Re: port-macppc/52712: Graphics corruption with Radeon 9600 card with or without radeonfb

[Michael <macallan%netbsd.org@localhost>]

The following reply was made to PR port-macppc/52712; it has been noted by GNATS.

From: Michael <macallan%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: port-macppc/52712: Graphics corruption with Radeon 9600 card
 with or without radeonfb
Date: Fri, 10 Nov 2017 10:55:47 -0500

 >  On 10/11/2017 04:30, Michael wrote:
 >  >  Please try
 >  >  options RADEONFB_ALWAYS_ACCEL_PUTCHAR  
 >  
 >  That sort of worked.
 >  When the kernel is probing, the display when through switching back
 > and forth between ok, corruption & lots of different colours.
 >  It then settles down by the time system goes multiuser but there's a
 >  large big blue bar an the bottom of the screen. The console is
 > offset by that.
 
 That's the debug code slowing everything down and leaving part of the
 glyph cache visible. So text output works with
 RADEONFB_ALWAYS_ACCEL_PUTCHAR?
 
 >  >  if that helps I know where to look at least.
 >  >  
 >  >  So, as usual, please send me your ofctl -p output, and dmesg output with
 >  >  options RADEONFB_DEBUG  
 >  
 >  Below, would it be worth bumping the dmesg buffer size if RADEONFB_DEBUG
 >  is set so the whole thing is obtainable? I notice there was a plist
 >  printed on the screen which is not obtainable via dmesg(1) at the mement.
 
 The plist is the EDID block read from the monitor ( or the firmware )
 But yeah, too much debug spam to be useful. I'm pretty sure I got rid
 of the DPRINTF in radeonfb_putpal() though.
 
 >  deq0 at ki2c1 Apple Digital Equalizer, addr 0x6a
 
 This tells me your kernel is a few weeks old ;)