Re: port-macppc/48255 Unreadable console output when netbooting macppc

[Michael <macallan1888%gmail.com@localhost>]

The following reply was made to PR port-macppc/48255; it has been noted by GNATS.

From: Michael <macallan1888%gmail.com@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: port-macppc/48255 Unreadable console output when netbooting
 macppc
Date: Sat, 7 Aug 2021 00:20:45 -0400

 On Sat,  7 Aug 2021 02:40:02 +0000 (UTC)
 David Holland <dholland-bugs%netbsd.org@localhost> wrote:
 
 > The following reply was made to PR port-macppc/48255; it has been
 > noted by GNATS.
 > 
 > From: David Holland <dholland-bugs%netbsd.org@localhost>
 > To: gnats-bugs%NetBSD.org@localhost
 > Cc: 
 > Subject: Re: port-macppc/48255 Unreadable console output when
 > netbooting macppc
 > Date: Sat, 7 Aug 2021 02:39:33 +0000
 > 
 >  On Wed, Jun 13, 2018 at 02:00:02AM +0000, Sevan Janiyan wrote:
 >   >  This issue was resolved by defining
 >   > RADEONFB_ALWAYS_ACCEL_PUTCHAR option in kernel and is set by
 >   > default in GENERIC kernel config in -HEAD. Note this along with
 >   > radeonfb were disabled in netbsd-8 as radeonfb causes different
 >   > issues with newer r9550 chipset graphics cards found in later
 >   > model system (specifically the last gen ibook g4s)  
 >  
 >  Does that mean the PR can be closed?
 
 In theory, yes.
 Sevan sent me an iBook which reproduced this bug and I figured out
 what's wrong - basically OF runs the radeon with tiling enabled which
 radeonfb didn't expect. It only disabled tiling for the video output
 logic and the blitter but not for video memory access via AGP, which is
 why drawing characters by hardware worked while software drawing did
 not, which ALWAYS_ACCEL_PUTCHAR worked around.
 I committed a proper fix a while ago - disable tiling translation for
 video RAM access from the host as well - I do not remember if it was
 pulled up into 9, I'll check.
 
 have fun
 Michael