http://bugs.winehq.org/show_bug.cgi?id=29907
Bug #: 29907 Summary: DungeonSeigeDemo Fullscreen missing upper part of screen Product: Wine Version: unspecified Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: SolisX007@yahoo.com Classification: Unclassified
Created attachment 38897 --> http://bugs.winehq.org/attachment.cgi?id=38897 Regression Test on Dungeon Seige Demo
Hello Everyone,
I'm reporting a bug on Dungeon Seige Demo. The problem is that when a play the game in "Fullscreen" the game is missing the upper part of the screen. The regression started in wine-1.4-rc1. On wine-1.3.37 the game was fine. I also did a regression test which i have attached to this post.
http://bugs.winehq.org/show_bug.cgi?id=29907
Lunknown SolisX007@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|unspecified |1.4-rc3
http://bugs.winehq.org/show_bug.cgi?id=29907
Lunknown SolisX007@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- URL| |http://www.fileplanet.com/8 | |7223/download/Dungeon-Siege | |-Demo
http://bugs.winehq.org/show_bug.cgi?id=29907
Vitaliy Margolen vitaliy-bugzilla@kievinfo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download, regression Component|-unknown |directx-d3d Regression SHA1| |8692ff48a953f368860967c1141 | |ec875adba9aaf
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #1 from Dmitry Timoshkov dmitry@baikal.ru 2012-02-15 23:33:52 CST --- 8692ff48a953f368860967c1141ec875adba9aaf is the first bad commit commit 8692ff48a953f368860967c1141ec875adba9aaf Author: Henri Verbeet hverbeet@codeweavers.com Date: Fri Jan 20 00:36:26 2012 +0100
wined3d: Don't load the draw_binding location in surface_unmap() unless the surface is on a swapchain.
It really makes no sense to do this just because a surface happens to be the current render target. In particular, this patch prevents needlessly bouncing the surface between the CPU and the GPU when the application does e.g. multiple colorkeyed blits to the current render target to draw a HUD.
:040000 040000 9016e62637f5a313868c17e4229441bf2f4f3779 cb294d3ecc22aa2285a5c78b9e74747523da8014 M dlls
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #2 from Henri Verbeet hverbeet@gmail.com 2012-02-16 02:57:37 CST --- Seems to work fine for me. What driver is this? Does it work any better in a virtual desktop?
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #3 from Lunknown SolisX007@yahoo.com 2012-02-16 04:19:32 CST --- Hello Henri Verbeet,
The Game works fine if i use virtual desktop. It only happens when i'm in Full screen. I'm using the lastest nvidia proprietary driver which is 295.20. My video card is NVIDIA GT430 with 1024mb running on Slackware 13.37.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #4 from Henri Verbeet hverbeet@gmail.com 2012-02-16 07:39:47 CST --- I asked a couple of other people to try, but it works for them as well. Is there anything uncommon about your setup? E.g. the window manager, X server or screen resolution?
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #5 from Lunknown SolisX007@yahoo.com 2012-02-16 16:02:34 CST --- Hello Henri Verbeet,
I using icewm-1.2.38pre2 as my window manager i can upgrade to 1.3.7 to test if it's the window manager. My resolution is 1024x768. I will get back to you after i test icewm 1.3.7.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #6 from Lunknown SolisX007@yahoo.com 2012-02-17 08:30:32 CST --- Hello Henri Verbeet,
I tested icewm-1.3.7. The game still did not work right in fullscreen. I then decide to try openbox. The Game works fine in openbox in fullscreen. However i know that it works fine on icewm on wine-1.3.37 so i guess it's a regression that only effects icewm.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #7 from Henri Verbeet hverbeet@gmail.com 2012-02-21 14:52:03 CST --- Could you create a +win,+event log?
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #8 from Lunknown SolisX007@yahoo.com 2012-02-21 16:28:38 CST --- Created attachment 39011 --> http://bugs.winehq.org/attachment.cgi?id=39011 DungeonSeigeDemo WINEBUG=+win,+event.txt
Hello Henri Verbeet,
Okay i attached a WINEBUG=+win,+event log. I exited the demo at the title screen where i saw the regression.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #9 from Henri Verbeet hverbeet@gmail.com 2012-02-22 02:26:36 CST --- (In reply to comment #8)
trace:event:X11DRV_ConfigureNotify win 0x20022/2000001 new X rect 0,0,1024x768 (event 0,0,1024x768) trace:event:X11DRV_ConfigureNotify 0x20022 resizing from (800x600) to (1024x768)
It looks like the window manager resizes the window (back) to 1024x768. I'd guess that the missing upper part of the screen is related to the height difference between 800x600 and 1024x768. It's perhaps also interesting to note that fvwm without MwmFunctions shows similar behaviour, although it's not clear to me why. The part that doesn't make sense to me is what commit 8692ff4 has to do with this. If this is really a regression I would have expected a commit touching something like winex11.drv / user32 / gdi32, or at least something related to the window handling in wined3d. The commit you found looks completely unrelated, does reverting it make the bug go away again?
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #10 from Lunknown SolisX007@yahoo.com 2012-02-22 13:14:05 CST --- Hello Henri Verbeet,
If you tell me how to revert i will try it. I'm still new to this stuff.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #11 from Austin English austinenglish@gmail.com 2012-02-22 13:22:32 CST --- (In reply to comment #10)
Hello Henri Verbeet,
If you tell me how to revert i will try it. I'm still new to this stuff.
$ cd $HOME/wine-git/ $ git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R $ ./configure && make $ cd $PROGRAM_LOCATION $ ~/wine-git/wine foo.exe
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #12 from Lunknown SolisX007@yahoo.com 2012-02-22 14:10:19 CST --- Hello Austin English,
I had already deleted my wine-git directory. I am starting it again from the begining so give me about 2 hours to try the last step you posted.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #13 from Lunknown SolisX007@yahoo.com 2012-02-22 16:42:48 CST --- Hello Everyone,
Ok i finished the regression test again and got the same result. Now when i tried
git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R
it asked me
Unreversed patch detected! Ignore -R? [n] n
I chose n
Which then asked me
Apply anyway? [n] y
I chose y.
which then showed
Hunk #1 FAILED at 953. 1 out of 1 hunk FAILED -- saving rejects to file dlls/wined3d/surface.c.rej
then i did ./configure && make
after it compile i tested it and the demo was working fine. So look through those steps to make sure i did not do anything incorrect.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #14 from Austin English austinenglish@gmail.com 2012-02-22 17:33:28 CST --- (In reply to comment #13)
Hello Everyone,
Ok i finished the regression test again and got the same result. Now when i tried
git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R
it asked me
Unreversed patch detected! Ignore -R? [n] n
I chose n
Which then asked me
Apply anyway? [n] y
I chose y.
which then showed
Hunk #1 FAILED at 953. 1 out of 1 hunk FAILED -- saving rejects to file dlls/wined3d/surface.c.rej
then i did ./configure && make
after it compile i tested it and the demo was working fine. So look through those steps to make sure i did not do anything incorrect.
What commit are you at? (git show). It's likely not current git, so the 'bad' patch wasn't yet applied.
In any case, you could try applying it on your current commit and see if it fails: git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 && ./configure && make
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #15 from Lunknown SolisX007@yahoo.com 2012-02-22 17:56:28 CST --- Hello Austin English,
What i did was right after i found the bad commit in my wine-git directory i copied "git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R" to terminal. then compile it.
I did git show on my wine-git directory and this what is what was shown.
commit d1254eddb598ad2a6bddaf09ddb5d73247d460cc Author: Henri Verbeet hverbeet@codeweavers.com Date: Fri Jan 20 00:36:25 2012 +0100
wined3d: Properly enable / disable register combiners.
diff --git a/dlls/wined3d/nvidia_texture_shader.c b/dlls/wined3d/nvidia_texture_ shader.c index dc3f8b9..1fef980 100644 --- a/dlls/wined3d/nvidia_texture_shader.c +++ b/dlls/wined3d/nvidia_texture_shader.c @@ -629,15 +629,31 @@ static void nvrc_texfactor(struct wined3d_context *context , const struct wined3d }
/* Context activation is done by the caller. */ -static void nvrc_enable(BOOL enable) {} +static void nvrc_enable(BOOL enable) +{ + if (enable) + { + glEnable(GL_REGISTER_COMBINERS_NV); + checkGLcall("glEnable(GL_REGISTER_COMBINERS_NV)"); + } + else + { + glDisable(GL_REGISTER_COMBINERS_NV); + checkGLcall("glDisable(GL_REGISTER_COMBINERS_NV)"); + } +}
/* Context activation and GL locking are done by the caller. */ static void nvts_enable(BOOL enable) { - if(enable) { + nvrc_enable(enable); + if (enable) + { glEnable(GL_TEXTURE_SHADER_NV); checkGLcall("glEnable(GL_TEXTURE_SHADER_NV)"); - } else { + } + else + { glDisable(GL_TEXTURE_SHADER_NV); checkGLcall("glDisable(GL_TEXTURE_SHADER_NV)"); }
This not bad commit. How can i get to bad commit? So i can patch it.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #16 from Lunknown SolisX007@yahoo.com 2012-02-22 18:05:36 CST --- Hello Everyone,
Did i have to do git reset --hard before git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #17 from Austin English austinenglish@gmail.com 2012-02-23 12:48:15 CST --- (In reply to comment #16)
Hello Everyone,
Did i have to do git reset --hard before git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R
That would put you back at current wine, then remove the broken patch, so yes.
http://bugs.winehq.org/show_bug.cgi?id=29907
--- Comment #18 from Lunknown SolisX007@yahoo.com 2012-02-23 14:56:55 CST --- Hello Everyone,
Ok i did
git reset --hard 8692ff48a953f368860967c1141ec875adba9aaf ./configure --verbose && make
DungeonSeigedemo had regression.
then
git show 8692ff48a953f368860967c1141ec875adba9aaf | patch -p1 -R ./configure --verbose && make
DungeonSeigedemo did not have regression.
http://bugs.winehq.org/show_bug.cgi?id=29907
Frédéric Delanoy frederic.delanoy@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED CC| |frederic.delanoy@gmail.com Resolution| |DUPLICATE
--- Comment #19 from Frédéric Delanoy frederic.delanoy@gmail.com 2012-04-02 17:30:49 CDT --- Dupe
*** This bug has been marked as a duplicate of bug 30213 ***
http://bugs.winehq.org/show_bug.cgi?id=29907
Frédéric Delanoy frederic.delanoy@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #20 from Frédéric Delanoy frederic.delanoy@gmail.com 2012-04-02 17:31:09 CDT --- Closing.