Thanks Marcela.
It's good to know that you and your team are keeping an eye on this.

For the rest, I was able to man-handle the Debian installer and finally have a running system. IT DID NOT render a bootable system, but after two or three restarts I was able to snag the kernel and INITRD from /boot and line those up in the Hercules reader with a hand-crafted parm file. Works.

Related: I have tried IPLing from FBA with mixed results.
At one time, I could IPL from unpartitioned 9336. That is, the EXT3 filesystem started at block zero of the disk (as Linux saw it), like /dev/dasda versus /dev/dasda1. The 'zipl' bootstrap code broke w/r/t CKD (in the unpartitioned context) a long time ago. I don't recall the details, but I think the boostrap simply went too far into the volume and clobbered the first part of the filesystem. Definitely NOT getting a clobbered FS with the latest unpartitioned FBA, but it won't IPL. I'll make a new thread if I learn more.

IPL from FBA is really handy because ISO (CD, DVD) is FBA.
Would be nice to have (e.g.) ...

        0CD0 9336 /the/installer/image.iso

 ... and then 'ipl cd0' in Hercules. So much simpler incantation than the current recipes.

For what it's worth, ISO 9660 FS has more room at the front than EXT2,3,4.
Interestingly, ISO 9660 is less often found in a partition. (Okay, yeah, on flash drives. [sigh])


-- R; <><




On 8/29/24 4:39 AM, Marcela Mašláňová wrote:
Hi,
I can't help with emulators, but I believe that Tumbleweed is at least
booting based on QA testing.
Tumbleweed is built even for older mainframes, so that shouldn't be an
issue. There is also a new installer
since July, which is not feature complete yet. I'm currently looking with
developers on storage management,
but it shouldn't lead to errors like this.
If you believe it's a kernel bug, please file a report with reproducer.
Thanks,
Marcela Maslanova

On Tue, Aug 27, 2024 at 1:52 AM Rick Troth<[email protected]>  wrote:

  > I think OpenSUSE Tumbleweed s390x does not run on Hercules due to
missing instructions.

Sadly, that might be correct.
After building Hyperion 4.7, I still cannot get either of them past
launching /init. Final trace at the end of this note.

I'll have to try a different distro. Been using SUSE variants as primary
for a long time.
Am skeptical of Red Hat because they're owned by IBM and IBM *does* push
for the latest hardware capability. (see end)
So prolly Debian. Ugh ... major learning curve there for me.

Here's the latest in case it says anything to anyone who can give me a
clue.
There's a bit of disassembly in there (nice!), and I grok assembler, but
I don't know quite how to interpret this.


     173.665796! T118! zram: Added device: zram0
     178.822340!   T117! zram0: detected capacity change from 0 to 2097152
     183.418660!     T1! EXT4-fs (zram0): mounting ext2 file system using
the ext4 subsystem
     183.437710!     T1! EXT4-fs (zram0): mounted filesystem
ff1ce7d0-6403-11ef-86c5-6f25d18d65fa r/w without journal. Quota mode: none.
     206.345791!     T1! ------------  cut here !------------
*   206.345992!     T1! kernel BUG at ../fs/dcache.c:2032 *
     206.346231!     T1! monitor event: 0040 ilc:2  #1! SMP
     206.346579!     T1! Modules linked in: ext4 crc16 mbcache jbd2 zram
     206.347159!     T1! Supported: Yes
     206.347354!     T1! CPU: 0 PID: 1 Comm: init Not tainted
6.4.0-150600.21-default #1 SLE15-SP6
2a55df39efb37a2b9dc9aa44909754d71b3a7867
     206.347707!     T1! Hardware name: HRC 2064 EMULATOR EMULATOR (LPAR)
     206.347907!     T1! Krnl PSW : 0704e00180000000 0000000023d126ae
(d_instantiate+0x26/0x28)
     206.348366!     T1!            R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0
AS:3 CC:2 PM:0 RI:0 EA:3
     206.348729!     T1! Krnl GPRS: 000000008498ddc8 0000000000000004
000000008040fd80 0000000085531b80
     206.348993!     T1!            0000000085531538 0000000066cd118c
0000037fffb65040 0000000000000000
     206.349272!     T1!            0000000000004000 0000000085531b80
000000008040fd80 000000008547d240
     206.349522!     T1!            000003ffb892cf98 0000037fffb13bff
00000000239add4a 0000037fffb13ad0
     206.350000!     T1! Krnl Code: 0000000023d1269e: c0f4ffffffb9
brcl 15,0000000023d12610
     206.350000!     T1!            0000000023d126a4: c0f40039c66a
brcl 15,000000002444b378
     206.350000!     T1!           #0000000023d126aa: af000000
mc 0,0
     206.350000!     T1!           >0000000023d126ae: 0707
bcr 0,%r7
     206.350000!     T1!            0000000023d126b0: c0040035d0ac
brcl 0,00000000243cc808
     206.350000!     T1!            0000000023d126b6: ebaff0680024
stmg %r10,%r15,104(%r15)
     206.350000!     T1!            0000000023d126bc: b90400ef
lgr %r14,%r15
     206.350000!     T1!            0000000023d126c0: e3f0ffc8ff71
lay %r15,-56(%r15)
     206.352274!     T1! Call Trace:
     206.352953!     T1!  126ae>! d_instantiate+0x26/0x28
     206.353329!     T1! (dcee>! hypfs_create_file+0x86/0x178)
     206.353635!     T1!  e51a>! hypfs_mkdir+0x2a/0x58
     206.353928!     T1!  ec6a>! hypfs_diag_create_files+0x1b2/0x480
     206.354242!     T1!  e45c>! hypfs_fill_super+0x94/0x128
     206.354538!     T1!  f6582>! vfs_get_super+0x5a/0x128
     206.354858!     T1!  f3f4c>! vfs_get_tree+0x3c/0x108
     206.355176!     T1!  26716>! path_mount+0x70e/0xa10
     206.355488!     T1!  26a7e>! do_mount+0x66/0x90
     206.355796!     T1!  26c6e>! __s390x_sys_mount+0xae/0x118
     206.356114!     T1! _do_syscall+0x1da/0x200
     206.356419!     T1!  f8>! system_call+0x70/0x98
     206.356737!     T1! Last Breaking-Event-Address:
     206.356900!     T1!  12694>! d_instantiate+0xc/0x28
     206.357287!     T1! Kernel panic - not syncing: Fatal exception:
panic_on_oops
HHC00809I Processor CP00: disabled wait state 0002000180000000
0000000023984418
HHC00826W Processor CP00: processor auto-stopped due to disabled wait
HHC00007I Previous message from function 'z900_process_interrupt' at
cpu.c(1875)

This is way better info than I was getting out of Hercules 3.13.
Thanks for the tip, Mark!

A request:
Distributors, _please avoid latest-and-greatest_ when building your
installers. Not everyone can afford to fork-lift a Z16 in a pinch.
Y'ALL KNOW how to build code which is broadly runnable. Just do so:
install first, optimize later.





On 8/26/24 5:06 PM, Paul Dembry wrote:
I think OpenSUSE Tumbleweed s390x does not run on Hercules due to
missing instructions.
-----Original Message-----
From: Linux on 390 Port<[email protected]>   On Behalf Of Rick
Troth
Sent: Monday, August 26, 2024 1:40 PM
To:[email protected]
Subject: Re: Clarification: OpenSUSE on Hercules

For clarity:
I am trying to run OpenSUSE Tumbleweed "s390x"
on top of Hercules
on top of OpenSUSE Tumbleweed "x86_64".


don't overthink it



On 8/26/24 4:29 PM, Paul Dembry wrote:
I think Rick is trying this

OpenSUSE running on
        ^
        |
        |
Hercules running on
        ^
        |
        |
Some os

IOW OpenSUSE S390x on top of Hercules, not Hercules on top of OpenSUSE.
IIRC the more recent versions of OpenSUSE and RedHat (I think RH 8.x)
require certain instructions that are not supported on Hercules. I goofed
around with RH 8.x and think I got it to work using qemu s390 but cannot
say for sure as it was a couple of years ago. I didn't really need anything
higher than RH 7.x so it was just an exercise.
Regards,
Paul



-----Original Message-----
From: Linux on 390 Port<[email protected]>   On Behalf Of Paul
Flint
Sent: Monday, August 26, 2024 1:19 PM
To:[email protected]
Subject: Clarification: OpenSUSE on Hercules

Greetings Truth

To clarify, you are operating SUSE Tumbleweed as the base system on
some random intel-type iron.  On top of this you want to operate
Hercules, apparently this is part of your studied plan for world
domination. (I am good with this :^)

This is where I get a bit confused.  I use Hercules to boot IBM VM 370.
This was enough of an entertainment that I folded Hercules and VM into
a Docker container and stuck the whole beta out there as "BaBy Bear VM"
(http://bbbvm.org/),

So the question I have is do you need code to run on top of Hercules on
top of Tumbleweed, or did I misunderstand?
If you are interested, this evening the Barre Open System Institute
(BOSI)  has our regular meeting at the York Library in East Barre, we shall
use meet.jit.si/bosi to connect. Come on to meet.Jit.si/bosi< 
http://meet.jit.si/bosi>   at 6PM!
Feel free to click the following link (around 6PM tonight) to join the
meeting:
https://meet.jit.si/bosi

Regards,

Paul

On Mon, Aug 26, 2024 at 2:06 PM Rick Troth<[email protected]>   wrote:

I need OpenSUSE on Hercules and have so far been unable to install
either of these:


       openSUSE-Leap-15.6-DVD-s390x-Build710.3-Media
       openSUSE-Tumbleweed-NET-s390x-Snapshot20240819-Media


Months ago, I tried other releases/snap-shots. No joy then either.

I have tried supplying the kernel, INITRD, and parm file via 00C
reader.
Didn't work. I have also tried Hercules IPL from ".ins" file.
After about 4 or 5 minutes, the run consistently ends with something
like ...


       T1! ------------ cut here !------------


     ... sometimes followed by (what appears to be) a wait state code,
sometimes not. In all cases the emulator stops with a disabled wait.

My preference is OpenSUSE Tumbleweed, but Leap works "if I gotta".
I'd like to install to un-partitioned FBA, but I've switched to CKD
in case FBA was/is causing the installer heartburn.
I've also tried to change the device addresses to match what the
samples have. Still nuttin, honey.

What am I doing wrong?





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

Reply via email to