Hi,
I know this has been asked before, but I did not come across any
solutions, so let me appologise if I missed them.
I have a: ibook 2 (white ibook from summer of 2001) 500MHz. It has a ATI
Rage 128 graphics card which my OF calls ATY,RageM3p29s. The aty128fb
module is used for video framebuffer, and I can't seem to get VGA mirror
to work.
case 1: Monitor connected at boot up
The external monitor is fine at depth 8, resolution 1024x768-8@60Hz.
Changing the depth (text mode) with fbset makes the font look like the
wrong depth. It feels like font depth has changed but not the display
depth. The internal-LCD display stays just fine.
case 2: Monitor not connected at boot up
The external monitor is "wavy". I turn the LCD off (m3mirror from benh)
and play around with timings, resolutions etc. The "waviness" gets more or
less but I could not find setting to make it stable. BTW, it is true for
both a CRT and flat-panel. Internal-LCD is fine (at 1024x768).
Has this been solved? If not, how can I help?
Regards,
manuj
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
case 1: Monitor connected at boot up
The external monitor is fine at depth 8, resolution 1024x768-8@60Hz.
Changing the depth (text mode) with fbset makes the font look like the
wrong depth. It feels like font depth has changed but not the display
depth. The internal-LCD display stays just fine.
Starting X will screw up the secondary display in this scenario too.
Has this been solved? If not, how can I help?
Nope. I wrote some tools to dump the Rage128's config registers as well
the PLL but haven't had a chance to actually get dumps.
Anyways, its located at http://www.stampflee.com/r128dump
It would useful for dumps (send them to me, not the list...) booting
with both the scenarios you provided and when booting with
"video=ofonly"
Owen
--
Owen Stampflee - owen@stampflee.com - http://www.stampflee.com
GnuPG Fingerprint: EDE6 48EC 4E65 BE83 D405 B091 B170 8AF0 F875 65E0
Download My Public Key: http://www.stampflee.com/owen.asc
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
From: Harald Welte <laforge@gnumonks.org> Date: 2003-02-09 10:44:14
On Mon, Jan 20, 2003 at 10:59:21AM -0500, Manuj Dhingra wrote:
I know this has been asked before, but I did not come across any
solutions, so let me appologise if I missed them.
I have a: ibook 2 (white ibook from summer of 2001) 500MHz. It has a ATI
Rage 128 graphics card which my OF calls ATY,RageM3p29s. The aty128fb
module is used for video framebuffer, and I can't seem to get VGA mirror
to work.
I have exactly the same model.
case 2: Monitor not connected at boot up
The external monitor is "wavy". I turn the LCD off (m3mirror from benh)
and play around with timings, resolutions etc. The "waviness" gets more or
less but I could not find setting to make it stable. BTW, it is true for
both a CRT and flat-panel. Internal-LCD is fine (at 1024x768).
Yes, I can confirm this behaviour. It's really annoying because I
cannot use my iBook2 for any of my presentations at conferences... and
I'm really too lazy to carry a second old x86 notebook with me.
From: Michael Schmitz <hidden> Date: 2003-02-10 10:06:04
quoted
case 2: Monitor not connected at boot up
The external monitor is "wavy". I turn the LCD off (m3mirror from benh)
and play around with timings, resolutions etc. The "waviness" gets more or
less but I could not find setting to make it stable. BTW, it is true for
both a CRT and flat-panel. Internal-LCD is fine (at 1024x768).
Yes, I can confirm this behaviour. It's really annoying because I
cannot use my iBook2 for any of my presentations at conferences... and
I'm really too lazy to carry a second old x86 notebook with me.
quoted
Has this been solved? If not, how can I help?
I'm not aware of any solution :(
Someone lately posted a message detailing the problem (either here or on
debian-powerpc): the hsync and vsync pulses appear to be OK (correct
frequency) but have a phase jitter. Does that ring a bell with anyone
here? Does anyone know which PLL setting the hsync position is derived
from?
Michael
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
On Feb 10, 2003 05:06 Monday Michael Schmitz wrote:
quoted
quoted
case 2: Monitor not connected at boot up
Someone lately posted a message detailing the problem (either here or on
debian-powerpc): the hsync and vsync pulses appear to be OK (correct
frequency) but have a phase jitter.
Unless somebody else posted it too, that would be me. And I now realise
that it was based on your suggestion (on debian-powerpc) a year back!
Owen asked: How was it using OFFB? The phase is steady with the OFFB.
Does that ring a bell with anyone
here? Does anyone know which PLL setting the hsync position is derived
from?
Michael
Is it possible that the lcd and crt *have to be* on different
controllers? From what I understand, OF and OSX set it up that way but we
don't. As I have no documentation, this is just a conjecture.
Still I would like to do just that. But I am constrained by the lack of
documentation. Pointers anybody? Maybe people who desire a dual-head would
like to help :)
regards,
-manuj
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
From: Michael Schmitz <hidden> Date: 2003-02-10 16:24:21
quoted
Someone lately posted a message detailing the problem (either here or on
debian-powerpc): the hsync and vsync pulses appear to be OK (correct
frequency) but have a phase jitter.
Unless somebody else posted it too, that would be me. And I now realise
that it was based on your suggestion (on debian-powerpc) a year back!
Possible - I've suggested someone take a scope and debug this more than
once :-)
Owen asked: How was it using OFFB? The phase is steady with the OFFB.
quoted
Does that ring a bell with anyone
here? Does anyone know which PLL setting the hsync position is derived
from?
Michael
Is it possible that the lcd and crt *have to be* on different
controllers? From what I understand, OF and OSX set it up that way but we
Yep, that's entirely possible. Maybe we can't get the proper specs for
both on one controller.
don't. As I have no documentation, this is just a conjecture.
Still I would like to do just that. But I am constrained by the lack of
documentation. Pointers anybody? Maybe people who desire a dual-head would
like to help :)
I'd talk to Michel Dänzer about this. The XFree people might have figured
something out on other (i386) architectures. XFree developers have access
to at least some ATI docs.
Michael
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
From: Michel Dänzer <hidden> Date: 2003-02-10 17:27:33
On Mon, 2003-02-10 at 17:24, Michael Schmitz wrote:
quoted
Owen asked: How was it using OFFB? The phase is steady with the OFFB.
quoted
Does that ring a bell with anyone
here? Does anyone know which PLL setting the hsync position is derived
from?
Michael
Is it possible that the lcd and crt *have to be* on different
controllers? From what I understand, OF and OSX set it up that way but we
Yep, that's entirely possible. Maybe we can't get the proper specs for
both on one controller.
quoted
don't. As I have no documentation, this is just a conjecture.
Still I would like to do just that. But I am constrained by the lack of
documentation. Pointers anybody? Maybe people who desire a dual-head would
like to help :)
I'd talk to Michel Dänzer about this. The XFree people might have figured
something out on other (i386) architectures. XFree developers have access
to at least some ATI docs.
The XFree86 radeon driver has indeed solved this by using both CRTCs.
The code is there for everyone to look at. ;)
--
Earthling Michel Dänzer (MrCooper)/ Debian GNU/Linux (powerpc) developer
XFree86 and DRI project member / CS student, Free Software enthusiast
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
I am running with external Cinema 23" display on my TiBook IV (Radeon
Mobility 9000).
The combination works great with earily 2.4.20 kernels, however with
2.4.20-ben5 my LCD stays on and goes white.
I managed to determine that LCD is not turned on by radeonfb, it turns own
earlier in the boot process. So even with video=ofonly the LCD stays on.
My knowledge of kernel workings is not sufficent to determine where
exactely the problem occurs. I would appreciate any lead into the right
direction...
Andrew
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
From: Benjamin Herrenschmidt <benh@kernel.crashing.org> Date: 2003-02-14 11:06:40
On Thu, 2003-02-13 at 23:52, daRonin wrote:
I am running with external Cinema 23" display on my TiBook IV (Radeon
Mobility 9000).
The combination works great with earily 2.4.20 kernels, however with
2.4.20-ben5 my LCD stays on and goes white.
I managed to determine that LCD is not turned on by radeonfb, it turns own
earlier in the boot process. So even with video=ofonly the LCD stays on.
My knowledge of kernel workings is not sufficent to determine where
exactely the problem occurs. I would appreciate any lead into the right
direction...
Please send me your latest working version of radeonfb.c
Ben.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/