 |
|
| on-board Millenium trouble |
|
|
By rudolfc - Posted on February 3, 2006 - 02:29:36 (#18990)
Current version when comment was posted: 0.26 |
|
|
Hi,
A screenshot would be interesting indeed. Could you mail that to me? And a logfile from the driver maybe?
BTW: Did you already try the Be driver instead to see if that works with your card?
Thanks, and sorry for the late reply :-/
Rudolf.
|
|
| Problems with old computer |
|
|
By edmundf - Posted on January 7, 2006 - 05:30:42 (#18788)
Current version when comment was posted: 0.26 |
|
|
The computer has an integrated video card.
And I can't find any names on chips.
So after I installed Windows I could read in the
settings it is an Matrox MGA millennium card.
If I manually select "fail-save video mode"
and resolution 800 600 16 everything is fine.
Of course I don't want to select that with every boot
so I tried Vesa mode, which doesn't work.
The problem seems to be that I don't know how to select
fail-save video mode in some script.
The 800 600 16 selection made by the vesa script does
work, but I still have to select "fail-save video mode"
by hand.
The Matrox driver gives color but there are vertical lines missing on the screen and things are not, or
very difficult to read.
If it is of any use I for someone ( like Rudolf :-))
I can make a screen capture or a photo from the screen.
|
|
| RE: G400 crashes |
|
|
By Eugenia - Posted on January 27, 2005 - 16:58:27 (#15600)
Current version when comment was posted: 0.20 |
|
|
Actually, I had installed ALT.Software's version of Matrox G400 driver. That might be the culprit. That driver from ALT.Software was never released to the public, but I was a beta tester and so I had it around. I guess that might be the problem, the two drivers colliding.
|
|
| G400 crashes |
|
|
By rudolfc - Posted on January 27, 2005 - 05:46:18 (#15586)
Current version when comment was posted: 0.20 |
|
|
Hi there Eugenia,
The driver should do that mode perfectly. If you were running Mark Watson's driver once however, please remove that first as otherwise my driver won't get loaded.
In the included readme HTML file with my driver there's an exact description on what to remove to get rid of that old driver.
Hope this helps.. ?
Best regards,
Rudolf.
|
|
| Driver does not work for me |
|
|
By Eugenia - Posted on January 26, 2005 - 16:15:04 (#15574)
Current version when comment was posted: 0.20 |
|
|
I am running BeOS 5.0.3 on my dual Celeron 2x533. I installed this driver and now BeOS crashes with Debugging Land right after it has loaded the desktop.
BeOS was set to run on 1600x1200x85 Hz on my Matrox Millenium G400-MAX AGP. Can your driver deliver this resolution/Hz? Why do I get crashes? Did I need to uninstall my previous driver first?
|
|
| requested! |
|
|
By rudolfc - Posted on October 5, 2004 - 15:14:16 (#14327)
Current version when comment was posted: 0.16 |
|
|
Hi,
Just for completeness I am letting you know I have inquired now about the status of the Parhelia series cards at Matrox devrel.
Once you complete the form and send it, you get a form onscreen which (only THEN) tells you that the form is only meant for professional developers. Freeware app developers and hobbyists so they say, may not even get a response...
Thanks for your inquiry in the techforum anyway, and thanks for letting me know!
I'll keep you posted if anything new happens. Keep your fingers crossed: this is not the first attempt that has been done..
Best regards,
Rudolf Cornelissen.
|
|
| matrox inc. support |
|
|
By an-dee - Posted on October 4, 2004 - 14:02:46 (#14316)
Current version when comment was posted: 0.16 |
|
|
Rudolf, check this: http://forum.matrox.com/mga/viewtopic.php?t=1689
According to Haig you just have to fill that form to get the neccessery docs to build P series drivers.
thanks
|
|
| Parhelia / other P-series cards |
|
|
By rudolfc - Posted on July 29, 2004 - 09:33:28 (#13650)
Current version when comment was posted: 0.16 |
|
|
No support for them yet, no docs whatsoever are available. For now your best hope would probably a VESA3.0 driver if someone would set that up.
This would still be unaccelerated though, just like a VESA2 'driver' as you are now using...
Sorry :-/
Best regards,
Rudolf.
|
|
| P serias (p650, p750, perhelia) |
|
|
By an-dee - Posted on July 29, 2004 - 09:03:10 (#13648)
Current version when comment was posted: 0.16 |
|
|
Hi
Is there any progress in supporting P serias cards?
I have a P650 with vesa mode only :(
An-Dee
|
|
| Now it works |
|
|
By biffuz - Posted on July 28, 2004 - 06:33:06 (#13632)
Current version when comment was posted: 0.16 |
|
|
Thank you Rudolf for pointing out that issue about the IRQ, now it works perfectly.
Keep up the good work
Gabriele
|
|
| Cursor and CPU load |
|
|
By rudolfc - Posted on July 27, 2004 - 04:23:27 (#13601)
Current version when comment was posted: 0.16 |
|
|
Hi again Jack!
The cursor takes more CPU time when it's on the bottom of the screen because I have to wait until retrace is done before updating. When it's higher, it can be done during the current drawing of the screen.
So actually you should see it just the other way around: the cursor takes less CPU if near the top of the screen, because I can speedup things by checking which line is getting drawn.. Normally you just have to wait until retrace is 'just' done. Though the right setup differs very much for card brands and types :-/
Hence: move to INT! to get a 'universal' solution in which I don't have to wait at all: I just let the cursor movement routines return, and assume the INT will update it's position at the proper time.
Bye!
Rudolf.
PS: I still have to test your BWindowScreen implementation for haiku. Sorry about not doing that yet, but you know how it is: always too less time to do everything you want :-/
Currently I am working on the Neomagic driver a bit, to add 2D acc for NM2200 and later cards: seems it's indeed starting to work! :-)
|
|
| IRQ Line |
|
|
By Jack Burton - Posted on July 27, 2004 - 02:35:44 (#13599)
Current version when comment was posted: 0.16 |
|
|
Thanks for the explanation.
Now that you told me, I remember BScreen::WaitForRetrace() didn't work with the be's driver.
About spawning a thread for every cursor movement... no, it's not a good practice, as spawning a thread isn't cheap enough for a thing as common as moving the cursor :) , and anyway, I think it would overly complicate things.
Better move the thing into the INT routine.
BTW, how come that currently it uses more cpu if the cursor is in the bottom part of the screen ?
And yeah, I think you shouldn't definitely worry about the INT support. Just put an hint into the readme saying it won't work without INT and you're done :=)
last but not least: thank for the great work :)
|
|
| IRQ-line dependancy |
|
|
By rudolfc - Posted on July 24, 2004 - 02:50:54 (#13547)
Current version when comment was posted: 0.16 |
|
|
Hi Jack!
Well, fortunately, as you already said, these days the IRQ is always assigned :-)
Anyway, the reason is simple: The driver has a function called WaitForRetrace(). This function is based on the IRQ's INT handler for retraces in the kerneldriver. This means Be's drivers don't have that which is a shame as it will result in some artifacts in video playback (for instance) not being possible to overcome.
It's possible to just shut down that feature, and then the drivers can do without that INT. For now.
I am planning to also move the cursor movement routines into the kerneldriver's INT routine partly, as that would lower CPU load a _lot_. Or maybe, I should spawn a new thread for every cursor move in the accelerant (is that even possible, or 'good practice'?) that then waits for the retrace to occur using the retrace semaphore..
I think I'll start the kerneldriver thingy (test) first: that sounds simpler to me, and easier to grasp for me as well :)
===
OK, anyway: I could still check for the INT being assigned, and drop both retrace and cursor support if not. But I don't think I am going to put time and effort in this stuff, as it very rarely comes up..
Hey: any thoughts about this stuff, while we are talking about it?
Bye!
Rudolf.
|
|
| Assign IRQ to PCI VGA |
|
|
By Jack Burton - Posted on July 23, 2004 - 06:12:23 (#13519)
Current version when comment was posted: 0.16 |
|
|
Rudolf, may I ask you why your drivers require this option to be set while the original drivers don't ?
I tought that option was only for PCI VGA, while AGP VGA always had an irq assigned.
I had a hard time making your nvidia driver work (long ago), till I tried to enable that option :))
|
|
| tip.. |
|
|
By rudolfc - Posted on July 23, 2004 - 05:13:44 (#13517)
Current version when comment was posted: 0.16 |
|
|
Hi,
If the driver doesn't load, and you fallback to the Be driver or B/W, VESA mode, chances are that your system BIOS does not assign an IRQ line to your graphicscard.
Please check it and set (if possible):
Assign IRQ to VGA: yes.
Hope this helps..
Rudolf.
|
|
| Doesn't work in my Dual P2 |
|
|
By biffuz - Posted on July 20, 2004 - 10:25:21 (#13456)
Current version when comment was posted: 0.16 |
|
|
Hi Rudolf, I just replaced my old PC's mobo with a dual P2 350, and your driver no longer works here. I don't know if this is a problem with the CPUs, the chipset (i440BX) or the mobo itself (Gigabyte GA-6BXD). The graphics card is a G200 8 Mb SGRAM (AGP, of course). The original driver works.
With 0.16, the system just tell me that my card isn't supported (???), in both BeOS 5.0.3 Pro and Zeta RC2; with 0.13beta2, BeOS hangs just when the last bootscreen's icon lights up, and Zeta goes into KDL.
Any tip?
|
|
| oops |
|
|
By Rene Gollent - Posted on July 9, 2004 - 14:59:03 (#13260)
Current version when comment was posted: 0.16 |
|
|
Hm, firefox double posted that for some reason, sorry.
|
|
| Re: test on BeBox |
|
|
By Rene Gollent - Posted on July 9, 2004 - 14:58:06 (#13259)
Current version when comment was posted: 0.16 |
|
|
You realize the posted driver is x86-specific right? It will not work correctly on a BeBox or PPC in its current form.
|
|
| Re: test on BeBox |
|
|
By Rene Gollent - Posted on July 9, 2004 - 14:58:05 (#13258)
Current version when comment was posted: 0.16 |
|
|
You realize the posted driver is x86-specific right? It will not work correctly on a BeBox or PPC in its current form.
|
|
| Test with bebox |
|
|
By mmt - Posted on July 9, 2004 - 07:40:25 (#13257)
Current version when comment was posted: 0.16 |
|
|
I tried this driver with my bebox + G200PCI (works fine in R5)
* Cold start (power off and on) does not work.
* Hot start (compile, install and restart) seems to work.
(but only 8bit mode, other color mode is strange color.)
|
|
| to MAX or not to MAX |
|
|
By rudolfc - Posted on February 23, 2004 - 14:28:54 (#11204)
Current version when comment was posted: 0.16 |
|
|
Hmm,
small correction: The MAX is not seperately recognized, but it's speeds _are_ used as this info is found seperately in the cards BIOS.
So in effect the driver is set to optimum speed at all times if all is right...
You can see if you have a MAX or not by looking at the logfile as well: the MAX has 360Mhz RAMDAC speed, while the normal G400 has 300Mhz RAMDAC speed.
Rudolf.
|
|
| ID 525 |
|
|
By rudolfc - Posted on February 23, 2004 - 14:18:21 (#11202)
Current version when comment was posted: 0.16 |
|
|
Hi,
Thanks for the compliment :-)
ID 525 can be a G400, a G400MAX or a G450 indeed. If you want to know, just enable full logging and have a look at the logfile. The driver will distinquish between the G400(MAX) and the G450.
G400 and G400MAX are the same, but G400MAX can do a bit higher refreshrates and is just a little bit faster than G400. The BeOS driver just uses the MAX as a normal G400 as it's still unknown how to distinquish between the two.
Rudolf.
|
|
| re: G450 PCI |
|
|
By memson - Posted on February 23, 2004 - 06:17:45 (#11191)
Current version when comment was posted: 0.16 |
|
|
I have a G450, and the driver works absolutely perfectly. Not tried the dual head mode yet though.
Dunno the exact card info, but the card ID is 525....
(It may be a G400, as the guy I bought it from seemed a little unsure himself. He assured me it was a G450 later in the conversation though..)
|
|
| G450 PCI |
|
|
By rudolfc - Posted on December 27, 2003 - 06:36:07 (#10184)
Current version when comment was posted: 0.15 |
|
|
Hi,
AFAIK it does not make any difference if you use a PCI or AGP based card with BeOS, regarding the (2D) driver. So, yes, it should work.
If it does not work after all, let me know so we can try to fix it..
(The cardID could be different in theorym though for Matrox this does not apply AFAIK)
Best regards,
Rudolf.
|
|
| g450 pci |
|
|
By an-dee - Posted on December 25, 2003 - 16:22:04 (#10171)
Current version when comment was posted: 0.15 |
|
|
Does this driver support the pci based g450 DH?
I would like to put one into a mini-itx based PC.
|
|
| Fantasy language |
|
|
By ryanknapper - Posted on December 18, 2003 - 10:07:25 (#10029)
Current version when comment was posted: 0.15 |
|
|
"Secondary head (maven) sync polarity programming"
Dude, you could make up whatever you want and I don't think anyone left in KDL could call you on it.
"Yeah, so I rotated the gonquax to feign the troc-processor."
|
|
| Parhelia / devrel |
|
|
By rudolfc - Posted on December 16, 2003 - 09:32:40 (#9989)
Current version when comment was posted: 0.14 |
|
|
Hi there,
Thanks for the tip: I did not know about this page.
I think I just might try contacting them again using this way.
You have to understand BTW that although they granted me access to some register level specs for older Matrox cards, they are ignoring normally sent mail to them from me about technical support. Also the developer pages, which list the registerlevel specs for older cards (if you are registered that is), still do not contain info about the Parhelia on the accesslevel I am at.
I'll keep checking that site every now and then, and I might try this form you pointed me at, as this is something they really should respond to. Although I think they should respond to normal mail as well :-/
Of course, if I am to start on a Parhelia driver, I am going to need a card to test with as well..
Rudolf.
|
|
| Parhelia Followup Question/Comment |
|
|
By Jace - Posted on December 16, 2003 - 01:14:32 (#9986)
Current version when comment was posted: 0.14 |
|
|
Howdy rudolfc
You said:
"As long as no register level specs are known BTW, I will not even try to support Parhelia cards. If I would have access to these specs, I would certainly take a (brief) look though.."
Matrox does not have the specs out there for some reason. However, they have a "request form" for developer support that might be useful if you do end up getting interested in this... I guess you already know about it, but here's the link anyway:
http://www.matrox.com/mga/support/tech_support/emaildevrel_form.cfm
Thanks again for the drivers you've made!
|
|
| BDirectWindow support hint... |
|
|
By rudolfc - Posted on November 5, 2003 - 12:25:51 (#9505)
Current version when comment was posted: 0.14 |
|
|
Hi there,
Here's a quick tip, which I just now found out due to some very sharp bug-report from someone (thanks!):
If you want to use windowed BDirectWindow modes (Chart demo app for instance), you have to make sure you have the hardcursor enabled.
BDirectWindow needs Parallel buffer access plus a hardcursor for this to work!!
I'll include this hint in the mga.settings documentation for the V0.15 release, which will also contain some bugfixes found during NeoMagic and nVidia driver development.
Rudolf.
|
|
| Re: BeBox |
|
|
By Rene Gollent - Posted on October 23, 2003 - 04:52:09 (#9325)
Current version when comment was posted: 0.14 |
|
|
The BeBox only supports the Millennium 1/2064W, the Mill2/2164W would definitely need a rewritten driver since Be aren't around to extend it any more.
|
|
| bebox.. |
|
|
By rudolfc - Posted on October 23, 2003 - 04:23:51 (#9324)
Current version when comment was posted: 0.14 |
|
|
Hi there,
Sorry I am slow in responding here at this time. I am afraid this driver only works on Intel machines currently. If openBeOS is going to run on a Pegasus(2) PPC system however, I think I just might try to update this driver to run on that.
It should be possible I think.
A BeBox will not be supported by me I'm afraid :-/
Best regards,
Rudolf.
|
|
| need Matrox driver for BeBox |
|
|
By digitalhap - Posted on September 12, 2003 - 02:08:55 (#8779)
Current version when comment was posted: 0.14 |
|
|
Hello,
I am trying to run a Matrox Millenium II card in my BeBox, but I need driver help. Can anyone assist?
Thanks
|
|
| MIL card owners |
|
|
By rudolfc - Posted on June 14, 2003 - 08:32:02 (#7734)
Current version when comment was posted: 0.14 |
|
|
If you use a Millenium 1 or 2 card with BeOS, and you have more than 2MB RAM on it, you need to specify this in mga.settings.
Millenium RAM amount detection is not yet implemented, the driver defaults to 2MB RAM. With 2MB RAM you cannot use high resolutions with high colorddepths. For instance 1024x768x32 bit color already needs 3MB RAM...
Rudolf.
|
|
| 1280x1024 |
|
|
By rudolfc - Posted on June 13, 2003 - 12:53:05 (#7711)
Current version when comment was posted: 0.14 |
|
|
Hi,
I'm afraid I don't understand what the problem is. AFAIK the driver is working perfectly. Just retested (once again) over here (R5.0.3) with driver 0.14: 1280x1024x60Hz is perfectly OK.
Uninstalling the driver can be done by changing to the folder you extracted the driver in before installation. If you type (in a terminal window):
make uninstall
The driver will be removed.
You can also do it by hand. Just remove the files the installation placed. The exact names and folders are mentioned on the driver's site (and are included with the driver as a HTML file) in the 'installation' part.
About the resolution problem: you will need to give more information to me for me to be able to say something more about it I'm afraid...
I could use a logfile with full logging enabled (via Email) and the cardtype you are using, the monitor type you are using, and the preferences app you are using. (Tested both Be's app and dualhead Setup: both are OK over here on the currently installed G550.)
Sorry I can't help more ATM.
BTW: I have no experience with the developer edition, not do I have it. The driver was written for the 'official' BeOS R5 version...
Best regards,
Rudolf.
|
|
| BeOS Matrox MGA driver |
|
|
By danten - Posted on June 13, 2003 - 05:59:59 (#7708)
Current version when comment was posted: 0.14 |
|
|
I have the same problem you have Yumurcak! And I couldn't undo the installation. I reinstalled BeOS, and I am using its standard driver which allow me 1280x1024. I have this resolutions on my system:
W98SE, 1280x1024 @ 60Hz
OS/2 Warp, 1400x1050 @ 60Hz
BeOS, 1280x1024 @ 60Hz
Linux, 1024x768 @ 60Hz
I have BeOS Developer Edition 1.1 installed.
|
|
| Screen modes |
|
|
By Yumurcak! - Posted on June 13, 2003 - 01:30:26 (#7705)
Current version when comment was posted: 0.14 |
|
|
Hello,
I have some problems i think. after downloading and installing your driver BeOS successfully initialized the display. But there's something:
W2K, 1280x1024 @ 70 Hz.
Linux, 1024x768 @ 70 Hz.
BeOS, 800x600 @ 75 Hz.
Can i use 1280x1024 on BeOS? How can i do it??
TIA
Yumurcak!
|
|
| kepla |
|
|
By Jess - Posted on April 8, 2003 - 17:14:56 (#6786)
Current version when comment was posted: V0.13 beta2 |
|
|
I doubt you can fix your first problem unless they change how it refreshes. I have the same problem but I just quickly resize & I'm off as I run 3 os's at 5 differant res's & refresh rates :) It does suck how the linux drivers for a given card draw to the left of where the windows drivers center off at a given refresh rate & then BeOS centers differant as well... Your only real choice atm is choose a higher or lower refresh & center it to that refresh rate... like 1280x 1024 @ 80ktz in linux & 1280x1024 @ 75htz in Be so
that each refresh rate is centered accordingly... if you can stand to lower or raise your current refresh that is.
|
|
| screen decay.. |
|
|
By rudolfc - Posted on April 8, 2003 - 13:44:14 (#6784)
Current version when comment was posted: V0.13 beta2 |
|
|
Hi,
Could you please enable full logging (in mga.settings config file) and send me a logfile?
Please delete the logfile mga.accelerant.log, make sure full logging is enabled, and reboot.
I would like to have a look at the newly created mga.accelerant.log that exists in your home folder then...
Please give me some details of your current setup also:
- Do you use 'usebios true' or 'usebios false'?
- Which driverversion are you running?
- If you use 'usebios false', can you change it to be 'usebios true', reboot and see if that helps please?
------
About the timing: I cannot really help you there. I do have one or two pointers though:
You could try changing the refreshrate in very small steps to modify the timing somewhat. You can do that by starting Be's screenprefs, and using the left and right arrows for that.
This will increase or decrease the refreshrate by 0.1Hz or so per step, probably moving the output a bit also.
You could also consider using a different refreshrate on BeOS compared to the other systems. Most monitors have a ROM inside them that's used to remember specific screen position settings for different resolutions _and_ refreshrates.
Using a few Hertz more or less is mostly sufficient.
Best regards,
Rudolf.
|
|
| Screen decay ! |
|
|
By kepla - Posted on April 8, 2003 - 10:40:37 (#6782)
Current version when comment was posted: V0.13 beta2 |
|
|
I'm running a 32 Mb G450, just using the primary display, at 1024 x 768 x 16 bit colour @ 75 Hz. Two small problems I'd could do with a little help, please:
1) The display on the monitor doesn't fit the screen - it is too wide and flat. I use the same resolution and refresh rate in both Linux & Windows without this problem, so don't really want to mess up the settings on the monitor. How can I adjust this within BeOS to match ?
2) Items displayed on the screen start off looking perfect and then gradually 'decay' a pixel at a time to the screen backgroud colour. It is almost as if some strange screen saver is running (it isn't). Any ideas of what might cause this and how to avoid it please ?
TIA
Kevin.
|
|
| extra G400 functionality compared to Be driver |
|
|
By rudolfc - Posted on February 17, 2003 - 03:47:22 (#5962)
Current version when comment was posted: V0.13 beta2 |
|
|
Hi,
Apart from dualhead and TVout functionality, this driver also adds hardware overlay capability for G400 on BeOS. So the Be G400 driver has none of these capabilities.
The Be driver for G200 already supports hardware overlay, but the openBeOS driver does it better... You'll see this clearly if you playback DVD on 1024x768 resolution for instance. Try downscaling to a small output window...
Of course G450 and G550 support does not exist via Be drivers, for these cards you'll indeed need this driver also. G450 and G550 have the same capabilities on BeOS as the G400 has with this driver. That is, TVout will be supported for these cards in the upcoming 0.14 driver.
For TVout and dualhead you need Dualhead Setup: also listed on BeBits.
Rudolf.
|
|
|
|
|
|
Comment Pages:
<< prev | 1 | 2 | next >> |
|
|
