@Geri

Hi, i finally find a null modle cable, but this is the only log that i can see....

U-Boot 1.1.1 (Mar 3 2005 - 16:42:53), Build: 03/03/05

CPU: MPC7450 v2.1 @ 800 MHz
Board: AmigaOne
DRAM:
Information for SIMM bank 0:
Number of banks: 2
Number of row addresses: 13
Number of coumns addresses: 11
SIMM is registered
Supported burst lenghts: 8 4 2 1
Supported CAS latencies: CAS 3 CAS 2
RAS to CAS latency: 3
Precharge latency: 3
SDRAM highest CAS latency: 750
SDRAM 2nd highest CAS latency: 1000
SDRAM data width: 4
Auto Refresh supported
Refresh time: 1042 clocks
Bank 0 size: 512 MB
Bank 1 size: 512 MB

Slot 1: vacant
DIMM0_B0_SCR0 = 0x00ff9f00
DIMM0_B1_SCR0 = 0x00ff9f00
DIMM0_B2_SCR0 = 0x007f9f00
DIMM0_B3_SCR0 = 0x007f9f00
Using CAS 3 (slow)
DRAM_GCR0 = 0x00550800
Refresh set to 2083 clocks, auto refresh on
DRAM_REFRESH0 = 0x00019823
Mode bank 0: 0x00008032
Mode bank 1: 0x00008032
1024 MB
FLASH: 0 kB
VGA: 0 ..
In: serial
Out: serial
Err: serial

_________________
--------------------------------------
Search and Destroy
---------------------------------------

@damned

Although it says serial already you may still need to redirect it like so in UBoot:
setenv stdout serial
setenv stdin serial

The last one may not be needed, but easy from terminal. You may also need to remove Linux console bootarg as it defaults to serial IIRC.

@Hypex
@Geri

Okay, I finally got around to doing a dpkg --install of the mol .deb files on my LENNY partition, and setting up uboot to add the uImage-2.6.18-6-book3s-amigaone as one of my many boot options. I rebooted and did the following:

startmol -a --test

and I get a bunch of tests that scroll by and then the banner indicating "Self-test successful"

So I am guessing that to install OSX from a CD/DVD, all I need to do is:

startmol -a -x cdboot

Is that correct?

Okay, now to buy a MAC OSX 10.4 CD cheap from somewhere....


Thanks!

_________________
AmigaOne uA1-c 512M RAM - Only Amiga Makes It Possible!
Check my blog AmigaOne Computing

@scabit

Quote:

scabit wrote: @Geri Quote: You need to use a raw binary image for U-Boot, as it should be created in arch/ppc/boot/images/vmlinux.bin. You are correct. I found the vmlinux.bin file dated 4-24-13 in my arch/ppc/boot/images directory - so I know I created it - and used it to build a proper image file...and it works! Wel,, one small problem, I get a lot of these types of warnings at bootup: snd_ac97_codec: disagrees about version of symbol struct_module gameport: disagrees about version of symbol struct_module snd_via82xx: disagrees about version of symbol struct_module

Did you also rebuild the modules for your kernel image?

Quote:

scabit wrote: @Hypex @Geri Okay, I finally got around to doing a dpkg --install of the mol .deb files on my LENNY partition, and setting up uboot to add the uImage-2.6.18-6-book3s-amigaone as one of my many boot options. I rebooted and did the following: startmol -a --test and I get a bunch of tests that scroll by and then the banner indicating "Self-test successful"

Good!

Quote:

So I am guessing that to install OSX from a CD/DVD, all I need to do is: startmol -a -x cdboot Is that correct?

If everything is configured correctly in /etc/mol/molrc.osx, then you can simply run MoL with:

$ startmol -X --cdboot

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@damned

Quote:

damned wrote: @Geri ... In: serial Out: serial Err: serial

You have to redirect the output to the serial port, as Hypex already mentioned.

Thanks for testing!

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@Geri


Quote:

But read more about this here under "AmigaOS4 slb_v2".

Excellent work Geri! Now, not to degrade performance, but perhaps I can put this setting in a preseed?

Quote:

Thus I wouldn't hold my breath that the bug report is ever dealt with.

Sorry I meant the bugs on amigaworld.net.

Quote:

I only have kernel headers for the 3.2.0 kernel

I saw that. What I'd like are 2.6.32 headers now.

Trying to build MoL on that kernel but it is giving me errors. Seems related to the headers. Which I couldn't make anyway as I always get errors. From the linux-headers-2.6.32 packages installed.

So hoping you'd have a 2.6.32 header or source. It's recent so where is the source? Perhpas I should download that instead.

BTW this kernel on my Debian makes the screen go funny with yellow around some ghosted gadgets. Don't know why except resolution isn't set correctly on my LCD. Fine with 2.6.31.

Quote:

My A1SE on the other side needs a sound card, but suffers from the buggy PCI bus 0, where any DMA operation fails. In this case the card wouldn't help, except for saving me a PCI slot.

Aren't there sound cards for OS4 that can't use DMA? Or am I just thinking of the Mediator which can't use DMA cards in OS4 mode? Either way would a non-DMA sound card help here?

Quote:

And the TQFP chip nearby the SATA ports could be a SiliconImage controller chip.

Okay that's interesting. I'm new to the TQFP acronym.

Quote:

Okay! Let me know, if something shows up!

I rebuilt my 2.6.32 image and it now boots! But as mentioned I get corrupted gfx. What should be ghosted gadgets have yellowing. They look old. Perhaps related to incorrect screen setting. I don't know. They need to have more default screen sizes. If I could choose TFT/16:9 and my resolution it would be fine

But 3,2.0 is no good. Rebuilt the image. Crashes everytime. Crashes UBoot. 3.2.0 is not stable for me Scabit!

Quote:

It would be good to know about the root cause of this behavior.

Linux is nore stable on the worse AmigaOne!

µA1 must be the worse.

Quote:

The build of the MoL kernel module stopped due to a problem with the tools provided by the kernel sources.

That's worse than an incompatible source then. In the changelog I read about fixes for later kernels, like MOL_BIT in place of BIT, which later kernel uses. But I don't see it in the archive! Must be svn only.

Quote:

Okay, so you tried to get the MoL kernel module to build under OS4 (but still for a Linux kernel)?

Yes I suppose I did. Even though I was starting with the main program first and didn't get up to the kernel module part.

Quote:

IMHO, patching Fusion sounds like a better approach than getting MoL to work under OS4.

It would be if we had the source code. In that case an OS4 version would be better. Also be good to run in Workbench window or screen.

Quote:

In this case the ide_core parameters wouldn't be necessary anymore.

How does kernel 2.6.32 stand with this? As it stands we can support both. Paying attention to what libata can do.

Quote:

For libata drivers you can only disable DMA for all harddisks and/or CDROM drives on all controllers, but not for a single device or driver.

Good point. This can also be a feature I can make use of.

Quote:

The new libata based pata_via driver however quickly runs over the empty IDE channels.

It seems my original DMA settings are following a legacy here.

Quote:

I think it was this command line: "sort -k 1.1,1.8 -k 1.9n"

Okay, thanks ,will try that out.

Quote:

But you have to use the sort command installed in the /target system (/target/usr/bin/sort), because busybox' version is buggy.

That's interesting. I always tested on a full desktop. Then later confirmed if it worked in the installer enviroment.

Quote:

I wouldn't do so. I'm not that happy with Wheezy yet.

I just downloaded latest Debian Squeeze to rebuild current installer with.

Quote:

The weekend ends at Monday 12:00 am CE[S]T here!

And you made it!

@Hypex @Geri
ok, so after i tiped
setenv stdout serial
the cursor blink and nothing i write does anything (like boota), so i reset but without do the saveenv command the stdout return to vga
so i changed this parameter from os4, in the prefs->uboot
after that i can see all the serial ouput..'yeah' i say..but after reboot all i can see on the A1 monitor is a blinking cursor..on the serial output i can see the uboot choice screen but i can't do nothing and it seems that the auto choice after count down don't work
ouch!
i think i must open the case and take away the battery....

_________________
--------------------------------------
Search and Destroy
---------------------------------------

@damned

Quote:

damned wrote: @Hypex @Geri ok, so after i tiped setenv stdout serial the cursor blink and nothing i write does anything (like boota), so i reset but without do the saveenv command the stdout return to vga so i changed this parameter from os4, in the prefs->uboot after that i can see all the serial ouput..'yeah' i say..but after reboot all i can see on the A1 monitor is a blinking cursor..on the serial output i can see the uboot choice screen but i can't do nothing and it seems that the auto choice after count down don't work ouch! i think i must open the case and take away the battery....

Actually you should be able to see everything you type on the keyboard on the serial output. That's how I do it to switch back to vga stdout.

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@Geri

but no linux or amigaos seem to start..

_________________
--------------------------------------
Search and Destroy
---------------------------------------

i opened the case and take away the cmos battery for at least 30sec but when i reassembled all the things and powered up the situation was same as before! ouch!!
so i go to uboot menu and a restore factory default saved my day....fiuuuuu!!
but i remain so sad that i can't use new kernels..

_________________
--------------------------------------
Search and Destroy
---------------------------------------

@damned

Quote:

damned wrote: @Geri but no linux or amigaos seem to start..

You could still start it manually without saving the U-Boot configuration to the NVRAM.

For this I simply break the U-Boot boot during the IDE device detection and switch to the serial output then. This way I can follow the SLB and U-Boot messages on the serial console:

Quote:

... VGA: 0 .. In: serial Out: serial Err: serial #### Here I press the escape key and enter "setenv stdout serial" on the U-Boot prompt. ] boota #### Typed on the AmigaOne, but displayed on the terminal program First-level bootloader: entered main First-level bootloader: got 2 valid boot sources AmigaOS4 first-level bootloader. Scanning HDD ST380011A 5JVZ2SB3 3.06... Entered find_sec_bl_start Reading sector 0 Reading sector 1 Found an HD FLB: found something FLB: SLB of length 27828; loading found AmigaOS4 second level bootloader. FLB: SLB loaded; now launching it Loading .text @ 0x014000a0 (22464 bytes) Loading .rodata @ 0x01405860 (2488 bytes) Loading .eh_frame @ 0x01406218 (52 bytes) Loading .got2 @ 0x0141624c (724 bytes) Loading .data @ 0x01416520 (208 bytes) Loading .sdata @ 0x014165f0 (16 bytes) Clearing .sbss @ 0x01416600 (168 bytes) Clearing .bss @ 0x014166a8 (3928 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. Booting configuration DebianLenny-2.6.18 ## Booting image at 00500000 ... Image Name: Linux-2.6.18-6-book3s-amigaone Created: 2013-04-24 21:24:33 UTC Image Type: PowerPC Linux Kernel Image (gzip compressed) Data Size: 1740668 Bytes = 1.7 MB Load Address: 00000000 Entry Point: 00000000 Verifying Checksum ... OK Uncompressing Kernel Image ... OK \0xff

Note that boota here automatically boots the default configuration, which is DebianLenny-2.6.18 in my case. You could temporarily change this before switching to the serial output.


Quote:

damned wrote: i opened the case and take away the cmos battery for at least 30sec but when i reassembled all the things and powered up the situation was same as before! ouch!! so i go to uboot menu and a restore factory default saved my day....fiuuuuu!!

There should always a way to recover. But yes, it's bothersome, if all settings are lost.

Quote:

but i remain so sad that i can't use new kernels..

I still have no explanation for this behavior. The only possibility I can think of at the moment would be a problem with the memory. I had problems with the memory modules on my A1SE, where Linux refused to boot, but OS4 was working...

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@Geri

it's the first thing i had done..but after i do the 'setenv stdout serial' command i could only see the cursor blink, even if i write boota nothing happen..boh!!

_________________
--------------------------------------
Search and Destroy
---------------------------------------

@damned

Which terminal program are you using to capture the output on the serial port?

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -

@damned

Get my PM?

@Hypex

Testing.

ADSL2 upload failing.

Last edited by Hypex on 16-May-2013 at 03:35 PM.

@damned

Quote:

it's the first thing i had done..but after i do the 'setenv stdout serial' command i could only see the cursor blink, even if i write boota nothing happen..boh!!

I'm sure you considered this already....but just to be sure...

1) Is the baud rate of your terminal program set to match that of your bootenv baudrate variable?

2) Are you using a proper cable (i.e. doesn;t this require a null modem cable?)

Okay...obvious things...just woundered if you might have missed something.

Scott

_________________
AmigaOne uA1-c 512M RAM - Only Amiga Makes It Possible!
Check my blog AmigaOne Computing

@damned

You should be able enter the menu option in the serial terminal corresponding to the one you want to boot. So cursor doesn't work but numbers should in serial terminal. Have you tried this?

What are you using for a serial terminal? Looks like your settings work but to just to make sure, port should be set to 115,200, 8N1.

ok..now it's works..the problem was that i had set the correct serial's parameter in putty (the program that i use to view the log) but not on the serial itselfs in the windows device manager
also i can only choose the boot selection with the number, the arrows don't work..so, here is the log with the latest kernel 3.2

] boota

First-level bootloader: entered main
First-level bootloader: got 3 valid boot sources
AmigaOS4 first-level bootloader.
Scanning HDD Corsair Nova 2 SSD 12176503000013450024 1.3.4... Entered find_sec_bl_start
Reading sector 0
Reading sector 1
Found an HD
FLB: found something
FLB: SLB of length 29036; loading
found AmigaOS4 second level bootloader.
FLB: SLB loaded; now launching it
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 bytes)
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.
Possible selections:
1 -> AmigaOS 4.x: SH0 AmigaOS 4.1 Update 6
2 -> Linux: Linux
3 -> Linux: Linux4
4 -> Linux: LiTest
5 -> Linux: Debian-6.0-2.6.32-NOSDA
6 -> Linux: Debian-3.2
7 -> Linux: Li2631Test
8 -> Linux: Lisda5Test
9 -> Linux: Lisda6Test
10 -> Linux: Debian-6.0-2.6.32
11 -> Linux: Debian-6.0-2.6.32-recovery
12 -> Linux: Debian-3.2
Choice ? 6
Booting configuration Debian-3.2
## Booting image at 00500000 ...
Image Name: Debian GNU/Linux 6.0 Squeeze
Created: 2013-04-02 23:39:15 UTC
Image Type: PowerPC Linux Multi-File Image (uncompressed)
Data Size: 8379335 Bytes = 8 MB
Load Address: 04000000
Entry Point: 04000454
Contents:
Image 0: 5787232 Bytes = 5.5 MB
Image 1: 2592091 Bytes = 2.5 MB
Verifying Checksum ... OK
OK
Loading Ramdisk to 0eced000, end 0ef65d5b ... OK
Bad trap at PC: 4000454, SR: 2030, vector=300
NIP: 04000454 (F3F98454 translated) XER: 20000000 LR: 0FFA1CA0 REGS: 0ef66a78 TRAP: 0300 DAR: 0FFAC800
MSR: 00002030 EE: 0 PR: 0 FP: 1 ME: 0 IR/DR: 11

GPR00: 0FFA1C70 0EF66B68 00000400 10000000 0ECED000 0EF65D5B 007FFC00 007FFC5D
GPR08: 1F8B0800 FE000080 FFFFFFFF 0EF66620 00000000 00000000 0FFD5900 04000454
GPR16: 00000001 FFFFFFFF 00000000 007FFC00 00000000 0EF66C68 0FFFF220 007FFBC0
GPR24: 00000001 00800000 00A84EAC 00278D5B 0ECED000 0EF66F88 0FFD71E4 0EF65D5B
Call backtrace:
0FFA1C70 (FFF39C70 translated) 0140571C (F139D71C translated) 01403068 (F139B068 translated) 0FF72348 (FFF0A348 translated) 0FFAE9E8 (FFF469E8 translated) 0FFAF064 (FFF47064 translated) 0FFAF1CC (FFF471CC translated)
0FF9FEE4 (FFF37EE4 translated) 0FF974C0 (FFF2F4C0 translated) 0FF6A6C8 (FFF026C8 translated) 0FF6A4EC (FFF024EC translated)
Exception in kernel pc 4000454 signal 0

_________________
--------------------------------------
Search and Destroy
---------------------------------------

and this with 2.6.32
http://dfiles.eu/files/yyrv34uus

_________________
--------------------------------------
Search and Destroy
---------------------------------------

@damned

Hmm, the 3.2 kernel seems to crash directly on its entry point, if I interpret the debug output correctly. Dunno what's going on here. I'll upload a complete boot image as soon as possible, which also works on my A1. Maybe you can try out this one then to rule out any image generation problems.

The 2.6.32 kernel actually boots fine, but for some reason can't find the root partition. Is the root partition really labeled as "squeeze"? Please try it with the real device node for now (e.g. /dev/sdxy).

Thanks for testing!

_________________
A1SE: G3@600MHz, 2GB, 1GBit network card
A1XE: G4@933MHz, 2GB, refitted AC'97 codec
microA1: G3@800MHz, 1GB

- A1 Linux support -