WinUAE 3.0.0 beta series (was 2.8.2/2.9.0)

[roomeo]

hi Guys.

The "distortion" comes when dragging windows..Not sure why this is happening. Any ideas?
Tried to use both driectdraw and direct3d. also fiddled with Tripple/Double buffering.. but with same result..

My config can follow if there are no obvious way to fix this..

Other than this its running aye ok with no lockups..

Great work Toni!

Cheers!

[zouzou]

Is this in AGA mode or picasso (it seems picasso, I guess AGA could not support this resolution)

Nice gfx corruption.
I never got this.

Anyway, I got that kind of messy drawing when I updated to Win 8.1 (in a game), I had to update the gfx drivers and suddently, the game was drawn ok again. No one seemed to complain about this strange drawing on the internet so....maybe it's you chance? (I had nvidia drivers)

[FOL]

Originally Posted by zouzou

Is this in AGA mode or picasso (it seems picasso, I guess AGA could not support this resolution)

Nice gfx corruption.
I never got this.

Anyway, I got that kind of messy drawing when I updated to Win 8.1 (in a game), I had to update the gfx drivers and suddently, the game was drawn ok again. No one seemed to complain about this strange drawing on the internet so....maybe it's you chance? (I had nvidia drivers)

I always get this type of corruption, not that bad though. I have also seen it on real machines. It seems to be down to GFX card / drivers.

[roomeo]

picassoIV emulated, yes.

So its not an emulation issue i presume? Never saw this when ran os4.0 classic native... But that is some years ago, so i might not remember correctly..

Cheers!

[roomeo]

I found solution for this... The corruption only occured when i used 1024x768 16bit. When using 1024x768 24bit BGR it works flawless..

[FOL]

Originally Posted by roomeo

I found solution for this... The corruption only occured when i used 1024x768 16bit. When using 1024x768 24bit BGR it works flawless..

Only problem is, you cant run Picasso IV in higher res as it crashes system.
Guessing was never ment to do what we are trying.
I can get it to run at 1366x768 native but only as 16bit.

[ancalimon]

Originally Posted by Michael Sykes

Can these strange D3D slowdowns be caused by windows side v-sync settings?
Just a wild guess...

You mean Aero? I'll try disabling that but it even happened with fullscreen modes. (Didn't try a full screen mode for a long time)

[Michael Sykes]

Originally Posted by ancalimon

You mean Aero? I'll try disabling that but it even happened with fullscreen modes. (Didn't try a full screen mode for a long time)

No I don't. Just check if v-sync is enabled by default in the gfx card settings.

[ancalimon]

Originally Posted by Michael Sykes

No I don't. Just check if v-sync is enabled by default in the gfx card settings.

ATI only has that option for OpenGL and that's off unless application specifies.

[Michael Sykes]

Originally Posted by ancalimon

ATI only has that option for OpenGL and that's off unless application specifies.

As far as I know Radeon drivers offer Dynamic V-Sync control DVC.
That feature is tinkering with the framerate depending on the monitors refresh rate.

Can't say more than that since I use a NV GTX570.

[ancalimon]

Originally Posted by Michael Sykes

As far as I know Radeon drivers offer Dynamic V-Sync control DVC.
That feature is tinkering with the framerate depending on the monitors refresh rate.

Can't say more than that since I use a NV GTX570.

That's only available if use use the discontinued RadeonPro.

[Toni Wilen]

You can attach your config and logs again but there is no guarantees that anything happens especially if it only happens in some alien operating system.. (=Must test if exact same problem happens using exact same RTG config on OS3)

[thomas]

Blizkick support is broken since Beta 17. Running Blizkick leads to HALT3 immediately. CPU% display still changes. Ctrl-A-A or F12 + Reset hangs WinUAE.

[Michael Sykes]

Originally Posted by ancalimon

That's only available if use use the discontinued RadeonPro.

I didn't use ATI in years but I cant imagine that they removed it.
There must be a selector like "wait for vertical refresh" in CCC's 3D settings....

EDIT: thinking abot it I googled some infos and found that video...
[ Show youtube player ]

[FOL]

Originally Posted by Michael Sykes

I didn't use ATI in years but I cant imagine that they removed it.
There must be a selector like "wait for vertical refresh" in CCC's 3D settings....

EDIT: thinking abot it I googled some infos and found that video...
[ Show youtube player ]

They have removed loads of stuff in newer drivers when using older hardware, .
Im using 2010 drivers cause they discontinued RadeonHD 3200.

[Toni Wilen]

I don't really see what this has to do with WinUAE anymore.

[Toni Wilen]

Originally Posted by thomas

Blizkick support is broken since Beta 17. Running Blizkick leads to HALT3 immediately. CPU% display still changes. Ctrl-A-A or F12 + Reset hangs WinUAE.

I think this is technically a blizkick bug, reset proof code goes to Z3 fast which is supposed to disappear at reset until autoconfig puts them back but this code runs before autoconfig. Previously Z3 RAM was not removed at reset but now that Z3 autoconfig is more flexible (does not anymore require ugly hack to force Z3 addresses), boards are not available until autoconfig has configured them.

I guess it works on real A4000s because all of them have onboard SIMMs (or accelerator board RAM) which have higher priority than Z3 boards and Z3 RAM boards are also quite rare.

[thomas]

Originally Posted by Toni Wilen

I think this is technically a blizkick bug, reset proof code goes to Z3 fast which is supposed to disappear at reset until autoconfig puts them back but this code runs before autoconfig. Previously Z3 RAM was not removed at reset but now that Z3 autoconfig is more flexible (does not anymore require ugly hack to force Z3 addresses), boards are not available until autoconfig has configured them.

Ok. But Reset should not cause WinUAE to hang.

I guess it works on real A4000s because all of them have onboard SIMMs (or accelerator board RAM) which have higher priority than Z3 boards and Z3 RAM boards are also quite rare.

Tried with chipset = A4000, accelerator = CSMKIII and 64MB accelerator RAM. Now it enters HALT5.

[Toni Wilen]

Originally Posted by thomas

Ok. But Reset should not cause WinUAE to hang.

I can't duplicate any hangs. Halted state (real CPU also halts in case of double fault) is better than letting it to execute code from undefined memory space. Only exception is compatible 68000 mode where at least one demo that does it as a protection method..

Tried with chipset = A4000, accelerator = CSMKIII and 64MB accelerator RAM. Now it enters HALT5.

Fixed. Technically also a bug in blizkick which first tries to reset CSPPC m68k CPU using method that can't work (it would only stop the CPU completely which it incorrectly did in emulation, real hardware does not allow it) HALT5 = both CPUs stopped.

[thomas]

Originally Posted by Toni Wilen

I can't duplicate any hangs.

If I press Ctrl-A-A when in HALT3 state, the CPU% display stops moving, F12 does not react and if I click the mouse buttons several times Windows makes the black window gray and says "no reaction" in the title bar.

Technically also a bug in blizkick which first tries to reset CSPPC m68k CPU using method that can't work (it would only stop the CPU completely which it incorrectly did in emulation, real hardware does not allow it) HALT5 = both CPUs stopped.

HALT5 state works as expected: when I press Ctrl-A-A it boots again. And surprisingly it continues as if Blizkick succeeded and didn't stop the CPU.