous text:
2003-03-19 23:24: Subject: I/O callbacks in blocking mode
You did, but it isn't enabled by default, it doesn't provide full protection against a race, and it hurts performance. Do you mean a previous blocking read on some systems makes poll/select return on the fd even when the data has been read out already?
/ Martin Stjernholm, Roxen IS