WinUAE 4.1.0 beta series

[AnnaWu]

Originally Posted by amiga_amigo

Hello - great work on the bridgeboard emulation!

I have a question though - the configfile for my bridgeboard machine used to work fine in beta4. Now, in beta5 the x86 hardfile isn't booting anymore.

I checked the config: in the harddrive properties the hardfile is no longer attached to any HD controller.
I can't select the "x86 AT primary IDE" since it's not in the dropdown list - just UAE (uaehf.device), IDE (auto) and SCSI (auto).
"AT IDE Primary" is enabled in x86 bridgeboard expansions though (as before).

Am i missing something? Is it a bug? Or do i have to create a completely new config file since the bridgeboard related expansions were moved and the old config files are no longer compatible?

Greetings, Michael

It seems, something is wrong on the last official and unofficial Beta version. I have similar problems.

[Toni Wilen]

Yeah, category change broke bridgeboard SCSI and IDE configuration. winuae.7z should already fix it.

[AnnaWu]

Originally Posted by Toni Wilen

Yeah, category change broke bridgeboard SCSI and IDE configuration. winuae.7z should already fix it.

Thanks for the fix, Toni.

[superfrog]

Shouldn't "Soundblaster ISA (Creative)" also be moved to the x86 Bridgeboard Expansions category?

[shaf]

Yesterday I tried creating a new Virtual HD 2.0gb Image from scratch and booted with the install disk but it appears no HDD image was generated. using default a4000 configuration with some added 64mb z3 ram. Can someone else please confirm.
To get around it I used an existing HDD image and overwrote it.

[Toni Wilen]

Originally Posted by shaf

Yesterday I tried creating a new Virtual HD 2.0gb Image from scratch and booted with the install disk but it appears no HDD image was generated. using default a4000 configuration with some added 64mb z3 ram. Can someone else please confirm.
To get around it I used an existing HDD image and overwrote it.

No and not enough information.

No information if it has anything to do with this beta either (it probably does not if it is not mentioned..).

What does "no HDD image was generated" mean? If you mean HDF file, it is really weird roundabout way to check if it exists by using emulator when you can easily check it with Windows explorer etc..

[shaf]

yes using latest beta , explorer doesn't see the file either, i'll try to create a smaller partition size and retest.

[Toni Wilen]

Originally Posted by shaf

yes using latest beta , explorer doesn't see the file either, i'll try to create a smaller partition size and retest.

But you also need to confirm it does not happen with previous official version. If it happens: this is wrong thread. Simple rule

[shaf]

I got it working, it was a USE. User Stupidity Error.

[Leandro Jardim]

- Added quick search text box to Configurations panel. Searches from config file name, description and new tags config entry (config_tags=<tag1,tag2,..>).
- Added quick directory select to Configurations panel. Filters by directory or category. Category is new config entry (config_category=<category name>).
- Moved Link and autoload options under Advanced information. Added Category and Tags fields.

Toni, sorry, but when you made that changes in WinUAE b4 GUI you forgot to put back the AutoLoad checkbox which was quite useful in previous versions. The only checkbox that you put under Advanced Information was the "Ignore Link".

EDIT: Ohh, now I found it, but it is hidden under the Category editbox. Ah, sorry Toni.

[Toni Wilen]

http://www.winuae.net/files/b/winuae_4100b6.7z
http://www.winuae.net/files/b/winuae64_4100b6.7z

Beta 6:

- Video recording from hardware emulated graphics boards recorded only black screen.
- Screenshot from hardware emulated graphics boards always captured visible output even if capture before filtering was ticked.
- Filter panel (D3D) scanlines work now properly in multimonitor mode. (Previously native/RTG selection was ignored)
- Taking screenshot/recording when in multithreaded RTG mode was not thread safe.
- x86 bridgeboard category move broke SCSI and IDE x86 expansions. Sound Blaster and NIC was not moved.
- x86 bridgeboard virtual serial mouse connected serial port was not reset at reboot.
- A2386SX bridgeboard chipset EMS banking emulation was not complete, chipset native EMS driver complained about invalid hardware.
- Creating state file when disk access was active and disk image included timing information (for example ipf) caused buffer overflow.
- Autoload checkbox was hidden under string field in Advanced information panel.
- Added untested ADKCON MSBSYNC disk synchronization mode support.

[Zarnal]

Hi Toni,

There seems to be a technical issue only with the latest betas (WinUae 4.0.1 work fine).

If you enable 68040 or 68060 + data cache emulation (with or without MMU+ FPU) some graphic corruption sometimes appears. Tiny launcher is a good example.

Everything works well with a 68030 + data cache emulation (with or without MMU + FPU).




Thanks.

Kick 3.1- AGA- 2Mo chip + 8Mo fast. Cycle exact full (chipset). CPU fastest possible. New UAE Indirect. Dx9

Edit : add small adf of tiny launcher.

Procedure :

Insert disk in df0:
And press Enter when tiny launcher is loaded.

[Toni Wilen]

Exactly which beta introduced it?

[Zarnal]

Beta 6


I've tested betas 1,2,3,4,5 they are ok (with my hdf).

[JimDrew]

Did you try this on a real 040 or 060 system? Cache issues like this were pretty common with 040/060 systems. It's why Commodore made CacheClearU().

[Toni Wilen]

Originally Posted by Zarnal

Hi Toni,

There seems to be a technical issue only with the latest betas (WinUae 4.0.1 work fine).

If you enable 68040 or 68060 + data cache emulation (with or without MMU+ FPU) some graphic corruption sometimes appears. Tiny launcher is a good example.

Everything works well with a 68030 + data cache emulation (with or without MMU + FPU).




Thanks.

Kick 3.1- AGA- 2Mo chip + 8Mo fast. Cycle exact full (chipset). CPU fastest possible. New UAE Indirect. Dx9

Edit : add small adf of tiny launcher.

Procedure :

Insert disk in df0:
And press Enter when tiny launcher is loaded.

Config file is required too. Does the glitch also appear in screenshot? (with "take screenshot before filtering" ticked?)

EDIT: This is quite strange because there is no CPU related changes, not since b1..

[Zarnal]

Yes, the glitch appears in screenshot with " take screenshot before filtering " ticked.

Note : WinUae.7z seems ok. I don't understand.

@Jim Drew

No, my real A1200 is a stock machine + 8 Mo.

[Toni Wilen]

Originally Posted by Zarnal

Yes, the glitch appears in screenshot with " take screenshot before filtering " ticked.

Note : WinUae.7z seems ok. I don't understand.

Does both 32-bit and 64-bit beta 6 have same problem? (If you can run both)

[Zarnal]

Yes, same issue with Winuae 64bits beta 6.

[Toni Wilen]

You don't have MMU emulation enabled, if datacache is enabled without MMU, it will cache everything, including chip ram which isn't really going to work if blitter is also used. Same would happen with real hardware if chip ram is marked as cacheable. This was the main reason for data cache emulation, ability to debug cache related problems using emulation. It is not very useful for games or normal WB use, it only slows everything down.

I can duplicate this with older versions and current version. (You sure you really have identical config?)

Anyway, I'll add workaround because this kind of config that can't exist in real world is mostly useless in this case: if no MMU emulation enabled and data cache emulation is enabled, chip ram is always forced not data cacheable.