Black screen under XFree86, AGP and Radeon 9500

zoomzoom · 11-29-2003, 12:10 PM

Hi folks,

Im not sure where to post this, as i don't know whether this is a hardware problem or not... So i will post it here

I'm using the Asus A7V8X (KT400) mobo with a Radeon 9500 (AGP3) graphics card and the fglrx driver from ATI (version 3.2.8)...

Well, i know it's difficult to get a AGP3 card on a kt400 working under linux...
So i did the following:

1. Downloaded the linux-2.4.23-rc5 kernel (which is now linux-2.4.23 stable, i think)

2. Downloaded a kt400-agp3-patch for linux-2.4.21-pre5 (do a google to find it)

3. Downloaded the linux-2.4.21-pre5

4. Applied the patch on it

5. Copied the two patched files (agp.h and apggart_be.c) from the linux-2.4.21-pre5 to the linux-2.4.23-rc5 kernel tree (replaced them)

6. Edited the Config.in so i can select the new AGP3-KT400 feature

7. Compiled and installed the linux-2.4.23-rc5 kernel with the AGP3 support ---> WORKED!

8. Rebooted and looked at dmesg: Detected AGP3 on VIA KT400 [...] Using apterture size 256MB. ---> cool! this also WORKED!

9. Downloaded and compiled the ATI drivers ----> WORKED!

10. Started X ----> DID NOT WORK... I just get a black screen...

So, does anybody know what the problem could be (i hope its not my "hack" or the driver)?

Here are the last "few" lines from my XFree86-log:

LoadModule: "fglrxdrm"
(II) Loading /usr/X11R6/lib/modules/linux/libfglrxdrm.a
(II) Module fglrxdrm: vendor="Fire GL - ATI Research GmbH, Germany"
compiled for 4.3.0, module version = 3.2.8
ABI class: XFree86 Server Extension, version 0.2
(II) fglrx(0): Depth moves disabled by default
(**) fglrx(0): Capabilities: 0x00000000
(**) fglrx(0): cpuFlags: 0x4000000f
(**) fglrx(0): cpuSpeedMHz: 0x000007c4
(==) fglrx(0): OpenGL ClientDriverName: "fglrx_dri.so"
(**) fglrx(0): using built in AGPGART module: no
(**) fglrx(0): UseFastTLS=0
(**) fglrx(0): BlockSignalsOnLock=1
(==) fglrx(0): EnablePrivateBackZ = NO
(II) fglrx(0): using CAIL version [ATI LIB=CAIL.LIB,IA32,2.0024]
(--) Depth 24 pixmap format is 32 bpp
(II) do I need RAC? No, I don't.
(II) resource ranges after preInit:
[0] 0 0 0xd2000000 - 0xd200ffff (0x10000) MX[B]
[1] 0 0 0xdc000000 - 0xdfffffff (0x4000000) MX[B]
[2] -1 0 0xffe00000 - 0xffffffff (0x200000) MX[B](B)
[3] -1 0 0x00100000 - 0x3fffffff (0x3ff00000) MX[B]E(B)
[4] -1 0 0x000f0000 - 0x000fffff (0x10000) MX[B]
[5] -1 0 0x000c0000 - 0x000effff (0x30000) MX[B]
[6] -1 0 0x00000000 - 0x0009ffff (0xa0000) MX[B]
[7] -1 0 0xcf000000 - 0xcf0000ff (0x100) MX[B]
[8] -1 0 0xd2800000 - 0xd2800fff (0x1000) MX[B]
[9] -1 0 0xd0000000 - 0xd001ffff (0x20000) MX[B]
[10] -1 0 0xd0800000 - 0xd0800fff (0x1000) MX[B]
[11] -1 0 0xd1000000 - 0xd10007ff (0x800) MX[B]
[12] -1 0 0xe0000000 - 0xdfffffff (0x0) MX[B]O
[13] -1 0 0xd1800000 - 0xd180ffff (0x10000) MX[B](B)
[14] -1 0 0xd4000000 - 0xd7ffffff (0x4000000) MX[B](B)
[15] -1 0 0xdbfe0000 - 0xdbffffff (0x20000) MX[B](B)
[16] -1 0 0xd2000000 - 0xd200ffff (0x10000) MX[B](B)
[17] -1 0 0xdc000000 - 0xdfffffff (0x4000000) MX[B](B)
[18] -1 0 0xd3000000 - 0xd3000fff (0x1000) MX[B](B)
[19] 0 0 0x000a0000 - 0x000affff (0x10000) MS[B]
[20] 0 0 0x000b0000 - 0x000b7fff (0x8000) MS[B]
[21] 0 0 0x000b8000 - 0x000bffff (0x8000) MS[B]
[22] 0 0 0x0000d800 - 0x0000d8ff (0x100) IX[B]
[23] -1 0 0x0000ffff - 0x0000ffff (0x1) IX[B]
[24] -1 0 0x00000000 - 0x000000ff (0x100) IX[B]
[25] -1 0 0x0000e000 - 0x0000e0ff (0x100) IX[B]
[26] -1 0 0x00009400 - 0x0000940f (0x10) IX[B]
[27] -1 0 0x00009800 - 0x0000981f (0x20) IX[B]
[28] -1 0 0x0000a000 - 0x0000a01f (0x20) IX[B]
[29] -1 0 0x0000a400 - 0x0000a41f (0x20) IX[B]
[30] -1 0 0x0000a800 - 0x0000a87f (0x80) IX[B]
[31] -1 0 0x0000b000 - 0x0000b00f (0x10) IX[B]
[32] -1 0 0x0000b400 - 0x0000b43f (0x40) IX[B]
[33] -1 0 0x0000b800 - 0x0000b87f (0x80) IX[B]
[34] -1 0 0x0000d800 - 0x0000d8ff (0x100) IX[B](B)
[35] 0 0 0x000003b0 - 0x000003bb (0xc) IS[B]
[36] 0 0 0x000003c0 - 0x000003df (0x20) IS[B]
(II) fglrx(0): UMM area: 0xdc500000 (size=0x03b00000)
(II) fglrx(0): driver needs XFree86 version: 4.3.x
(II) fglrx(0): detected XFree86 version: 4.3.0
(II) Loading extension ATIFGLRXDRI
(II) fglrx(0): doing DRIScreenInit
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmOpenDevice: minor is 0
drmOpenDevice: node name is /dev/dri/card0
drmOpenDevice: open result is 7, (OK)
drmGetBusid returned ''
(II) fglrx(0): [drm] created "fglrx" driver at busid "PCI:1:0:0"
(II) fglrx(0): [drm] added 8192 byte SAREA at 0xe0955000
(II) fglrx(0): [drm] mapped SAREA 0xe0955000 to 0x40016000
(II) fglrx(0): [drm] framebuffer handle = 0xdc000000
(II) fglrx(0): [drm] added 1 reserved context for kernel
(II) fglrx(0): DRIScreenInit done
(II) fglrx(0): Kernel Module Version Information:
(II) fglrx(0): Name: fglrx
(II) fglrx(0): Version: 3.2.8
(II) fglrx(0): Date: Sep 21 2003
(II) fglrx(0): Desc: ATI Fire GL DRM kernel module
(II) fglrx(0): Kernel Module version matches driver.
(II) fglrx(0): Kernel Module Build Time Information:
(II) fglrx(0): Build-Kernel UTS_RELEASE: 2.4.23-rc5
(II) fglrx(0): Build-Kernel MODVERSIONS: yes
(II) fglrx(0): Build-Kernel __SMP__: no
(II) fglrx(0): Build-Kernel PAGE_SIZE: 0x1000
(II) fglrx(0): [drm] register handle = 0xd2000000
(II) fglrx(0): [agp] Mode=0x1f000a0b bridge: 0x1106/0x3189
(II) fglrx(0): [agp] AGP disable mask 0x00000000
(II) fglrx(0): [agp] enabling AGP with mode=0x1f000b0a
(II) fglrx(0): [agp] AGP protocoll is NOT enabled for grafics board. (cmd=0x00000200)
(II) fglrx(0): [agp] grafics chipset has AGP v3.0 (native mode)
(II) fglrx(0): [drm] ringbuffer size = 0x00100000 bytes
(II) fglrx(0): [drm] DRM buffer queue setup: nbufs = 100 bufsize = 28672
(II) fglrx(0): [drm] texture shared area handle = 0xe4a6f000
(II) fglrx(0): VisualConfigs initialized
(II) fglrx(0): DRI initialization successfull!
(II) fglrx(0): FBADPhys: 0xdc000000 FBMappedSize: 0x00500000
(II) fglrx(0): Splitting WC range: base: 0xdc000000, size: 0x500000
(==) fglrx(0): Write-combining range (0xdc400000,0x100000)
(==) fglrx(0): Write-combining range (0xdc000000,0x500000)
(II) fglrx(0): FBMM initialized for area (0,0)-(1024,1280)
(II) fglrx(0): FBMM auto alloc for area (0,0)-(1024,768) (front color buffer - assumption)
(==) fglrx(0): Backing store disabled
(==) fglrx(0): Silken mouse enabled
(II) fglrx(0): Using hardware cursor (scanline 768)
(II) fglrx(0): Largest offscreen area available: 1024 x 504
(**) Option "dpms"
(**) fglrx(0): DPMS enabled
(II) fglrx(0): Using XFree86 Acceleration Architecture (XAA)
Screen to screen bit blits
Solid filled rectangles
Solid Lines
Dashed Lines
Offscreen Pixmaps
Setting up tile and stipple cache:
24 128x128 slots
(II) fglrx(0): Acceleration enabled
(II) fglrx(0): X context handle = 0x00000001
(II) fglrx(0): [DRI] installation complete

As you can see, there isn't an error... So what's wrong?!?!

Thank you!!

CU

zoomzoom · 11-30-2003, 07:58 AM

i just tried it with the older 2.4.21-pre5 kernel but it didn't help

Same error... PLEASE HELP ME!!!

UnTamed · 11-30-2003, 09:58 AM

Did you use #fglrxconfig to generate your XF86Config-4 file?
Did you select "UseInternalAGPGART" "no" to make sure the driver uses your patched
agpgart kernel module?

zoomzoom · 11-30-2003, 11:59 AM

both questions: yes...

i also tried it with useinternalagp=yes but no luck...

finegan · 11-30-2003, 02:20 PM

DPMS to automatically guess the monitor settings might have gone barfo kazoo, have you tried hand editting in your monitor Resolutions into the XF86Config file?

Cheers,

Finegan

UnTamed · 11-30-2003, 03:26 PM

Maybe you should patch ATI's agpgart_be.c instead and use internal.
Your chipset might be recognized as agp v2 but it's seems to work for most kt400 users.

/lib/modules/fglrx/build_mod/agpgart_be.c at line 6068 you'll find your chipset, just change the via_generic_setup for via_kt400_setup a few lines down and re-run the make scripts.

zoomzoom · 12-01-2003, 07:48 AM

I used the values from my monitor handbook (Iiyama VisionMaster Pro 450)
in the config file and i also tried to disable DPMS... No result...

I couldn't find either via_generic_setup or via_kt400_setup in that few lines down or up 6068...

UnTamed · 12-01-2003, 08:13 AM

Code:

    { PCI_DEVICE_ID_VIA_8377_0,
        PCI_VENDOR_ID_VIA,
        VIA_APOLLO_KT400,
        "Via",
        "Apollo KT400",
        via_generic_setup }, // AGP v2

Here's the whole section, mine starts at line 6068 ...

zoomzoom · 12-01-2003, 08:50 AM

WOOOOOOOOOOOOHOOOOOOOOOOOOOOOOOOOOO!!!!!!!!!!!!!!!!!

IT WORKS!!!!!!!!! I searched for that section and yep, found it (but not in line 6086, maybe because of an AMD patch i previously applied...). I changed it and it works great!!!!!!!!!!!!!!!!!!!!!!

THANK YOU !!!!!!!!!!

After about 3 weeks of patching, compiling and error chasing its good to finally see some result