Mac OS 9 booting on: Mac mini G4 (Detailed Posts)

[darthnVader]

ELN got USB working, we just don't have a valid download of his latest ROM.

I'm wondering if the conflict we are seeing with Sound and USB on some of these later Mac's is related to the Modem.

I think it is a USB device, but it has a connection to the sound device.

[RossDarker]

ELN got USB working, we just don't have a valid download of his latest ROM.

I'm wondering if the conflict we are seeing with Sound and USB on some of these later Mac's is related to the Modem.

I think it is a USB device, but it has a connection to the sound device.

Okay, nice. It will be good to have USB. I have managed to get a lot of extensions working, but even that is a small portion of the ones I am testing. I can use a whole install of the Unsupported G4's with ROM, System suitcase and most Extensions I have tested, but do you know why I have to delete "System Resources" to pass the address error? Every boot that file is re-created.

[MacOS Plus]

  Great news on the USB issue.  Makes me even more excited about my Mini that is currently in transit.  I'd be quite interested to hear ELN chime in here on an updated explanation of the error and of the nature of what was done to resolve it.

[ELN]

Hi guys,

I'm really sorry that last ROM gave you trouble! Uni has gone back for the year so I've been busy. Here's a repost in a couple of formats.

The technique I used to patch the necessary PowerPC code was a bit tricky. I have been working to document and harden my rather delicate ROM build system (now approaching its second birthday). Soon, I hope to have you all building and testing your own ROMs!

In broad terms, I just nopped out a function call from the USB interface driver to the NativePowerMgrLib.

But I have very little insight into this System Resources error! Can I suggest installing MacsBug on your boot drive and posting a picture of the error screen?

Best,

Elliot

[RossDarker]

My USB keyboard and mouse are working when it boots which is good with this new ROM, thanks. Now I have got an address error again, and it doesn't go away with the things I have tried last time. I installed MacsBug onto the drive and when it booted I typed StdLog and the images are attached. There is a picture of the address error before I installed MacsBug in there too. Also will note that I have tried to get back to what I had yesterday but just end up at an address error by doing exactly the same things so there must be something in the other partitions that are made when making os9 drivers, even though it was the same restore. <- restore then pram zap fixed it, but still address error with the USB rom.

[ELN]

https://github.com/elliotnunn/mac-rom
Output: BuildResults/RISC/Image/RomMondo

https://github.com/elliotnunn/powermac-rom
Input: RomMondo.bin
Output: BuildResults/PowerROM

https://github.com/elliotnunn/newworld-rom
Input: rom
Output: tbxi

Who's ready to get building?

[DieHard]

The technique I used to patch the necessary PowerPC code was a bit tricky. I have been working to document and harden my rather delicate ROM build system (now approaching its second birthday). Soon, I hope to have you all building and testing your own ROMs!

Best,

Elliot

Ahhh, teaching us how to fish instead of just handing out the fish... excellent

[darthnVader]

Has anyone tried the Apple CD/DVD driver on the Mini?

It's not working on the PowerBook.

Even disabling the extension and putting an HFS CD into the drive hangs the Finder before it starts, but that's not unexpected as an HFS CD has the Apple CD/DVD driver in it's partition map.

Only real reason it matters is it's going to be hard to make a bootable iso for people to test and install if the disc hangs the finder.

[darthnVader]

Good news, nobody sharked my bid for a Mini on Ebay, so I'll have one in about a week

I should be able to get 2D/3D working on it, I've figured out how all the ATI drivers work.

That's my part, I don't do assembly. I know some C C++, and I've made a few device drivers for OS X, but most of the stuff ELN and nanopico talk about is way over my head.

 

[Daniel]

That's my part, I don't do assembly. I know some C C++, and I've made a few device drivers for OS X, but most of the stuff ELN and nanopico talk about is way over my head.

You might be suprised how quickly you get used to it. I certainly was.

[TimothyHD]

Hello fellow Mac Mini users!  My name is Tim. I have recently acquired a 1.42 PowerPC Mac Mini, and have read up on the thread. I am getting it tomorrow, and I'm hoping to contribute to this cause! I see you've all made ground breaking work in the past few months!! Keep it up! 

[ELN]

You might be suprised how quickly you get used to it. I certainly was.

Daniel sure was quick!

Quick confession about the patch that I made to the USB class driver to get it to work without a PMU: I "saved" it in a scrollback buffer that I very quickly lost. But with some wrangling it has been recovered!

https://github.com/elliotnunn/mac-rom/commit/ac2112a

The complementary patch to NativePowerMgrLib is in the newworld-rom repo.

https://github.com/elliotnunn/newworld-rom/commit/0b73e80

[macStuff]

Good news, nobody sharked my bid for a Mini on Ebay, so I'll have one in about a week

congrats!

[MacOS Plus]

  I just snagged a 1.5GHz G4 Mini on eBay, so I'll officially be able to help with this project once it arrives!

I received my Mini today.  Mildly unfortunate - what showed up was a 1.42GHz instead of the 1.5GHz model it was listed as.  (I'll bitch at the seller tomorrow.)  On the up-side, it has 1GB of RAM.

  I went straight into testing a 9.2.2 drive that I've been using for awhile in another machine.  It was made from the "Unsupported G4s ASR", and I believe it's the one that I was using in the 9serve.  I attached it via firewire so I could leave the OS X install on the internal drive intact.  The only thing I swapped out was the Mac OS ROM file.  The disk was selected for boot using the option key method.

  Amazingly, only a single extension out of everything I normally have enabled broke the boot - "Apple Audio Extension".  After disabling that it booted properly, obviously without audio support, but it booted extremely fast.  Video detected a single resolution of 1920x1080, which is correct for this monitor, but predictably it's only offering 256-color mode.  Something is causing a Finder crash when it reaches the desktop but I have yet to determine what it is.  One weird thing is that the Apple menu won't enable the submenu for the Control Panels folder.

  There is one strange twist with the disabling of the PMU device.  For some reason the system identifies this Mini model as an iBook.  Without power management loading the system obviously complains when reaching the desktop that the software is missing.  The weird part is that after a few minutes of use a warning came up saying the system was low on power, as if it was running on a battery, and then promptly shut down because it thought it needed to protect itself.  I removed the Energy Saver control panel to prevent this from happening again, leading to the missing software warning every boot.

  Everything else seems to work.  The AppleTalk control panel even states there's a serial port available, presumably because it recognizes the modem.

  Now for the coolest part - I'm posting this from Classilla running on this Mini!  (This is getting a bit challenging on Classilla because compatibility with this forum is not as good as it used to be.)

Edit - One additional note - The device info for the Radeon 9200 in this thing is as follows:

Device name: "ATY,RockHopper2_A"
Chipset: "ATY,RV280"
Device ID: 5962
ROM: 113-xxxxx-116

[macStuff]

wonder if the device ID extension edit fix of user Bolle's would help you out there?
DeviceID: 5962 (instead of Bolle's 9200SE with the id of 5961?)

both of them are RV280 gpus

[MacOS Plus]

  No luck on the 2D acceleration yet.  I don't know if I'm doing it wrong or if something more has to be changed.

  I do have one minor thing to report on related to the Finder crash and warning message about missing system software.  (The exact text of that message was as follows - "Your computer can't sleep because some system software is missing.  Reinstall the software that came with your computer.  [Shut Down] [OK]")  Just to see what would happen, I did the "screamer" OF edit to the sound device using the commands from one of the iBooks.  The Apple Audio Extension no longer crashes the boot with this change in place.  Oddly though, the Finder crash and the missing system software warning at the desktop also went away.  The sound device itself doesn't appear in the Sound control panel though and obviously doesn't work.  Perhaps this experience will help someone sort out how to correct it properly though.

  I'm still really curious about why the AppleTalk control panel reports "Modem/Printer port" available and doesn't complain when I select it and save the setting.  System Profiler doesn't know the internal modem is present, but even after physically disconnecting the modem card from the motherboard, some other device is still recognized as this 'phantom' serial port.  I'd really like to know what's going on here and whether or not this means I could hack in a true serial port in place of the modem or via whatever other data path this detected port is showing up on.  You'll recall on the 9serve that I had to use the Stealth driver to make the built-in serial port appear to AppleTalk.  Serial ports only show up in AppleTalk without 3rd-party assistance if they are physically present as a standard serial device (or perhaps emulated, but I think that's unlikely).  I find it hard to believe that I'm seeing a 'false positive' here.  Is it a populated debug port?

  Another thing I'm wondering about is to do with the USB ports.  System Profiler reports three USB ports present - Ports 1 and 2 are the external ones, while Port 0 which comes first I can only assume is for the bluetooth card.  My Mini doesn't have the mezzenine board for Airport/bluetooth, so I can't prove it.  The bluetooth card that would go in a G4 Mini appears to be the same one that is in the MDD.  I've seen a reference online showing which pins are the USB data lines on the bluetooth connector, and the implication is that you could probably get a functional extra USB port wired from here in place of the bluetooth.  You just would have to pull the USB power from somewhere else, or at the very least dedicate a powered hub to this connection.

  Does anyone need to see NanoKernel Logs from my Mini or are we past that stage now?  If we don't need it anymore I'd like to be able to boot this ROM with the function disabled so it quits getting in the way.

[darthnVader]

Does anyone need to see NanoKernel Logs from my Mini or are we past that stage now?  If we don't need it anymore I'd like to be able to boot this ROM with the function disabled so it quits getting in the way.

I too was wanting to disable this, you can close it with Command+F2.

There are a couple of resources in the ATI Graphics Accelerator that have the Device ID for the 9200, make sure you change them all.

[MacOS Plus]

Does anyone need to see NanoKernel Logs from my Mini or are we past that stage now?  If we don't need it anymore I'd like to be able to boot this ROM with the function disabled so it quits getting in the way.

I too was wanting to disable this, you can close it with Command+F2.

There are a couple of resources in the ATI Graphics Accelerator that have the Device ID for the 9200, make sure you change them all.

  I'm pretty sure I changed those also, but I'll double-check.

[MacOS Plus]

I've been poking around at the ATI extensions with HexEdit, and have noted the following:

- ATI Graphics Accelerator - resource fork contains 5960
- ATI Resource Manager - data fork contains 5961
- ATI Video Accelerator - resource fork contains 5960

  Changing all the values to 5962 still hasn't got acceleration working.  The ATI Graphics Accelerator extension always has an X through it during startup.

  I was looking at a listing of ATI device entries online here:

https://lists.freedesktop.org/archives/xorg/2007-February/022012.html

  There are entries for 5960, 5961 and 5962 with official ATI device type titles.  I looked for other values in the list related to the RV280 to see if I could find them in the extensions.  Changing a couple others didn't help either.  So I'm stuck so far - hopefully someone else will have more luck.  It's probably something really stupidly simple.  I presume this isn't going to work until the driver loads properly in the first place, which would be indicated by having all the modes available in the Monitors control panel.  Nothing I've changed gave me any options other than the 1920x1080 / 256 color mode it is locked to.

[grhmhome]

I bought a G4 Mac Mini and upgraded the system to 1GB of ram in December. I was originally going to run only Morph OS, but I love Mac OS 9, and would rather run that. I miss playing Glider Pro. I used to run Mac OS 9 when I had a G4 eMac. This is the system I have. https://everymac.com/systems/apple/mac_mini/specs/mac_mini_g4_1.33.html

I'm willing to use my machine as a beta test machine to attempt to get Mac OS 9.x to install as I spent only $16.99 USD + shipping.