Time to let sun4m + sun4d RIP?

Time to let sun4m + sun4d RIP?

[Sam Ravnborg]

Hi all sparc lovers/users.

We all know that is is a long time ago that a sun4m or sun4d machine was produced,
and I wonder if there are anyone using these today.
The questions pops up as there are still stuff to do before the sparc32 port
is in good shape.

So the question to you guys...
Shall we keep support for sun4d and sun4m?
If yes then why?

LEON is in active use in several applications - so the sparc32 port is
as relevant as ever.

	Sam

Re: Time to let sun4m + sun4d RIP?

[Julian Calaby]

Hi Sam,

On Sat, Apr 19, 2014 at 9:48 PM, Sam Ravnborg <sam@ravnborg.org> wrote:
> Hi all sparc lovers/users.
>
> We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> and I wonder if there are anyone using these today.
> The questions pops up as there are still stuff to do before the sparc32 port
> is in good shape.
>
> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?

I have three sun4m boxen, all of which are unbootable for various
reasons (no hdd being the most common issue).

The last time I attempted to get "modern" Linux booting on them, the
major issue I ran into was the lack of modern distro support for
sparc32. - Gentoo was the only viable option (They last supported it
in 2006, however the last time I checked, that version was upgradable,
pre-emerge world, to the latest version). The last HDD set I had died
before I could start actually building it, so I've not actually seen
if the Linux kernel actually works properly on that hardware anymore.
(All three boxen need narrow-SCSI HDDs. I no-longer have any which I
know work. Netboot is an option, but I've not explored it enough yet.)

Sadly, getting them running again has been stuck on the bottom of my
TODO list due to a lot of other factors. (Lack of time being the main
one.)

If it's dropped, I won't mourn it's loss, however my vote will always
be to let it bitrot a bit more before we kill it.

IIRC, CPU-wise, sun4{m,d} is not _that_ much different from LEON,
however all the hardware stuff is, of course, significantly different.
If that's not holding LEON back....

Thanks,

-- 
Julian Calaby

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

Re: Time to let sun4m + sun4d RIP?

[Alan Perry]

> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?

I collect Sun lunchbox systems and also have an Axil SS20 clone.  I start up the lunchbox systems once a year to make sure they haven't died and regularly run the Axil. One of the lunchbox systems is running some Linux distro that I don't recall (it could be something old on a sun4c).

Number one problem keeping them running long term is the NVRAM battery (I have added external batteries to the chips in almost all of them). Replacement SCSI HDDs are now a problem to source locally; I couldn't get any from the computer recycler last year. I can still get them on eBay.

alan

Re: Time to let sun4m + sun4d RIP?

[Magnus Lindholm]

Hi all,

I'm still using a ss10 and a ss20 with a fairly current kernel and a
home-brew gentoo linux for sparc32. It's mostly for fun but I try to
keep my gentoo up to date and follow what's going on with the sparc32
kernel. I don't have access to LEON sparc hardware and LEON boards
seems to expensive to get just for fun. but would still love to keep
running current linux kernels on real sparc hardware. I guess SMP on
sparc32 has some issues and probably the most recent linux kernel is
even too big to boot on sparcstation hardware but UP still works fine.

I really hope that the Linux kernel can continue to run on sun4d and
sun4m hardware. Not sure if "because it's fun" is a good enough reason
but I hope so!

Regards


Magnus Lindholm


On Sat, Apr 19, 2014 at 1:48 PM, Sam Ravnborg <sam@ravnborg.org> wrote:
>
> Hi all sparc lovers/users.
>
> We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> and I wonder if there are anyone using these today.
> The questions pops up as there are still stuff to do before the sparc32 port
> is in good shape.
>
> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?
>
> LEON is in active use in several applications - so the sparc32 port is
> as relevant as ever.
>
>         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: Time to let sun4m + sun4d RIP?

[Olivier Danet]

On 19/04/2014 13:48, Sam Ravnborg wrote:
> Hi all sparc lovers/users.
> 
> We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> and I wonder if there are anyone using these today.
> The questions pops up as there are still stuff to do before the sparc32 port
> is in good shape.
> 
> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?
> 
> LEON is in active use in several applications - so the sparc32 port is
> as relevant as ever.
> 
> 	Sam

I wish Linux could keep the Sun4m port a bit longer, even if I cannot help much, I just posted a small patch last year.

There is a project for the recreation of Sun4m hardware in a FPGA : http://temlib.org.
Running current Linux kernel on this hardware could be interesting. It would also be preferable as a base for patchs to optimise
the kernel to TEMLIB's CPU which emulates legacy Sun CPUs (MicroSparc) but is not exactly identical to them
(different TLBs organisation, for example).

QEMU currently supports Sun4m quite well, despite some issues with SCSI emulation.
All operating systems can be run, except NextSTEP : Linux, *BSD, Solaris are supported.
It is now much faster than genuine Sun hardware (particulary for disk accesses !).

Regards
O.Danet

Re: Time to let sun4m + sun4d RIP?

[David Miller]

From: Sam Ravnborg <sam@ravnborg.org>
Date: Sat, 19 Apr 2014 13:48:02 +0200

> We all know that is is a long time ago that a sun4m or sun4d machine
> was produced, and I wonder if there are anyone using these today.
> The questions pops up as there are still stuff to do before the
> sparc32 port is in good shape.
> 
> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?
> 
> LEON is in active use in several applications - so the sparc32 port
> is as relevant as ever.

I think while qemu is around, it is feasible for people to try to keep
maintaining the port.

Re: Time to let sun4m + sun4d RIP?

[Jim MacKenzie]

On 14-04-22 07:21 PM, David Miller wrote:
> I think while qemu is around, it is feasible for people to try to keep 
> maintaining the port.

I still have two running sun4m machines and would love to get a working 
Linux distro on one or both.  SS10 and SS20, both well equipped.

Jim

Re: Time to let sun4m + sun4d RIP?

[Sam Ravnborg]

On Sat, Apr 19, 2014 at 01:48:02PM +0200, Sam Ravnborg wrote:
> Hi all sparc lovers/users.
> 
> We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> and I wonder if there are anyone using these today.
> The questions pops up as there are still stuff to do before the sparc32 port
> is in good shape.
> 
> So the question to you guys...
> Shall we keep support for sun4d and sun4m?
> If yes then why?
> 
> LEON is in active use in several applications - so the sparc32 port is
> as relevant as ever.

Just a fololow-up to my original question.

We still have a few brave souls running the kernels on sparc32 based boxes.
On top of this there are templig.org that looks like a one-man show to implment
sparc32 in VHDL. But very impressive!
Then there are qemu that today supports sun4m - if we drop sun4m we also drop
qemu support for sparc.

So in total there are still reasons to keep sun4m and sun4d around.

Thanks everyone for your responses!

	Sam

Re: Time to let sun4m + sun4d RIP?

[Sam Ravnborg]

On Tue, Apr 29, 2014 at 08:46:11PM +0200, Sam Ravnborg wrote:
> On Sat, Apr 19, 2014 at 01:48:02PM +0200, Sam Ravnborg wrote:
> > Hi all sparc lovers/users.
> > 
> > We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> > and I wonder if there are anyone using these today.
> > The questions pops up as there are still stuff to do before the sparc32 port
> > is in good shape.
> > 
> > So the question to you guys...
> > Shall we keep support for sun4d and sun4m?
> > If yes then why?
> > 
> > LEON is in active use in several applications - so the sparc32 port is
> > as relevant as ever.
> 
> Just a fololow-up to my original question.
> 
> We still have a few brave souls running the kernels on sparc32 based boxes.
> On top of this there are templig.org that looks like a one-man show to implment
> sparc32 in VHDL. But very impressive!
> Then there are qemu that today supports sun4m - if we drop sun4m we also drop
> qemu support for sparc.
> 
> So in total there are still reasons to keep sun4m and sun4d around.
> 
> Thanks everyone for your responses!

And then I read this on LWN:
Debian drops the SPARC architecture - http://lwn.net/Articles/596663/

Which point to: https://bugs.debian.org/cgi-bin/bugreport.cgi?bugt5938

	Sam

Re: Time to let sun4m + sun4d RIP?

[Josip Rodin]

On Tue, Apr 29, 2014 at 09:03:09PM +0200, Sam Ravnborg wrote:
> On Tue, Apr 29, 2014 at 08:46:11PM +0200, Sam Ravnborg wrote:
> > On Sat, Apr 19, 2014 at 01:48:02PM +0200, Sam Ravnborg wrote:
> > > Hi all sparc lovers/users.
> > > 
> > > We all know that is is a long time ago that a sun4m or sun4d machine was produced,
> > > and I wonder if there are anyone using these today.
> > > The questions pops up as there are still stuff to do before the sparc32 port
> > > is in good shape.
> > > 
> > > So the question to you guys...
> > > Shall we keep support for sun4d and sun4m?
> > > If yes then why?
> > > 
> > > LEON is in active use in several applications - so the sparc32 port is
> > > as relevant as ever.
> > 
> > Just a fololow-up to my original question.
> > 
> > We still have a few brave souls running the kernels on sparc32 based boxes.
> > On top of this there are templig.org that looks like a one-man show to implment
> > sparc32 in VHDL. But very impressive!
> > Then there are qemu that today supports sun4m - if we drop sun4m we also drop
> > qemu support for sparc.
> > 
> > So in total there are still reasons to keep sun4m and sun4d around.
> > 
> > Thanks everyone for your responses!
> 
> And then I read this on LWN:
> Debian drops the SPARC architecture - http://lwn.net/Articles/596663/
> 
> Which point to: https://bugs.debian.org/cgi-bin/bugreport.cgi?bugt5938

This simply means that the Debian sparc port will move to the category of
ports whose serious bugs aren't automatically assumed to be equally as
serious with regard to the preparation of new Debian releases.

That is indeed fair at this point, because even if we assume a generous
ratio of affected people to bug reporters, the demand just isn't there,
and hasn't really been there for quite a while now.

-- 
     2. That which causes joy or happiness.

Re: Time to let sun4m + sun4d RIP?

[Kaj-Michael Lang]

On Tue, 2014-04-22 at 21:49 -0600, Jim MacKenzie wrote: 
> On 14-04-22 07:21 PM, David Miller wrote:
> > I think while qemu is around, it is feasible for people to try to keep 
> > maintaining the port.
> 
> I still have two running sun4m machines and would love to get a working 
> Linux distro on one or both.  SS10 and SS20, both well equipped.

One more 32-bit/sun4m sparc user here, just to let you know.
Unfortunately all my old Sun boxes are suffering from the dead RTC
battery so that has limited my use of them lately.

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

Re: Time to let sun4m + sun4d RIP?

[Julian Calaby]

Hi Kaj,

On Fri, May 2, 2014 at 8:29 PM, Kaj-Michael Lang <milang@tal.org> wrote:
> On Tue, 2014-04-22 at 21:49 -0600, Jim MacKenzie wrote:
>> On 14-04-22 07:21 PM, David Miller wrote:
>> > I think while qemu is around, it is feasible for people to try to keep
>> > maintaining the port.
>>
>> I still have two running sun4m machines and would love to get a working
>> Linux distro on one or both.  SS10 and SS20, both well equipped.
>
> One more 32-bit/sun4m sparc user here, just to let you know.
> Unfortunately all my old Sun boxes are suffering from the dead RTC
> battery so that has limited my use of them lately.

I assume you're aware of the standard hacks for that?

Thanks,

-- 
Julian Calaby

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

Re: Time to let sun4m + sun4d RIP?

[Kaj-Michael Lang]

On Fri, 2014-05-02 at 23:53 +1000, Julian Calaby wrote:
> I assume you're aware of the standard hacks for that?

Yes, I fixed an ELC manymany years ago.

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