linuxconsole-dev

2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Yesterday I got two xservers running with 2.6.0-test6-ruby
Both screens are connected to the matrox G550.
The one running on the main console uses the normal
accelerated X server, and a keyboard connected to the mouse port.
The X config for this display isn't modified, it is
the single-user setup I always had.

The other screen uses the unaccelerated framebuffer driver
on /dev/fb/1, and a keyboard connected to the normal keyboard
connector.  Both mice are serial.

Framebuffer X was easy to set up.  The first attempt was right,
except it didn't work until I got the matroxset program
that must be used to decide which framebuffer is displayed
to which connector.  (The default is to see the first
framebuffer on both!)  I haven't made startup scripts
yet, so I start the second Xserver from an xterm.

I did nothing special with the sound setup.  Both
screens are positioned between a shared pair of
speakers.  The alsa drivers for my trident 4D NX supports
several simultaneous users and does its own mixing.

The long term plan is to use two accelerated X servers.
Ideally by getting acceleration on both heads - but that
seems very unlikely.  The second best solution seems
to be a pci graphichs card.  I am waiting, hoping to find
some pci card that supports DRI in such a way that
I can use DRI on both screens simultaneously.

The questions:
2.6.0-test6-ruby seems to hate framebuffer console, hanging the
machine before the first boot printk if I enable it.
Is framebuffer console possible somehow?
I currently need framebuffer in order to run X.  And a
console is nice to have in case I get X trouble.
I currently only have screen garbage until X starts.
Having some framebuffer consoles on the other screen
would be nice too.  fbcon also has the advantage
that it can be set up with the same video timing as X,
allowing console switching without waiting for the
screen to sync to another frequency.  

I use the kernel parameter "dumbcon=1" to tell ruby that I have an extra
console.  Is this documented anywhere?

Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

Hi,

> Yesterday I got two xservers running with 2.6.0-test6-ruby
> Both screens are connected to the matrox G550.
> The one running on the main console uses the normal
> accelerated X server, and a keyboard connected to the mouse port.
> The X config for this display isn't modified, it is
> the single-user setup I always had.
> 
> The other screen uses the unaccelerated framebuffer driver
> on /dev/fb/1, and a keyboard connected to the normal keyboard
> connector.  Both mice are serial.
> 
> Framebuffer X was easy to set up.  The first attempt was right,
> except it didn't work until I got the matroxset program
> that must be used to decide which framebuffer is displayed
> to which connector.  (The default is to see the first
> framebuffer on both!)  I haven't made startup scripts
> yet, so I start the second Xserver from an xterm.

How did you got it running?
i'm tring with no success quite a long time.

which version of XFree ?
which matrox drivers?

if i try to start accelerated X,
X doesn't want to start and is eating ~100% of the cpu
X on fbdev seems to start, but no output to Monitor 

> I did nothing special with the sound setup.  Both
> screens are positioned between a shared pair of
> speakers.  The alsa drivers for my trident 4D NX supports
> several simultaneous users and does its own mixing.
> 
> The long term plan is to use two accelerated X servers.
> Ideally by getting acceleration on both heads - but that
> seems very unlikely.  The second best solution seems
> to be a pci graphichs card.  I am waiting, hoping to find
> some pci card that supports DRI in such a way that
> I can use DRI on both screens simultaneously.

two things here:
1.) DRI doesn't support multiple cards
(and it doesn't seems this to be on the TODO list
 of the DRI developers)
only choice is to use Nvidia cards with the closed source driver

2.) even if DRI supported multiple cards
 all heads have to be addressable by the driver.
with Gxxx DH this doable only with fbdev driver
which doesn't support DRI, XV, .....
(G200/450 MMS series could work as they have
different BusID's, but are untested -- too expensive)


> 
> The questions:
> 2.6.0-test6-ruby seems to hate framebuffer console, hanging the
> machine before the first boot printk if I enable it.
> Is framebuffer console possible somehow?
> I currently need framebuffer in order to run X.  And a
> console is nice to have in case I get X trouble.
> I currently only have screen garbage until X starts.
> Having some framebuffer consoles on the other screen
> would be nice too.  fbcon also has the advantage
> that it can be set up with the same video timing as X,
> allowing console switching without waiting for the
> screen to sync to another frequency.  

probably it's the matroxfb driver,
Petr (the maintainer) still doesn't like the new FB API,
and is still working on a more complete driver using different fb API,
the in kernel driver is not that developed and with limited functionality,
try searching lkml on smth like "matroxfb 2.5" 

> I use the kernel parameter "dumbcon=1" to tell ruby that I have an extra
> console.  Is this documented anywhere?

1.) http://startx.times.lv
2.) the linuxconsole ml
3.) my howto 
in progres (merging ruby-2.6
stuff)
http://varna.demon.co.uk/~svetlio/ruby-contrib/how-to/XFree-Local-multi-user-HOWTO/index.html
stable (no mention of ruby-2.6) is on any tldp.org mirror

best,

svetljo

-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Svetoslav Slavtchev wrote:
> Hi,
> 
> 
>>Yesterday I got two xservers running with 2.6.0-test6-ruby
>>Both screens are connected to the matrox G550.
>>The one running on the main console uses the normal
>>accelerated X server, and a keyboard connected to the mouse port.
>>The X config for this display isn't modified, it is
>>the single-user setup I always had.
>>
>>The other screen uses the unaccelerated framebuffer driver
>>on /dev/fb/1, and a keyboard connected to the normal keyboard
>>connector.  Both mice are serial.
>>
>>Framebuffer X was easy to set up.  The first attempt was right,
>>except it didn't work until I got the matroxset program
>>that must be used to decide which framebuffer is displayed
>>to which connector.  (The default is to see the first
>>framebuffer on both!)  I haven't made startup scripts
>>yet, so I start the second Xserver from an xterm.
> 
> 
> How did you got it running?
I used this machine single-headed for a long time.
All I did was to create a second config file (xf86config-4) 
for fbdev, and use matroxset-0.3 from Petrs site to
set up which framebuffer connects to what output.

I can set up a guest account on the machine, if you
want to look more at the setup.  

> i'm tring with no success quite a long time.

> 
> which version of XFree ?
debian packaged xserver-xfree86-dri-trunk from http://people.debian.org/~daenzer/dri-trunk-sid/
because I want accelerated opengl on at least one screen.
I got xdm working too.
From my /etc/X11/xdm/Xservers:

:0 local /usr/X11R6/bin/X :0 vt7  -deferglyphs 16 -nolisten tcp
:1 local /usr/X11R6/bin/X :1 vt17 -deferglyphs 16 -nolisten tcp -xf86config xf86config-4

The first server use the standard XF86Config-4
This runs the servers, but I had to modify /etc/X11/xdm/Xsetup
too to actually make the second framebuffer visible.  Whenever the
accelerated server starts, it configures the G550 to show the same
screen on both outputs.  Fortunately, xdm then runs the Xsetup script,
so that's where I put the following lines:
/usr/local/bin/matroxset -f /dev/fb/1 -m 0
/usr/local/bin/matroxset -f /dev/fb/0 -m 2
/usr/local/bin/matroxset -f /dev/fb/1 -m 1
1. line: disconnect /dev/fb/1 from whatever output it uses (in case it was using&blocking  output 2)
2. line: connect /dev/fb/0 to output 2, which is the vga connector
3. line: connect /dev/fb/1 to output 1, which is the DVI connector with a VGA converter.
The output numbers are really bits, so connecting a framebuffer to output 3 shows
it on both screens.  There's also a third output, bitnumber 4, but there is no such plug.
Perhaps the video out?  "matroxset" seems undocumented.

This setup means both servers get a screen.  A minor inconvenience is the way
both screens blink whenever one of the xdm's are reset. 

> which matrox drivers?
I believe I'm using the mga_drv.o (and fbdev_drv.o) that comes with xserver-xfree86-dri-trunk.
From /var/log/XFree86.0.log:
(II) LoadModule: "mga"
(II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/mga_drv.o
(II) Module mga: vendor="The XFree86 Project"
        compiled for 4.3.99.12, module version = 1.1.1
        Module class: XFree86 Video Driver
        ABI class: XFree86 Video Driver, version 0.7

From /var/log/XFree86.1.log:
(II) LoadModule: "fbdev"
(II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/fbdev_drv.o
(II) Module fbdev: vendor="The XFree86 Project"
        compiled for 4.3.0, module version = 0.1.0
        ABI class: XFree86 Video Driver, version 0.6

From /etc/X11/XF86Config-4  (accelerated)
Section "Device"
        Identifier      "Matrox G550"
        Driver          "mga"
        BusID           "PCI:01:00:0"
        Option          "UseFBDev"              "true"

        Option          "HWcursor" "off"   #avoid cursor corruption
        Option          "AGPMode" "2"      #Faster 3D...
        Option          "DRIReinit"        #Possibly several servers with DRI
EndSection

The last option because someone on the net got DRI working on two
matrox cards, (PCI+AGP) but only one at a time, unfortunately.

From /etc/X11/xf86config-4 (fbdev)
Section "Device"
        Identifier      "Matrox G550"
        Driver          "fbdev"
#       BusID           "PCI:01:00:1"
        Option          "UseFBDev"              "true"
        Option          "fbdev"                 "/dev/fb/1"

        Option          "HWcursor" "off"   #avoid cursor corruption        
EndSection

Using BusID here doesn't work.  The two files are otherwise
identical except that they specify different mice.

> if i try to start accelerated X,
> X doesn't want to start and is eating ~100% of the cpu
Ouch.  I haven't seen that happen.

> X on fbdev seems to start, but no output to Monitor 
Make sure you use matroxset so both framebuffers are visible.
You can do this without running X, cat a few megs of garbage
to both framebuffers first so they won't be all black.
Then experiment with matroxset.
> 
> 
>>I did nothing special with the sound setup.  Both
>>screens are positioned between a shared pair of
>>speakers.  The alsa drivers for my trident 4D NX supports
>>several simultaneous users and does its own mixing.
>>
>>The long term plan is to use two accelerated X servers.
>>Ideally by getting acceleration on both heads - but that
>>seems very unlikely.  The second best solution seems
>>to be a pci graphichs card.  I am waiting, hoping to find
>>some pci card that supports DRI in such a way that
>>I can use DRI on both screens simultaneously.
> 
> 
> two things here:
> 1.) DRI doesn't support multiple cards
> (and it doesn't seems this to be on the TODO list
>  of the DRI developers)
Too bad.  Theoretically this should be _trivial_ for
dual-seat two-card setups, because the two xservers are
fully independent and so are the two graphichs cards.
I wonder what they are running up against.

Of course the much more common case of one seat, several
cards aren't that easy - but I'm not trying for that.
> only choice is to use Nvidia cards with the closed source driver
 
> 2.) even if DRI supported multiple cards
>  all heads have to be addressable by the driver.
> with Gxxx DH this doable only with fbdev driver
> which doesn't support DRI, XV, .....
Which is why I'll buy a card if I believe it'll work.  Accelerated
2D on both screens is nice too, of course.
> (G200/450 MMS series could work as they have
> different BusID's, but are untested -- too expensive)
> 
Are these effectively two cards in a single slot then?
 
>>The questions:
>>2.6.0-test6-ruby seems to hate framebuffer console, hanging the
>>machine before the first boot printk if I enable it.
>>Is framebuffer console possible somehow?
>>I currently need framebuffer in order to run X.  And a
>>console is nice to have in case I get X trouble.
>>I currently only have screen garbage until X starts.
>>Having some framebuffer consoles on the other screen
>>would be nice too.  fbcon also has the advantage
>>that it can be set up with the same video timing as X,
>>allowing console switching without waiting for the
>>screen to sync to another frequency.  
> 
> 
> probably it's the matroxfb driver,

Well, 2.6.0-test6 (without ruby) works fine with
framebuffer console.  It dies only if I use ruby.

> Petr (the maintainer) still doesn't like the new FB API,
> and is still working on a more complete driver using different fb API,
> the in kernel driver is not that developed and with limited functionality,
> try searching lkml on smth like "matroxfb 2.5" 
> 
> 
>>I use the kernel parameter "dumbcon=1" to tell ruby that I have an extra
>>console.  Is this documented anywhere?
> 
> 
> 1.) http://startx.times.lv
> 2.) the linuxconsole ml
> 3.) my howto 

Thanks!


Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

> Svetoslav Slavtchev wrote:
> > Hi,
> > 
> > 
> >>Yesterday I got two xservers running with 2.6.0-test6-ruby
> >>Both screens are connected to the matrox G550.
> >>The one running on the main console uses the normal
> >>accelerated X server, and a keyboard connected to the mouse port.
> >>The X config for this display isn't modified, it is
> >>the single-user setup I always had.
> >>
> >>The other screen uses the unaccelerated framebuffer driver
> >>on /dev/fb/1, and a keyboard connected to the normal keyboard
> >>connector.  Both mice are serial.
> >>
> >>Framebuffer X was easy to set up.  The first attempt was right,
> >>except it didn't work until I got the matroxset program
> >>that must be used to decide which framebuffer is displayed
> >>to which connector.  (The default is to see the first
> >>framebuffer on both!)  I haven't made startup scripts
> >>yet, so I start the second Xserver from an xterm.
> > 
> > 
> > How did you got it running?
> I used this machine single-headed for a long time.
> All I did was to create a second config file (xf86config-4) 
> for fbdev, and use matroxset-0.3 from Petrs site to
> set up which framebuffer connects to what output.
> 
> I can set up a guest account on the machine, if you
> want to look more at the setup.  

it would be nice :-)
i'll try to get it running in the weekend 

> > i'm tring with no success quite a long time.
> 
> > 
> > which version of XFree ?
> debian packaged xserver-xfree86-dri-trunk from
> http://people.debian.org/~daenzer/dri-trunk-sid/
> because I want accelerated opengl on at least one screen.

do they already include the XFree-PrefBusID patch ?
(would be really nice addition in the howto)
or do you use the video hack ?

> I got xdm working too.
> >From my /etc/X11/xdm/Xservers:
> 
> :0 local /usr/X11R6/bin/X :0 vt7  -deferglyphs 16 -nolisten tcp
> :1 local /usr/X11R6/bin/X :1 vt17 -deferglyphs 16 -nolisten tcp
> -xf86config xf86config-4
> 
> The first server use the standard XF86Config-4
> This runs the servers, but I had to modify /etc/X11/xdm/Xsetup
> too to actually make the second framebuffer visible.  Whenever the
> accelerated server starts, it configures the G550 to show the same
> screen on both outputs.  Fortunately, xdm then runs the Xsetup script,
> so that's where I put the following lines:
> /usr/local/bin/matroxset -f /dev/fb/1 -m 0
> /usr/local/bin/matroxset -f /dev/fb/0 -m 2
> /usr/local/bin/matroxset -f /dev/fb/1 -m 1
> 1. line: disconnect /dev/fb/1 from whatever output it uses (in case it was
> using&blocking  output 2)
> 2. line: connect /dev/fb/0 to output 2, which is the vga connector
> 3. line: connect /dev/fb/1 to output 1, which is the DVI connector with a
> VGA converter.
> The output numbers are really bits, so connecting a framebuffer to output
> 3 shows
> it on both screens.  There's also a third output, bitnumber 4, but there
> is no such plug.
> Perhaps the video out?  "matroxset" seems undocumented.
> 
> This setup means both servers get a screen.  A minor inconvenience is the
> way
> both screens blink whenever one of the xdm's are reset. 

i've beiing using it for a long time 
mga accelled + Nvidia TNT2 Nvdia GL libs,
but when i tried mga acceled + fb/ 2xfb
i can not get it running :(
probably i didn't played enough with matroxset
(
i remmember trying to get Miguel solutions before a year may be,
and X did started, but i loosed my mices ?
)

> > which matrox drivers?
> I believe I'm using the mga_drv.o (and fbdev_drv.o) that comes with
> xserver-xfree86-dri-trunk.
> >From /var/log/XFree86.0.log:
> (II) LoadModule: "mga"
> (II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/mga_drv.o
> (II) Module mga: vendor="The XFree86 Project"
>         compiled for 4.3.99.12, module version = 1.1.1
>         Module class: XFree86 Video Driver
>         ABI class: XFree86 Video Driver, version 0.7
> 
> >From /var/log/XFree86.1.log:
> (II) LoadModule: "fbdev"
> (II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/fbdev_drv.o
> (II) Module fbdev: vendor="The XFree86 Project"
>         compiled for 4.3.0, module version = 0.1.0
>         ABI class: XFree86 Video Driver, version 0.6
> 
> >From /etc/X11/XF86Config-4  (accelerated)
> Section "Device"
>         Identifier      "Matrox G550"
>         Driver          "mga"
>         BusID           "PCI:01:00:0"

>         Option          "UseFBDev"              "true"

using the matrox driver
with this turned off i got X using 100% cpu and nothing on the screen
with this turned on not sure, but the same or imediate lock up

> 
>         Option          "HWcursor" "off"   #avoid cursor corruption
>         Option          "AGPMode" "2"      #Faster 3D...

>         Option          "DRIReinit"        #Possibly several servers with
> DRI

never tried that

> EndSection
> 
> The last option because someone on the net got DRI working on two
> matrox cards, (PCI+AGP) but only one at a time, unfortunately.
> 
> >From /etc/X11/xf86config-4 (fbdev)
> Section "Device"
>         Identifier      "Matrox G550"
>         Driver          "fbdev"
> #       BusID           "PCI:01:00:1"
>         Option          "UseFBDev"              "true"
>         Option          "fbdev"                 "/dev/fb/1"
> 
>         Option          "HWcursor" "off"   #avoid cursor corruption       
> 
> EndSection

same here

> 
> Using BusID here doesn't work.  The two files are otherwise
> identical except that they specify different mice.
> 
> > if i try to start accelerated X,
> > X doesn't want to start and is eating ~100% of the cpu
> Ouch.  I haven't seen that happen.
> 
> > X on fbdev seems to start, but no output to Monitor 
> Make sure you use matroxset so both framebuffers are visible.
> You can do this without running X, cat a few megs of garbage
> to both framebuffers first so they won't be all black.
> Then experiment with matroxset.

will experiment with this :-)

> > 
> > 
> >>I did nothing special with the sound setup.  Both
> >>screens are positioned between a shared pair of
> >>speakers.  The alsa drivers for my trident 4D NX supports
> >>several simultaneous users and does its own mixing.
> >>
> >>The long term plan is to use two accelerated X servers.
> >>Ideally by getting acceleration on both heads - but that
> >>seems very unlikely.  The second best solution seems
> >>to be a pci graphichs card.  I am waiting, hoping to find
> >>some pci card that supports DRI in such a way that
> >>I can use DRI on both screens simultaneously.
> > 
> > 
> > two things here:
> > 1.) DRI doesn't support multiple cards
> > (and it doesn't seems this to be on the TODO list
> >  of the DRI developers)
> Too bad.  Theoretically this should be _trivial_ for
> dual-seat two-card setups, because the two xservers are
> fully independent and so are the two graphichs cards.
> I wonder what they are running up against.
> 
> Of course the much more common case of one seat, several
> cards aren't that easy - but I'm not trying for that.

Hi Hi

Theoretically there are about 15-20 users(known to me) of such system,
and i doubt that someone complained about the problem.

there was a guy who wanted to get it running for some kind of commersial
system/ project,
i think he didn't needed multiple users, not sure about Xinerama, but
on the dri-devel list he got advised to ask for commersial support
from one of the developers's X releated company

IIRC the dri developer mentioned that no one thought of developing
such thing, and generally the core itself shouldn't be changed much,
but the X drivers needed major chages, and may be he could make it work with
certain driver
against a paiment


> > only choice is to use Nvidia cards with the closed source driver
>  
> > 2.) even if DRI supported multiple cards
> >  all heads have to be addressable by the driver.
> > with Gxxx DH this doable only with fbdev driver
> > which doesn't support DRI, XV, .....
> Which is why I'll buy a card if I believe it'll work.  Accelerated
> 2D on both screens is nice too, of course.
> > (G200/450 MMS series could work as they have
> > different BusID's, but are untested -- too expensive)
> > 
> Are these effectively two cards in a single slot then?

i assume almost yes,
and i always wanted to get one of the 4 headed one's 
they do have 2/4 G200/G450 chips with separate BusID's on them
  
> >>The questions:
> >>2.6.0-test6-ruby seems to hate framebuffer console, hanging the
> >>machine before the first boot printk if I enable it.
> >>Is framebuffer console possible somehow?
> >>I currently need framebuffer in order to run X.  And a
> >>console is nice to have in case I get X trouble.
> >>I currently only have screen garbage until X starts.
> >>Having some framebuffer consoles on the other screen
> >>would be nice too.  fbcon also has the advantage
> >>that it can be set up with the same video timing as X,
> >>allowing console switching without waiting for the
> >>screen to sync to another frequency.  
> > 
> > 
> > probably it's the matroxfb driver,
> 
> Well, 2.6.0-test6 (without ruby) works fine with
> framebuffer console.  It dies only if I use ruby.

which ruby-2.6 is this ?
the one with real multi-user console, or the earlier
with only multiuser X ?

if it's the first one may be fbcon might get confused
somehow by the 2 fb's on a single card, which might be
not properly initialized without matroxset that early
or might be Aivils did missed smth(but i doubt it)

can you try the earlier patch?


best,

svetljo

-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Svetoslav Slavtchev wrote:

> 
> do they already include the XFree-PrefBusID patch ?
> (would be really nice addition in the howto)
> or do you use the video hack ?
> 
I use neither.  I believe these things are for multi-card
issues. (X server stupidly disabling other cards)
But I have only one card at this time, there are no
other cards there to disable via pci programming.

I use a BusID (not prefBusID) for the accelerated
server - it  is probably not necessary.  Just something I
added because I read about it, preparing for a future with
two cards.  Specifying the same BusID for fbdev goes wrong,
the server notices that it can't grab that card because
it is in use, and quits. 
Fortunately, the fbdev driver doesn't need to grab a card
at all, it merely grabs a framebuffer device and uses its
memory.  

>>>From /var/log/XFree86.1.log:
>>(II) LoadModule: "fbdev"
>>(II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/fbdev_drv.o
>>(II) Module fbdev: vendor="The XFree86 Project"
>>        compiled for 4.3.0, module version = 0.1.0
>>        ABI class: XFree86 Video Driver, version 0.6
>>
>>>From /etc/X11/XF86Config-4  (accelerated)
>>Section "Device"
>>        Identifier      "Matrox G550"
>>        Driver          "mga"
>>        BusID           "PCI:01:00:0"
> 
> 
>>        Option          "UseFBDev"              "true"
> 
> 
> using the matrox driver
> with this turned off i got X using 100% cpu and nothing on the screen
> with this turned on not sure, but the same or imediate lock up

Come to think of it, "UseFBDev" doesn't say anything
about _which_ fbdev to use.  So I guess it uses the first one.
That could cause all sorts of strange bugs if the first one
isn't the right one.

> 
> 
>>        Option          "HWcursor" "off"   #avoid cursor corruption
>>        Option          "AGPMode" "2"      #Faster 3D...
> 
> 
>>        Option          "DRIReinit"        #Possibly several servers with
>>DRI
> 
> never tried that
It shouldn't make a difference as long as DRI only is used on one display.
> 
> 
>>EndSection
>>
>>The last option because someone on the net got DRI working on two
>>matrox cards, (PCI+AGP) but only one at a time, unfortunately.

>>>two things here:
>>>1.) DRI doesn't support multiple cards
>>>(and it doesn't seems this to be on the TODO list
>>> of the DRI developers)
>>
>>Too bad.  Theoretically this should be _trivial_ for
>>dual-seat two-card setups, because the two xservers are
>>fully independent and so are the two graphichs cards.
>>I wonder what they are running up against.
>>
>>Of course the much more common case of one seat, several
>>cards aren't that easy - but I'm not trying for that.
> 
> 
> Hi Hi
> 
> Theoretically there are about 15-20 users(known to me) of such system,
> and i doubt that someone complained about the problem.
> 
Time to complain then. :-)
They'll probably just tell me to do it myself, perhaps I
can get enough information to try.

> there was a guy who wanted to get it running for some kind of commersial
> system/ project,
> i think he didn't needed multiple users, not sure about Xinerama, but
> on the dri-devel list he got advised to ask for commersial support
> from one of the developers's X releated company
> 
> IIRC the dri developer mentioned that no one thought of developing
> such thing, and generally the core itself shouldn't be changed much,
> but the X drivers needed major chages, and may be he could make it work with
> certain driver
> against a paiment

Hm.  X already works well in a dual seat setup, wonder what those major
changes are.  DRI already uses a device driver, using another driver
shouldn't be a problem.  I am tempted to get a pci card and use a
chroot so both X servers can get their own "/dev/dri/card0" which 
happens to be different devices.   The problem is time, and
which pci card to try.  I happen to like matrox for their good
picture quality (sharp and no flickering), and history of good dual
screen support.

>>>only choice is to use Nvidia cards with the closed source driver
I guess I could live with a closed-source X driver, if it
works.  But not a closed kernel driver, I run experimental
kernels and the module interfaces breaks all the time.
>>
>>Which is why I'll buy a card if I believe it'll work.  Accelerated
>>2D on both screens is nice too, of course.
>>
>>>(G200/450 MMS series could work as they have
>>>different BusID's, but are untested -- too expensive)
>>>
>>
>>Are these effectively two cards in a single slot then?
> 
> 
> i assume almost yes,

Any information about the "almost" part?  Are they sharing
something (such as 3D hw) that might prevent fully independent
use of the two heads?
The one thing I don't mind so much is shared clock/ramdac, I have two
identical screens so the ideal frequency and resolution is
the same for both.

> and i always wanted to get one of the 4 headed one's 
> they do have 2/4 G200/G450 chips with separate BusID's on them
>   
Wow - the strange things that exists...  
I have little use for more than 2 seats, but I see how a 4-way
solution could be great when there is that many people.
> 
>>
>>Well, 2.6.0-test6 (without ruby) works fine with
>>framebuffer console.  It dies only if I use ruby.
> 
> 
> which ruby-2.6 is this ?
It is ruby-260t6-20030930.diff.bz2 on top of plain 2.4.0-test6
A search for ruby-260t6 turns up:
ruby-260t6-20031015.diff.bz2      (This is the newer one?)
ruby-260t9-CVS-20031028.tar.bz2   (test9 ruby? exciting!)

> the one with real multi-user console, or the earlier
> with only multiuser X ?
> 
I don't know the differences between them, I don't usually
find much version-specific information.  I don't
think the 20031015 was available when I downloaded the patch.

> if it's the first one may be fbcon might get confused
> somehow by the 2 fb's on a single card, which might be
> not properly initialized without matroxset that early
> or might be Aivils did missed smth(but i doubt it)

I don't think matroxset makes a difference for _using_
the framebuffers, it only decides which one is visible
on what connector.  Framebuffer X (and accelerated)
works just fine "blind" if I mess up a matroxset command.

> can you try the earlier patch?
I suppose you mean the later one as I
seem to be using the earlier one?

That 260t9-CVS version seems even more interesting, unless
it is "work in progress" with known problems.  I'll
probably try that one first.

Helge Hafting

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Svetoslav S. <sv...@gm...>]

Hi,

i think i almost got it :-)

almost, cause my main monitor died in the weekend 
and i used a switch, and the accelerated head gives
pretty strange output 

Helge please correct me if i'm wrong somewhere,
or if smth differs from your setup
 
> > How did you got it running?
> I used this machine single-headed for a long time.
> All I did was to create a second config file (xf86config-4) 
> for fbdev, and use matroxset-0.3 from Petrs site to
> set up which framebuffer connects to what output.
> 
> I can set up a guest account on the machine, if you
> want to look more at the setup.  
> 
> > i'm tring with no success quite a long time.
> 
> > 
> > which version of XFree ?
> debian packaged xserver-xfree86-dri-trunk from
> http://people.debian.org/~daenzer/dri-trunk-sid/
> because I want accelerated opengl on at least one screen.

that was my first problem, i used the matrox drivers
which obviously doesn't play nice with linux-2.6
i got all kind of agp errors and tring to activate "UseFBDev" 
caused lock ups
switching to the plain mga drivers fixed the problem,
but also the distored screen :( i'll have to play more with it


> I got xdm working too.
> >From my /etc/X11/xdm/Xservers:
> 
> :0 local /usr/X11R6/bin/X :0 vt7  -deferglyphs 16 -nolisten tcp
> :1 local /usr/X11R6/bin/X :1 vt17 -deferglyphs 16 -nolisten tcp
> -xf86config xf86config-4
> 
> The first server use the standard XF86Config-4
> This runs the servers, but I had to modify /etc/X11/xdm/Xsetup
> too to actually make the second framebuffer visible.  Whenever the
> accelerated server starts, it configures the G550 to show the same
> screen on both outputs.  Fortunately, xdm then runs the Xsetup script,
> so that's where I put the following lines:
> /usr/local/bin/matroxset -f /dev/fb/1 -m 0
> /usr/local/bin/matroxset -f /dev/fb/0 -m 2
> /usr/local/bin/matroxset -f /dev/fb/1 -m 1

that's basicly the script swapped from the matroset tarball, correct ?

> 1. line: disconnect /dev/fb/1 from whatever output it uses (in case it was
> using&blocking  output 2)
> 2. line: connect /dev/fb/0 to output 2, which is the vga connector
> 3. line: connect /dev/fb/1 to output 1, which is the DVI connector with a
> VGA converter.
> The output numbers are really bits, so connecting a framebuffer to output
> 3 shows
> it on both screens.  There's also a third output, bitnumber 4, but there
> is no such plug.
> Perhaps the video out?  "matroxset" seems undocumented.
> 
> This setup means both servers get a screen.  A minor inconvenience is the
> way
> both screens blink whenever one of the xdm's are reset. 
> 
> > which matrox drivers?
> I believe I'm using the mga_drv.o (and fbdev_drv.o) that comes with
> xserver-xfree86-dri-trunk.
> >From /var/log/XFree86.0.log:
> (II) LoadModule: "mga"
> (II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/mga_drv.o
> (II) Module mga: vendor="The XFree86 Project"
>         compiled for 4.3.99.12, module version = 1.1.1
>         Module class: XFree86 Video Driver
>         ABI class: XFree86 Video Driver, version 0.7
> 
> >From /var/log/XFree86.1.log:
> (II) LoadModule: "fbdev"
> (II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/fbdev_drv.o
> (II) Module fbdev: vendor="The XFree86 Project"
>         compiled for 4.3.0, module version = 0.1.0
>         ABI class: XFree86 Video Driver, version 0.6
> 
> >From /etc/X11/XF86Config-4  (accelerated)
> Section "Device"
>         Identifier      "Matrox G550"
>         Driver          "mga"
>         BusID           "PCI:01:00:0"
>         Option          "UseFBDev"              "true"

without "UseFBDev here i get the "mga" screen of both fb's
and if i run matroxset after "mga" is started ," mga" goes blank
or distored, if i start the "fb" i get lock up

with "UseFBDev activated everything works,
running matroxset doesn't affect "mga"
(there might be short blanking but i couldn't see as
i used the switch with only one monitor )
starting "fb" works OK  

>         Option          "HWcursor" "off"   #avoid cursor corruption
>         Option          "AGPMode" "2"      #Faster 3D...
>         Option          "DRIReinit"        #Possibly several servers with
> DRI
> EndSection
> 
> The last option because someone on the net got DRI working on two
> matrox cards, (PCI+AGP) but only one at a time, unfortunately.
> 
> >From /etc/X11/xf86config-4 (fbdev)
> Section "Device"
>         Identifier      "Matrox G550"
>         Driver          "fbdev"
> #       BusID           "PCI:01:00:1"
>         Option          "UseFBDev"              "true"
>         Option          "fbdev"                 "/dev/fb/1"
> 
>         Option          "HWcursor" "off"   #avoid cursor corruption       
> 
> EndSection
> 
> Using BusID here doesn't work.  The two files are otherwise
> identical except that they specify different mice.


So, in a recap :-)
(for the howto )

1.)
don't use matrox's closed source driver (mga_hall_drv)
it doesn't support linux-2.6 && the new FB API
do use the original XFree driver

2.)
do activate 
Option "useFBDev" "true" 
for the accelerated head 
(? does this option affect "fbdev" unaccelerated head ? )

3.)
get matroxset, build it and install it
install also the swapped script from the tarball

4.)
convince xdm/kdm/gdm to run swapped  at every X server
start/restart
for xdm/kdm Xsetup
?for gdm ? wrapper? is there smth like Xsetup used by gdm ?
 

best,

svetljo

-- 
NEU FÜR ALLE - GMX MediaCenter - für Fotos, Musik, Dateien...
Fotoalbum, File Sharing, MMS, Multimedia-Gruß, GMX FotoService

Jetzt kostenlos anmelden unter http://www.gmx.net

+++ GMX - die erste Adresse für Mail, Message, More! +++

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Svetoslav Slavtchev wrote:
> Hi,
> 
> i think i almost got it :-)
> 
> almost, cause my main monitor died in the weekend 
> and i used a switch, and the accelerated head gives
> pretty strange output 
> 
> Helge please correct me if i'm wrong somewhere,
> or if smth differs from your setup
>  
Some tips:
The framebuffer driver is supposed to start the second head
in 640x480, but the accelerated driver seems to set it to
1280x1024 while setting the first head to that resolution.
Quitting X seems to reset this as "old" settings are restored.
Try using the same resolution on both heads.
Also, the framebuffer drivers claim to support only 16 or 32
bit color on the second head.  So don't bother trying
8 or 24 bit.
 
OpenGL seems to work only with 16bit, so that's what I use.
16-bit color and the same resolution on both screens.
I happen to have two identical monitors, others should
set up the accelerated driver with timings that both
monitors can handle.  I believe the second head uses 
the same timing as the first.  Other setups may be
possible - this is the "easy" one.

>>>How did you got it running?
>>
>>I used this machine single-headed for a long time.
>>All I did was to create a second config file (xf86config-4) 
>>for fbdev, and use matroxset-0.3 from Petrs site to
>>set up which framebuffer connects to what output.
>>
>>I can set up a guest account on the machine, if you
>>want to look more at the setup.  
>>
>>
>>>i'm tring with no success quite a long time.
>>
>>>which version of XFree ?
>>
>>debian packaged xserver-xfree86-dri-trunk from
>>http://people.debian.org/~daenzer/dri-trunk-sid/
>>because I want accelerated opengl on at least one screen.
>  
> that was my first problem, i used the matrox drivers
> which obviously doesn't play nice with linux-2.6
> i got all kind of agp errors and tring to activate "UseFBDev" 
> caused lock ups
> switching to the plain mga drivers fixed the problem,
> but also the distored screen :( i'll have to play more with it
> 
The matrox drivers (beta 3) were fine in single-screen setup,
I haven't tried them with two screens.  I gave up on them
because the dri-trunk driver was better at doing opengl.
If agp cause trouble, consider forcing agp 1x until
everything else works.  And turn off any kind of
overclocking.  (Experiment with such things after
your setup works...)
I use agp 2x, as my old motherboard (bx) isn't supposed
to support more than that.  I can specify agp4x, but then
I have to change the bios from agp=1/1 cpuclk to
agp=2/3 cpucl, where the cpu clock is 100MHz.
This don't buy anything though, so I use the supported agp 2x
and use agp=1/1 cpuclk.

Just write down whatever bios settings you have, and test
if some really slow settings work better.  
> 
>>I got xdm working too.
>>>From my /etc/X11/xdm/Xservers:
>>
>>:0 local /usr/X11R6/bin/X :0 vt7  -deferglyphs 16 -nolisten tcp
>>:1 local /usr/X11R6/bin/X :1 vt17 -deferglyphs 16 -nolisten tcp
>>-xf86config xf86config-4
>>
>>The first server use the standard XF86Config-4
>>This runs the servers, but I had to modify /etc/X11/xdm/Xsetup
>>too to actually make the second framebuffer visible.  Whenever the
>>accelerated server starts, it configures the G550 to show the same
>>screen on both outputs.  Fortunately, xdm then runs the Xsetup script,
>>so that's where I put the following lines:
>>/usr/local/bin/matroxset -f /dev/fb/1 -m 0
>>/usr/local/bin/matroxset -f /dev/fb/0 -m 2
>>/usr/local/bin/matroxset -f /dev/fb/1 -m 1
> 
> 
> that's basicly the script swapped from the matroset tarball, correct ?
Yes, except that I was surprised that the first screen was the
second output.  It was easy enough to find out though.
The xfree driver for g550 really ought to have options for
what output port to use.  (first, second, both.. use the same numbers
as for matroxset.) I currently have to re-run those lines
whenever X changes resolution.  I.e. when my girlfriend starts
tuxracer on the accelerated screen.  Any resolution change
shows /dev/fb/0 on both screeens, but only the correct one
changes resolution.  Looks weird, and I have to
run matroxset blind to get my own screen back.
> 
>>1. line: disconnect /dev/fb/1 from whatever output it uses (in case it was
>>using&blocking  output 2)
>>2. line: connect /dev/fb/0 to output 2, which is the vga connector
>>3. line: connect /dev/fb/1 to output 1, which is the DVI connector with a
>>VGA converter.
>>The output numbers are really bits, so connecting a framebuffer to output
>>3 shows
>>it on both screens.  There's also a third output, bitnumber 4, but there
>>is no such plug.
>>Perhaps the video out?  "matroxset" seems undocumented.
>>
>>This setup means both servers get a screen.  A minor inconvenience is the
>>way
>>both screens blink whenever one of the xdm's are reset. 
>>
>>
>>>which matrox drivers?
>>
>>I believe I'm using the mga_drv.o (and fbdev_drv.o) that comes with
>>xserver-xfree86-dri-trunk.
>>>From /var/log/XFree86.0.log:
>>(II) LoadModule: "mga"
>>(II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/mga_drv.o
>>(II) Module mga: vendor="The XFree86 Project"
>>        compiled for 4.3.99.12, module version = 1.1.1
>>        Module class: XFree86 Video Driver
>>        ABI class: XFree86 Video Driver, version 0.7
>>
>>>From /var/log/XFree86.1.log:
>>(II) LoadModule: "fbdev"
>>(II) Loading /usr/X11R6/lib/modules-dri-trunk/drivers/fbdev_drv.o
>>(II) Module fbdev: vendor="The XFree86 Project"
>>        compiled for 4.3.0, module version = 0.1.0
>>        ABI class: XFree86 Video Driver, version 0.6
>>
>>>From /etc/X11/XF86Config-4  (accelerated)
>>Section "Device"
>>        Identifier      "Matrox G550"
>>        Driver          "mga"
>>        BusID           "PCI:01:00:0"
>>        Option          "UseFBDev"              "true"
> 
> 
> without "UseFBDev here i get the "mga" screen of both fb's
> and if i run matroxset after "mga" is started ," mga" goes blank
> or distored, if i start the "fb" i get lock up
> 
> with "UseFBDev activated everything works,
> running matroxset doesn't affect "mga"
> (there might be short blanking but i couldn't see as
> i used the switch with only one monitor )
> starting "fb" works OK  
> 
It is only the unaccelerated screen that blinks for me,
because that is the only one that is turned off for a moment.
The accelerated screen is on all the time with that script,
assuming it was on from the beginning, so no blinking there.

> 
>>        Option          "HWcursor" "off"   #avoid cursor corruption
>>        Option          "AGPMode" "2"      #Faster 3D...
>>        Option          "DRIReinit"        #Possibly several servers with
>>DRI
>>EndSection
>>
>>The last option because someone on the net got DRI working on two
>>matrox cards, (PCI+AGP) but only one at a time, unfortunately.
>>
>>>From /etc/X11/xf86config-4 (fbdev)
>>Section "Device"
>>        Identifier      "Matrox G550"
>>        Driver          "fbdev"
>>#       BusID           "PCI:01:00:1"
>>        Option          "UseFBDev"              "true"
>>        Option          "fbdev"                 "/dev/fb/1"
>>
>>        Option          "HWcursor" "off"   #avoid cursor corruption       
>>
>>EndSection
>>
>>Using BusID here doesn't work.  The two files are otherwise
>>identical except that they specify different mice.
> 
> 
> 
> So, in a recap :-)
> (for the howto )
> 
> 1.)
> don't use matrox's closed source driver (mga_hall_drv)
> it doesn't support linux-2.6 && the new FB API
> do use the original XFree driver
> 
> 2.)
> do activate 
> Option "useFBDev" "true" 
> for the accelerated head 
> (? does this option affect "fbdev" unaccelerated head ? )
> 
> 3.)
> get matroxset, build it and install it
> install also the swapped script from the tarball
> 
> 4.)
> convince xdm/kdm/gdm to run swapped  at every X server
> start/restart
> for xdm/kdm Xsetup
> ?for gdm ? wrapper? is there smth like Xsetup used by gdm ?
>  
This seems fine.

I don't know what to do with gdm/kdm, other than
read the docs and see if they have a similiar file that
runs as root after X server initialization before presenting
the login screen.  It has to be at that moment,
after the accelerated server initialization that
scrambles the output setup, and before the
login screen or you won't see it!

xdm seems to work fine for this purpose though.
I can restart the fbdev X server anythime without affecting
the accelerated one.  When running matroxset "blind" fails I
simply press ctrl+alt+backspace, killing the fbdev server
and then xdm restarts it and runs the script for me.
I haven't tried this with the accelerated server yet.  It will
at least blink the wrong screen as it briefly take over
both outputs.

Setting up some hotkey for the matroxset script is
highly advisable - I haven't bothered figuring it out
yet though.  This may depend on the window manager used.

The framebuffer driver works surprisingly well.  Browsing,
word processing and 2D games performs fine even with
this unaccelerated driver on 500MHz dual cpus.  It is still
fine with tuxracer running on the other screen using 100%
of one cpu. 

Helge Hafting

Re:2.6.0-test6 ruby success with matrox G550, and some questions

[Aivils S. <Aiv...@un...>]

>The questions:
>2.6.0-test6-ruby seems to hate framebuffer console, hanging the
>machine before the first boot printk if I enable it.
>Is framebuffer console possible somehow?

seems matroxfb low level device drivers is broken.

Under ruby patched only hardware independ abstract layer -
frambuffer console fbcon.c .
If matroxfb will run under vanilla 2.6.0-test6, then it
should do same under 2.6.0-test6-ruby

>I currently need framebuffer in order to run X.

This is only way for dual head video adapters.
To start accelerated xf86 on both heads indepent
input You should write Your own XFree86 5.0.0 .

>And a
>console is nice to have in case I get X trouble.
>I currently only have screen garbage until X starts.
>Having some framebuffer consoles on the other screen
>would be nice too.  fbcon also has the advantage
>that it can be set up with the same video timing as X,
>allowing console switching without waiting for the
>screen to sync to another frequency.

Current ruby support straight in kernel compiled fb device
drivers and fbcon. To test Your matrox figure fbcon into
kernel and reboot.
Without parameters /dev/fb/0 take over VGA and use
/dev/tty1 - /dev/tty16 device files.
/dev/fb/1 create extra console with /dev/tty17 - /dev/tty33
You schould add additional mingetty in /etc/inittab to get
login prompt on /dev/fb/1 .
For test I use Voodoo1 as secondary card and have got
success. Of course Voodoo1 is very special video card, but
suitable for tests.

Aivils

Re: 2.6.0-test6 ruby success with matrox G550, and some questions

[Helge H. <hel...@ai...>]

Aivils Stoss wrote:
>>The questions:
>>2.6.0-test6-ruby seems to hate framebuffer console, hanging the
>>machine before the first boot printk if I enable it.
>>Is framebuffer console possible somehow?
> 
> 
> seems matroxfb low level device drivers is broken.
Well, it works with plain 2.6.0-test6.
> 
> Under ruby patched only hardware independ abstract layer -
> frambuffer console fbcon.c .
> If matroxfb will run under vanilla 2.6.0-test6, then it
> should do same under 2.6.0-test6-ruby
> 
Unfortunately it fails with ruby. :-(
> 
>>I currently need framebuffer in order to run X.
> 
> 
> This is only way for dual head video adapters.
> To start accelerated xf86 on both heads indepent
> input You should write Your own XFree86 5.0.0 .
Seems an extra card is cheaper then. :-)
 
>>And a
>>console is nice to have in case I get X trouble.
>>I currently only have screen garbage until X starts.
>>Having some framebuffer consoles on the other screen
>>would be nice too.  fbcon also has the advantage
>>that it can be set up with the same video timing as X,
>>allowing console switching without waiting for the
>>screen to sync to another frequency.
> 
> 
> Current ruby support straight in kernel compiled fb device
> drivers and fbcon. To test Your matrox figure fbcon into
> kernel and reboot.
2.6.0-test6-ruby dies before the first printk if I do that, 
while 2.6.0-test6 works.  fbcon on matrox breaks somehow
with ruby.  I don't know what to do to debug this
further - adding printk's is probably not an option.

> Without parameters /dev/fb/0 take over VGA and use
> /dev/tty1 - /dev/tty16 device files.
> /dev/fb/1 create extra console with /dev/tty17 - /dev/tty33
> You schould add additional mingetty in /etc/inittab to get
> login prompt on /dev/fb/1 .
Thanks!  I'll do this if I ever get those
fbconsoles working.

Helge Hafting