Problematic Demos. - Page 14

amilo3438 · 26 October 2013, 16:05

Stuff to Love (Intro) by Savage

The scroler (pic.1), with option remove int. artefacts (pic.2) ... on link picture it seems to look different (pic.3).

EDIT: Oh, it looks to be similar problem as Shed Tears/Ozone ... so the answer would be probably the same. Case closed!

amilo3438 · 26 October 2013, 20:41

Quote:

Originally Posted by Toni Wilen

Bootblock is run with sprite DMA off. This is probably bad bootblock rip that does not "emulate" bootblock environment fully.

Boot Intro by Digitech -> http://eab.abime.net/showpost.php?p=...&postcount=249

Well, here is another bootblock intro with almost same behaving (see on pic.):

Boot Intro (Intro) by Vortex 42 -> http://janeway.exotica.org.uk/release.php?id=23867

What now dont understand is - what does it mean "bootblock enviroment"?!

How I understand - if all sprite DMA channels are OFF before loading anything into memory and then the bootblock program is loaded into memory and running, the only possibility is that the running code enabled the sprite DMA channels after or maybe "something" else is - but what, still stay unknown?!

Otherwise, are the sprite DMA channels really set OFF, mean already on WB hand shown, before anything is loaded into memory?!

Toni Wilen · 26 October 2013, 20:44

Quote:

Originally Posted by amilo3438

What now dont understand is - what does it mean "bootblock enviroment"?!

Linked files are executables, not floppy images with code in bootblock.

They have been ripped from bootblock and converted to executable (and conversion has not been very good)

amilo3438 · 26 October 2013, 21:14

Wonderboy's First (Intro) by Vortex 42

A500 OCS/ECS ... some dashes running?!

Quote:

Originally Posted by Toni Wilen

Linked files are executables, not floppy images with code in bootblock.

They have been ripped from bootblock and converted to executable (and conversion has not been very good)

OK, so there is some difference if loading from a real bootblock or ripped bootblock program file. Thanks.

Toni Wilen · 26 October 2013, 21:43

Quote:

Originally Posted by amilo3438

Wonderboy's First (Intro) by Vortex 42

A500 OCS/ECS ... some dashes running?!

A500 confirmed. (Nice colors!)

Quote:

Originally Posted by amilo3438

Balls & Motorboats (Demo) by Savage

A500 OCS/ECS ... glitches (pic.1,2)?!

Grey dot between logo and equalizers (left side) can be seen on A500 test but garbage below equalizers won't appear. No idea so far.

Quote:

Red It Intro by Savage

The logo is black (pic.3), but not on the above link picture (pic.4) ... it looks also that top of screen has some scroler text but now invisible (pic.3)?!

A500 ECS: logo is white A1000: logo is black. Reason unknown. (as in reason how color changes, colors "should" stay same as WB colors except COLOR0 and 1)

Demo has weird copperlist (lots of moves to register zero before color changes) that should never execute color changes, leaving the logo black unless COPDANG is set and I don't see how it can happen.

Never seen bluish logo.. Could be another version. (or different CLI colors or menu or whatever)

amilo3438 · 26 October 2013, 23:04

Quote:

Originally Posted by Toni Wilen

Grey dot between logo and equalizers (left side) can be seen on A500 test but garbage below equalizers won't appear. No idea so far.

Balls & Motorboats (Demo) by Savage

worked in test Beta 9 2013-10-22 16:17 ... broken in test Beta 9 2013-10-23 17:19 !!!!!!

Toni Wilen · 27 October 2013, 10:33

Quote:

Originally Posted by amilo3438

Balls & Motorboats (Demo) by Savage

worked in test Beta 9 2013-10-22 16:17 ... broken in test Beta 9 2013-10-23 17:19 !!!!!!

Ah, 68000/68010 cpu core merge had a bug, some 68000 special instructions (clr memory and scc memory) lost their 68000 special behavior (read before writing). I think this changed timing just enough.

amilo3438 · 27 October 2013, 16:27

Quote:

Originally Posted by Toni Wilen

It does not happen on real A500. Probably the usual some instruction still being wrong, too time consuming to find without some really simple test cases.

Hi Freaks (Intro) by Supplex & The Silents

last worked fine in WinUAE2.3.1 2011-02-26 ... broken in WinUAE2.3.2 2011-06-02 !!!!!!

Toni Wilen · 27 October 2013, 20:05

Quote:

Originally Posted by amilo3438

Hi Freaks (Intro) by Supplex & The Silents

last worked fine in WinUAE2.3.1 2011-02-26 ... broken in WinUAE2.3.2 2011-06-02 !!!!!!

It didn't really work any better. It just timing difference, it wasn't correct in old version, it isn't correct now. It is just differently wrong..

Corruption happens because CPU modifies bitplane pointers in copperlist after copper has fetched high word but before copper fetches low word of bitplane.

EDIT: It is only few clock cycle difference between corruption or not having corruption which means timing change in any instruction, even really tiny, will "fix" this again..

amilo3438 · 27 October 2013, 21:54

Quote:

Originally Posted by Toni Wilen

It didn't really work any better. It just timing difference, it wasn't correct in old version, it isn't correct now. It is just differently wrong..

Corruption happens because CPU modifies bitplane pointers in copperlist after copper has fetched high word but before copper fetches low word of bitplane.

EDIT: It is only few clock cycle difference between corruption or not having corruption which means timing change in any instruction, even really tiny, will "fix" this again..

Great finding! But probably will not be easy to find what instruction has bad timing ... so maybe one day, right ... anyway, thanks for your effort Toni.

amilo3438 · 27 October 2013, 22:37

Anarchy Party Demo by Symbiosis

It looks similar to Funky Big-Mac (Intro) by Lsd problem, except here dots has same colors!?

amilo3438 · 28 October 2013, 11:42

Ting Man Aldrig Maa Goere (Demo) by Rednex

EDIT:
Amiga Expo 92 (Slideshow) by Trygon

Black text on screen (pic.2)?! ... does it happens same on real OCS/ECS?

EDIT2:
New Demo by British Swapping Syndicate

KS1.2/1.3 , OCS/ECS 512mb ... at first work slowly with high cpu usage ... but when music starts work faster and cpu usage drops?! Weird?!

Toni Wilen · 28 October 2013, 19:11

Quote:

Originally Posted by amilo3438

Anarchy Party Demo by Symbiosis

It looks similar to Funky Big-Mac (Intro) by Lsd problem, except here dots has same colors!?

No ugly dots on A500. I can't see much common in code but I did only quick check, no idea if it is actually same problem or not.

Quote:

Originally Posted by amilo3438

Ting Man Aldrig Maa Goere (Demo) by Rednex

White flash is there, but position is random on real A500. Also random in emulation if you boot without disk in drive and then insert the disk manually (to introduce randomness).

Quote:

EDIT:
Amiga Expo 92 (Slideshow) by Trygon

Black text on screen (pic.2)?! ... does it happens same on real OCS/ECS?

Yes

Quote:

EDIT2:
New Demo by British Swapping Syndicate

KS1.2/1.3 , OCS/ECS 512mb ... at first work slowly with high cpu usage ... but when music starts work faster and cpu usage drops?! Weird?!

Probably does something stupid to slowdown it ("Atari emulation")

I can't see any obvious difference on my desktop (overclocked 4770k), probably need to test it on slower PC...

amilo3438 · 28 October 2013, 22:09

A little statistics:
---------------------------------------
>3200 demos/intros already verified
20 issues already fixed
16 open issues left (<0,5% of all tested!)

---------------------------------------

What does it mean now ... one needs to verify approx. at least 200 Amiga OCS/ECS demos/intros to expect to find a new problem (if any) in latest WinUAE beta 10 versions.

So, it looks that has been reached a point where further searching for any new problem is more like searching for a needle in a haystack.

Of course, if in near/distant future bump on some new/old problem/issue, for sure it will be reported here ... but 7 weeks of solo intensive work should be quite enough.

amilo3438 · 29 October 2013, 12:22

Ups, it seems forgot to report this findings earlier, so here it is now:

Marchewki (Trackmo) by Alchemy -> http://janeway.exotica.org.uk/release.php?id=8680
A500 OCS/ECS ... see on pic.1 below ...

And maybe also this one (pic.2):
Lethal Formula Final (Trackmo) by Cult -> http://janeway.exotica.org.uk/release.php?id=20650

Also, does it happens on a real A500 (pics.3,4,5)?... (note: used fix version)
Deformations (Demo) by Deform -> http://janeway.exotica.org.uk/release.php?id=8393

See arrows on Pic.6 ...
No Reality (Trackmo) by Devils -> http://janeway.exotica.org.uk/release.php?id=3676

Pic.7 ... glitches?!
Alternative Reality (Demo) by Suspect -> http://janeway.exotica.org.uk/release.php?id=12343

amilo3438 · 29 October 2013, 13:51

What does it mean in a log window: "DS: underflow (3840 8192)" !?

When it happens I got a short stuttering, and it happens not always but sometimes. (EDIT: not me but WinUAE

)
What could be a reason?

heavy · 29 October 2013, 15:18

Quote:

Originally Posted by amilo3438

Also, does it happens on a real A500 (pics.3,4,5)?... (note: used fix version)
Deformations (Demo) by Deform -> http://janeway.exotica.org.uk/release.php?id=8393

exactly the same things happens on A500. (pic4 blinking? hmm no)
fix is for the end part sound corrupted due to bugged trackloader !! seems to work with b10, but it's for real A500 : i always had the sound corrupt : see the post above !

amilo3438 · 29 October 2013, 19:30

Quote:

Originally Posted by heavy

exactly the same things happens on A500. (pic4 blinking? hmm no)

Thanks for verifying ... so, no blinking on a real A500 (pic.4)! But it does blinking on WinUAE (all green area) ... at least on my PC laptop running WinUAE.

Now need to wait to see what will Toni tell.

heavy · 29 October 2013, 19:34

Quote:

Originally Posted by amilo3438

But it does blinking on WinUAE (all green area) ... at least on my PC laptop running WinUAE.

no blinking on winuae (rested with 2700b10 - A500 cycle exact enable)

Toni Wilen · 29 October 2013, 19:34

I don't see anything blinking in that part. (Do not use beta10, reason explained few posts ago)

26 October 2013, 16:05	#261
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	Stuff to Love (Intro) by Savage The scroler (pic.1), with option remove int. artefacts (pic.2) ... on link picture it seems to look different (pic.3). EDIT: Oh, it looks to be similar problem as Shed Tears/Ozone ... so the answer would be probably the same. Case closed! Attached Thumbnails Last edited by amilo3438; 26 October 2013 at 17:02.

27 October 2013, 22:37	#271
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	Anarchy Party Demo by Symbiosis It looks similar to Funky Big-Mac (Intro) by Lsd problem, except here dots has same colors!? Attached Thumbnails

28 October 2013, 11:42	#272
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	Ting Man Aldrig Maa Goere (Demo) by Rednex EDIT: Amiga Expo 92 (Slideshow) by Trygon Black text on screen (pic.2)?! ... does it happens same on real OCS/ECS? EDIT2: New Demo by British Swapping Syndicate KS1.2/1.3 , OCS/ECS 512mb ... at first work slowly with high cpu usage ... but when music starts work faster and cpu usage drops?! Weird?! Attached Thumbnails Last edited by amilo3438; 28 October 2013 at 18:26.

28 October 2013, 22:09	#274
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	A little statistics: --------------------------------------- >3200 demos/intros already verified 20 issues already fixed 16 open issues left (<0,5% of all tested!) --------------------------------------- What does it mean now ... one needs to verify approx. at least 200 Amiga OCS/ECS demos/intros to expect to find a new problem (if any) in latest WinUAE beta 10 versions. So, it looks that has been reached a point where further searching for any new problem is more like searching for a needle in a haystack. Of course, if in near/distant future bump on some new/old problem/issue, for sure it will be reported here ... but 7 weeks of solo intensive work should be quite enough. Last edited by amilo3438; 28 October 2013 at 22:20.

29 October 2013, 12:22	#275
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	Ups, it seems forgot to report this findings earlier, so here it is now: Marchewki (Trackmo) by Alchemy -> http://janeway.exotica.org.uk/release.php?id=8680 A500 OCS/ECS ... see on pic.1 below ... And maybe also this one (pic.2): Lethal Formula Final (Trackmo) by Cult -> http://janeway.exotica.org.uk/release.php?id=20650 Also, does it happens on a real A500 (pics.3,4,5)?... (note: used fix version) Deformations (Demo) by Deform -> http://janeway.exotica.org.uk/release.php?id=8393 See arrows on Pic.6 ... No Reality (Trackmo) by Devils -> http://janeway.exotica.org.uk/release.php?id=3676 Pic.7 ... glitches?! Alternative Reality (Demo) by Suspect -> http://janeway.exotica.org.uk/release.php?id=12343 Attached Thumbnails Last edited by amilo3438; 29 October 2013 at 14:09.

29 October 2013, 13:51	#276
amilo3438 Amiga 500 User Join Date: Jun 2013 Location: EU Posts: 1,506	What does it mean in a log window: "DS: underflow (3840 8192)" !? When it happens I got a short stuttering, and it happens not always but sometimes. (EDIT: not me but WinUAE ) What could be a reason? Attached Thumbnails Last edited by amilo3438; 29 October 2013 at 15:37.

29 October 2013, 19:34	#280
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,516	I don't see anything blinking in that part. (Do not use beta10, reason explained few posts ago)

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problematic GVP A530 Turbo	Mogsy	support.Hardware	22	13 November 2012 19:12
Problematic thread?	prowler	project.EAB	5	28 October 2011 00:06
Problematic Holiday Lemmings 94 demo disk	MethodGit	support.Games	19	08 November 2010 03:56
Problematic slaves list	keropi	project.Killergorilla's WHD packs	44	28 November 2008 12:46
Looking for two demos...	eternalcrusader	request.Demos	3	01 August 2006 08:57