Compression

[Neale Ferguson]

I gave the instructions at 
https://linux-on-z.blogspot.com/2019/10/howto-exploiting-hardware-compression.html
 to exercise the hardware compression and it certainly gives things a boost:

Without – 

$ time python test.py
7.23user 0.33system 0:07.57elapsed 99%CPU (0avgtext+0avgdata 
1278228maxresident)k

With (the new .so was in my cwd) –
 
$ LD_LIBRARY_PATH=$PWD time python test.py
1.49user 0.27system 0:01.77elapsed 99%CPU (0avgtext+0avgdata 
1282176maxresident)k

One issue I did have during the build was that __STDC_VERSION__ wasn’t defined 
so I had to do it manually via -D

Neale Ferguson


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Rob van der Heij]

On Fri, 20 Dec 2019 at 02:49, Neale Ferguson  wrote:

> Here are some birthday-inspired nostalgia:
>

And we were cross-compiling things like glibc, which means using the
compiler and tool chain on x86 to build the executable code for s390. You
tell the tools to put those s390 executable parts in a separate directory
from where you package it in the initial ram disk. Except when you make a
typo late at night and overwrite your Linux x86 glibc with the s390 code
that does not run there. One by one things were falling over and I had to
re-install Linux on the PC next morning...

Rob

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Neale Ferguson]

Here are some birthday-inspired nostalgia:

1. Being part of the bigfoot project (to which TrothR alluded) where we got the 
kernel to boot and getting it to launch a shell by 1998. Then going to 
Poughkeepsie to write a Redbook on OpenEdition where I met Boas Betzler (sp?) 
who was in town to pitch to the software council about their skunkworks 
project. I managed to get a private briefing and was blown away by their 
progress. One of the group (no names Erich) expressed some hope I may be able 
to take a tape away with me but I've never underestimated IBM legal and thought 
we'd be luck to see anything within 18-24 months. I went back to Australia and 
continued with Bigfoot just so we could publish progress and hope to support 
the Skunkworks folks get it past the objections. (I was privy to some email 
exchanges where some were calling for the excision and display of certain 
anatomical features of the developers.) So I was very pleasantly proved wrong 
when the announcement was made less than 8 months later on 18 December. 
Appropriately, I was at a dinner at the home of Ira and Pam Bland and a host of 
other VM old hands. 

2. One of my first contributions was the result of trying to run the code on a 
9672-G2 at Sterling. The kernel used the Compare-swap-and-purge instruction 
which was G5 and above. I wrote some code that did a SIGP to get the other CPUs 
to achieve the same effect.

3. I wanted to understand how device drivers worked so wrote the CPINT device 
driver to talk to the hypervisor.. It seemed like a good place to start. It was 
made obsolete by vmcp. I later adapted the signal shutdown code that Romney had 
written for CMS and put it in the kernel. I think this code was also cleaned, 
dryed, and ironed.

4. Userland code:
- Adding s390 support to config.sub and config.guess so that packages would 
build with autoconf 
- Getting postgersql running around February or March by providing the atomic 
locking code (prior to gcc supporting it as a builtin function) and providing 
similar code for OpenAFS that helped Sine Nomine on its way
- Wrote the s390x interpreter for OpenJDK before it was completely rewritten 
when Oracle implemented HotSpot. Thanks to SAS and AdoptJDK/IBM it's now 
available in JIT form. 

Neale Ferguson


 


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Paul Flint]

Greetings Rick Truth,

I will need to chime in here, because this day, December 18, is also my 
65th birthday.  As some of you may be aware, despite attempting to play 
guitar, my only claim to fame is as follows:


I was the first person ever to boot zLinux under IBM VM on an IBM 9672 R24 
Mainframe in the basement of the Pentagon.


This happened prior to 9 September 2001.

I used a custom kernel developed by Karl Shultze-Stracle (spelling?) and 
cut-to-fit by Romney White who I met out at the old IBM Center off 270 in 
Gaithersberg, and may not remember this, as he has done so much fantastic 
stuff.


It took 45 Minutes to boot, and I distinctly recall when the 3279 console 
emulator at the time went from EBCDIC to ASCII.  From there we telnet'ed 
in.


I suppose the dates matter, so I may need to ascertain specifically when 
before 9/11 we actually pulled this off...


Thanks and...

Kindest Regards,

Paul Flint

On Thu, 19 Dec 2019, Rick Troth wrote:


Date: Thu, 19 Dec 2019 10:59:27 -0500
From: Rick Troth 
Reply-To: Linux on 390 Port 
To: LINUX-390@VM.MARIST.EDU
Subject: Re: Happy birthday

The first native run of S/390 Linux outside of IBM was at BMC in
Houston. (And I do mean NATIVE, not LPAR.)
Mike Martin and I were tapped because we both knew S/390 and we both
knew Linux.

The big day ... er, uh ... the big night came when we got native time on
the 600S. We had already IPLed the Marist distro on VM. We had no doubt
it would "just work". So we pointed the HMC at that disk and clicked go.

It crashed.

We quickly learned to *not* include every device in the IO gen.
Jan Ott was with Mike and me and has a knack for disassembling in his
head. He shot the dump right then and there. Jan is the kind of
character that makes me think of Lynn Wheeler, to give y'all an idea. I
remember him mumbling, "there's that Intel instruction", when seeing the
then-new relative and immediate op codes. Jan recognized that we had
blown the device table. Smoking gun. And easy enough to fix w/o
re-building the kernel.

We re-genned, re-IPLed, and it ran!
Total nerd that I am, I just started compiling stuff. Mike was more
practical, giving the system a real workload: he built and ran DOOM.
(Yes, *that* DOOM.) Nearest graphical display was a Sun workstation via
X windows. Worked great! So there you have it, the first native use of
S/390 Linux outside of IBM was DOOM.

I was thankful to be working with a system a little closer to home. Off
and on ran Linux on family laptops. (Not S/390, but more practical for
them than CMS, and wwwaaayyy more practical than TSO.) I aggressively
treated all Linux the same, brought everything I could from PC Linux up
to S/390. Took concepts from Linux on VM down to Linux on Xen and then
to Linux on KVM. (And yes you *can* share virtual disks with Xen, KVM,
even VMware.) But as Melinda has well written, such things can't be
mandated.

Kinda needed this today. Thanks GNeale for making mention of it. Nice to
reflect ... I for one have had a fun career, in large part due to Linux
on S/390 and Z.

-- R; <><


On 12/19/19 8:26 AM, Cornelia Huck wrote:

On Thu, 19 Dec 2019 13:06:24 +
Neale Ferguson  wrote:


And thus was born cio_ignore?

Well, that was my second kernel patch :) (And the oldest one still
existing in remnants -- /proc/subchannels only lived during the 2.4 era
-- although the current cio_ignore implementation is a far cry from the
original one.)

The main problem cio_ignore was actually trying to solve were LPAR
definitions including every device, their friends, and some random ones
picked up on the street -- including devices shared with other LPARs.
You don't want all of these, and especially you don't want to write to
other people's disks by accident.



 Original message 
From: Ingo Adlung 
Date: 12/19/19 23:55 (GMT+10:00)
To: LINUX-390@VM.MARIST.EDU
Subject: Re: [LINUX-390] Happy birthday

And I remember the legendary Install Fest parties we had with clients in
groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
that time and was overwhelmed when people booted (IPLed) Linux into OS/390
partitions with 10s of thousands of I/O devices defined/attached. In order
for not having to allocate so much static memory I first limited the number
of addressable devices to 1024, following the VSE/ESA model, and only when
clients wanted more (many did start with Linux in LPAR and only partitions
configured for OS/390 at hand) I back-ported dynamic boot memory allocation
back from the 2.3 kernel to 2.2, still avoiding large memory allocations
when all a client wanted to operate where just a handful of devices ...

Ha, I thought 1024 devices would be enough for anybody :)

Funnily enough, I ran into limitations of the procfs when I tried to
create a /proc entry for every device discovered in 2.4. We only got
per-device entries with the advent of sysfs (and the driver model) in
2.5.

--
For L

Re: Happy birthday

[Rick Troth]

The first native run of S/390 Linux outside of IBM was at BMC in
Houston. (And I do mean NATIVE, not LPAR.)
Mike Martin and I were tapped because we both knew S/390 and we both
knew Linux.

The big day ... er, uh ... the big night came when we got native time on
the 600S. We had already IPLed the Marist distro on VM. We had no doubt
it would "just work". So we pointed the HMC at that disk and clicked go.

It crashed.

We quickly learned to *not* include every device in the IO gen.
Jan Ott was with Mike and me and has a knack for disassembling in his
head. He shot the dump right then and there. Jan is the kind of
character that makes me think of Lynn Wheeler, to give y'all an idea. I
remember him mumbling, "there's that Intel instruction", when seeing the
then-new relative and immediate op codes. Jan recognized that we had
blown the device table. Smoking gun. And easy enough to fix w/o
re-building the kernel.

We re-genned, re-IPLed, and it ran!
Total nerd that I am, I just started compiling stuff. Mike was more
practical, giving the system a real workload: he built and ran DOOM.
(Yes, *that* DOOM.) Nearest graphical display was a Sun workstation via
X windows. Worked great! So there you have it, the first native use of
S/390 Linux outside of IBM was DOOM.

I was thankful to be working with a system a little closer to home. Off
and on ran Linux on family laptops. (Not S/390, but more practical for
them than CMS, and wwwaaayyy more practical than TSO.) I aggressively
treated all Linux the same, brought everything I could from PC Linux up
to S/390. Took concepts from Linux on VM down to Linux on Xen and then
to Linux on KVM. (And yes you *can* share virtual disks with Xen, KVM,
even VMware.) But as Melinda has well written, such things can't be
mandated.

Kinda needed this today. Thanks GNeale for making mention of it. Nice to
reflect ... I for one have had a fun career, in large part due to Linux
on S/390 and Z.

-- R; <><


On 12/19/19 8:26 AM, Cornelia Huck wrote:
> On Thu, 19 Dec 2019 13:06:24 +
> Neale Ferguson  wrote:
>
>> And thus was born cio_ignore?
> Well, that was my second kernel patch :) (And the oldest one still
> existing in remnants -- /proc/subchannels only lived during the 2.4 era
> -- although the current cio_ignore implementation is a far cry from the
> original one.)
>
> The main problem cio_ignore was actually trying to solve were LPAR
> definitions including every device, their friends, and some random ones
> picked up on the street -- including devices shared with other LPARs.
> You don't want all of these, and especially you don't want to write to
> other people's disks by accident.
>
>>
>>  Original message 
>> From: Ingo Adlung 
>> Date: 12/19/19 23:55 (GMT+10:00)
>> To: LINUX-390@VM.MARIST.EDU
>> Subject: Re: [LINUX-390] Happy birthday
>>
>> And I remember the legendary Install Fest parties we had with clients in
>> groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
>> that time and was overwhelmed when people booted (IPLed) Linux into OS/390
>> partitions with 10s of thousands of I/O devices defined/attached. In order
>> for not having to allocate so much static memory I first limited the number
>> of addressable devices to 1024, following the VSE/ESA model, and only when
>> clients wanted more (many did start with Linux in LPAR and only partitions
>> configured for OS/390 at hand) I back-ported dynamic boot memory allocation
>> back from the 2.3 kernel to 2.2, still avoiding large memory allocations
>> when all a client wanted to operate where just a handful of devices ...
> Ha, I thought 1024 devices would be enough for anybody :)
>
> Funnily enough, I ran into limitations of the procfs when I tried to
> create a /proc entry for every device discovered in 2.4. We only got
> per-device entries with the advent of sysfs (and the driver model) in
> 2.5.
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Richard Lynch]

Certainly, and I was one of them.  I was totally amazed that Linux was 
booting and working on our IBM 9672-r63.  If I remember right I had ours 
up and running on Christmas Eve.


Richard Lynch
WVNET Systems Programming Manager (retired)



On 12/18/19 9:34 PM, Rich Smrcina via vm.marist.edu [*] wrote:

I’m sure there were more than a few of us installing the ‘Marist’ distribution 
on our mainframes over Christmas.

I know I was working on the install at Grede Foundries.

Rich Smrcina
Sr. Systems Engineer

Velocity Software Inc.
Main: (650) 964-8867
Main: (877) 964-8867
r...@velocitysoftware.com 



On Dec 18, 2019, at 8:25 PM, Neale Ferguson  wrote:

https://tech-insider.org/linux/research/1999/1218.html

"Dec 18, 1999

According to Alan Cox's 2.2.14pre14 patch, Linux now has at least preliminary 
support for running on the IBM S/390 Mainframe. Rumors that IBM has been engaged in 
a port have been circulating for some time, but this is the first concrete code that 
has surfaced to date..."

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390



--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Martha McConaghy]

Reading all these emails, especially this one, brings back a lot of 
memories.   Especially, how we were rushing to get the web site up to 
make the "Marist distro" available before Y2K brought the whole world 
downdoh!


Martha

On 12/19/2019 8:20 AM, Rob van der Heij wrote:

On Thu, 19 Dec 2019 at 03:34, Rich Smrcina 
wrote:
  

I’m sure there were more than a few of us installing the ‘Marist’
distribution on our mainframes over Christmas.

  
For me the journey started a bit earlier, working on Melinda's system in

Princeton. I told my wife this was significant for z/VM and I would be
spending more spare time on it, for a few months. As all projects it took
longer. A few months later, at the office wearing my Marist sweat shirt,
someone pointed saying "Hey, that's a Mainframe Linux sweat shirt!"
  
Rob
  
--

For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
Martha McConaghy
Marist: System Architect/Technical Lead
SHARE: Vice President
Marist College IT
Poughkeepsie, NY 12601

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Cornelia Huck]

On Thu, 19 Dec 2019 13:06:24 +
Neale Ferguson  wrote:

> And thus was born cio_ignore?

Well, that was my second kernel patch :) (And the oldest one still
existing in remnants -- /proc/subchannels only lived during the 2.4 era
-- although the current cio_ignore implementation is a far cry from the
original one.)

The main problem cio_ignore was actually trying to solve were LPAR
definitions including every device, their friends, and some random ones
picked up on the street -- including devices shared with other LPARs.
You don't want all of these, and especially you don't want to write to
other people's disks by accident.

>
>
>  Original message 
> From: Ingo Adlung 
> Date: 12/19/19 23:55 (GMT+10:00)
> To: LINUX-390@VM.MARIST.EDU
> Subject: Re: [LINUX-390] Happy birthday
>
> And I remember the legendary Install Fest parties we had with clients in
> groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
> that time and was overwhelmed when people booted (IPLed) Linux into OS/390
> partitions with 10s of thousands of I/O devices defined/attached. In order
> for not having to allocate so much static memory I first limited the number
> of addressable devices to 1024, following the VSE/ESA model, and only when
> clients wanted more (many did start with Linux in LPAR and only partitions
> configured for OS/390 at hand) I back-ported dynamic boot memory allocation
> back from the 2.3 kernel to 2.2, still avoiding large memory allocations
> when all a client wanted to operate where just a handful of devices ...

Ha, I thought 1024 devices would be enough for anybody :)

Funnily enough, I ran into limitations of the procfs when I tried to
create a /proc entry for every device discovered in 2.4. We only got
per-device entries with the advent of sysfs (and the driver model) in
2.5.

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Ingo Adlung]

Linux on 390 Port  wrote on 19/12/2019 14:06:24:

> From: Neale Ferguson 
> To: LINUX-390@VM.MARIST.EDU
> Date: 19/12/2019 14:07
> Subject: [EXTERNAL] Re: [LINUX-390] Happy birthday
> Sent by: Linux on 390 Port 
>
> And thus was born cio_ignore?
>

Yes, that was the quick remedy :-) But in the end wanted to allow for as
many as the clients actually needed and/or desired ...

>
>  Original message 
> From: Ingo Adlung 
> Date: 12/19/19 23:55 (GMT+10:00)
> To: LINUX-390@VM.MARIST.EDU
> Subject: Re: [LINUX-390] Happy birthday
>
> And I remember the legendary Install Fest parties we had with clients in
> groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
> that time and was overwhelmed when people booted (IPLed) Linux into
OS/390
> partitions with 10s of thousands of I/O devices defined/attached. In
order
> for not having to allocate so much static memory I first limited the
number
> of addressable devices to 1024, following the VSE/ESA model, and only
when
> clients wanted more (many did start with Linux in LPAR and only
partitions
> configured for OS/390 at hand) I back-ported dynamic boot memory
allocation
> back from the 2.3 kernel to 2.2, still avoiding large memory allocations
> when all a client wanted to operate where just a handful of devices ...
>
> Mit freundlichem Gruß / Best regards
> Ingo Adlung
>
>
>
>Ingo AdlungIBM Deutschland Research &
>IBM Distinguished Engineer Development GmbH
>Chief Architect, and CTO   Vorsitzender des Aufsichtsrats:
>IBM Z and LinuxONE Virtualization  Matthias Hartmann
>& LinuxGeschäftsführung: Dirk Wittkopp
>mail: adl...@de.ibm.comSitz der Gesellschaft: Böblingen
>phone: +49-7031-16-4263Registergericht: Amtsgericht
>   Stuttgart, HRB 243294
>
>
>
>
>
>
>
>
>
> From:   Mike Riggs 
> To: LINUX-390@VM.MARIST.EDU
> Date:   19/12/2019 13:34
> Subject:[EXTERNAL] Re: [LINUX-390] Happy birthday
> Sent by:Linux on 390 Port 
>
>
>
> And thus started the new future of judicial technology here. Not too long
> after, a distro was running (well, sorta walking) on a 9672. Been a blast
> ever since.
>
> Mike Riggs
> OES/SCV
>
> -Original Message-
> From: Linux on 390 Port  On Behalf Of Rich
Smrcina
> Sent: Wednesday, December 18, 2019 9:34 PM
> To: LINUX-390@VM.MARIST.EDU
> Subject: Re: Happy birthday
>
> I’m sure there were more than a few of us installing the ‘Marist’
> distribution on our mainframes over Christmas.
>
> I know I was working on the install at Grede Foundries.
>
> Rich Smrcina
> Sr. Systems Engineer
>
> Velocity Software Inc.
> Main: (650) 964-8867
> Main: (877) 964-8867
> r...@velocitysoftware.com 
>
>
> > On Dec 18, 2019, at 8:25 PM, Neale Ferguson 
wrote:
> >
> >
> https://urldefense.proofpoint.com/v2/url?
>
u=https-3A__tech-2Dinsider.org_linux_research_1999_1218.html&d=DwIGaQ&c=jf_iaSHvJObTbx-

> siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-
> y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-
> CZzwY375fuXlEuAgDrNo&s=QoPSQ9tkxyy7_dafKnVSmtqWoHB6nZaezrAdL2ji_MY&e=
>
> >
> > "Dec 18, 1999
> >
> > According to Alan Cox's 2.2.14pre14 patch, Linux now has at least
> preliminary support for running on the IBM S/390 Mainframe. Rumors that
IBM
> has been engaged in a port have been circulating for some time, but this
is
> the first concrete code that has surfaced to date..."
> >
> > --
> > For LINUX-390 subscribe / signoff / archive access instructions, send
> > email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> > visit
> >
> https://urldefense.proofpoint.com/v2/url?
>
u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-

> siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-
> y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-
> CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=
>
>
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions, send
email
> to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
> https://urldefense.proofpoint.com/v2/url?
>
u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-

> siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-
> y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-
> CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=
>
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions,
> send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
> https://urldefense.proofpoint.com/v2/url?
>
u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-

> siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-
> y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-
> CZz

Re: Happy birthday

[Rob van der Heij]

On Thu, 19 Dec 2019 at 03:34, Rich Smrcina 
wrote:

> I’m sure there were more than a few of us installing the ‘Marist’
> distribution on our mainframes over Christmas.
>

For me the journey started a bit earlier, working on Melinda's system in
Princeton. I told my wife this was significant for z/VM and I would be
spending more spare time on it, for a few months. As all projects it took
longer. A few months later, at the office wearing my Marist sweat shirt,
someone pointed saying "Hey, that's a Mainframe Linux sweat shirt!"

Rob

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Neale Ferguson]

And thus was born cio_ignore?


 Original message 
From: Ingo Adlung 
Date: 12/19/19 23:55 (GMT+10:00)
To: LINUX-390@VM.MARIST.EDU
Subject: Re: [LINUX-390] Happy birthday

And I remember the legendary Install Fest parties we had with clients in
groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
that time and was overwhelmed when people booted (IPLed) Linux into OS/390
partitions with 10s of thousands of I/O devices defined/attached. In order
for not having to allocate so much static memory I first limited the number
of addressable devices to 1024, following the VSE/ESA model, and only when
clients wanted more (many did start with Linux in LPAR and only partitions
configured for OS/390 at hand) I back-ported dynamic boot memory allocation
back from the 2.3 kernel to 2.2, still avoiding large memory allocations
when all a client wanted to operate where just a handful of devices ...

Mit freundlichem Gruß / Best regards
Ingo Adlung



   Ingo AdlungIBM Deutschland Research &
   IBM Distinguished Engineer Development GmbH
   Chief Architect, and CTO   Vorsitzender des Aufsichtsrats:
   IBM Z and LinuxONE Virtualization  Matthias Hartmann
   & LinuxGeschäftsführung: Dirk Wittkopp
   mail: adl...@de.ibm.comSitz der Gesellschaft: Böblingen
   phone: +49-7031-16-4263Registergericht: Amtsgericht
  Stuttgart, HRB 243294









From:   Mike Riggs 
To: LINUX-390@VM.MARIST.EDU
Date:   19/12/2019 13:34
Subject:[EXTERNAL] Re: [LINUX-390] Happy birthday
Sent by:Linux on 390 Port 



And thus started the new future of judicial technology here. Not too long
after, a distro was running (well, sorta walking) on a 9672. Been a blast
ever since.

Mike Riggs
OES/SCV

-Original Message-
From: Linux on 390 Port  On Behalf Of Rich Smrcina
Sent: Wednesday, December 18, 2019 9:34 PM
To: LINUX-390@VM.MARIST.EDU
Subject: Re: Happy birthday

I’m sure there were more than a few of us installing the ‘Marist’
distribution on our mainframes over Christmas.

I know I was working on the install at Grede Foundries.

Rich Smrcina
Sr. Systems Engineer

Velocity Software Inc.
Main: (650) 964-8867
Main: (877) 964-8867
r...@velocitysoftware.com 


> On Dec 18, 2019, at 8:25 PM, Neale Ferguson  wrote:
>
>
https://urldefense.proofpoint.com/v2/url?u=https-3A__tech-2Dinsider.org_linux_research_1999_1218.html&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=QoPSQ9tkxyy7_dafKnVSmtqWoHB6nZaezrAdL2ji_MY&e=

>
> "Dec 18, 1999
>
> According to Alan Cox's 2.2.14pre14 patch, Linux now has at least
preliminary support for running on the IBM S/390 Mainframe. Rumors that IBM
has been engaged in a port have been circulating for some time, but this is
the first concrete code that has surfaced to date..."
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions, send
> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
>
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=



--
For LINUX-390 subscribe / signoff / archive access instructions, send email
to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=





--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvin

Re: Happy birthday

[Ingo Adlung]

And I remember the legendary Install Fest parties we had with clients in
groups of 10-15 and eventually doing 1:1 calls. I wrote the I/O layer at
that time and was overwhelmed when people booted (IPLed) Linux into OS/390
partitions with 10s of thousands of I/O devices defined/attached. In order
for not having to allocate so much static memory I first limited the number
of addressable devices to 1024, following the VSE/ESA model, and only when
clients wanted more (many did start with Linux in LPAR and only partitions
configured for OS/390 at hand) I back-ported dynamic boot memory allocation
back from the 2.3 kernel to 2.2, still avoiding large memory allocations
when all a client wanted to operate where just a handful of devices ...

Mit freundlichem Gruß / Best regards
Ingo Adlung


   
   Ingo AdlungIBM Deutschland Research &   
   IBM Distinguished Engineer Development GmbH 
   Chief Architect, and CTO   Vorsitzender des Aufsichtsrats:  
   IBM Z and LinuxONE Virtualization  Matthias Hartmann
   & LinuxGeschäftsführung: Dirk Wittkopp  
   mail: adl...@de.ibm.comSitz der Gesellschaft: Böblingen 
   phone: +49-7031-16-4263Registergericht: Amtsgericht 
  Stuttgart, HRB 243294
   








From:   Mike Riggs 
To: LINUX-390@VM.MARIST.EDU
Date:   19/12/2019 13:34
Subject:[EXTERNAL] Re: [LINUX-390] Happy birthday
Sent by:Linux on 390 Port 



And thus started the new future of judicial technology here. Not too long
after, a distro was running (well, sorta walking) on a 9672. Been a blast
ever since.

Mike Riggs
OES/SCV

-Original Message-
From: Linux on 390 Port  On Behalf Of Rich Smrcina
Sent: Wednesday, December 18, 2019 9:34 PM
To: LINUX-390@VM.MARIST.EDU
Subject: Re: Happy birthday

I’m sure there were more than a few of us installing the ‘Marist’
distribution on our mainframes over Christmas.

I know I was working on the install at Grede Foundries.

Rich Smrcina
Sr. Systems Engineer

Velocity Software Inc.
Main: (650) 964-8867
Main: (877) 964-8867
r...@velocitysoftware.com 


> On Dec 18, 2019, at 8:25 PM, Neale Ferguson  wrote:
>
>
https://urldefense.proofpoint.com/v2/url?u=https-3A__tech-2Dinsider.org_linux_research_1999_1218.html&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=QoPSQ9tkxyy7_dafKnVSmtqWoHB6nZaezrAdL2ji_MY&e=

>
> "Dec 18, 1999
>
> According to Alan Cox's 2.2.14pre14 patch, Linux now has at least
preliminary support for running on the IBM S/390 Mainframe. Rumors that IBM
has been engaged in a port have been circulating for some time, but this is
the first concrete code that has surfaced to date..."
>
> --
> For LINUX-390 subscribe / signoff / archive access instructions, send
> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
> visit
>
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=



--
For LINUX-390 subscribe / signoff / archive access instructions, send email
to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=


--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
https://urldefense.proofpoint.com/v2/url?u=http-3A__www2.marist.edu_htbin_wlvindex-3FLINUX-2D390&d=DwIGaQ&c=jf_iaSHvJObTbx-siA1ZOg&r=jQ4IiHbzZ0l-wFKuUHMHvPIsi5vD8MZZCyI-y49pWL0&m=6Vdtn9czQLmEFCLh41YmBe-CZzwY375fuXlEuAgDrNo&s=SB0o_Utp2GmkAmFiGILQpRbYiW1Vmnz9Jh2eAtxvc80&e=





--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

Re: Happy birthday

[Mike Riggs]

And thus started the new future of judicial technology here. Not too long 
after, a distro was running (well, sorta walking) on a 9672. Been a blast ever 
since. 

Mike Riggs
OES/SCV

-Original Message-
From: Linux on 390 Port  On Behalf Of Rich Smrcina
Sent: Wednesday, December 18, 2019 9:34 PM
To: LINUX-390@VM.MARIST.EDU
Subject: Re: Happy birthday

I’m sure there were more than a few of us installing the ‘Marist’ distribution 
on our mainframes over Christmas.

I know I was working on the install at Grede Foundries.

Rich Smrcina
Sr. Systems Engineer

Velocity Software Inc.
Main: (650) 964-8867
Main: (877) 964-8867
r...@velocitysoftware.com 


> On Dec 18, 2019, at 8:25 PM, Neale Ferguson  wrote:
> 
> https://tech-insider.org/linux/research/1999/1218.html
> 
> "Dec 18, 1999
> 
> According to Alan Cox's 2.2.14pre14 patch, Linux now has at least preliminary 
> support for running on the IBM S/390 Mainframe. Rumors that IBM has been 
> engaged in a port have been circulating for some time, but this is the first 
> concrete code that has surfaced to date..."
> 
> --
> For LINUX-390 subscribe / signoff / archive access instructions, send 
> email to lists...@vm.marist.edu with the message: INFO LINUX-390 or 
> visit
> http://www2.marist.edu/htbin/wlvindex?LINUX-390


--
For LINUX-390 subscribe / signoff / archive access instructions, send email to 
lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390

--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www2.marist.edu/htbin/wlvindex?LINUX-390