ATI Catalyst 7.5 hotfix for ArmA leak

[fasad 1]

Thanks for the info Maruk.

[Horus 83]

So I take it j_f_oko and me are the only ones this hotfix didn't work for? Anyone else have problems with the 7.5 hotfix out there?

Maybe u need turn off WHQL certification...start>control panel>system>hardware>driver signing...select 'Ignore - Install the software..' and tick the 'Make this action the system default' box.

[berghoff 11]

Can you describe the "ATI beach bug"?

Sure there is a clear visible line between transistion where wave hits the shore. Doesn't happen on all ATi cards though.

Like this? I refer to the sharp change in reflected light just to the left of the middle of the screen. It appears to be the edge of the water plane itself, the water to the right of this line is painted onto the terrain.

[ig]http://i164.photobucket.com/albums/u36/aghiawie/ArmA_beach_line.jpg[/img]

Does anyone not get this line, ATI or otherwise?

Ah yes its fixed, on previous version of video driver the shore didn't have the ripples on water and you could look through the water with ease.

[Jack-UK 0]

Downloaded it and its fixed Sweet

[scarz^ 0]

Thats awesome,finaly i can play the game wihtout "hendicap",now its time to change the old trash(r9800pro,1gb ddr,3200 barton) with new stuff hehe

Greetings from Croatia!

[Easy_tiger 0]

Still doesn't work for me.

After 30min the handle count had risen to 22k and the gfx bug was back.

I have also had 2 occasions now where ive had a BSOD referring to ati3aug.dll

[Richieb0y 0]

whit that fix(lol) its unplayable for me im going back 7.5

im waiting for my ohter parts then i can fit my evga 8800GTS and never turn my back at nvidia

[Easy_tiger 0]

I hope that this hotfix doesnt pass WHQL testing , (would drivers that cause windows stop messages pass testing anyway??? ) as i suspect ATI will then sign this handle leak off as complete and im stuck with a brand new bit of ATI TRASH

[Dwarden 1124]

tiger it's just one of many driver branches AMD.ATI driver team(s) work on ...

if the ArmA fix/change is complete and stable then they implement it into next stable WHQL branch ...

[USSRsniper 0]

It feels like an old driver.... this is strange fix, maybe they are using old files jsut to fix errors? Because with this fix control center becames old looking, when 6.x drivers were used.... Or this is because, its just a quickfix?

[Ilaafi 0]

Well this hotfix perhaps fixed the handle leak but it added some weird stuttering to my 1900xtx, like loading pauses or something. And I went back to 6.7 and stuttering is gone. So this fix is useless for me.

[mad_rabbit 0]

It seems I picked the worse combination possible for ArmA:

i.e. Widescreen @ 1440x900 Max 60Hz + ATI x1950Pro

737-27932: Radeon series - Black Display with the Error Out of Range When Selecting 1440x900/60Hz Resolution in Catalyst Control Center

Other reports:

1440x900 #1

1440x900 #2

1440x900 #3

Forum thread on my local server about this as well:

OGN Forum - Handle Leaks

GOD DAMN ATI PULL YOUR FINGER OUT!!!

EDIT: ATI Link fixed

[Spectre-63 0]

It seems I picked the worse combination possible for ArmA:

i.e. Widescreen @ 1440x900 Max 60Hz + ATI x1950Pro

737-27932: Radeon series - Black Display with the Error Out of Range When Selecting 1440x900/60Hz Resolution in Catalyst Control Center

Other reports:

1440x900 #1

1440x900 #2

1440x900 #3

Forum thread on my local server about this as well:

OGN Forum - Handle Leaks

GOD DAMN ATI PULL YOUR FINGER OUT!!!

mad_rabbit -

your link to the ATI board is for the 7.5 hotfix, not a 1440x900 issue

unable to duplicate your experience with the 7.5 Cats. I am running an X1950XT though, instead of the Pro. Have you verified that the card isn't setting the refresh rate too high? You may need to grab PowerStrip and modify the refresh rate settings for that resolution.

[Dwarden 1124]

7.5 hotfix is bigger than 7.5 whql because hotfix contains help files for all languages

also hotfix was compiled independently on 7.5 whql because AMD.ATI needs to be sure this feature/fix is final before merging it with main driver branch ...

consider hotfix as public beta (and be glad ATI done it as they rarely release any hotfixes)

btw.

to overcome issue with blank screen on 1440x900 You may try use Catalyst COntrol Center or PowerStrip or Ati Tray Tools to FORCE 75Hz (try also 60Hz in PS/ATT) for that resolution ...

[mad_rabbit 0]

@Spectre-63

I fixed the link for the 1440x900 issue, thanks for the heads-up.

@ALL

I currently use ATI tray tools instead of powerstrip.  In terms of having my refresh rate too high, I was under the impression from reading around the net that setting the refresh higher than 60Hz WAS the solution as opposed to the problem.  My question is this:

How do I set the refresh rate higher than 60Hz if I can't see the screen to do it?

i.e. Install new drivers, restart windows, screen black therefore can't set refresh rate, go into VGA only mode to reset drivers?!

Also I may be wrong but I think my monitors max refresh rate at 1440x900 is 60Hz ?!

LG L194WT

EDIT: Yep.  My monitors max refresh @ 1440x900 is 60Hz digital. An although I could switch to analog for 1440x900@75Hz I shouldn't have to!

----------

On a final note, although we should be lucky that ATI have released a hotfix, should we also not be angry that as paying customer we have a 'faulty' product for want of a better term?  Please no plight of the poor technicians speech because we all know it's not them, it's ATI as a company responsibility and therefore fault.

[memnoch 0]

Well this hotfix perhaps fixed the handle leak but it added some weird stuttering to my 1900xtx, like loading pauses or something. And I went back to 6.7 and stuttering is gone. So this fix is useless for me.

I had exactly this problem when I went back to the other game I'm playing at the moment, Lord of the Rings Online, and the performance was atrocious. For no reason at all the game would just stall for a few seconds and then continue at full speed. It didn't matter where I was and didn't coincide with HD access either. Put the normal 7.5 and it works perfectly. Guess I'll be waiting for 7.6

[Spectre-63 0]

mad_rabbit -

a couple of points:

1.) re: your monitor's limitation of 60Hz for 1440x900 digital - not anything you can do about it, simply a limitation of the monitor.  LCD's, unlike CRT's, don't look bad at 60Hz though.  My Gateway FPD2485-W runs at 60Hz at 1920x1200 and looks fantastic.  If I'm not mistaken, Dell's high-end 24" widescreen runs at the same refresh.  As I said though, 60Hz isn't a problem, especially since you can't see the refresh with an LCD (at least I can't)

2.) I wouldn't necessarily agree that the hotfix provided by ATI indicates a fault BY ATI.  This article at Extreme Tech goes into some depth about what's going on with ATI's Vista driver development and, if you have an interest in some of the stuff going on behind the scenes, is a very interesting read.  Pay special attention to the third page where Phil Rogers discusses the Delayed Surface Delete issue...one that's can be aggravated by programming decisions made by the software developer - in this case, CryTek.  Unfortunately, software developers aren't in a position to fix the issue once the software is released - the time for modifying how the graphics engine works has come and gone.  That leaves ATI in a position of having to provide a fix for an issue that is not of their own creation.  They do so for the benefit of their customers, not out of a responsibility to correct the error.

I most definitely don't know enough to say whether this driver handle issue was one caused by BI or ATI...I'm leaning towards the former but have no evidence either way.  All I'm saying is that just because ATI provided the resolution doesn't mean that the source of the issue lies with them.  It could come down to decisions made by BI during the graphics engine development.

[mad_rabbit 0]

@Spectre-63

1) Yeah I've learnt a lot over the last couple of months in regards to LCDs and what I know about refresh rates, std. resolutions, etc.  And I definitely don't notice it at 60Hz.  

But again the only work around in order to allow me to use the new 7.5 drivers is to increase the refresh rate whilst at 1440x900 which is against the physical limitations of the monitor and alos practical i.e. can't change via windows if I can't SEE windows.

2) I had a brief scan through the article you posted on ATI driver development.  The long and short of it for me is that I, like others, am the customer.  What happens behind the scene is 'behind the scenes'.  If BI and/or ATI want to shift the blame back and forth, which they haven't been to the best of my knowledge, this does not help me as a customer.

So whilst I agree, or at least can spectulate, that disagreements between developers, technicians at both ends can be lead to problems during driver development, this should not be my problem as a customer.  It becomes my problem when the concentration is on who to blame rather than just fixing the problem, regardless of the reasons or fro the customers prespective, excuses.

That aside, I would not usually voice this opinion as many people have it and also the corresponding sympathy for the plight of many game or driver developers (yes I have this too believe it or not).  

What's #&@% me off is that the solution to one problem i.e. leaky handles, has created another problem, unable to use the leaky handle fix and/or 7.5 drivers on 1440x900 widscreen.  Now the first problem with the leaky handles, fair enough, it's game (ArmA) specific = development conflict.  The second, well...doesn't anyone have a widescreen monitor to test the drivers out at ATI?  IT just appears to be a massive oversight, hence ATI should pull their finger out.

------

I'm trying to make this rant at least semi-constructive. I've attached links to Omega and DNA modified driver homepages below:

Omega

DNA

Now as far as I know, the latest Omega drivers still use the Catalyst 7.4 as a base whilst the DNA drivers now have the 7.5 hotfix as a base.

I'll try test them tonight to see if they fix the widescreen problem with 1440x900@60Hz but as far as I understand things, if ATI can't fix the problem yet then what hope to these mod teams have.

[Easy_tiger 0]

Well after a proper attempt at installing the hotfixed drivers i can confirm that the handle leak has been sorted, handle count now is stable at around a couple of hundred handles.

However ive still got bugs mainly the gfx corruption problem and BSODs referring o ati3uag.dll

The gfx bug is a strange one, and it always starts with a temporary freeze of about 1 second before the shadows start going crazy and the random solid color polygons appear (which seem to stem from the centre of the screen).

The handle count remains unchanged whilst this bug is occuring.

a ALT-TAB or screen res change will get rid of the corruption for a bit, but its always back after a while (if a ati3ug.dll BSOD doesnt occur first!

[Dwarden 1124]

keep reporting crash bugs via support tickets on AMD.ATI support website ...

some of crashbugs (in general) are already on known list but more reports = faster fix

[Warpy 0]

Arma is unplayable for me with the hotfix driver if you can even call it that... Quiet clearly this bug hasn't been fixed and they do not know what they are doing - if they did we wouldn't have a thread loaded with problems...

[Warpy 0]

keep reporting crash bugs via support tickets on AMD.ATI support website ...

some of crashbugs (in general) are already on known list but more reports = faster fix

What planet are you on?

This bug was noted by Suma back in February this year - thats almost 6mths ago (many driver releases from ATI since then, in fact they release a new set once a month!! ) and the issue still hasn't been fixed. I believe hundreds of people contributed to lodging tickets with ATI and the issue hasn't be resolved... What is the point in reporting the same crash bug over and over again, are you assuming ATI fix things on the basis of mass volume tickets? I'd have thought that Bohemia's Programmer lodging ticket a ticket in Feb would have been enough.

[Dwarden 1124]

keep reporting crash bugs via support tickets on AMD.ATI support website ...

some of crashbugs (in general) are already on known list but more reports = faster fix

What planet are you on?  

This bug was noted by Suma back in February this year - thats almost 6mths ago (many driver releases from ATI since then, in fact they release a new set once a month!! ) and the issue still hasn't been fixed.  I believe hundreds of people contributed to lodging tickets with ATI and the issue hasn't be resolved... What is the point in reporting the same crash bug over and over again, are you assuming ATI fix things on the basis of mass volume tickets?  I'd have thought that Bohemia's Programmer lodging ticket a ticket in Feb would have been enough.  

what bug ? ... if You speak about leak it's fixed already...

also that bug was fixed in multiple stages ... first leak speed was decreased then it was fixed for owners of AMD CPUs then for rest ....

what i was reffering to was about reporting these crashes and other issues ASAP ...

crashing ? where SUMA confirmed any crashes to originate from/to ATI drivers ?

fixes delay? they are in priority list ... if You get 100 ArmA and 5000 Stalker and 60000 WoW users of ATI card complaining about error then of course top priority is first WoW then Stalker then ArmA ...

i'm not saying about reporting same bug multiple times from 1 person, but FROM multiple users and each user should keep that ticket open (aka tick 'this autoanswer =! fix) as unresolved till it's really FIXED

be sure ATI fix bugs first based on volume and critical type (vulnerability, stability) this also gets hand in hand with limited resources (programmers etc)

[ManDay 0]

Well. I never encountered any problems but I will d/l it now. Maybe I am going to experience a great performance afterwards - which would be awesome

-edit-

is that fix also included in the latest 7.5 now or do i have to d/l it seperatly?

-edit-

W_T_H? The hotfix is two times bigger than the full package with cat ctrl-center and drivers itsself? so. explain that?

[Dwarden 1124]

already said that here like 2 times ... it contain all help files for all languages ...