Nabble - debian-68k

[fixed] Re: Dead link on http://www.debian.org/ports/m68k/

2008-10-11T05:46:58Z

Charles, Ingo,

On Sat, Oct 11, 2008 at 01:33:30PM +0200, Ingo Juergensmann wrote:

> On Sat, Oct 11, 2008 at 08:14:16PM +0900, Charles Plessy wrote:
> > I would like to report the following dead link on
> > http://www.debian.org/ports/m68k/ :
> > <p>The <a href="http://crest.debian.org/">Debian/68k autobuild
> > system</a> contains up to date information about the porting effort.
> > Since crest.debian.org is not down, I guess that it does not host the
> > page anymore...
>
> http://unstable.buildd.net/index-m68k.html has replaced that page on Crest
> years ago.

Thanks, the URL has been fixed on ports/m68k/.
The change will be visible in a few hours on the website.

Best regards,

--
Simon Paillard

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: Dead link on http://www.debian.org/ports/m68k/

2008-10-11T04:33:30Z

On Sat, Oct 11, 2008 at 08:14:16PM +0900, Charles Plessy wrote:

> I would like to report the following dead link on
> http://www.debian.org/ports/m68k/ :
> <p>The <a href="http://crest.debian.org/">Debian/68k autobuild
> system</a> contains up to date information about the porting effort.
> Since crest.debian.org is not down, I guess that it does not host the
> page anymore...

http://unstable.buildd.net/index-m68k.html has replaced that page on Crest
years ago.

--
Ciao... // Fon: 0381-2744150
Ingo \X/ http://blog.windfluechter.net

gpg pubkey: http://www.juergensmann.de/ij_public_key.asc

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Dead link on http://www.debian.org/ports/m68k/

2008-10-11T04:14:16Z

Hello,

I would like to report the following dead link on
http://www.debian.org/ports/m68k/ :

<p>The <a href="http://crest.debian.org/">Debian/68k autobuild
system</a> contains up to date information about the porting effort.

Since crest.debian.org is not down, I guess that it does not host the
page anymore...

Have a nice day,

--
Charles Plessy
Tsurumi, Kanagawa, Japan

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-10T19:41:52Z

On Sat, Oct 11, 2008 at 03:38:13AM +0200, Michael Schmitz wrote:
> Hi,
>
>> Yes, you are right. Tried it several times again with different
>> networking cards (EtherNEC, EtherNAT and NetUsBee) but none of them
>> worked. It stops at "loading optitude" (6%).
>
> That means it got past the point where the EtherNEC load failed
> previously (right at kernel boot time, before the installer gets
> started).

Hopefully it's a module now?

>>> Looks like we established that running the bootstrap as GEM program is
>>> somehow fatal, but the real question is why that is? The difference in
>>> ramdisk load position is minor, really.
>>
>> Well, don't know why as well (guess you knew that) but it only works
>> with bootstra.tos instead of bootstra.prg. Perhaps this can be fixed in
>> the next version of the netinstall iso.
>
> The file rename sure can be done in the nect ISO, fixing the cause of the
> hang may required changes to the bootstrap itself...

Hopefully scheduled for tonight. Already fixed in the d-i build.

>>>> So atm I try the procedure again with my NetUsBee (EtherNEC compatible)
>>>> installed.
>>> Hope my last EtherNEC bugfix is included, at least.
>>
>> Doesn't seem so. So any ideas how to get any of my networking cards
>> running to continue the installation procedure?
>
> Do you get to select any modules at all? That would be the first step to
> provide a custom built module ...

Yikes, custom built stuff for the cdrom? Right now the whole thing is
custom. ;)

Peace,

Stephen

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<stephen@...>

signature.asc (196 bytes) Download Attachment

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-10T18:38:13Z

Hi,

> Yes, you are right. Tried it several times again with different
> networking cards (EtherNEC, EtherNAT and NetUsBee) but none of them
> worked. It stops at "loading optitude" (6%).

That means it got past the point where the EtherNEC load failed previously
(right at kernel boot time, before the installer gets started).

>> Looks like we established that running the bootstrap as GEM program is
>> somehow fatal, but the real question is why that is? The difference in
>> ramdisk load position is minor, really.
>
> Well, don't know why as well (guess you knew that) but it only works
> with bootstra.tos instead of bootstra.prg. Perhaps this can be fixed in
> the next version of the netinstall iso.

The file rename sure can be done in the nect ISO, fixing the cause of the hang
may required changes to the bootstrap itself...

>>> So atm I try the procedure again with my NetUsBee (EtherNEC compatible)
>>> installed.
>> Hope my last EtherNEC bugfix is included, at least.
>
> Doesn't seem so. So any ideas how to get any of my networking cards
> running to continue the installation procedure?

Do you get to select any modules at all? That would be the first step to provide
a custom built module ...

Can you switch to an alternate console (alt-f2 or ctrl-alt-f2 etc.)? The
installer used to log its internal state and progress to one of the consoles (4,
or 7, ??), the output there might be helpful.

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-10T11:28:30Z

Hi Michael,

>> Just a little self-comment. Seems that I've been too impatient. Tried it
>> again today and it worked until Linux wants to recognize an ethernet
>> card.
> :-) Well, it would need the network card fopr a net install I guess.

So we are on-topic again. :-)
Yes, you are right. Tried it several times again with different
networking cards (EtherNEC, EtherNAT and NetUsBee) but none of them
worked. It stops at "loading optitude" (6%).

> Looks like we established that running the bootstrap as GEM program is
> somehow fatal, but the real question is why that is? The difference in
> ramdisk load position is minor, really.

Well, don't know why as well (guess you knew that) but it only works
with bootstra.tos instead of bootstra.prg. Perhaps this can be fixed in
the next version of the netinstall iso.

>> I tried it only with EtherNat installed and Linux unfortunately didn't
>> recognize it.
> The EtherNAT patch may not be in the install kernel yet (hey, I only got
> it working a month ago or so).
>> So atm I try the procedure again with my NetUsBee (EtherNEC compatible)
>> installed.
> Hope my last EtherNEC bugfix is included, at least.

Doesn't seem so. So any ideas how to get any of my networking cards
running to continue the installation procedure?

Regards,

Frank

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [martinwguy@yahoo.it: Bug#497165: amiga-fdisk-cross is not being built on any architecture other than i386]

2008-10-09T18:30:49Z

> It's not upto buildd admins, it's upto packages-arch-specific[1]
> maintainers (cc:'d).
>
> Please change the atari-fdisk p-a-s entry to match what's on the
> package sources:
>
> amiga-fdisk-cross: !m68k !poweprc # Everything but m68k/ppc
> amiga-fdisk: m68k powerpc # m68k/ppc specific

Actually this was about amiga-fdisk, but I have no objection to changing the
Atari fdisk p-a-s entry as well.

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: openjdk-6 (and fastjar) build ping?

2008-10-09T06:43:11Z

On Thu, Oct 09, 2008 at 12:30:41PM +0200, Matthias Klose wrote:

> would it be possible to start builds of these two?

fastjar_0.95-4 was installed on Sept. 23.

openjdk-6 is dep-waited on iceape-dev-bin.

iceape just seems to need a log signed (pinged).

Peace,

Stephen

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<stephen@...>

signature.asc (196 bytes) Download Attachment

Re: fastjar 2:0.95-4 needs to be uploaded from thing2 buildd machine. required for successful OpenJDK 6 build.

2008-10-09T06:32:58Z

On Mon, Oct 06, 2008 at 10:44:11AM +0200, Xerxes Rånby wrote:

> Greetings
>
> I noticed that fastjar 2:0.95-4 have been built sucessfully on thing2 on
> Tue Sep 23 15:36:16 2008
> http://buildd.debian.org/fetch.cgi?pkg=fastjar;ver=2%3A0.95-4;arch=m68k;stamp=1222184176
>
> The m68k OpenJDK-6 package segfaults during build when built with the
> current fastjar found on the ftp sites 2:0.95-3
> 2:0.95-4 is a bugfix for this specific fastjar segfault but i can't find
> it on the current debian sid ftp's. upload needed from thing2 ?

m68k sid is now at debian-ports.org. It seems to be there [0].

[0] <http://ftp.debian-ports.org/debian/pool-m68k/main/f/fastjar/>

hth,

Stephen

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<stephen@...>

signature.asc (196 bytes) Download Attachment

openjdk-6 (and fastjar) build ping?

2008-10-09T03:30:41Z

Hi,

would it be possible to start builds of these two?

thanks, Matthias

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [buildd] Configuring buildds for dports

2008-10-08T19:38:52Z

Hi,

> I updated kullervo's configuration but cannot get access to the database. Do
> I need to generate a new key and have it installed on debian-ports-org?

Forget that - I found the key installed but it still did not seem to work. I'll
look into that some more later.

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-08T19:34:11Z

Hi Frank,

>> The installation procedure run beyond the point where you had to enter
>> the country you live in, the keyboard layout, then it recognized my
>> cd-roms correctly (2 scsi drives connected externally).
>
>> Then the screen went blue and nothing else happened.
>
> Just a little self-comment. Seems that I've been too impatient. Tried it
> again today and it worked until Linux wants to recognize an ethernet card.

:-) Well, it would need the network card fopr a net install I guess.

Looks like we established that running the bootstrap as GEM program is somehow
fatal, but the real question is why that is? The difference in ramdisk load
position is minor, really.

> I tried it only with EtherNat installed and Linux unfortunately didn't
> recognize it.

The EtherNAT patch may not be in the install kernel yet (hey, I only got it
working a month ago or so).

> So atm I try the procedure again with my NetUsBee (EtherNEC compatible)
> installed.

Hope my last EtherNEC bugfix is included, at least.

> Keep you informed.
>
> Regards,
>
> Frank
>
> P.S.: Just in case, a nullmodem cable is available ;-)

Well, now that you found out your Falcon behaves just as mine, I'll just use my
own nullmodem cable :-)

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [martinwguy@yahoo.it: Bug#497165: amiga-fdisk-cross is not being built on any architecture other than i386]

2008-10-08T12:35:44Z

On Wed, Oct 08, 2008 at 08:52:31PM +0200, Christian T. Steigies wrote:
> could the buildd admins please enable building for amiga-fdisk and upload
> the resulting packages (powerpc has been built, but it seems it was never
> uploaded).

It's not upto buildd admins, it's upto packages-arch-specific[1]
maintainers (cc:'d).

Please change the atari-fdisk p-a-s entry to match what's on the
package sources:

amiga-fdisk-cross: !m68k !poweprc # Everything but m68k/ppc
amiga-fdisk: m68k powerpc # m68k/ppc specific

[1] http://cvs.debian.org/srcdep/Packages-arch-specific?cvsroot=dak&rev=HEAD

signature.asc (196 bytes) Download Attachment

Re: [martinwguy@yahoo.it: Bug#497165: amiga-fdisk-cross is not being built on any architecture other than i386]

2008-10-08T12:18:55Z

On Wed, Oct 08, 2008 at 08:52:31PM +0200, Christian T. Steigies wrote:
> could the buildd admins please enable building for amiga-fdisk and upload
> the resulting packages (powerpc has been built, but it seems it was never
> uploaded).

First off, debian-ports is not the right contact address for buildd
admins, there is no guarantee the buildd admins of a particular port are
subscribed, AFAIK.

Second, you need to get amiga-fdisk removed from packages-arch-specific
to have it built by the buildds, see
http://cvs.debian.org/srcdep/Packages-arch-specific?root=dak&view=markup
and the contact addresses at the top of that file.

cheers,

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

[martinwguy@yahoo.it: Bug#497165: amiga-fdisk-cross is not being built on any architecture other than i386]

2008-10-08T11:52:31Z

Hi,
could the buildd admins please enable building for amiga-fdisk and upload
the resulting packages (powerpc has been built, but it seems it was never
uploaded).

thanks,
Christian

----- Forwarded message from Martin Guy <martinwguy@...> -----

Subject: Bug#497165: amiga-fdisk-cross is not being built on any architecture other than i386

Not a bug in amiga-fdisk itself, but a Debian buildd issue that needs sorting

On 16 Jan 2008, armel was added to the architecture list for
amiga-fdisk-cross (bug #461081)
but, despite being enabled for 12 architectures, the package has not
been built for anything but i386.

I am guessing this is due to being included in the various buildd
admins' Not-For-Us lists - can you poke them?

Cheers!

----- End forwarded message -----

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-08T09:54:50Z

> The installation procedure run beyond the point where you had to enter
> the country you live in, the keyboard layout, then it recognized my
> cd-roms correctly (2 scsi drives connected externally).

> Then the screen went blue and nothing else happened.

Just a little self-comment. Seems that I've been too impatient. Tried it
again today and it worked until Linux wants to recognize an ethernet card.

I tried it only with EtherNat installed and Linux unfortunately didn't
recognize it.

So atm I try the procedure again with my NetUsBee (EtherNEC compatible)
installed.

Keep you informed.

Regards,

Frank

P.S.: Just in case, a nullmodem cable is available ;-)

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: Re: No telnet into SE/30?

2008-10-08T08:38:58Z

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-07T02:05:50Z

>> O.K. Copied everything to harddisc. What happened then is that after I
>> started bootstra.prg from partition c:\ (GEMDOS) with MiNT the computer
>> just freezed. When I started from partition c:\ with TOS the computer
>> reboots (coldboot).
> NVDI included, or not?

Without NVDI, just plain TOS but I guess you found the solution (see below).

> I just remembered something else that may be important - after the first
> unsuccesful boot attempt from HD, I renamed bootstrap.prg to
> bootstrap.tos. This appears to have some effect on how TOS handles the
> bootstrap, because reading the kernel image from disk and decompressing
> it happens a lot less fast than it does when using bootstrap.prg as name
> for the binary.

And that did the trick. I renamed it to bootstrap.tos as well and voilà
- Linux boots.

> :-) Understood (that's why I left the kernel image on CD). Could you
> please try that as well?

Well, I guess that is not necessary because Linux now boots.

The installation procedure run beyond the point where you had to enter
the country you live in, the keyboard layout, then it recognized my
cd-roms correctly (2 scsi drives connected externally).

Then the screen went blue and nothing else happened.

Regards,

Frank

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [buildd] Configuring buildds for dports

2008-10-07T00:01:26Z

Hi Stephen,

>> Here are some notes from zeus, who is now buildding on dports.
>
> First package uploaded.

I updated kullervo's configuration but cannot get access to the database. Do I
need to generate a new key and have it installed on debian-ports-org?

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-06T09:57:06Z

On Mon, Oct 06, 2008 at 02:31:18AM +1100, Finn Thain wrote:

>
> On Sun, 5 Oct 2008, Stephen R Marenka wrote:
>
> > I think the fix for both of these is fixed by using modules=etch-support
> > on the kernel command line.
> >
> > http://wiki.debian.org/DebianInstaller/M68k#issues
>
> Ah yes, that rings a bell... will test that tomorrow.
>
> (BTW, diff tells me that the initrd.gz on the CD is the same as the daily
> initrd.gz that I used. Which is a good to know.)
>
> > > Only apt-get update says 404 not found for
> > > http://ftp.debian-ports.org/debian/dists/etch/main/binary-m68k/{Packages.gz,Sources.gz}
> > > and craps out... not sure why.
> >
> > debian-ports hosts sid. For the regular debian mirrors, you'd need
> > etch-m68k.
>
> Doh! That fixed it, thanks.
>
> >
> > Thanks for the report! It sounds like we're making progress.
>
> Absolutely.
>
> Finn
>
> PS. We didn't CC the list, but probably should do?

Sorry, didn't even notice. cc'ing now just for the record. :)

Thanks,

Stephen

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<stephen@...>

signature.asc (196 bytes) Download Attachment

fastjar 2:0.95-4 needs to be uploaded from thing2 buildd machine. required for successful OpenJDK 6 build.

2008-10-06T01:44:11Z

Greetings

I noticed that fastjar 2:0.95-4 have been built sucessfully on thing2 on
Tue Sep 23 15:36:16 2008
http://buildd.debian.org/fetch.cgi?pkg=fastjar;ver=2%3A0.95-4;arch=m68k;stamp=1222184176

The m68k OpenJDK-6 package segfaults during build when built with the
current fastjar found on the ftp sites 2:0.95-3
2:0.95-4 is a bugfix for this specific fastjar segfault but i can't find
it on the current debian sid ftp's. upload needed from thing2 ?

Have a great day!
Xerxes Rånby

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-06T00:54:32Z

On Mon, 6 Oct 2008, Finn Thain wrote:

>
> Ignore that post -- it occurred to me that I probably have the daily
> initrd instead of the one from the CD... I will try again.

It turns out that they are the same initrd. Stephen helpfully pointed out
that I needed the "suite=etch-m68k modules=etch-support" bootloader
arguments (and the distribution is of course etch-m68k, hence the 404 when
I changed sid to etch in sources.list). The installer needs the
debian-ports archive selected at the beginning, but the target needs a
normal debian mirror selected at the end.

I tried again with the correct boot args today and now sources.list is
fine (it says etch-m68k not sid) but the other problems remain:
--dlwaypoint and safe-upgrade options are still used though they don't
work in the etch chroot, so the "configure package manager" step got stuck
and "select and install software" failed. Despite that I ended up with a
working system.

Finn

>
> Finn
>
> On Sun, 5 Oct 2008, Finn Thain wrote:
>
> >
> > On Wed, 24 Sep 2008, Stephen R Marenka wrote:
> >
> > >
> > > Both images use lenny d-i. The etch-m68k one installs etch-m68k. The
> > > sid one might install sid. I haven't added the appropriate locales yet
> > > for sid, so who knows.
> >
> > I tried the etch installer on my quadra 630 today (debian mirror at
> > ftp.debian-ports.org).
> >
> > It went well up to the "configure the package manager" step. A couple of
> > times it hung until I killed the stuck process. I looked into it and found
> > that apt-setup-verify passes the --dlwaypoint option which causes a
> > debconf-apt-progress child in the etch chroot to stall. I edited out
> > "--dlwaypoint 100" and re-ran the step. I think it tried to upgrade
> > source.list to sid. It complained about a missing public key with which to
> > verify the sid Release file.
> >
> > After that, the "select and install software" step fails too, because
> > aptitude in the etch chroot doesn't seem to understand "safe-upgrade". So
> > I basically ended up doing "finish installation" at this point.
> >
> > I also found that, on the initrd, busybox seems to have a bug. If you
> > switch to a shell prompt and run ip/wget/tar with incorrect options or no
> > options they all segfault. But they work fine when you pass correct
> > arguments.
> >
> > In the end, I have a bootable etch system, so the installer is 99%
> > working.
> >
> > Only apt-get update says 404 not found for
> > http://ftp.debian-ports.org/debian/dists/etch/main/binary-m68k/{Packages.gz,Sources.gz}
> > and craps out... not sure why.
> >
> > Finn
> >
> >
> >
>

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-05T06:18:30Z

Ignore that post -- it occurred to me that I probably have the daily
initrd instead of the one from the CD... I will try again.

Finn

On Sun, 5 Oct 2008, Finn Thain wrote:

>
> On Wed, 24 Sep 2008, Stephen R Marenka wrote:
>
> >
> > Both images use lenny d-i. The etch-m68k one installs etch-m68k. The
> > sid one might install sid. I haven't added the appropriate locales yet
> > for sid, so who knows.
>
> I tried the etch installer on my quadra 630 today (debian mirror at
> ftp.debian-ports.org).
>
> It went well up to the "configure the package manager" step. A couple of
> times it hung until I killed the stuck process. I looked into it and found
> that apt-setup-verify passes the --dlwaypoint option which causes a
> debconf-apt-progress child in the etch chroot to stall. I edited out
> "--dlwaypoint 100" and re-ran the step. I think it tried to upgrade
> source.list to sid. It complained about a missing public key with which to
> verify the sid Release file.
>
> After that, the "select and install software" step fails too, because
> aptitude in the etch chroot doesn't seem to understand "safe-upgrade". So
> I basically ended up doing "finish installation" at this point.
>
> I also found that, on the initrd, busybox seems to have a bug. If you
> switch to a shell prompt and run ip/wget/tar with incorrect options or no
> options they all segfault. But they work fine when you pass correct
> arguments.
>
> In the end, I have a bootable etch system, so the installer is 99%
> working.
>
> Only apt-get update says 404 not found for
> http://ftp.debian-ports.org/debian/dists/etch/main/binary-m68k/{Packages.gz,Sources.gz}
> and craps out... not sure why.
>
> Finn
>
>
>

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-05T02:10:05Z

On Wed, 24 Sep 2008, Stephen R Marenka wrote:

>
> Both images use lenny d-i. The etch-m68k one installs etch-m68k. The
> sid one might install sid. I haven't added the appropriate locales yet
> for sid, so who knows.

I tried the etch installer on my quadra 630 today (debian mirror at
ftp.debian-ports.org).

It went well up to the "configure the package manager" step. A couple of
times it hung until I killed the stuck process. I looked into it and found
that apt-setup-verify passes the --dlwaypoint option which causes a
debconf-apt-progress child in the etch chroot to stall. I edited out
"--dlwaypoint 100" and re-ran the step. I think it tried to upgrade
source.list to sid. It complained about a missing public key with which to
verify the sid Release file.

After that, the "select and install software" step fails too, because
aptitude in the etch chroot doesn't seem to understand "safe-upgrade". So
I basically ended up doing "finish installation" at this point.

I also found that, on the initrd, busybox seems to have a bug. If you
switch to a shell prompt and run ip/wget/tar with incorrect options or no
options they all segfault. But they work fine when you pass correct
arguments.

In the end, I have a bootable etch system, so the installer is 99%
working.

Only apt-get update says 404 not found for
http://ftp.debian-ports.org/debian/dists/etch/main/binary-m68k/{Packages.gz,Sources.gz}
and craps out... not sure why.

Finn

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-04T22:00:14Z

Hi again,

> Output from Bootstra.prg:
>
> Linux/68k AtariBootstrap version 6.0snapshot Dec 13 2004
> Copyright 1993-2004 by Arjan Knor, Robert de Vries, Roman Hodek, Andreas
> Schwab, Petr Stehlik
>
> Kernel command line: root=/dev/ram video=atafb:vga16 console=tty
> load_ramdisk=1 fb=false BOOT_IMAGE=c:\linux\vmlinuz
> CPU: 68060; FPU: 68060
> Model: Falcon
> TT-RAM: 510.51050 MB at 0x01000000
> ST-RAM: 14 MB at 0x00000000
> Total: 524.52450 MB

Same here.

> Decompressing c:\linux\vmlinuz
> - 2711552 gunzip - 1407533 file
>
> | 2775312 file
>
> Bootstrap's bootinfo version: 2.1
> Kernel's bootinfo version : 2.1
> ramdisk src at 0x130fbc, size is 2775312

src at 0x14779f0 (TOS) or 0x14789f0 (PRG), size 2604787

(we seem to use different ramdisk images)

> ramdisk dest is 0x20bda6f0 ... 0x20e7ffff

Different due to size - 0x20c0410d ... same.

> Kernel segment 0 at 0x1000, size 2796276

Size is 2808648, so we use different kernel images as well.

> Kernel segment 1 at 0x2ac000, size 122880

seg. 1 at 0x2af000 size 126976

> boot_info is at 0x2ca000

boot info at 0x2ce000

See the difference in ramdisk source address above - it might be a good exercise
to try renaming the bootstrap to .tos (tried that and again, it hung
indefinitely if I rename it to .prg)

Regarding differences in kernel and ramdisk sizes - I guess you used a later
version of the CD than mine (or a different ISO). I'll go fetch a current one
...

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-04T20:20:50Z

Hi,

> O.K. Copied everything to harddisc. What happened then is that after I
> started bootstra.prg from partition c:\ (GEMDOS) with MiNT the computer
> just freezed. When I started from partition c:\ with TOS the computer
> reboots (coldboot).

NVDI included, or not?

I just remembered something else that may be important - after the first
unsuccesful boot attempt from HD, I renamed bootstrap.prg to bootstrap.tos. This
appears to have some effect on how TOS handles the bootstrap, because reading
the kernel image from disk and decompressing it happens a lot less fast than it
does when using bootstrap.prg as name for the binary.

>>> bootargs:
>>> =========
>>> -s -d -k o:\install\kernels\vmlinuz-2.6.26-1-atari -r
>>> o:\install\cdrom\initrd.gz root=/dev/ram video=atafb:vga16 console=tty
>>> load_ramdisk=1 fb=false
>> That's the one from the CD, except for fixed paths?
>
> Right. This time modified as follows (had to shorten kernel's name due
> to gemdos restrictions (maxpath 8+3):

:-) Understood (that's why I left the kernel image on CD). Could you please try
that as well?

> -s -d -k c:\linux\vmlinuz -r c:\linux\initrd.gz root=/dev/ram
> video=atafb:vga16 console=tty load_ramdisk=1 fb=false
>
> Output from Bootstra.prg:
>
> Linux/68k AtariBootstrap version 6.0snapshot Dec 13 2004
> Copyright 1993-2004 by Arjan Knor, Robert de Vries, Roman Hodek, Andreas
> Schwab, Petr Stehlik
>
> Kernel command line: root=/dev/ram video=atafb:vga16 console=tty
> load_ramdisk=1 fb=false BOOT_IMAGE=c:\linux\vmlinuz
> CPU: 68060; FPU: 68060
> Model: Falcon
> TT-RAM: 510.51050 MB at 0x01000000
> ST-RAM: 14 MB at 0x00000000
> Total: 524.52450 MB
>
> Decompressing c:\linux\vmlinuz
> - 2711552 gunzip - 1407533 file
>
> | 2775312 file
>
> Bootstrap's bootinfo version: 2.1
> Kernel's bootinfo version : 2.1
> ramdisk src at 0x130fbc, size is 2775312
> ramdisk dest is 0x20bda6f0 ... 0x20e7ffff
> Kernel segment 0 at 0x1000, size 2796276
> Kernel segment 1 at 0x2ac000, size 122880
> boot_info is at 0x2ca000
>
> Type a key to continue the Linux boot...
> Booting Linux...

And a while after this, it reboots?

I'll check your data against my boot results. I hope we'll get there yet ...

Do you have a nullmodem serial cable to attach the Falcon to another computer,
and capture kernel debug output there?

Thanks,

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-04T10:20:22Z

> I had that screen corruption one - tried it again and it worked. Only
> difference: the initrd was located on the HD as well as the bootargs.

O.K. Copied everything to harddisc. What happened then is that after I
started bootstra.prg from partition c:\ (GEMDOS) with MiNT the computer
just freezed. When I started from partition c:\ with TOS the computer
reboots (coldboot).

>> bootargs:
>> =========
>> -s -d -k o:\install\kernels\vmlinuz-2.6.26-1-atari -r
>> o:\install\cdrom\initrd.gz root=/dev/ram video=atafb:vga16 console=tty
>> load_ramdisk=1 fb=false
> That's the one from the CD, except for fixed paths?

Right. This time modified as follows (had to shorten kernel's name due
to gemdos restrictions (maxpath 8+3):

-s -d -k c:\linux\vmlinuz -r c:\linux\initrd.gz root=/dev/ram
video=atafb:vga16 console=tty load_ramdisk=1 fb=false

Output from Bootstra.prg:

Linux/68k AtariBootstrap version 6.0snapshot Dec 13 2004
Copyright 1993-2004 by Arjan Knor, Robert de Vries, Roman Hodek, Andreas
Schwab, Petr Stehlik

Kernel command line: root=/dev/ram video=atafb:vga16 console=tty
load_ramdisk=1 fb=false BOOT_IMAGE=c:\linux\vmlinuz
CPU: 68060; FPU: 68060
Model: Falcon
TT-RAM: 510.51050 MB at 0x01000000
ST-RAM: 14 MB at 0x00000000
Total: 524.52450 MB

Decompressing c:\linux\vmlinuz
- 2711552 gunzip - 1407533 file

| 2775312 file

Bootstrap's bootinfo version: 2.1
Kernel's bootinfo version : 2.1
ramdisk src at 0x130fbc, size is 2775312
ramdisk dest is 0x20bda6f0 ... 0x20e7ffff
Kernel segment 0 at 0x1000, size 2796276
Kernel segment 1 at 0x2ac000, size 122880
boot_info is at 0x2ca000

Type a key to continue the Linux boot...
Booting Linux...

Regards,

Frank

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-02T01:22:51Z

Hi,

> O.K. Even though I still didn't receive my ct63 yet I gave it another
> try. Tried the etch netinstall iso from the source above mentioned.
>
> After I started bootstra.prg from harddisk in 16 colour vga mode on my
> ct60 falcon with the following bootargs the ide led started to flash,
> after a while the screen got corrupted and that's it.

I had that screen corruption one - tried it again and it worked. Only
difference: the initrd was located on the HD as well as the bootargs.

> bootargs:
> =========
> -s -d -k o:\install\kernels\vmlinuz-2.6.26-1-atari -r
> o:\install\cdrom\initrd.gz root=/dev/ram video=atafb:vga16 console=tty
> load_ramdisk=1 fb=false

That's the one from the CD, except for fixed paths?

Cheers,

Michael

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: Changes for m68k port

2008-10-01T11:09:45Z

On Tue, Sep 30, 2008 at 10:47:20PM +0200, Helge Kreutzmann wrote:
> Hello Stephen,
> I read your e-mail on -release[1]. Are you planing on updating the
> entry on www.debian.org[2] as well?

This is a great point. Who has access?

> [1] http://lists.debian.org/debian-release/2008/09/msg01003.html
> [2] http://www.debian.org/ports/m68k and http://www.debian.org/ports

Thanks,

Stephen

--
Stephen R. Marenka If life's not fun, you're not doing it right!
<stephen@...>

signature.asc (196 bytes) Download Attachment

Re: [patch 3/2] m68k: Atari EtherNAT - add writew_be for data push

2008-10-01T10:40:39Z

>> Where are the current isos located?
> Ahhhhhh, iso's (runs screaming into the night).
> <http://people.debian.org/~smarenka/d-i/m68k/cds/daily/> is the
> canonical place for the daily cd buildds.
> I'm in the process of getting those working again.
> Hopefully I'll have something reasonable for you to test about the
> time you're ready to test. :)

O.K. Even though I still didn't receive my ct63 yet I gave it another
try. Tried the etch netinstall iso from the source above mentioned.

After I started bootstra.prg from harddisk in 16 colour vga mode on my
ct60 falcon with the following bootargs the ide led started to flash,
after a while the screen got corrupted and that's it.

bootargs:
=========
-s -d -k o:\install\kernels\vmlinuz-2.6.26-1-atari -r
o:\install\cdrom\initrd.gz root=/dev/ram video=atafb:vga16 console=tty
load_ramdisk=1 fb=false

Regards,

Frank

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-01T09:36:42Z

On Wed, 1 Oct 2008, Lance Tagliapietra wrote:

> Summary: debian install CD image for etch-m68k works on my Amiga!
>
> On Tue, Sep 30, 2008 at 08:57:05AM -0500, Lance Tagliapietra wrote:
> >
> > c) The decompressed kernel did not boot with amiboot-5.6 when specifying
> > the compressed initrd included on the install CD. Tonight I will try to
> > boot the decompressed 2.6 kernel with a decompressed initrd.
>
> Correction - the decompressed kernel *does* boot with amiboot-5.6 when
> specifying the compressed initrd included on the install CD! You just
> have to be very, very patient. I found that I had to wait at a blank
> screen almost 2 minutes from the time I pressed return to tell amiboot
> to boot the linux kernel until the boot messages started scrolling up
> the screen. I never had to wait that long for a kernel to
> start putting boot messages on to the screen before, so I had been
> thinking that it had failed.

How long does it take to decompress the ramdisk under AmigaOS? ;-)

> Also, it seems that boot fails if the initrd is not compressed. So the
> required combination seems to be uncompressed kernel and compressed
> initrd.

Good to know...

> I also observed that a memfile was not needed. The 2.6 kernel properly
> identifies my Zorro II RAM, it seems the Zorro II RAM device support is
> compiled in.

Yes, 2.6 doesn't use Z2 RAM as system memory on Z3 capable machines.

> The clgen kernel support seems to be broken. Adding the video=clgen:
> does not switch video to the EGS Spectrum video output, and using the
> video=clgen... parameters that I use in booting my 2.4 kernel does cause
> the EGS Spectrum video to be used, but the monitor capabilities and
> signal parameters for the 800x600 mode I request do not operate the
> same. The screen was not viewable, the horizontal rate was all messed
> up.

Hmm, I was already afraid clgen got broken somewhere...

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: m68k netinst cd's

2008-10-01T08:15:51Z

Summary: debian install CD image for etch-m68k works on my Amiga!

On Tue, Sep 30, 2008 at 08:57:05AM -0500, Lance Tagliapietra wrote:
>
> c) The decompressed kernel did not boot with amiboot-5.6 when specifying
> the compressed initrd included on the install CD. Tonight I will try to
> boot the decompressed 2.6 kernel with a decompressed initrd.

Correction - the decompressed kernel *does* boot with amiboot-5.6 when
specifying the compressed initrd included on the install CD! You just
have to be very, very patient. I found that I had to wait at a blank
screen almost 2 minutes from the time I pressed return to tell amiboot
to boot the linux kernel until the boot messages started scrolling up
the screen. I never had to wait that long for a kernel to
start putting boot messages on to the screen before, so I had been
thinking that it had failed.

Also, it seems that boot fails if the initrd is not compressed. So the
required combination seems to be uncompressed kernel and compressed
initrd.

> d) I would suggest that the uncompressed kernel be placed on the CD, or
> an AmigaDos native gzip should be included on the CD if not already
> there.

There is a native Amigados gzip program. I tested it with the
compressed kernel and initrd.gz from the CD and it worked.

>
> e) The amiboot on the install CD may need to be upgraded, or regressed
> back to the version 5.6, as the version currently on the CD does not
> work. Is there a hint to make that work that I'm just not getting?

I have not been able to get the amiboot on the CD to work. Testing above
was done with amiboot version 5.3. It just seems to lock up the
machine, maybe kills interrupts improperly, as I run a hi res mouse
program, which drops the mouse back to the standard resolution, my
indication that something wrong happened.

> In my tests, I modify the startinstall script to add a -m memfile
> option, where my memfile hides the 6M of Zorro II memory on my GVP
> memory card. I assume that I still need to do this with 2.6 kernel. I
> will also add a video=clgen: parameter and try booting with my EGS
> Spectrum video card.

I also observed that a memfile was not needed. The 2.6 kernel properly
identifies my Zorro II RAM, it seems the Zorro II RAM device support is
compiled in.

The clgen kernel support seems to be broken. Adding the video=clgen:
does not switch video to the EGS Spectrum video output, and using the
video=clgen... parameters that I use in booting my 2.4 kernel does cause
the EGS Spectrum video to be used, but the monitor capabilities and
signal parameters for the 800x600 mode I request do not operate the
same. The screen was not viewable, the horizontal rate was all messed
up.

Looking at the kernel messages, with the initrd loaded, the kernel ran
out of memory during boot, but observed this and killed a low priority
process. Also, another kernel message indicated that it could not
access the hardware real time clock.

Once the kernel booted, the installer started to run. It immediately
informed me that it was operating in a low memory mode. I let the
installer run until it identified my CD ROM drive, found the
installation media, and started verifying the pool files.

That's as far as I could get last night.

Regards,

--Lance

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: exim performance hint

2008-09-30T22:49:38Z

Ingo,

I cleaned out my whole /var/spool tonight on my Fedora box. Lots of
print jobs from months ago were abandoned there. It was good to clean
the spool folders out. Not something I regularly attend to. So a
cleanout script run by cron would be a great idea.

Richard
(the guy with a few SE/30's)

Ingo Juergensmann wrote:

> On Tue, Sep 30, 2008 at 05:19:05PM -0500, Lance Tagliapietra wrote:
>
>
>> Looking closer, exim is run by a cron task periodically, to try to
>> re-send email that could not be sent, or was queued for some reason. If
>> the message could not be delivered, it sits in the queue. Due to a
>> mis-configuration of exim, I had system messages to root sitting in
>> there, all the mail to support popularity-contest, and others, for about
>> 6 years! There were about 1000 files there to be processed, each time,
>> which simply took time.
>> [...]
>> Note that all this time I was able to send email/ receive email and
>> didn't really know about much system generated email.
>>
>
> You could run exim -d -qff <msgid> to see why those mails in the queue are
> failing.
>
> I suffered from a similar problem once and wrote a little Python script to
> deal with this problem. It tries to flush the queue and depending on the
> error code it gets back, it either tries to deliver the mail again, spool it
> again or simply delete it. If it can't send mails for a certain amount of
> time in the queue, it deletes the mail as well.
> The script is attached and as always: use it on your own risk! ;)
>
>

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...

Re: exim performance hint

2008-09-30T22:07:36Z

On Tue, Sep 30, 2008 at 05:19:05PM -0500, Lance Tagliapietra wrote:

> Looking closer, exim is run by a cron task periodically, to try to
> re-send email that could not be sent, or was queued for some reason. If
> the message could not be delivered, it sits in the queue. Due to a
> mis-configuration of exim, I had system messages to root sitting in
> there, all the mail to support popularity-contest, and others, for about
> 6 years! There were about 1000 files there to be processed, each time,
> which simply took time.
> [...]
> Note that all this time I was able to send email/ receive email and
> didn't really know about much system generated email.

You could run exim -d -qff <msgid> to see why those mails in the queue are
failing.

I suffered from a similar problem once and wrote a little Python script to
deal with this problem. It tries to flush the queue and depending on the
error code it gets back, it either tries to deliver the mail again, spool it
again or simply delete it. If it can't send mails for a certain amount of
time in the queue, it deletes the mail as well.
The script is attached and as always: use it on your own risk! ;)

--
Ciao... // Fon: 0381-2744150
Ingo \X/ http://blog.windfluechter.net

gpg pubkey: http://www.juergensmann.de/ij_public_key.asc

#!/usr/bin/python
# (c) 2003 by Ingo Juergensmann <ij (a) ${year}.bluespice.org>
# License: GPLv2, see http://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# for details.

import os
from string import *
import popen2

cmd = "mailq"
r, w = popen2.popen2(cmd)

while 1:
line = r.readline()
if not len(line): break
print "!", line
if count(line, "*** frozen ***"):
msgid = split(line)[2]
print "!! Examing MsgID", msgid,
cmd2 = "exim4 -Mrm %s" % msgid
os.system(cmd2)
elif (count(line, "-") and (count(line,"<") and count(line,">"))):
msgid = split(line)[2]
try:
timeout = int(replace(split(line)[0],"h",""))
except:
try:
timeout = int(replace(split(line)[0],"m",""))
except:
try:
timeout = int(replace(split(line)[0],"d","")) * 24
except:
pass
print "++", timeout,
print "!! Examing MsgID", msgid
print "!! Trying to deliver MsgID", msgid,"...",
cmd2 = "exim4 -N -M %s" % msgid
r2, w2 , e2= popen2.popen3(cmd2)
while 1:
line2 = e2.readline()
#print "!!", line2
if len(line2)<1: break
print "!!", line2,
if (((count(line2, "(-19)") or count(line2, "(113)") or count(line2, "(110)") or count(line2, "(111)")) and timeout>=42) or count(line2, "(-44)") or count(line2, "(-1)") or count(line2, "(-42)")):
cmd2 = "exim4 -Mrm %s" % msgid
os.system(cmd2)
print "!!! Message %s deleted." % msgid
r2.close()
w2.close()
e2.close()

r.close()
w.close()

exim performance hint

2008-09-30T15:19:05Z

Hello,

Summary: clean exim dead letter database

For the past several months I have seen my Debian 68K system run slower
and slower. Checking is ps ax showed that every 5 minutes or so, exim
would be running for about 5 minutes. I mean, I thought I got a low of
spam, but..

Looking closer, exim is run by a cron task periodically, to try to
re-send email that could not be sent, or was queued for some reason. If
the message could not be delivered, it sits in the queue. Due to a
mis-configuration of exim, I had system messages to root sitting in
there, all the mail to support popularity-contest, and others, for about
6 years! There were about 1000 files there to be processed, each time,
which simply took time.

So, take a look at

#du -a /var/spool/exim

output, and if thre are any files in the /var/spool/exim/input or
/var/spool/exim/msglog directories, there is probably a
mis-configuration in exim that may not be immediately obvious. (note,
the spool directory is used by exim in the process of sending email, so
it does get used. Files will set there while waiting to be sent, or if
exim has given up on trying to send them).

Note that all this time I was able to send email/ receive email and
didn't really know about much system generated email.

So, now, hopefully, my popularity-contest support for m68k will start
being counted.

--Lance

--
To UNSUBSCRIBE, email to debian-68k-REQUEST@...
with a subject of "unsubscribe". Trouble? Contact listmaster@...