I just got Rhapsody 1.0 booting!

Started by neozeed, April 22, 2017, 02:58:59 am

Previous topic - Next topic

neozeed

Issues with the driver kit & PCI persist.

BUT



booting up


at the desktop


shutting down.

---- UPDATE ----

PCI issue has been resolved!

# include <wittycomment.h>

eagle

Wow, that's so cool!  Can you post a howto?
My NeXTs:
NeXT Computer prototype (68030-25 x2, 68040-25)
Two NeXTstations (68040-25)
All mono

neozeed

Quote from: "eagle"Wow, that's so cool!  Can you post a howto?


Sure!

First you need THIS ISO IMAGE.

I'm assuming you have Rhapsody DR2 installed somewhere.  Of course make a backup, lol etc etc, don't blame me if this doesn't work.

On your DR2 system, boot it up in single user mode (-s) at the boot prompt.

We are going to mount the CD-ROM image, and restore the new kernel and compiled userland on top of your existing system.  If this goes wrong it will hopelessly brick your machine.  Do this in a VM first to get the feel of it.

fsck -y /dev/hd0a
mount -w /
mkdir /tmp/CD
mount_cd9660 /dev/sd0a /tmp/CD
gzip -dc /tmp/CD/darwin01.gz | restore -rf -
reboot


That's really all there is to it.  I've gone through the hell of compiling the kernel, and all the command tools from Darwin 0.1, and just verified a bunch of times with Qemu 0.90 that I can install DR2, then restore this CD on top of it, and unlock 512MB of RAM ...  which I think is the only real noticeable change right now.
# include <wittycomment.h>

eagle

Wow, that's awesome - thanks! Almost all of my NeXTs are VMs (my 3 real ones are in storage), and I'll give this a try sometime this weekend.  Cool!
My NeXTs:
NeXT Computer prototype (68030-25 x2, 68040-25)
Two NeXTstations (68040-25)
All mono

neozeed

Quote from: "eagle"Wow, that's awesome - thanks! Almost all of my NeXTs are VMs (my 3 real ones are in storage), and I'll give this a try sometime this weekend.  Cool!


Sure thing!

I need to follow NCommander's adventure with putting the OPENSTEP kernel on top of NeXTSTEP, to see if I can put this kernel onto NeXTSTEP .... If that works it'll be amazing!

Luckily mine are virtual too, so it's really alll zero risk, it just takes up time.

I should try this on VMWare as well...
# include <wittycomment.h>

Daxziz

This is really exciting news :)

My personal opinion is that the unlock of 512 MB of RAM is more than enough to call this a great success :)

Thanks for all the work you've put into this.


-Dax

neozeed

Quote from: "Daxziz"This is really exciting news :)

My personal opinion is that the unlock of 512 MB of RAM is more than enough to call this a great success :)

Thanks for all the work you've put into this.


-Dax


Yeah it's great right now, lots of room for improvement, which we as a community (the royal we) can make.

I got the EIDE driver to build!  It fails on new hardware, but it works on old.  It's a stepping stone.
# include <wittycomment.h>

Daxziz

Quote from: "neozeed"I got the EIDE driver to build!  It fails on new hardware, but it works on old.  It's a stepping stone.


Meaning more specifically? Old hardware as in date or old hardware as in already installed with original driver and "updated" to new?


-Dax

neozeed

Quote from: "Daxziz"
Quote from: "neozeed"I got the EIDE driver to build!  It fails on new hardware, but it works on old.  It's a stepping stone.


Meaning more specifically? Old hardware as in date or old hardware as in already installed with original driver and "updated" to new?


-Dax


well old hardware like fake stuff qemu 0.90 to be exact.  It' detects and runs fine.  But this Windows98 era PC I bought used, along with any modern qemu both have errors right after the IDE probe.

Take this example:



In this case I changed the driver to return only single sector read/write in the detect, and yet it still bombs out when it tries to read the disk.  Since I'm already installed the next things to do is to gut features in the driver, or dump out what it detects as in this case it's sectCnt (I fixed the typo to at least match the variable name) is 0xfe which is just WAY WAY too high, so obviously some value is being detected far too high on what the drive is actually capable of.
# include <wittycomment.h>