Author Topic: Next step with SImagixII  (Read 1939 times)

Offline (S)ATAman

  • Veteran Member
  • ****
  • Posts: 161
  • New Member
Next step with SImagixII
« on: January 21, 2020, 03:18:36 PM »
... get rid of "old cats" build, it belongs to the ROM.
FCode was updated a while ago to reflect the 6Gbit port multiplier quirks (JMicron and Marvell).
Flasher  (that one is a big story).

I am sick of every card having a flasher.

Instead, there should be one flasher, which opens immediately.
Than, you can drag and drop the ROM or whatever firmware image you have.

That firmware image is encoded, check-summed and has the matching hardware + software info in it's header.
So it will "help" the flasher find the proper hardware, whatever that hardware could be.

So for Vitesse 7174, Silicon Image 3124, Frodo, AHCI or who-knows-what there will be only one executable.
What card is addressed is determined by the firmware image itself.

This was done years ago... but not the drag and drop.

I never was a Cocoa (or corresponding "9" niceties) person, but it has to be.
The number of flash utilities is frightening.

The worst is the life under "9".

The way it currently works is that the FCode will ultimately "swallow" both "9" and "X" drivers, encoding them inside.
The FCode image is generated by the good olde Apple Tokenizer, with the "save-image" command.

That will produce pure data file as data fork.
The current flasher needs it in resource, so there is a separate application called "fork mover" (you never ever see it, it's mine internally).
One fork mover for TurboMax/33, one fork mover for Promise/66, one fork mover for HighPoint 370-372, one fork mover for Promise/100 (legacy), one fork mover for Promise/133 for Sonnet, one fork mover for Promise/100 for Sonnet, one fork mover for Silicon Image 3112, one fork mover for Silicon Image 3114, one fork mover for CMD/SiliconImage 680 PATA. And one for Vitesse 7174.
And each fork mover has a separate, rather ugly icon.

That's it - there is no "9" flasher for 3124, 3132, Marvell 7042 and AHCI cards (yet). But wait, how about Frodo, and other obsolete PCI-X cards?
And what if there will be a correctly working PCI-X to PCIe bridge?

There is no "fork mover" for these yet. AND NEVER WILL BE.
I am not going to make even more fork movers, I am sick with already what is out there.

Instead I will change the flasher, as described above. One flasher for every need, no fork movers - just dragging and dropping the firmware image file.

But it needs some time - remember, this stuff is not for money, just for exercising the mind and it has to be done in spare time!

Since the very begin (1998) all the drivers were designed the way that they do accommodate the flash interface protocol. That makes the driver slightly larger, but the flash utility is just a wrapper.

On "9" it does a bit more, but I will tuck away what I can in SIM.

The single product which is extremely tight is the Sonnet ATA-100/133 and the TurboMax/33.

Actually... TurboMax/33 did not even have "X" drivers. But space inside of Sonnet cards is a real problem.
« Last Edit: January 21, 2020, 03:30:33 PM by (S)ATAman »