intelfb not working on a 945GM

intelfb not working on a 945GM

[Paride Legovini]

I'm an owner of a Thinkpad x60s, that is provided with an Intel 945GM
(details: http://www.thinkwiki.org/wiki/Category:X60s). This is the
relevant lspci line:

00:02.0 VGA compatible controller: Intel Corporation Mobile
945GM/GMS/940GML Express Integrated Graphics Controller (rev 03).

This is not working for me. I tried booting with a lot of vga=xxx
options, but I always get a scrambled screen and the system freezes.
Yes, it doesn't continue the boot, so no log is written. If I attach an
external CRT monitor to the VGA port it doesn't get scrambled, but shuts
down as there were no signal, and the system freezes.

Moreover, I can't try to boot the system and THEN change video mode, as
the driver prevents this when an LVDS port is in use. For the same
reason specifying a 'mode:XxY-D@R' option is unuseful: the driver only
accept the vga= option. I've also tried put both (matching, I hope), but
nothing changed (scrambled screen). Here is the output of a dmesg | grep
intelfb (with the debug kernel option enabled, IIRC):

intelfb: intelfb_init
intelfb: Framebuffer driver for Intel(R)
830M/845G/852GM/855GM/865G/915G/915GM/945G/945GM chipsets
intelfb: Version 0.9.4
intelfb: intelfb_setup
intelfb: no options
intelfb: intelfb_pci_register
intelfb: fb aperture: 0xd0000000/0x10000000, MMIO region: 0xee100000/0x80000
intelfb: 00:02.0: Intel(R) 945GM, aperture size 256MB, stolen memory 7932kB
intelfb: fb: 0xd0000000(+ 0x0)/0x7bf000 (0xf8a00000)
intelfb: MMIO: 0xee100000/0x80000 (0xfc200000)
intelfb: ring buffer: 0xd3001000/0x10000 (0xfba01000)
intelfb: HW cursor: 0x0/0x0 (0x00000000) (offset 0x0) (phys 0x0)
intelfb: options: vram = 4, accel = 1, hwcursor = 0, fixed = 0, noinit = 0
intelfb: options: mode = ""
intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode switching.
intelfb: Video mode must be programmed at boot time.
intelfb: cleanup

Both the vesafb and the X11 'intel' driver work really fine. What could
this be? If you need mode informations or want me to make any sort of
test just tell me.

Thank you,
p.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: intelfb not working on a 945GM

[Jesse Barnes]

On Thursday, October 18, 2007, Paride Legovini wrote:
> intelfb: intelfb_init
> intelfb: Framebuffer driver for Intel(R)
> 830M/845G/852GM/855GM/865G/915G/915GM/945G/945GM chipsets
> intelfb: Version 0.9.4
> intelfb: intelfb_setup
> intelfb: no options
> intelfb: intelfb_pci_register
> intelfb: fb aperture: 0xd0000000/0x10000000, MMIO region:
> 0xee100000/0x80000 intelfb: 00:02.0: Intel(R) 945GM, aperture size
> 256MB, stolen memory 7932kB intelfb: fb: 0xd0000000(+ 0x0)/0x7bf000
> (0xf8a00000)
> intelfb: MMIO: 0xee100000/0x80000 (0xfc200000)
> intelfb: ring buffer: 0xd3001000/0x10000 (0xfba01000)
> intelfb: HW cursor: 0x0/0x0 (0x00000000) (offset 0x0) (phys 0x0)
> intelfb: options: vram = 4, accel = 1, hwcursor = 0, fixed = 0,
> noinit = 0 intelfb: options: mode = ""
> intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode
> switching. intelfb: Video mode must be programmed at boot time.
> intelfb: cleanup
>
> Both the vesafb and the X11 'intel' driver work really fine. What
> could this be? If you need mode informations or want me to make any
> sort of test just tell me.

The intelfb driver doesn't support much in the way of modesetting, and 
in fact it can be problematic if you try to use it, X and the DRM at 
the same time, so I'd recommend avoiding it unless you have specific 
embedded requirements or something...

Jesse

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: intelfb not working on a 945GM

[Paride Legovini]

Jesse Barnes wrote:
> On Friday, October 19, 2007 4:20 am Paride Legovini wrote:
>> Jesse Barnes wrote:
>>> The intelfb driver doesn't support much in the way of modesetting,
>>> and in fact it can be problematic if you try to use it, X and the
>>> DRM at the same time, so I'd recommend avoiding it unless you have
>>> specific embedded requirements or something...
>> But the problem I described come out during the very first stages of
>> the boot, before drm or the X driver are loaded (and, moreover, I
>> don't start X at boot, I use startx, and then X loads the modules it
>> needs). So this doesn't seem to be the answer...
> 
> intelfb doesn't support LVDS afaik, so unless you specifically pass it a 
> mode to set your display won't work at all with it.
> 
[...]
> 
> Though again, if you have specific embedded requirements (e.g. using fb 
> applications directly in some specific environment) it might be fine 
> for you.  But even then, you may prefer something like what we have in 
> the DRM modesetting-101 tree, since it can also give you 3D 
> acceleration.

I have no embedded requirements, I just don't always need X, and I like
the framebuffer console in general. Thank you for the modesetting-101
hint, maybe I'll give it a try (I didn't ever know of its existence).

Oh, sorry for writing to you directly and not to the list, I'm too used
to lists where the Reply-To: header is set to the list's address.

p.

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

Re: intelfb not working on a 945GM

[Jesse Barnes]

On Saturday, October 20, 2007 8:53 am Paride Legovini wrote:
> I have no embedded requirements, I just don't always need X, and I
> like the framebuffer console in general. Thank you for the
> modesetting-101 hint, maybe I'll give it a try (I didn't ever know of
> its existence).

Yeah, we're trying to provide a more complete (i.e. 3d accel and other 
advanced features available) alternative to fb for people who don't 
want X (or want other features).  The fb layer is nice and simple, so 
for a lot of applications and devices it's a perfect fit.  If you want 
to use all the features of your card though, you're stuck with X and 
all its baggage.  The modesetting-101 tree is one big step away from 
that; it allows you to run a full 3d stack w/o X and w/o any of the 
conflicts that occur when you use DRM with the fb layer.

> Oh, sorry for writing to you directly and not to the list, I'm too
> used to lists where the Reply-To: header is set to the list's
> address.

No problem, it's conventional to reply to both the list and the 
individual anyway...

Jesse

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/