WinUAE 2.6.0 beta series

[Mad-Matt]

I use ide0 (scsi.device) for my hardfiles to appear like on an actual Amiga. But uaescsi.device for mounting cdroms. idefix is used in most of my configs.

[Toni Wilen]

Confirmed, 2 IDE drives seems to cause issues, fortunately reads are also broken = nothing should get corrupted.

[Toni Wilen]

http://www.winuae.net/files/b/winuae.zip should fix both problems.

[Mad-Matt]

Originally Posted by Toni Wilen

http://www.winuae.net/files/b/winuae.zip should fix both problems.

Confirmed, the above issues are now solved.

030 MMU mode, CPU Fastrom command appears to work.

When using 040/060, CPU Fastrom doesnt appear to work, is this still normal for the current state of mmu emulation?

I am using the mmulib 040/060 libs. (loaded by setpatch before cpu command). The mmulib fastrom command is also not working

[Toni Wilen]

Originally Posted by Mad-Matt

When using 040/060, CPU Fastrom doesnt appear to work, is this still normal for the current state of mmu emulation?

What does "doesn't appear to work" mean?

[Mad-Matt]

Originally Posted by Toni Wilen

What does "doesn't appear to work" mean?

After Having typed 'CPU FASTROM', The output doesnt show fastrom as installed.

Ive attached screengrabs of 68030 and 68060 (same happens with 040).

EDIT: MuFastrom is working now (didnt in Beta1)

[Toni Wilen]

http://www.winuae.net/files/b/winuae_2600b3.zip

Beta 3:

- Programmed display modes disable normal DMA start/end display limits. (NetBSD AGA console screen)
- ATAPI CD emulation compatibility improved, should fix detection problems.
- IDE HD emulation returned corrupt data if 2 drives enabled at the same time (b2).
- CD image mounter crashed if CD image selection was empty. (b2)

MMU status:

68030 MMU works the best, may still have some rare issues left.
68040 MMU has unknown bug(s). (NetBSD hangs etc..)
68060 MMU seems to have same problem (both are quite similar so nothing surprising) and non-locked read-modify-write status is not emulated, any instruction that reads and then writes to same address has special bus error status bit setup. This is not yet emulated, it is annoying to handle properly.

[Neil79]

OMG!

Eye of the beholder 1 and 2 AGA versions are working perfectly with BETA 2.6.0 BETA 3 !, they never did for me with previous ones. Always complained about MMU or loading up and stuttering all over the shop.

Nice work Toni

[neozeed]

I know this is hardly a pressing issue, but NetBSD 1.0 locks when you try to bring the ethernet adapter online..

Still, again pretty amazing work!!!

[Toni Wilen]

Originally Posted by Neil79

Eye of the beholder 1 and 2 AGA versions are working perfectly with BETA 2.6.0 BETA 3 !, they never did for me with previous ones. Always complained about MMU or loading up and stuttering all over the shop.

Hmm.. My guess is that MMU emulation is now active (and you have MMU setting in s:whdload.prefs) which causes emulation to slow down just enough to make the game work better

Originally Posted by neozeed

I know this is hardly a pressing issue, but NetBSD 1.0 locks when you try to bring the ethernet adapter online..

Does 1.1 installer do the same? (I have test image but I haven't tested a2065 emulation)

Try -a2065log2 -log command line to enable full a2065 emulation logging.

[neozeed]

I suspect this is what you are looking for:

A2065: INIT.
A2065: Initialization block2:
.00.00.00.80.10.28.5D.00.00.00.00.00.00.00.00.00.20.80.00.80.A0.80.00.40
A2065: 0000 008000 0 0 16 4 008020 0080A0 00:80:10:28:5D:00
uaenet_win32 initialized
MMU: page fault (logical addr=0003ED39 SSW=0351 read=1 size=1 fc=1 pc=0000b1d0)
MMU: page fault (logical addr=0005003C SSW=0311 read=0 size=1 fc=1 pc=00039bfe)
MMU: page fault (logical addr=0003EC5A SSW=0301 read=0 size=4 fc=1 pc=0000c0b6)
setmouseactive 1->0 showcursor=1 focus=1 recap=0
releasecapture
A2065<!DST:FF.FF.FF.FF.FF.FF SRC:08.00.45.00.00.8A E=0800 S=152
A2065<!DST:FF.FF.FF.FF.FF.FF SRC:08.00.45.00.00.8A E=0800 S=152

Also with scsi ID 6, NetBSD 1.1 freaks out thinking that the scsi disk is on all possible ID's.. but even though the ethernet doesn't work, it doesn't lock like 1.0 ..



And if it matters here is my HD image & Config ..

[Toni Wilen]

Originally Posted by neozeed

I suspect this is what you are looking for

This does not show any obvious problems. I'll test the image soon..

Also with scsi ID 6, NetBSD 1.1 freaks out thinking that the scsi disk is on all possible ID's.. but even though the ethernet doesn't work, it doesn't lock like 1.0 ..

This is already fixed, check other recent MMU/SCSI related threads

[neozeed]

Originally Posted by Toni Wilen

This is already fixed, check other recent MMU/SCSI related threads

Awesome, dude you rock!

[Toni Wilen]

Originally Posted by neozeed

HD image

How do you boot it? It only boots to WB = I immediately lost interest completely

[mark_k]

If it's any help I will upload a NetBSD 1.0 installed HDF image shortly.

On that subject, the NetBSD 1.0 installation process requires using a floppy disk as the initial filesystem. As the kernel reads from the disk, WinUAE reports this several times:
doreaddma() fifo overflow detected, retrying..

Would that indicate some problem with floppy drive emulation? I noticed that sometimes the NetBSD 1.0 installer shows a different error message, which could be explained by the data read from floppy disk only being corrupted some of the time.

[Toni Wilen]

Originally Posted by mark_k

If it's any help I will upload a NetBSD 1.0 installed HDF image shortly.

If it shows A2065 hang

On that subject, the NetBSD 1.0 installation process requires using a floppy disk as the initial filesystem. As the kernel reads from the disk, WinUAE reports this several times:
doreaddma() fifo overflow detected, retrying..

I think this is NTSC-only issue I haven't really bothered to examine too much..

[mark_k]

I tried running the NetBSD 1.0 installer, but possibly due to the floppy problem, after rebooting into multiuser mode the system hangs at the login prompt. I'll retry things with an emulated PAL machine and see if the installer can complete then.

If you want to look into the floppy problem, I have uploaded a pre-installation HDF to http://www..com/file/wKboQhEj...installta.html. It's about 15MB compressed.
Download the inst-10.fs floppy image from http://ftp.netbsd.org/pub/NetBSD-arc...ies/inst-10.fs. Load the attached config and put inst-10.fs into DF0:. Boot the HDF and type gobsd ASKROOT. At the root device prompt enter fd0.

If you let it boot several times you should see different error messages.

The NetBSD installer sets NTSC mode, so to avoid the floppy bug you have to set chipset type to OCS PAL.

[Toni Wilen]

Try turbo floppy speed, it won't trigger this problem and I don't think this loader needs accurate emulation.

[mark_k]

I re-did the installation using emulated PAL OCS machine and there were no fifo overflow messages. I have uploaded the installed HDF to http://www..com/file/3CEJIst3...alled_A30.html. ~23MB file.

Using the config I attached above, boot it and type gobsd MULTIUSER. For me, with A2065 enabled it hangs at the login prompt. (So that problem probably wasn't related to the floppy bug.) With no A2065 you can login.

Notice that on the initial boot there are some messages about filesystem checks. I don't know whether that's normal on the first boot after installing NetBSD 1.0, or if it indicates some issue with WD33C93 emulation.

neozeed: can you check out this HDF? I don't know anything about configuring networking in NetBSD and couldn't test it anyway since I'm using Wine.

[emufan]

mark_k: out of curiosity i want to give it a go - want me to make an account. are there no normal filehoster around anymore?