Gerard R. Tyra (gerry++at++sa-tech.com)
Mon, 08 Dec 1997 13:32:46 -0800
Okay, correct me if I am wrong (the finer points of the graphics pipe
are a moving target). I am under the impression that graphics commands
are put on a FIFO while waiting for the graphics hardware to execute
them. While multiple SWAPBUFFERs can be issued within micro-seconds of
each other, they are not executed by the hardware until the FIFO has
cleared. Hence, there can be a significant (measured in micro-seconds)
difference in the execution time of the SWAPBUFFERs on the various
pipes. The greater the difference in pipe loading the greater the
difference in the SWAPBUFFERs.
If this difference occurs across a vertical retrace, you do get a
momentary break between the channels. Even one break every couple of
seconds is extremely noticeable.
Using a software lock on the swap reduces the probability of an async
swap, but only hardware (swap ready) guarantees a lock.
Gerry
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:56:21 PDT