Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: [1]   Go Down

Author Topic: New G4 Mini audio CSM + breakthrough fix for G4 Mini startup mouse freeze  (Read 172 times)

UnexpectedBomb

  • 4 MB
  • **
  • Posts: 5
    • GitHub account

Hi All,

Stumbling upon his forum post a few weeks ago, I was inspired by the progress of "x7x" in solving some of the Mini's OS 9 audio issues: https://tinkerdifferent.com/threads/mac-mini-g4-sound-on-os-9-2.5173/

Using his foundational work, I did some vibe coding of my own and developed a Control Strip Module that does the following:
- automatically enables audio on boot
- provides a mute/unmute function
- auto-detects when headphones are plugged/unplugged in the 3.5mm audio jack
- displays corresponding graphics to indicate mute/speaker/headphones

During development of that CSM, however, I was annoyed by the now-infamous "1-in-5 boot mouse cursor freeze" and decided to tackle that next. It was discovered that the freeze was due a VBL-interrupt issue; I have developed a small app that sits in Startup Items that, upon reaching the desktop, will UNFREEZE the cursor, effectively neutralizing the problem.

I would like your help in testing both of these solutions further, particularly the latter to ensure that my VBL-interrupt theory with the mouse-freeze is accurate. You can review technical notes and download both solutions here:

Audio CSM - https://github.com/UnexpectedBomb/G4-Mac-Mini-Audio
Mouse freeze fix - https://github.com/UnexpectedBomb/G4-Mac-Mini-VBL-Fix

BIG thanks to Elliot Nunn for sending me in the right direction on the VBL-interrupt. When testing this, please keep track of whether you are booting into a native or non-native resolution; if a native resolution causes this freeze even a single time, that would mean I need to re-evaluate my theory.

Additional OS 9 G4 Mini projects that are currently underway:
- A true audio driver that does everything the CSM does, plus enables variable audio levels (via the slider) and fixes keyboard volume controls (such as on the A1048 keyboard)
- Drivers to enable true USB 2.0 speeds both on the Mini directly and also from PCI cards on G3/G4 towers
- Drivers to enable eSATA compatibility when using PCI cards such as the LaCie 130823
- Drivers to enable the Mini's internal Bluetooth module as well as related USB Bluetooth dongles

All feedback is welcome so I can continue to improve these solutions. Thanks in advance.

-UB
Logged

smilesdavis

  • 1024 MB
  • ******
  • Posts: 1242
  • ...

awsome awsome work thank you so very much

i happen to have come across the Lacie 130823 in one of my PMG4's and I put it in my Mac OS X box.
https://www.seagate.com/content/dam/seagate/migrated-assets/lacie-content/datasheet/pci_esata_sismo_de.pdf

© 2008, LaCie / Mac® OS X 10.2.8 +

did I miss someone getting eSATA to work on OS9?
Logged
Gestalt: 40 47 54 55 58(new!) 65 67 69 112 514 PowerMac: 1,1 1,2 2,1 2,2 3,1 3,3 3,4 3,5 3,6 4,1 4,2 4,4 4,5 5,1 PowerBook: 1,1 2,1 3,1 3,2 3,3 3,4 3,5 4,1 4,2 4,3

n8blz

  • 16 MB
  • ***
  • Posts: 26
  • New Member

If the mouse freeze bug is fixed that would be huge. I’ll test it out tomorrow. Thanks!
Logged

UnexpectedBomb

  • 4 MB
  • **
  • Posts: 5
    • GitHub account

did I miss someone getting eSATA to work on OS9?
That is the same exact LaCie card I have in my MDD currently. And no, I don't know of any solution that has gotten eSATA to work in OS 9, which is why I'm spearheading it now. End goals are reading/writing external SATA drives and hopefully even booting from one as well.
Logged

IIO

  • Staff Member
  • 4096 MB
  • *******
  • Posts: 4837
  • just a number

SeriTek/1VE2 is to my knowledge the only original eSATA solution with OS9 ROM Drivers - including boot support.

What i do not know is if its eSATA connectors are actually delivering a higher voltage compared to SATA (which is the only actualy difference between to two, right?)

Logged
insert arbitrary signature here

smilesdavis

  • 1024 MB
  • ******
  • Posts: 1242
  • ...

look into FirmTek SeriTek/1S2 also sold/rebadged as Sonnet Tempo SATA / TSATA

edit: IIO beat me to it

edit2: a eSATA port carries data only - the disk still needs power from an enclosure, except for the 2006 eSATAp or eSATApd (w/USB) who i have never come across but there is a sticker on my SATA museum box that i want one since last week actually :D - they are also no offical specification of the SATA-IO they are vendor proprietory to come across the limitations of eSATA
Logged
Gestalt: 40 47 54 55 58(new!) 65 67 69 112 514 PowerMac: 1,1 1,2 2,1 2,2 3,1 3,3 3,4 3,5 3,6 4,1 4,2 4,4 4,5 5,1 PowerBook: 1,1 2,1 3,1 3,2 3,3 3,4 3,5 4,1 4,2 4,3

UnexpectedBomb

  • 4 MB
  • **
  • Posts: 5
    • GitHub account

SeriTek/1VE2 is to my knowledge the only original eSATA solution with OS9 ROM Drivers - including boot support.

What i do not know is if its eSATA connectors are actually delivering a higher voltage compared to SATA (which is the only actualy difference between to two, right?)

Interesting, I did not know about the FirmTek cards! I looked into this and they use different chips compared to the LaCIe cards, and the LaCie cards never got OS 9 drivers. Thankfully, I am not reinventing the wheel after all  ;D

And yeah I think eSATA just carries data, not power, correct? I plan to eventually have a USB-to-SATA power cable plugged into an MDD USB port to power a SATA drive (if that actually works). We'll see shortly. The USB 2.0 project is coming along nicely as well, hoping to have that one done very soon.
Logged

n8blz

  • 16 MB
  • ***
  • Posts: 26
  • New Member

Good afternoon gentlemen

I have a Mini G4 1.5 GHz connected to a 23” aluminum cinema display (running at its native 1680*1050). It has a Kingston SSD in it connected via a cheap amazon msata>ide adapter. My keyboard is an Apple white plastic one connected via USB, and my mouse is a logitech M-BJ58. I have a Ugreen USB>headphone audio adapter connected as well. About one in every seven boots I get the frozen mouse.

I dl’d VBLFix_v1.bin, expanded it on my Mac Mini, gave it 500 kb of RAM instead of the default 1xx kb (force of habit from the old days, I just always give everything more RAM), and placed it in the startup items folder. I have been restarting continuously for 30 mins now. I have not experienced one mouse freeze, which is unusual for this Mini.

I did experience one crash on boot but it occurred before startup items would have loaded and so I don’t think it was related. But if it was, one crash in 30 mins of booting is preferable to a mouse freeze every seven boots.

This is an enormous, earthshaking development.

Edit: I have been restarting this mini for an hour now. Grand total of zero mouse freezes and that crash only happened the once.

Edit again: I carried on restarting a bit more and never experienced a mouse freeze. So I made it possible to launch the VBLFix application with the keyboard and removed it from startup items. First reboot I got a mouse freeze. Launching the VBLFix application with the keyboard immediately unfroze the mouse (the first time I have ever seen a Mini running OS9 recover from that state) and reset the mouse position to absolute top left.

Congratulations, I believe you have reliably fixed the Mac Mini G4 mouse freezing issue.
« Last Edit: Today at 06:25:54 PM by n8blz »
Logged

UnexpectedBomb

  • 4 MB
  • **
  • Posts: 5
    • GitHub account

Congratulations, I believe you have reliably fixed the Mac Mini G4 mouse freezing issue.

This is very exciting to hear, thanks so much for your thorough testing and feedback! Eventually it would be nice if I can somehow get this fix gracefully integrated into the ATI driver itself to eliminate the need of the VBLFix app altogether, but at least we can safely say we are in a better place than we were before.

My personal test case was a 1.42 GHz Mini using a 20" aluminum Cinema Display running at its native 1680x1050, same as your 23". Interestingly, I never got a single mouse freeze in 30+ boots at that resolution; it only happened at the lower non-native resolutions. Perhaps it was a statistical anomaly.
Logged
Pages: [1]   Go Up

Recent Topics