Request for help/advice in investigation of one interesting "huge FPS regression" bug

[Alexey Loukianov]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

15.04.2012 07:22, Alexey Loukianov wrote:
> Good day to all.
> 
> Henri, Stefan, I address this message to you at the first place as to a
> main developers of wined3d/opengl stuff... ---8<--- strip ---8<--- ... an
> app (localized version of "Perfect World" MMORPG game client from "Mail.Ru
> Games Corp") that seems to suffer huge FPS regression which I believe to be
> a bug in nVIDIA drivers rather than a regression in Wine...
> 

So I'm still trying to investigate this issue and have some additional info to
share.

Recently Mail.Ru had updated the game client so now its version almost matches
one that is available on Perfect World International site. I was hoping that
this update might made the issue go away but that didn't work out. Good news
here are that - most probably - it should be possible to reproduce this issue
using PWI game client. My sacred hopes are that Henri would be able to find
some time and took a look into it.

Another hope was that the issue would be magically auto-fixed in fresh nVIDIA
drivers, like 302.07 beta which had been released yesterday. It turned out not
to be a case.

Trying to pinpoint the cause using oprofile produced no valuable results: it
either me not able to use this wonderful profiler correctly or the issue is of
such kind that isn't easily tracked by oprofile.

Next bet was to try to start up each and every D3D app I have in windowed mode
and check if there's any that experience similar problem. I had tried running
a bunch of demos from D3D SDK and several well-known game titles (namely "UFO:
ET", "Trine", "Portal", "King's Bounty: The Legend", "Osmos", "LIMBO" and
"Braid") but none of them had suffered the issue.

What I had noticed though is a big difference in behavior that shows up when I
move an app window. When I setup Wine prefix to use virtual desktop mode,
launch "windowed" D3D app and try to move app's window around inside virtual
desktop window - the only app I've got having window contents updated while
I'm dragging it is the PW game client. All other apps tested - and they are
not affected by this bug - do not update the contents of their window while I
move it around. Don't know what does it mean per se, but it well might be a
hint that would allow someone more clever than me to understand what's
happening and why does nVIDIA driver versions 280+ acts as a trigger for this bug.

P.S. Sorry once again for making noise on devel list but I don't want to open
up a garbage bug report which would have decent chances to be closed as "not a
Wine bug, blame nVIDIA".

- -- 
Best regards,
Alexey Loukianov                          mailto:mooroon2 at mail.ru
System Engineer,                            Mob.:+7(926)218-1320
*nix Specialist

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.16 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJPohTuAAoJEPB9BOdTkBUL+v4H/iyXRKfjQ/Zkz/jljDyOMZX8
maWPGGRqqyX72P+0WJuzgQLnj26dbAP1hKJS6BciJsDBBvzdwd2rPZqQ/Yj7DduE
6PCBRQ99MIrtVHW1aMaIEhPMubC6FMazisfG87Ndi3RjnorlfQJJz41O9s27oFIa
zmXQc9OhPUNPa2JNUux3AlwtyKgD/gLDoHFN139XAp2cy6HFmhzCm4iyxufhqtT1
BNW0ZOUXWzdQqBTR0h/v3th6r+9pNLgDn4KQl2qCk2J87+sDdK5cWCo4H8yhomIF
s/uZST7HKfo1CLl/LVzNd96S0Vp2K5JIzdX3wtVNSrovTHis62NhgqL9mMllHU4=
=Avzg
-----END PGP SIGNATURE-----