Click Here
home features news forums classifieds faqs links search
6071 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
32 crawler(s) on-line.
 108 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 OlafS25:  48 mins ago
 Bruce72:  50 mins ago
 Rob:  1 hr 22 mins ago
 MEGA_RJ_MICAL:  1 hr 38 mins ago
 t0lkien:  1 hr 43 mins ago
 amigakit:  2 hrs 3 mins ago
 OneTimer1:  2 hrs 17 mins ago
 Troels:  3 hrs 18 mins ago
 Gunnar:  3 hrs 33 mins ago
 zipper:  4 hrs 2 mins ago

/  Forum Index
   /  Amiga PPC Linux
      /  First Debian Etch Installer CD release!!
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 Next Page )
PosterThread
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 3-Dec-2012 22:33:39
#181 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@Hypex

Quote:
Hypex wrote:
@Geri

Quote:
The right way would be to add this functionality also to U-Boot.


That's interesting. the closest we have is the bootloader. But a delay caused by an empty controller isn't a big deal if it gets past it.

What's more important is being able to set DMA. Needed for install as well as installed system.

Does the ub2lb bootloader have direct access to U-Boot variables (because it is executed within the U-Boot environment)? Then it could translate a specific U-Boot variable to (a) corresponding kernel module option(s).

The Debian installer then automatically stores kernel module options during the installation in /etc/modprobe.d/local*.conf, from where they go into the initramfs. On an installed system, the user simply edits the files in /etc/modprobe.d and regenerates the initramfs.

Or is there still an use case, where it would be necessary to change the settings for an installed system by an U-Boot variable?

Quote:
Quote:
Maybe I can hack something into the platform code to clear the progif register before the driver is loaded.


Possible but setting DMA is easier?

I think the DMA settings are also read from a PCI register in the IDE controller, which is usually setup by the BIOS according to the IDE cable configuration (40/80 pin - if there is a way to autodetect it, another feature that the A1 U-Boot lacks).

On the other side it would still be necessary to specify a kernel option so that the A1 platform code knows, if it should change the IDE configuration. So this would just reduce the amount of kernel module options that need to be specified, but the main problem persists.

Quote:
Fresh install would need it. But it also needs to install without errors!

Well, I would say it is needed at least for the installation to avoid a chicken/egg problem.

Quote:
Quote:
The NVRAM drivers usually only provide a block interface for userspace access, not for the kernel.


Okay if we can read it and deal with it on time.

That's the problem. With a standard NVRAM driver reading and parsing of U-Boot variables would have to be done in userspace, before any drivers is loaded (and it wouldn't work at all for a kernel with compiled in modules)! IMHO that would require a lot of modifications to the installer's initramfs.

Otherwise we have to replicate a lot of code in the kernel (which certainly won't find its way into the official kernel source), where the bootloader probably can simply call an U-Boot function to retrieve the value of an environment variable.

Quote:
Sure. Because my fix is a solderless one the capacitors stay on!

It starts out:
[ 38.781915] hub 3-0:1.0: USB hub found
[ 38.808774] hub 3-0:1.0: 2 ports detected
[ 38.810668] uhci_hcd 0000:00:09.0: calling quirk_via_vlink+0x0/0xc4
[ 38.810712] uhci_hcd 0000:00:09.0: UHCI Host Controller
[ 38.836480] uhci_hcd 0000:00:09.0: new USB bus registered, assigned bus number 4
[ 38.837747] uhci_hcd 0000:00:09.0: irq 10, io base 0x00802440
[ 38.847353] usb usb4: New USB device found, idVendor=1d6b, idProduct=0001
[ 38.848454] usb usb4: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[ 38.849446] usb usb4: Product: UHCI Host Controller
[ 38.850393] usb usb4: Manufacturer: Linux 2.6.38.3+ uhci_hcd
[ 38.851396] usb usb4: SerialNumber: 0000:00:09.0

If usb4 corresponds to hub 3. Log doesn't make it clear to me. But this kind of message constantly appears every second or so as it boots:

[ 41.762992] hub 3-0:1.0: unable to enumerate USB device on port 2

Okay, I have to see, if there is a way to debug this. I'm not an expert on the USB subsystem.

Quote:
Then I'll guess I'll never know what OS4 is doing wrong.

I'm afraid you're right.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 9-Dec-2012 16:53:15
#182 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Not sure if you ever answered my question about this. There is a specific patch for the USB2 EHCI issues written for kernel 2.6.18 on sourceforge with this comment:

( 29/06/2008 ) Patch written by Marcus Comstedt. Makes the USB EHCI driver work on non cache coherent PPC 74xx systems

Was this particular patch applied to the kernel you made for me with the SATA drivers? If not, is there a way for me to apply this patch and rebuild the kernel?

Thanks!

Scott

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

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 9-Dec-2012 19:27:05
#183 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Quote:
scabit wrote:
@Geri

Not sure if you ever answered my question about this. There is a specific patch for the USB2 EHCI issues written for kernel 2.6.18 on sourceforge with this comment:

( 29/06/2008 ) Patch written by Marcus Comstedt. Makes the USB EHCI driver work on non cache coherent PPC 74xx systems

Was this particular patch applied to the kernel you made for me with the SATA drivers? If not, is there a way for me to apply this patch and rebuild the kernel?

It was applied! Otherwise the kernel would "panic", but not lockup completely.

BTW: Could you answer these questions? Maybe can hack something into U-Boot to get your VIA SATA controller card running. No promises though.

Last edited by Geri on 09-Dec-2012 at 07:47 PM.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 10-Dec-2012 2:00:41
#184 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

proc/iomem shows this:

80000000-fcffffff : PCI host bridge
80000000-87ffffff : PCI Bus #01
80000000-87ffffff : 0000:01:00.0
80000000-87ffffff : radeonfb framebuffer
88000000-880fffff : PCI Bus #01
88000000-8800ffff : 0000:01:00.0
88000000-8800ffff : radeonfb mmio
88020000-8803ffff : 0000:01:00.0
88100000-8810007f : 0000:00:06.0
88100100-881001ff : 0000:00:08.2
88100100-881001ff : ehci_hcd

proc/ioports shows this:

00000000-00bfffff : PCI host bridge
00000000-0000001f : dma1
00000020-00000021 : 8259 (master)
00000040-0000005f : timer
00000060-0000006f : i8042
00000080-0000008f : dma page reg
000000a0-000000a1 : 8259 (slave)
000000c0-000000df : dma2
000001f0-000001f7 : ide0
000002f8-000002ff : serial
00000330-00000333 : 0000:00:07.5
00000378-0000037a : parport0
0000037b-0000037f : parport0
000003c0-000003df : vga+
000003e8-000003ef : serial
000003f6-000003f6 : ide0
000003f8-000003ff : serial
000004d0-000004d1 : 8259 edge control
00002000-00002fff : PCI Bus #01
00002000-000020ff : 0000:01:00.0
00005000-0000500f : 0000:00:07.4
00005000-00005007 : vt596_smbus
00006000-0000607f : 0000:00:07.4
00006000-0000607f : via686a
0000cc00-0000cc0f : 0000:00:07.1
0000cc00-0000cc07 : ide0
0000cc08-0000cc0f : ide1
0000dc00-0000dcff : 0000:00:07.5
0000e000-0000e003 : 0000:00:07.5
0000e400-0000e4ff : 0000:00:07.6
0000ff00-0000ffff : 0000:00:05.0
0000ff00-0000ffff : CMI8738-MC6
00802000-0080207f : 0000:00:06.0
00802080-0080209f : 0000:00:07.2
00802080-0080209f : uhci_hcd
008020a0-008020bf : 0000:00:07.3
008020a0-008020bf : uhci_hcd
00802400-0080241f : 0000:00:08.0
00802400-0080241f : uhci_hcd
00802420-0080243f : 0000:00:08.1
00802420-0080243f : uhci_hcd
00802440-0080244f : 0000:00:08.3
00802440-0080244f : sata_via
00802450-0080245f : 0000:00:08.3
00802450-0080245f : sata_via
00802460-0080246f : 0000:00:08.3
00802460-0080246f : sata_via
00802470-0080247f : 0000:00:08.3
00802470-0080247f : sata_via
00802480-0080249f : 0000:00:08.3
00802480-0080249f : sata_via
00802500-008025ff : 0000:00:08.3
00802500-008025ff : sata_via
fc002000-fc00207f : 0000:00:06.0


Does that make any sense?

Thanks!

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

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 10-Dec-2012 6:06:31
#185 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Yes, that makes sense. I'll try to come up with an U-Boot script until the weekend to change some of the PCI addresses.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 10-Dec-2012 14:54:05
#186 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Quote:
I'll try to come up with an U-Boot script until the weekend to change some of the PCI addresses.


Thank you! I really appreciate your help with this!

Scott

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

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 12-Dec-2012 1:15:33
#187 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri
@hypex

A question about Mac On Linux....

I have the two files, mol_0.9.72-2_powerpc.deb and mol-kmods_0.9.72-2_powerpc.deb. Will these work with the Debian kernel you provided for me that has the SATA driver build in?
I assume that to install these I just use:

dpkg --install mol-kmods_0.9.72-2_powerpc.deb
dpkg --install mol_0.9.72-2_powerpc.deb

Then what? I have looked at the MOL web site and cannot find instructions on how to actually get MOL up and running. I assume I need a hard drive partition that has Mac OSX on it that I can add to the linux configuration somehow....

Thanks,

Scott

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

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 16-Dec-2012 22:37:34
#188 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Quote:
scabit wrote:
@Geri
@hypex

A question about Mac On Linux....

I have the two files, mol_0.9.72-2_powerpc.deb and mol-kmods_0.9.72-2_powerpc.deb. Will these work with the Debian kernel you provided for me that has the SATA driver build in?

They should still work. But simply install the packages and run "startmol --test" (IIRC). That should tell you, if the kernel modules are loaded correctly.

Quote:
I assume that to install these I just use:

dpkg --install mol-kmods_0.9.72-2_powerpc.deb
dpkg --install mol_0.9.72-2_powerpc.deb

Right!

Quote:
Then what? I have looked at the MOL web site and cannot find instructions on how to actually get MOL up and running. I assume I need a hard drive partition that has Mac OSX on it that I can add to the linux configuration somehow....

Yes, the easiest way is to use a dedicated harddisk or partition for OSX. Here's my OSX MoL config:

Quote:
# -*- perl -*- (for emacs syntax coloring)
#
# /etc/mol/molrc.osx
#
# Mac-on-Linux master configuration file for MacOS X booting
#


#------------------------------------------------------------------------------
# Parse other config files (these files are in /etc/mol/)
#------------------------------------------------------------------------------

include ${etc}/molrc.video # video configuration
include ${etc}/molrc.input # mouse and keyboard
include ${etc}/molrc.net # networking
include ${etc}/molrc.sound # sound


#------------------------------------------------------------------------------
# Misc Settings
#------------------------------------------------------------------------------

ram_size: 256
disable_altivec: yes


#------------------------------------------------------------------------------
# OS X Volumes
#------------------------------------------------------------------------------
#
# Volumes to be mounted in MacOS X are specified through the
# blkdev keyword:
#
# blkdev: [flags]
#
# where can be a complete disk (/dev/hda), a single
# partition (/dev/hda7) or a disk image.
#
# The following flags are supported
#
# -rw/-ro read-write / read-only [default]
# -force force usage of device (be careful!)
# -boot boot from this disk
# -boot1 boot from this disk (ignore other -boot flags)
# -whole export the entire device (including
# any non-HFS partitions). BE CAREFUL!
# -cd CDROM/DVD
#
# MOL will boot from CD if it invoked through 'startmol -X --cdboot'.

blkdev: /dev/cdrom1 -cd
blkdev: /dev/cdrom -cd ${cdboot}

ifempty ${altconfig} {
# default configuration
blkdev: /dev/hda14 -rw -force
#blkdev: /dev/hda -rw
#blkdev: /dev/hdb -rw
#blkdev: /dev/sda -rw
#blkdev: /dev/sdb -rw

} else {
# alternate configuration
blkdev: /tmp/osx.image -rw
#blkdev: /dev/loop0 -rw
}


#------------------------------------------------------------------------------
# SCSI configuration (also relevant for IDE CDs and other SCSI-like devices)
#------------------------------------------------------------------------------
#
# By default, MOL automatically scans for non-harddisk type SCSI devices.
# If only specified SCSI devices are to be used by mol, change the
# the autoprobe setting below.
#
# Note: IDE, firewire and USB devices typically show up as SCSI devices.


autoprobe_scsi: yes

# If SCSI autoprobing is disabled then devices must be exported manually

#scsi_dev: "0:0:0" # host/channel/id
#scsi_dev: "1:0:0" #


#------------------------------------------------------------------------------
# USB Configuration
#------------------------------------------------------------------------------
#
# Generic USB support. Devices that are _not_ claimed by specific
# USB drivers are seen by MOL. Note: Make sure the kernel is compiled
# with 'Preliminary USB device filesystem' support. Also, for best effect,
# turn off drivers for specific USB devices (like 'USB Scanner support').

enable_usb: yes

#------------------------------------------------------------------------------
# PCIproxy Configuration
#------------------------------------------------------------------------------
#
# MOL can be configured to make PCI devices of the host machine available to the
# guest OS. You will have to make sure the linux kernel isn't using the device
# itself by disabling/unloading the driver (if it exists). Devices to be proxied
# have an entry here that gives the pci address as first argument and a an
# optional second argument that is a filename containing a template for the
# device tree node of the device to be proxied (MOL format, get a template by
# running the ofdump utility included in the MOL package). This template can be
# used to specify address mappings by hand, setting additional properties etc.

# enable the pciproxy module
#enable_pciproxy: yes

# Airport Extreme (Broadcom 4320) Card (Second is with optional ofdump)
#pci_proxy_device: 0001:10:12.0
#pci_proxy_device: 0001:10:12.0 ${etc}/pci80211.dnode

You see that I use partition /dev/hda14 for OSX. Just create a partition with MediaToolBox that is not used by OS4 (for example another Linux partition). There is no need to format it, as OSX will see it as a (virtual) harddisk. Thus you will create other partitions within this partition, when you install OSX.

/dev/cdrom is used with the "--cdboot" option, which you have to use when you're going to install OSX. It may also be a good idea to disable MoL's USB support for your A1 (set "enable_usb" to "no").

Here's also my video configuration file (molrc.video):

Quote:

# -*- perl -*- (for emacs syntax coloring)
#
# Mac-on-Linux video configuration
#


# ----------------------------------------------------------------------
# Startup Resolution (this is used as a hint only)
# ----------------------------------------------------------------------
#

resolution: 1600/1200/60 # width/height/Hz
depth: 32


# ----------------------------------------------------------------------
# X11 Settings
# ----------------------------------------------------------------------
#
# Some config options:
#
# xdisplay: remote.host:0.0
# use_backing_store: yes/no
#
# Backing store is primarily useful if the X-connection is remote.
# Note: X11-video is not as fast as full-screen video.

enable_xvideo: yes


# ----------------------------------------------------------------------
# Fullscreen (Console/fbdev) Settings
# ----------------------------------------------------------------------
#
# The tool 'molvconfig' is used to configure the fullscreen mode.
#
# Some config options:
#
# vt: 9 # VT-number to use
# gamma: 1.0 # gamma (parameter between 0.1 and 1.0)
# gamma: 1.0 1.0 1.0 # gamma (r,g,b)
#
# Ctrl-Cmd-Space switches back to the orignal VT

enable_console_video: yes

start_on_console: no # switch to console initially
autoswitch_console: yes # allow automatic switching to console

#fb_modes: /etc/fb.modes # extra video modes for molvconfig


# ----------------------------------------------------------------------
# XDGA (X11 fullscreen) Settings
# ----------------------------------------------------------------------
#
# At startup, all available XDGA modes are listed. Unfortunately,
# some of the modes might be broken (depending on the video card
# and the XFree86 version).
#
# By default, MOL uses just a single mode with high
# probability of working. However, it is possible to manually
# specify multiple modes. For instance
#
# xdga_modes: 3 5 9
#
# instructs MOL to use mode 3, 5 and 9.
#
# Cntrl-Cmd-Space leaves XDGA mode and switches back to X

enable_xdga_video: no


# ----------------------------------------------------------------------
# VNC Settings
# ----------------------------------------------------------------------
#
# For VNC information, visit http://www.uk.research.att.com/vnc/
#

enable_vncvideo: no # Enable the VNC server
vnc_port: 5900 # port to use for VNC

You may want to change the resolution for your monitor. I only use fullscreen video, as this should be faster than running MoL in a window on the Linux desktop.

I think I left the other config files unchanged. But let me know, if you have problems with sound or something else.

I hope that helps a little bit.

PS: Unfortunately I didn't have time yet to look at the workaround for the VIA SATA controller. I hope I have more time in the next days.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
Hypex 
Re: First Debian Etch Installer CD release!!
Posted on 17-Dec-2012 14:17:50
#189 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@scabit

Quote:
Will these work with the Debian kernel you provided for me that has the SATA driver build in?


MOL can be very picky about kernels. I remember I had this annoying error about not syncing on the HD or something. And ended up recompiling it.

But if it is made for kernel 2.6.18 it won't work on kernel 2.6.31. It needs to be compiled for the kernel you are using. I don't know what kernel your MOL was built for but you will see error straight away if problem.

Quote:
Then what?


Unfortunately it has no GUI. So as Geri told you need to use a CLI and run with with a startmol command. Use "startmol -X" to boot OSX or "startmol -X --cdboot" to boot OSX CD. From top of my head.

There is also PegXMac which installs a MOL booter onto a partition or two using a custom Linux installer.

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 17-Dec-2012 15:08:12
#190 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Okay, installed the two MOL .deb files, then ran startmol --test. It asked me to run molvconfig, which I did and it seemed to work okay.
Running startmol --test again gives me this:

debian:/etc/mol# startmol --test
Mac-on-Linux 0.9.72 [Jän 25 2008 22:55]
Copyright (C) 1997-2004 Samuel Rydh
Starting MOL session 0
Loading Mac-on-Linux kernel module:
/usr/lib/mol/0.9.72/modules/2.6.18/mol.ko
insmod: error inserting '/usr/lib/mol/0.9.72/modules/2.6.18/mol.ko': -1 Invalid module format
====================================================================
Failed to load the module - try recompiling the MOL kernel
module. Instructions (and information about common problems)
are available at .
====================================================================

So I assume I will need to recompile the MOL kernel modules somehow? I edited the /etc/molrc.osx file - it defaulted to altivec yes, usb yes, and for some reason RAM 96.

Out of curiosity, what version of OSX was the latest one that MOL supports for PowerPC processors? I see comments about supporting Darwin in the MOL files.....

Thanks!

Scott

Last edited by scabit on 17-Dec-2012 at 03:13 PM.

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

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 17-Dec-2012 16:24:51
#191 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Hypex

So if I need to rebuild the MOL kernel .ko file, can I just download the MOL source from sourceforge and run make install?
Interestingly, the one available on sourceforge is 9.72-1, where I am using the ones I got from the AmigaOne page which are 9.72-2.
Are there any specific source code changes that need o be made to build the MOL.ko to make it work on the micro A1? I know nothing about building linux kernel modules...although I want to learn.

Thanks

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

 Status: Offline
Profile     Report this post  
Hypex 
Re: First Debian Etch Installer CD release!!
Posted on 18-Dec-2012 13:33:26
#192 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@scabit

Quote:
/usr/lib/mol/0.9.72/modules/2.6.18/mol.ko


There's your clue. 2.6.18.

Quote:
So if I need to rebuild the MOL kernel .ko file, can I just download the MOL source from sourceforge and run make install?


You should be able too. I've done it on Ubuntu. But last I checked Debian didn't have all the deveoper tools needed. So if it fails you may need to install gcc, make amd possibly some header files. apt-get install gcc is a good start.

Quote:
Are there any specific source code changes that need o be made to build the MOL.ko to make it work on the micro A1?


I don't know why we have a 0.9.72-2. But Geri made some changes 5 years ago:

2007-01-26 Gerhard Pircher
* Remove cache coherency requirement for AmigaOne support

AFAIK there are no changes needed for the Micro. But you'll find out when you test. The BUILDING doc has info on how to compile. Good luck!

Oh BTW, I have run 10.4 Tiger under MOL. I borrowed 10.5 from a friend but it is like some hybrid x86/PPC DVD and didn't work IIRC. It has some Windows partition that gets in the way.

Last edited by Hypex on 18-Dec-2012 at 01:36 PM.

 Status: Offline
Profile     Report this post  
Hypex 
Re: First Debian Etch Installer CD release!!
Posted on 18-Dec-2012 14:05:54
#193 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@Geri

Quote:
Does the ub2lb bootloader have direct access to U-Boot variables


Yes it does. It can read and write.

Quote:
The Debian installer then automatically stores kernel module options during the installation


That's good. We can also read them elsewhere in some kernel command line variable.

Quote:
Or is there still an use case, where it would be necessary to change the settings for an installed system by an U-Boot variable?


Possibly for ignoring onboard IDE or turning off DMA. Or even cache.

Quote:
I think the DMA settings are also read from a PCI register in the IDE controller,


Okay I see.. So, really UBoot doesn't check cable type?

Quote:
So this would just reduce the amount of kernel module options that need to be specified, but the main problem persists.


Yes but we are limited what we can do. And more so with kennels that don't accept settings for external modules.

Quote:
Well, I would say it is needed at least for the installation to avoid a chicken/egg problem.


Yes it is. What comes first, the boot or the arguments?

Quote:
With a standard NVRAM driver reading and parsing of U-Boot variables would have to be done in userspace


In that case we can read them before hand and pass them as a string which can be parsed later. Possibly by a module.

Quote:
U-Boot function to retrieve the value of an environment variable.


Yes that can be done and looks like the best move.

Quote:
I'm not an expert on the USB subsystem.


The EHCI moduile is still being loaded last and I still see complaints.

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 27-Dec-2012 18:35:43
#194 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Quote:

scabit wrote:
@Geri

Quote:
I'll try to come up with an U-Boot script until the weekend to change some of the PCI addresses.


Sorry, better late than never.

Okay let's see, if we can convince the SATA controller to work. I still think that the controller may only work within the 64k I/O address space (this was also the problem with the onboard sound). Thus we're going to move the I/O base address registers (BARs) to a free I/O window with a base address below 0x10000. The addresses below 0x1000 are reserved for the southbridge, 0x2000 is for the graphics card, 0x4000 (it doesn't show up in /proc/ioports, but I know it is reserved ), 0x5000, 0x6000 and 0xc000 to 0x10000 are already assigned to other PCI devices. So address 0x8000 should be free and we can try to relocate the BARs with some U-Boot commands (without the "# ..." comments):

Quote:
] pci write.l 00.08.03 10 00008000 # BAR 0
] pci write.l 00.08.03 14 00008020 # BAR 1
] pci write.l 00.08.03 18 00008040 # BAR 2
] pci write.l 00.08.03 1c 00008060 # BAR 3
] pci write.l 00.08.03 20 00008080 # BAR 4
] pci write.l 00.08.03 24 00008200 # BAR 5


You can save these commands as an U-Boot variable, so that you don't have to reenter everything on bootup:

Quote:
] setenv via_sata_quirk "pci write.l 00.08.03 10 00008000; pci write.l 00.08.03 14 00008020; pci write.l 00.08.03 18 00008040; pci write.l 00.08.03 1c 00008060; pci write.l 00.08.03 20 00008080; pci write.l 00.08.03 24 00008200"

Note that the double quotes and semicolons are important!

If you put this variable in front of the bootcmd variable, then the pci commands are executed on every boot:

Quote:
] setenv bootcmd "run via_sata_quirk; ...old content of your bootcmd variable..."


A final "saveenv" command will permanently store this setup. But note that OS4's U-Boot preferences may overwrite it! Also note that the pci write commands have to be adapted, if you move the VIA card to another slot, because the 00.08.03 number will change!

But let's see first, if the commands actually have an effect. Please post the kernel log and the content of /proc/ioport. If it doesn't work out, then we have to come up with another theory.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 27-Dec-2012 18:46:53
#195 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Quote:
scabit wrote:

So I assume I will need to recompile the MOL kernel modules somehow?

Yes, the MoL kernel module needs to be recompiled.

Quote:
I edited the /etc/molrc.osx file - it defaulted to altivec yes, usb yes, and for some reason RAM 96.

Make sure "disable_altivec" is set to "yes"!

Quote:
Out of curiosity, what version of OSX was the latest one that MOL supports for PowerPC processors? I see comments about supporting Darwin in the MOL files.....

I think I had MacOSX Tiger running here. AFAIK Darwin is just the name of the MacOSX kernel.

Quote:
Hypex wrote:
@scabit

Quote:
/usr/lib/mol/0.9.72/modules/2.6.18/mol.ko


There's your clue. 2.6.18.

Quote:
So if I need to rebuild the MOL kernel .ko file, can I just download the MOL source from sourceforge and run make install?


You should be able too. I've done it on Ubuntu. But last I checked Debian didn't have all the deveoper tools needed. So if it fails you may need to install gcc, make amd possibly some header files. apt-get install gcc is a good start.

IIRC I used a previous version of MoL, which at least compiles on Debian Etch. But I would have to check, which version it was. Unfortunately I don't have my full Linux environment available on my A1 at the moment so I can't take a look.

Quote:
Quote:
Are there any specific source code changes that need o be made to build the MOL.ko to make it work on the micro A1?


I don't know why we have a 0.9.72-2. But Geri made some changes 5 years ago:

2007-01-26 Gerhard Pircher
* Remove cache coherency requirement for AmigaOne support

I can't remember right now, if I had to patch the MoL version I used or if this patch was already integrated back then. I think the latter one is the case.

Quote:
AFAIK there are no changes needed for the Micro.

Yes, there are no specific changes required for the Micro.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 27-Dec-2012 19:02:53
#196 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@Hypex

Quote:
Hypex wrote:
@Geri

Quote:
The Debian installer then automatically stores kernel module options during the installation.


That's good. We can also read them elsewhere in some kernel command line variable.

Okay. You mean /proc/cmdline of the U-Boot bootargs variable?

Quote:
Or even cache.

True, the cache setting (l2cr=) would make sense.

Quote:
Quote:
I think the DMA settings are also read from a PCI register in the IDE controller,


Okay I see.. So, really UBoot doesn't check cable type?

Right. But it is questionable, if the hardware even supports cable detection.

Quote:
Quote:
Well, I would say it is needed at least for the installation to avoid a chicken/egg problem.


Yes it is. What comes first, the boot or the arguments?

The "boot arguments"!

Quote:
Quote:
With a standard NVRAM driver reading and parsing of U-Boot variables would have to be done in userspace


In that case we can read them before hand and pass them as a string which can be parsed later. Possibly by a module.

In this case we would have to make sure that all the other modules are loaded after the U-Boot variables have been read and processed. Which means we have to do more changes to the initial boot process of the debian-installer.

Quote:
The EHCI moduile is still being loaded last and I still see complaints.

IIRC this was an udev specific problem. I still don't know, why it happens for the 2.6.31 kernel. On a normal installed system we could work around it by only putting the ehci_hcd module in the initramfs, which would defer the loading of the ohci/uhci_hcd modules until the root filesystem becomes available.

Last edited by Geri on 27-Dec-2012 at 07:05 PM.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 28-Dec-2012 14:46:45
#197 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Quote:
Sorry, better late than never.


No problem! I appreciate your taking the time no matter how long it takes for ou to get around to it! I was busy having fun building Opus Magellan this past week anyway.

Here is what I did....I plugged in my China combo card - note the micro A1 has only one slot - no chance I will be able to plug things into another slot! Anyway, I atached an 80 Gig IDE drive to the PATA interface of the RAID/SATA board. I added a command VSATA to my uboot vars as you indicated - being sure of exact punctuation. I used setenv bootcmd "run VSATA; menu; run menuboot_cmd" then did a saveenv. Reboot into linux - Debian.
Here is my proc/iomem:

0000000-fcffffff : PCI host bridge
80000000-87ffffff : PCI Bus #01
80000000-87ffffff : 0000:01:00.0
80000000-87ffffff : radeonfb framebuffer
88000000-880fffff : PCI Bus #01
88000000-8800ffff : 0000:01:00.0
88000000-8800ffff : radeonfb mmio
88020000-8803ffff : 0000:01:00.0
88100000-8810007f : 0000:00:06.0
88100100-881001ff : 0000:00:08.2
88100100-881001ff : ehci_hcd

Here is my proc/ioports:

00000000-00bfffff : PCI host bridge
00000000-0000001f : dma1
00000020-00000021 : 8259 (master)
00000040-0000005f : timer
00000060-0000006f : i8042
00000080-0000008f : dma page reg
000000a0-000000a1 : 8259 (slave)
000000c0-000000df : dma2
000001f0-000001f7 : ide0
000002f8-000002ff : serial
00000330-00000333 : 0000:00:07.5
00000378-0000037a : parport0
0000037b-0000037f : parport0
000003c0-000003df : vga+
000003e8-000003ef : serial
000003f6-000003f6 : ide0
000003f8-000003ff : serial
000004d0-000004d1 : 8259 edge control
00002000-00002fff : PCI Bus #01
00002000-000020ff : 0000:01:00.0
00005000-0000500f : 0000:00:07.4
00005000-00005007 : vt596_smbus
00006000-0000607f : 0000:00:07.4
00006000-0000607f : via686a
00008000-0000800f : 0000:00:08.3
00008000-0000800f : sata_via
00008020-0000802f : 0000:00:08.3
00008020-0000802f : sata_via
00008040-0000804f : 0000:00:08.3
00008040-0000804f : sata_via
00008060-0000806f : 0000:00:08.3
00008060-0000806f : sata_via
00008080-0000809f : 0000:00:08.3
00008080-0000809f : sata_via
00008200-000082ff : 0000:00:08.3
00008200-000082ff : sata_via
0000cc00-0000cc0f : 0000:00:07.1
0000cc00-0000cc07 : ide0
0000cc08-0000cc0f : ide1
0000dc00-0000dcff : 0000:00:07.5
0000e000-0000e003 : 0000:00:07.5
0000e400-0000e4ff : 0000:00:07.6
0000ff00-0000ffff : 0000:00:05.0
0000ff00-0000ffff : CMI8738-MC6
00802000-0080207f : 0000:00:06.0
00802080-0080209f : 0000:00:07.2
00802080-0080209f : uhci_hcd
008020a0-008020bf : 0000:00:07.3
008020a0-008020bf : uhci_hcd
00802400-0080241f : 0000:00:08.0
00802400-0080241f : uhci_hcd
00802420-0080243f : 0000:00:08.1
00802420-0080243f : uhci_hcd
fc002000-fc00207f : 0000:00:06.0

I am about to unplug the IDE drive from the PATA port of the combo card and plug in the IDE to SATA converter to hook the drive to the SATA bus. Right now, when I run gparted it only sees my hda drive and nothing else that is new.
Does the SATA kernel support the PATA part of the combo IDE/SATA VIA chip, or only the SATA part?

I'll reconnect to the SATA port and see what I find.

Thanks!

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

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 28-Dec-2012 15:01:47
#198 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Hooray! It works. I connected my drive to the SATA port of my multicard and boot into Debian. Using gparted I now see my 80 Gig SATA drive (sda) as well as my 160Gig IDE drive (hda).
Now I can hopefully select a partition to format as EXT3 and backup my entire Debian partition to it .... is there any special trick to that...that is, making a "clone" of a Debian partition on another drive?
Anyway, now I just need to format a partition and see if I can mount it and access it....

Thanks,

Scott

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

 Status: Offline
Profile     Report this post  
scabit 
Re: First Debian Etch Installer CD release!!
Posted on 28-Dec-2012 15:39:35
#199 ]
Super Member
Joined: 8-Jan-2005
Posts: 1667
From: Satellite Beach, FL USA

@Geri

Okay, your help in getting the VIA SATA controller to work in Debian on my multicard now opens up a lot of questions.

1) The memory addressing area of the multicard being too high seems to have caused the problems with the SATA I/O. Is the same true for the USB2 I/O? That is, if I were to add: pci write.l 00.08.02 XX 00009000 to my uboot variables as say USB2 and then add that to my bootcmd variable, would that allow the USB2 to work without locking up? I am not sure what XX should be, I thought at first you were specifying the length of the I/O area, but that is not the case. What would I put for XX? I know the USB2 device is 00.08.02, and I think we could use address 0x00009000, or am I wrong?
Would that work?

2) Now that I have the SATA working in Debian, how exactly does one go about writing or hacking in a device driver for OS4? I'd love to add a new driver to the kickstart modules that supports the VIA VT6421 SATA device. The card came with the "C" source code driver for the SATA port as well, perhaps that would help. Any thoughts on this?

Thanks again for your help. I'm slowly learning some things here that are very useful......

Scott

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

 Status: Offline
Profile     Report this post  
Geri 
Re: First Debian Etch Installer CD release!!
Posted on 28-Dec-2012 20:45:37
#200 ]
Elite Member
Joined: 7-Oct-2003
Posts: 2038
From: ST/AT

@scabit

Quote:
scabit wrote:
@Geri
Here is what I did....I plugged in my China combo card - note the micro A1 has only one slot - no chance I will be able to plug things into another slot!

True, I didn't think about it, when I replied.

Quote:
Here is my proc/ioports: ...

Perfect!

Quote:
Does the SATA kernel support the PATA part of the combo IDE/SATA VIA chip, or only the SATA part?

Good question. I only see one mass storage controller PCI device in your lspci output. Can you post the output of "lspci -nn"? Then I can check, which drivers match the PCI vendor/device ID.


Quote:
Hooray! It works.



Quote:
I connected my drive to the SATA port of my multicard and boot into Debian. Using gparted I now see my 80 Gig SATA drive (sda) as well as my 160Gig IDE drive (hda).

I guess the hda drive is connected to the onboard IDE controller, right?

Quote:
Now I can hopefully select a partition to format as EXT3 and backup my entire Debian partition to it .... is there any special trick to that...that is, making a "clone" of a Debian partition on another drive?

You can use "dd" to make an exact 1:1 copy of a partition, if the target partition has the same size or is bigger. In this case you don't need to format the target partition before the copy.

Quote:
Okay, your help in getting the VIA SATA controller to work in Debian on my multicard now opens up a lot of questions.



Quote:
1) The memory addressing area of the multicard being too high seems to have caused the problems with the SATA I/O. Is the same true for the USB2 I/O? That is, if I were to add: pci write.l 00.08.02 XX 00009000 to my uboot variables as say USB2 and then add that to my bootcmd variable, would that allow the USB2 to work without locking up? I am not sure what XX should be, I thought at first you were specifying the length of the I/O area, but that is not the case. What would I put for XX? I know the USB2 device is 00.08.02, and I think we could use address 0x00009000, or am I wrong?
Would that work?

I'm sorry to disappoint you, but this will not work. The SATA controller uses PCI I/O address space, whereas the VIA EHCI controller needs PCI memory space. The distinction between PCI I/O and PCI memory address space is a leftover of the old ISA architecture, as x86 processors originally could access I/O devices only with specific instructions (experts should correct me, if I'm wrong).

BTW: You can see that the EHCI controller is listed in /proc/iomem, but not in /proc/ioports. The "XX" in the U-Boot pci commands are the addresses of the BaseAddressRegisters of the PCI device. Each BAR is 4 bytes long so that a 32 bit address can be written to it.

Quote:
2) Now that I have the SATA working in Debian, how exactly does one go about writing or hacking in a device driver for OS4? I'd love to add a new driver to the kickstart modules that supports the VIA VT6421 SATA device. The card came with the "C" source code driver for the SATA port as well, perhaps that would help. Any thoughts on this?

Uhh, I guess it would be better to ask an OS4 developer about this, as I'm not an expert on AmigaOS programming. At least a developer needs to know about the PCI addressing problem and should have a datasheet for the VIA SATA controller or maybe some BSD driver source code for it, which he could use as a starting point.

Quote:
Thanks again for your help. I'm slowly learning some things here that are very useful......

You're welcome! And be prepared to learn more things.

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

- A1 Linux support -

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle