Tales of the Abyss crashes in battle (with the gamefix enabled, started happened with a recent version)

[GoogleCodeExporter]

1) Did the game ever work correctly? 
Yes. PP319 has no issues with the gamefix (slow DVD access) enabled. Not
sure what the cutoff revision is.

2) What steps will reproduce the problem?
1. Open the game with a more recent revision (for example, 353 and 362)
with the fix enabled.
2. enter a battle, a crash will most likely occur pretty quickly.

5) Please provide any additional information below.
Without the fix, by the way, crashes still happen (they're random crashes
in or after battle)

Original issue reported on code.google.com by Nne...@gmail.com on 25 Nov 2008 at 6:22

[GoogleCodeExporter]

Well, with recent revisions you should prolly leave the slow dvd fix off.
The crashes that one fixed are already gone in recent builds.

We'll have to wait a bit how coding on IOP comes along, then decide what to do 
with
that gamefix.

Original comment by ramapcsx2 on 25 Nov 2008 at 9:57

[GoogleCodeExporter]

I've reduced the EE_WAIT_CYCLES to 512 in r363.  Please test it and let us know 
if it
fixes this problem.

Original comment by Jake.Stine on 27 Nov 2008 at 1:42

• Changed state: Accepted

[GoogleCodeExporter]

It didn't - with the fix still causes a crash at battle entry; without the fix 
still
causes random crashes at battle exit.

Original comment by Nne...@gmail.com on 27 Nov 2008 at 10:07

[GoogleCodeExporter]

THe game doesnt crash here. And this is without the game fix enabled.
Im running it using a .ISO file.

Original comment by Wagnar...@gmail.com on 27 Nov 2008 at 11:00

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Without the fix, it doesn't always crash - it seems to only happen after 
battles in some
areas (and never happen in other areas). 
I'll attach a memory card save - from the save go left and fight a wolf. seems 
to
always (almost always?) crash after that battle (not in the battle).

Original comment by Nne...@gmail.com on 28 Nov 2008 at 12:22

Attachments:

• abyss.rar

[GoogleCodeExporter]

Humm indeed it does. I was only checking at the beginning of the game.

error I got in console is :  a lot of vmhack thing.
and :
etAddr: memLUT PFN warning
# DMAC(15) Handler does not exist.
Exception write.

I even tested with a EE_WAIT_CYCLES of 200 and it didnt fix it,

Original comment by Wagnar...@gmail.com on 28 Nov 2008 at 12:35

[GoogleCodeExporter]

For me, the crash are in R364, R363 and less seem to work fine / better

Original comment by Wagnar...@gmail.com on 28 Nov 2008 at 3:10

[GoogleCodeExporter]

Well, even with the gamefix the game would hang after battles in older 
revisions.
What I think happens now is the game crashing right away, due to the same 
problem.
IOP timing are still not perfect, and i guess they mess up the game again..

Original comment by ramapcsx2 on 28 Nov 2008 at 7:58

[GoogleCodeExporter]

Still happens in r393 for me.

Original comment by Nne...@gmail.com on 5 Dec 2008 at 12:22

[GoogleCodeExporter]

Recent fix to a counters bug looks like it may have solved the crashing and slow
loading problems.  Please give the latest r440 svn a try and let me know if you 
still
have crashes after battles or slow loading times.

Original comment by Jake.Stine on 16 Dec 2008 at 8:06

[GoogleCodeExporter]

On r441, the game seems to crash with TLB miss errors (in the TLB build, regular
crash in the VM build) on entering battle.

Original comment by Nne...@gmail.com on 16 Dec 2008 at 12:55

[GoogleCodeExporter]

Give r448 a try.  It should be significantly more stable now, and (I think) the
SlowCDVD hack shouldn't be needed anymore.  Please report if you still 
experience any
crashes.

Original comment by Jake.Stine on 18 Dec 2008 at 4:36

[GoogleCodeExporter]

Looks stable over here. Thank you.

Original comment by Nne...@gmail.com on 18 Dec 2008 at 11:19

[GoogleCodeExporter]

Issue (finally!) resolved in r448 by implementing more accurate CDVD seektime 
delays.

Details: TOTA performs "pre-reads" where it initiates subsequent CDVD block 
reads
before copying the data from previous reads.  When the reads occur too quickly 
the
game hasn't time to copy the data before the new data comes in, resulting in 
random
crashes.  This is the same problem that it suffers from when running off a PS2 
HDD,
and is fixed by adding a sufficient delay to the cdvd's seek time.

Case Closed. :)

Original comment by Jake.Stine on 19 Dec 2008 at 3:01

• Changed state: Verified