| [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.
|
|
| Hyperion finished AmigaOne firmware : Comment 51 of 212 | ANN.lu |
| Posted by xisp on 29-Sep-2002 21:29 GMT | In reply to Comment 45 (Anonymous):
Sometimes I think you only say provocative things to retrieve new information.
Now you've just done it again. You are waiting for Hyperion to answer you the question:
-"Can you boot OS4 on PPC without using any of the 68k remaining libraries?".
Which translates as: The ported modules are the most central and basic to succesfully boot and run the system?
You really hope the answer is not. So, they have beed developing an OS that, until today, remains untested. Well. Use logic again. Hyperion has a lot of work to do after OS4, like porting new games and releasing them to get some revenue of their hard work. I'm sure they don't want to waste another whole year testing OS4 until it reliably works on AmigaONE, so they have ported the exact libraries required to make the testing from the begining.
Exec+DOS+Shell+layers+RTG+some other libraries=A bootable PPC that, maybe lacks some functionality, but is enough for testing.
Thats the way I would do it. Maybe Hyperion have done it differently, but they were interested in acomplish it in a year and I'm sure they've done it in a way they can test it without petunia integrated.
Oh! and... You have no clue. It feels like a stab, don't you think? |
|
| Hyperion finished AmigaOne firmware : Comment 52 of 212 | ANN.lu |
| Posted by Adam Kowalczyk on 29-Sep-2002 21:31 GMT | In reply to Comment 47 (Anonymous):
Hehe....yes I guess I did leave out the steps I took to do it. 12hour shifts seven days a week are starting to get to me I guess. When I'm not working my day job, I'm helping out with PPCBoot. But for the sake of completeness here are all the steps.
Assumption: I'm starting with an AmigaOne with a Softex ROM
1) Boot linux on AmigaOne and login under root.
2) Connect to ftp server where latest sources are located and download appropriate file.
3) Decompress and untar said files.
4) Compile ppcboot.
5) Transfer ppcboot.bin to my laptop where the flash burner is.
6) Halt linux and turn off AmigaOne.
7) Use software and hardware on my laptop that was obtained at www.willem.org to flash AM29F040 flash chip after erasing it first.
8) Ground myself and then remove AmigaOne flash chip with approprate PLCC remover.
9) Install newly flashed chip and turn on AmigaOne and watch PPCBoot do its' thing.
10) Stop and ponder when OS 4.0 or a Linux kernel will be available.
11) Realizing that Hyperion is doing a fantastic job with OS 4.0, move on to getting a Linux kernel to work.
That is the amazing technology I use....smartass. |
|
| Hyperion finished AmigaOne firmware : Comment 53 of 212 | ANN.lu |
| Posted by Anonymous on 29-Sep-2002 21:37 GMT | "...capable of initialising and using nearly all PCI and AGP graphic cards "
I think this feature is very nice...It will tease/lure 3rd party coders to make better drivers for OS4. (Instead of basic VESA driver)
I wouldn't mind having a Geforce4 / Radeon 9700 drivers or even "XYZ Hype 3d" driver for OS4. ;) |
|
| Hyperion finished AmigaOne firmware : Comment 54 of 212 | ANN.lu |
| Posted by Anonymous on 29-Sep-2002 21:53 GMT | In reply to Comment 51 (xisp):
> Now you've just done it again. You are waiting for Hyperion to answer you the question:
>
> -"Can you boot OS4 on PPC without using any of the 68k remaining libraries?".
>
> Which translates as: The ported modules are the most central and basic to
> succesfully boot and run the system?
There are two ways to interpret the silence: Either
a) they have it working and won't break the silence (for whatever reasons), or
b) they don't have it working and obviously they can't spill it out.
Honestly, which you think is more probable?
Another thing is, if you know AmigaOS (which you obviously don't, as You Have No Clue, and You Need To Go Learn Your Lesson), it is impossible to make the OS to work completely without 68K (need to have the emulation).
Basically this answer the question, there is no need for Hyperion to reply.
> You really hope the answer is not. So, they have beed developing an OS that,
> until today, remains untested.
If you know your stuff (You...blabla), everything points to to the fact the answer is "no".
Feel free to prove me wrong. Show us OS4 working on PPC.
> Well. Use logic again.
Ho-ho. Look who's talking.
> Hyperion has a lot of work to do after OS4, like porting new games and
> releasing them to get some revenue of their hard work. I'm sure they don't
> want to waste another whole year testing OS4 until it reliably works on
> AmigaONE, so they have ported the exact libraries required to make the
> testing from the begining.
Now that is wishful thinking from YOU. YOU really hope that is the case.
> Exec+DOS+Shell+layers+RTG+some other libraries=A bootable PPC
You make me laugh. For the nth time, Go Learn Your Lesson, dude.
> that, maybe lacks some functionality, but is enough for testing.
See above.
> Thats the way I would do it. Maybe Hyperion have done it differently, but
> they were interested in acomplish it in a year and I'm sure they've done it
> in a way they can test it without petunia integrated.
So you acknowledge the possibility that they might not have done it your way?
So you didn't base your clueless rambling on some information from Hyperion?
Oh dear, what a colossal surprise here...
Oh! and... You have no clue. It feels like a stab, don't you think? |
|
| Hyperion finished AmigaOne firmware : Comment 55 of 212 | ANN.lu |
| Posted by Anonymous on 29-Sep-2002 21:56 GMT | In reply to Comment 52 (Adam Kowalczyk):
I couldn't help myself. I feel so sorry. |
|
| Hyperion finished AmigaOne firmware : Comment 56 of 212 | ANN.lu |
| Posted by Lando^Trinity on 29-Sep-2002 22:08 GMT | Just wanted to say that what Hyperion have accomplished in the last year is remarkable.
Many thanks to all involved for their hard work. Looking forward to getting my hands on OS4 (best christmas present since my A1200 9 years ago). |
|
| Hyperion finished AmigaOne firmware : Comment 57 of 212 | ANN.lu |
| Posted by Mike Veroukis on 29-Sep-2002 22:22 GMT | In reply to Comment 52 (Adam Kowalczyk):
Thanks for putting him in his place, he was getting annoying! Btw, good job and can't wait for all the pieces to come together. Ciao.
- Mike |
|
| Hyperion finished AmigaOne firmware : Comment 58 of 212 | ANN.lu |
| Posted by Anonymous on 29-Sep-2002 22:29 GMT | In reply to Comment 57 (Mike Veroukis):
Yep, it's always annoying when someone ask the tricky questions.
That "amazing tech" bit was stupid of me though, sorry about that. |
|
| Hyperion finished AmigaOne firmware : Comment 59 of 212 | ANN.lu |
| Posted by Mike Veroukis on 29-Sep-2002 22:31 GMT | Wow, the signal to noise ratio on this thread is through the floor. Lotsa morons stirring up shit they no nothing about. So what else is new, right? Well, I have some questions that perhaps have been answered somewhere already but I'll ask anyway.
1) Will the end user be able to upgrade the BIOS rom by flashing an updated version?
2) Does the BIOS support any kind of power management or monitoring?
3) How about wake-up on keyboard/modem/lan?
4) Password protection?
5) Any "Advanced User" or "Performance Tweaking" options?
I understand that Hyperion is probably not going to spend much time on all the extras, just wondering if there are any considerations for them. I think my first question is the most important. Anyway, keep up the good work.
- Mike |
|
| Hyperion finished AmigaOne firmware : Comment 60 of 212 | ANN.lu |
| Posted by KenH on 29-Sep-2002 22:37 GMT | In reply to Comment 54 (Anonymous):
Take this petty argument elsewhere. |
|
| Hyperion finished AmigaOne firmware : Comment 61 of 212 | ANN.lu |
| Posted by xisp on 29-Sep-2002 22:43 GMT | In reply to Comment 55 (Anonymous):
I go to sleep. It was funny. I laughed a lot with you.
Next time we will have another "I'm right even if I'm not right" contest, when Hyperion releases more news.
Note: Man, I reckon it: I have no clue. All I do is guessing. Now, leave alone Hyperion making their job. |
|
| Hyperion finished AmigaOne firmware : Comment 62 of 212 | ANN.lu |
| Posted by Anonymous on 29-Sep-2002 22:48 GMT | In reply to Comment 61 (xisp):
Agreed. Sleep well. |
|
| Hyperion finished AmigaOne firmware : Comment 63 of 212 | ANN.lu |
| Posted by Kjetil on 29-Sep-2002 23:27 GMT | In reply to Comment 34 (Steffen Haeuser):
How long time do you expect to be testing this product before you ship it to the customers, really hope you are not shiping an equal unstabel product as Windows95 or some thing, as Hyperian has pointed out before unexpected issues allways appers in projects development like this, I expect X number of issues involving memory protection Exec communication ports and 68k program taking to PPC programs issues be an real mess, I do expect may otter unknown issues involving making the emulator work integrated whit the ExecSG, that even won’t be discovered on the first relies of AmigaOS4.0, how are you going to dele whit this? You know there is only tree mounts to December? Yes I’m not pulling you down, I only think by the reports you are giving, that they do not correspond to estimate time for relies. Way have you not tested the 68k modules on the new ExecSG you have talked about it being finished for a long time now. I do believe that recompiling programs are not an big issue how ever the Exec is not ExecSG you have talked about new features added to ExecSG, you can’t simply recompile the 68k modules this will not take advantage of the new features can you? |
|
| Hyperion finished AmigaOne firmware : Comment 64 of 212 | ANN.lu |
| Posted by Kjetil on 29-Sep-2002 23:40 GMT | In reply to Comment 53 (Anonymous):
>"...capable of initialising and using nearly all PCI and AGP graphic cards "
>
>I think this feature is very nice...It will tease/lure 3rd party coders to make better drivers >for OS4. (Instead of basic VESA driver)
>
>I wouldn't mind having a Geforce4 / Radeon 9700 drivers or even "XYZ Hype 3d"
>driver for OS4. ;)
I believe this is to be able to se any thing at boot up before the driver is loaded in AmigaOS4.0 or Linux, this is needed or at lest on card with no PPC native firmware. |
|
| Hyperion finished AmigaOne firmware : Comment 65 of 212 | ANN.lu |
| Posted by Kjetil on 30-Sep-2002 00:01 GMT | In reply to Comment 54 (Anonymous):
Talking about the Open firmware, ye we should not be suppressed if there is a setback, as the Open firmware imputation that Hyperion has done is where not supposed to be done by Hyperion at all, this affects the workload on the developers and the resources that can be put on otter things, We can not hold Hyperion reassemble for this, I believe they have done an good job considering that the Open firmware where supposed to be done many months be fore to day. This set back makes impossible even to test booting the ExecSG kernel on the AmigaOne, The only thing that is frustrating is way they can’t admit it. |
|
| Hyperion finished AmigaOne firmware : Comment 66 of 212 | ANN.lu |
| Posted by amigammc on 30-Sep-2002 02:43 GMT | In reply to Comment 35 (Steve):
Yes, we saw that with Win2000, but that was a team of 2000 people, this is hardly the same problem |
|
| Hyperion finished AmigaOne firmware : Comment 67 of 212 | ANN.lu |
| Posted by CodeSmith on 30-Sep-2002 02:50 GMT | To the crew at Hyperion: great work guys! looking forward to using OS4
To everytone who is wondering about the 68K emulation stuff:
I admit up-front that I'm not "in the loop", so what I'm saying here is my own speculation.
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. 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. 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? 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). Sure, I'd do some integration testing to make sure nothing broke, but if Hyperion followed their own rules everything should just work. So that's one way in which the emulator can be "work in progress" without the entire project being hopelessly stalled. Again, this is just my speculation, but I don't think it's too far from the truth. |
|
| Hyperion finished AmigaOne firmware : Comment 68 of 212 | ANN.lu |
| Posted by CodeSmith on 30-Sep-2002 03:09 GMT | In reply to Comment 67 (CodeSmith):
Hmm, re-reading my posting above I realized that I used the computer term "userland", which non-English speakers probably call something else...
A CPU operates in one of two modes: system and application. System mode lets you do "dangerous" things like stopping interrupts, and user mode only lets you do "safe" things. In all the CPUs I know application mode is a subset of system mode. "Userland" is just another term for user mode. System mode is highly customized for the CPU type (eg system mode on a 68000 is different from system mode on a 68030), while application mode tends to be more generic even across CPU families - that's why you can do something extreme like taking zlib for Linux on an x86 and just recompile it on Irix on an R10000, and expect it to work. Zlib is userland code. On the other hand, if you look at the Linux kernel for the x86 and for MIPS, you'll see very many differences - the Linux kernel has a lot of system mode code in it. |
|
| Hyperion finished AmigaOne firmware : Comment 69 of 212 | ANN.lu |
| Posted by CodeSmith on 30-Sep-2002 03:12 GMT | In reply to Comment 68 (CodeSmith):
Nuts, I did it again - in the above, I meant to say "userland" is another term for "application mode" |
|
| Hyperion finished AmigaOne firmware : Comment 70 of 212 | ANN.lu |
| Posted by Alkemyst on 30-Sep-2002 03:46 GMT | golly ANN is just as bad as moobunny now.
its bad anough the alot of good amigauser only read Ann & not post because of the annon postings fools.
& now more & more are not even reading Ann anymore.
more are going to Amiga.org & the only reason i can see why Amiga.org is much better is cos there are no annon postings. |
|
| Hyperion finished AmigaOne firmware : Comment 71 of 212 | ANN.lu |
| Posted by Timothy De Groote on 30-Sep-2002 04:00 GMT | In reply to Comment 70 (Alkemyst):
hi Alcemyst(060/80MHz power user),
Just wanted to let people know that I have PPCBOOT aswell on my A1g3DEV and it's running just fine.
THat's all :)P |
|
| Hyperion finished AmigaOne firmware : Comment 72 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 05:16 GMT | In reply to Comment 59 (Mike Veroukis):
>1) Will the end user be able to upgrade the BIOS rom by flashing an updated >version?
Yes.
>2) Does the BIOS support any kind of power management or monitoring?
I'm not quite sure. It's not implemented at the moment and I don't know if PPCBoot supports this.
A new build of PPCBoot was released just yesterday (2.0) and we are still evaluating what is in there.
>3) How about wake-up on keyboard/modem/lan?
Not sure what you mean here. You mean the ability to get the machine to "wake up" upon pressing a key, getting a packet over the network etc?
>4) Password protection?
That's worth considering.
>5) Any "Advanced User" or "Performance Tweaking" options?
Not at the moment. We are planning to incorporate more features as time progresses including but not limited to adding support for more graphic-boards and CPU types.
And some kind of configuration menu. |
|
| Hyperion finished AmigaOne firmware : Comment 73 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 05:21 GMT | In reply to Comment 67 (CodeSmith):
Like I pointed out before, we are currently at work on integrating the 68K emulation layer into the system.
It should be pointed out however that the following components have been tested thoroughly:
- Exec SG (PPC native)
- 68K emulation with real life applications
- new OS modules as 68K code (*)
(*) It's clear that any bugs present in the 68K version will with a likelihood of close to 100% also be present in a PPC native version.
This is why it makes sense to first test the modules as 68K code so that any bugs that are detected can be attributed solely to the code itself rather than the 68K emulation or Exec SG+emulation layer. |
|
| Hyperion finished AmigaOne firmware : Comment 74 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 05:23 GMT | >X86 Emulator inside the PPCBoot BIOS.
Is this emulator related to BOCHS (open source X86 Emulator)?
http://bochs.sourceforge.net
Is it possible for OS / application program to use the X86 emulator inside the BIOS?
PPCBoot is targeted at embedded application, similar to AmigaDE / Intent/Elate.
http://ppcboot.sourceforge.net
Is PPCBoot suitable for Desktop computer (AmigaOne)?
Is there any plan to support AmigaDE / Intent/Elate?
Thanks. |
|
| Hyperion finished AmigaOne firmware : Comment 75 of 212 | ANN.lu |
| Posted by Don Cox on 30-Sep-2002 05:39 GMT | In reply to Comment 24 (Bladerunner):
"Sorry, maybe I didn`t get it, English is not my mother language, but as i understand ist,
there WILL be 68k OS modules in OS 4.0... Or what else does it mean? "
You haven't grasped that the Amiga OS is modular. There are many
modules which may or may not come on the OS floppies or CDs (depending
on whather you have OS 3.0, 3.9 or 4.0). There isn't a single thing
called AmigaOS.
It really does not matter whether or not some modules are running in
68k or PPC, so long as the emulator is bug free.
For example, the clock and the calculator. Do you really care whether
those modules are PPC or 68k ? Would it make the slightest difference
in practice? No.
How about SnoopDOS? Everyone has it installed as a key module in the
OS. Is it "part of AmigaOS"? Does it have to run native?
There is absolutely nothing wrong with emulation on a fast modern
processor. |
|
| Hyperion finished AmigaOne firmware : Comment 76 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 05:42 GMT | | Good job! Now get this remaining thing finished :) |
|
| Hyperion finished AmigaOne firmware : Comment 77 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 05:43 GMT | In reply to Comment 74 (Anonymous):
>X86 Emulator inside the PPCBoot BIOS.
>Is this emulator related to BOCHS (open source X86 Emulator)?
No.
>Is it possible for OS / application program to use the X86 emulator inside the >BIOS?
Possibly but you don't want to do that. The x86 Bios emulator was not written with speed in mind, it serves a very special purpose.
You'd be better off with Bochs and the likes.
>PPCBoot is targeted at embedded application, similar to AmigaDE / Intent/Elate.
Yes.
>Is PPCBoot suitable for Desktop computer (AmigaOne)?
Absolutely. It offers everything a firmware needs to offer.
Take a look at the feature list of PPCBoot as implemented for the AmigaOne.
The purpose of a firmware is to initialise the hardware prior to booting.
>Is there any plan to support AmigaDE / Intent/Elate?
Not at this stage. Note that this would not require any dramatic changes in the firmware.
One single firmware may be used to boot up several different operating systems. |
|
| 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? |
|
| Hyperion finished AmigaOne firmware : Comment 79 of 212 | ANN.lu |
| Posted by anarchic_teapot on 30-Sep-2002 05:50 GMT | In reply to Comment 38 (Anonymous):
Anonymous Troll wrote:
"So, did Ole-Egil and others already receive the source code for the ROMs that have been distributed to beta testers / developers/ whatever Alan is calling them at the moment?"
What makes you assume these people actually *want* the source code? They do have other things to test, you know. Ever heard of team work? |
|
| Hyperion finished AmigaOne firmware : Comment 80 of 212 | ANN.lu |
| Posted by Ole-Egil on 30-Sep-2002 05:59 GMT | In reply to Comment 38 (Anonymous):
Memo to self. When leaving the office to have a chat with colleagues, don't write jibberish in open "Add a new comment" windows from ANN.
:-)
(I have no idea why I was gonna write something here, but name and email was already filled out :-) ) |
|
| Hyperion finished AmigaOne firmware : Comment 81 of 212 | ANN.lu |
| Posted by CD32freak on 30-Sep-2002 06:05 GMT | | Please, if you are not interested in the new AmigaOS 4.0 and AmigaOneG3-SE, simply don't buy it. It's your right as a consumer to do so. However, don't poison the Amiga forums with your rhetoric... |
|
| Hyperion finished AmigaOne firmware : Comment 82 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 06:06 GMT | In reply to Comment 78 (Anonymous):
>this MMU driven method has several performance and other issues, that eventually >might just force total redesign of the whole emulation system
The MMU method does not present any issues with regards to performance.
Even a total redesign would be very easy to accomplish as we could just go the MorphOS route and go for "emulation traps".
But that's exactly what we don't want to do. |
|
| Hyperion finished AmigaOne firmware : Comment 83 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:11 GMT | In reply to Comment 81 (CD32freak):
Here is a fun one!
Hyperion report solid progress, lamers get scared. Notice Ben Hermans following
up to a reasonable question and start to post lame barely technical clueless crap about C/C++ source code portability ( or claimed lack thereof of it clearly demonstrating no knowledge of compilers ).
It is of course technically infeasable to take C/C++ source decks and use a compiler on a different platform which provides API compatibility but has a different ( shock ) processor to compile it to the target platform. Yep. You have to hand code the ASM bytecode by bytecode. Well, you do if the target platform is not MorphOS. ;-)
Now insert lame off topic comments and attempt to start a MOS/AOS fLame war.
It's gotta be A N N!
Keep going Hyperion, you seem to have the childish ones rattled - you know the ones that use childish invective and think that Visual Basic is the pinnacle of development environments. |
|
| Hyperion finished AmigaOne firmware : Comment 84 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:13 GMT | In reply to Comment 82 (Ben Hermans/Hyperion):
> The MMU method does not present any issues with regards to performance.
Right, so CPU exceptions are free now? (oh why do I bother...) |
|
| Hyperion finished AmigaOne firmware : Comment 85 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:14 GMT | In reply to Comment 82 (Ben Hermans/Hyperion):
LOL!
"Oh I don't really understand how they do that, go the MOS route..."
aka
"I am here to try and spread FUD about the ability of AOS4 to work properly
and when I have spread my clueless muck Ill say /"go the MOS route/"."
Transparent knobwits unite, on ANN.lu. |
|
| Hyperion finished AmigaOne firmware : Comment 86 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:15 GMT | In reply to Comment 83 (Anonymous):
Now who is acting lame here...? |
|
| Hyperion finished AmigaOne firmware : Comment 87 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:15 GMT | In reply to Comment 84 (Anonymous):
Because you do not really have a position and therefore do not know when to stop idle speculation? |
|
| Hyperion finished AmigaOne firmware : Comment 88 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:16 GMT | In reply to Comment 86 (Anonymous):
Ah, a rhetorical question I believe. |
|
| Hyperion finished AmigaOne firmware : Comment 89 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 06:19 GMT | In reply to Comment 84 (Anonymous):
Ofcourse CPU exceptions are not free.
But the overhead involved with an exception is far, far less on PPC than on 68K.
Thomas Frieden has studied this for quite some time and as he's the one who wrote Exec SG, I'm sure he knows what he is talking about.
Remember, he wrote the exception handling.
Sure, there is a performance penality but this has to be balanced against the fact that it obviates the need for emulation traps that need to be inserted into the source-code and which pollute the code-base and API/ABI.
The idea is to move away from 68K entirely at one point and the MMU approach allows us to do so without leaving any clutter behind once this is accomplished. |
|
| Hyperion finished AmigaOne firmware : Comment 90 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:22 GMT | In reply to Comment 87 (Anonymous):
Heh, well if you read Mr Hermans reply, he told us that the components are being tested under 68K. He did give a different reason for that, though. I still claim that going from 68K->PPC is not that trivial he wants ppl to think.
Speculation? Maybe. Valid questions? Definetely. |
|
| Hyperion finished AmigaOne firmware : Comment 91 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:23 GMT | In reply to Comment 89 (Ben Hermans/Hyperion):
"Of course CPU exceptions are not free"
Lets see, is this an OS or a sad scene demo? Haggling over this is like a return to the S/370 days which is the last time anyone cared about this making
a visible difference.
However put a bunch of Amiga users selected at random in a room and they will
talk balls at each other until the cows come home, ergo why there are so few
new applications being developed. |
|
| Hyperion finished AmigaOne firmware : Comment 92 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 06:31 GMT | In reply to Comment 90 (Anonymous):
>I still claim that going from 68K->PPC is not that trivial he wants ppl to think.
This was already asked and answered so many times.
Sure, if you are using SAS C as opposed to GCC, you will need to invest a bit more time to migrate your code but then this would be the same irrespective of whether you would be targeting the same CPU or not.
If you are already using GCC, you can look forward to near perfect compatibility at the source-code level.
Let me remind you again what Thomas Frieden told me the previous time somebody came up with these issues.
(I do find it particularly funny BTW that some people are always bashing us for being "game porters" and then turn around and claim that we would be in the dark about exactly that - moving code from one platform to another. These past few years we have moved code from Windows to Linux x86 and PPC, Mac OS 9/X and AmigaOS and from Linux x86 to Mac, from Amiga to Mac etc. etc.)
Here goes again:
>"Okay, so explain how callback hooks, inputhandlers, interrupt code and
> similar work?
Hooks and input handlers are just 68k code that is called from PPC. No
special precautions are needed, the emulator will catch instruction
reads from illegal code segments and just assume they are 68k.
For interrupts, there's a different mechanism. Interrupt code is called
through specialised hooks (struct Interrupt), which contain a full
struct Node (including type). There are now two types that actually
define an interrupt, NT_INTERRUPT, which is an old 68k interrupt, and
NT_EXINTERRUPT, which is a PPC native interrupt.
> They must be 68k code or else old 68k apps could not run at all. If you
> compile your app for PPC, all code will be PPC -> problem.
That's the usual bullshit, and that's the prime reason for the exception
based cpu switches: With our approach, there is *NO DISTINCTION* between
68k code and PPC code. All code is _assumed_ to be PPC, and is jumped to
accordingly. But PPC code is only loaded into specific virtual segments
(these are flagged as executable). If the CPU tries to fetch
instructions from a non-execute segment, an exception is triggered, and
the emulator will pick up. Likewise, the emulator knows which segments
contain PPC code, so if a 68k program wants to jump into PPC code, the
emulator just transfers control to the normal operation." |
|
| Hyperion finished AmigaOne firmware : Comment 93 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:39 GMT | In reply to Comment 92 (Ben Hermans/Hyperion):
Yes, I most definetely remember this stuff, and yes, I still claim it isn't that easy. The best way to prove me wrong is to show it working in real life.
That should be quite soon now, if you're going to get it all done before end of the year... |
|
| Hyperion finished AmigaOne firmware : Comment 94 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 06:40 GMT | In reply to Comment 93 (Anonymous):
Why don't you specify exactly why you think it isn't so easy?
You offer no justification for this claim whatsoever. |
|
| Hyperion finished AmigaOne firmware : Comment 95 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 06:53 GMT | In reply to Comment 94 (Ben Hermans/Hyperion):
True. I will not point out the possible problems for you. It's up to you to figure out why I won't do that.
I bet the crowd will go wild now, and claim I don't have have a clue about this, but I am sure you know better than that. |
|
| Hyperion finished AmigaOne firmware : Comment 96 of 212 | ANN.lu |
| Posted by Nicholai Benalal on 30-Sep-2002 06:56 GMT | In reply to Comment 89 (Ben Hermans/Hyperion):
"But the overhead involved with an exception is far, far less on PPC than on 68K.
Thomas Frieden has studied this for quite some time and as he's the one who wrote Exec SG, I'm sure he knows what he is talking about."
We will all believe you when there is a working implementation. Until then, neither you, mr Frieden nor anyone else can be sure that your scheme will work reliably and have decent performance. The issues involved are too many for a paper solution to be worth trusting. Let's wait with this discussion 3 or 4 months. At that point you might know more.
Friendly,
Nicholai |
|
| Hyperion finished AmigaOne firmware : Comment 97 of 212 | ANN.lu |
| Posted by Ben Hermans/Hyperion on 30-Sep-2002 07:12 GMT | In reply to Comment 95 (Anonymous):
Okay, so you don't want to help us out by giving us pointers.
I hope you can at least reserve judgement until the product is actually available then.
This would save us a lot of time and effort here. |
|
| Hyperion finished AmigaOne firmware : Comment 98 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 07:18 GMT | In reply to Comment 95 (Anonymous):
>et the crowd will go wild now, and claim I don't have have a clue about this
LOL!!!! You don't have a clue mate. C is a portable language and there is no reason why most software can't easily be ported from one OS to another or from different CPU architectures. Take Intuition. If it's written in pure C now, then why would it be a problem to recompile it under the same compiler (GCC) for a different CPU? If the code doesn't do crazy stuff like access 68k registers or blah blah low level hardware stuff (which a GUI wouldn't need to do) then you're a fucking moron...
Whatever |
|
| Hyperion finished AmigaOne firmware : Comment 99 of 212 | ANN.lu |
| Posted by Ole-Egil on 30-Sep-2002 07:23 GMT | In reply to Comment 79 (anarchic_teapot):
Ah, now I remember I was gonna say:
I have all the source code I'll ever need :-) |
|
| Hyperion finished AmigaOne firmware : Comment 100 of 212 | ANN.lu |
| Posted by Anonymous on 30-Sep-2002 07:29 GMT | In reply to Comment 98 (Anonymous):
Your ignorance is so evident I won't go into details. Have fun with yourself. |
|
|
- 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.
|
