Re: Anyone ever tried Debian on a Fujitsu/Oracle M3000 ?

[Chris Newport]

On 15/05/2022 14:52, Dennis Clarke wrote:

On 5/14/22 17:52, Rick Leir wrote:
 From folks who have tried? Sorry, no. If the info below is not 
helpful, sorry and please just delete my message.

Is illumos or SmartOS an option? Info from Bryan Cantrill:
http://dtrace.org/blogs/bmc/2017/09/04/the-sudden-death-and-eternal-life-of-solaris/ 

Following a few links from that article to 
https://illumos.org/docs/about/distro/ , I see that

derived systems claim to have SPARC builds: Tribblix, DilOS, and V9os.

cheers -- Rick


Yep .. been there and read that. Even if any of those did work at all we
are going in a circle and landing back on some Solaris thing. That is
not really of any value any more. It is beginning to sound a lot like
Oracle sold a whack of machines to people about ten years ago and they
are all dead pieces of trash and junk today. Nothing runs there any
more.


They should still run just fine on older versions of Solaris (8 or 10 ?) 
and similar

aged Debian and BSD variants. Not trash, just stuck in their own time.

I have customers still running Ultra 60s and a few still on ancient 
stuff running
SunOs 4.1.4, they still run the tasks they were bought for and will be 
replaced

with modern kit when they finally die. The killer will be the SCSI disks.

Re: Fan speed control on T2000

[Chris Newport]

The fans are controlled by the open boot prom, Solaris monitors the CPU 
temperatures and sends control calls to the OBP. Not sure if anyone has 
implemented this for Linux.


Running the fans at max will do no harm apart from the noise.



On 04/14/2019 11:31 AM, Alexandre Bencz wrote:

Hi!
I have installed Debian 9 on my T2000, but, the front fans are sniping 
in the max, all the time, there is some tool to install to control the 
fans ?

Re: OBP patch upgrade for Ultra 2.

[Chris Newport]

On Wed, 8 Apr 2009, Chris Andrew wrote:


I've been lucky enough to get myself an Ultra Enterprise 2.  I want to
make sure the OBP patch is as up to date as possible.

This site:

http://www.sunshack.org/data/bootroms.html

Suggests that 3.25.0 was the latest patch.  Can anyone confirm this?
If I download this patch, can someone remind me how to install it.  I
did this ages ago for a U10, but I've lost my instructions.

Any help appreciated.  I also intend to squeeze as much CPU and RAM
into this as I can (currently 296megs).


The patch is usually a zip file containing a README and an executable
which you may need to run under Solaris. You need to move a jumper
on the main board to allow the EEPROM to be written.

I have a spare pair of 300 MHz procs here, remove the NOSPAM to email me.
You can fit a total of 16 * 128 MB sticks in sets of 4 to get 2Gb.


--
To UNSUBSCRIBE, email to debian-sparc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Re: Debian Installer support for Sparc

[Chris Newport]

On Sunday 13 July 2008, Chris Newport wrote:

An underlying issue here is that the soc.c and socal.c drivers are not
maintained and have not worked since 2.2.x
This means that the disk packs ( 3 drawers of 10 disks each ) usually
found on these machines will not work.


In that case I would say that the request to have them included in D-I was 
not correct and that we should remove them again. If I see no objections 
to that over the next few days, I will make that change.


http://www.mail-archive.com/debian-sparc@lists.debian.org/msg20892.html
says socal works from vanilla sources, if the poster was building a kernel 
current at the time of his post then he can confirm they do work and should 
stay.


1) The original poster needed the soc.c (Sun FC/25) driver which 
definitely has not worked since 2.2.x
2) soc.c and socal.c (Sun FC/100) are both unmaintained 
3) The above msg20892.html does not specify which kernel version 
worked with socal.c, the poster only says "when I hand built kernels"

4) This mess has been discussed several times on the sparclinux list

Is there any evidence that socal.c has worked with modern kernels ?.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Debian Installer support for Sparc (was: Sun Enterprise 3500 - FC)

[Chris Newport]

On Sun, 13 Jul 2008, Frans Pop wrote:


Date: Sun, 13 Jul 2008 08:01:52 +0200
From: Frans Pop <[EMAIL PROTECTED]>
To: debian-sparc@lists.debian.org
Subject: Debian Installer support for Sparc (was: Sun Enterprise 3500 - FC)

Giorgio Davanzo wrote:

we're stuck with the FC disk problem. We've found the alleged solution
to the problem, but it's not working:
http://www.mail-archive.com/debian-sparc@lists.debian.org/msg20888.html

Someone suggested using microcode from SUN firmware (?)

Did anyone ever solved the problem? If yes, could you give us a hint?


From that thread it seems the main issue was missing drivers. This should
been fixed since the Lenny Beta1 release of the installer [1]:
http://bugs.debian.org/443504


An underlying issue here is that the soc.c and socal.c drivers are not
maintained and have not worked since 2.2.x
This means that the disk packs ( 3 drawers of 10 disks each ) usually
found on these machines will not work.

The solution is to use SCSI disk packs.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Intention to drop sparc32 support for Lenny

[Chris Newport]

Uwe Hermann wrote:


Well, I just saw three or more sparc32 patches being committed to Linus'
git tree today or yesterday, so that may not be quite correct.

 


You are missing the point. Those patches were created by enthusiastic
users fixing the problems that they have experienced.

Until someone volunteers to become the official maintainer Sparc32
is effectively dead. An official maintainer is essential as the point of
contact and the person who contantly tests to ensure that changes
elsewhere in the kernel have not caused regressions.

Without an official maintainer who is going to be responsible for
saying that the port is or is not ready for the next kernel release ?.
The Sun4d subset of Sparc32 is an example of cumulative bitrot
over many years which will take a major effort to resolve.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Retiring the sparc32 port

[Chris Newport]

andrew holway wrote:


Its only wasteful if they end up in a landfill. If they could be
recycled? Another important factor to consider is power consumption.
If you have several V8's (circa early 90's?) running I shudder think
how much juice they draw.


32 bit Sparc systems draw far LESS power than modern machines.
For example, the PSU in my SS10 is rated at 60 watts MAX.
In reality it sits there as a firewall drawing around 28 watts (measured).



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: [ss20]cdrom

[Chris Newport]

Chris Andrew wrote:


Ludovic,

That's interesting.  Do you know where I can download an ISO that has 
the re-written module?  I prefer debian, but at the moment, any linux 
would do.


Many thanks,

Chris.


The latest gregbeta release of splack should work.
Be warned, however, that it is still in the beta stage and the non-usa
keyboard mappings are broken.

ftp://ftp.splack.org/pub/splack/gregbeta/bobware-10.2/iso/bobware-10.2.rc3.iso



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Retiring the sparc32 port

[Chris Newport]

> My own experience with sun4d and (late) 2.4 suggests that some
> versions of gcc might work better than others, but I don't have
> methodical notes.

Sun4d SMP has never worked. I spent many hours trying to figure out
why, and never managed to achieve a stable system. In the end I
concluded that the only viable strategy was to go back to the 2.3.x
release where sun4d/smp was first included, fix some bugs, and then
sync the evolving sun4m specific code into sun4d one step at a time.
This is unfortunately what happens when a minority interest system
is allowed to bitrot for so many years. Part of the problem is the lack
of Sun4d systems in the hands of kernel developers.

AFAIK Sun4m/SMP and Sun4c are currently in a reasonable state
and should not be difficult to keep going.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Retiring the sparc32 port

[Chris Newport]

BERTRAND Joël wrote:


andrew holway wrote:


just thinkin, I don't think a sparc32 chip has been released in more
than 12 years. Surely these cannot be energy efficient machines ;)



And LEON processor ? A sparc V8 that can be written in a FPGA ? It 
runs with Linux. Berkeley university has a work in progress on a super 
computer that uses sparc32 too.


Why does a Linux distribution need the latest bleeding edge kernel ?
With no new hardware to support it should be easy to put together a
distribution with the last known good kernel and the latest applications.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Dropping sparc32 for lenny

[Chris Newport]

Mark Morgan Lloyd wrote:


I would be if I had the time and the reference materials to allow me to build up
the skills, neither of which are true.

 


Most if not all of the reference materials are available online, David
knows where to find them.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Dropping sparc32 for lenny

[Chris Newport]

Mark Morgan Lloyd wrote:


/Why/ are there problems with sparc32? Because nobody really understands the
architecture except Sun. Why are there MMU miss issues on sparc64? Same reason.
What chance do we have of a fix? Slim since Sun says the hardware's obsolete.
Who does this affect? First time users, and casual developers who want to test
that their code works on SPARC. What will the result be? Development tools
languish, Linux on SPARC dies.

Maybe I'm being pessimistic, and in a few months time everybody finds that
they've got enough cash in their pockets to go out and buy T1 systems.

 


You are misinterpreting Sun's position.
Sun provides hardware and Solaris support for all Sun machines
until at least 10 years after the last of the type was offered for sale.
This is one helluva lot longer than most other companies.

I have Ultra 2 and Ultra 5 machines here, both are supported in both
Solaris 10 and Linux. Both can be found on EBay, usually selling for
less than the shipping cost.

All 32 bit Sun machines have been obselete for more than 10 years,
the last example being the Sparc 5 which was kept in production
artificially until the Ultra 5 had been qualified for use as the SSP on
E1 systems.

Despite this, the architecture is well documented and Sun have been
very helpfull in tracking down any information that has been requested.

Sun assists the 64 bit Sparc port of Linux, and has done for
several years.

The MMU miss issues come in 2 varieties, some 400MHz processor
modules contain iffy cache control chips which were supplied by IBM,
these modules should have been replaced. In virtually all other cases
there will be a bug in the kernel code. I am not aware of any current
bugs so this should not be an issue.

The real problem with Sparc32 is simple, the maintainer is not doing the
job and nobody has volunteered to take over. Without an active kernel
maintainer the port is suffering from bit rot and is doomed.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: md (software) RAID missing in the installer for sparc

[Chris Newport]

On Tue, 17 Apr 2007, Josip Rodin wrote:


4) If the partition table on the disk was first written by an old
   version of the Solaris format command the maximum useable size of
   any partition will be either 1Gb or 2Gb depending on version.
   Linux will probably not notice and allow you to create a larger
   partition which will confuse the OBP even if the OBP does not have
   the limitation in 2) above.
   The workaround for this is to destroy cylinder zero by filling it
   with zeros using dd and then use Linux fdisk to create a new BSD
   table.  [ Affects disks initially formatted with old utilities ]


I had Solaris on it previously, but I only had Linux parted and fdisk
trample over it several times, I never filled it with zeros. Is that really
necessary? :)


Yes, The table is only created when the disk is first labelled.
The table does not change, only the entries in the table.
The table contains version information which must be preserved because
the data format changed between versions to allow larger pointers.
The same problem also shows up on IDE disks, if older utilities were
used to label the disk it will always have a maximum size of 8Gb or 37Gb
until you zap the table and make a new one.

There is also a problem with SILO which I do not fully understand, but
it can also have problems looking more than 2047 Mb into a partition.
This can sneak up on you if you edit any of SILO's files, the new file
can be past the limit and make the system unbootable.

It is always best practice to have a small /boot partition at the
beginning of the disk.

BTW - If you are using RAID for your data you should also mirror swap.

Do not use anything other than a simple mirror for swap and do not have
more than one swap area on the same spindle. Both Solaris and Linux will
try to optimise performance by striping swap data across as many swaps
as you have, so having more than one swap per spindle will guarantee
serious head thrashing.

Maybe someone more knowledgeable will explain the SILO problems.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: md (software) RAID missing in the installer for sparc

[Chris Newport]

On Tue, 17 Apr 2007, Josip Rodin wrote:


I wasn't able to reproduce the working condition with a 3.5 GB root (/) md0
partition, or a 2 GB root md0. :( I've finally created a 256 MB /boot md0,
and a 5 GB root (/) md1 partition, and finally PROM found SILO.

Someone who knows PROM and/or SILO could explain which of those variables
causes the boot process to fail.


There are several factors involved.

1) The OBP does not understand raid, but it should be able to
   cope with a simple mirror.

2) Older OBPs cannot see beyond either 1Gb or 2Gb into a partition
   depending on version. If your installer puts anything needed for
   the initial boot process beyond this point you will not be able
   to boot. [ Affects Sparc32 ]

3) The first cylinder of the disk contains the partition information.
   Most filesystems work around this by not actually using the first
   cylinder, but Linux Raid does not understand this and it will
   overwrite the tables. The solution is simple - just create a small
   partition of one cylinder at cylinder zero and do not use it.
   [ Affects Linux Raid and possibly oddball filesystems such as XFS
 only when they start at cylinder zero ]

4) If the partition table on the disk was first written by an old
   version of the Solaris format command the maximum useable size of
   any partition will be either 1Gb or 2Gb depending on version.
   Linux will probably not notice and allow you to create a larger
   partition which will confuse the OBP even if the OBP does not have
   the limitation in 2) above.
   The workaround for this is to destroy cylinder zero by filling it
   with zeros using dd and then use Linux fdisk to create a new BSD
   table.  [ Affects disks initially formatted with old utilities ]

You could have any combination of the above - best to avoid them all.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Please test the new kernel (2.6.18.dfsg.1-9)

[Chris Newport]

Stuart Brady wrote:


#394047 seems to include a report of the same problem I'm having.  My
suspicion is that udev is getting confused by the two lance interfaces
using the same MAC address.

If I remove z25_persistent-net.rules and z45_persistent-net-generator.rules
from /etc/udev/rules.d, the problem goes away.

 


Try setting local_mac_address to true in the OBP.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Installation manual updates -- still open

[Chris Newport]

On Sun, 31 Dec 2006, Frans Pop wrote:




3.5.1 mentions the 1GB magic boundary -- silo will sometimes refuse to
boot the kernel, if it is physically located beyond the 1GB from the
start of the disk. I believe, however that this applies only to
sparc32 machines, is that correct?


Anyone know this?


Early Sparc32 machines assume an early disk VTOC limitation. The root 
filesystem must be on the first slice and no slice should exceed 2Gb.
This is an OBP issue. The only cure is to put the root (including /boot) 
filesystem on the first slice and limit its size to 1024Mb.
AFAIK this only affects Sun4C machines which may not work with 2.6 kernels 
anyway.



5.1.4: How relevant is Sun NVRAM FAQ linked from this section for us?


Any comments?


Most older machines will have expired NVRAM batteries which should be 
fixed using this FAQ before attempting to install an operating system.

Probably the most frequent problem with older machines.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Boot sparc 60

[Chris Newport]

Martin wrote:


FWIW I'm not sure Ultra 60s can boot from floppy.

 


Only pre-ultrasparc machines can (officially) boot from floppy.
Modern linux kernels are also too big to fit on a floppy.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: fast data access MMU miss

[Chris Newport]

[EMAIL PROTECTED] wrote:


Hello

I have a fast data access MMU miss
since this morning when I am trying to
boot with the cdrom of my U60.
Searching the net , it seems the motherboard
is defective, Is someone can confirm me the this
state of fact ?

 


This is most often caused by a faulty CDROM or drive.
The OBP has minimal intelligence, so it reads the initial boot sector 
and executes it without checking for corruption. This  inevitably leads 
to an illegal memory reference  and thus to the message that you see.


Try using a known good Sun pressed CD, not a CDR.
If this fails your drive has probably expired, if you do not have a 
spare try booting from the network.
The older drives in the U60 will probably have problems reading CDRs 
written at high speed and will almost certainly not read a CDRW.





--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Booting "Etch" on older hardware

[Chris Newport]

Mark Morgan Lloyd wrote:


Chris Newport wrote:

 


I did some work on 2.4, so late versions should run in UP mode.
SMP was still broken but bootable with kbd&mouse, serial console broken.
The FC stuff is broken since the 2.2 to 2.4 transition, so you will have
to use SCSI disks.
   



2.4.32 appears to run OK unmodified with 8-way SMP on my sun4d (SPARCserver
1000E) over a serial tty although I've had a watchdog timeout, as expected Pluto
doesn't work.
 

My patches made it into the kernel somewhere around 2.4.29, there were 
serial issues which may have been helped by later sparc32 work.
Pluto can be made to work with 2.2.x , you will need to read my writeup 
at http://www.splack.org/fibrechan to fix the pcode firmware.
You might then be able to track down the changes needed to get to 2.4, 
but it will not be easy.


Others have since warned me not to boot from the FC interfaces, the OBP 
will overwrite the pcode with an old version, wrecking the above update. 
This might also happen if you probe-fc-all, so be warned that you may 
need to reapply the pcode fix.


I used 3 SCSI disks when I was working on this, with Solaris 8 for the 
pcode, 2.2 stable, and 2.4 tests.



Performance is significantly better than 2.2. I'm tempted to try getting Sarge
onto the system but I don't know whether there is a prebuilt ISO and the
tightest kernel I can build is very close to the maximum bootable.

 

This is an issue with all sparc32 systems, you need to be rather 
ruthless in pruning your kernel config and initrd. There have been calls 
to drop sparc32 from 2.6 for this reason, but some people have managed 
to make it work.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Running Vmware onto Debian R31r2

[Chris Newport]

Kevin Fullerton wrote:


Hi,

I believe that VMWare will only work on x86 processors, not on SPARC
processors.  Same goes for Xen.

 

Linux-VServer is a kernel-based virtualisation which should work OK on 
Sparc. YMMV, I have not personally tested it.

http://linux-vserver.org/


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: X on ultra sparc 30

[Chris Newport]

Jeffrey Maples wrote:

have a Sun Ultra 30 creator with a strange graphics card. I have been 
head over heals trying to figure out what it is. I have taken it out, 
and found nothing on it that tells me what chipset it is. 



Take the card out and look for barcode stickers. Under the barcode will 
be a readable number, usually starting 501 or 370.

Add a dash after the first 3 digits then add the next 4 digits.
This gives you a part number (like 501-2325) which you can search for at 
http://sunsolve.sun.com/  or post here and ask.


When we know what card you have we should be able to suggest the 
appropriate driver.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Xorg 7 on sparc64

[Chris Newport]

Ludovic Courtès wrote:


The patch in question fixes problems encountered for applications using
`mremap ()'.  In the case of X, it may be a lower-level, specific
problem (I'd have said DMA or some such, but I am not familiar at all
with how X interacts with the kernel and video RAM).

 


FWIW there are 2 variants of Ultra5 video.
Early models only had enough video RAM for 8 bit video, later ones were 
24 bit capable.


Making the wrong assumptions in a driver could cause problems, is this 
bug reproducible on both variants ?.


AFAIK all 270MHz models are 8 bit, but I am not sure about when the 
change happened.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Video Card

[Chris Newport]

Earl Violet wrote:


I am relatively new to Debian and Sparc. I recently obtained 2 Tatung
COMPstation U10s. The first part of the install of Debian stable went
well; hooked to the home network, got security updates, etc.. When I
tried to install x I found Debian couldn't identify the Video board. 


It is a Sun board. It fits into the video bus on the mother board. I
can't find any identification I recognize as such. Here is the
information I have.

The board is 12" long and 4" high. It has a Sun Monitor adapter and a
female round adapter with 7 round pin receptors around a rectangular
pin recepticle.

These are the following numbers I can find on tags on the board:
04301U2903
5014788164377
06REV50
 


501-4788 is a Creator 3D Series 3 FFB2+ 24 bit color frame buffer.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: SS2 with Boot 2.6

[Chris Newport]

Dieszel wrote:

   Can anyone tell me if Sarge can be installed on an SS2 (it will be 
a system logger) with Boot 2.6 ?  Or can I upgrade it to Open Boot ? 
If neither what is the newest version of SunOS or Saolaris that will 
run on it. 
Thanx


Your biggest problem is memory.
If you have 16 or 32Mb Solaris 2.6 is the limit
With 64Mb you can go up to Solaris 7.

64Mb is a full house of 4Mb SIMMs, which can be hard to find.

In general, Linux is horribly slow on Sun4c due to task switching 
issues, you would be better advised to stay with Solaris or BSD.




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Sun Enterprise E450 and Sun Storage A1000, troubles on install...

[Chris Newport]

Do you have the A1000 connected to a DIFFERENTIAL interface ?.
Your A1000 will not work with the standard Fast/SE interfaces in
the E450. Differential interfaces are clearly marked on the
connector panel.

The A1000 is HVD (high voltage differential). You can damage both
the A1000 and the interface by connecting it to a single ended or LVD
connector.


Evgeny Yugov wrote:


Hello.

On boot:
...
SILO Version 1.4.9
\
boot: linux ramdisk_size=16000
Allocated 8 Megs of memory at 0x4000 for kernel
Loaded kernel version 2.4.27
Loading initial ramdisk (3041649 bytes at 0x87802000 phys, 0x40C0 virt)...
-
Remapping the kernel... Booting Linux...
...

System freeze. :o(

Anyone help me?

 




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Ultra 10 processor support

[Chris Newport]

Also you need to check the main board and OBP revision levels for 
compatibility with faster processors. Memory speed is also an issue with 
400MHz processors.


The 333 MHz processor is one of the fastest due to a large internal cache.


Scott Walker wrote:


Simple web search would have brought up the fact that no you cannot add
2 CPU's into a Ultra10.

Chris Andrew wrote:
| Hi, all.
|
| I have a 333 Ultra 10.  I was just wondering whether it is possible to
| add more processors.  I think the max speed processor I can get is 440,
| but 2 processors (or more) would be good.
|
| Any thoughts?
|
| Thanks,
|
| Chris.





--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Sun Enterprise 4000 insmod problem

[Chris Newport]

Chris Brandstetter wrote:


Dear All,
   I have a Sun Enterprise 4000 (e4000, e4K) that I am trying to get
Debian Sarge to run on.  Now I installed debian on the hard drive
using a SS2 Dual Proc so that it would install the smp kernel.  I did
a base install and only added a couple things like Kernel Source and
Tree.  It boot's fine, but won't allow me to insmod the sunhme or any
net driver as it gives this error unresolved symbol in $driver
crc32_le_Rsmp_92ea4ae4.  It is running the generic
2.4.27-2-sparc64-smp.  Version gives me "Debian 1:3.3.5-13".  The
config of the e4K is 6 processors (It's kinda col having six penguins
at the top of the boot screen :-) )  and 3GB Ram over 3 Mainboards. 
There are two SBUS boards, and it is the 66MHz backplane and

clockboard.  I have 2 Disk boards but they are not installed.  Any
help would be appreciated.
 

Try the sunlance driver. Early E4000 systems had le ether ports, later 
models had hme.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Ultra 10 max HDD capacity?

[Chris Newport]

The physical limit is 137Gb, which effectively limits you to the nearest 
normal size of 120Gb. I have a 120Gb drive working OK here.


For more capacity you would need to go to external SCSI devices.

Chris Andrew wrote:


Hi, all.

Sun suggest that the max HDD capacity for Ultra 10's is 20.4 Gb, is 
there any reason why (under linux), we can't use bigger?  I don't need 
bigger, but it's easier to buy big than small, these days.


Cheers,

Chris.




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Drivers socal on E3500

[Chris Newport]

Mark Morgan Lloyd wrote:


Chris Newport wrote:

 


The last I heard the soc and socal code is unmaintained and did not make
the scsi changes from 2.2.x to 2.4.x.
IOW, if you want to use soc or socal you need to use a 2.2.x kernel or
fix the drivers.

The easy workaround is to fit a SCSI card and forget about soc or socal.
You will need to tell the OBP to boot from the scsi interface.
   



Noted and aware of that but does anybody have soc working at all? I've got
2.2.20 on a SPARCserver here and even with the firmware patch (assuming I got it
right, and noting the quoting problems in Sebastien's posting) I get

 


I booted from SCSI and then loaded the soc and pluto modules.
You also need to read my instructions at
http://www.splack.org/fibrechan

This worked for me, YMMV. Here be dragons.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: saving sparc for etch requalification

[Chris Newport]

David Arnold wrote:


 Chris> The Ecache problems only affected 400MHz Ultrasparc II
 Chris> modules as used in Enterprise systems.  300 MHz modules were
 Chris> not affected. UltrasparcIIi modules as used in the Ultra 5/10
 Chris> were not affected.

We had large numbers of U10s installed on desktops in my previous
organisation.

We didn't use the 300MHz CPUs, found the 333MHz CPU modules to be
stable and the 440MHz were good for about a year of 24/7 and then
they'd start to fail.

 


This is a different issue caused by a bad batch of electrolytic capacitors
on the CPU module leaking and drying out. The culprits are easy to see
by visual inspection and fairly easy to replace.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: saving sparc for etch requalification

[Chris Newport]

Romain Dolbeau wrote:


If the instability affects a 300 Mhz UII module (typically
in an U10 or an U5), then it might very well be faulty hardware.
A lot of modules fitting that description suffered serious
cache problems ; the symptoms were transient uncorrectable
failure in the L2 cache, resulting in a kernel panic.


The Ecache problems only affected 400MHz Ultrasparc II  modules
as used in Enterprise systems. 
300 MHz modules were not affected.

UltrasparcIIi modules as used in the Ultra 5/10 were not affected.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Drivers socal on E3500

[Chris Newport]

Sebastien LANGE wrote:


Hi,

I try to install Sarge with net boot but the installer doesn't find the
driver for fibre channel scsi controller.
It's normal because the module is not present so I've compile the module
socal with the microcode.

I've extract the microcode:

$ objdump -x socal

socal: file format elf64-sparc
socal
architecture: sparc:v9, flags 0x0011:
HAS_RELOC, HAS_SYMS
start address 0x

Sections:
Idx Name  Size  VMA   LMA   File off
Algn
0 .text 6e44      0040  2**2
   CONTENTS, ALLOC, LOAD, RELOC, READONLY, CODE
1 .rodata1  0452      6e88  2**3
   CONTENTS, ALLOC, LOAD, READONLY, DATA
2 .data  000103d0      _72e0_  2**3
   CONTENTS, ALLOC, LOAD, RELOC, DATA
3 .data10782      000176b0  2**3
   CONTENTS, ALLOC, LOAD, DATA
|
|
|

03cc g O .data  0001 socal_ucode
000103cc g O .data  0004 socal_ucode_end


$ dd if=socal of=socal.ucode bs=1 skip=30380 count=64K

72e0 + 03cc = 76ac in hexa and in decimal => 30380 (skip)

Then I have to do:

cat socal.ucode | od -t "x1" | grep ` ` | sed `1i static __u32
socal_ucode [ ] __ initdata = { ;s/^[0-9a-f]* /0x/;s/
/,0x/g;s/$/,/;$a};` > socal_asm.h

After I uncomment HAVE_SOCAL_UCODE in socal.c and I've recompile the module.

If load the module, I've not message and I can only to make stop-a.

Where is the problem ?

Excuse me for my bad english :-(
 

The last I heard the soc and socal code is unmaintained and did not make 
the scsi changes from 2.2.x to 2.4.x.
IOW, if you want to use soc or socal you need to use a 2.2.x kernel or 
fix the drivers.


The easy workaround is to fit a SCSI card and forget about soc or socal.
You will need to tell the OBP to boot from the scsi interface.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Problems with an Ultra 5

[Chris Newport]

What monitor are you using ?
This sounds typical of a monitor which cannot handle Sun's default 
resolution of 1172 x (796 ? not sure).


Try using a serial console (you must unplug the keyboard for this to work)
or a different monitor. About half of my PC monitors will not work with 
Sun machines.


On Fri, 5 Aug 2005, Tyler wrote:


Date: Fri, 05 Aug 2005 01:44:19 -0700
From: Tyler <[EMAIL PROTECTED]>
To: debian-sparc@lists.debian.org
Subject: Problems with an Ultra 5
Resent-Date: Fri,  5 Aug 2005 03:46:16 -0500 (CDT)
Resent-From: debian-sparc@lists.debian.org

I'm trying to set up an Ultra 5 that I purchased last year, but am having 
problems getting anything out of it.  I've hooked it up to power, network, 
video and sun type 6 keyboard/mouse (not usb).  I've then turned it on, a few 
seconds goes by with the monitor not doing anything, then, the monitor gets a 
signal, and comes out of power save mode, but this only lasts 5 or 10 
seconds, before it goes back to power save mode (no signal) (and never gets a 
signal again), so this isn't long enough to see anything on the screen 
because the screen isn't on long enough (doesn't warm up) to actually see the 
image.


The system sounds like its doing something still, as the 8.4gb IDE hard drive 
chugs away for 30 seconds to a minute, before becoming quiet again.  At this 
point.. I don't know what (if anything) I can do .. without rebooting it 
again (is there a key combo for rebooting?).  I've tried turning it on with a 
debian sarge net install bootable CD, and a solaris 9 install cd, both do the 
same thing as above.


At this point, I'm looking for help (documentation) on setting up, and 
troubleshooting an ultra5/10 system, lists of key combinations that may help 
me (I've read that STOP + A during power on may get me to the OBP open boot 
prompt, however, I haven't had the chance to try this option yet, and am not 
really sure what settings may need changing once i'm there, to stop the video 
from disabling itself, or whatever its doing).


Thanks,
Tyler.

PS - My goal is to get debian installed.  If someone can point me to some 
*specific* docs to troubleshoot such linux un-related problems on a sun 
machine, so I can actually get to installing linux, I would appreciate it. 
Also, there's been a few messages on the mailing list that seem to point out 
that the standard debian net install CD is not useable on an ultra5 system... 
if this is true, can someone please point specifically to one that will work, 
for when I actually get to the point of installing :)







--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Debian Woody running SMP on a SPARCserver 1000E (sun4d)

[Chris Newport]

Mark Morgan Lloyd wrote:


Around a year ago I acquired a number of SPARCserver 1000E machines,
[snip]
I have not made similar progress with a scratch-compiled 2.4 kernel.
[snip]
 

I put rather a lot of work into this a while back and my patches are in 
2.4.30, there are still some outstanding issues :-


Serial console does not work - use a frame buffer and keyboard.
UP is stable but SMP still crashes, looks like an IRQ handling problem 
in SMP mode but ICBW.


I ran out of time and skills, so I have not done any more work, if you 
would like to pick up the ball and try running with it I will try to 
help you get started.


In general Sun4d is very similar to sun4m, if you look at the patches 
that have been applied for sun4m but not sun4d this will probably fix a 
few things.




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: problems using a printer at my ultra5

[Chris Newport]

Dr. Zimmermann wrote:


Hi,

Id'd like to connect a printer (via parallel or usb connectors) to my 
ultra5.


Both connections (parallel and usb) lead to problems  ...

Parallel:
After modprobing parport_pc the printer is successfully  autoprobed
(/proc/sys/dev/parport/parport0/autoprobe  shows the  correct printer 
type)

The printer can be installed in cups without errors ... If I try to use
it there were messages at the console and in the syslog:
lp0: ECP mode
... but nothing is received by the printer ('data' light stays dark)

As stated in the kernel documentation I switched off the FIFO option
and try to:
insmod parport_pc irq=none

still no data arrives via parallel port ...
The parallel port on the ultra5 seems to be a black whole !

USB:
Trying the USB-port (cat ... > /dev/usb/lp0) transmits some data to 
the printer (i.e.
'data' light is blinking) - due to incomplete ioctl calls on sparc64 
linux I receive

errors when I try printing using cups.
(see also http://lists.debian.org/debian-sparc/2004/11/msg00193.html)


I've used severel 2.6 kernels for testing ...
Any help ?

Roger


The parallel printer port driver has not worked since maybe 2.2.x, this 
is a known problem. I use a network connected printer (HP2250CN) here OK.

Sorry - I have not tested USB so I cannot help there.

First make sure that you have not installed lprng, if you install lprng 
and cups then neither will work. You will need to remove both and then 
re-install cups.



--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: apache and CGI :segfault !

[Chris Newport]

Lindo Nepi wrote:


on my debian/sparc:
i try to launch a simple CGI (write in bash) from apache...
this CGI launch gradsc (a scientific application that generate gif 
images) ...but in my webpage i obtain an exit code 11


on my debian/x86
no problem with the same configfuration.. gradsc works fine.


Did you recompile the GrADS programs for Linux-Sparc ?.
I cannot see a supported version on their website.
Forget about your webpages and get the GrADS suite working properly 
first. You will need to recompile from the sources.




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Sarge Release Notes - Architecture specific news?

[Chris Newport]

Chris Newport wrote:
I will try Sarge in the next few days and report back, but I suspect 
that I will need to upgrade to 2.4.29 or later.
AFAIK nobody has done any work on 2.6.x for Sun4d.

Sarge does not boot on Sun4d - this looks like a SILO issue, but I could 
be wrong. My SS1000E has never booted from a SILO  CD correctly, but it 
does boot  from a TILO tftpboot image  for Splack and it does boot a 
Solaris CD.

Loaded kernel version 2.4.27
Data Access Exception
Type help for more information
<#0> ok

--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Sarge Release Notes - Architecture specific news?

[Chris Newport]

inkubus wrote:
Sun4d is the Sparcserver 1000 and 2000 series, they are 32 bit SMP
boxes. Very similar to Sun4m, they work fine in single processor mode
with 2.4.27 and above, but SMP is still badly broken. More kernel work
is required.
   

Have you directly tested >= 2.4.27 ?  Last time I checked everything from
about 2.4 onwards was kinda broken.  There is a fix for the 2.2.20 kernel
to get SMP working on machines with uniform clock speeds (IIRC each board
must have consistant speed processors but it can vary between boards).
In the past there has been quite a bit of interest on this list about
resurrecting sun4d support, all that appears to be missing is someone to
co-ordinate the effort.
When I tested it woody did not run 'out of the box' on sun4d.  I do not
know of anyone trying sarge.
 

All of my work was done in Splack, rather than Debian.
Looking at my notes - Sun4d UP needs my patches which are in 2.4.29
Sun4d SMP is still broken, and fixing it is beyond my skills.
There are 2 known issues, possibly related.
1) Serial console corruption
2) SMP instability seems to be IRQ related
I will try Sarge in the next few days and report back, but I suspect 
that I will need to upgrade to 2.4.29 or later.
AFAIK nobody has done any work on 2.6.x for Sun4d.


--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Re: Sarge Release Notes - Architecture specific news?

[Chris Newport]

Hendrik Sattler wrote:
This also affects the install-manual where "sun4cdm and sun4u" is mentioned as 
supported (What's sun4d and is it really supported?)
 

Sun4d is the Sparcserver 1000 and 2000 series, they are 32 bit SMP 
boxes. Very similar to Sun4m, they work fine in single processor mode 
with 2.4.27 and above, but SMP is still badly broken. More kernel work 
is required.

--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]