[ltp] T21 -- RH 7.1 no MPEGs with xmms

Tue, 08 May 2001 13:20:18 -0400

Tim,

Thanks for responding.

I must say that I am disappointed in IBM's response towards the resolution of
these bugs. It has been a month since we were told that LinuxCare was going to
look at a 1400x1050 ThinkPad. We havent heard from IBM since. There are at least
two graphics problems that I think are unique to our brand of T21s

1) Simultaneous CRT + LCD display.
2) xvideo related problems that doesnt allow us to play MPEGs or DVDs. (This
wasnt broken before xvideo support in the savage driver).

(There is also a third less serious issue where thin lines (1 pixel wide or
so?) are rendered as much thicker lines. This makes for weird looking plots in
applications such as gnuplot, matlab, maple......tools that I use all the time).

If IBM isnt willing to resolve Linux problems on their older models, the least
they could do is divulge information so others can resolve things. I say "older"
models, when it has only been 5 months since IBM started selling T21s. Yet, they
seem to be concentrating their efforts on newer models. These are expensive
machines and if this kind of lacklusture support continues I am going to
seriously consider changing brands when it comes to upgrading my machine in a
couple of years time. Lack of modem support is one thing (we could always use an
external modem), but graphics chip support is an entirely different ball game.

Tim Roberts wrote:

> On Mon, 07 May 2001 18:31:22 -0400, D. Sen wrote:
> >
> >Hmm Markus,
> >
> >Switching between X and VCs (running FB) seems to be much better with the
> >latest BIOS.
> >
> >All this silence from Tim Roberts and IBM makes me wonder if they arent
> >willing to divulge information about the SXGA+ chipset based T21s. I
> >wonder if Dirk Husemann's T21 has a SXGA (1024x768 resolution) chip.
> >Those dont seem to have many of the bugs we are observing. (IBM sells
> >those preloaded with Caldera..........).
>
> This is the first e-mail I've seen on this thread, and then only by accident.
> Apparently, I am being blind-copied on these, but my mailer automatically
> discards as spam any e-mail where my name is not specifically mentioned in
> the headers.
>
> There is no intent on my part to conceal information.  I just do not have the
> information I need.
>
> There are two primary reasons why there has been no resolution on these
> issues.  First, I do not own a T21.  The 1400x1050 T21 has some kind of magic
> in the video BIOS to generate the 1400x1050 display (I do not know why; it
> shouldn't be required), and that magic interferes with the normal operation
> of the X driver.  I do not how I can make any meaningful progress without
> more information on the exact process they are using, and since IBM probably
> considers it to be proprietary, it is extremely unlikely they will be willing
> to share their T21 BIOS modifications with me.  It might be possible to
> reverse engineer some of it, but that requires a significant investment of
> time.
>
> Second, since their buyout by VIA, S3 has lost interest in Linux.  When S3
> renegotiated my contract, they specifically prohibited me from doing any work
> without their express authorization.  In the six weeks since then, there has
> been no communication at all.  I'm a consultant; I need to have someone pay
> for my time.  If they are no longer willing to pay, then the Savage driver
> might be at a dead end.  This remains to be seen.
>
> What you have here is good information, and I will file it in case S3 gets a
> query from IBM.  That's probably the only path that will lead to a
> resolution.
>
> >Markus Alt wrote:
> >
> >> Friedemann Baitinger wrote:
> >> >
> >> > On Mon, 7 May 2001, D. Sen wrote:
> >> >
> >> > > Friedemann Baitinger wrote:
> >> > >
> >> > > > I agree. And to complete the list of problems: Once in X you can't
> >> > > > switch to consoles anymore. More presicely, once you switch to any
> >> > > > of the consoles all you get is a black screen. The same is true
> >> > > > when X is terminated. I have just verified that this still happens
> >> > > > on a freshly installed RedHat 7.1.
> >> > >
> >> > > Well, the latest BIOS seems to have improved the situation a bit with
> >> > > regard to this bug. Markus Alt posted (
> >> > > http://www.bm-soft.com/~bm/archives/200103/msg00170.html )
> >> > > that he got better results by enabling screen expansion during boot
> >> > > time.  This
> >> > > helped until you connected an external monitor (at which point you
> >> > > couldnt get the
> >> > > virtual consoles again). However, with the latest BIOS.......I have
> >> > > had the virtual
> >> > > consoles up for a week.
> >> >
> >> > Good point, I had seen Markus' posting but never followed up. I have now
> >> > just tried both configurations "HV Expansion" Disabled and Enabled and
> >> > guess what: With expansion _enabled_ I get the black screens and with
> >> > expansion _disabled_ I can terminate X and find myself back on a working
> >> > console. Of course it is a small console then but still better than just
> >> > a black screen.
> >>
> >> That's exactly what I meant in my posting back in March, but after
> >> re-reading it, I think that maybe I haven't made myself that clear ...
> >> :-/
> >>
> >> BTW, I have the same effect when trying to watch MPEGs (using KDE's
> >> media player "noatun" and also its predecessor "kaiman"). And IIRC, it
> >> has worked some months ago, so it should be related to the Savage driver
> >> version. (Meanwhile I'm also using 1.1.15 with XFree86 4.0.3 on SuSE
> >> 7.1.)
> >>
> >> I may have older version of the driver on some of my backup tapes, so if
> >> you are interested in getting them for testing, let me know.
> >>
> >> As we are on the topic: There's also still that annoying behaviour of
> >> the broken framebuffer virtual consoles when connecting an external
> >> display and switching between X and the VCs. I hoped that this would be
> >> solved someday, but as you already mentioned, development of the driver
> >> seems to be stuck at the moment.
> >>
> >> Markus
>
> --
> - Tim Roberts, timr@probo.com
>   Providenza & Boekelheide, Inc.

--
D. Sen, Room E167
AT&T Labs-Research
Shannon Laboratory
180 Park Ave.
Florham Park NJ 07932-0971
Ph: 973-360-8546
http://www.research.att.com/~dsen

----- The Linux ThinkPad mailing list -----
The linux-thinkpad mailing list home page is at:
http://www.bm-soft.com/~bm/tp_mailing.html