 |
 |
|
Ati Radeon Graphics Driver |
|
|
|
|
|
|
| ID 4a69 does not work |
|
|
By Garjala - Posted on January 21, 2005 - 20:08:50 (#15499)
Current version when comment was posted: 5.1 |
|
|
I have HIS Radeon X800 vivo (ID 4a69). Does not work (not yet at least).
|
|
| RE: PB with radeon 9600 pro (Dell inspiron 8600c) |
|
|
By ljr - Posted on January 21, 2005 - 16:39:03 (#15493)
Current version when comment was posted: 5.1 |
|
|
Ok, problem solved with detect.c modification. I associated the device ID to a rt350 with mobility flag to true...
Nice to have BeOS with 1680x1050 ;-P
PS: Radeon Screen does not show 1280x800 ! any idea to have it ??
|
|
| PB with radeon 9600 pro (Dell inspiron 8600c) |
|
|
By ljr - Posted on January 12, 2005 - 23:12:57 (#15387)
Current version when comment was posted: 5.1 |
|
|
I try to install Zeta Neo on a dell 8600c (P-m 1.6, 855GM, 768Mb, 9600 pro) and i can't have the desktop managed by radeon driver 5.1 . After last icons, screen is weird and Be seems to be in kdl....
After desactiving user add-ons, i check irq and almost of the devices uses IRQ11 (USBx4, VGA, ETH, FWire, Wifi and sound). This Dell Bios is bullsit ! :-(
The radeon id: 1002 / 4e50
Any idea ?
|
|
| Radeon 9600 0x4e51 |
|
|
By biffuz - Posted on January 11, 2005 - 23:38:19 (#15382)
Current version when comment was posted: 5.1 |
|
|
I got a new Sapphire Atlantis Radeon 9600 128 Mb (SKU# 11019-00-20). Its PCI ID is 0x4e51.
Some reports on the 'net says this is for the Mobility Radeon 9600 (M10), but I added this ID in the driver as a RV350 (as it should be) and it works.
Another duplicate PCI ID perhaps?
|
|
| cursors |
|
|
By Euan - Posted on January 4, 2005 - 21:54:49 (#15288)
Current version when comment was posted: 5.1 |
|
|
Mike, I think windows dows the high-colour / true-colour cursors via software. IE not using the hardware cursor function of the graphics processor unit. Additionally for the Alpha-transparency effects that 2K/XP brought I think these require DirectX, rather than using basic Alpha Transparency-bit-blt copies. Due to performance reasons I would suspect. I guess the simplest way to do it is 2 bitmaps the size of the cursor, one a snapshot the framebuffer region before drawing the cursor, and the second the cursor. When ever the cursor moves, you remove the cursor by using the first bitmap to redraw the original pixels, move the region, make new "backup", bltbit cursor image over new region. You can probably also do it using additonal bit-mask images, AND/XOR/OR copies and so on...
You most likely need v-sync ebabled to prevent the cursor flickering, if you go this route.
Sorry, don't speak russian, other than what babelfish can translate!
|
|
| Re: New Starter |
|
|
By SD - Posted on January 4, 2005 - 18:00:03 (#15285)
Current version when comment was posted: 5.1 |
|
|
If you use Windows OS, go to toshiba or ati support page (support.ati.com)
This driver is for BeOS and Zeta OS-es, and in case you use it - just double click on *.pkg file
|
|
| New Starter/ATI Mobility Radeon 7000 IGP driver |
|
|
By tynesider - Posted on January 4, 2005 - 17:43:22 (#15283)
Current version when comment was posted: 5.1 |
|
|
Hi,
I have the ATI Mobility Radeon 7000 IGP driver installed on my new Toshiba laptop, I cannot get the s-video to work so i can see the screen on my TV. I tried to download the RadeonDriver_5.1.pkg, but how do I run a pkg file?
Sorry if all this sounds a bit bit basic but I'm a novice.
Thanks
Tynesider
|
|
| yeah, clipping well |
|
|
By mike.dld - Posted on January 4, 2005 - 16:58:53 (#15282)
Current version when comment was posted: 5.1 |
|
|
Now got clipping to work too :)
Thanks alot Euan for pointing me out. What about cursor (see question above), how to make it 32bpp? Windows does it somehow...
---
BTW do you know Russian?
|
|
| clipping |
|
|
By Euan - Posted on January 3, 2005 - 07:11:29 (#15255)
Current version when comment was posted: 5.1 |
|
|
Clipping is not the responsibility of a graphics driver. The graphics driver simply receives drawing requests from the OS, and draws to the framebuffer. In beos the appserver manages windows and layers. X does it in unix land, and the GDI does in windows. Basically the "window manager" does it.
Congrats on the progress. Personally I think your mad!
|
|
| Thanks |
|
|
By mike.dld - Posted on January 2, 2005 - 22:06:03 (#15252)
Current version when comment was posted: 5.1 |
|
|
Got it working in 32bpp :)
Another question (hope I'm not too boring). What should I do when one (or 10) windows are on top (overlap) of another? Radeon has only 1 clipping rectangle, or there's something I don't understand?
|
|
| no 24bpp is not available |
|
|
By Euan - Posted on January 2, 2005 - 15:09:37 (#15251)
Current version when comment was posted: 5.1 |
|
|
No the 2D accelerator has no support for 24bbp. It specifically states 8,16,and 32 bpp in the developer documents (NDA contract from ATI required). AFAIK ATI's opinion is that 24bpp has no logic to it what so ever when you could use 32bbp, and have your memmory alighned to 4 and 8 byte memory accesses. So 24bpp is a redundant surface format. If I wrote the driver, I wouldn't support 8, or 16 either. Mainly because I'm really lazy. :)
If you contact ATI (devrel at ati dot com) they may give you access to their developer docs, but it's really difficult to let them give you access. It has DOS and Xfree code examples of each stage of the driver proces. IE detecting the card, setting up a display mode, drawing GDI objects, etc.
|
|
| 24bpp |
|
|
By mike.dld - Posted on January 2, 2005 - 14:43:30 (#15249)
Current version when comment was posted: 5.1 |
|
|
Thanks, Euan, that helps.
Another question: is there any possibility to force Radeon to work in 24bpp mode when pixel depth is 24? Looking to the xfree driver sources I met the following lines:
if (info->pix24bpp == 24) {
xf86DrvMsg(pScrn->scrnIndex, X_ERROR,
"Radeon does NOT support 24bpp\n");
return FALSE;
}
Any other suggestion?
|
|
| hehe, it's acually quite complicated... |
|
|
By Euan - Posted on January 2, 2005 - 09:04:52 (#15246)
Current version when comment was posted: 5.1 |
|
|
You probably want to email the author directly. As he's usually very busy, and not often around here.
I can give you some help however. The driver is based of the Xfree86 code, just majorly tidied up, and alot of redundant mess removed, and some very fancy memory management stuff. Thomas is good at that. ;)
The Radeon cards can operate using two different 2D mechanisms. PIO, and DMA. The driver you see here exclusively uses DMA bus mastering "ring buffers" to queue the 2d command instructions (draw rectangle, screen-to-screen copy, etc) via system memory. It is much more complex than doing standard Programed IO. In PIO mode you set some registers to enable 2d acceleration, then punch in registers for drawing. You may want to consider that in what you expect the driver to do. Of course DMA performs better and is more efficient. Just many times more complex. In DMA mode, you need to download microcode to the card, initialise the buffer memory regions, and setup various functions for buffer syncronisation etc.
The best way to see the flow of the driver is to turn the debug output to the highest limit, and save the syslog. I can provide you with one if you wish. THe debug output is *very* verbose, and provides the function name, as well as the action taken, for absolutely everything the driver / accelerant does.
From memory I think this is the order of things:
done by the kernel driver:
Map PCI memory / register regions
Save the VGA bios area into memory
process the bios for PLL clock settings, connected devices, and special features (Video In/Out TV-out etc).
done by the 2D accelerant:
setup the output device registers (CRTC / DVI / TV-out)
calculate the clock settings (PLL deviders)
set the clock registers to setup up the screen resolution (setdisplaymode.c)
initialise the 2d Engine (command processor if in DMA mode).
The hardest task I think you will have is unraveling code that is specific to the architecture of BeOS' display driver system, ie the driver / accelerant relationship, multiple adapaters, multiple heads on single cards, and also how the accelerant is cloned everytime a new workspace / app requires it.
|
|
| Help on porting |
|
|
By mike.dld - Posted on January 1, 2005 - 14:18:17 (#15243)
Current version when comment was posted: 5.1 |
|
|
I'm new to drivers developing and my first project is the attempt to port your driver for MenuetOS. As this OS written in assembly, drivers must also be of that kind. I've already made HW cursor support (but still B/W, is there's possibility for 32-bit??) and now have stuck with 2D acceleration. If you could post there a list of functions in order they're called from driver initialization to possibility of drawing smth. it'd be cool. Also you may e-mail me... And please, If you have some special info about ATi Radeon 9000 (my card) mail it too. Hope you'll answer ;)
|
|
| Unfortunately I have to use BONE |
|
|
By llama - Posted on January 1, 2005 - 03:13:36 (#15242)
Current version when comment was posted: 5.1 |
|
|
I would use standard R5 if I could but my network card (Broadcom 570x) needs BONE for Nathan W's (excellent) driver, and dano (although being buggy) is less buggy than R5 + BONE IMHO (things like DHCP work in dano but not in R5 + BONE for me etc.)
Well it looks like I'll just have to live with the crashing (it's actually not THAT big a deal anyway as I usually run the laptop plugged in)
Cheers,
Andrew
|
|
| acpi |
|
|
By Euan - Posted on January 1, 2005 - 02:36:59 (#15241)
Current version when comment was posted: 5.1 |
|
|
the dano, phos, bone versions of beos have extra ACPI support, which all drivers are supposed to support. AFAIK. Becuase DANO is incomplete, I don't think it works correctly, and the Graphic drivers don't implement the ACPI commands. I never use DANO / BONE / PHOS as it's just too buggy. IMHO
|
|
| PS |
|
|
By llama - Posted on December 31, 2004 - 16:26:23 (#15237)
Current version when comment was posted: 5.1 |
|
|
Sorry, forgot to say:
I'm using the radeon 340 igp on an IBM Thinkpad R40e. This laptop has an incredibly broken bios (linux will not install unless you disable ACPI, and then fix things afterwards) so most likely it is my laptop.
Cheers,
Andrew
|
|
| Strange problem with radeon driver |
|
|
By llama - Posted on December 31, 2004 - 16:23:08 (#15236)
Current version when comment was posted: 5.1 |
|
|
I'm having a strange problem with the radeon driver on my laptop. The driver works fine but if I plug or unplug the laptop into the power socket in the wall (or shut the lid, basically any ACPI event) the system freezes with no error. That means if my battery is getting flat I have to shut down, plug it in, and then reboot! The strange thing is this happens with BeOS DevEd 1.1 and PhOS (dano), but NOT with MAX. Different kernels? I know dano has a different kernel but I thought DevEd and MAX were both based on 5PE? DevEd gives a choice of kernels when installing but it happens no matter which one I choose.
It's definitely the radeon driver(or something else having a knock-on effect) because in VESA mode there is no problem with ACPI events, the system keeps on ticking along.
Not really a big problem, but a hassle. Any ideas on how the MAX kernel differs from DevEd, and if that difference can be applied to the dano kernel? (I'd much rather use PhOS then either of the others - it's just sooo much slicker!)
Cheers,
Andrew
|
|
| ID fix |
|
|
By Blafasel - Posted on December 30, 2004 - 01:16:16 (#15222)
Current version when comment was posted: 5.1 |
|
|
Seems I have the same problem with my 9700 Mobility, so I'd appreciate to see any update to this driver and would be glad if someone could send me a patch/updated driver..
Actually I've never compiled anything for beos..
TIA, Ben
eMail address would be benjamin dot podszun at gmail dot com
|
|
| Former post |
|
|
By Ding - Posted on December 29, 2004 - 14:30:44 (#15218)
Current version when comment was posted: 5.1 |
|
|
Have to change the virtual to 1280 960 as well otherwise the menus will be rendered partly outside the edge of the screen
|
|
| Change of vertical resolution |
|
|
By Ding - Posted on December 29, 2004 - 14:18:46 (#15217)
Current version when comment was posted: 5.1 |
|
|
I successfully changed the vertical resolution to 960 instead of 1024 just as was described a few posts back.
Changing the app_server_settings was easy.
A few things to be aware of:
- app_server_settings text file cannot be saved to as "Save" from within styled edit but only as "Save as"...no problem though.
- prior to changing the settings and reboot, move the start menu to the upper region because it will disappear if going from 1024 back to 960 and the start menu is at the bottom.
------------
Device /dev/graphics/1002_4e48_010000
Workspaces 9 {
Workspace 0 {
timing 157640 1280 1344 1504 1728 960 1025 1028 1072 0x00000000
colorspace 0x00000008
virtual 1280 1024
flags 0x00000000
color 336698
}
Workspace 1 {
...
}
---------------
Timing of 157640 is equivalent to 85.1 Hz
Now it would be nice if I could choose 1280x960 in the screen preferences.
|
|
| but not M10! |
|
|
By Euan - Posted on December 25, 2004 - 23:43:19 (#15171)
Current version when comment was posted: 5.1 |
|
|
AFAIK It should be M11, and RV360.
But there's practically no difference between M10 and M11 anyway. :)
|
|
| Yeah looks like the ID needs updated. |
|
|
By Euan - Posted on December 25, 2004 - 23:33:31 (#15169)
Current version when comment was posted: 5.1 |
|
|
It looks like this ID code was just reserved by the devs), and so it was later re-assigned. Radeon_NP is correct as NP is ascii for 0x4e50! So we know that is right. Also the NP model is the model for the 15" Powermac range according to google.
However the link below shows that the 4e50 ID has been changed at sometime:
http://pciids.sourceforge.net/iii/?i=1002
Look for 4e50:
RV350 [Mobility Radeon 9600 M10]
New PCI ID provided by ATI developer relations
You can also see some devices it's used on:
http://pciids.sourceforge.net/iii/?i=10024e50
|
|
| ID |
|
|
By DaaT - Posted on December 25, 2004 - 21:02:12 (#15165)
Current version when comment was posted: 5.1 |
|
|
|
| To 7 |
|
|
By DaaT - Posted on December 25, 2004 - 21:01:07 (#15164)
Current version when comment was posted: 5.1 |
|
|
Hey, my M9700 has that ID as well, could you send me the driver to daat at cybernaute dot com? Appreciate it thx :)
|
|
| Radeon 9700 Mobility |
|
|
By 7 - Posted on December 25, 2004 - 15:30:51 (#15158)
Current version when comment was posted: 5.1 |
|
|
I was able to make my radeon 9700 Mobilty work using the exact same method as the previous poster (identifying as a rv360 and changing the params of it in the source).
The odd thing is that my card id is listed in the source as being DEVICE_ID_RADEON_NP (0x4e50), which is either wrong in the source, or ATI has re-used a device ID.
Given a comment in the source, I am willing to guess that the device detection constants might have been wrong, and am going to add an m11 monicer to the source with my ID and the one of the previous poster.
Thomas, I would like to know if you have any thoughts on this, or want my updated source.
Alan
|
|
| RE: ATI Radeon 9700 SE Mobility 128Mb |
|
|
By xeD - Posted on December 24, 2004 - 17:54:19 (#15154)
Current version when comment was posted: 5.1 |
|
|
ok i reply to myself:
Got it working by a changing the source code.
My card as a M11 chip that is similar to the R360.
I've modified the RV360 cardID definition to match mine then modified the rv360 struct by setting mobility to true and 'hasvip' to false.
Reboot and now got laptop fully supported!
|
|
| ATI Radeon 9700 SE Mobility 128Mb |
|
|
By xeD - Posted on December 24, 2004 - 08:29:03 (#15150)
Current version when comment was posted: 5.1 |
|
|
This card found on a new laptop (Fujitsu/Siemens 7424)
does not work.
Looking in syslog I found:
KERN 'app_server'[24]: Radeon - Radeon_CardDetect: no supported devices found
According to Devices the CardID is 4e52
|
|
| X800 support? |
|
|
By mconrad - Posted on December 20, 2004 - 07:31:47 (#15115)
Current version when comment was posted: 5.1 |
|
|
Any chance of getting support for the newest cards? Thx.
|
|
| 2 Ding |
|
|
By SD - Posted on December 18, 2004 - 14:03:58 (#15087)
Current version when comment was posted: 5.1 |
|
|
For experiment you can try to set required mode manually.
Open file
/boot/home/config/settings/app_server_settings
with text editor.
Try to edit
line which look like
timing 78750 1024 1040 1136 1312 768 769 772 800 0x60000000
for some workspace (not 0).
2nd and 6th parameters are resolution numbers (my card uses 1024*768 atm, as you see)
3,4,5 are adjustments for horizontal resolution,
7,8,9 - for vertical
1st is related to refresh rate, but i don't remember now real formula. Maybe you can get more info, investigating parameters and manuals for Linux XFree server. But you can guess some bits of that, changing refresh rate in BeOS and looking back at that file.
So, after setting some xperimental workspace to use those parameters, reboot, and switch to that workspace. Then back to 0, with Alt+F1, if you dislike result:)
If and when you've get proper result, copy those settings for all workspaces.
|
|
| More support for graphics modes |
|
|
By Ding - Posted on December 18, 2004 - 12:02:30 (#15085)
Current version when comment was posted: 5.1 |
|
|
Hellow,
I would very much like to see more modes supported.
I have a Radeon 9800pro and set Windows at 1280x960 at 85hz and its showing a very crisp image on a Iiyama Vision Master Pro 410, which is quite old by now.
This mode is used for years now on Windows and frankly and don't know why its not supported on BeOS.
The reason for 960 pixels vertically is because you could still draw concentric circles, but not with 1024 pixels vertical. The latter screen resolution gives a distorted look when using BeOS and that does not look nice.
|
|
| 9700 Mobility fails to work (fallback to vesa apparently) |
|
|
By Blafasel - Posted on December 15, 2004 - 10:36:18 (#15055)
Current version when comment was posted: 5.1 |
|
|
Hi there.
I'm trying to play with Zeta Neo on my Asus M6 Laptop, but - after I switched to this driver from the one YT delivers (and which didn't work at all -> crash during boot) - it seems that the driver has some problems/issues with my card. I can boot up without a problem, but I'm stuck with a really slow and sluggish UI and a resolution fixed at 800x600.. Is there anything I can provide to help locating the problem? A guy on irc told me, that he assumes a failure in the driver's initialization and therefor an automatic fallback to vesa. Any help? TIA, Ben
|
|
| R9800Pro/XT |
|
|
By BePage - Posted on December 14, 2004 - 19:28:53 (#15050)
Current version when comment was posted: 5.1 |
|
|
Some 9800Pro got shipped with R360 chips. (Meaning they are basically down-clocked XT's)
|
|
| 9800XT |
|
|
By BeAR - Posted on December 14, 2004 - 04:22:59 (#15047)
Current version when comment was posted: 5.1 |
|
|
Just want to say that it works fine with ASUS Radeon 9800XT 256Mb VIVO.
PS. As a reply to some post below: Radeon 9800Pro is based on R350, 9800XT - on R360. But I don't know if there is much difference exept hardware temperature monitoring.
|
|
| This is a great driver! |
|
|
By Randall - Posted on December 13, 2004 - 23:06:06 (#15044)
Current version when comment was posted: 5.1 |
|
|
I have been useing this driver for quite som time on my HP Pavilion ZE4115 laptop it on BeOS & Zeta. Sure beats working in gray scale or VESA mode. I can now play DVD's & AVI's smoothly with out maxing out my cpu. I just think credit should be given where credit is do. You make BeOS Rock!
Thanks Thomas
|
|
| problem! |
|
|
By dragossido - Posted on December 10, 2004 - 12:00:28 (#15009)
Current version when comment was posted: 5.1 |
|
|
I have problem with my notebook.
I am not able to speack well english, but i try!
Is fujitsu-siemens amlio A (CY 26) amd 2600mhz 250mb of ram video raedon igp 320M and audio ac 97
I have try to istall the beos max edition, but i can't.
So i have install the cd edition and the developer edition.
The video was in the centre of my video an not all the video, so with the help of the people of itbug i have download your video driver, but now i have biger problem because the immagin on the video are bad! I have already ask to the people of itbug, but they speack that you are the only that can help me!
Now on the video there is on the right a black zone of about 400 pixel and before these zone the clock on the menu there is two times... and when i go with mouse some pixel before the black zone there is two little hand (normal is one little hand) and the voice on the menu is write like when you try to see the computer on the tv! Can you help me?
|
|
| 677001 |
|
|
By dragossido - Posted on December 10, 2004 - 12:00:05 (#15008)
Current version when comment was posted: 5.1 |
|
|
I have problem with my notebook.
I am not able to speack well english, but i try!
Is fujitsu-siemens amlio A (CY 26) amd 2600mhz 250mb of ram video raedon igp 320M and audio ac 97
I have try to istall the beos max edition, but i can't.
So i have install the cd edition and the developer edition.
The video was in the centre of my video an not all the video, so with the help of the people of itbug i have download your video driver, but now i have biger problem because the immagin on the video are bad! I have already ask to the people of itbug, but they speack that you are the only that can help me!
Now on the video there is on the right a black zone of about 400 pixel and before these zone the clock on the menu there is two times... and when i go with mouse some pixel before the black zone there is two little hand (normal is one little hand) and the voice on the menu is write like when you try to see the computer on the tv! Can you help me?
|
|
| 2 Rossi |
|
|
By s_d - Posted on December 6, 2004 - 01:36:57 (#14967)
Current version when comment was posted: 5.1 |
|
|
http://www.bebits.com/app/731
did you try that?
|
|
| oops ... 4c4d seems to be an ATI Rage Mobility-P |
|
|
By rossi@aixplosive.de - Posted on December 5, 2004 - 22:39:23 (#14965)
Current version when comment was posted: 5.1 |
|
|
... I guess I'm out of luck, it's not even a radeon chipset ... sorry!!!
|
|
| 9800XT |
|
|
By rossi@aixplosive.de - Posted on December 5, 2004 - 21:20:54 (#14963)
Current version when comment was posted: 5.1 |
|
|
Just for the record, yes the 9800XT is just a higher clocked version of the 9800Pro.
ATI usually has different clock speeds for all ASICS, usually they are suffixed as follows:
XT for the highest clocked variant (performance segment only, i.e. 9800XT, X800XT)
Pro second highest clocked variant /again performance segment)
LE mainstream segment, this suffix is usually dropped (9600)
SE entry segment, often only uses a 64bit memory interface
Hope this helps a bit! Also note, that actual clockspeeds may vary if you have purchased a board within an OEM system, as OEM variants are not identical to their retail counterparts and often use different specifications.
Cheers,
Rossi
|
|
|
|
|
|
|
|
|
 |
|
|
|
|
|
|
|
|
 |
| |
Recent Downloads - # 2
Total Downloads - # 2
Total Views - # 2
User Ratings - # 629
|
|
|
|
 |
| |
Video Drivers
|
|
|
|
 |
|
1. BePodder - 9.80
2. QEMU - 9.68
3. ScummVM - 9.57
4. cpu_fix - 9.42
5. Jukebox - 9.40
6. libdl.so - 9.40
7. Haiku AGP busm... - 9.35
8. vim6 - 9.31
9. Beezer - 9.26
10. HandBrake - 9.25
|
|
|
|
 |
|
1. BeOS 5 Perso... - 13,554
2. Realtek RTL8... - 13,089
3. Ati Radeon G... - 12,541
4. Ensoniq Audio... - 7,538
5. ATI Rage 128... - 7,432
6. USB Joystick... - 5,643
7. Broadcom 440x... - 5,399
8. S3 Trio 64 v2... - 4,705
9. USB Serial dr... - 4,703
10. Intel Extreme... - 4,461
|
|
|
|
 |
| You are not logged in.
Login
|
|
|
|
 |
|
