WinUAE 1.5.0 beta series

[Toni Wilen]

Read the sticky above this one before replying as usual.

Note that some big changes will be implemented in future betas. I only released this before change log and number of bugs gets too big

http://www.winuae.net/files/b/winuae_1500b1.zip

Beta 1:

- MSVC 2008 compiled. Does not run on Windows 98/ME anymore.
- debugger sprite dump command added (sp <addr> [<addr2>] [<size])
- A3000 2.04 revision number fixed. It really is 37.175.
- DOS HD floppy image crash fixed
- PCMCIA SRAM card emulation. SCRAM "HD controller" in harddrive panel (other HD parameters are ignored) Hot-swapping supported. Note 1: Amiga PCMCIA slot supports max 4M of directly addressable RAM and you can't have more than 4M of (non-32bit) fast RAM at the same time. Note 2: real PCMCIA SRAM card in PCMCIA/CardBus slot is not supported yet. I need SRAM card to implement it (or good documentation)
- rom scanner odd/even split and byteswapped rom support. Does not include: AR2/3/1200 ROMs (no real rom dumps available), Arcadia ROMs (not worth the trouble) or "non-rom roms" like Cloanto ROMs.
- rom scanner data in registry changed, added human-readable names NOTE: You need to manually rescan roms if you need to use 1.4.6 or older WinUAE versions
- EXTB.L was 68000+ when it should have been 68020+ (bug since the beginning and more)
- Gayle interrupt handling rewritten to match preliminary datasheet. This solved remaining PCMCIA problems. This can also fix possible problems with IDE emulation with specific device driver versions
- uaenet.device NSCMD_DEVICEQUERY fixed (didn't work and/or corrupted memory) S2_ADDMULTICASTADDRESSES and S2_DELMULTICASTADDRESSES had wrong command number
- uaenet.device random deadlock fixed
- really ugly uaenet.device interrupt handler setup replaced with proper exec.library AddIntServer() call
- added NSCMD_DEVICEQUERY to uaeserial.device
- added some sanity checks to emulated devices' open function
- Quickstart boot disk check does not complain anymore if track zero has errors but sectors 0 and 1 are ok
- multithreaded avioutput (finally). Video/audio encoding now runs in separate thread, MUCH faster encoding on 2+ core CPUs. Note: null filter recommended (or any other 2d filter), non-filter mode requires reads from display RAM which is very slow
- GUI debugger improvements (Karsten Bock)

[Toni Wilen]

http://www.winuae.net/files/b/winuae_1500b2.zip

Beta 2:

- experimental Picasso96 speedup code. Default enabled, temporary option added to RAM-panel (stored to registry, not config file) May be slower on older PCs, I get up to 8x speedup in P96Speed benchmark (both desktop and laptop) Post your results here. Including PC specs, "old" and "new" P96Speed or other test results.
- harddrive detection now also includes more devices, even CDROMs.. (perhaps this allows PCMCIA SRAM direct access)
- Weird Dreams sound problem (hospital scene when game is starting)
- GUI debugger patch I forgot to merge

(Too quiet.. should I even bother with these anymore?)

[Retro-Nerd]

Looks like you need another temp link for this beta.

[sink]

it's me or i can't download the 2 betas

and u course like usual , u rox toni toni

[thor]

Originally Posted by Retro-Nerd

Looks like you need another temp link for this beta.

Shouldn't take long according to this post

Originally Posted by Toni Wilen

(Too quiet.. should I even bother with these anymore?)

Most definitely. Sorry that I can't test anything besides GUI debugger currently, but thanks for merging and the betas

[Ian]

Originally Posted by Toni Wilen

(Too quiet.. should I even bother with these anymore?)

I tried to download it yesterday, and today....

Service Temporarily Unavailable

 The server is temporarily unable to service your request due to maintenance downtime or capacity problems. Please try again later.

No joy....

[Toni Wilen]

Temporary attachment added to first post.

[Anubis]

Danke!

[turrican3]

i tested amikit with 1.5b2, it works very well.
Thank you toni great work

[Toni Wilen]

Originally Posted by turrican3

i tested amikit with 1.5b2, it works very well.
Thank you toni great work

Can you post old and new mode p96speed (available in aminet) results?

[Retro-Nerd]

My PC:

Windows XP/SP2
CPU: AMD Sempron 2800
Ram: 1GB DDR
Graphics adapter: GeForce 6600

[Ian]

Don't know why mine says 7mhz, but f it.

Athlon x64 3500+
2 gb DDR
Windows Vista Ultimate x64
ATI x1950 xt

Considerable gains. Only tested with a standard Amigasys install/config.

Erm, whoops.

Tested in the new version and didn't notice the option for old, the new one is using the P96-OLD setting, odd that it is far faster across the board though, any chance you have linked the new setting to the old option?

Old = Last offical release
New = Beta 2 old setting
Proper-new = Beta 2 new setting

Proper-new and old roughly even, new far better.

[turrican3]

xp sp2
amd athlon 64 3800 +
geforce 7800 gs 256 mb agp

[Toni Wilen]

Anyone with dual core? (amd or intel). These don't really match my results. I get huge increases (except in few tests)

Not that dual core shouldn't have any effect but perhaps GetWriteWatch() win32-function works faster with dual cores..

[BenniP]

AMD 64 4000+
2GB DDR
Radeon X1600 512MB
XP SP2

No real difference between my results. (As far as I see )

[_Steve_]

Originally Posted by Toni Wilen

Anyone with dual core? (amd or intel). These don't really match my results. I get huge increases (except in few tests)

Not that dual core shouldn't have any effect but perhaps GetWriteWatch() win32-function works faster with dual cores..

I am attaching my results, that do show approximately a 7 times increase in some operations:

Intel E6700
NVidia 8800GTX (forceware 158.22 - downgraded after 16x.xx gave me issues)
Windows XP (32Bit) SP2
2GB DDR2 (PC-6400)

[Ed Cruse]

I tried B2 and I'm getting some pretty strange things happening with the P96-New on.

The first picture in the attachment shows how things come up initially. A full normal looking WB screen comes up first, then within maybe half a second a half width WB screen opens on the full WB screen with funny colors. The mouse stays inside the half width WB screen.

If I do F12 and hit OK or Cancel the WB screen looks normal again. The second pictures shows what happens as I move the mouse up and down. If I move the mouse all the way up and down it rebuilds the half width WB screen. I only moved the mouse up and down a little bit to make it easier to see the difference.

I never got past this part to check the speed.

I have a new computer, 2Ghz, AMD Athlon 64X2 dual core, XP SP2

[turrican3]

I did the test once again and this time the new p96 leads.
What change ? i overclocked my gfx board, i disabled some windows services and this time no other applications was launched before the test.
I don't know if it will help, perhaps more you have free power more the results are good.
I don't know ?

[Toni Wilen]

Originally Posted by Ed Cruse

I tried B2 and I'm getting some pretty strange things happening with the P96-New on.

The first picture in the attachment shows how things come up initially. A full normal looking WB screen comes up first, then within maybe half a second a half width WB screen opens on the full WB screen with funny colors. The mouse stays inside the half width WB screen.

Must be some utility/application/patch that does this.. Can't really help without more information. (I have tested with my own basic OS installation and AmiKit and both work fine)

Do you have 32-bit display? (16-bit may not work correctly yet) Does same happen if you switch resolution?

[Anubis]

turrican3,

what did you set your CPU Emulation Speed? (how did you get 28MHz and other times only 7MHz)

I don't have AmiKit, so no test.