@olsen

I'm not sure how much impact MKBADDR has in real world code.

I believe that it would be OK until you subtract to determine the distance between two BPTRs and one is located in the upper 2GB.

I'm not sure how often this is done, but considering the primary task (IMHO) of at least Amithlon is running legacy software, combined with the lack of updates to said software, it could be a major issue.

There are a few other places where AOS has problems, such as KickTags, but until you have chip ram located above 2GB those are non-issues.

@intangybles

Quote:

intangybles wrote: @SnkBitten Thanks for the reply, I should have been more exact on the model, sorry. Rushing to find the little free time available as usual. Its an R51 (1829EM2) Pentium M (1.6 Ghz) Chipset - Intel 855PM Graphics - ATI Mobility Radeon 7500 Not looked at the exact chips at this time but In my impatience I had a go at install and I got it base installed, not had time to go further yet but Kernel 3.10 seems ok, no trouble partitioning drives Graphics is 1024x768 and 16M colours.. This took several goes at adjusting video and memory settings. I will let you know how I go from here... Thanks again..

It looks like a good candidate then, the ATI Mobility Radeon 7500 has been found by others to be supported in kernel 3.10 and above.

@kolla

Quote:

Are there still any real advantages in using Amithlon over using for example FS-UAE on Linux?

I've been wondering the same thing, not knowing much about Amithlon, other than it's an impressive technical achievement. If FS-UAE on Linux does about the same thing, I can't see any reason to use outdated software which almost certainly has security issues.

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison

Once you use Amithlon, it's really hard to go back to just FS-UAE.

The mouse is laggy and while it's fast compared to a real Amiga, it not even close to Amithlon.

If you want to use it as a real OS like OS4 or MorphOS, FS-UAE is a poor substitute.

IMHO, FS-UAE is actually a step backwards from UAE. It has only seemed to get more bloated and slow over time. More features, sure, but at a cost.

@Wumpus

I've not noticed any mouse lag with FS-UAE, but maybe it's there and I got used to it.

Other than mouse lag, what other benefits does Amithlon have over FS-UAE?

Update

I just fired up FS-UAE, and I see what you mean. It's not really mouse lag, at least not on my system; it's more an issue of ghost images of the mouse pointer. Sometimes there are two or three of them visible at once. This doesn't bother me too much, since the "real" pointer doesn't seem to have any lag. It would of course be better without the ghost images.

Last edited by bison on 01-May-2018 at 07:45 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison

With vsync turned on it's comical, like moving the mouse with a rubber band, but with it off there is still a definite delay in response.

It's hard to say exactly what's wrong. In a lot of ways it just doesn't quite feel like you're using the computer itself, even full screen with the Amiga screen set to the exact monitor and xorg screen mode.

When you run Amithlon though it feels like it's a 1:1 connection like any other OS and everything feels snappy and responsive.

Until I saw it work right, UAE wasn't so bad. Now I almost regret it!

I mostly run FS-UAE on my Macbook and Mac Mini at work, under macOS, and I don't see any of what you guys describe.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Maybe it's better on Mac than Linux? I'm not sure where it's developed.

Also, have you run Amithlon (not VESA, but accelerated RTG) for an extended period of time on similarly fast hardware recently?

I didn't realize it until after I had something to compare it to.

@Wumpus

Quote:

Maybe it's better on Mac than Linux?

It's probably an issue with the graphics driver, not the OS. I'm running on Kaby Lake integrated graphics.

_________________
"Unix is supposed to fix that." -- Jay Miner

@SnkBitten

Yes, basic install went well in the end, RAM setting was critical and VESA does not play well with this beast - had to not use the vesa option and away it boots. After the normal 3.1 drivers install and Amithlon updates it seems to be loading picasso96 fine, you get the picasso stage flag on boot.

A question for you if you don't mind, its been a long time since I installed Amithlon but I never remember updating Picasso from the pre-loaded version - I see you have added this to your tutorial - excuse the stupid question here - but how do you run the installer to update? I have had no luck, it keeps saying no installed version found (or the like) - my brain is still a bit rusty on Amiga OS, I have only just restored my old A2000 and thats been a strain enough!

Thanks in advance!

@intangybles

Copied from the install guide. It's using the expert option. I've not had it not install, so I'm not sure what it's doing differently on your setup. Make sure you pick Expert mode, that is very important.

*Picasso96Install
I realized I never covered updating the built-in installation of Picasso96 until doing a new Amithlon install recently and this is a key piece regardless of a hardware accelerated supported video card or not.
This is very specific in what to install. I’m using Picasso96 2.1e so download it and extract it to RAM: Run the Setup from the Picasso96Install folder and make sure to choose “Update an installed version”, then make sure to choose “Expert User” as we want to specifically pick what is being installed. Choose ONLY the following selections:
Libraries
24bit picture.datatype
Picasso96Mode
Picasso96 Variable Setter (PVS)
Continue installing with until completed then browse the Picasso96Install/Devs/Monitors/ and copy the ‘powerfb’ monitor file to your SYS:Devs/Monitors/ folder overwriting the one currently there. Reboot AmigaOS and then go to Prefs and launch Picasso96Mode. From the menus you can select the “attach settings to” option, and if your graphics card is supported, select your card, mine is “nVidia-1” as I’m using an NVidia FX5500 PCI card which is supported in the kernel. You’ll have to tweak the settings and that is covered in Part 2 of the Installation Guide. If your card is not listed it will most likely show “powerfb” as the current board and “VESA-1” as a choice. I’m not sure if you should try to manipulate the settings if using powerfb or VESA-1. I have only made changes for nVidia-1 and my monitor’s sync/refresh rates.

@SnkBitten

Thanks for that, yes I was trying to follow the steps you added, can you tell me where is the correct target for the install? I think that is where I am going wrong? is it SYS: as thats what I have been trying without luck?

Thanks again.

@intangybles

Duh! This was recently pointed out to me by someone else and I just need to make the addition/correction in the guide.

It's looking for the P96 install directory......so point it to the folder you are running the installer from. It's looking for it's drivers to copy to the appropriate places (which you can't change). Example......If extracted to ram: and run from the p96 folder there, point it to ram:p96

@SnkBitten

Thanks again for the assistance, the brain cells are starting to warm up now!

Well, its been reasonably successful, not done anything productive with it yet but its running well with everything operating that is supported by the 3.10 kernel.

Picasso96 (Updated)
Sound (Kernel AHI)
USB

May even dig out the old ThinkCentre, which has had Amithlon installed on it for years and see if it still works!



And the story continues....

Last edited by intangybles on 03-May-2018 at 07:12 AM.

@intangybles

Nice, glad I could help.

Have you tried networking yet?

@SnkBitten

Hi, The system reports that it has an Intel PRO/100 VE Network card, which I am assuming is not supported by Kernel 3.10

In all honesty, networking is not really a great requirement for how I want to use this beast but of course being a laptop it would normally be a must have!

USB was my main want, my restored A2000 now has a scsi2sd card in it and I have a USB cable connected to it, gives me direct mountable access to the Amiga's virtual scsi drives - works a treat!

I understand that Kernel 4 has expanded Intel card support but I have not ever successfully setup Kernel 4 on the other 2 Amithlon machines I have had over the years, I never could get it to boot on the ThinkCentre machine - which I did check - and is still booting and running, although it did have a dead CMOS battery.

The 3.1 seems to be operating well, so i will see what else I can get installed and running on it! I am inspired by your F18 demo!

@intangybles

No problem, kernel 3.10 works well and if it suits your needs/wants then there is no reason to change.

Kernel 4 needs the init string to change from 'root=/dev/ram0' to 'root=/dev/ram1' to work properly. I think it was stated as needed for bigird.gz to boot by the kernel dev but I've found it needs to be set to that for smallird.gz as well to allow the modules to be inserted.

@umisef

Quote:

For mixed-mode programming to be anything other than a complete nightmare, the first two have to match.

Okay. What about other concerns such as the VBR? And where ever x86 stores its own vectors?

Quote:

Unfortunately, in a linux x86 system, that's where the kernel expects to be mapped. So for a 1:1 mapping, the PCI space has to be moved.

This couldn't be configured so the kernel could be moved? Like how a Linux kernel can be massively configured today.

Quote:

Exactly, 1GB was almost unheard of amounts of actual memory, so moving the PCI space to the 1GB to 2GB range (the 2GB to 3GB was used for the actual executable, in the only bit of memory which wasn't 1:1 mapped) was a perfectly sensible thing to do.

Makes sense. Even on an actual Amiga 1GB would be hard to come by in one continuous address space. And on the AmigaOne 1GB was rare when it cane out and 2GB even rarer years later.

Quote:

These days, my approach would be very different --- use a 64 bit kernel, with the kernel safely tucked away a long way beyond the 4GB, and leave the PCI where it is (and where space for it has been set aside by the BIOS/firmware). But Amithlon was 18 years ago, so that wasn't an option.

A 64-bit CPU and kernel would be a different approach to emulating a 32-bit OS. But would this need 4GB RAM for headroom? Can't believe it's been so long.

@olsen

Quote:

They did, by virtue of the 68020 being available in 1984/1985, after Apple had already committed to the humble 68000 and shipped the Macintosh. Amiga, being "late", still had the chance to avoid getting locked into the 24 bit address bus properties (although whoever designed AmigaBASIC didn't get the memo).

I have a theory that MS took AppleBASIC, since it was on 68K, and ripped that to make AbominationBASIC bundled with the Amiga.

You may say this is circumstantial evidence, as they say on TV, but on my A1200 BASIC failed to load files. And I knew they were there. So I ran the code through a debugger and found that every string address it looked at had the top 8-bits chopped off. Not only wasting CPU time doing nothing of value but corrupting the pointer. So all pointers are wrong and could crash. My A1200 had RAM above the 24-bit area in 32-bit space when I upgraded it. The Mac has pointer tagging. Addresses would be masked to 24-bits. I put those together and am very suspicious they they simply ported it from the Mac version and left in the pointer hacks.

I wrote a patch later on or some kind of hack to fix it on my machine.

Quote:

The use of InitStruct() with 24 bit APTRs was nipped in the bud, but AllocEntry() somehow got away

I'm not aware of anything using InitStruct() bar the system itself, since most programmers like to set up the library base with startup code. And with AllocEntry() even less aware. In any case name fudges like OldAllocEntry() or NewAllocEntry() to fix this stuff was popular.

The early semaphores is probably the most well known example. But also confusing as it looked like one set of functions was replaced by a completely different set of functions.

@bison

Quote:

Other than mouse lag, what other benefits does Amithlon have over FS-UAE?

For one thing, FS-UAE is an image based emulator. It likes to use images. So is limited compared to UAE. Amithlon uses the direct hardware on your machine. Put a CD in the CDROM Drive and it can read it directly.