Click Here
home features news forums classifieds faqs links search
6071 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
» Home
» Features
» News
» Forums
» Classifieds
» Links
» Downloads
Extras
» OS4 Zone
» IRC Network
» AmigaWorld Radio
» Newsfeed
» Top Members
» Amiga Dealers
Information
» About Us
» FAQs
» Advertise
» Polls
» Terms of Service
» Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
13 crawler(s) on-line.
 147 guest(s) on-line.
 0 member(s) on-line.



You are an anonymous user.
Register Now!
 blmara:  24 mins ago
 miggymac:  1 hr 7 mins ago
 Gunnar:  1 hr 35 mins ago
 pixie:  2 hrs 56 mins ago
 DiscreetFX:  3 hrs 35 mins ago
 DWolfman:  3 hrs 44 mins ago
 cncparts:  5 hrs 18 mins ago
 saipaman4366:  6 hrs 4 mins ago
 Beajar:  6 hrs 23 mins ago
 Rob:  6 hrs 25 mins ago

/  Forum Index
   /  Amiga General Chat
      /  AmigaOS4 runs on MacMini
Register To Post

Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 Next Page )
PosterThread
Turrican3 
Re: AmigaOS4 runs on MacMini
Posted on 30-Jun-2008 18:28:29
#201 ]
Regular Member
Joined: 20-Jun-2003
Posts: 386
From: Italy

@Hypex

Quote:
Did Escom really manufacture A1200 motherboards? They weren't old Commodore stock? If so I can't believe they would do that without changing the design.

Yes, they did manufacture A1200's, in France maybe if I recall correctly. As far as I remember they were identical to plain good old C= A1200 motherboards, I believe they even left intact the signatures of the engineers near the internal expansion slot.

The differences were the kickstart (Escom A1200's had 3.1, C= A1200's had 3.0) and the infamous PC-drive that made some games incompatible.

Don't know about Quikpak though, by that time I was still following the platform but my main interest had switched to Playstation.

Quote:
I don't consider the AmigaOne to be an Amiga in the sense of being a real Amiga.

Me too.

I think the closest thing that could "feel" like a real Amiga would be a PS3, but are there really any chances OS4 will be ported to that platform? We can only hope so.

Quote:
So in this case whatever CPU they use has to be compatible with the 68k to some extent. Remember they are porting the OS and not a copy and so in doing so they use the same system structures.

Again, I think WindowsNT proves quite the opposite regarding (un?)necessary CPU similarities.

Quote:
It would then end up like Linux I think. AmigaOS-PPC. AmigaOS-x86. Etc.

Naaa, highly unlikely IMHO, it would be a waste of resources even if developers were not as few as they are.

There's no need to release the OS to multiple architecture: hardware abstraction is nice of course (I would say is a must nowadays), but I believe is mostly useful to the developers to make the transition to different HW easier/faster/whatever.

Which turn us back to the original point: OS4 absolutely, desperately needs a cheap, affordable HW platform to run on, no matter what it is (IMHO).

Last edited by Turrican3 on 30-Jun-2008 at 06:30 PM.

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 6:52:02
#202 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@michalsc

Quote:
Let's have a look. The very first portion of AROS PPC kernel is written in assembly - it initializes the very basic MMU mapping (AROS kernel is mapped at 0xff800000 address), sets the stack, clears bss and such. It uses 26 PPC instructions.

Exception prologue on sam440 AROS port is a macro, 33 lines, expanded 16 times (for each exception entry one). Prologue calls the trampoline code which saves the rest of the CPU context and changes few bits in MSR - 34 lines. Return from exception restores the CPU context and returns back to user mode - 48 lines. The rest of exception handlers is in C.


Okay, so that's not much so far. Quite few really. I notice the PPC looks like a pain when it comes to exceptions. 68k would need a few lines right? a MOVEM and an RTE mostly.

Quote:
Additionally, the exec.library/StackSwap is written in assembly. Additionally, few lines of the scheduler (and whole schedule chain of AROS) are in assembler, eg:


Okay, nothing too huge yet.

Quote:
There are some calls of this kind in AROS kernel, perhaps 100 or less... The userland code may use the C interface of AROS kernel, that is, calling functions of kernel.resource and exec.library.


Now I understand the point of having the kernel.resource, good idea.

Quote:
x86_64 AROS is similar. A tiny startup code which prepares the stack and calls C entry, tiny portions of assembly in prologue of exceptions and interrupts entry/exit. StackSwap is, as in case of PPC, written in assembly. The biggest portion of asm code in case of x86_64 is the smp trampoline used to wake up the CPUs - beginning of this code is 16-bit assembly which turns on 32-bit mode, which sets up MMU and turns on 64-bit code, which calls C function :) The very rest of assembly is hidden behind inline functions and macros.


That reminds me, do you use standard ABIs like SystemV for PPC? And what exists for x86?

It seems funny, on the x86, that bit of work to go from 16 to 32 to 64. Hehe.

Now what I am wondering about with 64-bit, apart from AROS being the first Amiga type OS to natively support 64-bit CPUs, would it have to use 64-bit pointers? Of course, 64-bits refers to data width, but I also seem to think it should be pointer width as well. In this case, can the 32-bit pointers in AROS Exec suffice? Or are there compiler tricks to extend pointers to 64-bits?

Apart from that, can AROS data structures be extended to 64-bits? And what inside the AROS kernel itself can make use of 64-bit data lengths?

Quote:
Yes, because of me :) The x86 core had much more assembly code in it. It was written in the old days, when I was still assembler freak. Now I grew up ;)




I've been testing AROS on an old PC donated by a family member. It was made in the time of Windows '95. It can't boot from CD, has 48MB and an "Amiga, Inc." Matrox Millennium gfx card. I can't get AROS to work on it, according to the specs of AROS it should work.

It can do 1024x768-16 and 1280x1024-8 which was selected in VESA menu. The floppy boots, then starts loading from CD, then I hear the monitor clock and nothing else happens. I don't think the CD drive light goes out. I tried the VGA-4bpp nodma modes and got text on screen. All was fine until it said IIRC that sfs and ffs handlers failed. At the end it says it can't open the trackdisk.device, twice. Why would that be? What would a rackdisk.device, be doing in a PC OS that can't read Amiga native disks?

Quote:
Accessing unaligned data is slower comparing to the PPC-native Big Endian mode.


Okay, so is little-endian on PPC considered unaligned?

Quote:
As I have already pointed it - you cannot have less assembly lines that easy. Besides, 800 lines of code is really just a bit. I took me much less than a month to have aros' exec running and multitasking on sam440. Keep in mind that I was not working full time on it: usually two or three hours every second or third evening.


Okay, so it doesn't sound too bad. Do you have to write less then 800 lines for Sam port?

Last edited by Hypex on 07-Jul-2008 at 06:53 AM.

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 7:09:21
#203 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@Turrican3

Quote:
Yes, they did manufacture A1200's, in France maybe if I recall correctly. As far as I remember they were identical to plain good old C= A1200 motherboards, I believe they even left intact the signatures of the engineers near the internal expansion slot.


No way! Cheapskates!

Quote:
The differences were the kickstart (Escom A1200's had 3.1, C= A1200's had 3.0) and the infamous PC-drive that made some games incompatible.


Okay, yes. I didn't see the point, we still didn't get HD drives, how hard can it be? IIRC they removed the diskready line which stuffed things up. They should have known better.

Quote:
Don't know about Quikpak though, by that time I was still following the platform but my main interest had switched to Playstation.


I was following the Playstation but my main was still the Amiga. IIRC QuikPak did the A4000/060's.

Quote:
Me too.

I think the closest thing that could "feel" like a real Amiga would be a PS3, but are there really any chances OS4 will be ported to that platform? We can only hope so.


Not without a license! Unless a "Project Playana" comes along. Um, that sounds crap.

I heard there is a MOS port to MacMini, are they always copying us? Will we see an AROS port to any-Mac? If MOS or AROS could be ported to the PS3 or Wii or 360 then it opens the doors and shows how realistic it would be of OS4 being able to actually run on the hardware.

Quote:
Again, I think WindowsNT proves quite the opposite regarding (un?)necessary CPU similarities.


That depends, it has been said PPC gets around this by using it's little-endian addressing mode. So, bearing that in mind, does the x86 also have a big-endian addressing mode or instruction? If so, then problem solved.

Otherwise, it isn't so easy. If you keep the endianess out of it and only run the OS and software for that native CPU it is compiled for then it is a non-issue.

Quote:
Naaa, highly unlikely IMHO, it would be a waste of resources even if developers were not as few as they are.

There's no need to release the OS to multiple architecture: hardware abstraction is nice of course (I would say is a must nowadays), but I believe is mostly useful to the developers to make the transition to different HW easier/faster/whatever.

Which turn us back to the original point: OS4 absolutely, desperately needs a cheap, affordable HW platform to run on, no matter what it is (IMHO).


Yes, agreed,

Hopefully the Bill Mc daddy can sort something out.

 Status: Offline
Profile     Report this post  
Turrican3 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 8:06:29
#204 ]
Regular Member
Joined: 20-Jun-2003
Posts: 386
From: Italy

@Hypex

Quote:
No way! Cheapskates!

Sorry, I'm not sure I understand this.

Quote:
Okay, yes. I didn't see the point, we still didn't get HD drives, how hard can it be? IIRC they removed the diskready line which stuffed things up. They should have known better.

HD drives were custom built because they needed to run at half speed due to Paula being unchanged since the launch of the Amiga (and so it was unable to pump data at rate required by a 1,76 MB drive). I guess it wasn't exactly cheap -if possible at all!!- to fund a small production run at the time.

Quote:
Not without a license! Unless a "Project Playana" comes along. Um, that sounds crap.

Indeed.

Quote:
If MOS or AROS could be ported to the PS3 or Wii or 360 then it opens the doors and shows how realistic it would be of OS4 being able to actually run on the hardware.

It is unrealistic IMHO to hope for a port to Wii or 360, especially the latter since Microsoft is quite known for being a bit contrary ( ) to any other OS whose name does *not* start with "Win" and end with "dows".

The only possible (and IMHO smart) chance would be porting to PS3, assuming a license is granted as you said of course.

Quote:
Hopefully the Bill Mc daddy can sort something out.

I'm afraid an AmigaOS 4 Xbox360 port officially endorsed by Microsoft is more likely than that.

But one can always hope, right?

 Status: Offline
Profile     Report this post  
michalsc 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 10:01:16
#205 ]
AROS Core Developer
Joined: 14-Jun-2005
Posts: 377
From: Germany

@Hypex

Quote:
Okay, so that's not much so far. Quite few really. I notice the PPC looks like a pain when it comes to exceptions. 68k would need a few lines right?


Yes, m68k would need a little bit less assembly. The rule of thumb: the more complex cpu is, the more context has to be saved in exceptions. Besides, on PPC I could have used stmw (store multiple word) to save more than one register at once.

Quote:
That reminds me, do you use standard ABIs like SystemV for PPC? And what exists for x86?


Yes, I use standard SysV ABI on PPC. However, I will probably change it a bit in the future, so that the library base is passed in %r13 register. x86 and x86_64 use SysV ABI at the moment.

Quote:
It seems funny, on the x86, that bit of work to go from 16 to 32 to 64.


Well, GRUB does for me the transition to 32 bit mode on Bootstrap Processor by itself. I need to do the complete 16->32->64 transition only once for every Application Processor. Since x86_64 AROS does not even allow 32-bit code to run on it, I do not have to deal with mode switching anymore.

Quote:
Now what I am wondering about with 64-bit, apart from AROS being the first Amiga type OS to natively support 64-bit CPUs, would it have to use 64-bit pointers? Of course, 64-bits refers to data width, but I also seem to think it should be pointer width as well. In this case, can the 32-bit pointers in AROS Exec suffice? Or are there compiler tricks to extend pointers to 64-bits?


All pointers in x86_64 are 64-bit in length. There are no tricks to achieve that. The sizeof(APTR) equals 8. The sizes of datatypes for x86_64 aros are following:
sizeof(UBYTE)=sizeof(uint8_t)=1
sizeof(UWORD)=sizeof(uint16_t)=2
sizeof(ULONG)=sizeof(uint32_t)=4
sizeof(UQUAD)=sizeof(uint64_t)=8
sizeof(APTR)=sizeof(IPTR)=sizeof(void *)=sizeof(uintptr_t)=8

as a comparison, the sizes of x86 types are *almost* the same. The only difference is the last line, which on x86 is as following:
sizeof(APTR)=sizeof(IPTR)=sizeof(void *)=sizeof(uintptr_t)=4.

Quote:
Apart from that, can AROS data structures be extended to 64-bits?


They can. They ARE :)

Quote:
And what inside the AROS kernel itself can make use of 64-bit data lengths?


Everything where the computed result exceeds 32-bit boundary.

Quote:
I've been testing AROS on an old PC donated by a family member. It was made in the time of Windows '95. It can't boot from CD, has 48MB and an "Amiga, Inc." Matrox Millennium gfx card. I can't get AROS to work on it, according to the specs of AROS it should work.


We have had some major issues with our ata.device recently. Try the great vmwaros live 0.8 distribution (Website link) and boot in "emergency mode".

Quote:
What would a rackdisk.device, be doing in a PC OS that can't read Amiga native disks?


Reading PC-formatted floppies? Like mfm.device on Amiga.

Quote:
Do you have to write less then 800 lines for Sam port?


Nope. It wasn't that easy:)
- ub2lb - my own opensource version of slb known from OS4 - it is responsible for loading and relocating aros kernel and all rom modules into memory. The ub2lb has been now massively extended by Giuseppe Coviello who added harddrive support and ability to boot both AROS and linux with it.
- kernel resource: ca. 800 lines of assembly and much more in C
- replacing some exec functions with PPC-aware variants
- Changing the way timer.device works
- Fixing some drivers (eg. radeon, ata, ohci and some other portions of USB stack) so that they work on BigEndian system
- Worst thing: making many parts of AROS aware of cache incoherent system
- much, much more ;)

 Status: Offline
Profile     Report this post  
Crumb 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 12:38:24
#206 ]
Elite Member
Joined: 12-Mar-2003
Posts: 2209
From: Zaragoza (Aragonian State)

@Hypex

I'm only interested in PPC AmigaOSes as long as they run m68k AmigaOS apps (and MorphOS/OS4 apps).

Unfortunately, as long as PPC AROS has no m68k compatibility and/or PPC MorphOS/OS4 compatibility it will run much better in any PC.

Last edited by Crumb on 07-Jul-2008 at 12:39 PM.

_________________
The only spanish amiga news web page/club: CUAZ

 Status: Offline
Profile     Report this post  
bert64 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 12:54:55
#207 ]
Member
Joined: 6-Jun-2007
Posts: 18
From: Unknown

@jorkany

Yeah, pointless to port to a "dead" platform, much better to keep it restricted to running on another dead platform, that has been dead a lot longer and has far less existing units out in the wild.

That's the most pathetic excuse i've heard sofar.

 Status: Offline
Profile     Report this post  
bert64 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 13:01:26
#208 ]
Member
Joined: 6-Jun-2007
Posts: 18
From: Unknown

@jack

OS4 for "classic" ppc hardware? And in the same post complain that macs aren't that common? Come on, there are many times more ppc macs out there than ppc amigas.

Who will license new hardware? There's pretty much no market for it... Noone is going to build new PPC systems except for certain very large markets (see games consoles), and all the PPC boards i've seen available are slower than "discontinued" macs, they all seem to be G4 based and not even clocked as high as the later G4 macs, never mind the G5 based ones.

 Status: Offline
Profile     Report this post  
bert64 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 13:05:10
#209 ]
Member
Joined: 6-Jun-2007
Posts: 18
From: Unknown

@jack

The used mac market is far more healthy than the used amiga market...
A used ppc mac will cost you a lot less than a used ppc capable amiga, be a lot easier to find, and be a lot faster. A temporary solution while far from ideal is much better than no solution at all.

 Status: Offline
Profile     Report this post  
bert64 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 13:17:53
#210 ]
Member
Joined: 6-Jun-2007
Posts: 18
From: Unknown

@Turrican3

Assuming a license is granted?
You don't need a license to write software for PS3 unless you want to use accelerated video.. Linux runs on it just fine.

As for requiring a license to run OS4 on particular hardware, that's just stupid and artificially limits the available hardware... Take a leaf out of Microsoft's book, let anyone who wants to buy your OS and run it on any hardware they see fit. Just do that and provide enough information to produce drivers.

If you're business is making software, then don't try to be anal about the hardware, it will only serve to decrease sales. Microsoft know this, and this is part of the reason they've been so successful.

 Status: Offline
Profile     Report this post  
Turrican3 
Re: AmigaOS4 runs on MacMini
Posted on 7-Jul-2008 13:56:02
#211 ]
Regular Member
Joined: 20-Jun-2003
Posts: 386
From: Italy

@bert64

Quote:
You don't need a license to write software for PS3 unless you want to use accelerated video.. Linux runs on it just fine. As for requiring a license to run OS4 on particular hardware, that's just stupid and artificially limits the available hardware...

I was talking about the "legal" requirements (in bold). Unfortunately, as far as we know there is currently no developer allowed to write a port of OS4 for PS3 hardware.

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 14:44:38
#212 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@Turrican3

Quote:
Sorry, I'm not sure I understand this.


Escom, they were cheap not improving the design I mean.

Quote:
HD drives were custom built because they needed to run at half speed due to Paula being unchanged since the launch of the Amiga (and so it was unable to pump data at rate required by a 1,76 MB drive). I guess it wasn't exactly cheap -if possible at all!!- to fund a small production run at the time.


Another thing that could have been made use of. I wonder if Paula could have run at double speed with a VGA display, she did with sound. Even at half speed it still would have been good to use HD disks.

I still think it is possible to connect a standard drive with a simple circuit. I have a PCB here someone made with some resistors and a logic IC that allows you to connect up a standard PC floppy drive to an Amiga floppy port. I've tested it, it works. IIRC I think it could even be used to write HD disks up to 1.5MB.

Quote:
It is unrealistic IMHO to hope for a port to Wii or 360, especially the latter since Microsoft is quite known for being a bit contrary ( ) to any other OS whose name does *not* start with "Win" and end with "dows".

The only possible (and IMHO smart) chance would be porting to PS3, assuming a license is granted as you said of course.


I was trying to find console hardware all with similar CPUs. All we need is an SDK for each to open up the possibility. In that case AROS is the most likely candidate. Since license issues are just going to drag on...

Quote:
I'm afraid an AmigaOS 4 Xbox360 port officially endorsed by Microsoft is more likely than that.


Hehe.

Quote:
But one can always hope, right?


Yep. Who knows, maybe Bill hopes for these things as well.

Last edited by Hypex on 08-Jul-2008 at 02:46 PM.

 Status: Offline
Profile     Report this post  
number6 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 15:07:57
#213 ]
Elite Member
Joined: 25-Mar-2005
Posts: 11540
From: In the village

@thread

Just a restatement, since we have new members in the last few years.
Some people seem to believe that Moana was the 1st attempt at persuading AI to look at Mac hardware. Not so.
I'm just mentioning this so people don't relate all this to the legal activity.

2006 question/comment from @Pulsating Quasar:

Quote:
It's Apple right.They allow the MiniMac G4 design to be reused for an Amiga system. Tons of second hand hardware thingies will then be available.


ACKcontrols (who had discussed this option with AI long before this):

Quote:
Several solutions based on previously existing Apple hardware were looked at, but the logistics of acquring the hardware for non-DIYers was going to be difficult to manage. Personally, the documentation aspect for driver development wasn't a big issue, but gathering sufficient quantities of motherboards efficiently and cost effectively wasn't going to work out. It may have served as a short term solution, but the goal has always been to produce a board that could be manufactured to meet RoHS compliance and


As I understood it, there was a warehouse with 500 units ready to go, until the idea was killed.

Certainly you could postulate that "any" ideas getting killed were due the uncertainty that people had over whom would have the "rights" in the future, but we do have this earlier instance of an attempt.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 15:58:23
#214 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@michalsc

Quote:
Yes, m68k would need a little bit less assembly. The rule of thumb: the more complex cpu is, the more context has to be saved in exceptions. Besides, on PPC I could have used stmw (store multiple word) to save more than one register at once.


Yes, true, true.

Quote:
Yes, I use standard SysV ABI on PPC. However, I will probably change it a bit in the future, so that the library base is passed in %r13 register. x86 and x86_64 use SysV ABI at the moment.


Okay. Now what I don't understand is how SysV can also be used on x86. AFAIK an ABI specifies what registers are used for the stack and parameters. The specs for SysV looks like it needs 16 base registers to round up the ones needed for stack, system and parameters. However, AFAIK x86-64 has 16 but x86-32 still only has 8. So how can an ABI made for specific PowerPC CPUs work on x86 CPUs? I've only read about SysV PPC.

Quote:
Well, GRUB does for me the transition to 32 bit mode on Bootstrap Processor by itself. I need to do the complete 16->32->64 transition only once for every Application Processor. Since x86_64 AROS does not even allow 32-bit code to run on it, I do not have to deal with mode switching anymore.


I see. So can GRUB be customised with a boot message? It does seem funny on the floppy when it says "GRUB loading stage 2..." What happened to stage 1?

Quote:
All pointers in x86_64 are 64-bit in length. There are no tricks to achieve that. The sizeof(APTR) equals 8. The sizes of datatypes for x86_64 aros are following:
sizeof(UBYTE)=sizeof(uint8_t)=1
sizeof(UWORD)=sizeof(uint16_t)=2
sizeof(ULONG)=sizeof(uint32_t)=4
sizeof(UQUAD)=sizeof(uint64_t)=8
sizeof(APTR)=sizeof(IPTR)=sizeof(void *)=sizeof(uintptr_t)=8

as a comparison, the sizes of x86 types are *almost* the same. The only difference is the last line, which on x86 is as following:
sizeof(APTR)=sizeof(IPTR)=sizeof(void *)=sizeof(uintptr_t)=4.


I am starting to see how convenient a C pointer is here. I do wonder how close to the original structures like Exec, Gfx and Intuition AROS is? In any case, the length would be bigger on 64-bit and since that only runs 64-bit code it must make things a lot easier not supporting 32-bit apps.

Quote:
They can. They ARE :)


Much more than simple pointer extensions?

Quote:
Everything where the computed result exceeds 32-bit boundary.


Okay I am thinking date structures, DOS, utility, anything where a 64-bit value would benefit. Actually making full use of that could be a lot of work. Like for functions.

Quote:
We have had some major issues with our ata.device recently. Try the great vmwaros live 0.8 distribution (Website link) and boot in "emergency mode".


It sounds great, and huge. But the machine I am testing it on is underspecced! It requires 500Mhz and 196MB. I have only 166Mhz and 48MB! Apart from that I don't see a vmwaros live floppy image. My BIOS doesn't CD boot.

I did find out that my CD drive was connected as slave only on second bus. So corrected that to master. Either way, the floppy boots, and then it looks like it is about to load off CD but it just stops. It resets the keyboard and starts reading the CD. I took the CD out once and got al this text scrolling up the screen telling me to put in a bootable disk.

The CDRW I burnt does work, I have tested it on a better PC.

Quote:
Reading PC-formatted floppies? Like mfm.device on Amiga.


So it does, but trackdisk wasn't meant to read PC floppies, since it can't read Amiga floppies on a normal PC. Just picking on the name here, it just seems inaccurate. For example, you didn't choose scsi.device for the harddisk, you have the more accurate ata.device.

What's with that rackdisk.device? Sounds like it's for floppy drives that are rack mounted. What will I invent next, the stackdisk.device, where disks are stacked up in a multi stacking floppy drive?

Quote:
Nope. It wasn't that easy:)
- ub2lb - my own opensource version of slb known from OS4 - it is responsible for loading and relocating aros kernel and all rom modules into memory. The ub2lb has been now massively extended by Giuseppe Coviello who added harddrive support and ability to boot both AROS and linux with it.
- kernel resource: ca. 800 lines of assembly and much more in C
- replacing some exec functions with PPC-aware variants
- Changing the way timer.device works
- Fixing some drivers (eg. radeon, ata, ohci and some other portions of USB stack) so that they work on BigEndian system
- Worst thing: making many parts of AROS aware of cache incoherent system
- much, much more ;)


Wow, that's a lot! Looks like they did possibly need all those assembler lines in the OS4 kernel. Wonder how much assembly Linux needs for each arch?

I am most interestied in ub2lb. How closely does it follow slbv2? Could it try to boot an OS4 partition on the Sam440?

That last point is interesting, cache incoherency, I though the AmigaOne was the only PPC platform that had that problem? Oh no not again!!

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 16:07:58
#215 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@Crumb

Quote:
I'm only interested in PPC AmigaOSes as long as they run m68k AmigaOS apps (and MorphOS/OS4 apps).


Yes, mainly that is my interest too. Things like AROS do fascinate me otherwise.

Quote:
Unfortunately, as long as PPC AROS has no m68k compatibility and/or PPC MorphOS/OS4 compatibility it will run much better in any PC.


Sure. Although how hard would it be to make it compatible with MOS so MOS programs could load? In some sense they are rather similar OSes.

The PPC AROS builds could possibly be made compatible with 68k AmigaOS so that a 68k emulator could be built in. Open source 68k emulators exist, so this could be grafted on. At least then, PPC hardware like the Sam could have a new lease on life without an OS4 being able to run. And, PPC could be the only AROS arch that could run 68k apps out of the box. Then, PPC would have something the PC doesn't, a rare opportunity these days.

 Status: Offline
Profile     Report this post  
Hypex 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 16:11:37
#216 ]
Elite Member
Joined: 6-May-2007
Posts: 11180
From: Greensborough, Australia

@number6

Quote:
Some people seem to believe that Moana was the 1st attempt at persuading AI to look at Mac hardware. Not so.
I'm just mentioning this so people don't relate all this to the legal activity.


It was an attempt to persuade Amiga? I thought Moana was done outside of Amiga, Inc.

 Status: Offline
Profile     Report this post  
number6 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 16:35:46
#217 ]
Elite Member
Joined: 25-Mar-2005
Posts: 11540
From: In the village

@Hypex

Quote:
It was an attempt to persuade Amiga? I thought Moana was done outside of Amiga, Inc.


conclude what you will.

Seems exceptionally clear that someone was trying to gauge interest, no?
Done in a similar way as Adam, through proper channels.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

 Status: Offline
Profile     Report this post  
Turrican3 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 17:07:52
#218 ]
Regular Member
Joined: 20-Jun-2003
Posts: 386
From: Italy

@Hypex

Quote:
Escom, they were cheap not improving the design I mean.

Understood now (but please be patient, remember I'm no native English speaker)

Well, to be honest I somehow respect Escom; IMHO it was the last IP owner who actually tried to build custom HW, in line with the Amiga legacy/philosophy. Hey, they even had Dave Haynie on-board. After that we have Jim Collas' dreams and then... well, McBill & Co. Bah!

But talking about A1200's I can hardly blame them: at the time IIRC they just wanted to resume production of well known models, as soon as possible. I guess that if they wanted to improve something this would have taken more time.

Quote:
I wonder if Paula could have run at double speed with a VGA display, she did with sound.

Hmm, IIRC actually it wasn't Paula running at double speed: maximum sound playback frequency was simply tied to video in the Amiga architecture, so with a 31KHz mode like VGA you were able to play samples at double the speed of traditional PAL/NTSC 15KHz modes. Maybe this is not the most technical (and correct!) explanation of how things worked, but I have to clarify that this is what I remember having read many, many years ago in the documentation of a popular sound player (EaglePlayer or Delitracker I think).

Quote:
Even at half speed it still would have been good to use HD disks.

Absolutely.

 Status: Offline
Profile     Report this post  
michalsc 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 19:50:26
#219 ]
AROS Core Developer
Joined: 14-Jun-2005
Posts: 377
From: Germany

@Hypex

Quote:
Okay. Now what I don't understand is how SysV can also be used on x86 ....... I've only read about SysV PPC.


x86_64 SysV ABI
x86 SysV ABI
M32R SysV ABI
MIPS SysV ABI
Motorola 6800 EABI (8 and 16-bit)
PowerPC SysV ABI
PPC64 SysV ABI
Sparc SysV ABI

Some questions? ;)

Quote:
So can GRUB be customised with a boot message?


huh?

Quote:
It does seem funny on the floppy when it says "GRUB loading stage 2..." What happened to stage 1?


Stage 1 is 512-byte long, it says "GRUB laoding stage 2..." and loads it :)

Quote:
I do wonder how close to the original structures like Exec, Gfx and Intuition AROS is?


From C programmer's point of view there are almost no differences. The alignment of data may differ (2-byte alignment of m68k does not have to be valid on other architectures) as well as size (eg. on x86_64 struct MinNode is 16 bytes long and consists of two 8-byte long pointers: mln_Succ and mln_Pred).

Quote:
Much more than simple pointer extensions?


Dunno. There are some places in AROS code where 64-bit types are used explicitly on every architecture (like eg. ata.device which has to deal with TD64 and NSCD 64-bit extensions).

Quote:
Wow, that's a lot! Looks like they did possibly need all those assembler lines in the OS4 kernel.

I guess they did :)

Quote:
I am most interestied in ub2lb. How closely does it follow slbv2?


It does not follow slbv2. SLB is a closed source proprietary code and I know nothing about. Ub2lb uses it's own configuration file, called menu.lst, which follows the spec of GRUB menu file.

Quote:
Could it try to boot an OS4 partition on the Sam440?

It even if you would write proper menu.lst file for ub2lb, it would not know what to do with OS4 kernel actually. Since I do not own OS4 I have no knowledge on loading OS4 and it's kmods. Of course, I would have nothing against adding OS4 load support.

BTW. Even if ub2lb would be able to load OS4, it would fail immediately since the PPC4x0 family of processors is, in case of supervisor mode, incompatible with 604/G3/G4.

Quote:
That last point is interesting, cache incoherency, I though the AmigaOne was the only PPC platform that had that problem?


AmigaOne had a hardware issue. The ArticaS chip used there contains so called "floating buffer" inside. This part of chip is responsible for traffic maintenance between CPU, memory and PCI bus (including AGP). It should maximize the bandwidth. As a bonus, floating buffer was supposed to maintain the cache coherency. It was supposed to be smart enough to merge the "updates" coming from several devices which occur even within one line of cache (eg. cpu updating bytes 0-3 and PCI device updating bytes 4-7 of the very same line of cache). The whole behavior and set of cool features of the "floating buffer" are described in ArticiaS developer manual (which I happen to have ;))). Unfortunately, the real ArticiaS happen to behave differently.

In case of AMCC440, the cache incoherency is a well known fact. Do not forget, that this chip is meant for embedded systems, where every spared microwatt of energy and every hundred transistors removed from CPU core do matter. Here, I just have to deal with it.

Quote:
Oh no not again!


Well, automatic update and flush of caches done in CPU core logic would be much much better of course. Well, it seems I can live with it, even if I need to do massive flushes and invalidate some bits of cache quite often.

 Status: Offline
Profile     Report this post  
billt 
Re: AmigaOS4 runs on MacMini
Posted on 8-Jul-2008 21:20:48
#220 ]
Elite Member
Joined: 24-Oct-2003
Posts: 3205
From: Maryland, USA

@bert64
Quote:
Assuming a license is granted?
You don't need a license to write software for PS3 unless you want to use accelerated video.. Linux runs on it just fine.


You need license from Amiga Inc. and Hyperion to port OS4 to ANYTHING

Last edited by billt on 08-Jul-2008 at 09:21 PM.

_________________
All glory to the Hypnotoad!

 Status: Offline
Profile     Report this post  
Goto page ( Previous Page 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle