--Signature_Tue__1_Nov_2005_11_20_44_-0500_l1+OmE+V=uC=c4KP
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

Hello,

> I've had it again. It looked quite a bit different this time, no
> "meltdown" effect, but headache-inducing, flickering, changing,
> repetitive, colourful, rectangular patterns, looks like DAC-register
> messup, too. Unlike last time, though, this time the display didn't
> recover after restarting the Xserver, and I had to login remotely
> and shutdown the machine. Is the difference between microSPARC (GX)
> and TURBOsparc (TX) likely to influence the behaviour?

Well, the TurboSPARC has L2 caches, the microSPARC doesn't, the cores
should be more or less the same. I'll check if anything was mapped
cacheable, maybe that's the reason. Did anything like this ever happen
in the console?

> For the time being I'm using software cursor. If it happens again
> we know it's something else at least.

I'm pretty sure it's caused by accessing DAC registers from userland.
Maybe I need to move the cursor sprite handling into the kernel, there
at least I can be sure that no interrupt or anything else gets in the
way.

> > Before it happened sooner or later, usually I've been able to
> > reproduce it within minutes.
>=20
> Yes, it's very different now, and happens only occasionally, whereas
> it used to be very frequently.

It shouldn't stomp over all DAC registers anymore, this may explain the
different look.

have fun
Michael

--Signature_Tue__1_Nov_2005_11_20_44_-0500_l1+OmE+V=uC=c4KP
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (NetBSD)

iQEVAwUBQ2eV58pnzkX8Yg2nAQIpngf/cvs/CU3nkfV5wbVCFPUWnwXo997CbhrV
/NdqMDrm9JYvV9/XDNgZSoTExUzeT+qw4nC7Wh4KCrexCogefkRPFIgbYWu2tOpQ
o77EbGtDPSXWk0DDnaDGBVeqyDr7f2wdoZjymV4tG7s4qx+k+B+vMvykOqYFmQPM
wkJ0BYS3TwnnyJljdHr9jryG59mO4z00BkRW7lGsQm3U7GOpjDmMAIM7WjWCPhJN
L/dugfuYN4SUrAW7MjLTeDlETJg/2egdxYZcssf863CQEuFx6qQ9QNZEUES84jbB
i92YdThYOw9/z8S5J1Ja+Fl0sXuaib3jG2lahuaIETRJSN90HJgGGA==
=im0C
-----END PGP SIGNATURE-----

--Signature_Tue__1_Nov_2005_11_20_44_-0500_l1+OmE+V=uC=c4KP--