Here is a video showing how to install the system on QEMU A1 with firmware using QEMU-3Dfx-Voodoo-3-Port. You don’t need to modify the CD for ATI and SM501 drivers.
Yes, you only need to compile QEMU with patches for that. I still think the easiest is the BBoot method I have on my page. I also mentioned there that you need to redo Kickstart changes and zip after each update. Or if you use sam460ex then you can install with SM501 then replace the ATIRadeon.chip on the installed system.
What mouse pointer issue do you get with latest QEMU 11? I don't see a problem with the startup screen here but maybe it depends on the -display backend so could be something specific to cocoa? Errors with older QEMU versions I don't intend to fix only with current QEMU versions are meant to work but of course still could have bugs as I could not test everything.
How would I know if I don't even know what's the proper API? I had to reconstruct the header file from P96 so I don't know what different versions of AmigaOS 4 had only that it roughly corresponds to older P96 versions. (For which the headers are also removed from their site and not available any more.)
Quote:
If not I might finally dig out some of the ancient backups of the latest AmigaOneXE version of AmigaOS 4.1 I have (probably Update 3 or 4, not "FE" or an update of it!) and use it with QEmu.
The driver is really simple and only implements the few functions listed in the README so I guess it might work with older AmigaOS versions too but I only tested 4.1FE install CD so I guess you'd have to try.
Quote:
You can ignore the overlay/PIP parts, at least in case you are able to implement HW/QEmu accelerated compositing, which can be even faster than overlay/PIP YUV modes for video playback.
Everything else needs improving ati-vga emulation in QEMU first so I'm also going to ignore transparency for now. It should be simple as long as pixman works but unfortunately pixman does not work on all systems and it's also optional in QEMU so that means I should implement the whole thing in fallback too which I might do eventually but not now. The software fallback within AmigaOS seems to work well enough but you might get better results using the "2nd look themes" with less transparency effects.
If not I might finally dig out some of the ancient backups of the latest AmigaOneXE version of AmigaOS 4.1 I have (probably Update 3 or 4, not "FE" or an update of it!) and use it with QEmu
Just for the sake of testing, I found an old version of AOS 4.0 for the A1 tucked away somewhere on the internet. After testing it, I deleted all the files from the drive. The ATI driver for QEMu works. You just need to set the mode. If it works on the old 4.0, it will work on 4.1 too.
I also checked out the QEMU-3Dfx-Voodoo-3-Port by Falke_34 It’s a bit slow but it works (only 24-bit).
I watched your video and looks like this Voodoo emulation has endianness problem with the color palette (seen in the boot splash screen), it has problems with mouse pointer which did not appear on your video with the ati-vga driver. The rendering problems in 24-bit with transparency may be a problem in AmigaOS and not the emulation. I get the same with ati-vga in little endian 24-bit mode and I think AmigaOS just does not handle this correctly. Maybe this is the reason why SM502 driver does not support more than 16bits? In my ati-vga driver I just disabled these modes, the big endian modes seem to work.
Tested with Qemu 9.x—I'm not exactly sure which version I'm using. It works very well. The 24-bit mode seems to be a bit broken, but that's not a problem. What we want is 32-bit
Since you tried with older than QEMU 11.0 this does not matter unless you can reproduce with latest QEMU but how did you even test 24bit mode? If a 32bit mode is available it does not even show up for me in ScreenMode prefs, I can only force it by disabling 32 bit. But maybe if you edit the monitor icon and disable DDC and manually add modes you can get that. In any case 24 bit mode does not have fill accelerated (limitation of real ATI VGA too) so it's not recommended to use and 32 bit mode is preferred.
What mouse pointer issue do you get with latest QEMU 11? I don't see a problem with the startup screen here but maybe it depends on the -display backend so could be something specific to cocoa?
The problem is that a square appears in the top-left corner. From what I’ve noticed, it’s the pointer/cursor. It isn’t rendered properly and a square appears instead. It occurs on QEMU 11, Fedora 43, Linux Wayland. It is possible that this is not entirely a problem with the driver itself. It occurs in GTK and SDL displays. It does not occur with VNC. It also occurs in older versions of QEMU on Linux with X11 and macOS. You can easily check this by disabling ‘guest_hwcursor’. On the welcome screen with the ‘AmigaOS 4’ logo, move the mouse. You will see this square instead of the cursor. The colour may blend into the black background but is slightly lighter. It is very visible against a white background. I have such an AOS4 start-up logo on one installation. The mouse pointer should not appear at all at this point (before the system has fully loaded). It should be hidden/masked/invisible. The problem does not disappear when you click in the window or switch the ‘QEMU’ window to the active one. This is not a major issue. Some people might not even notice it.
edit: The video I recorded yesterday was actually from a Fedora 43 Linux system and QEMU 11
edit2: There’s another way to open it. Restart AmigaOS using the option in the main menu. On the start-up screen, you’ll see the correct mouse pointer in the top-left corner, which was a square on the first boot.
Tested with Qemu 9.x—I'm not exactly sure which version I'm using. It works very well. The 24-bit mode seems to be a bit broken, but that's not a problem. What we want is 32-bit
Since you tried with older than QEMU 11.0 this does not matter unless you can reproduce with latest QEMU but how did you even test 24bit mode? If a 32bit mode is available it does not even show up for me in ScreenMode prefs, I can only force it by disabling 32 bit. But maybe if you edit the monitor icon and disable DDC and manually add modes you can get that. In any case 24 bit mode does not have fill accelerated (limitation of real ATI VGA too) so it's not recommended to use and 32 bit mode is preferred.
I just double-checked: I'm using Qemu 10.2.1. The only thing I did was replace the ati-vga driver in Kickstart.zip. As for AmigaOS 4.1, I didn't create a new board in Sys:Devs/Monitors/—it was already there and was created by AmigaOS 4.1 itself using the SM501 driver.
The problem, as @smarkusg also describes, also exists on macOS and appears immediately upon booting and in the boot log in the top-left corner.
MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
@smarkusg So the problem only occurs without guest_hwcursor=on? I did not try to fix that too much because it has some known problem that wasn't fixed in QEMU yet and also hardware cursor is currently always enabled because otherwise there would be flickering screen due to how it's implemented in QEMU now so it would need fixing that first to enable the function in the driver to allow AmigaOS to disable/enable the hardware cursor. So maybe it's possible that HW cursor is enabled when not yet fully set up by AmigaOS. If otherwise it works I consider this a minor problem and flickering on every mouse pointer shape change (e.g. wait cursor) would be much worse.
@Maijestro I don't see that even in your screen shot so maybe it's only with host side hw cursor. I think your monitor icon is likely ignored as I found it has to match the driver name so it should be Radeon with CMPLENGTH=6 (I have no idea where 17 came from when you did not even enter that many chars in BOARDNAME). Is your ScreenMode prefs an Enhancer replacement? With the default one 24 bit mode did not show up for me but it as I said it should not be used anyway.
balaton wrote:@Maijestro I don't see that even in your screen shot so maybe it's only with host side hw cursor. I think your monitor icon is likely ignored as I found it has to match the driver name so it should be Radeon with CMPLENGTH=6 (I have no idea where 17 came from when you did not even enter that many chars in BOARDNAME). Is your ScreenMode prefs an Enhancer replacement? With the default one 24 bit mode did not show up for me but it as I said it should not be used anyway.
When I look at the screenshot you posted on my computer, you can clearly see the square in the top left corner. When I use my phone, you can’t see it, but trust me, it’s there.
As for the monitor settings, yeah, you might be right—I’ve installed a few enhancer software programs, so the board settings might have been created there. What would happen if I delete the monitor I created? Would AmigaOS 4.1 create a new one on its own when I reboot? You can also see that in the Screen Mode preferences, there’s nothing about ATI-VGA—just the settings for the board.
Otherwise, I could probably copy the original monitor settings for ati from AmigaOS 4.1 Final Edition, which might also work and would match the AmigaOS 4.1 standard.
MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
When I look at the screenshot you posted on my computer, you can clearly see the square in the top left corner. When I use my phone, you can’t see it, but trust me, it’s there.
If I look at it magnified there's some very faint dark gray there. I checked and it's color #070606 so very dark gray against #000000 black background. Maybe you can only see it if you move it over the logo or have your screen brightness way up. If this is the only place it shows and only shows without guest_hwcursor=on then I don't think it's worth trying to fix. It may resolve when mode change is improved in ati-vga and disabling HW cursor can be enabled in the driver.
Quote:
As for the monitor settings, yeah, you might be right—I’ve installed a few enhancer software programs, so the board settings might have been created there. What would happen if I delete the monitor I created? Would AmigaOS 4.1 create a new one on its own when I reboot?
I think you'd see no change without a monitor icon as the driver supports DDC/EDID so modes are read from there. You probably only need a monitor icon if you want to override some settings which is not needed in most cases.
Quote:
You can also see that in the Screen Mode preferences, there’s nothing about ATI-VGA—just the settings for the board.
The prefs does not say card name but since you have no other cards used here I think Board 0 refers to the Radeon driver and the other modes not marked Board 0 are the default built in modes. The modes shown are what qemu-edid generates so this seems to work.
@joerg Quote: Does it depend on anything from AmigaOS 4.1 "FE"?
How would I know if I don't even know what's the proper API? I had to reconstruct the header file from P96 so I don't know what different versions of AmigaOS 4 had only that it roughly corresponds to older P96 versions.
It just depends on the min. library version you are using in IExec->OpenLibrary() calls in your ATIRadeon.chip driver, if you'd use for example V54 for opening expansion.library (for the PCI functions) it will AFAIK only work with AmigaOS "FE" Update 3 (or newer), if you use V52 or V53 it should work with any AmigaOS 4.1 version and when using V50, and you don't use any functions added later, it will even work with AmigaOS version 4.0 (incl. developer pre-releases of it).
The only relevant change from AmigaOS 4.0 to current versions for ATIRadeon.chip, and only theoretically as you don't support it yet, is the compositing function which IIRC was added in AmigaOS 4.1 (not "FE") in the .chip driver API and may not have been available in AmigaOS 4.0 yet. For gfx drivers not supporting it in HW, i.e. all of the about 20 P96 .chip drivers except for the original ATIRadeon.chip, RadeonHD.chip and RadeonRX.chip, the AmigaOS 4.x graphics.library uses a software fallback function.
Quote:
(For which the headers are also removed from their site and not available any more.)
It just depends on the min. library version you are using in IExec->OpenLibrary() calls in your ATIRadeon.chip driver, if you'd use for example V54 for opening expansion.library (for the PCI functions) it will AFAIK only work with AmigaOS "FE" Update 3 (or newer), if you use V52 or V53 it should work with any AmigaOS 4.1 version and when using V50, and you don't use any functions added later, it will even work with AmigaOS version 4.0 (incl. developer pre-releases of it).
I don't have any OpenLibrary calls. It's a chip driver and doesn't do much more than take the parameters from P96 API and write it to the appropriate ati-vga registers where I don't even really have to care about cache or PPC MMIO either as it's only for QEMU where this isn't an issue. Everything else is handled by the default AmigaOS PCIGraphics.card which I haven't replaced so it might work with any version that recognises the Radeon 7000 PCI card which is the oldest Radeon so should work with most versions.
This is just the normal P96 header one can get from iComp with the P96 driver docs but the AmigaOS 4 version seems to differ at the end and in some functions may correspond to older versions than tis header. There's nothing in this about transparency so unless I reverse that I cannot support it but since it's not yet emulated in ati-vga so it's also not a problem yet. The AmigaOS fallback works except for little endian 24 bit mode but nobody should use that anyway.
Everything else is handled by the default AmigaOS PCIGraphics.card which I haven't replaced so it might work with any version that recognises the Radeon 7000 PCI card which is the oldest Radeon so should work with most versions.
The MAI Teron Mini/EyeTech AmigaOne µA1-C (the µA1-I version was never produced) includes a Radon 7000 with 32 MB VRAM, but no BIOS ROM, it's included in the U-Boot binary of it instead, probably a GZip or Bzip2 compressed image. Not sure any more, but the µA1-C might have used AGP instead of PCI for the onboard R7000 chip. The only way to get even worse gfx support on AmigaOS 4.x is using the SM501/502 with QEmu (16 bit only) or on a real Sam4x0 without any PCI(e) gfx card installed, or Classic Amiga emulation (WinUAE) with Voodoo3/4/5: (Warp)3D supported, but like on original hardware only in 16 bit modes, and limited to 4-16 MB VRAM...
On a black background/“boot logo,” you can also see it depending on the monitor. The square is gray and the entire background is black. After restarting, you can see the mouse pointer instead of the square. It’s displayed correctly, but it shouldn’t be there. As I mentioned, there’s nothing significant affecting how it works. It’s behaving strangely, which is why I let you know
On a black background/“boot logo,” you can also see it depending on the monitor. The square is gray and the entire background is black. After restarting, you can see the mouse pointer instead of the square. It’s displayed correctly, but it shouldn’t be there. As I mentioned, there’s nothing significant affecting how it works. It’s behaving strangely, which is why I let you know
Thanks for confirming that. Since you're using a different, very bright startup logo, it's clearly visible. It doesn't affect the driver's functionality, but it doesn't match the original and shouldn't be like that.
MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
@Maijestro There seems to be some other issue here besides the ATI driver for QEMU itself. We found a bug that occurs in the latest version of QEMU as well as in older versions. I checked a few things regarding how ATI (QEMU) works on versions < 11 and... @balaton created patches for ati-vga, and they are available in QEMU-11, which ensure that the driver itself works fully correctly only in QEMU-11. That’s why I understand why he’s pushing so hard to report bugs only on QEMU-11. On older versions QEMU, you may encounter endianness issues with sdl1, crashes in native applications, etc. The fully functional version works correctly only with the latest version of QEMU-11
@smarkusg Yes there were quite a few fixes during QEMU 11.0 freeze and some bigger changes between QEMU 10.2 and 11 in ati-vga so if you find a problem with older QEMU versions it may already be fixed in the latest QEMU version. The problem with mouse pointer showing on boot screen is probably due to the disable/enable of HW cursor not emulated because it conflicts with a work around for another OS driver so this would need fixing the mode changing in ati-vga to work better. I have an idea how but would need some time and couldn't be in this QEMU release, so currently when AmigaOS wants to disable the cursor it will show anyway but this is less annoying than potentially flickering screen at mouse pointer changes that would be the other option now. So this needs some improvement in the ati-vga emulation such as most likely for any other problem you may find as there isn't much in the driver itself so most problems are likely with emulation instead. So it's known to be incomplete and not perfect but probably still useful as it is. I'd be more interested if there are any serious problems that makes it unusable with latest QEMU or if it's any better than the SM502 driver (apart from allowing 32 bit modes).
I think your monitor icon is likely ignored as I found it has to match the driver name so it should be Radeon with CMPLENGTH=6
That's only for Hans' RadeonHD.chip and RadeonRX.chip drivers. Using CMPLENGTH=6 only checks the "Radeon" part and therefore works with both of them.
With the AmigaOS 4.x ATIRadeon.chip it may be BOARDNAME=ATIRadeon instead, and with your version it seems to be BOARDNAME=Qemu ati-vga according to @Maijestro's screenshot. Or maybe BOARDNAME=qemu ati-vga (file name in DEVS:Monitors), I don't know if it's a case sensitive or case insensitive compare.
@joerg No its board name is "Radeon (QEMU ati-vga)" overriding the Radeon 7000 name PCIGraphics.card sets and before I had DDC working I tested it with a monitor icon named Radeon with BOARDNAME=Radeon and CMPLENGTH=6 and that worked and picked up modes from it so I think that should work. But now with DDC you probably don't need a monitor icon at all and would get monitor info and modes via DDC. I've also found before that AmigaOS 4 does not really use BOARDNAME but the name of the monitor icon file has to match the driver name (which may not be the same as the kernel module name but what it sets as the board name). So no, CMPLENGTH does not compare to ATIRadeon.chip file name but to the "Radeon (QEMU ati-vga)" name the driver sets as the board name (or to the names set by PCIGraphics.card which seems to begin with Radeon for all ATI cards). Therefore a monitor icon named Radeon with CMPLENGTH=6 should work.
News feed
OS4 Feedback forum
OS4Depot Feedback forum
AmiCygnix forum
ABC shell forum
AmiKit forum
Cinnamon Writer forum
CodeBench forum
Digital Universe forum
Dopus 5 forum
E-UAE forum
Gnash forum
Ibrowse forum
JAmiga forum
Odyssey forum
OWB forum
Qt forum
SmartFileSystem forum
Timberwolf forum
TouchDevice forum
TuneNet forum
Unsatisfactory Software forum
Donate
