* Linux on the O2
@ 1997-12-02 21:50 cypher
1997-12-02 22:10 ` Ralf Baechle
` (2 more replies)
0 siblings, 3 replies; 60+ messages in thread
From: cypher @ 1997-12-02 21:50 UTC (permalink / raw)
To: linux
Hey,
I've been looking for some information to start working on an O2 port and
have basically gotten the response that managmement at SGI is not willing
to release hardware specs for the O2 (a mistake IMHO). Don't get me wrong
I think it's great that they allowed the Indy hardware specs to go out the
door, even if in a limitied fashion.
(I ran into the same problem with a Nintendo, but don't you agree that the
64 would make a great little "Network Computer"?)
I'm guessing that they're looking at this as if an Indy running Linux
won't really be that competetive with the O2s runnning IRIX, which I can
almost understand.
I can speak from my personal experience. I working with various research
and production departments in a university that uses UNIX heavily. I can
honestly say that we have several platforms (Alpha, Sparc, and Intel) that
are running linux in production capacity. Also we have a lot of people
using linux on there workstations (the above platforms plus a slew of
others).. I think reallity is that people like Linux because they can get
it to do usefull things on so many platrforms.
Certain organizations within the school, who run anything from O2s to
Octanes, to a Cave/Reallity Enginge, and an Origin 2000 (ok that
probably gave away who I was talking about) are considering Linux on
high-end Intel boxes as a inexpensive replacement for a $6000 dollar O2
running IRIX. I think this is partly due to the added cost of OS licenses
and various other software packages they have to pay for. It would be
ashamed if SGI lost a share of it's hardware sales (even if it is
the only in workstation market) market becasue some freeware wasn't
available to run on it.
I realize that SGI has a vested interest in seeing IRIX succeed, and IRIX
is probably (it hurts to say this) better than Linux when running 128 (or
soon 4096) CPUs. Linux however, is great lightweight OS for workstations
especially. There's even more value added to am SGI running Linux that can
be introduced into an evironment where multiple high-preformance platform
type are desired (ie, Sparcs, Ultras, Alphas, SGIs. HPs, etc.) but a
single operating platform is needed.
This makes perfect sense to me, but I don't know if I'm explaining
everything in the most effecient manner. I'd be _extremely_ happy to
discuss this with someone, especially people from SGI that have some sort
of influence on getting us the specs we need.
Right now I am basically in position (not officially supported by the
university of course, but there not telling me not to do it either, if you
know what I mean) where I've been given and Indy and an O2 to work on
porting Linux while on the clock. I don't know how much better it gets,
but I'm guessing not much... I won't have the hardware forever,
unfortunately, but if I can start making some sort of progress I can
possibly convince them of the need to let me keep it.
I hope we can somehow figure out a way for SGI to help us out on this one.
;-)
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 21:50 Linux on the O2 cypher
@ 1997-12-02 22:10 ` Ralf Baechle
1997-12-02 22:31 ` Ariel Faigon
1997-12-02 22:33 ` William J. Earl
1997-12-03 0:09 ` Alan Cox
1997-12-03 7:46 ` Chris Carlson
2 siblings, 2 replies; 60+ messages in thread
From: Ralf Baechle @ 1997-12-02 22:10 UTC (permalink / raw)
To: cypher; +Cc: linux
On Tue, Dec 02, 1997 at 04:50:40PM -0500, cypher wrote:
> I've been looking for some information to start working on an O2 port and
> have basically gotten the response that managmement at SGI is not willing
> to release hardware specs for the O2 (a mistake IMHO). Don't get me wrong
> I think it's great that they allowed the Indy hardware specs to go out the
> door, even if in a limitied fashion.
Yes, I think the way we will have to go is to demonstrate success of the Linux
port to the Indy. If Linux on the Indy is a success then future projects
will be alot easier.
> (I ran into the same problem with a Nintendo, but don't you agree that the
> 64 would make a great little "Network Computer"?)
The 4mb RAM or so that the N64 has are almost rediculous in context of a
real operating system ...
> Certain organizations within the school, who run anything from O2s to
> Octanes, to a Cave/Reallity Enginge, and an Origin 2000 (ok that
> probably gave away who I was talking about) are considering Linux on
> high-end Intel boxes as a inexpensive replacement for a $6000 dollar O2
> running IRIX. I think this is partly due to the added cost of OS licenses
> and various other software packages they have to pay for. It would be
> ashamed if SGI lost a share of it's hardware sales (even if it is
> the only in workstation market) market becasue some freeware wasn't
> available to run on it.
The sad thing is many people see NT as the solution of this problem ...
> I realize that SGI has a vested interest in seeing IRIX succeed, and IRIX
> is probably (it hurts to say this) better than Linux when running 128 (or
> soon 4096) CPUs. Linux however, is great lightweight OS for workstations
> especially. There's even more value added to am SGI running Linux that can
> be introduced into an evironment where multiple high-preformance platform
> type are desired (ie, Sparcs, Ultras, Alphas, SGIs. HPs, etc.) but a
> single operating platform is needed.
Indeed, you're right and the ambiguousity of the aim IRIX has to solve is
what makes the chance for Linux.
> This makes perfect sense to me, but I don't know if I'm explaining
> everything in the most effecient manner. I'd be _extremely_ happy to
> discuss this with someone, especially people from SGI that have some sort
> of influence on getting us the specs we need.
Actually aside of the Linux developers like me there are also SGI engineers
interested in Linux as well as some ties on this list, so posting here was
a right thing to do.
I can well imagine that way can be found under which an O2 port could become
reality. It's probably most important that customers (== $$$) ask for Linux
for their boxes and that word has to spread upto the higher managment.
> Right now I am basically in position (not officially supported by the
> university of course, but there not telling me not to do it either, if you
> know what I mean) where I've been given and Indy and an O2 to work on
> porting Linux while on the clock.
I suggest to work on the Indy. There is Indy code that just needs a bit of
polishing and debugging, but essentially we're relativly close. Working for
the O2 would mean that you'll aside of a bigger technical task you'll also
have to solve a political problem.
Ralf
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:10 ` Ralf Baechle
@ 1997-12-02 22:31 ` Ariel Faigon
1997-12-02 22:31 ` Ariel Faigon
` (2 more replies)
1997-12-02 22:33 ` William J. Earl
1 sibling, 3 replies; 60+ messages in thread
From: Ariel Faigon @ 1997-12-02 22:31 UTC (permalink / raw)
To: Ralf Baechle; +Cc: cypher, linux
Ralf wrote:
:
:On Tue, Dec 02, 1997 at 04:50:40PM -0500, cypher wrote:
:
:I can well imagine that way can be found under which an O2 port could become
:reality. It's probably most important that customers (== $$$) ask for Linux
:for their boxes and that word has to spread upto the higher managment.
:
:> Right now I am basically in position (not officially supported by the
:> university of course, but there not telling me not to do it either, if you
:> know what I mean) where I've been given and Indy and an O2 to work on
:> porting Linux while on the clock.
:
:I suggest to work on the Indy. There is Indy code that just needs a bit of
:polishing and debugging, but essentially we're relativly close. Working for
:the O2 would mean that you'll aside of a bigger technical task you'll also
:have to solve a political problem.
:
It is nice to see interest in Linux from big SGI customers
(and I wish there was more of this :-)
I'll second Ralf. Once we have an Indy port that's stable,
easy to install and maybe supported by Red-Hat or other
Linux distributors. Once we have something that's actually used
by many Indy owners (rather than just a handful), and once SGI
real customers (who bring SGI real money) seem to like it and
talk to their SGI reps about their interest in Linux on SGI,
only then, management may be open to opening more HW specs
to the public.
I believe there are a few hundreds of thousands Indy's out there.
If we can make one in ten of them (10%) switch to Linux, it would be
a great achievement and one that would certainly be noted
by everyone at SGI, including those who can approve releasing
further specs.
And let's not waste too many resources on the N64 port either :-)
As Mao Zedong used to say... hint, hint, nudge, nudge.
--
Peace, Ariel
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:31 ` Ariel Faigon
@ 1997-12-02 22:31 ` Ariel Faigon
1997-12-03 0:04 ` Alan Cox
1997-12-03 1:51 ` Douglas F. Elznic
2 siblings, 0 replies; 60+ messages in thread
From: Ariel Faigon @ 1997-12-02 22:31 UTC (permalink / raw)
To: Ralf Baechle; +Cc: cypher, linux
Ralf wrote:
:
:On Tue, Dec 02, 1997 at 04:50:40PM -0500, cypher wrote:
:
:I can well imagine that way can be found under which an O2 port could become
:reality. It's probably most important that customers (== $$$) ask for Linux
:for their boxes and that word has to spread upto the higher managment.
:
:> Right now I am basically in position (not officially supported by the
:> university of course, but there not telling me not to do it either, if you
:> know what I mean) where I've been given and Indy and an O2 to work on
:> porting Linux while on the clock.
:
:I suggest to work on the Indy. There is Indy code that just needs a bit of
:polishing and debugging, but essentially we're relativly close. Working for
:the O2 would mean that you'll aside of a bigger technical task you'll also
:have to solve a political problem.
:
It is nice to see interest in Linux from big SGI customers
(and I wish there was more of this :-)
I'll second Ralf. Once we have an Indy port that's stable,
easy to install and maybe supported by Red-Hat or other
Linux distributors. Once we have something that's actually used
by many Indy owners (rather than just a handful), and once SGI
real customers (who bring SGI real money) seem to like it and
talk to their SGI reps about their interest in Linux on SGI,
only then, management may be open to opening more HW specs
to the public.
I believe there are a few hundreds of thousands Indy's out there.
If we can make one in ten of them (10%) switch to Linux, it would be
a great achievement and one that would certainly be noted
by everyone at SGI, including those who can approve releasing
further specs.
And let's not waste too many resources on the N64 port either :-)
As Mao Zedong used to say... hint, hint, nudge, nudge.
--
Peace, Ariel
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:10 ` Ralf Baechle
1997-12-02 22:31 ` Ariel Faigon
@ 1997-12-02 22:33 ` William J. Earl
1997-12-02 23:32 ` Christian Reisch
1 sibling, 1 reply; 60+ messages in thread
From: William J. Earl @ 1997-12-02 22:33 UTC (permalink / raw)
To: Ralf Baechle; +Cc: cypher, linux
Ralf Baechle writes:
> On Tue, Dec 02, 1997 at 04:50:40PM -0500, cypher wrote:
...
> Actually aside of the Linux developers like me there are also SGI engineers
> interested in Linux as well as some ties on this list, so posting here was
> a right thing to do.
I am one of the SGI engineers, and I would like to be able to release
the O2 specifications. Releasing the Indy specifications basically happened
when we had started shipping O2, and the Indy was thus obsolete as a hardware
product. I expect it would be easy to get approval when O2 is no longer
shipping, but doing it now will take some persuasion.
> I can well imagine that way can be found under which an O2 port could become
> reality. It's probably most important that customers (== $$$) ask for Linux
> for their boxes and that word has to spread upto the higher managment.
This is precisely correct. If enough large customers ask for it,
it will happen. The problem so far has been persuading management that
very many people would consider SGI boxes with Linux. Part of the
problem is that a simple Linux port leaves much of the special features
of the SGI system unused, and really exploiting the graphics and digital
media features is a major project, when SGI already has too many projects
underway.
> > Right now I am basically in position (not officially supported by the
> > university of course, but there not telling me not to do it either, if you
> > know what I mean) where I've been given and Indy and an O2 to work on
> > porting Linux while on the clock.
>
> I suggest to work on the Indy. There is Indy code that just needs a bit of
> polishing and debugging, but essentially we're relativly close. Working for
> the O2 would mean that you'll aside of a bigger technical task you'll also
> have to solve a political problem.
O2 is also a bigger technical problem, even with specifications in hand.
The drivers are actually easier than for Indy, but really exploiting the
O2 capabilities in graphics and digital media required a lot of enhancements
to the memory management, since O2 is a UMA system. You could do a dumb
port, with a fixed frame buffer size, but then most of the O2 is wasted.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:33 ` William J. Earl
@ 1997-12-02 23:32 ` Christian Reisch
1997-12-02 23:45 ` William J. Earl
0 siblings, 1 reply; 60+ messages in thread
From: Christian Reisch @ 1997-12-02 23:32 UTC (permalink / raw)
To: linux
William J. Earl wrote:
O2 is also a bigger technical problem, even with specifications in
hand.
> The drivers are actually easier than for Indy, but really exploiting the
> O2 capabilities in graphics and digital media required a lot of enhancements
> to the memory management, since O2 is a UMA system. You could do a dumb
> port, with a fixed frame buffer size, but then most of the O2 is wasted.
Actually I suspect the exact opposite would happen and hence
management's fear. There are ALOT more people in the linux community
than we have working at SGI. It is very concievable that if the specs
were
available linux on O2 would massively out perform irix on the O2. This
would
be a major embarrassment to the company. Compare and contrast as Unixes
Solaris X86 and Linux on a given x86 box or Digital Unix and linux on an
alpha.
Quality wise free software always beats proprietary software (if for
no other reason than it has to make fewer concesions to release
schedules
and politics) often embarassingly so.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 23:32 ` Christian Reisch
@ 1997-12-02 23:45 ` William J. Earl
0 siblings, 0 replies; 60+ messages in thread
From: William J. Earl @ 1997-12-02 23:45 UTC (permalink / raw)
To: Christian Reisch; +Cc: linux
Christian Reisch writes:
> William J. Earl wrote:
> O2 is also a bigger technical problem, even with specifications in
> hand.
> > The drivers are actually easier than for Indy, but really exploiting the
> > O2 capabilities in graphics and digital media required a lot of enhancements
> > to the memory management, since O2 is a UMA system. You could do a dumb
> > port, with a fixed frame buffer size, but then most of the O2 is wasted.
>
> Actually I suspect the exact opposite would happen and hence
> management's fear. There are ALOT more people in the linux community
> than we have working at SGI. It is very concievable that if the specs
> were
> available linux on O2 would massively out perform irix on the O2. This
> would
> be a major embarrassment to the company. Compare and contrast as Unixes
> Solaris X86 and Linux on a given x86 box or Digital Unix and linux on an
> alpha.
Many of us who want to see Linux on SGI systems expect that Linux
will be substantially faster on a number of primitive functions, especially
on uniprocessor systems. We think of this as a nice club to persuade people
that low-level performance (a cache miss here and 100 instructions there)
really matters.
What I was talking about in regard to O2 was exploiting the
hardware functionality, such as the ability to use 100 MB of memory
for Photoshop, and then switch in seconds to using 100 MB of memory
for texture mapping, including real-time texture mapping of live
video. Some of this is fairly generic library work, but some of it
requires fundamental changes to the operating system memory management
(assuming you do not impose a fixed partitioning of memory).
> Quality wise free software always beats proprietary software (if for
> no other reason than it has to make fewer concesions to release
> schedules
> and politics) often embarassingly so.
That is not always the case. It is the case for GNU and Linux
software, due to the constraints imposed on the development and release
process, which tend to keep untested changes out of the main tree.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:31 ` Ariel Faigon
1997-12-02 22:31 ` Ariel Faigon
@ 1997-12-03 0:04 ` Alan Cox
1997-12-03 0:46 ` John Wiederhirn
1997-12-03 1:51 ` Douglas F. Elznic
2 siblings, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-03 0:04 UTC (permalink / raw)
To: ariel; +Cc: ralf, cypher, linux
> I'll second Ralf. Once we have an Indy port that's stable,
> easy to install and maybe supported by Red-Hat or other
> Linux distributors. Once we have something that's actually used
> by many Indy owners (rather than just a handful), and once SGI
> real customers (who bring SGI real money) seem to like it and
> talk to their SGI reps about their interest in Linux on SGI,
> only then, management may be open to opening more HW specs
> to the public.
Even if they don't its far far easier to reverse engineer the O2 working
from a stable Indy base than from a wobbly base platform. I don't personally
believe there is anything SGI could do to stop a Linux hacker making Linux
run on the O2
> As Mao Zedong used to say... hint, hint, nudge, nudge.
Yep. I don't see a big problem. Im not privy to SGI accounts but I refuse
to believe Irix itself does anything more than break even at best just by
taking published information on developer team size, and sales. In fact
I suspect its a loss maker like Solaris. If Linux sells SGI hardware Im sure it
will get us into good books 8)
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 21:50 Linux on the O2 cypher
1997-12-02 22:10 ` Ralf Baechle
@ 1997-12-03 0:09 ` Alan Cox
1997-12-03 0:09 ` Alan Cox
1997-12-03 7:46 ` Chris Carlson
2 siblings, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-03 0:09 UTC (permalink / raw)
To: cypher; +Cc: linux
> I'm guessing that they're looking at this as if an Indy running Linux
> won't really be that competetive with the O2s runnning IRIX, which I can
> almost understand.
Ah conspiracy theories abound. It could just be they didnt want anyone
cloning their hardware
> I realize that SGI has a vested interest in seeing IRIX succeed, and IRIX
> is probably (it hurts to say this) better than Linux when running 128 (or
> soon 4096) CPUs. Linux however, is great lightweight OS for workstations
Right now Linux scales to about 2-3 CPU's for generic stuff in 2.1.x maybe
4 if the hardware is sane (eg Ultrasparc). To go to 128-4096 CPU's really
involves changing the fundamental rules and building a sort of very tightly
interlinked multicomputer - AP1000+ on steroids as it were. There is some
interest in this from various people, notably in sticking 9 or 10 EBSA285
Digital strongarm boards in a PCI backplane and using PCI bus as the
backbone network.
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 0:09 ` Alan Cox
@ 1997-12-03 0:09 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-03 0:09 UTC (permalink / raw)
To: cypher; +Cc: linux
> I'm guessing that they're looking at this as if an Indy running Linux
> won't really be that competetive with the O2s runnning IRIX, which I can
> almost understand.
Ah conspiracy theories abound. It could just be they didnt want anyone
cloning their hardware
> I realize that SGI has a vested interest in seeing IRIX succeed, and IRIX
> is probably (it hurts to say this) better than Linux when running 128 (or
> soon 4096) CPUs. Linux however, is great lightweight OS for workstations
Right now Linux scales to about 2-3 CPU's for generic stuff in 2.1.x maybe
4 if the hardware is sane (eg Ultrasparc). To go to 128-4096 CPU's really
involves changing the fundamental rules and building a sort of very tightly
interlinked multicomputer - AP1000+ on steroids as it were. There is some
interest in this from various people, notably in sticking 9 or 10 EBSA285
Digital strongarm boards in a PCI backplane and using PCI bus as the
backbone network.
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 0:04 ` Alan Cox
@ 1997-12-03 0:46 ` John Wiederhirn
1997-12-03 3:06 ` William J. Earl
1997-12-03 18:28 ` cypher
0 siblings, 2 replies; 60+ messages in thread
From: John Wiederhirn @ 1997-12-03 0:46 UTC (permalink / raw)
To: linux
On Dec 3, 12:04am, Alan Cox wrote:
>
> > I'll second Ralf. Once we have an Indy port that's stable,
> > easy to install and maybe supported by Red-Hat or other
> > Linux distributors. Once we have something that's actually used
> > by many Indy owners (rather than just a handful), and once SGI
> > real customers (who bring SGI real money) seem to like it and
> > talk to their SGI reps about their interest in Linux on SGI,
> > only then, management may be open to opening more HW specs
> > to the public.
>
> Even if they don't its far far easier to reverse engineer the O2 working
> from a stable Indy base than from a wobbly base platform. I don't personally
> believe there is anything SGI could do to stop a Linux hacker making Linux
> run on the O2
As I've commented before, the problem isn't so much in getting the
kernel working (and working quickly). The problem comes the second
you need to deal with graphics. Given the O2's onboard multimedia
hardware, the situation only gets more complex.
Right now, there is no effective hardware abstraction mechanism in
Linux for graphics/multimedia. With the Indy's rather simple gfx
capabilities, that isn't a tremendous hardship. The hardship comes
when you are managing things like graphics pipelines, or hardware
that has very specific requirements for memory allocation and
alignment (and access). Many of the necessary pieces simply do not
currently exist in the Linux kernel services.
While it would be possible to move these needs out to the X server
to some extent, not all of them can be handled that way (in specific,
the memory issues). Further, there isn't really anything like
uniform functionality in Linux X servers for handling things such
as video I/O, genlocking, etc. They exist in 3rd party servers,
but then the cooperation requirement rises again.
The proposal has come up multiple times to use the IRIX X server on
top of a Linux kernel. Unfortunately, once again, that gets very
problemmatic on the more advanced gfx platforms such as O2 and Octane.
They both contain hardware-dependant sections of code which RELY on
certain kernel functionalities being present. Factor in OpenGL and
the whole complexity level rises again.
Now, having said all that, I do believe the problem is solvable.
I think that with some modifications of the Linux kernel, it would
be quite possible to get the necessary abstraction to support
either an IRIX or customized third-party X server. Further, it
would be nice if something equivalent to the SGI DMedia libraries
were available on Linux (with similar hardware abstraction), since
this is an area where simply put, Linux is FAR behind Windows{95,NT}
and SGI (and others).
Getting Linux on O2 probably isn't too difficult. Getting USEFUL
O2/Octane/Onyx2 2D/3D graphics and multimedia under Linux would
be god-awful difficult with the current kernel/X situation.
My $0.02.
--
John Wiederhirn (DSD, Graphics Kernel MTS) jwiede@engr.sgi.com
"I've always said there's nothing an agnostic can't do if he
really doesn't know whether he believes in anything or not."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 22:31 ` Ariel Faigon
1997-12-02 22:31 ` Ariel Faigon
1997-12-03 0:04 ` Alan Cox
@ 1997-12-03 1:51 ` Douglas F. Elznic
1997-12-03 1:51 ` Douglas F. Elznic
` (4 more replies)
2 siblings, 5 replies; 60+ messages in thread
From: Douglas F. Elznic @ 1997-12-03 1:51 UTC (permalink / raw)
To: Ariel Faigon, Ralf Baechle; +Cc: cypher, linux
At 02:31 PM 12/2/97 -0800, Ariel Faigon wrote:
>Ralf wrote:
>And let's not waste too many resources on the N64 port either :-)
You must admit that the n64 would make a neat toy project.
--
Douglas F. Elznic
Email:
delznic@acm.org
Pager:
3154492502.2423685@pagenet.net
"The tree of liberty must be refreshed from time to time with the blood of
patriots and tyrants."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 1:51 ` Douglas F. Elznic
@ 1997-12-03 1:51 ` Douglas F. Elznic
1997-12-03 7:01 ` ralf
` (3 subsequent siblings)
4 siblings, 0 replies; 60+ messages in thread
From: Douglas F. Elznic @ 1997-12-03 1:51 UTC (permalink / raw)
To: Ariel Faigon, Ralf Baechle; +Cc: cypher, linux
At 02:31 PM 12/2/97 -0800, Ariel Faigon wrote:
>Ralf wrote:
>And let's not waste too many resources on the N64 port either :-)
You must admit that the n64 would make a neat toy project.
--
Douglas F. Elznic
Email:
delznic@acm.org
Pager:
3154492502.2423685@pagenet.net
"The tree of liberty must be refreshed from time to time with the blood of
patriots and tyrants."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:05 ` David Chatterton
@ 1997-12-03 2:34 ` Brendan Black
1997-12-03 2:58 ` Richard Offer
` (3 more replies)
0 siblings, 4 replies; 60+ messages in thread
From: Brendan Black @ 1997-12-03 2:34 UTC (permalink / raw)
To: linux
David Chatterton wrote:
>
> Maybe then my wife would let me buy one: "Yes dear, I'm doing real work" :)
I can imagine all sorts of 'uses' for something like this (linux on the
N64) - but one thing crops
up, I thought there WAS an N64 port.
I seem to remember something, may have been an article on IX magazine...
--
Brendan Black - Network Engineer, Telecom Internet Services
email: ratfink@xtra.co.nz (personal) phone: +649 3555238
mob: +6425 2752667
The word "spine" is, of course, an anagram of "penis". This is true in
almost fifty percent of the languages of the Galaxy, and many people
have attempted to explain why. Usually these explanations get bogged
down in silly puns about "standing erect".
-- Douglas Adams, "The Hitchhiker's Guide to the Galaxy"
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 2:34 ` Brendan Black
@ 1997-12-03 2:58 ` Richard Offer
1997-12-03 3:19 ` Douglas F. Elznic
` (2 subsequent siblings)
3 siblings, 0 replies; 60+ messages in thread
From: Richard Offer @ 1997-12-03 2:58 UTC (permalink / raw)
To: Brendan Black; +Cc: linux
* $ from ratfink@xtra.co.nz at "3-Dec: 3:34pm" | sed "1,$s/^/* /"
*
*
* David Chatterton wrote:
* >
* > Maybe then my wife would let me buy one: "Yes dear, I'm doing real work" :)
*
* I can imagine all sorts of 'uses' for something like this (linux on the
* N64) - but one thing crops
* up, I thought there WAS an N64 port.
*
* I seem to remember something, may have been an article on IX magazine...
It was a fairly obvious April Fool joke.
* --
* Brendan Black - Network Engineer, Telecom Internet Services
* email: ratfink@xtra.co.nz (personal) phone: +649 3555238
* mob: +6425 2752667
* The word "spine" is, of course, an anagram of "penis". This is true in
* almost fifty percent of the languages of the Galaxy, and many people
* have attempted to explain why. Usually these explanations get bogged
* down in silly puns about "standing erect".
* -- Douglas Adams, "The Hitchhiker's Guide to the Galaxy"
Never seen or heard that bit before.
*
richard.
-----------------------------------------------------------------------
Richard M. Offer Widget FAQ --> http://reality.sgi.com/widgetFAQ
MTS-Core Design (Motif)
___________________________________________http://reality.sgi.com/offer
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 0:46 ` John Wiederhirn
@ 1997-12-03 3:06 ` William J. Earl
1997-12-03 3:06 ` William J. Earl
1997-12-03 18:28 ` cypher
1 sibling, 1 reply; 60+ messages in thread
From: William J. Earl @ 1997-12-03 3:06 UTC (permalink / raw)
To: John Wiederhirn; +Cc: linux
John Wiederhirn writes:
...
> Now, having said all that, I do believe the problem is solvable.
> I think that with some modifications of the Linux kernel, it would
> be quite possible to get the necessary abstraction to support
> either an IRIX or customized third-party X server. Further, it
> would be nice if something equivalent to the SGI DMedia libraries
> were available on Linux (with similar hardware abstraction), since
> this is an area where simply put, Linux is FAR behind Windows{95,NT}
> and SGI (and others).
>
> Getting Linux on O2 probably isn't too difficult. Getting USEFUL
> O2/Octane/Onyx2 2D/3D graphics and multimedia under Linux would
> be god-awful difficult with the current kernel/X situation.
...
Having implemented a good bit of the system-specific code for
Indy and O2, I agree entirely. Without the above features, O2 is just
another, not especially distinguished, general purpose workstation,
except for having fairly fast I/O for its price. The most interesting
thing about O2 is what you can do with graphics and multimedia, which
is otherwise possible only with far more expensive hardware. In
different problem domains, something similar can be said of Octane and
the various SGI servers.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 3:06 ` William J. Earl
@ 1997-12-03 3:06 ` William J. Earl
0 siblings, 0 replies; 60+ messages in thread
From: William J. Earl @ 1997-12-03 3:06 UTC (permalink / raw)
To: John Wiederhirn; +Cc: linux
John Wiederhirn writes:
...
> Now, having said all that, I do believe the problem is solvable.
> I think that with some modifications of the Linux kernel, it would
> be quite possible to get the necessary abstraction to support
> either an IRIX or customized third-party X server. Further, it
> would be nice if something equivalent to the SGI DMedia libraries
> were available on Linux (with similar hardware abstraction), since
> this is an area where simply put, Linux is FAR behind Windows{95,NT}
> and SGI (and others).
>
> Getting Linux on O2 probably isn't too difficult. Getting USEFUL
> O2/Octane/Onyx2 2D/3D graphics and multimedia under Linux would
> be god-awful difficult with the current kernel/X situation.
...
Having implemented a good bit of the system-specific code for
Indy and O2, I agree entirely. Without the above features, O2 is just
another, not especially distinguished, general purpose workstation,
except for having fairly fast I/O for its price. The most interesting
thing about O2 is what you can do with graphics and multimedia, which
is otherwise possible only with far more expensive hardware. In
different problem domains, something similar can be said of Octane and
the various SGI servers.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 2:34 ` Brendan Black
1997-12-03 2:58 ` Richard Offer
@ 1997-12-03 3:19 ` Douglas F. Elznic
1997-12-03 8:21 ` Oliver Frommel
1997-12-03 10:51 ` Alan Cox
1997-12-03 18:32 ` cypher
3 siblings, 1 reply; 60+ messages in thread
From: Douglas F. Elznic @ 1997-12-03 3:19 UTC (permalink / raw)
To: Brendan Black, linux
At 03:34 PM 12/3/97 +1300, Brendan Black wrote:
>David Chatterton wrote:
>>
>> Maybe then my wife would let me buy one: "Yes dear, I'm doing real work" :)
>
>I can imagine all sorts of 'uses' for something like this (linux on the
>N64) - but one thing crops
>up, I thought there WAS an N64 port.
>
>I seem to remember something, may have been an article on IX magazine...
I found the article but nothing about any progress. I am looking for some
code samples or anything else on the project.
http://www.ix.de/ix/artikel/E/1997/04/036
--
Douglas F. Elznic
Email:
delznic@acm.org
Pager:
3154492502.2423685@pagenet.net
"The tree of liberty must be refreshed from time to time with the blood of
patriots and tyrants."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 1:51 ` Douglas F. Elznic
1997-12-03 1:51 ` Douglas F. Elznic
@ 1997-12-03 7:01 ` ralf
[not found] ` <Pine.LNX.3.95.971203132942.15095B-100000@vertigo.cs.indian a.edu>
` (2 subsequent siblings)
4 siblings, 0 replies; 60+ messages in thread
From: ralf @ 1997-12-03 7:01 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: Ariel Faigon, cypher, linux
On Tue, Dec 02, 1997 at 08:51:46PM -0500, Douglas F. Elznic wrote:
> At 02:31 PM 12/2/97 -0800, Ariel Faigon wrote:
> >Ralf wrote:
> >And let's not waste too many resources on the N64 port either :-)
> You must admit that the n64 would make a neat toy project.
Maybe - but I personally would cut my fingers off for other things like
doing the R10k port or SMP.
Ralf
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-02 21:50 Linux on the O2 cypher
1997-12-02 22:10 ` Ralf Baechle
1997-12-03 0:09 ` Alan Cox
@ 1997-12-03 7:46 ` Chris Carlson
1997-12-04 5:39 ` Lige Hensley
2 siblings, 1 reply; 60+ messages in thread
From: Chris Carlson @ 1997-12-03 7:46 UTC (permalink / raw)
To: linux
cypher wrote:
> Hey,
>
> I've been looking for some information to start working on an O2 port and
> have basically gotten the response that managmement at SGI is not willing
> to release hardware specs for the O2 (a mistake IMHO). Don't get me wrong
> I think it's great that they allowed the Indy hardware specs to go out the
> door, even if in a limitied fashion.
>
> (I ran into the same problem with a Nintendo, but don't you agree that the
> 64 would make a great little "Network Computer"?)
>
> I'm guessing that they're looking at this as if an Indy running Linux
> won't really be that competetive with the O2s runnning IRIX, which I can
> almost understand.
>
> I can speak from my personal experience. I working with various research
> and production departments in a university that uses UNIX heavily. I can
> honestly say that we have several platforms (Alpha, Sparc, and Intel) that
> are running linux in production capacity. Also we have a lot of people
> using linux on there workstations (the above platforms plus a slew of
> others).. I think reallity is that people like Linux because they can get
> it to do usefull things on so many platrforms.
>
> Certain organizations within the school, who run anything from O2s to
> Octanes, to a Cave/Reallity Enginge, and an Origin 2000 (ok that
> probably gave away who I was talking about) are considering Linux on
> high-end Intel boxes as a inexpensive replacement for a $6000 dollar O2
> running IRIX. I think this is partly due to the added cost of OS licenses
> and various other software packages they have to pay for. It would be
> ashamed if SGI lost a share of it's hardware sales (even if it is
> the only in workstation market) market becasue some freeware wasn't
> available to run on it.
>
> I realize that SGI has a vested interest in seeing IRIX succeed, and IRIX
> is probably (it hurts to say this) better than Linux when running 128 (or
> soon 4096) CPUs. Linux however, is great lightweight OS for workstations
> especially. There's even more value added to am SGI running Linux that can
> be introduced into an evironment where multiple high-preformance platform
> type are desired (ie, Sparcs, Ultras, Alphas, SGIs. HPs, etc.) but a
> single operating platform is needed.
>
> This makes perfect sense to me, but I don't know if I'm explaining
> everything in the most effecient manner. I'd be _extremely_ happy to
> discuss this with someone, especially people from SGI that have some sort
> of influence on getting us the specs we need.
>
> Right now I am basically in position (not officially supported by the
> university of course, but there not telling me not to do it either, if you
> know what I mean) where I've been given and Indy and an O2 to work on
> porting Linux while on the clock. I don't know how much better it gets,
> but I'm guessing not much... I won't have the hardware forever,
> unfortunately, but if I can start making some sort of progress I can
> possibly convince them of the need to let me keep it.
>
> I hope we can somehow figure out a way for SGI to help us out on this one.
> ;-)
>
I also would like to see Linux ported to the O2 but also don't expect it to
happen in the near future. We have to remember that SGI is a "hardware
company" and did not get where it is because of software. To ask them to
divulge the inner workings of their bread-and-butter product, thus allowing
competitors to reproduce and improve upon the design is asking for disaster.
SGI does not fabricate their own chips and they have done their best to
provide access to their source (as long as it doesn't compromise their
hardware investment). The only thing they have to sell is their design.
Since MIPS is not faring very well against competitors and there seems to me
a bit of doubt that they can stay ahead, what else does SGI have to sell?
Just their internal architecture and IRIX.
Giving the Linux community access to the internal design of the O2, Octane
and/or Origin would give any competitor the information they need to put SGI
out of business. I don't blame the company for not wanting Linux on the
newer products until the next generation is released.
--
+----------------------------------------+
| Christopher W. Carlson |
| e-mail home: cwcarlson@home.com |
| WWW: http://members.home.net/cwcarlson |
+----------------------------------------+
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 3:19 ` Douglas F. Elznic
@ 1997-12-03 8:21 ` Oliver Frommel
0 siblings, 0 replies; 60+ messages in thread
From: Oliver Frommel @ 1997-12-03 8:21 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: Brendan Black, linux
> >I can imagine all sorts of 'uses' for something like this (linux on the
> >N64) - but one thing crops
> >up, I thought there WAS an N64 port.
> >
> >I seem to remember something, may have been an article on IX magazine...
> I found the article but nothing about any progress. I am looking for some
> code samples or anything else on the project.
> http://www.ix.de/ix/artikel/E/1997/04/036
>
sorry to disappoint you, but this was an April's joke (even confirmed in the
next issue of iX). IMHO the most advanced projects on N64 in terms of hacking
it without docs are some demos (you might take a look at www.dextrose.com)
o.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 2:34 ` Brendan Black
1997-12-03 2:58 ` Richard Offer
1997-12-03 3:19 ` Douglas F. Elznic
@ 1997-12-03 10:51 ` Alan Cox
1997-12-03 10:51 ` Alan Cox
1997-12-03 18:32 ` cypher
3 siblings, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-03 10:51 UTC (permalink / raw)
To: Brendan Black; +Cc: linux
> I can imagine all sorts of 'uses' for something like this (linux on the
> N64) - but one thing crops
> up, I thought there WAS an N64 port.
>
> I seem to remember something, may have been an article on IX magazine...
April issue I believe ...
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 10:51 ` Alan Cox
@ 1997-12-03 10:51 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-03 10:51 UTC (permalink / raw)
To: Brendan Black; +Cc: linux
> I can imagine all sorts of 'uses' for something like this (linux on the
> N64) - but one thing crops
> up, I thought there WAS an N64 port.
>
> I seem to remember something, may have been an article on IX magazine...
April issue I believe ...
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
[not found] ` <Pine.LNX.3.95.971203132942.15095B-100000@vertigo.cs.indian a.edu>
@ 1997-12-03 18:04 ` Douglas F. Elznic
1997-12-03 18:39 ` Alan Cox
` (2 more replies)
0 siblings, 3 replies; 60+ messages in thread
From: Douglas F. Elznic @ 1997-12-03 18:04 UTC (permalink / raw)
To: cypher; +Cc: Ariel Faigon, Ralf Baechle, linux
At 01:31 PM 12/3/97 -0500, cypher wrote:
>
>Speaking of the N64,
>
>There now exists an emulator for it. Now if one were to port the kernel to
>the emulator and then burn that to a rom.... hmmm
>
>I guess you would at least want the 8mb memory upgrade, but 4mb would be a
>nice enough for an embeded linux.
>
What are the specs on Sun's javastation? They can not be a whole lot
greater that the N64.
--
Douglas F. Elznic
Email:
delznic@acm.org
Pager:
3154492502.2423685@pagenet.net
"The tree of liberty must be refreshed from time to time with the blood of
patriots and tyrants."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 1:51 ` Douglas F. Elznic
` (2 preceding siblings ...)
[not found] ` <Pine.LNX.3.95.971203132942.15095B-100000@vertigo.cs.indian a.edu>
@ 1997-12-03 18:05 ` David Chatterton
1997-12-03 2:34 ` Brendan Black
1997-12-03 18:31 ` cypher
4 siblings, 1 reply; 60+ messages in thread
From: David Chatterton @ 1997-12-03 18:05 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: linux
On Dec 2, 8:51pm, Douglas F. Elznic wrote:
| Subject: Re: Linux on the O2
| At 02:31 PM 12/2/97 -0800, Ariel Faigon wrote:
| >Ralf wrote:
| >And let's not waste too many resources on the N64 port either :-)
| You must admit that the n64 would make a neat toy project.
|
Maybe then my wife would let me buy one: "Yes dear, I'm doing real work" :)
David
--
David Chatterton (61-3) 9882 8211 (Tel)
R&D Software Engineer (61-3) 9882 8030 (Fax)
Performance Tools Group http://reality.sgi.com/chatz
Silicon Graphics Pty.Ltd., 357 Camberwell Rd, Melbourne, Australia
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 0:46 ` John Wiederhirn
1997-12-03 3:06 ` William J. Earl
@ 1997-12-03 18:28 ` cypher
1 sibling, 0 replies; 60+ messages in thread
From: cypher @ 1997-12-03 18:28 UTC (permalink / raw)
To: John Wiederhirn; +Cc: linux
On Tue, 2 Dec 1997, John Wiederhirn wrote:
> As I've commented before, the problem isn't so much in getting the
> kernel working (and working quickly). The problem comes the second
> you need to deal with graphics. Given the O2's onboard multimedia
> hardware, the situation only gets more complex.
>
> Right now, there is no effective hardware abstraction mechanism in
> Linux for graphics/multimedia. With the Indy's rather simple gfx
> capabilities, that isn't a tremendous hardship. The hardship comes
> when you are managing things like graphics pipelines, or hardware
> that has very specific requirements for memory allocation and
> alignment (and access). Many of the necessary pieces simply do not
> currently exist in the Linux kernel services.
I would definately agree here, but don't get me wrong. I'm not looking for
a quick solution. I think part of what makes Linux so great is that when
it does get ported to a new platform, the kernel learns something new.
What I mean is that by porting to the 02 and allowing Linux to develope a
good "hardware abstraction mechanism" or better memory management or the
various other internals that allow it to run well on an 02 would also give
Linux as UNIX an advantage over the other UNIXes that only run on one or
two platforms. Don't mistake what I say to mean that ever change we make
the kernel for the O2 will improve Linux on every other platform, but now
and then something will improve other code, just like some code used for
other platforms will help improve the Linux on the various SGI platforms.
> While it would be possible to move these needs out to the X server
> to some extent, not all of them can be handled that way (in specific,
> the memory issues). Further, there isn't really anything like
> uniform functionality in Linux X servers for handling things such
> as video I/O, genlocking, etc. They exist in 3rd party servers,
> but then the cooperation requirement rises again.
If there is one thing I have learned it is never to say never or isn't or
can't when talking to a group of Linux hackers...
> Now, having said all that, I do believe the problem is solvable.
> I think that with some modifications of the Linux kernel, it would
> be quite possible to get the necessary abstraction to support
> either an IRIX or customized third-party X server. Further, it
> would be nice if something equivalent to the SGI DMedia libraries
> were available on Linux (with similar hardware abstraction), since
> this is an area where simply put, Linux is FAR behind Windows{95,NT}
> and SGI (and others).
I'm glad you think it is...
> Getting Linux on O2 probably isn't too difficult. Getting USEFUL
> O2/Octane/Onyx2 2D/3D graphics and multimedia under Linux would
> be god-awful difficult with the current kernel/X situation.
But not impossible ;-)
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
---
find my pgp public key at http://www.cs.indiana.edu/~tshrider
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 1:51 ` Douglas F. Elznic
` (3 preceding siblings ...)
1997-12-03 18:05 ` David Chatterton
@ 1997-12-03 18:31 ` cypher
1997-12-03 21:19 ` Warner Losh
4 siblings, 1 reply; 60+ messages in thread
From: cypher @ 1997-12-03 18:31 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: Ariel Faigon, Ralf Baechle, linux
Speaking of the N64,
There now exists an emulator for it. Now if one were to port the kernel to
the emulator and then burn that to a rom.... hmmm
I guess you would at least want the 8mb memory upgrade, but 4mb would be a
nice enough for an embeded linux.
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
On Tue, 2 Dec 1997, Douglas F. Elznic wrote:
> At 02:31 PM 12/2/97 -0800, Ariel Faigon wrote:
> >Ralf wrote:
> >And let's not waste too many resources on the N64 port either :-)
> You must admit that the n64 would make a neat toy project.
>
> --
> Douglas F. Elznic
> Email:
> delznic@acm.org
> Pager:
> 3154492502.2423685@pagenet.net
> "The tree of liberty must be refreshed from time to time with the blood of
> patriots and tyrants."
>
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 2:34 ` Brendan Black
` (2 preceding siblings ...)
1997-12-03 10:51 ` Alan Cox
@ 1997-12-03 18:32 ` cypher
3 siblings, 0 replies; 60+ messages in thread
From: cypher @ 1997-12-03 18:32 UTC (permalink / raw)
To: Brendan Black; +Cc: linux
On Wed, 3 Dec 1997, Brendan Black wrote:
> I can imagine all sorts of 'uses' for something like this (linux on the
> N64) - but one thing crops
> up, I thought there WAS an N64 port.
>
> I seem to remember something, may have been an article on IX magazine...
I think if you take another look at that magazine you will notice that it
was published on April, 1st. I've read the same article...
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:04 ` Douglas F. Elznic
@ 1997-12-03 18:39 ` Alan Cox
1997-12-03 18:39 ` Alan Cox
1997-12-03 18:42 ` cypher
1997-12-03 18:49 ` cypher
2 siblings, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-03 18:39 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: cypher, ariel, ralf, linux
> What are the specs on Sun's javastation? They can not be a whole lot
> greater that the N64.
The existing one is a Sun4 box with funny keyboard and other oddments. DaveM
has one if someone knows how to create time by magic ;)
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:39 ` Alan Cox
@ 1997-12-03 18:39 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-03 18:39 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: cypher, ariel, ralf, linux
> What are the specs on Sun's javastation? They can not be a whole lot
> greater that the N64.
The existing one is a Sun4 box with funny keyboard and other oddments. DaveM
has one if someone knows how to create time by magic ;)
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:04 ` Douglas F. Elznic
1997-12-03 18:39 ` Alan Cox
@ 1997-12-03 18:42 ` cypher
1997-12-03 18:49 ` cypher
2 siblings, 0 replies; 60+ messages in thread
From: cypher @ 1997-12-03 18:42 UTC (permalink / raw)
To: Douglas F. Elznic; +Cc: Ariel Faigon, Ralf Baechle, linux
Now your thinking... ;-)
I have a friend who is doing a trace of N64 as we speak... Nintendo won't
even consider talking to use, let alone supporting this, so were doing it
on our own. As soon as I have info I'll make it available...
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
On Wed, 3 Dec 1997, Douglas F. Elznic wrote:
> At 01:31 PM 12/3/97 -0500, cypher wrote:
> >
> >Speaking of the N64,
> >
> >There now exists an emulator for it. Now if one were to port the kernel to
> >the emulator and then burn that to a rom.... hmmm
> >
> >I guess you would at least want the 8mb memory upgrade, but 4mb would be a
> >nice enough for an embeded linux.
> >
> What are the specs on Sun's javastation? They can not be a whole lot
> greater that the N64.
>
> --
> Douglas F. Elznic
> Email:
> delznic@acm.org
> Pager:
> 3154492502.2423685@pagenet.net
> "The tree of liberty must be refreshed from time to time with the blood of
> patriots and tyrants."
>
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:04 ` Douglas F. Elznic
1997-12-03 18:39 ` Alan Cox
1997-12-03 18:42 ` cypher
@ 1997-12-03 18:49 ` cypher
2 siblings, 0 replies; 60+ messages in thread
From: cypher @ 1997-12-03 18:49 UTC (permalink / raw)
To: linux
I'm really glad I got such a great response to my message. Everyone's
input is greatly appreciated.
I'll try and summarize everyones point in another mail as soon as I get a
chance hoping that the "Big Brother" at SGI is listening...
Until then, here I come Indy...
(Although wouldn't it be nice to be able to have IU ask for Linux the
next time they decide to add more nodes to there Cray Origin 2000. I
mean, we could always just use that Intel Paragon we have sitting in
the cornere) ;-)
---
Todd M. Shrider Unix Workstation Support Group
(812)855-2627 2711 E. 10th Street
tshrider@indiana.edu Indiana University,
http://www.uwsg.indiana.edu/ Bloomington, IN 47408-2671
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 18:31 ` cypher
@ 1997-12-03 21:19 ` Warner Losh
1997-12-03 22:16 ` Chad Campbell
[not found] ` <Pine.LNX.3.95.971203161250.15277A-100000@dv123s34.lawrence .ks.us>
0 siblings, 2 replies; 60+ messages in thread
From: Warner Losh @ 1997-12-03 21:19 UTC (permalink / raw)
To: cypher; +Cc: linux
In message <Pine.LNX.3.95.971203132942.15095B-100000@vertigo.cs.indiana.edu> cypher writes:
: There now exists an emulator for it. Now if one were to port the kernel to
: the emulator and then burn that to a rom.... hmmm
Hmmm. Interesting. How much does it cost to burn a EEPROM for the
N64? I suspect the anser is $500 for the first one, and EEPROM cost +
$0.01 for each additional one due to startup costs.
Also, do you have a URL for the N64 emulator?
If memory serves, the N64 has a VR4700 chip in it made by IDT or some
such. I have an IDT eval board that has a R4700 on it right now,
which would be enough to get Linux running on the chip.
But one does have to ask the question of WHY? What can you get from
an N64 that would make it worth running Linux on it?
Warner
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 21:19 ` Warner Losh
@ 1997-12-03 22:16 ` Chad Campbell
1997-12-03 22:31 ` William J. Earl
[not found] ` <Pine.LNX.3.95.971203161250.15277A-100000@dv123s34.lawrence .ks.us>
1 sibling, 1 reply; 60+ messages in thread
From: Chad Campbell @ 1997-12-03 22:16 UTC (permalink / raw)
To: Warner Losh; +Cc: cypher, linux
On Wed, 3 Dec 1997, Warner Losh wrote:
> But one does have to ask the question of WHY? What can you get from
> an N64 that would make it worth running Linux on it?
I don't know much about the N64, but it should have at least two game
controller ports for a keyboard and a mouse, and I think I heard that
there were storage devices available for it. With an upgrade to 8 MB, it
sounds like it could possibly be made into a cheap NC or WebTV type of
product. Of course, as others have mentioned, any kernel port is valuable
at least in the sense that we might learn something new which could aid in
the design of kernels for other architectures.
Chad
Chad Campbell
Web Manager, University of Kansas Natural History Museum
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 22:16 ` Chad Campbell
@ 1997-12-03 22:31 ` William J. Earl
1997-12-03 22:48 ` Alan Cox
0 siblings, 1 reply; 60+ messages in thread
From: William J. Earl @ 1997-12-03 22:31 UTC (permalink / raw)
To: linux
Chad Campbell writes:
> On Wed, 3 Dec 1997, Warner Losh wrote:
> > But one does have to ask the question of WHY? What can you get from
> > an N64 that would make it worth running Linux on it?
> I don't know much about the N64, but it should have at least two game
> controller ports for a keyboard and a mouse, and I think I heard that
> there were storage devices available for it. With an upgrade to 8 MB, it
> sounds like it could possibly be made into a cheap NC or WebTV type of
> product. Of course, as others have mentioned, any kernel port is valuable
> at least in the sense that we might learn something new which could aid in
> the design of kernels for other architectures.
I have heard of several companies using Linux as a sort of embedded
OS, as would be the case when using it on N64 to build an NC or WebTV device.
One interesting area for development, both in the embedded space, and on
the SGI workstations, would be in regard to realtime support, which is
needed for digital media. By realtime, I mean not fast interrupt response,
but rather precisely scheduled repetitive events (such as driving the Indy
audio ports every 1 ms. with at most 0.9 ms delay no matter what else is
going on).
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
[not found] ` <Pine.LNX.3.95.971203161250.15277A-100000@dv123s34.lawrence .ks.us>
@ 1997-12-03 22:38 ` Douglas F. Elznic
0 siblings, 0 replies; 60+ messages in thread
From: Douglas F. Elznic @ 1997-12-03 22:38 UTC (permalink / raw)
To: Chad Campbell, Warner Losh; +Cc: cypher, linux
At 04:16 PM 12/3/97 -0600, Chad Campbell wrote:
>On Wed, 3 Dec 1997, Warner Losh wrote:
>> But one does have to ask the question of WHY? What can you get from
>> an N64 that would make it worth running Linux on it?
>I don't know much about the N64, but it should have at least two game
>controller ports for a keyboard and a mouse, and I think I heard that
>there were storage devices available for it. With an upgrade to 8 MB, it
>sounds like it could possibly be made into a cheap NC or WebTV type of
>product. Of course, as others have mentioned, any kernel port is valuable
>at least in the sense that we might learn something new which could aid in
>the design of kernels for other architectures.
>
>
Namely the MIPS arch.
--
Douglas F. Elznic
Email:
delznic@acm.org
Pager:
3154492502.2423685@pagenet.net
"The tree of liberty must be refreshed from time to time with the blood of
patriots and tyrants."
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 22:31 ` William J. Earl
@ 1997-12-03 22:48 ` Alan Cox
1997-12-03 22:48 ` Alan Cox
0 siblings, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-03 22:48 UTC (permalink / raw)
To: William J. Earl; +Cc: linux
> One interesting area for development, both in the embedded space, and on
> the SGI workstations, would be in regard to realtime support, which is
> needed for digital media. By realtime, I mean not fast interrupt response,
> but rather precisely scheduled repetitive events (such as driving the Indy
> audio ports every 1 ms. with at most 0.9 ms delay no matter what else is
> going on).
This exists although not for the MIPS platform and not directly accessible
from user space. NMT have a RTLinux kernel that allows real time drivers
to hit targets well well under 0.1ms.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 22:48 ` Alan Cox
@ 1997-12-03 22:48 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-03 22:48 UTC (permalink / raw)
To: William J. Earl; +Cc: linux
> One interesting area for development, both in the embedded space, and on
> the SGI workstations, would be in regard to realtime support, which is
> needed for digital media. By realtime, I mean not fast interrupt response,
> but rather precisely scheduled repetitive events (such as driving the Indy
> audio ports every 1 ms. with at most 0.9 ms delay no matter what else is
> going on).
This exists although not for the MIPS platform and not directly accessible
from user space. NMT have a RTLinux kernel that allows real time drivers
to hit targets well well under 0.1ms.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-03 7:46 ` Chris Carlson
@ 1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:39 ` Lige Hensley
` (3 more replies)
0 siblings, 4 replies; 60+ messages in thread
From: Lige Hensley @ 1997-12-04 5:39 UTC (permalink / raw)
To: Chris Carlson; +Cc: linux
On Tue, 2 Dec 1997, Chris Carlson wrote:
>
> I also would like to see Linux ported to the O2 but also don't expect it to
> happen in the near future. We have to remember that SGI is a "hardware
> company" and did not get where it is because of software. To ask them to
> divulge the inner workings of their bread-and-butter product, thus allowing
> competitors to reproduce and improve upon the design is asking for disaster.
> SGI does not fabricate their own chips and they have done their best to
> provide access to their source (as long as it doesn't compromise their
> hardware investment). The only thing they have to sell is their design.
> Since MIPS is not faring very well against competitors and there seems to me
> a bit of doubt that they can stay ahead, what else does SGI have to sell?
> Just their internal architecture and IRIX.
I was under the impression that MIPS was doing pretty well. Someone
correct me if I am wrong, but I remember reading that the vast majority of
MIPS chips wound up in things other than SGI computers (i.e. Sony
Playstation, Nintendo 64's, etc.).
>From what I read about the R12K, MIPS would appear to be near the top of
the CPU game as far as raw procesing power. When compared to a Pentium-II
350 MHz, a 300MHz R12K was about equal in integer ops and nearly 5 times
the speed with floating point ops. It's not an Alpha, but it far better
than anything cooked up at Intel. (ok, maybe I am a little biased, but
real men use RISC).
-lige
----------------------------------------------------------------
Slam Foot Neck! Ride the wave. Touch Touch Touch. I'm cereal.
e-mail: lige.hensley@carpediem.com
www: http://www.iusd.iupui.edu/~henslelf
o__
,>/' Network Coordinator. Compression, Inc.
(_)\(_) http://www.carpediem.com
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 5:39 ` Lige Hensley
@ 1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:48 ` David S. Miller
` (2 subsequent siblings)
3 siblings, 0 replies; 60+ messages in thread
From: Lige Hensley @ 1997-12-04 5:39 UTC (permalink / raw)
To: Chris Carlson; +Cc: linux
On Tue, 2 Dec 1997, Chris Carlson wrote:
>
> I also would like to see Linux ported to the O2 but also don't expect it to
> happen in the near future. We have to remember that SGI is a "hardware
> company" and did not get where it is because of software. To ask them to
> divulge the inner workings of their bread-and-butter product, thus allowing
> competitors to reproduce and improve upon the design is asking for disaster.
> SGI does not fabricate their own chips and they have done their best to
> provide access to their source (as long as it doesn't compromise their
> hardware investment). The only thing they have to sell is their design.
> Since MIPS is not faring very well against competitors and there seems to me
> a bit of doubt that they can stay ahead, what else does SGI have to sell?
> Just their internal architecture and IRIX.
I was under the impression that MIPS was doing pretty well. Someone
correct me if I am wrong, but I remember reading that the vast majority of
MIPS chips wound up in things other than SGI computers (i.e. Sony
Playstation, Nintendo 64's, etc.).
From what I read about the R12K, MIPS would appear to be near the top of
the CPU game as far as raw procesing power. When compared to a Pentium-II
350 MHz, a 300MHz R12K was about equal in integer ops and nearly 5 times
the speed with floating point ops. It's not an Alpha, but it far better
than anything cooked up at Intel. (ok, maybe I am a little biased, but
real men use RISC).
-lige
----------------------------------------------------------------
Slam Foot Neck! Ride the wave. Touch Touch Touch. I'm cereal.
e-mail: lige.hensley@carpediem.com
www: http://www.iusd.iupui.edu/~henslelf
o__
,>/' Network Coordinator. Compression, Inc.
(_)\(_) http://www.carpediem.com
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:39 ` Lige Hensley
@ 1997-12-04 5:48 ` David S. Miller
1997-12-04 16:55 ` Greg Chesson
1997-12-04 16:51 ` Greg Chesson
1997-12-04 22:08 ` David Chatterton
3 siblings, 1 reply; 60+ messages in thread
From: David S. Miller @ 1997-12-04 5:48 UTC (permalink / raw)
To: ligeh; +Cc: cwcarlson, linux
Date: Thu, 4 Dec 1997 00:39:07 -0500 (EST)
From: Lige Hensley <ligeh@carpediem.com>
From what I read about the R12K, MIPS would appear to be near the top of
the CPU game as far as raw procesing power. When compared to a Pentium-II
350 MHz, a 300MHz R12K was about equal in integer ops and nearly 5 times
the speed with floating point ops. It's not an Alpha, but it far better
than anything cooked up at Intel.
Isn't R12K pre-production still, out in a few months or something like
this? Anyways, if this is the case you might as well compare in the
predicted Ultra III figures in your analysis as well if this is a
vapor ware cpu performance comparison ;-)
(ok, maybe I am a little biased, but real men use RISC).
Indeed.
Later,
David S. Miller
davem@dm.cobaltmicro.com
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 22:08 ` David Chatterton
@ 1997-12-04 14:06 ` Lige Hensley
1997-12-04 17:03 ` Greg Chesson
1 sibling, 0 replies; 60+ messages in thread
From: Lige Hensley @ 1997-12-04 14:06 UTC (permalink / raw)
To: David Chatterton; +Cc: Chris Carlson, linux
On Thu, 4 Dec 1997, David Chatterton wrote:
> On Dec 4, 12:39am, Lige Hensley wrote:
> | >From what I read about the R12K, MIPS would appear to be near the top of
> | the CPU game as far as raw procesing power. When compared to a Pentium-II
> | 350 MHz, a 300MHz R12K was about equal in integer ops and nearly 5 times
> | the speed with floating point ops. It's not an Alpha, but it far better
> | than anything cooked up at Intel. (ok, maybe I am a little biased, but
> | real men use RISC).
> |
>
> But the best we have right now are 200MHz R10Ks, so we are relying on our
> server architecture to compete with everyone else. We all hope that MIPS can
> continue to produce chips that are competitive (and the R12 looks great on
> paper), but Intel has a lot more engineers and R&D budget to leap frog ahead in
> the long run. That's life.
As someone mentioned earlier (and what I am lead to understand) the R12K
is in pre-production now, which is much more than a concept on paper. Is
this correct?
As far as Intel being able to leap-frog ahead in the long-run, I wouldn't
count on it. I've always believed that money can't buy everything,
especially in engineering.
Or maybe I like to root for the underdog :-)
-lige
----------------------------------------------------------------
Slam Foot Neck! Ride the wave. Touch Touch Touch. I'm cereal.
e-mail: lige.hensley@carpediem.com
www: http://www.iusd.iupui.edu/~henslelf
o__
,>/' Network Coordinator. Compression, Inc.
(_)\(_) http://www.carpediem.com
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:48 ` David S. Miller
@ 1997-12-04 16:51 ` Greg Chesson
1997-12-04 22:08 ` David Chatterton
3 siblings, 0 replies; 60+ messages in thread
From: Greg Chesson @ 1997-12-04 16:51 UTC (permalink / raw)
To: Lige Hensley, Chris Carlson; +Cc: linux
Mips does quite well with cpus in the embedded systems area: about
35 Million components/year at recent count. This area accounts for
good revenue for Mips and its vlsi partners.
The R12000, which is an improved R10000, and its planned shrinks
and process upgrades are competitive with other processors in all areas
except for extreme floating point. Mips had a best-of-breed raw floating
point cpu project that was cancelled because it was deemed not competitive
enough in integer performance to serve in all but a few platforms.
Mips processors tend to be employed with more powerful cache and memory
and io subsystems than equivalent Intel parts. This natually leads
to differences in cost/price.
Another little-appreciated factoid: in order to comply with Intel specs
for system transceiver components used with high-performance Intel cpus,
one needs to use a $25 component.
Nearly all cheap/fast motherboards use a $2 component that works
"most of the time."
g
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 5:48 ` David S. Miller
@ 1997-12-04 16:55 ` Greg Chesson
0 siblings, 0 replies; 60+ messages in thread
From: Greg Chesson @ 1997-12-04 16:55 UTC (permalink / raw)
To: David S. Miller, ligeh; +Cc: cwcarlson, linux
First silicon R12Ks exist and are being tested in the lab.
It is premature to make any predictions about eventual clock rate.
Production manufactured systems are usually available about 12 months
from first silicon of the processor chip. In the case of the R12K
this time will probably be reduced because the chips are going into
the "stable" R10K hardware/software environment.
g
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 22:08 ` David Chatterton
1997-12-04 14:06 ` Lige Hensley
@ 1997-12-04 17:03 ` Greg Chesson
1997-12-04 17:22 ` William J. Earl
1997-12-04 18:54 ` Alan Cox
1 sibling, 2 replies; 60+ messages in thread
From: Greg Chesson @ 1997-12-04 17:03 UTC (permalink / raw)
To: David Chatterton, Lige Hensley, Chris Carlson; +Cc: linux
I've been reading the O2 thread with great interest.
It seems to me that a "vanilla" port of Linux on the O2 is quite feasible.
Several writers have agreed.
But full support in Linux for extremem platform-dependent multimedia
hardware is problematic. This has also been mentioned by several writers,
some of whom developed the OS code. The problems are twofold: it would
be a non-trivial task even with full documentation to provide complete
support for O2 hardware since the job would require as much new invention
as it would require porting, and second- full disclosure
of the inner workings is viewed as too much of a giveaway to present
and future competitors.
Makes me wonder whether the Indy port of Linux could be ratcheted into
a vanilla port for the O2, and then perhaps SGI could produce binary
loadable drivers for the missing bits.
g
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 17:03 ` Greg Chesson
@ 1997-12-04 17:22 ` William J. Earl
1997-12-04 18:59 ` Alan Cox
1997-12-04 22:01 ` ralf
1997-12-04 18:54 ` Alan Cox
1 sibling, 2 replies; 60+ messages in thread
From: William J. Earl @ 1997-12-04 17:22 UTC (permalink / raw)
To: Greg Chesson; +Cc: David Chatterton, Lige Hensley, Chris Carlson, linux
Greg Chesson writes:
> I've been reading the O2 thread with great interest.
>
> It seems to me that a "vanilla" port of Linux on the O2 is quite feasible.
> Several writers have agreed.
>
> But full support in Linux for extremem platform-dependent multimedia
> hardware is problematic. This has also been mentioned by several writers,
> some of whom developed the OS code. The problems are twofold: it would
> be a non-trivial task even with full documentation to provide complete
> support for O2 hardware since the job would require as much new invention
> as it would require porting, and second- full disclosure
> of the inner workings is viewed as too much of a giveaway to present
> and future competitors.
>
> Makes me wonder whether the Indy port of Linux could be ratcheted into
> a vanilla port for the O2, and then perhaps SGI could produce binary
> loadable drivers for the missing bits.
For IRIX, most of the O2 work was really in areas which, in
linux, would be reasonably general purpose. In particular, being able
to allocate large pages (64 KB tiles on O2) at any time, no matter how
long the system has been running, was a major change, and would be
useful for performance reasons on any system with variable page size,
since it helps ordinary programs with large address spaces on
large-memory systems, as well as supporting the O2 graphics and
digital media. Similarly, having solid time-scheduled realtime
support is broadly useful. For digital media, one needs a reasonably
portable API, and applications on top of it, which supports high
performance. With those facilities in hand, doing the drivers is not
all that difficult. Providing those facilities, however, is a substantial
project, and not especially tied to SGI hardware.
In graphics, if linux winds up with an IRIX-compatible kernel
graphics driver on some SGI boxes, the IRIX Xsgi, libgl.so, and
libGL.so platform-specific binaries could run on linux. Note that
each of our systems comes with a license for the then-current version
of these components, so loading linux does not preclude using the IRIX
components, although, at present, they may not be shipped in a linux
distribution. Since they are only useful on SGI hardware, perhaps SGI
could make arrangements to permit "stable" binaries to be included in
linux distributions for selected platforms.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 17:03 ` Greg Chesson
1997-12-04 17:22 ` William J. Earl
@ 1997-12-04 18:54 ` Alan Cox
1997-12-04 18:54 ` Alan Cox
` (2 more replies)
1 sibling, 3 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-04 18:54 UTC (permalink / raw)
To: Greg Chesson; +Cc: chatz, ligeh, cwcarlson, linux
> Makes me wonder whether the Indy port of Linux could be ratcheted into
> a vanilla port for the O2, and then perhaps SGI could produce binary
> loadable drivers for the missing bits.
If that is SGI's attitude to the thing then theres an Indy on my desk waiting
for someone to ship it back.
Actually Ariel there sort of is - unless anyone knows why the monitor would
have decided to fold the middle third of the picture in really oddly ..
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 18:54 ` Alan Cox
@ 1997-12-04 18:54 ` Alan Cox
1997-12-04 19:41 ` Alex deVries
1997-12-05 0:17 ` ralf
2 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-04 18:54 UTC (permalink / raw)
To: Greg Chesson; +Cc: chatz, ligeh, cwcarlson, linux
> Makes me wonder whether the Indy port of Linux could be ratcheted into
> a vanilla port for the O2, and then perhaps SGI could produce binary
> loadable drivers for the missing bits.
If that is SGI's attitude to the thing then theres an Indy on my desk waiting
for someone to ship it back.
Actually Ariel there sort of is - unless anyone knows why the monitor would
have decided to fold the middle third of the picture in really oddly ..
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 17:22 ` William J. Earl
@ 1997-12-04 18:59 ` Alan Cox
1997-12-04 18:59 ` Alan Cox
1997-12-04 22:01 ` ralf
1 sibling, 1 reply; 60+ messages in thread
From: Alan Cox @ 1997-12-04 18:59 UTC (permalink / raw)
To: William J. Earl; +Cc: greg, chatz, ligeh, cwcarlson, linux
> linux, would be reasonably general purpose. In particular, being able
> to allocate large pages (64 KB tiles on O2) at any time, no matter how
I assume you mean physical tiles. 64K tiles on Linux is actually quite tricky
- creating them and allocating them is ok its the fact theres no current
vm way to say move this and make me a tile. (the kernel allocated objects
arent a problem as the kernel can allocate drawing from tiles itself)
> In graphics, if linux winds up with an IRIX-compatible kernel
> graphics driver on some SGI boxes, the IRIX Xsgi, libgl.so, and
> libGL.so platform-specific binaries could run on linux. Note that
This is certainly a very good starting basis but its not a "complete"
answer. Also of course one day Indy + IRIX always together wont be an
automatic fact - I dont actually believe it is in the EEC right now anyway
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 18:59 ` Alan Cox
@ 1997-12-04 18:59 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-04 18:59 UTC (permalink / raw)
To: William J. Earl; +Cc: greg, chatz, ligeh, cwcarlson, linux
> linux, would be reasonably general purpose. In particular, being able
> to allocate large pages (64 KB tiles on O2) at any time, no matter how
I assume you mean physical tiles. 64K tiles on Linux is actually quite tricky
- creating them and allocating them is ok its the fact theres no current
vm way to say move this and make me a tile. (the kernel allocated objects
arent a problem as the kernel can allocate drawing from tiles itself)
> In graphics, if linux winds up with an IRIX-compatible kernel
> graphics driver on some SGI boxes, the IRIX Xsgi, libgl.so, and
> libGL.so platform-specific binaries could run on linux. Note that
This is certainly a very good starting basis but its not a "complete"
answer. Also of course one day Indy + IRIX always together wont be an
automatic fact - I dont actually believe it is in the EEC right now anyway
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 18:54 ` Alan Cox
1997-12-04 18:54 ` Alan Cox
@ 1997-12-04 19:41 ` Alex deVries
1997-12-04 22:28 ` Eric Kimminau
1997-12-05 0:17 ` ralf
2 siblings, 1 reply; 60+ messages in thread
From: Alex deVries @ 1997-12-04 19:41 UTC (permalink / raw)
To: SGI Linux
On Thu, 4 Dec 1997, Alan Cox wrote:
>
> Actually Ariel there sort of is - unless anyone knows why the monitor would
> have decided to fold the middle third of the picture in really oddly ..
Not to be too critical, but my monitor only sort-of works. Everyonce in
awhile it'll black out, and I won't see anything until I give it a whack.
I think there's a wire loose in there. It's not too troublesome, but my
neighbours are starting to wonder.
Anyway, I'm still happy for free hardware.
- A
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 17:22 ` William J. Earl
1997-12-04 18:59 ` Alan Cox
@ 1997-12-04 22:01 ` ralf
1997-12-04 23:33 ` William J. Earl
1 sibling, 1 reply; 60+ messages in thread
From: ralf @ 1997-12-04 22:01 UTC (permalink / raw)
To: William J. Earl
Cc: Greg Chesson, David Chatterton, Lige Hensley, Chris Carlson,
linux
On Thu, Dec 04, 1997 at 09:22:12AM -0800, William J. Earl wrote:
> Greg Chesson writes:
> > I've been reading the O2 thread with great interest.
> >
> > It seems to me that a "vanilla" port of Linux on the O2 is quite feasible.
> > Several writers have agreed.
> >
> > But full support in Linux for extremem platform-dependent multimedia
> > hardware is problematic. This has also been mentioned by several writers,
> > some of whom developed the OS code. The problems are twofold: it would
> > be a non-trivial task even with full documentation to provide complete
> > support for O2 hardware since the job would require as much new invention
> > as it would require porting, and second- full disclosure
> > of the inner workings is viewed as too much of a giveaway to present
> > and future competitors.
> >
> > Makes me wonder whether the Indy port of Linux could be ratcheted into
> > a vanilla port for the O2, and then perhaps SGI could produce binary
> > loadable drivers for the missing bits.
>
> For IRIX, most of the O2 work was really in areas which, in
> linux, would be reasonably general purpose. In particular, being able
> to allocate large pages (64 KB tiles on O2) at any time, no matter how
> long the system has been running, was a major change, and would be
> useful for performance reasons on any system with variable page size,
> since it helps ordinary programs with large address spaces on
> large-memory systems, as well as supporting the O2 graphics and
> digital media. Similarly, having solid time-scheduled realtime
> support is broadly useful. For digital media, one needs a reasonably
> portable API, and applications on top of it, which supports high
> performance. With those facilities in hand, doing the drivers is not
> all that difficult. Providing those facilities, however, is a substantial
> project, and not especially tied to SGI hardware.
Indeed - and you're pointing to what I consider _the_ problem with
current Linux kernels. Linux uses the ``buddy system'' described by
Donald Knuth's ``Algorithems And Data Structures'' to maintain it's
free pages. This algorithem results in massive fragmentation even
after a short uptime.
On a Linux box you can press <shift> + <scroll-lock> and the kernel will
print some memory statistics to the console or syslog. For the
purpose of illustrating things I've appended those numbers from a
couple of systems. As it is easy to see Linux doesn't do a good job
at all.
As a result certain kernel subsystems that need large buffers may fail
on a loaded machine. Typical example is the floppy driver allocating
a track sized buffer. Much worse - the networking code sometimes does
atomic allocations. Atomic means it cannot sleep for swapping etc.
On machines that have sufficiently fragment memory as a result network
services like NFS using >=4kb blocks (results in allocation of 8kb
blocks) are stalling.
For my 8mb PC this may mean it will sometime stop NFS serving when
nothing else is running at the same time just because the compile I
was running before resulted in completly fragmented memory leaving
no chunks bigger than a single 4kb page. Unfortunately there is nothing
that 100% prevents this to happen under all circumstances on machines
with more memory. I've seen reports of machines with >128mb of memory
having 2mb of free memory left all in 4kb chunks.
> In graphics, if linux winds up with an IRIX-compatible kernel
> graphics driver on some SGI boxes, the IRIX Xsgi, libgl.so, and
> libGL.so platform-specific binaries could run on linux. Note that
> each of our systems comes with a license for the then-current version
> of these components, so loading linux does not preclude using the IRIX
> components, although, at present, they may not be shipped in a linux
> distribution. Since they are only useful on SGI hardware, perhaps SGI
> could make arrangements to permit "stable" binaries to be included in
> linux distributions for selected platforms.
This sounds like a possible way to go. From the technical point of view
I'd prefer to rebuild those binaries - if possible - as native Linux
binaries. It'd reduce memory usage significantly.
As far as technical documentation about the O2 goes - does a model like
- full disclosure of documentation under a NDA to some Linux developers
- the NDA would contain an expiration date, maybe one or two years, after
which the covered information can be disclosed because SGI does no
longer consider nondisclosure a vital for it's bussines.
- disclosure as binaries:
Parts that are considered to be based on information that SGI thinks
should not be disclosed could be made available as binaries.
- disclosure as source code:
This would be drivers for disk/network (as I understand SGI doesn't
consider this part of the O2 to be too valuable) and a part of the
technical details about GFX that allow to implement a reasonable ASCII
console, maybe a simple X server and to spread these components as
source.
sound acceptable? While I believe we should currently not try to go for
an O2 port because we still have same homework to do, I also think it's
important that we find a solution that allows Linux to be used on SGI
boxes before they can be considered old iron like the Indy.
Another important thing to consider is that porting Linux to the O2 takes
time. So until we Linux developers have produced an acceptable O2
implementation an NDA wouldn't even hurt the idea of Free Software very
much. It's also time during which the value of the non-disclosed
information for SGI decreases, so why shouldn't we - unlike for the
Indy port - try to overlap time of non-disclosure of technical information
with working on the port?
Another argument for or port to current SGI hardware - many people
want to prototype their embedded applications on a desktop using the same
operating system and CPU type. For many embedded applications Linux is
one of the operating system of choice, after all Linux has many points
where it beats IRIX. Most important the price per license. For many
embedded applications MIPS is the CPU of choice. So for developers of
such appliances running Linux, not IRIX, on an O2 could well be the OS of
choice even at the price of loosing IRIX's advantages.
One of the people interested in Linux for such an MIPS based appliance
being prototyped on a UNIX system was Kevin Kissel (kevink@neu.sgi.com).
So I guess even for SGI having a Linux desktop should be interesting.
Ralf
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 5:39 ` Lige Hensley
` (2 preceding siblings ...)
1997-12-04 16:51 ` Greg Chesson
@ 1997-12-04 22:08 ` David Chatterton
1997-12-04 14:06 ` Lige Hensley
1997-12-04 17:03 ` Greg Chesson
3 siblings, 2 replies; 60+ messages in thread
From: David Chatterton @ 1997-12-04 22:08 UTC (permalink / raw)
To: Lige Hensley, Chris Carlson; +Cc: linux
On Dec 4, 12:39am, Lige Hensley wrote:
| Subject: Re: Linux on the O2
| On Tue, 2 Dec 1997, Chris Carlson wrote:
|
| >
| > I also would like to see Linux ported to the O2 but also don't expect it to
| > happen in the near future. We have to remember that SGI is a "hardware
| > company" and did not get where it is because of software. To ask them to
| > divulge the inner workings of their bread-and-butter product, thus allowing
| > competitors to reproduce and improve upon the design is asking for
disaster.
| > SGI does not fabricate their own chips and they have done their best to
| > provide access to their source (as long as it doesn't compromise their
| > hardware investment). The only thing they have to sell is their design.
| > Since MIPS is not faring very well against competitors and there seems to
me
| > a bit of doubt that they can stay ahead, what else does SGI have to sell?
| > Just their internal architecture and IRIX.
|
| I was under the impression that MIPS was doing pretty well. Someone
| correct me if I am wrong, but I remember reading that the vast majority of
| MIPS chips wound up in things other than SGI computers (i.e. Sony
| Playstation, Nintendo 64's, etc.).
|
That's the point, MIPS is making a healthy profit from _non-SGI_ products
though like Windows CE, N64, Playstation as you suggested. Not that I have a
problem with that...
| >From what I read about the R12K, MIPS would appear to be near the top of
| the CPU game as far as raw procesing power. When compared to a Pentium-II
| 350 MHz, a 300MHz R12K was about equal in integer ops and nearly 5 times
| the speed with floating point ops. It's not an Alpha, but it far better
| than anything cooked up at Intel. (ok, maybe I am a little biased, but
| real men use RISC).
|
But the best we have right now are 200MHz R10Ks, so we are relying on our
server architecture to compete with everyone else. We all hope that MIPS can
continue to produce chips that are competitive (and the R12 looks great on
paper), but Intel has a lot more engineers and R&D budget to leap frog ahead in
the long run. That's life.
David
--
David Chatterton (61-3) 9882 8211 (Tel)
R&D Software Engineer (61-3) 9882 8030 (Fax)
Performance Tools Group http://reality.sgi.com/chatz
Silicon Graphics Pty.Ltd., 357 Camberwell Rd, Melbourne, Australia
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 19:41 ` Alex deVries
@ 1997-12-04 22:28 ` Eric Kimminau
1997-12-04 22:44 ` Alan Cox
0 siblings, 1 reply; 60+ messages in thread
From: Eric Kimminau @ 1997-12-04 22:28 UTC (permalink / raw)
To: Alex deVries; +Cc: SGI Linux
Alex deVries wrote:
>
> On Thu, 4 Dec 1997, Alan Cox wrote:
> >
> > Actually Ariel there sort of is - unless anyone knows why the monitor would
> > have decided to fold the middle third of the picture in really oddly ..
>
> Not to be too critical, but my monitor only sort-of works. Everyonce in
> awhile it'll black out, and I won't see anything until I give it a whack.
> I think there's a wire loose in there. It's not too troublesome, but my
> neighbours are starting to wonder.
>
> Anyway, I'm still happy for free hardware.
>
> - A
To both Alex and Alan: Both of those systems "should" be under SGI
support. If you will run netstat -ia and give me the serial numbers
(08:00:69:xx:xx:xx) Ill look them up to make sure. If they do, you can
just call 1-800-800-4SGI (4744) and have them ship you brand new
monitors.
Eric.
--
Eric Kimminau System Engineer/RSA
eak@detroit.sgi.com Silicon Graphics, Inc
Voice: (248) 848-4455 39001 West 12 Mile Rd.
Fax: (248) 848-5600 Farmington, MI 48331-2903
VNet Extension - 6-327-4455
"I speak my mind and no one else's."
http://www.dcs.ex.ac.uk/~aba/rsa/perl-rsa-sig.html
When confronted by a difficult problem, solve it by reducing
it to the question, "How would the Lone Ranger handle this?"
"I am the great supportfolio, do you have http?"
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 22:28 ` Eric Kimminau
@ 1997-12-04 22:44 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-04 22:44 UTC (permalink / raw)
To: eak; +Cc: adevries, linux
> To both Alex and Alan: Both of those systems "should" be under SGI
> support. If you will run netstat -ia and give me the serial numbers
> (08:00:69:xx:xx:xx) Ill look them up to make sure. If they do, you can
> just call 1-800-800-4SGI (4744) and have them ship you brand new
> monitors.
Ok I'll chase your UK office. Firstly it might be something dumb I can do to
make it come back right 8)
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 22:01 ` ralf
@ 1997-12-04 23:33 ` William J. Earl
1997-12-05 1:33 ` ralf
0 siblings, 1 reply; 60+ messages in thread
From: William J. Earl @ 1997-12-04 23:33 UTC (permalink / raw)
To: ralf
Cc: William J. Earl, Greg Chesson, David Chatterton, Lige Hensley,
Chris Carlson, linux
ralf@uni-koblenz.de writes:
...
> Indeed - and you're pointing to what I consider _the_ problem with
> current Linux kernels. Linux uses the ``buddy system'' described by
> Donald Knuth's ``Algorithems And Data Structures'' to maintain it's
> free pages. This algorithem results in massive fragmentation even
> after a short uptime.
Are you saying that linux uses the buddy system for all of memory,
or just for the kernel heap? (I would be surprised if it were used
for other than the kernel heap, although that is bad enough.)
...
> For my 8mb PC this may mean it will sometime stop NFS serving when
> nothing else is running at the same time just because the compile I
> was running before resulted in completly fragmented memory leaving
> no chunks bigger than a single 4kb page. Unfortunately there is nothing
> that 100% prevents this to happen under all circumstances on machines
> with more memory. I've seen reports of machines with >128mb of memory
> having 2mb of free memory left all in 4kb chunks.
One thing we do in IRIX is to always allocate kernel heap buffers
of 1 page or larger as an integral set of pages, mapped into kernel
virtual space, but not part of the main kernel heap (which is used only
for smaller buffers). Except for fragmentation of the kernel mapped
space pool (a pool of address space, not real memory), this guarantees
you can always get large buffers if pages are available. Fragmentation
of smaller buffers of course requires a better heap manager, although
using zone allocation (where a zone has blocks all of the same size,
and there are zones for most popular sizes) helps a lot and is also
faster than using the regular heap manager.
...
> As far as technical documentation about the O2 goes - does a model like
>
> - full disclosure of documentation under a NDA to some Linux developers
> - the NDA would contain an expiration date, maybe one or two years, after
> which the covered information can be disclosed because SGI does no
> longer consider nondisclosure a vital for it's bussines.
> - disclosure as binaries:
> Parts that are considered to be based on information that SGI thinks
> should not be disclosed could be made available as binaries.
> - disclosure as source code:
> This would be drivers for disk/network (as I understand SGI doesn't
> consider this part of the O2 to be too valuable) and a part of the
> technical details about GFX that allow to implement a reasonable ASCII
> console, maybe a simple X server and to spread these components as
> source.
>
> sound acceptable? While I believe we should currently not try to go for
> an O2 port because we still have same homework to do, I also think it's
> important that we find a solution that allows Linux to be used on SGI
> boxes before they can be considered old iron like the Indy.
We (the SGI people on the list) will have to talk with the
general manager for the O2 platform. I personally think that would be
ok. As I mentioned in my earlier message, a lot of the IRIX value
added for O2 is in essentially generic facilities, not in the
device-dependent drivers, and we already tell customers in high level
terms about the overall architecture (such as the unified memory for
graphics and I/O), so there is not much in the hardware documentation
which needs to be treated as a trade secret. Conceptually, the O2 graphics
pipeline is fully described by the OpenGL reference (software) implementation
and specification, so the hardware value added is in the internals
of the implementation, not in the interface (which is essentially
a large part of the OpenGL pipeline).
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 18:54 ` Alan Cox
1997-12-04 18:54 ` Alan Cox
1997-12-04 19:41 ` Alex deVries
@ 1997-12-05 0:17 ` ralf
1997-12-05 9:06 ` Alan Cox
2 siblings, 1 reply; 60+ messages in thread
From: ralf @ 1997-12-05 0:17 UTC (permalink / raw)
To: Alan Cox; +Cc: Greg Chesson, chatz, ligeh, cwcarlson, linux
On Thu, Dec 04, 1997 at 06:54:18PM +0000, Alan Cox wrote:
> > Makes me wonder whether the Indy port of Linux could be ratcheted into
> > a vanilla port for the O2, and then perhaps SGI could produce binary
> > loadable drivers for the missing bits.
>
> If that is SGI's attitude to the thing then theres an Indy on my desk waiting
> for someone to ship it back.
Actually I think there is a lot that can be negotiated. In one of my
previous postings I've already gone a bit into this. Some people at
SGI are *alot* more Linux-friendly than you think. Furthermore given the
fact that way more UNIX systems run Linux than anything else definately
has impressed a couple of people in the higher levels at SGI.
Given that the way the Free Software comunity and the world of the
proprietary software think is radically different we have to accept
that a big company like SGI doesn't take make the way from proprietary
software to Free Software in one big jump. Between black and white
there is grey ...
And after all - resistance is futile. "You can't outrun Penguin from
Borg. You can't destroy him. If you damage him, the essence of what
he is remains ... he regenerates and keeps coming. Eventually you'll
weaken. Your reserves will be gone. His is relentless"
- Q, Star Date 42761.3
;-)
Ralf
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-04 23:33 ` William J. Earl
@ 1997-12-05 1:33 ` ralf
1997-12-05 3:25 ` Chad Campbell
0 siblings, 1 reply; 60+ messages in thread
From: ralf @ 1997-12-05 1:33 UTC (permalink / raw)
To: William J. Earl
Cc: Greg Chesson, David Chatterton, Lige Hensley, Chris Carlson,
linux
On Thu, Dec 04, 1997 at 03:33:01PM -0800, William J. Earl wrote:
> > Indeed - and you're pointing to what I consider _the_ problem with
> > current Linux kernels. Linux uses the ``buddy system'' described by
> > Donald Knuth's ``Algorithems And Data Structures'' to maintain it's
> > free pages. This algorithem results in massive fragmentation even
> > after a short uptime.
>
> Are you saying that linux uses the buddy system for all of memory,
> or just for the kernel heap? (I would be surprised if it were used
> for other than the kernel heap, although that is bad enough.)
The buddy system is only used to maintain the pool of free pages. The
buddy system has the advantage that it is very fast. On top of this
lowest level we've got additional layers:
- The ``slab allocator''. It's basically what Jeff Bonwick from Sun
describes in his USENIX paper from '94. Miguel has the paper on his
homepage. The slab allocator has been added during the Linux 2.1.x
series.
- The ``simp allocator''. Yet another memory allocator for cached objects
that has been added during 2.1.x. Pretty fast and still under
development.
- kmalloc() is the kernel equivallent to malloc(3). It's mostly used
for allocations smaller than a single page but can be used to allocate
upto 128kb. For Linux 2.0 kmalloc() is getting it's pages directly from
the free page pool. For Linux 2.1.x kmalloc is implemented on top of
the slab allocator.
Both implementations are directly or indirectly getting the memory from
the pool of free pages (that's KSEG0 on MIPS), therefore have to
live with the advantages and disadvantages of the buddy system.
> One thing we do in IRIX is to always allocate kernel heap buffers
> of 1 page or larger as an integral set of pages, mapped into kernel
> virtual space, but not part of the main kernel heap (which is used only
> for smaller buffers). Except for fragmentation of the kernel mapped
> space pool (a pool of address space, not real memory), this guarantees
> you can always get large buffers if pages are available. Fragmentation
> of smaller buffers of course requires a better heap manager, although
> using zone allocation (where a zone has blocks all of the same size,
> and there are zones for most popular sizes) helps a lot and is also
> faster than using the regular heap manager.
Actually we also have this type of kernel virtual memory in Linux. On
MIPS the address space >=KSEG2 is being used for that purpose. The
functions vmalloc(9) and vfree(9) are used for that purpose. However
vmalloc is rarely being used in the kernel. Among the reasons is that
vmalloc() is slower than other types of memory allocation. Furthermore
the primitive PC-style DMA hardware often does not have the required
support to use vmalloc'ed memory as scatter/gather buffer. Finally
accessing vmalloced memory may result in TLB reloads on some architectures
while pages allocated from the pool of free pages don't. So accessing
vmalloced memory may imply a performance penalty.
> We (the SGI people on the list) will have to talk with the
> general manager for the O2 platform. I personally think that would be
> ok. As I mentioned in my earlier message, a lot of the IRIX value
> added for O2 is in essentially generic facilities, not in the
> device-dependent drivers, and we already tell customers in high level
> terms about the overall architecture (such as the unified memory for
> graphics and I/O), so there is not much in the hardware documentation
> which needs to be treated as a trade secret. Conceptually, the O2 graphics
> pipeline is fully described by the OpenGL reference (software) implementation
> and specification, so the hardware value added is in the internals
> of the implementation, not in the interface (which is essentially
> a large part of the OpenGL pipeline).
Well, sounds good. You're making the differenciation between the interface
and the implementation of the O2, something that so far people I've been
discussing with haven't done. As a software guy all need is the
interface and maybe a peek into internals for a better understanding of
the interface.
Ralf
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-05 1:33 ` ralf
@ 1997-12-05 3:25 ` Chad Campbell
0 siblings, 0 replies; 60+ messages in thread
From: Chad Campbell @ 1997-12-05 3:25 UTC (permalink / raw)
To: ralf; +Cc: linux
This is my second to last semester, and I'm now in the OS class. I have a
question about page miss resolution. Does the kernel use a global or
local strategy to determine which pages get swapped out? My teacher,
Victor Wallace, has done research on using global resolution strategies,
and has found it to be supperior, but he didn't know of any OS that
actually uses it.
Chad Campbell
Web Manager, University of Kansas Natural History Museum
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: Linux on the O2
1997-12-05 0:17 ` ralf
@ 1997-12-05 9:06 ` Alan Cox
0 siblings, 0 replies; 60+ messages in thread
From: Alan Cox @ 1997-12-05 9:06 UTC (permalink / raw)
To: ralf; +Cc: alan, greg, chatz, ligeh, cwcarlson, linux
> > If that is SGI's attitude to the thing then theres an Indy on my desk waiting
> > for someone to ship it back.
Ok before this goes too far let me point of that there were a certain number
of invisible humour/sarcasm tags on that comment as a lead in to the monitor
info.
To start with if Dave can figure out an ultrasparc and we can crack the
Macintosh 68K nobody in SGI could stop us 8)
Alan
^ permalink raw reply [flat|nested] 60+ messages in thread
end of thread, other threads:[~1997-12-05 9:06 UTC | newest]
Thread overview: 60+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
1997-12-02 21:50 Linux on the O2 cypher
1997-12-02 22:10 ` Ralf Baechle
1997-12-02 22:31 ` Ariel Faigon
1997-12-02 22:31 ` Ariel Faigon
1997-12-03 0:04 ` Alan Cox
1997-12-03 0:46 ` John Wiederhirn
1997-12-03 3:06 ` William J. Earl
1997-12-03 3:06 ` William J. Earl
1997-12-03 18:28 ` cypher
1997-12-03 1:51 ` Douglas F. Elznic
1997-12-03 1:51 ` Douglas F. Elznic
1997-12-03 7:01 ` ralf
[not found] ` <Pine.LNX.3.95.971203132942.15095B-100000@vertigo.cs.indian a.edu>
1997-12-03 18:04 ` Douglas F. Elznic
1997-12-03 18:39 ` Alan Cox
1997-12-03 18:39 ` Alan Cox
1997-12-03 18:42 ` cypher
1997-12-03 18:49 ` cypher
1997-12-03 18:05 ` David Chatterton
1997-12-03 2:34 ` Brendan Black
1997-12-03 2:58 ` Richard Offer
1997-12-03 3:19 ` Douglas F. Elznic
1997-12-03 8:21 ` Oliver Frommel
1997-12-03 10:51 ` Alan Cox
1997-12-03 10:51 ` Alan Cox
1997-12-03 18:32 ` cypher
1997-12-03 18:31 ` cypher
1997-12-03 21:19 ` Warner Losh
1997-12-03 22:16 ` Chad Campbell
1997-12-03 22:31 ` William J. Earl
1997-12-03 22:48 ` Alan Cox
1997-12-03 22:48 ` Alan Cox
[not found] ` <Pine.LNX.3.95.971203161250.15277A-100000@dv123s34.lawrence .ks.us>
1997-12-03 22:38 ` Douglas F. Elznic
1997-12-02 22:33 ` William J. Earl
1997-12-02 23:32 ` Christian Reisch
1997-12-02 23:45 ` William J. Earl
1997-12-03 0:09 ` Alan Cox
1997-12-03 0:09 ` Alan Cox
1997-12-03 7:46 ` Chris Carlson
1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:39 ` Lige Hensley
1997-12-04 5:48 ` David S. Miller
1997-12-04 16:55 ` Greg Chesson
1997-12-04 16:51 ` Greg Chesson
1997-12-04 22:08 ` David Chatterton
1997-12-04 14:06 ` Lige Hensley
1997-12-04 17:03 ` Greg Chesson
1997-12-04 17:22 ` William J. Earl
1997-12-04 18:59 ` Alan Cox
1997-12-04 18:59 ` Alan Cox
1997-12-04 22:01 ` ralf
1997-12-04 23:33 ` William J. Earl
1997-12-05 1:33 ` ralf
1997-12-05 3:25 ` Chad Campbell
1997-12-04 18:54 ` Alan Cox
1997-12-04 18:54 ` Alan Cox
1997-12-04 19:41 ` Alex deVries
1997-12-04 22:28 ` Eric Kimminau
1997-12-04 22:44 ` Alan Cox
1997-12-05 0:17 ` ralf
1997-12-05 9:06 ` Alan Cox
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox