sparc sun4c support - time to drop it?

sparc sun4c support - time to drop it?

[Sam Ravnborg]

Is anyone having any sun4c boxes alive and running Linux?

I have a nice box myself that does not boot - I do not know why.
But we are talking abot slow and old stuff here.

Is it time to drop support for sun4c - or is anyone still
having an interest keeping that platform alive?

Looking into recent history of the kernel only one person
(Mark - copied) seems to have contributed
solely for the purpose of getting a sun4c box working.


Sunstuff has a list of boxes based on sun4c here:
http://sunstuff.org/hardware/systems/sun4/sun4c/

SPARCstation 1, SLC, IPC, 1+, 2, ELC, IPX
SPARCengine 1


	Sam

Re: sparc sun4c support - time to drop it?

[Alan Perry]

On 5/5/12 9:21 AM, Sam Ravnborg wrote:
> Is anyone having any sun4c boxes alive and running Linux?
>
> I have a nice box myself that does not boot - I do not know why.
> But we are talking abot slow and old stuff here.
>
> Is it time to drop support for sun4c - or is anyone still
> having an interest keeping that platform alive?
I do.

I collect the Sun lunchbox systems (IPC, IPX, Classic, Classic/X and 
LX), which includes sun4c machines.  I run Linux and *BSD on them.  But 
the Linux that I am running on them is pretty old.

I have interest in keeping the platform alive, but my day job (Solaris 
kernel developer at Oracle) prevents me from working on the platform.

alan

>
> Looking into recent history of the kernel only one person
> (Mark - copied) seems to have contributed
> solely for the purpose of getting a sun4c box working.
>
>
> Sunstuff has a list of boxes based on sun4c here:
> http://sunstuff.org/hardware/systems/sun4/sun4c/
>
> SPARCstation 1, SLC, IPC, 1+, 2, ELC, IPX
> SPARCengine 1
>
>
> 	Sam
>
> --
> To unsubscribe from this list: send the line "unsubscribe sparclinux" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: sparc sun4c support - time to drop it?

[Mark Fortescue]

On Sat, 5 May 2012, Sam Ravnborg wrote:

> Is anyone having any sun4c boxes alive and running Linux?

I have not had much time to do any more work on Linux + sun4c but I do 
have a sun4c box waiting for a new external disk to put Linux on (the last 
one stopped, I thing due to a ded PSU).

>
> I have a nice box myself that does not boot - I do not know why.
> But we are talking abot slow and old stuff here.
>
> Is it time to drop support for sun4c - or is anyone still
> having an interest keeping that platform alive?
>
> Looking into recent history of the kernel only one person
> (Mark - copied) seems to have contributed
> solely for the purpose of getting a sun4c box working.
>
>
> Sunstuff has a list of boxes based on sun4c here:
> http://sunstuff.org/hardware/systems/sun4/sun4c/
>
> SPARCstation 1, SLC, IPC, 1+, 2, ELC, IPX
> SPARCengine 1
>
>
> 	Sam
>
>

Re: sparc sun4c support - time to drop it?

[Sam Ravnborg]

On Sun, May 06, 2012 at 05:01:20AM +0100, Mark Fortescue wrote:
> On Sat, 5 May 2012, Sam Ravnborg wrote:
>
>> Is anyone having any sun4c boxes alive and running Linux?
>
> I have not had much time to do any more work on Linux + sun4c but I do  
> have a sun4c box waiting for a new external disk to put Linux on (the 
> last one stopped, I thing due to a ded PSU).

I have come a long way using tftpboot.img and a ram-dsik.
This allows me to keep Solaris on the disk, and booting
Linux. It is a bit slow as I always fetch the full kernel via the
network but this is acceptable.

I hacked busybox to support sparc32 - patches are upstream.

	Sam

Re: sparc sun4c support - time to drop it?

[Julian Calaby]

Hi Mark,

On Sun, May 6, 2012 at 2:01 PM, Mark Fortescue <mark@mtfhpc.demon.co.uk> wrote:
> On Sat, 5 May 2012, Sam Ravnborg wrote:
>
> I have not had much time to do any more work on Linux + sun4c but I do have
> a sun4c box waiting for a new external disk to put Linux on (the last one
> stopped, I thing due to a ded PSU).

I have one of the standard Sun 411 external narrow-scsi CD-ROM boxes.

If you're in Australia, I could ship it to you.

I don't need it anymore as I'm finding different solutions to my
storage issues with my sun4m boxen.

Thanks,

-- 
Julian Calaby

Email: julian.calaby@gmail.com
Profile: http://www.google.com/profiles/julian.calaby/
.Plan: http://sites.google.com/site/juliancalaby/

Re: sparc sun4c support - time to drop it?

[Kaj-Michael Lang]

On Sat, 2012-05-05 at 18:21 +0200, Sam Ravnborg wrote:
> Is it time to drop support for sun4c - or is anyone still
> having an interest keeping that platform alive? 

I'd probably be running my sun4c from time to time, unfortunately that
Really-Bad-Choice of clock+battery module that Sun made has "killed"
many of my older suns.

-- 
Kaj-Michael Lang <milang@tal.org>

Re: sparc sun4c support - time to drop it?

[Julian Calaby]

Hi Kaj,

On Sun, May 6, 2012 at 7:53 PM, Kaj-Michael Lang <milang@tal.org> wrote:
> On Sat, 2012-05-05 at 18:21 +0200, Sam Ravnborg wrote:
>> Is it time to drop support for sun4c - or is anyone still
>> having an interest keeping that platform alive?
>
> I'd probably be running my sun4c from time to time, unfortunately that
> Really-Bad-Choice of clock+battery module that Sun made has "killed"
> many of my older suns.

You know, with a few simple hand tools, a replacement battery, and
some knowledge of OpenFirmware, you can fix them.

See here:

http://www.squirrel.com/sun-nvram-hostid.faq.html

Thanks,

-- 
Julian Calaby

Email: julian.calaby@gmail.com
Profile: http://www.google.com/profiles/julian.calaby/
.Plan: http://sites.google.com/site/juliancalaby/

Re: sparc sun4c support - time to drop it?

[David Miller]

From: Sam Ravnborg <sam@ravnborg.org>
Date: Sat, 5 May 2012 18:21:04 +0200

> Is it time to drop support for sun4c - or is anyone still
> having an interest keeping that platform alive?

I have no objections and if we go this route I can make the the 32-bit
sparc configurations of GCC under Linux start generating V8 code by
default.

Re: sparc sun4c support - time to drop it?

[Kaj-Michael Lang]

On Sun, 2012-05-06 at 20:29 +1000, Julian Calaby wrote:
> You know, with a few simple hand tools, a replacement battery, and
> some knowledge of OpenFirmware, you can fix them. 

Yes, I know, I've done it for an ELC once. But I'm lazy :)

-- 
Kaj-Michael Lang <milang@tal.org>

Re: sparc sun4c support - time to drop it?

[crn]

On Sun, 6 May 2012, Kaj-Michael Lang wrote:

> On Sun, 2012-05-06 at 20:29 +1000, Julian Calaby wrote:
>> You know, with a few simple hand tools, a replacement battery, and
>> some knowledge of OpenFirmware, you can fix them.
>
> Yes, I know, I've done it for an ELC once. But I'm lazy :)

I am in favour of dropping Sun4c (and possibly Sun4d but thats a different 
story) for all new kernels.
All of the possible hardware combinations for Sun4c have been covered 
since 2.2.x and nothing new will happen. Modern kernels need more 
memory and these machines are all limited to 64Mb or less.

Those who want to keep thse antiques alive should just use the older 
kernels. BSD is faster on Sun4c anyway due to task swapping issues.

Re: sparc sun4c support - time to drop it?

[Mark Fortescue]

Hi,

Replacing the NVRAM is not that dificult and replacements are still 
available.

Reprogramming the hostid/mac address is a pain. I have some notes from the 
web as I can never remember how to do it.

If you re really adventurous you can just replace the battery. I replaced
one of mine with a rechargeable version so it will probably outlast
the rest of the hardware.

Regards
 	Mark.

On Sun, 6 May 2012, Kaj-Michael Lang wrote:

> On Sat, 2012-05-05 at 18:21 +0200, Sam Ravnborg wrote:
>> Is it time to drop support for sun4c - or is anyone still
>> having an interest keeping that platform alive?
>
> I'd probably be running my sun4c from time to time, unfortunately that
> Really-Bad-Choice of clock+battery module that Sun made has "killed"
> many of my older suns.
>
> -- 
> Kaj-Michael Lang <milang@tal.org>
>
> --
> To unsubscribe from this list: send the line "unsubscribe sparclinux" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: sparc sun4c support - time to drop it?

[Mark Fortescue]

On Mon, 7 May 2012 crn@netunix.com wrote:

>
>
> On Sun, 6 May 2012, Kaj-Michael Lang wrote:
>
>> On Sun, 2012-05-06 at 20:29 +1000, Julian Calaby wrote:
>>> You know, with a few simple hand tools, a replacement battery, and
>>> some knowledge of OpenFirmware, you can fix them.
>> 
>> Yes, I know, I've done it for an ELC once. But I'm lazy :)
>
> I am in favour of dropping Sun4c (and possibly Sun4d but thats a different 
> story) for all new kernels.
> All of the possible hardware combinations for Sun4c have been covered since 
> 2.2.x and nothing new will happen. Modern kernels need more memory and these 
> machines are all limited to 64Mb or less.

Not strictly true. My SPARCStation 1+ clone has a 4 port SBUS serial card 
that is not supported by Linux. :).

>
> Those who want to keep thse antiques alive should just use the older kernels. 
> BSD is faster on Sun4c anyway due to task swapping issues.
>
> --
> To unsubscribe from this list: send the line "unsubscribe sparclinux" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: sparc sun4c support - time to drop it?

[Sam Ravnborg]

On Sat, May 05, 2012 at 06:21:04PM +0200, Sam Ravnborg wrote:
> Is anyone having any sun4c boxes alive and running Linux?
> 
> I have a nice box myself that does not boot - I do not know why.
> But we are talking abot slow and old stuff here.
> 
> Is it time to drop support for sun4c - or is anyone still
> having an interest keeping that platform alive?
> 
> Looking into recent history of the kernel only one person
> (Mark - copied) seems to have contributed
> solely for the purpose of getting a sun4c box working.
> 
> 
> Sunstuff has a list of boxes based on sun4c here:
> http://sunstuff.org/hardware/systems/sun4/sun4c/
> 
> SPARCstation 1, SLC, IPC, 1+, 2, ELC, IPX
> SPARCengine 1

To summarize this thread...

There are unsuprisingly some that for sentimental
reasons would like to keep sun4c support in-kernel.

But sun4c is outdated and if anyone want to run something
on a sun4c box the *BSD variants has better performance.

Also considering that sun4c support has been broken for
some time and there is a few changes pending where sun4c
does not make things simpler we will drop sun4c support.

I already took a quick look at it and in the process
I think I uncovered a bug in NMI handling for sun4m
in UP mode. We would call the sun4c nmi trap despite
being a sun4m box - this would not go well.

When I have something tested I will forward my
patches.

	Sam

Re: sparc sun4c support - time to drop it?

[Tom Callaway]

On 05/07/2012 01:46 PM, Sam Ravnborg wrote:

> There are unsuprisingly some that for sentimental
> reasons would like to keep sun4c support in-kernel.
> 
> But sun4c is outdated and if anyone want to run something
> on a sun4c box the *BSD variants has better performance.
> 
> Also considering that sun4c support has been broken for
> some time and there is a few changes pending where sun4c
> does not make things simpler we will drop sun4c support.

I haven't powered on any of my sun4c (or older) kit in several years, so
I'm also supportive of this change.

~tom

=
Fedora Project

Re: sparc sun4c support - time to drop it?

[Mark Fortescue]

On Mon, 7 May 2012, Sam Ravnborg wrote:

> On Sat, May 05, 2012 at 06:21:04PM +0200, Sam Ravnborg wrote:
>> Is anyone having any sun4c boxes alive and running Linux?
>>
>> I have a nice box myself that does not boot - I do not know why.
>> But we are talking abot slow and old stuff here.
>>
>> Is it time to drop support for sun4c - or is anyone still
>> having an interest keeping that platform alive?
>>
>> Looking into recent history of the kernel only one person
>> (Mark - copied) seems to have contributed
>> solely for the purpose of getting a sun4c box working.
>>
>>
>> Sunstuff has a list of boxes based on sun4c here:
>> http://sunstuff.org/hardware/systems/sun4/sun4c/
>>
>> SPARCstation 1, SLC, IPC, 1+, 2, ELC, IPX
>> SPARCengine 1
>
> To summarize this thread...
>
> There are unsuprisingly some that for sentimental
> reasons would like to keep sun4c support in-kernel.
>
> But sun4c is outdated and if anyone want to run something
> on a sun4c box the *BSD variants has better performance.
>
> Also considering that sun4c support has been broken for
> some time and there is a few changes pending where sun4c
> does not make things simpler we will drop sun4c support.
>

Will Sun4c be dropped for 3.4 or 3.5 ?

(Just so that I can get my hands on a 'last supported' kernel to build 
when the network for my remote build host wakes up. :).)

> I already took a quick look at it and in the process
> I think I uncovered a bug in NMI handling for sun4m
> in UP mode. We would call the sun4c nmi trap despite
> being a sun4m box - this would not go well.
>
> When I have something tested I will forward my
> patches.
>
> 	Sam
> --
> To unsubscribe from this list: send the line "unsubscribe sparclinux" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: sparc sun4c support - time to drop it?

[Sam Ravnborg]

> Will Sun4c be dropped for 3.4 or 3.5 ?
>
> (Just so that I can get my hands on a 'last supported' kernel to build  
> when the network for my remote build host wakes up. :).)

The merge window for 3.4 is closed long time ago - so for
sure it is not dropped in 3.4.
It will be dropped in 3.5 I think - if I get the patches
prepared fast enough.

	Sam

Re: sparc sun4c support - time to drop it?

[Sam Ravnborg]

Hi Tom.

On Mon, May 07, 2012 at 01:59:11PM -0400, Tom Callaway wrote:
> On 05/07/2012 01:46 PM, Sam Ravnborg wrote:
> 
> > There are unsuprisingly some that for sentimental
> > reasons would like to keep sun4c support in-kernel.
> > 
> > But sun4c is outdated and if anyone want to run something
> > on a sun4c box the *BSD variants has better performance.
> > 
> > Also considering that sun4c support has been broken for
> > some time and there is a few changes pending where sun4c
> > does not make things simpler we will drop sun4c support.
> 
> I haven't powered on any of my sun4c (or older) kit in several years, so
> I'm also supportive of this change.

Thanks for your feedback!

	Sam

Re: sparc sun4c support - time to drop it?

[Julian Calaby]

Hi Sam,

On Tue, May 8, 2012 at 3:46 AM, Sam Ravnborg <sam@ravnborg.org> wrote:
> On Sat, May 05, 2012 at 06:21:04PM +0200, Sam Ravnborg wrote:
> Also considering that sun4c support has been broken for
> some time and there is a few changes pending where sun4c
> does not make things simpler we will drop sun4c support.
>
> I already took a quick look at it and in the process
> I think I uncovered a bug in NMI handling for sun4m
> in UP mode. We would call the sun4c nmi trap despite
> being a sun4m box - this would not go well.
>
> When I have something tested I will forward my
> patches.

Would it be possible to hide sun4c away behind a kconfig variable that
is marked as BROKEN with a entry in the feature-removal file?

Or is that just too much work for code which has been broken for a long time.

Thanks,

-- 
Julian Calaby

Email: julian.calaby@gmail.com
Profile: http://www.google.com/profiles/julian.calaby/
.Plan: http://sites.google.com/site/juliancalaby/

Re: sparc sun4c support - time to drop it?

[David Miller]

From: Julian Calaby <julian.calaby@gmail.com>
Date: Tue, 8 May 2012 09:07:34 +1000

> Would it be possible to hide sun4c away behind a kconfig variable
> that is marked as BROKEN with a entry in the feature-removal file?
> 
> Or is that just too much work for code which has been broken for a long time.

We just want to kill all of this code off, there is zero value keeping
it around at all.

It's in the revision history if you just want to look at it, and
nobody is going to make changes that try to keep it working so letting
this stuff linger has zero value.

Re: sparc sun4c support - time to drop it?

[Julian Calaby]

Hi David,

On Tue, May 8, 2012 at 9:16 AM, David Miller <davem@davemloft.net> wrote:
> From: Julian Calaby <julian.calaby@gmail.com>
> Date: Tue, 8 May 2012 09:07:34 +1000
>
>> Would it be possible to hide sun4c away behind a kconfig variable
>> that is marked as BROKEN with a entry in the feature-removal file?
>>
>> Or is that just too much work for code which has been broken for a long time.
>
> We just want to kill all of this code off, there is zero value keeping
> it around at all.
>
> It's in the revision history if you just want to look at it, and
> nobody is going to make changes that try to keep it working so letting
> this stuff linger has zero value.

I figured that would be the consensus.

I have no objection to annihilating the code - I just don't like
seeing code being thrown away.

Thanks,

-- 
Julian Calaby

Email: julian.calaby@gmail.com
Profile: http://www.google.com/profiles/julian.calaby/
.Plan: http://sites.google.com/site/juliancalaby/