5.1 - should I?

[Kevin Oberman]

> From: Michael Collette <metrol at metrol.net>
> Date: Mon, 16 Jun 2003 14:24:58 -0700
> 
> On Monday 16 June 2003 01:59 pm, Kevin Oberman wrote:
> > > 5-CURRENT
> > >
> > > 1. Will it properly talk with this ATI card?
> >
> > No better than 4.8, although I have not seen your problem with this
> > chip. The vtys never fail on my system. Only the graphics port has a
> > problem. The system can't deal with the display powering on while
> > running 1400x1050. Do you know what the R40 screen runs at in vty? If
> > it's 1400x1050, that would explain a lot!
> 
> Not sure what you're asking here.  vty?  I can say that the native resolution
> for this display is 1400x1050.

As it is for my T30. But, on the T30 and at least several other models
from Dell and Compaq, the text mode (which is what I was calling vty)
is at only half that resolution, 700x525. The chip has not issues with
the display being cycled on at this resolution.

I (and some others) call it vty mode because that is the mode you are
in when running on vty0-7 (Alt-F0 to Alt-F8). You must add CTRL if you
are in X as X grabs all ALT keys. E.g. <CTRL-ALT-F1>.

The standard trick to get the display back on these units is to
either:
1. Switch to a text display before suspending the unit (see man apmd)

2. Switch to a text display and cycle the display off and on (Fn-F3 to
turn it off and any key to turn it on). This only works on my T30 with
APM.

> > > 2. Will it resolve my power management issues?
> >
> > Not sure what power management issues you have, so I can't say much.
> > APM is about the same. ACPI has a poor track record on ThinkPads, but
> > I have reports that it is working on some models. The R40 is very new,
> > so, maybe.
> 
> When I suspend this box it has a problem with waking up.  It will start going
> through the motions of waking up, then lock up hard shortly after reseting 
> the ata devices.  Also having a problem, perhaps related, with the screen 
> being turned off.  I can see the LCD backlight is getting power, but no 
> display is rendered.  In both cases a hard power down is required to get the 
> box running again.

Does this happen when in text mode or only in X?

> > I do know that I can't put my units to sleep when they are on AC
> > power. I have to unplug the AC and then hit suspend. I THINK that is
> > the only problem I've seen.
> 
> I can put it to sleep just fine.  Either with the Fn-F4 combo, or with the zzz 
> command.  The only problem I'm seeing here is the waking up portion of the 
> equation.

I'd say that is a fairly serious limitation! I wonder if it's related
to the BIOS configuration? I'd suggest going through the power
management stuff with ps2. (ps2 is available on Windows as well as as
a bootable floppy on the IBM web site.)

> > > 3. Built in wireless working?
> >
> > Maybe. If you have the Cisco card, it will work. I belive that the
> > Intel card uses  a Symbol chip and is not yet supported. Maybe someone
> > can confirm this. (My T30 uses a Prism 2.5 chip.)
> 
> Not sure what this thing has in it.  I'm not seeing anything showing
> up on the dmesg.  Honestly, I haven't been looking that hard as I
> don't have a wireless hub handy to test this thing with.  It'd be
> nice if it works is all.

The one thing that does not work now is the modem. It is being worked
on, but it may be a while. I suspect support for the Symbol chips
might be on the horizon, as well, but I have not heard any clear words
that it's being worked on.

> > > 4. Anyone out there with a similar setup on 5?
> >
> > Yes, but running on T30 and 600E systems. I can send you the config,
> > XF86Config, device.hints, and loader.conf files.
> 
> Consider the flood gates wide open.  I'd love to see any and all applicable 
> configuration files that might help me through this.  Might also be helpful 
> to see your kernel config as well.
> 
> I do have X up and running.  KDE and a set of applications all compiled and 
> just about configured for regular use.  If I had managed to muddle through a 
> workable power management solution I would have been ready to hand this box 
> on over.  As it is, things just aren't stable enough for me to do that at 
> this time.

OK. I'll mail them out in a separate message. The whole list probably
does not care. I suspect that the XF86Config won't help, but the rest
might.
-- 
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net			Phone: +1 510 486-8634