Re: Antialiasing and other RE stuff

New Message Reply Date view Thread view Subject view Author view

Allan Schaffer (aschaffe)
Thu, 1 Dec 1994 13:24:49 -0800

>
> I have had similar problems on our ONYX (OS 5.2, 2 CPU's, 2 RM's),
> except it appears as a transparency problem (maybe AA too, it's hard
> to tell in NTSC). I ran at 4++at++640x486_30i. I have seen the same
> problem when I run at 1600x1200 (small pixel depth), so for some
> reason the machine thinks it's in small pixel depth mode when no one
> is logged in to the console. We are running the visuallogin. The
> problem is not there if I log on to the console.

This is something that has been mentioned a few times before; I'll
add it to the FAQ list. This is a bug of sorts; it is caused by
'clogin' and 'chost' using the ImageVision library to display bitmap
images of users & machine types. On RealityEngine the IL uses
hardware acceleration to scale the images, which gobbles up
framebuffer memory.

The symptoms show up when Performer (or any other GL application)
tries to configure the framebuffer. Since the framebuffer memory is
already in use (by clogin or chost) Performer has to back off & ask
for something less.

It's possible to work around this, somewhat:

    setenv IL_HW_ACCELERATE 0

should force the IL to do its' computations on the host. Or to
work around it for just 'clogin' you can 'chkconfig visuallogin off'.
Or just log in.

Allan

-- 
Allan Schaffer
Silicon Graphics
aschaffe++at++sgi.com

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:50:43 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.