http://bugs.winehq.org/show_bug.cgi?id=9717
Summary: Gothic 3 - Product: Wine Version: unspecified Platform: PC-x86-64 URL: http://www.gothic3.com/index.php?do=09709911610511111006 1109101100105097 OS/Version: Linux Status: UNCONFIRMED Severity: enhancement Priority: P2 Component: wine-misc AssignedTo: wine-bugs@winehq.org ReportedBy: kgbricola@web.de
Created an attachment (id=8141) --> (http://bugs.winehq.org/attachment.cgi?id=8141) Graficissue Gothic 3
There is a black "shadow" of the landscape mirrored along the middle of the screen.
My System: Fedora 7 x86_64, Nvidia 8800GTS, Proprietary drivers 100.14.11
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Gothic 3 - |Gothic 3 - shadow on the | |upper half of the screen
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|wine-misc |wine-directx-d3d
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #1 from Rico kgbricola@web.de 2007-10-03 06:23:35 --- There seems to be an issue with the shadows. They are placed at the wrong position!
http://bugs.winehq.org/show_bug.cgi?id=9717
Yannik Hampe yankee@cipher-code.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #2 from Yannik Hampe yankee@cipher-code.de 2007-10-06 09:36:36 --- *** This bug has been confirmed by popular vote. ***
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #3 from Yannik Hampe yankee@cipher-code.de 2007-10-06 10:04:49 --- Created an attachment (id=8420) --> (http://bugs.winehq.org/attachment.cgi?id=8420) Log: Start Gothic, Start new game, wait a couple of seconds, exit gothic
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #4 from Rico kgbricola@web.de 2007-10-11 16:51:39 --- So I only got a lot of this:
- fixme:d3d:IWineD3DDeviceImpl_ValidateDevice (0x1342f8) : stub - fixme:d3d:scissorrect >>>>>>>>>>>>>>>>> GL_INVALID_VALUE (0x501) from glScissor @ state.c / 3608
So some width and some height values are negative!
http://bugs.winehq.org/show_bug.cgi?id=9717
Norbert Lataille nonal@freesurf.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |nonal@freesurf.fr
--- Comment #5 from Norbert Lataille nonal@freesurf.fr 2007-11-22 13:45:14 ---
fixme:d3d:scissorrect >>>>>>>>>>>>>>>>> GL_INVALID_VALUE (0x501) from glScissor @ state.c / 3608
Yes, glScissor is supposed to be set in screen coordiates, and wine receives a lot of huge/negatives numbers. In fact these bad cord could well be an app bug (unnoticed on windows).
Whatever, this does not seem to be the source of our graphic corruption (to confirm this I suppressed these glScissors calls => no change).
This corruption which looks more stencil buffer issue to me. The game may be doing multipass rendering, and may be trying to stencil-protect the ground. It seems the area which is black/"not written to" on the sky is exactly the mirror image of the ground shape. Do not mirror it and you are fine.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #6 from Norbert Lataille nonal@freesurf.fr 2007-11-24 09:53:05 --- Hi,
More info:
At least 3 bugs so far.
1/ Wine does not load properly the openmp.manifest file, which causes the game not to load.
Easy fix, patch ready.
2/ "fixme:d3d:scissorrect >>>>>>>>>>>>>>>>> GL_INVALID_VALUE (0x501)"
These messages are caused by a wine bug in wined3d, not by an app bug. However, as I suspected, fixing it did not solve at all the "shadow" graphic corruption.
Patch ready.
3/ Graphic corruption.
Some progress. Contrary to what I thought, this is not caused a stencil buffer issue, but by the ZBuffer being filled upside down, during the first phase of frame drawing.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #7 from Norbert Lataille nonal@freesurf.fr 2007-11-24 10:05:59 --- Created an attachment (id=9312) --> (http://bugs.winehq.org/attachment.cgi?id=9312) The backbuffer
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #8 from Norbert Lataille nonal@freesurf.fr 2007-11-24 10:07:34 --- Created an attachment (id=9313) --> (http://bugs.winehq.org/attachment.cgi?id=9313) The ZBuffer for the screenshot
The ZBuffer for the screenshot (this is looking at the cause of our garphic corruption)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #9 from H. Verbeet hverbeet@gmail.com 2007-11-25 08:47:26 --- Does using a different offscreen rendering mode make any difference?
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #10 from Norbert Lataille nonal@freesurf.fr 2007-11-25 09:23:46 --- Hi,
I will test different registry options, and report the difference.
This seems clearly linked to the issue as: - the drawing phase when my Zbuffer is updated "upside down" is the only one which has an "on-screen context" (Projection matrix is not multiplied by invymat) - rest of the frame is drawn in an "off-screen" context AFAIK.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #11 from Norbert Lataille nonal@freesurf.fr 2007-11-25 14:47:45 --- Unfortunately, backbuffer and pbuffers both produce zbuffer glitches. FBO produce a full black screen.
However, with pbuffers sky is no longer full of "black" sections, reflecting the ground shape (probably means the first rendering phase is now correct), but there are other glitches... I will have a look a projection matrix to see how it changes between both trys.
I need to study more the depth/stencil buffer implementation in wine before going on...
http://bugs.winehq.org/show_bug.cgi?id=9717
Tobias Lieske samusin@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |samusin@gmx.net
--- Comment #12 from Tobias Lieske samusin@gmx.net 2007-11-26 07:35:02 --- nice to hear about your progress. go on. your my hero :D
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #13 from Rico kgbricola@web.de 2007-12-01 11:36:57 --- @Norbert Lataille nice to hear, too. If you need some help, please feel free to say this.
There seem to be the same issue in Crysis and 3DMark05 Game1 test. There are also some glitches, which could look like the same bug.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #14 from Norbert Lataille nonal@freesurf.fr 2007-12-04 01:13:53 --- Hi,
Some news here.
I have a patch looking like a workaround for this bug (on this game). Still looking for a complete fix, though. I will have a look at Crysis bug reports to see if this is looking like the same issue.
Now that I can test it, I am disappointed by performance of this game when looking at some directions/zones.
It does not seem to be related to the workaround, though (my windows box is also showing slowdowns in some regions/directions). So, possibly the game is un-optimized, and wine layer is making things worse.
I am installing oprofile to get a status report, and see if I can do something on this front.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #15 from Tobias Lieske samusin@gmx.net 2007-12-04 14:08:51 --- yeeeeha! thanks a lot man! now i gotta hurry up finishing gothic 2 + addon :D you rock man!
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #16 from Rico kgbricola@web.de 2007-12-09 08:14:12 --- @ Norbert Lataille Could you put the workaround here so we can take a look at it ;-).
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #17 from Tobias Lieske samusin@gmx.net 2007-12-12 13:22:37 --- so i just found out that norbert already has uploaded his patch ... i was wondering why he didnt give us a link ... then i remembered a comment with link to wine-patches and i took a closer look at the patches in decemeber and ah there it is :D http://www.winehq.org/pipermail/wine-patches/2007-December/047462.html i couldnt try it out cause i found it one minute ago ^^
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #18 from Rico kgbricola@web.de 2007-12-12 15:52:33 --- @Tobias Lieske This patch copy the right size of data to the scissorRect so that this fixmes disappear.
fixme:d3d:scissorrect >>>>>>>>>>>>>>>>> GL_INVALID_VALUE (0x501) from glScissor @ state.c / 3714
So this is a good patch but doesn't solve the main problem.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #19 from Norbert Lataille nonal@freesurf.fr 2007-12-13 01:11:01 --- Sorry guys, I am busy on evenings these days...
What I need to do is to move my work to the last wine revision before sending a patch (everything was based on .48 here). At least to check the workaround still works on .50/HEAD, or that the underlying issue has not changed.
I will try to reserve time this evening, and attach the workaround patch this evening so that you can check if it is also fixing your sky issue.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #20 from Norbert Lataille nonal@freesurf.fr 2007-12-13 14:38:04 --- Created an attachment (id=9618) --> (http://bugs.winehq.org/attachment.cgi?id=9618) actxtx patch (if you have DLL load errors)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #21 from Norbert Lataille nonal@freesurf.fr 2007-12-13 14:38:57 --- Created an attachment (id=9619) --> (http://bugs.winehq.org/attachment.cgi?id=9619) workaround for sky issues (not a clean fix !)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #22 from Norbert Lataille nonal@freesurf.fr 2007-12-13 14:40:46 --- Created an attachment (id=9620) --> (http://bugs.winehq.org/attachment.cgi?id=9620) Result with workaround, in medium quality
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #23 from Norbert Lataille nonal@freesurf.fr 2007-12-13 14:44:26 --- When I have more time,I will probably focus on the performance of this game (is frame rendering using wine slow path ?), as I don't believe that I can play this game today on my hardware :(
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #24 from Tobias Lieske samusin@gmx.net 2007-12-13 14:51:50 --- ah ... and i thought i did an mistake when i started gothic after patching and recompiling wine and the sky was black :D ... i ll try the new patch tommorow and again thanks to robert for your great work!
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #25 from Rico kgbricola@web.de 2007-12-14 05:30:57 --- This workaround is great and soleved the main issue particularly in Gothic3 and Crysis.
The speed is acceptable @ high details. On low it is also good. But the problem isn't fixed completely. I'll attach a screen shot later.
In Gothic3 there are more issues to get this running. There seems to be a memory issue (when Gothic3 takes about 1.4GB of memory in high details - after playing time < 5 min): err:d3d:CreateIndexBufferVBO Failed to initialize the index buffer with error GL_OUT_OF_MEMORY (0x505)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #26 from Rico kgbricola@web.de 2007-12-14 05:48:34 --- Created an attachment (id=9622) --> (http://bugs.winehq.org/attachment.cgi?id=9622) mirror seen in the smoke
Mirror in the smoke of the flame (at low detail) ... on this screen shot it isn't seen so good, but in game you can see it when you are moving the mouse.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #27 from Norbert Lataille nonal@freesurf.fr 2007-12-14 12:08:09 ---
This workaround is great and soleved the main issue particularly in Gothic3 and Crysis.
This is interesting news... these 2 graphic corruption may have the same underlying bug.
The speed is acceptable @ high details. On low it is also good. But the problem isn't fixed completely. I'll attach a screen shot later.
Yes, I was expecting it. This is clearly not a fix for the cause, just a hack to help making the game playable while the bug is being investigated.
On the perf front, I found 1/2 things in the oprofile report I would like to test in order to improve rendering performance. But it does not seem this game is providing a native FPS counter, it will be hard to test.
It seems that most of my slowdowns are due to hard drive loading phases. I am wondering why this game needs to load that many entities while I am just spinning on myself. Worth a look...
In Gothic3 there are more issues to get this running. There seems to be a memory issue (when Gothic3 takes about 1.4GB of memory in high details - after playing time < 5 min): err:d3d:CreateIndexBufferVBO Failed to initialize the index buffer with error GL_OUT_OF_MEMORY (0x505)
May be easy to look at (is this related to a memory leak ? I noticed that Heap management routines are called A LOT !!!) On my side I always end up with an exception.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #28 from Rico kgbricola@web.de 2007-12-14 13:42:59 --- The game has a native frame counter. Therefore you have to aktivate the cheatmode. This can be realized by typing "marvin" very fast while the game runs. Then you can select in the menu "Profile" -> "FPS" and the fps are displayed in the left bottom corner. But don't forget to disable the cheat afterwards (type "marvin" again). More information can be found here: http://www.worldofgothic.de/gothic3/index.php?go=cheats (german)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #29 from Rico kgbricola@web.de 2007-12-14 13:54:54 ---
May be easy to look at (is this related to a memory leak ? I noticed that Heap management routines are called A LOT !!!) On my side I always end up with an exception.
Easy? Where do I have to look at? Or how could I check if it is a memory leak? Played on low details there was no crash (over 30 minutes).
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #30 from Norbert Lataille nonal@freesurf.fr 2007-12-14 16:26:07 --- (In reply to comment #28)
The game has a native frame counter. Therefore you have to aktivate the cheatmode. This can be realized by typing "marvin" very fast while the game runs. Then you can select in the menu "Profile" -> "FPS" and the fps are displayed in the left bottom corner. But don't forget to disable the cheat afterwards (type "marvin" again). More information can be found here: http://www.worldofgothic.de/gothic3/index.php?go=cheats (german)
Hum, thanks... Looking at the logs there are a few things worth optimizing. But I would prefer to check this is not over-engineering.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #31 from Norbert Lataille nonal@freesurf.fr 2007-12-14 16:32:34 ---
Easy? Where do I have to look at? Or how could I check if it is a memory leak? Played on low details there was no crash (over 30 minutes).
First thing you can do is from time to time check the memory size of the process. top, ps auxw, cat /proc/PID/status all show vmRSS exploding => few gigs.
On my side, I can confirm there is a memory leak in wined3d, visible in wine logs. I am trying to spot which source files are concerned.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #32 from Rico kgbricola@web.de 2007-12-15 04:06:28 --- Gothic3 crashes when the memory is about VIRT=4094mb, RES=1.3gb SHR=21mb. (high details)
It crashes on middle details at about the same amount of memory ... but this takes around 30 minutes.
I will try it on low details but this could take some hours.
So is there a maximum of VIRTUAL MEMORY for 32bit apps (4gb)?
Here is the backtrace: =>1 0x7bc390ac __regs_RtlRaiseException+0x4c(rec=0xeadbc8, context=0xead858) [/home/ricola/Desktop/wine/wine-git/dlls/ntdll/exception.c:396] in ntdll (0x00ead848) 2 0x7bc739e3 __wine_call_from_32_regs+0xc3() in ntdll (0x00eadba4) 3 0x7bc38676 RtlRaiseException+0x6() in ntdll (0x00eadc1c) 4 0x445d73ac in ge3dialogs (+0x173ac) (0x00eadc54) 5 0x445c6f13 in ge3dialogs (+0x6f13) (0x00eadc68) 6 0x44610338 (0x00eadc98) 7 0x445c209b in ge3dialogs (+0x209b) (0x00eadd10) 8 0x445d597e in ge3dialogs (+0x1597e) (0x00eadd38) 9 0x445d55b6 in ge3dialogs (+0x155b6) (0x00eadd78) 10 0x445d5620 in ge3dialogs (+0x15620) (0x00eadda0) 11 0x7bc43dcd MODULE_InitDLL+0x8d(wm=<register EDI not in topmost frame>, reason=<register ESI not in topmost frame>, lpReserved=0x1) [/home/ricola/Desktop/wine/wine-git/dlls/ntdll/loader.c:909] in ntdll (0x00eade30) 12 0x7bc4420f process_detach+0x6f(bForceDetach=0x1, lpReserved=<register EDI not in topmost frame>) [/home/ricola/Desktop/wine/wine-git/dlls/ntdll/loader.c:1054] in ntdll (0x00eade50) 13 0x7b86f48f ExitProcess+0x1f(status=0x1bf) [/home/ricola/Desktop/wine/wine-git/dlls/kernel32/process.c:1947] in kernel32 (0x00eade70) 14 0x00f525f1 in sharedbase (+0xa25f1) (0x00eadeb0) 15 0x00f5294f in sharedbase (+0xa294f) (0x00eadef8)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #33 from Rico kgbricola@web.de 2007-12-15 04:36:18 --- After warping around I often (always) get this issue:
err:seh:setup_exception stack overflow 0 bytes in thread 0025 eip 7e606e13 esp 6583e000 stack 0x6583e000-0x6604e000 err:ntdll:RtlpWaitForCriticalSection section 0x7e67ee68 "d3d9_main.c: d3d9_cs" wait timed out in thread 0010, blocked by 0025, retrying (60 sec)
Should I make an extra bug report for this thinks which doesn't have to do with the mirror?
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #34 from Norbert Lataille nonal@freesurf.fr 2007-12-15 09:11:09 --- Looks like this is an extra bug report, yes. There are two different memory leaks in the d3d9, wined3d code.
I've got something looking like half a patch, but wine is still leaking.
I am impressed how many mallocs/free are done by the wine layer ! No surprise that oprofile tells me how big an offender this is...
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #35 from Rico kgbricola@web.de 2007-12-15 09:40:16 --- Say a word if I should do some tests or logs ;-).
I opened an extra bug for the warp. http://bugs.winehq.org/show_bug.cgi?id=10778
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #36 from Rico kgbricola@web.de 2007-12-17 07:07:41 --- I have done another try ... at hight details ... run the game about 30 min. When entering a second town (big one) the game crashed with this message: (0) : fatal error C9008: out of memory - malloc failed Cg compiler terminated due to fatal error (I think for now the message is irrelevant because it differs from time to time and has to do where the OUT_OF_MEMORY appears first.
Here are the cat /proc/4789/status (short before the crash): Name: Gothic3.exe State: R (running) Tgid: 4789 Pid: 4789 PPid: 3162 TracerPid: 0 Uid: 500 500 500 500 Gid: 500 500 500 500 FDSize: 256 Groups: 500 VmPeak: 4181804 kB VmSize: 4175684 kB VmLck: 6104 kB VmHWM: 1592296 kB VmRSS: 1592296 kB VmData: 4126348 kB VmStk: 84 kB VmExe: 8 kB VmLib: 40532 kB VmPTE: 3468 kB StaBrk: 7c003000 kB Brk: 7c379000 kB StaStk: ffadf9a0 kB Threads: 11 SigQ: 0/53248 SigPnd: 0000000000000000 ShdPnd: 0000000000000000 SigBlk: 0000000000000000 SigIgn: 0000000000011000 SigCgt: 00000001800046fa CapInh: 0000000000000000 CapPrm: 0000000000000000 CapEff: 0000000000004000 Cpus_allowed: 00000000,0000000f Mems_allowed: 00000000,00000001
and here is a second one: cat /proc/5683/status Name: Gothic3.exe State: S (sleeping) Tgid: 5683 Pid: 5683 PPid: 3162 TracerPid: 0 Uid: 500 500 500 500 Gid: 500 500 500 500 FDSize: 256 Groups: 500 VmPeak: 4178684 kB VmSize: 4172564 kB VmLck: 6104 kB VmHWM: 1584748 kB VmRSS: 1583072 kB VmData: 4122948 kB VmStk: 84 kB VmExe: 8 kB VmLib: 40784 kB VmPTE: 3476 kB StaBrk: 7c003000 kB Brk: 7c379000 kB StaStk: ffe35d00 kB Threads: 11 SigQ: 0/53248 SigPnd: 0000000000000000 ShdPnd: 0000000000000000 SigBlk: 0000000000000000 SigIgn: 0000000000011000 SigCgt: 00000001800046fa CapInh: 0000000000000000 CapPrm: 0000000000000000 CapEff: 0000000000004000 Cpus_allowed: 00000000,0000000f Mems_allowed: 00000000,00000001 voluntary_ctxt_switches: 673663 nonvoluntary_ctxt_switches: 1276990
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #37 from Rico kgbricola@web.de 2007-12-17 10:20:11 --- After waiting 2h in the menu the VmRss increased up to 1.6GB and the games doesn't crash. At about 1.6GB the VmData and VmSize are also increasing.
After 2:15 the the game crashed width this stats: cat /proc/12744/status Name: Gothic3.exe State: R (running) Tgid: 12744 Pid: 12744 PPid: 3162 TracerPid: 0 Uid: 500 500 500 500 Gid: 500 500 500 500 FDSize: 256 Groups: 500 VmPeak: 4171620 kB VmSize: 4166640 kB VmLck: 4900 kB VmHWM: 1834488 kB VmRSS: 1834488 kB VmData: 4117520 kB VmStk: 84 kB VmExe: 8 kB VmLib: 40548 kB VmPTE: 3792 kB StaBrk: 7c003000 kB Brk: 7c379000 kB StaStk: ffe40500 kB Threads: 10 SigQ: 0/53248 SigPnd: 0000000000000000 ShdPnd: 0000000000000000 SigBlk: 0000000000000000 SigIgn: 0000000000011000 SigCgt: 00000001800046fa CapInh: 0000000000000000 CapPrm: 0000000000000000 CapEff: 0000000000004000 Cpus_allowed: 00000000,0000000f Mems_allowed: 00000000,00000001 voluntary_ctxt_switches: 6553959 nonvoluntary_ctxt_switches: 6767157
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #38 from Rico kgbricola@web.de 2007-12-30 09:01:20 --- @Norbert
Could you send this patch ( http://www.winehq.org/pipermail/wine-patches/2007-December/047462.html )to the wine-patches. It isn't in the git yet? Or am I wrong?
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #39 from Rico kgbricola@web.de 2008-01-16 14:41:54 --- Are there any news for this bug or any hints where to search? I am running out of idears.
@Norbert The patch isn't in git yet. Should I resend it or would you like to do this?
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #40 from Rico kgbricola@web.de 2008-02-02 13:17:30 --- Looking around in the bugs I've found an intresting post http://bugs.winehq.org/show_bug.cgi?id=6351#c18 which could be the reason for this glitch.
http://bugs.winehq.org/show_bug.cgi?id=9717
Florian mcflow-forum@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mcflow-forum@gmx.net
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #41 from Rico kgbricola@web.de 2008-03-12 15:50:48 --- Created an attachment (id=11343) --> (http://bugs.winehq.org/attachment.cgi?id=11343) flicker in the screen
Switching to fbo produces now a good picture, but there seems to be a problem with the z-buffer. The screen flickers and some objects were drawn/ not drawn when the player moves around.
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #11343|0 |1 is obsolete| |
--- Comment #42 from Rico kgbricola@web.de 2008-03-12 15:54:19 --- Created an attachment (id=11344) --> (http://bugs.winehq.org/attachment.cgi?id=11344) licker in the screen
Sorry, the last picture was wrong ...
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #43 from Florian mcflow-forum@gmx.net 2008-03-14 12:04:03 --- Created an attachment (id=11379) --> (http://bugs.winehq.org/attachment.cgi?id=11379) Transparent Textures
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #44 from Florian mcflow-forum@gmx.net 2008-03-14 12:06:46 --- Sometimes I get something like "transparent" Textures. Some objects can be seen through other objects, like the mushrooms in this picture. This happens with all kinds of objects. An annoying thing in Forrests is, that trees and other plants appear and disappear. It's a kind of "blinking". This seems to kill performance, too. I am not quite sure if this is related to this bug here.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #45 from Rico kgbricola@web.de 2008-04-01 15:48:12 --- Created an attachment (id=11790) --> (http://bugs.winehq.org/attachment.cgi?id=11790) use the number of instances from the first stream not from the last one
An annoying thing in forrests is, that trees and other plants appear and
disappear. It's a kind of "blinking".
Could you try the attached patch? I am working on a test which figures out the right behavior, but this could take a while.
Some objects can be seen
through other objects, like the mushrooms in this picture.
"Set the in game setting "DistanceLow" to 0 (off) otherwise there are graphic glitches." Could you try this, too?
Both have nothing to do with this bug!
http://bugs.winehq.org/show_bug.cgi?id=9717
Mariusz magmp1@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |magmp1@gmail.com
--- Comment #46 from Mariusz magmp1@gmail.com 2008-05-24 12:34:25 --- (In reply to comment #45)
Created an attachment (id=11790)
--> (http://bugs.winehq.org/attachment.cgi?id=11790) [details]
use the number of instances from the first stream not from the last one
An annoying thing in forrests is, that trees and other plants appear and
disappear. It's a kind of "blinking".
Could you try the attached patch? I am working on a test which figures out the right behavior, but this could take a while.
Some objects can be seen
through other objects, like the mushrooms in this picture.
"Set the in game setting "DistanceLow" to 0 (off) otherwise there are graphic glitches." Could you try this, too?
Both have nothing to do with this bug!
Hello. Do you know if this patch can be implemented to wine 1.0 rc2? If so, could you tell me how?
Thanks.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #47 from Rico kgbricola@web.de 2008-05-24 15:37:01 --- The patch is in wine since 0.9.60 and solves the "blinking" effect . Nothing more! The problem in this bug isn't fixed! I really don't know if this could be fixed due to the limitations of the backbuffer. So switching OffscreenRenderingMode to fbo ( http://wiki.winehq.org/UsefulRegistryKeys ) solves the problem in this bug.
If you have any questions regarding the game please look in the AppDB ( http://appdb.winehq.org/objectManager.php?sClass=version&iId=6012 ) or use the Wine Forum ( http://forum.winehq.org/ ).
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #48 from Florian mcflow-forum@gmx.net 2008-11-02 08:47:39 --- Created an attachment (id=17045) --> (http://bugs.winehq.org/attachment.cgi?id=17045) workaround for sky issues (not a clean fix !)
The Patch now will compile with wine-1.1.7 again.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #49 from H. Verbeet hverbeet@gmail.com 2008-11-02 09:09:46 --- So what is the current status of this bug?
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #50 from Rico kgbricola@web.de 2008-11-02 09:33:49 --- This bug is only with OffscreenRenderingMode = backbuffer and it isn't solved yet. This could be the reason http://bugs.winehq.org/show_bug.cgi?id=6351#c18 . Probably we should mark it as wont fix because there is only a dirty workaround and no real way to solve it?
Using fbo let the game look fine (some polygons are flickering).
Using pbuffer then the black shadows are in a total wrong position.
http://bugs.winehq.org/show_bug.cgi?id=9717
Christoph Korn c_korn@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |c_korn@gmx.de
--- Comment #51 from Christoph Korn c_korn@gmx.de 2009-03-26 16:24:55 --- Is there a proper patch to be expected in the near future?
I would try to code it myself if I had any knowledge :-(
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #52 from NSLW lukasz.wojnilowicz@gmail.com 2009-05-17 02:00:07 --- Created an attachment (id=21137) --> (http://bugs.winehq.org/attachment.cgi?id=21137) workaround for sky issues (not a clean fix !) for >=Wine 1.1.21
The patch still fixes sky issue
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #53 from NSLW lukasz.wojnilowicz@gmail.com 2009-05-17 02:06:17 --- Created an attachment (id=21138) --> (http://bugs.winehq.org/attachment.cgi?id=21138) actxtx patch (if you have DLL load errors) for >=Wine 1.1.21
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #54 from Tobias Lieske samusin@gmx.net 2009-05-21 07:26:38 --- Hi NSLW when i use your recommended patches to play gothic 3 i cant build wine. when make reaches the mouse patch it stops, because it expects some ")" ";" or "(". then next try i left out the mouse patch then it stopped at the device.c file. building wine without patches workes fine. i am using wine source 1.1.21 now i dont know what to do :(
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #55 from Tobias Lieske samusin@gmx.net 2009-05-21 11:59:16 --- Created an attachment (id=21219) --> (http://bugs.winehq.org/attachment.cgi?id=21219) make error with patch #21137 applied
this is the error i get, when i try to compilie with the sky workaround patch
http://bugs.winehq.org/show_bug.cgi?id=9717
Zerg2000 zerg2000@astral.org.pl changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |zerg2000@astral.org.pl
--- Comment #56 from Zerg2000 zerg2000@astral.org.pl 2009-05-21 13:21:55 --- Quotes in patch #21137 are nice UTF encoded characters. :-P
Replace them by "normal" ASCII characters in the following line: checkGLcall(”glClearDepth”);
” => "
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #57 from Tobias Lieske samusin@gmx.net 2009-05-21 15:16:24 --- oooooooh noooooowez ^^
okay... after editing, the patch works... and gothic 3 works AND looks soooo amazing! nice work NSLW!!!! thanks for the patch
and thanks to Zerg2000 ^^
http://bugs.winehq.org/show_bug.cgi?id=9717
NSLW lukasz.wojnilowicz@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |lukasz.wojnilowicz@gmail.co | |m
--- Comment #58 from NSLW lukasz.wojnilowicz@gmail.com 2009-05-22 12:54:14 --- (In reply to comment #57)
Credits belong to Norbert Lataille. I only updated the patch to be compatible with 1.1.21 and nothing more.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #59 from Christoph Korn c_korn@gmx.de 2009-08-07 16:25:53 --- (In reply to comment #52)
Created an attachment (id=21137)
--> (http://bugs.winehq.org/attachment.cgi?id=21137) [details]
workaround for sky issues (not a clean fix !) for >=Wine 1.1.21
The patch still fixes sky issue
The patch does not work in wine-1.1.27 any longer: device.c: In function ‘IWineD3DDeviceImpl_UpdateTexture’: device.c:5485: error: ‘glMask’ undeclared (first use in this function) device.c:5485: error: (Each undeclared identifier is reported only once device.c:5485: error: for each function it appears in.)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #60 from Zerg2000 zerg2000@astral.org.pl 2009-08-07 18:42:39 --- Created an attachment (id=22894) --> (http://bugs.winehq.org/attachment.cgi?id=22894) Workaround for sky issues (not a clean fix !) for >=Wine 1.1.27
Workaround for sky issues adapted to Wine 1.1.27. It compiles now but I didn't check if it works in game.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #61 from Christoph Korn c_korn@gmx.de 2009-08-07 18:47:27 --- (In reply to comment #60)
Created an attachment (id=22894)
--> (http://bugs.winehq.org/attachment.cgi?id=22894) [details]
Workaround for sky issues (not a clean fix !) for >=Wine 1.1.27
Workaround for sky issues adapted to Wine 1.1.27. It compiles now but I didn't check if it works in game.
Yes, it works. Thanks.
The old patch just was applied in the wrong line. Does not happen very often that a patch perfectly fits in a wrong line.
http://bugs.winehq.org/show_bug.cgi?id=9717
knan-wine@anduin.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |knan-wine@anduin.net
--- Comment #62 from knan-wine@anduin.net 2009-11-06 17:13:24 --- The sky issues fixhack also fixes graphics issues for me in Dragon Age. For entirely the wrong reasons, I suspect. I get tons and tons of
fixme:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 6172
but in exchange, the misplaced shadowy artefacts are gone. (You need the hack from #20469 first, to avoid various flickering blackness.)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #63 from NSLW lukasz.wojnilowicz@gmail.com 2010-03-08 03:53:36 --- This is still an issue in Wine 1.1.40. Workaround still works.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #64 from NSLW lukasz.wojnilowicz@gmail.com 2010-03-21 05:54:28 --- Created an attachment (id=26944) --> (http://bugs.winehq.org/attachment.cgi?id=26944) Quarter of the sky
The shadow showed in attachment #8141 sometimes looks like in attached image. It started to look like this in Wine 1.1.41. Workaround for sky issues doesn't work no more and shows following error at compilation:
device.c: In function ‘dirtify_p8_texture_samplers’: device.c:4975: error: ‘else’ without a previous ‘if’ device.c:4979: error: ‘glMask’ undeclared (first use in this function) device.c:4979: error: (Each undeclared identifier is reported only once device.c:4979: error: for each function it appears in.) device.c:4980: error: ‘This’ undeclared (first use in this function) make[1]: *** [device.o] Error 1 make[1]: Leaving directory `/home/nslw/Desktop/wine-1.1.41/dlls/wined3d' make: *** [dlls/wined3d] Error 2
ORM=backbuffer NVIDIA 9xxx with 195.36.15
http://bugs.winehq.org/show_bug.cgi?id=9717
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #22894|0 |1 is obsolete| |
--- Comment #65 from Austin English austinenglish@gmail.com 2010-03-30 03:47:34 --- Created an attachment (id=27101) --> (http://bugs.winehq.org/attachment.cgi?id=27101) updated patch
Rediff'ed patch.
http://bugs.winehq.org/show_bug.cgi?id=9717
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |patch
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #66 from NSLW lukasz.wojnilowicz@gmail.com 2010-04-03 06:56:10 --- (In reply to comment #65)
Created an attachment (id=27101)
--> (http://bugs.winehq.org/attachment.cgi?id=27101) [details]
updated patch
Rediff'ed patch.
It's still an issue in Wine 1.1.42. Your rediff'ed patch works around that issue. Thanks
http://bugs.winehq.org/show_bug.cgi?id=9717
Xavier Vachon xvachon@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |xvachon@gmail.com
--- Comment #67 from Xavier Vachon xvachon@gmail.com 2010-07-10 12:12:57 --- Does this bug happens in particular instances or throughout the entire game?
When I tested in current git (1.2-rc7), I did not experience this problem.
Can some of the testers try again and report?
Using Gothic 3 1.73 with winetricks d3dx9 and nvidia-drivers 256.35.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #68 from Rico kgbricola@web.de 2010-07-10 13:09:07 --- This happens the entire game, but only with ORM=backbuffer and it is still an issue.
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Gothic 3 - shadow on the |Gothic 3 - shadow on the |upper half of the screen |upper half of the screen | |(ORM=backbuffer)
http://bugs.winehq.org/show_bug.cgi?id=9717
kostindima@yandex.ru kostindima@yandex.ru changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |kostindima@yandex.ru
--- Comment #69 from kostindima@yandex.ru kostindima@yandex.ru 2010-10-10 04:07:37 CDT --- (In reply to comment #68)
This happens the entire game, but only with ORM=backbuffer and it is still an issue.
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #70 from kostindima@yandex.ru kostindima@yandex.ru 2010-10-10 04:09:12 CDT --- Created an attachment (id=31205) --> (http://bugs.winehq.org/attachment.cgi?id=31205) Patch rediffed for 1.3.4
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #71 from t.kijas@gmail.com 2011-05-12 20:11:05 CDT --- Created an attachment (id=34671) --> (http://bugs.winehq.org/attachment.cgi?id=34671) Debug information from wine 1.3.4
I think it doesn't work anymore. I compiled 1.3.4 with updated patch, installed it and run - via PlayOnLinux it does show still top half of screen in black, via wine in terminal it doesn't even show main menu.
I've added debug output as attachment.
http://bugs.winehq.org/show_bug.cgi?id=9717
t.kijas@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #34671|debug |debug.txt filename| |
http://bugs.winehq.org/show_bug.cgi?id=9717
joaopa jeremielapuree@yahoo.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jeremielapuree@yahoo.fr
--- Comment #72 from joaopa jeremielapuree@yahoo.fr 2011-05-12 21:12:27 CDT --- Maybe this bug could be marked as WONTFIX, since the default ORM is FBO now.
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #34671|application/octet-stream |text/plain mime type| |
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |WONTFIX
--- Comment #73 from Rico kgbricola@web.de 2011-05-13 02:40:48 CDT --- Marking as WONTFIX, because it seems to be impossible to do the correct rendering with ORM=backbuffer. The attached hack doesn't solve the problem, it only moves the problem in a way, that it couldn't be seen that well. The preferred rendering method is fbo. Thus you won't even see this bug.
http://bugs.winehq.org/show_bug.cgi?id=9717
t.kijas@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |t.kijas@gmail.com
--- Comment #74 from t.kijas@gmail.com 2011-05-13 03:17:43 CDT --- Yes but it is still the only solution how to make the game playable, because for example for me the fbo with disabled low distance view doesn't work. (proprietary NVIDIA drivers in latest version, NV GT240M)
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #75 from Rico kgbricola@web.de 2011-05-13 04:57:28 CDT --- The bug which should be solved by "distance low" is somehow gone (see http://bugs2.winehq.org/attachment.cgi?id=20131 ). Do you still get a picture like that? What you are seeing now is probably bug 23012 .
http://bugs.winehq.org/show_bug.cgi?id=9717
--- Comment #76 from t.kijas@gmail.com 2011-05-13 05:01:42 CDT --- Yes you are right. But I must say that I gave a second chance to use a patch in this bug report and run the game and voila - it works - no more black sky:) just some strange fog around the main hero, but nothing as bad as the black sky was. With backbuffer and latest wine 1.3.19 patched with rediffed patch for 1.3.4.
http://bugs.winehq.org/show_bug.cgi?id=9717
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #77 from Austin English austinenglish@gmail.com 2011-05-16 04:34:25 CDT --- Closing.
http://bugs.winehq.org/show_bug.cgi?id=9717
Rico kgbricola@web.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #11344| licker in the screen |flicker in the screen description| |
-
wine-bugs＠winehq.org