Xorg r128 driver errors filling "/var"

["John D. Baker" <jdbaker%mylinuxisp.com@localhost>]

Recently, following an update of netbsd-7/i386, my big build system began
running out of space on "/var".  Looking closer, "/var/log/xdm.log" and
"/var/log/Xorg.0.log" had grown to enormous size, consuming some 80% of
my 5GB "/var" filesystem.

These log files end with a continuous stream of:

[...]
(EE) R128(0): Idle timed out, resetting engine...
(EE) R128(0): R128CCEWaitForIdle: CCE stop -9
(EE) R128(0): R128CCEWaitForIdle: CCE start -9
(EE) R128(0): R128CCEWaitForIdle: CCE idle -9
(EE) R128(0): Idle timed out, resetting engine...
(EE) R128(0): R128CCEWaitForIdle: CCE stop -9
(EE) R128(0): R128CCEWaitForIdle: CCE start -9
(EE) R128(0): R128CCEWaitForIdle: CCE idle -9
[...]

The machine uses an AGP ATI Rage128-based video card.  I first noticed
this behavior following an update just before 1 Jan 2015.  The previous
kernel has a date of 31 Dec. 2014, so it was likely a change made to
netbsd-7 xsrc just before then.

I'm using a custom kernel based on GENERIC with all drm drivers except
the "r128drm" UMS driver commented out.

Before that, such messages were seen occasionally, but didn't repeat
so frequently as to fill up "/var"...

So far, my only workaround is to kill -9 xdm (the rc.d script waits for
the PID forever) and X and delete the log files.

Anyone else with an ancient Rage128 video card on i386 seeing anything
like this?

-- 
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645