Super/HyperSPARC (2.6.18.2)

Super/HyperSPARC (2.6.18.2)

[BERTRAND Joël]

	Hello,

	I use a SparcSTATION 20 to do some tests (network configuration and some
	others...). On one of these, I have two SuperSPARC-II/75 and this
	workstation runs a Linux 2.6.18.2 kernel (smp) without any trouble.
	The last week, I have changed both processors by a dual RT-626 module.
	The workstation boots and is usable, but, when I try to install a new
	package with apt-get, I receive :

hilbert:~# LANG=C apt-get install fbset
Reading package lists... Done
Building dependency tree... Done
The following NEW packages will be installed:
  fbset
0 upgraded, 1 newly installed, 0 to remove and 0 not upgraded.
Need to get 0B/103kB of archives.
After unpacking 336kB of additional disk space will be used.
Preconfiguring packages ...
dpkg: error processing /var/cache/apt/archives/fbset_2.1-19_sparc.deb
(--unpack):
parse error, in file `/var/lib/dpkg/tmp.ci/control' near line 3 package 'fbset':
field name `Section: admin
Priority: optional
Architecture: sparc
Depends: libc6' must be followed by colon
Erreur de segmentation
hilbert:~# 

	I have six different modules and I receive the same error with all modules.
	I have tested all modules under Solaris9 and they work fine and this error
	does not come from SMP code (I have tried with only one CPU).

	A bug in hypersparc MMU code ? But why the only error I can see is with
	apt-get ?

	Regards,

	JKB

Super/HyperSPARC (2.6.18.2)

[BERTRAND Joël]

	Hello,

	I use some SS20 to tests (as routers). All of these workstations have 2
	SSII/75, but I have some single and dual RT626.

	On one SS20, I have changed SuperSPARC by HyperSPARC. It runs with a
	2.6.18.2 linux kernel and a debian/testing. With ROOS, I can see a very
	strange mistake : apt-get returns a random error (only occurs with
	hypersparc, UP and SMP).

	I have tested all motherboards and modules I have used for test with Solaris
	and netbsd. All work fine. I don't understand why I only see this bug with
	apt-get. Is there any issue ? A bug in hypersparc specific code ?

	Regards,

	JKB

Re: Super/HyperSPARC (2.6.18.2)

[Ulrich Teichert]

Hi,

>	I use some SS20 to tests (as routers). All of these workstations have 2
>	SSII/75, but I have some single and dual RT626.
>
>	On one SS20, I have changed SuperSPARC by HyperSPARC. It runs with a
>	2.6.18.2 linux kernel and a debian/testing. With ROOS, I can see a very
>	strange mistake : apt-get returns a random error (only occurs with
>	hypersparc, UP and SMP).
>
>	I have tested all motherboards and modules I have used for test with Solaris
>	and netbsd. All work fine. I don't understand why I only see this bug with
>	apt-get. Is there any issue ? A bug in hypersparc specific code ?

What PROM version do you use? I had strange errors as well until I updated
the PROMs,

HTH,
Uli
-- 
Dipl. Inf. Ulrich Teichert|e-mail: Ulrich.Teichert@gmx.de
Stormweg 24               |listening to: Cauchemar (Opération S)
24539 Neumuenster, Germany|Good Looks, Big Deal (Sweatmaster)

Re: Super/HyperSPARC (2.6.18.2)

[BERTRAND Joël]

Ulrich Teichert a écrit :
> Hi,
> 
> 
>>	I use some SS20 to tests (as routers). All of these workstations have 2
>>	SSII/75, but I have some single and dual RT626.
>>
>>	On one SS20, I have changed SuperSPARC by HyperSPARC. It runs with a
>>	2.6.18.2 linux kernel and a debian/testing. With ROOS, I can see a very
>>	strange mistake : apt-get returns a random error (only occurs with
>>	hypersparc, UP and SMP).
>>
>>	I have tested all motherboards and modules I have used for test with Solaris
>>	and netbsd. All work fine. I don't understand why I only see this bug with
>>	apt-get. Is there any issue ? A bug in hypersparc specific code ?
> 
> 
> What PROM version do you use? I had strange errors as well until I updated
> the PROMs,

	I use a regular 2.25R from ROSS technology. Same trouble with 2.25 
(that doesn't officialy support RT626). I have remounted the SS20 with 
two SS-II and I shall test on another SS20 with a 2.25W from ROSS 
(hyperstation).

	Regards,

	JKB