Graphics Card -Sam 440ep-Flex PCI slot

[xenic]

There is also a problem on the U-Boot and Hardware side.
The Sam440ep-flex wil not boot with a Radeon HD card in slot number 1 and a Radeon 9250 in slot 2.
The rverse will boot, a 9250 in slot 1 and a Radeon HD in slot two but goes into an endless loop after the OS 4.1 boot logo.
With Amiga Os 4.1 update 2 Uboot displays on the Radeon HD Monitor and the OS 4.1 boot logo after on the 9250 Monitor.

Are you saying that both graphics cards work with Update 2? That's what your last line indicates to me.

[Spectre660]

It is possible that my dual monitor setup in Devs:Monitors is incorrect for a dual display so I cannot rule out that it should work under Update 2 yet.
To this point it does not though.These the the Monitor files that I have .

list WorkBench:Devs/Monitors/
Directory "WorkBench:Devs/Monitors" on Sunday 05-Aug-12
PCIGraphics.info 15834 ----rw-d Yesterday 13:57:18
Radeon 9250.info 15834 ----rw-d Yesterday 13:56:56
Radeon.info 15828 ----rw-d Yesterday 13:56:42
PCIGraphics 4096 ----rwed 20-Sep-09 00:02:22
Radeon 4096 ----rwed 20-Sep-09 00:02:22
Radeon 9250 4096 ----rwed 20-Sep-09 00:02:22
6 files - 58K bytes - 123 blocks used

There is also a problem on the U-Boot and Hardware side.
The Sam440ep-flex wil not boot with a Radeon HD card in slot number 1 and a Radeon 9250 in slot 2.
The rverse will boot, a 9250 in slot 1 and a Radeon HD in slot two but goes into an endless loop after the OS 4.1 boot logo.
With Amiga Os 4.1 update 2 Uboot displays on the Radeon HD Monitor and the OS 4.1 boot logo after on the 9250 Monitor.

Are you saying that both graphics cards work with Update 2? That's what your last line indicates to me.

[Spectre660]

These are the tool types of the three monitor files

Radeon

CMPLENGTH=6
BOARDNAME=Radeon
DDCMODE=All
(VSYNCMIN=60)
(VSYNCMAX=70)
(HSYNCMIN=31000)
(HSYNCMAX=70000)
(MODE=640x480@60)
(MODE=1280x720@60)
(MODE=800x600@60)
(MODE=1280x720@60)
(MODE=1024x768@60)
(MODE=1280x720@60)
(FAKENATIVEMODES=No)
INTERRUPT=Yes

Radeon 9250

CMPLENGTH=11
BOARDNAME=Radeon 9250
DDCMODE=All
(VSYNCMIN=60)
(VSYNCMAX=70)
(HSYNCMIN=31000)
(HSYNCMAX=70000)
(MODE=640x480@60)
(MODE=1280x720@60)
(MODE=800x600@60)
(MODE=1280x720@60)
(MODE=1024x768@60)
(MODE=1280x720@60)
(FAKENATIVEMODES=No)
INTERRUPT=Yes

PCIGraphics

CMPLENGTH=11
BOARDNAME=PCIGraphics
DDCMODE=All
(VSYNCMIN=60)
(VSYNCMAX=70)
(HSYNCMIN=31000)
(HSYNCMAX=70000)
(MODE=640x480@60)
(MODE=1280x720@60)
(MODE=800x600@60)
(MODE=1280x720@60)
(MODE=1024x768@60)
(MODE=1280x720@60)
(FAKENATIVEMODES=No)
INTERRUPT=Yes

[xenic]

@Spectre660

With Amiga Os 4.1 update 2 Uboot displays on the Radeon HD Monitor and the OS 4.1 boot logo after on the 9250 Monitor.

In the line above you say that you see UBoot on one monitor and OS4.1 on the other monitor. To me that means that both cards are working. What do you expect to see?

You listed 3 monitor drivers for 2 graphics cards. I've never seen a PCIGraphics monitor driver in OS 4.1. Where did that come from? How can you have a monitor driver for a non-existant graphics card?

Returning to your original problem with the Radeon 9250 not working in other PCI slots, I do remember a similar problem I encountered with one of the OS 4.0 updates on my µA1. For some reason the monitor MODEIDs changed and the MODEIDs saved for my screenmodes were invalid and caused some programs to crash. You might try renaming your WorkBench screenmode prefs (ENVARC:Sys/screenmode.prefs) so that WorkBench opens in a default mode (probably 640x480) with your Radeon 9250 in a different PCI slot. I can't think of any other possibilities and can only suggest that you start moving Update4 files one-at-a-time to your Update2 boot SD to determine what is causing the problem.

[Spectre660]

@xenic

Both Cards work if in the right slots. Slot 1 for the Radeon 9250 and Slot 2 for the other card. (I also tried it with 2 Radeon 9250's)
The PCIGraphics monitor driver is on the Acube site. It is needed for multiple graphics card usage. If you read the X1000 threads you will see it mentioned there. I have not found any official documentation though. Mine was installed by an early beta of the Radeon HD driver.The Sam460ex also has this driver installed .
My problem is not the mode id with updates 3 and 4 as the Workbench screen comes up with one or two Hard drive icons then freezes.
I also tried from hardrive to eliminate any USB issues causing it and got the same results.
In am not up to replacing kickstart components at this point but hope that both Acube and Hyperion will look into these issues as multiple
displays could become useful on Amiga OS 4.x in the future.

[m3x]

Just tested the following configuration on a Sam440ep-flex:

Radeon 9000 PCI into slot 1 (66 Mhz, the one near the CPU)
Radeon HD 2400 PCI into slot 3
RadeonHD.chip 0.32

In order to have both gfx cards working you need to:

- download the dual graphic cards setup from here:

http://www.acube-systems.biz/download/D ... dSetup.lha

- place a Radeon 9xxx monitor driver in Devs/Monitors
- place a PCIGraphics monitor driver in Devs/Monitors
- copy Picasso96Settings into DEVS:
- run Picasso96Mode, select the Radeon HD board from the "Attach settings to" menu, then Save
- disable rtg.library DMA with "setenv save Picasso96/UseDMA 0"
- if running the Sam440ep_setup, run it with the "SNDGFX" argument

With this setup, both gfx cards work, but there will be a performace degratation due missing DMA support.

Hope this helps.

[Spectre660]

@m3x

Thanks for the info.

I tried but still could not get it to work.
Ended up with a Guru after the OS 4.1 boot screen of the 9250 Monitor
Even got the Kickstart DIsk Animation again .
Uboot came up on the Radeon HD monitor.

Disabled the Sam440ep_setup.
Booted with the Radeon HD card only and did the Picasso96mode Attach setting and saved. tried using both slots 2 and 3 for the Radeon HD. and tried both a PCI x1300
and a HD5670 with a PCIE adapter.
Using the Radeon HD driver version 0.49 and PCIGraphics.card version 53.4 .
the disabling of the rtg.library DMA did allow booting from PCI slots 2 and 3 with a single card.

Just tested the following configuration on a Sam440ep-flex:

Radeon 9000 PCI into slot 1 (66 Mhz, the one near the CPU)
Radeon HD 2400 PCI into slot 3
RadeonHD.chip 0.32

In order to have both gfx cards working you need to:

- download the dual graphic cards setup from here:

http://www.acube-systems.biz/download/D ... dSetup.lha

- place a Radeon 9xxx monitor driver in Devs/Monitors
- place a PCIGraphics monitor driver in Devs/Monitors
- copy Picasso96Settings into DEVS:
- run Picasso96Mode, select the Radeon HD board from the "Attach settings to" menu, then Save
- disable rtg.library DMA with "setenv save Picasso96/UseDMA 0"
- if running the Sam440ep_setup, run it with the "SNDGFX" argument

With this setup, both gfx cards work, but there will be a performace degratation due missing DMA support.

Hope this helps.

[Radov]

Hello!

I had a very similar problem and would like to bring this topic back. My friend has already reported my case to hyperion some time ago, but this time I'm able to give you more information - thanks to the null-modem cable.

Crash after the boot screen happens with two of my Radeon cards:
- Radeon 7500 (64MB) Low profile
- Radeon 9250 (128MB) Low profile
Luckily – I’m able to boot Amiga OS 4.1 update 5 with (another) Radeon 9200 128MB (Low profile).
This is the log taken with Radeon 7500, that made AmigaOS 4.1 (updates 3-5) to crash after the boot screen:
[hidden="AmigaOS + R7500 log]U-Boot 1.3.1d (Feb 15 2011 - 17:02:59)

CPU: AMCC PowerPC 440EP Rev. C at 799.999 MHz (PLB=133, OPB=66, EBC=66 MHz)
I2C boot EEPROM enabled
Bootstrap Option H - Boot ROM Location I2C (Addr 0x52)
Internal PCI arbiter enabled, PCI async ext clock used
32 kB I-Cache 32 kB D-Cache
Board: Sam440ep-flex
I2C: ready
DRAM: 512 MB
PCI: Bus Dev VenId DevId Class Int
01 04 1013 6005 0401 00
01 05 1033 0035 0c03 00
01 05 1033 0035 0c03 00
01 05 1033 00e0 0c03 00
00 0a 12d8 8150 0604 00
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0c 1002 5157 5157 58
00 0e 1095 3114 0180 00
In: serial
Out: serial
Err: serial
Net: ppc_4xx_eth0, ppc_4xx_eth1
SM502: not found
VGA: NO CARDS
USB: OHCI pci controller (1033, 0035) found @(1:5:0)
scanning bus for devices... 1 USB Device(s) found

Last Switch off: Date: 2012-08-15 (Wednesday)

IDE Device 0: OK
Model: SAMSUNG HD253GJ Firm: 1AJ10001 Ser#: S24JJ9BZ601007
Type: Hard Disk
Supports 48-bit addressing
Capacity: 238474.1 MB = 232.8 GB (488395055 x 512)
not available
AOS4 FLB
found AOS4 SLB
Loading .text @ 0x014000a0 (24080 bytes)
Loading .rodata @ 0x01405eb0 (2482 bytes)
Loading .data @ 0x01416870 (208 bytes)
Loading .got2 @ 0x01416940 (596 bytes)
Loading .fixup @ 0x01416b94 (104 bytes)
Loading .sdata @ 0x01416bfc (16 bytes)
Clearing .sbss @ 0x01416c0c (44 byt
Clearing .bss @ 0x01416c38 (4104 bytes)
A1 Second-level booter V2.
Idea, design and main code: Andrea Vallinotto.
Testing and fixes: Stefan Burstroem
SFS and ISO9660 reading routines: Joerg Strohmayer.
Linux booting extensions: Andrea and the GRUB team.
A1 Second-level booter V2.
Idea, design and main code: Andrea Vallinotto.
Testing and fixes: Stefan Burstroem
SFS and ISO9660 reading routines: Joerg Strohmayer.
Linux booting extensions: Andrea and the GRUB team.



Machine Check Exception.
Caused by (from msr): regs 0df83e70 Data Write PLB Error
NIP: 0FF90A2C XER: 00000000 LR: 0FF908E4 REGS: 0df83e70 TRAP: 0200 DEAR: 0000000
0
MSR: 00029000 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 00

GPR00: C0C0C0C0 0DF83F60 0000051B 00000000 00000008 0DF83FC0 00000000 00000276
GPR08: 00000000 00000000 C0C0C0C0 00000000 3F400000 82800430 0FFDCD00 0DF84200
GPR16: 01416C20 0DF8420C 014063F8 014063E4 0140640C 01416C20 01416C20 0FFD8BF8
GPR24: 0FFE9C08 00000000 00000000 0FFCC220 0FFE9C14 0DF84B38 0FFDCF20 0FFE9C20
Call backtra
0DF83FC0 0FF90BC8 014028BC 01402CEC 0FF91B78 0FFB8DA8 0FFB8ED4

CPU: AMCC PowerPC 440EP Rev. C at 799.999
0FFB8DA8 0FFB9884 0FFB08F4 0FF876B0
machine check[/hidden]

If needed, I can provide You with another log taken with R9250.

[Spectre660]

Either your card is defective or does not fit properly in the PCI Slot or U-boot is confused by this card as U-Boot can't find any video cards. You can eliminate the card being defective if you can try it on a PC and it is ok.

SM502: not found
VGA: NO CARDS

[Radov]

Yes, I have noticed this "no vga" entry too, but can assure You, that this card was perfectly working with AmigaOS 4.1 update 2.
This week I'll get new RS232 connector, maybe another logs will say more...

PS. On the other hand - yes, I do have problem with PCI slots. Radeon 9200 must fit pefectly or screen gets some weird colored lines/trash. Those lines doesn't seem to make system to be less stable though...
PS2. Both the other Radeons: 7500 and 9250 were unable to boot update 3 & 4 a long before I've noticed first problems with PCI slots. Oh... and I've already them with update2 installed again: without any problems ...