[News] Hyperion finished AmigaOne firmware | ANN.lu |
Posted on 29-Sep-2002 18:55 GMT by xisp | 212 comments View flat View list |
Hyperion is pleased to report that it has completed the AmigaOne firmware based on the GPL'ed PPCBoot project.
Source: AmigArt.
Leuven, Belgium - September 27, 2002.
Hyperion is pleased to report that it has completed the AmigaOne firmware based on the GPL'ed PPCBoot project. Below you will find a description of the implemented hardware support and functionality. We want to thank everyone who has contributed to this project.
Hardware support for the following devices:
-Serial port
-USB UHCI
-USB keyboard
-USB storage devices
-PS/2 compatible keyboard
-i8259 interrupt controller
-3com 3c920 ethernet with full busmastering
-VIA IDE controller (A and B revision)
-Real time clock (RTC)
Features of PPCBoot (as implemented by Hyperion):
-x86 BIOS emulator capable of initialising and using nearly all PCI and AGP graphic cards based on the chipsets of ATI, nVidia, 3DFX, S3, 3Dlabs (Permedia), Trident etc.
-Auto detection of SDRAM modules
-Auto detection of bus and CPU speed
-Support for PCI and AGP
-General ArtiticaS support
-General VIA686 southbridge support (686A and 686B)
-IDE support (harddisk and CD-ROM booting)
-Floppy support
-Booting over network"
Ben Hermans, Hyperion.
|
|
List of all comments to this article |
Hyperion finished AmigaOne firmware : Comment 78 of 212 | ANN.lu |
Posted by Anonymous on 30-Sep-2002 05:48 GMT | In reply to Comment 67 (CodeSmith): > As I see it, the "68K problem" crowd are expressing two views: Firstly,
> some are saying that since the 68K emulation code is not yet built into
> the OS, the 68K binaries not only are not tested but they can't be, not
> until the kernel is 100% complete.
Uhm. Sure you can test 68K binaries. You can't test PPC binaries though, and that is the point.
Assuming OS components will just work by magic when compiled for PPC is quite stupid. (No doubt some "expert" will prove me wrong about this... their main argument being you just need to add some compiler switch and It Will Work By Magic...)
> Secondly, others are saying that the parts of OS4 that we've seen so far
> are 68K code and this proves that OS4 on PPC is nothing but a cruel hoax.
It doesn't prove anything, but strongly suggest that the scenario I've proposed is true: They simple can't build and test any of the PPC built components until the 68K emulation is working fully.
> I wonder if you guys have considered that maybe the 68K components we've
> seen running within OS3.9 are fully tested (and that's how they were
> tested), and are just waiting for the emulator to be integrated so they
> can be dropped into OS4?
So you're saying that the OS components will remain 68K compiled? Err, they should be PPC native... Or maybe you're trying to suggest OS4 will be 68K, after all?
> Think about it - OS components are most likely going to be OS-friendly,
> so they don't need to be tested inside OS4 itself, 3.9 is good enough
> (this applies even to code that will be released as PPC native, like the
> new hdtoolbox - userland code *really* should not care what CPU you're
> compiling for, and with a microkernel like Exec/ExecSG 90% of the OS
> lives in userland).
Yeah, this works for the applications most of the time. But we're talking about OS components here, which mostly can't be considered "userland" in AOS (can't be considered application).
In fact, the whole concept of kernel/userland doesn't apply as-is to AOS, as there are no separate memory spaces or privilege separation. Everything run superuser in AOS (and I am not speaking about supervisor mode here!).
> Sure, I'd do some integration testing to make sure nothing broke, but if
> Hyperion followed their own rules everything should just work.
Problem is they can't test anything PPC native until they can't put the system together, that is, until the 68K emulation is working.
All they're saying is that they have this magic emulation system that will magically make everything to work, and they have not tested it in real life, ever. Sure on paper it looks "nice" for average User Joe, but really, this MMU driven method has several performance and other issues, that eventually might just force total redesign of the whole emulation system. I am sure that I'll get flamed for claiming this, but until you see it running it's just a wild dream and a paper system...
Again, yet more wild speculation from anonymous, but it's "easy" to prove me wrong, just show us OS4 running on PPC. It should run on PPC already, right? |
|
List of all comments to this article (continued) |
|
- User Menu
-
- About ANN archives
- The ANN archives is powered by #AmigaZeux. It was updated daily (news last: 22-Oct-2004; comments last: 18-May-2005).
ANN.lu was created, previously owned and maintained by Christian Kemp, www.ckemp.com.
- Contribute
- Not possible at this time!
- Search ANN archives
- Advanced search
- Hosting
- ANN.lu was hosted by Dreamhost. Sign up through this link, mention "ckemp" as referrer and he will get a 10% commission on any account you purchase.
Please show your appreciation for any past, present and future work on ANN.lu by making a contribution via PayPal.
|