WinUAE 2.3.1 beta series

[Toni Wilen]

2.3.1 beta series.

Re-recorder discussion goes here: http://eab.abime.net/showthread.php?p=710917

Quick rules:

Allowed posts in this thread:

- bug reports but make sure bug is NOT in previous offical version. (Old bug -> separate thread)
- comments and suggestion about new features in betas or changed feature in beta. (missing feature or old feature which has not been changed in latest beta series -> separate thread)

Everything else will be deleted, no questions asked. (includes offtopic posts, discussion about pros and cons of public beta testing etc..)

Misc notes:

- "not working" is not a bug report -> instadelete
- always include name of the program. "Everything" or "most games" fail is not a bug report!
- always use quickstart settings. Non-QS setting reports will be ignored unless there is reason why quickstart isn't good idea but still test using quickstart just to confirm! (unless it needs some HD highend config)
- always check logs (both winuaebootlog.txt and winuaelog.txt, don't forget to enable logging in misc-panel) for possible error or warning messages.
- remember to include PC specs + Windows version (if asked)
- graphics error -> include screenshot.
- disk based program -> include CRC32 (mouse over disk history arrow down button will show CRC32) of the image(s) (or SPS ID or full tosec name if from latest set). Plain name of the game is NOT generally enough! (usually there are different cracks, trainers etc..)

As usual, betas may destroy your PC and your sanity

This thread is for general discussion (bad/good/stupid idea/whatever, no trolling or flaming) about public beta testing.

Beta test related PMs or emails will be generally ignored.

[Toni Wilen]

http://www.winuae.net/files/b/winuae_2310b1.zip

Beta 1

!!!!!! NOTE: Re-recorder thread is here: http://eab.abime.net/showthread.php?p=710917 All posts related to new input recorder in this thread will disappear. !!!!! ->

- directory filesystem ACTION_LOCK_RECORD and ACTION_FREE_RECORD dos packets implemented (very rarely used)
- CDTV CDA playback didn't work if play ending track was set to last track (Prehistorik CDTV)
- reset CD support internal flags when reseting or loading new configuration
- save chipset extra setting to statefile, previously state restore always restored full advanced chipset configuration and reset chipset extra setting
- 3-stage 68020 cycle-exact prefetch pipeline (not much difference compared to old one)
- do not emulate very small audio periods exactly in non-cycle exact modes, this combined with very short samples used huge amounts of CPU power (used by some stupid audio routines) (2.3.0)
- when restoring state, sound was initialized after first frame, could have caused glitches or missing sound if audio registers were modified during first frame
- mouse counters (JOYxDAT) were not restored from statefile, fixes random statefile restore mouse jump
- VPOSW ECS/AGA-only LOL-bit (long line) is hardwired to zero (can't be modified) in standard PAL mode
- SCSI emulation START/STOP UNIT translated to physical drive load/eject command
- combined volume and period modulation was broken in 2.3 audio update (probably never used in real world..)
- added "gamepad" device type to gameports panel, has two differences compared to "joystick", 3 buttons mapped and pullup resistors emulated (most Amiga compatible gamepads have pullups). It is not possible to have pullups automatically in joystick mode because it can break some badly programmed single-button games, some detect always pressed button in this situation.. (fixes Aladdin and other games that only have working 2nd button if buttons have pullup resistors)
- allow scanning of plain roms (no archives) in winuae root directory
- A2065 statefile support (autoconfig location saved), max allowed transmit size was too small, big packets were silently dropped, random lost transmit interrupts, broadcast packets being ignored if NIC chip hardware multicast filter was disabled
- change emulated board filesystem and Z3 board IDs because of ID conflicts with QuickNet Z2 board (2011/2)
- set default filter to 1x (was FS) and enable new "default" scaling method that only scales if window size is set to "large enough" or so small that image wouldn't fit, always scales in fullscreen modes.
- "autoresolution" checkbox added to Display panel, automatically selects lowest used resolution setting (lores screen = lores/nondoubled, lores and hires at the same time = hires/nondoubled etc..) Positioning may not work correctly yet, glitches possible.. Enable some autoscaling settings if you don't want changing display size..
- do not attempt to steal focus in magic mouse mode if WinUAE is not active, it only messes up focus state
- A500 power led fade tricks should really work now
- less accurate CIA register access timing in JIT modes (not really much point in accuracy..)
- "re-recorder" implemented (completely rewritten combined input/state recorder). More information coming later.

- many statefile related changes for input/state recorder compatibility, may break something else..
* WARNING: statefiles created with betas may not be compatible with future versions!
* save current track's data and density in statefile if disk drive motor was active (regenerated
raw track data/density after state restore may not be identical if using ipf or fdi images)
* state save/restore while disk DMA is active is now fully supported
* save strings in utf8 (should have been done long time ago during unicode updates..)
* useless DMAC chunk was saved in non-CDTV configurations (2.3.0)
* CD state chunks were saved even if selected configuration didn't have any CD drives (2.3.0)
* 68020 cache and prefetch pipeline saved and restored
* debugger memory watchpoints saved and restored
* save active (very rare) interrupt delays (interrupts that have been triggered but have not yet been noticed by Paula)
* mid-instruction CPU state save support, restored state CPU state is now exactly same as when state was saved (68000 CE only, 68020 CE later), previously current instruction was simply restarted (this was really complex and non-trivial task, it is only easy if you accept much higher CPU usage. Which isn't good idea...)
* full blitter statefile support in cycle exact mode, state capture/input recorder only currently. This saves blitter emulation variables, not real hardware state (internal statemachine state, latches, counters etc..) because internal structure is unknown. State capture only because it requires 100% identical state after restore (normal blitter state forces current blit to finish). This is not enabled in normal state saving because it is very difficult to keep state saving compatibility between versions when state is hardcoded to current emulation implementation.

[Maren]

This build seems to have problems running the "Guardian Dragon II" demo by Kefrens when CE is on. It reaches a point (somewhat early, during the intro) where it just burns a picture on the screen while looping the latest second of the background music forever. Basically demo playback's halted, yet SND/CPU/FPS values remain stable.

This happened with every version I tried, .dms and .adf, and each disk "froze" at a slightly different time.

GuardianDrago1.dms (CRC32:25F56DC4)
GuardianDrago2.dms (CRC32:A9AD16A0)

Needless to say, they all work ok with CE off and 2.3.0 remains unaffected. Default config was used for the test.

Get the example savestate for the .dms version here: http://filesurf.ru/210577/gd.uss.html

P.S. kudos for fixing the Zool prob

[Toni Wilen]

Originally Posted by Maren

This build seems to have problems running the "Guardian Dragon II" demo by Kefrens when CE is on. It reaches a point (somewhat early, during the intro) where it just burns a picture on the screen while looping the latest second of the background music forever. Basically demo playback's halted, yet SND/CPU/FPS values remain stable.

Too much CIA emulation optimization..

[dlfrsilver]

ok, i have found a way to get out of the freezing !!!

When it happens, i break in with my Action replay 3 under winuae.

Then i type X to exit, and i come back in the demo and it continues to run.

[Maren]

The .ipf version of the game "Zoom" crashes this beta on startup. Nothing much to say except 2.3.0 doesn't crash, the default config was used and regular .adf's work just fine.

Zoom.ipf (CRC32:4E206F8A)

Get the crash dump here: http://filesurf.ru/210675/winuae_231...01245.dmp.html

Regards.

[Toni Wilen]

http://www.winuae.net/files/b/winuae_2310b2.zip

Beta 2:

- CIA timing issue in some situations (b1)
- some 68020+ exceptions had wrong return address (b1)
- do not attempt to save log window position if it is minimized
- do not set vblank bit during state restore, old code that broke interrupt delaying and rerecorder timing

[Maren]

Does "CIA timing issue in some situations (b1)" have anything to do with "Too much CIA emulation optimization"? just curious, given that the aforementioned demo is still behaving the same way.

Something else I just noticed: zoom.ipf isn't the only ipf image to crash WinUAE here, as a matter of fact, all of them do, I just didn't know as I rarely run games.

[hexaae]

Originally Posted by Toni Wilen

http://www.winuae.net/files/b/winuae_2310b2.zip

Beta 2:

- CIA timing issue in some situations (b1)
- some 68020+ exceptions had wrong return address (b1)
- do not attempt to save log window position if it is minimized
- do not set vblank bit during state restore, old code that broke interrupt delaying and rerecorder timing

MAPROM emu doesn't work anymore:
BlizKick failed with returncode 20
in StartupSequence...

[Toni Wilen]

Zoom works fine here (now, it was broken in b1). Hmm..

EDIT: currently only SPS 4.1 library (recently released) works correctly. Fixed in next beta.

[Retroplay]

Originally Posted by Maren

Something else I just noticed: zoom.ipf isn't the only ipf image to crash WinUAE here, as a matter of fact, all of them do, I just didn't know as I rarely run games.

BC Kid, ipf 0435 also crashes beta 2.

[mr.vince]

Originally Posted by Toni Wilen

EDIT: currently only SPS 4.1 library (recently released) works correctly. Fixed in next beta.

Was it? I think it's the one you had for testing... would need to prepare a public package and put it online...

[Toni Wilen]

Originally Posted by mr.vince

Was it? I think it's the one you had for testing... would need to prepare a public package and put it online...

I thought it was already released. Someone was talking about about new library being available for download but I guess I misunderstood it

http://www.winuae.net/files/b/winuae.zip fixes it. Real beta coming later.

[mr.vince]

We'll try and make it public soon, just wanted to try it first. You were the only one atm that could do this.

Since we now have the new beta of WinUAE, we can try some things and then push it out of the door.

[LocalH]

Just for the record, I posted about the forthcoming new input recorder over at TASvideos. If anyone can help Toni make the rerecorder rock solid, it's those guys (maybe he's already been in touch with them? =P). Please feel free to move this post (rather than delete it) after the actual input recorder thread has been established, I just wanted to go ahead and get a dialog going between the two communities, and let the guys over at TASvideos go ahead and get their hands on the 2.3.1 beta.

[Toni Wilen]

There is very good reason why no talk about rerecorder yet: it is too early for public testing. It still in private testing (it already works, looking for last few glitches)

Any so called "bug report" I see will only make me annoyed. Especially if I see even one so called "recording" without using very specific methods to record it. It has very strict requirements which will be explained soon.

Don't do it if you want to see finished rerecorder.

[LocalH]

Originally Posted by Toni Wilen

There is very good reason why no talk about rerecorder yet: it is too early for public testing. It still in private testing (it already works, looking for last few glitches)

Any so called "bug report" I see will only make me annoyed. Especially if I see even one so called "recording" without using very specific methods to record it. It has very strict requirements which will be explained soon.

Don't do it if you want to see finished rerecorder.

I merely meant this to bring those who are the most experienced at using rerecording functions over here to help you make the rerecorder the best possible. I already advised them that this is early days, not to shit up this thread, and not to use the rerecorder yet for published movies, but only for developing TAS runs, or for playing around with and getting used to it. These guys are high-quality and won't just be posting random bug reports, I will ensure that they know how to file proper bug reports with you should they come across one, when the time becomes necessary.

Basically, just trying to set up the underlying dialogue here so that when proper public testing begins, you'll already have some highly qualified people willing to test and strengthen the rerecorder.

As a courtesy to you, I will post a copy of your statement over on TASvideos, just to let them know how you plan on handling it, and make sure that there are no useless "bug reports" or anything. I've already ensured that they know that the rerecorder (as a branch of the new input recorder as a whole) will ONLY work in cycle-exact mode (especially for TAS purposes), and not to rely on the way things are until the recorder is ready for public consumption. These guys sometimes share recordings amongst themselves, but only for TAS development purposes. I can assure you that noone will be coming here and posting useless bug reports with an input file recorded with JIT/P96/immediate blitter/any of the other numerous things that can desync an input movie), nor will they be using WinUAE to publish runs until it's ready, essentially.

[Toni Wilen]

http://www.winuae.net/files/b/winuae_2310b3.zip

Beta 3:

- ipf support only worked with not yet released library version (b1)
- use original autoconfig IDs if maprom is enabled (blizkick detection checks for IDs, update probably needed in future)
- stop blitter linedraw if bltcon0 is accessed while line draw is active (workaround for Boundless Void / Nemol)
- Direct3D + LARGEADDRESSAWARE fix (possible fix for D3D blank screen problem?)

[geoslake]

Still got blackscreen in d3d if i use any of of the "complex" filters.
I use bilinear + png mask so i dont really care though.
Btw thanks for supporting Mame mask files, its amazing and i dont understand why all emulators dont use that.

Cheers

[geoslake]

Ok, I retried this morning and got black screen in d3d (any mask or filter setting).
I switched to ddraw (that works flawless), then back to d3d and then d3d worked !!!
I have a radeon 5750, catalyst 10.9, latest dx from 6/2010.

Cheers