PS2 [Open PS2 Loader] Game Bug Reports

[TnA]

Actually that 'contradiction' is quite good, because it proves that NEITHER an SSD was used (so the supposed argument of @Grahf regarding that is clearly void now), NOR a Clone-HDD-Adapter!!!

It is also proof that this is not related to SATA, but could happen on IDE-Drives as well, because the SONY-NWA has no native SATA and the maxdiypower-adapter 'translates' it to IDE-'ops'/'functions'/'calls'...


THX @Algol!
Btw.: You mean 'maxdiypower' (like 'maximum do it yourself power'), right?!

 

[TnA]

There is one thing which just came to my mind, which might help in that regard! It could be a foolish idea as well tho'... Lol


I believe the HDD delivers data so fast on a request, that the timing and interrupts made in-between could cause this issues... (oversimplified short explanation, a.k.a. 'tl;dr')

Essentially, due to varying threading during runtime the 'answer-times' might 'jitter' a bit, thus leading to these weird test-results!
Just like the ping-example (websites), I posted before... Once the answer-time could be 47ms, then 53ms, then 28ms, then... I suppose you get, what I mean.


I SUPPOSE that some part of the code @sp193 actually wrote for the streaming of BGM (in the SND-Thread) could be a good example.

The idea is to split the transfer-buffer into at least 2 (or 3) 'partitions' (possibly for all devices, not just iHDD) and change some things to the general IO-Management to be a bit less interrupt-driven (if applicable).

But this is rather a suggestion for a rather complex test and currently solely based on the tests since page 6.

This suggestion would not DIRECTLY change timings of the cdvd-emu-driver, but might affect it due to a bit different transfer-structure and thus different threading (thus possibly leading to less 'jittering' timings IF THAT IS THE CAUSE).


tl;dr2 If that is the case, than those issues could be related to possible 'race conditions' from some (partially) bugged game-code, which is triggered by too high 'answer-times'...

Again! This is a conclusion partially based on the facts and tests and partially on assumptions, because we have not (for 100% certainty) identified the culprit(s) of it(/those issues), in most cases (yet).

Yes, I state it when something of me is an assumption... lol
But it's still an assumption based on observations (the tests since page 6)!

 

[Peppe90]

One thing I noticed:
I boot this game, then IGR (immediately or after it freeze, it doesn't make difference), start Ferrari challenge trofeo pirelli and got the freeze at the second loading screen (I can solve this cold loading the HDD with Ule), it's the only game that cause this.

But I'm now thinking that it can be related to the MDMA change… I'll do some tests.

As I thought it was related to the MDMA changing…

Normally Ferrari Challenge trofeo pirelli works without any setting change (with any OPL/HDLoader version), so in UDMA 4 with OPL.
But if I start a game with a different DMA, then start again Ferrari Challenge TP, it freezes on the second loading time (after the MC check).

To solve I need to cold load the HDD with Ule (accessing hdd0). Even cold booting to certain OPL versions works (it works with OPL_1330_DB, doesn't work with OPL 0.9.3_all).

@Grahf Tried your settings (with SONY NA and maxdiypower board, OPLB_1330) and the game still freeze (it loads a little more, about 95%, but eventually freezes). I'll try the same with OPL 0.9.3 to make sure it's not the OPL version causing the problem.

 

[Grahf]

@Grahf Tried your settings (with SONY NA and maxdiypower board, OPLB_1330) and the game still freeze (it loads a little more, about 95%, but eventually freezes). I'll try the same with OPL 0.9.3 to make sure it's not the OPL version causing the problem.

Can you try a standard, internal HDD (IDE)? I never tested it with a SATA adaptor, nor would I since the compatibility with those is all over the place (hence so much confusion in threads like this).

Regarding State of Emergency, it sure is an interesting case, but it doesn't mean that the game is doing something all that weird. I wonder if the timing with this one is so annoying because it comes as a 4.5 GB disc, but the game itself is only around 550 MB? The rest is a dummy file, it seems. Maybe this is screwing up the seek times?

 

[Peppe90]

Can you try a standard, internal HDD (IDE)? I never tested it with a SATA adaptor, nor would I since the compatibility with those is all over the place (hence so much confusion in threads like this).

This is my HDD: http://ps2drives.x-pec.com/?p=drive&did=6245
Anyway I've just launched HDLoader (I added some games since the last time I used it, it takes 50 minutes to create the list!). State of Emergency boot perfectly...

Regarding State of Emergency, it sure is an interesting case, but it doesn't mean that the game is doing something all that weird. I wonder if the timing with this one is so annoying because it comes as a 4.5 GB disc, but the game itself is only around 550 MB? The rest is a dummy file, it seems. Maybe this is screwing up the seek times?

Ferrari challenge TP too is quite interesting. The game freeze at the second loading time (after the MC check), but if you cold boot your HDD with Ule (booting Ule from anywhere but the iHDD obviously), the game won't freeze anymore. At least until you do some "determined things", like booting a game with a DMA that's not UDMA4, installing a game, making certain modifications to certain MC files and even leaving the Ps2 power cord disconnected for a while… It's all clear no??

And even more interesting is that this game (Always Ferrari Challenge TP) never freeze at the loading time if you boot HDL/OPL (any version, from the oldest to the newest) from USB.

I'll try to figure out the DMA settings of the HDLoader version I just used.

 

[TnA]

As I thought it was related to the MDMA changing…

Normally Ferrari Challenge trofeo pirelli works without any setting change (with any OPL/HDLoader version), so in UDMA 4 with OPL.
But if I start a game with a different DMA, then start again Ferrari Challenge TP, it freezes on the second loading time (after the MC check).

To solve I need to cold load the HDD with Ule (accessing hdd0). Even cold booting to certain OPL versions works (it works with OPL_1330_DB, doesn't work with OPL 0.9.3_all).

These seem to be 2(-3) issues actually...

1. A timing-issue triggered by slow transfer-modes...
2. (Possibly) The HDD is still 'initialized' with the old transfer-mode and after a reset, it doesn't re-initialize the HDD completely (or at least the settings like transfer-mode). This is similar to the issue of an SMB-Share or a USB-Device not being 'unmounted' (not 100%ly the correct technical term) on a reset/poweroff (poweroff solely affects SMB/LAN). --> 2. would/could be the trigger for 1. in that case!
3. (IMO unlikely but still possible) some stuff is remaining in RAM, after a reset...

One question: Does this also happen, when you configure the game directly to use that transfer-mode?!?

@Grahf Tried your settings (with SONY NA and maxdiypower board, OPLB_1330) and the game still freeze (it loads a little more, about 95%, but eventually freezes). I'll try the same with OPL 0.9.3 to make sure it's not the OPL version causing the problem.

THX for these tests...
I marked something interesting, which (IMO) points to a timing-issue again.

Can you try a standard, internal HDD (IDE)? I never tested it with a SATA adaptor, nor would I since the compatibility with those is all over the place (hence so much confusion in threads like this).

Yes, that would be interesting (especially if the rest of his setup remains the same).

Even better would be a test of a faster/newer 500GB IDE HDD and a test with an old IDE HDD (40GB or so) as well!

Regarding State of Emergency, it sure is an interesting case, but it doesn't mean that the game is doing something all that weird.

You are correct on that!
The symptoms however (not only of this game), make it quite likely because it is hard to explain it in another way seen from the technical perspective (hard-&software-combination).

I wonder if the timing with this one is so annoying because it comes as a 4.5 GB disc, but the game itself is only around 550 MB? The rest is a dummy file, it seems. Maybe this is screwing up the seek times?

Well, that MIGHT be the case indeed! Yes, that could be correct, but only under the premise that the game's code is expecting a specific kind of 'timing' due to these longer seeks (i.e. when it would 'jump' from the middle to the outer disc or vice versa).

If that's the case, the solution of patching it would be quite easy in that case, because we would only need to patch the delays to lower (approximate) values.
It would be harder to even trace the function which causes this, than the patching of the values itself...


On another note... Did someone ever tried to rip SoE CORRECTLY to a CD?

If it doesn't have any LBA-Checks and still freezes, that could be an indication that this game indeed expects high seek-times (on purpose).

Does it work on HDLoader?

 

[TnA]

I edited my post quite a bit. I hope you've (all) seen it!

 

[TnA]

Why the heck do Peppe90's posts appear after a while in the midst of the thread?

Hm... So it works with HDLoader... Interesting!

Have you tried using on of these other transfer-modes on the Ferrari-Game directly? Does it cause the same issue?

 

[Peppe90]

Why the heck do Peppe90's posts appear after a while in the midst of the thread?

lol it's because my posts need to be approved and you guys write too fast!!

Hm... So it works with HDLoader... Interesting!

Yes and it seems that the speed doesn't matter. I tried the game in UDMA6 (133,3mbps), UDMA4 (66,6mbps), MDMA0 (4,2mbps) and with another HDLoader version of which I don't know the DMA.
Always cold booting to HDLoader. I can notice differencies in the in filling the loading bar speed, but it never freeze.

Have you tried using on of these other transfer-modes on the Ferrari-Game directly? Does it cause the same issue?

I tried this:
booted a game with MDMA2 then booted FC setting it the same (MDMA 2) and it freeze. Never tried what you're saying, I'll make you know.

 

[Peppe90]

Have you tried using on of these other transfer-modes on the Ferrari-Game directly? Does it cause the same issue?

I cold boot to OPL, set Ferrari Challenge to MDMA0 and it doesn't freeze.
Then I IGR, launched State of Emergency (that obiously freeze), IGR, launched FC (UDMA4) and it freeze.

About State of Emergency…
I forgot to mention it doesn't need any MODE enabled with HDLoader, it just runs fine.

 

[Grahf]

I forgot to mention it doesn't need any MODE enabled with HDLoader, it just runs fine.

I wrote this a few pages ago.

The specific combination of modes is needed in OPL 0.9.3 and the betas I tried (1200 an 1220).

 

[TnA]

I cold boot to OPL, set Ferrari Challenge to MDMA0 and it doesn't freeze.

Why MDMA0? You've mentioned using MDMA2 before for SoE and after restarting FC froze...
So would you mind testing FC with this very same transfer-mode, MDMA2?! Does it still not freeze on a cold boot?

About State of Emergency…
I forgot to mention it doesn't need any MODE enabled with HDLoader, it just runs fine.

THX!

@Grahf: I forgot it, sorry...

 

[Peppe90]

Why MDMA0? You've mentioned using MDMA2 before for SoE and after restarting FC froze...
So would you mind testing FC with this very same transfer-mode, MDMA2?! Does it still not freeze on a cold boot?

I think that FC will work in any DMA mode… I'll try MDMA2 anyway to be sure.

P.S. I still can't pass through the State of Emergency loading screen with OPL...

 

[Grahf]

P.S. I still can't pass through the State of Emergency loading screen with OPL...

With that 5400RPM internal HDD you should be able to using the combination I posted. At least that's how it works for me with the PAL version. Version 1.02, I might add, but I think the only difference is that it has more language options than 1.01.

I did check this in all ways. Adding or removing modes, changing the transfer speed, but it only works with that specific combination and MDMA2/UDMA0.

 

[Peppe90]

With that 5400RPM internal HDD you should be able to using the combination I posted. At least that's how it works for me with the PAL version. Version 1.02, I might add, but I think the only difference is that it has more language options than 1.01.

I did check this in all ways. Adding or removing modes, changing the transfer speed, but it only works with that specific combination and MDMA2/UDMA0.

I've your same version, SLES_506.06 MD5: bb2e77b3a2061b035726978a5e3a900a

As I said it doesn't work for me with your settings. It work with HDLoader though, with any DMA mode, from MDMA 0 to UDMA 6

@TnA Ferrari Challenge TP perfectly works with MDMA 2 too.
This morning happened a weird thing though. I cold boot to OPL, started Ferrari challenge TP with my normal settings (No modes UDMA4), it works without problems. Then I IGR and lost the video signal. I had to hard-power-off the Ps2.

The lost of video signal happen after IGR with some games, but until now for me it's Always been a stable thing. Games that lost the video signal, Always do it, but FC never done this joke before…
After power-off, I restart the Ps2, gone to ULe and after a while it freezes (maybe my Ps2 is tired of all that testing )

This remind me a little to State of Emergency, when I played it with the Gamestar adaptor. After the IGR 90% of times I got an empty page in OPL HDD page. I had to restart the Ps2 to make the HDD list reappear.

 

[Tupakaveli]

I concur with @Peppe90

Tested State of Emergency SLES_506.06
MD5: bb2e77b3a2061b035726978a5e3a900a

With OPL r1220, Sony NA with maxdiypower SATA board and 5400rpm HDD.

With MDMA2 + Mode 2 + Mode 3 + Mode 6 + Mode 8 the game freezes.

Tried UDMA0 and MDMA1 with the same result.

 

[Peppe90]

I concur with @Peppe90

Tested State of Emergency SLES_506.06
MD5: bb2e77b3a2061b035726978a5e3a900a

With OPL r1220, Sony NA with maxdiypower SATA board and 5400rpm HDD.

With MDMA2 + Mode 2 + Mode 3 + Mode 6 + Mode 8 the game freezes.

Tried UDMA0 and MDMA1 with the same result.

I just enter to ask for someone having my same board

...The game worked for me with exactly same equipment (same HDD, MC, Ps2...) but with the Gamestar NA (and it needed only MDMA 0).
I seems to remember that this game worked even with the SONY NA as long as I used the BITFUNX board.

Therefore, OPL (versions after 0.9.2) could have some incompatibilities with the maxdiypower board??

 

[Tupakaveli]

I still have the IDE board I took out of the NA so tomorrow night I will try putting it back in and using an IDE HDD just to see if it makes any difference.

 

[Grahf]

I still have the IDE board I took out of the NA so tomorrow night I will try putting it back in and using an IDE HDD just to see if it makes any difference.

I tested it on an IDE HDD, with no mods to the official Sony NA.

Also, when you're switching between OPL versions, make sure to clear ALL settings. Not just Game Settings, but OPL's settings as well. Then turn on the four compatibility modes (2+3+6+8) and pick MDMA2 or UDMA0.

 

[Peppe90]

I tested it on an IDE HDD, with no mods to the official Sony NA.

Also, when you're switching between OPL versions, make sure to clear ALL settings. Not just Game Settings, but OPL's settings as well. Then turn on the four compatibility modes (2+3+6+8) and pick MDMA2 or UDMA0.

What do you mean by "clear all settings"?? To delete the opl cfg file??

I thought about it. That maybe some OPL settings could affect the compatibility, since the game wotked for me before. I tried to set my OPL as I had it last year, but I can't remember exactly…

I tried with another MC too with different settings and it doesn't work.

With MODE3, MODE6, MODE 8, MDMA2 the loading bar fill-up a little more (95%)