http://bugs.winehq.org/show_bug.cgi?id=18728
Summary: Black screenshots in wow d3d with OffScreenRendering=fbo Product: Wine Version: 1.1.22 Platform: PC URL: http://www.worldofwarcraft.com OS/Version: Linux Status: UNCONFIRMED Severity: minor Priority: P2 Component: directx-d3d AssignedTo: wine-bugs@winehq.org ReportedBy: adys.wh+winehqdotorg@gmail.com
When you do an ingame screenshot in WoW running under d3d, if offscreenrendering is set to fbo, the following error is spit out: fixme:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 6994
The produced screenshot is completely black. This happens at least as early as wine-1.1.11 - haven't tested with earlier versions but I don't think it's a regression.
http://bugs.winehq.org/show_bug.cgi?id=18728
Adys adys.wh+winehqdotorg@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download
--- Comment #1 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-01 00:43:47 --- This happens even on the login screen and doesn't require an account to test, so adding download keyword.
http://bugs.winehq.org/show_bug.cgi?id=18728
Ken Sharp kennybobs@o2.co.uk changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|directx-d3d |-unknown Severity|minor |trivial
--- Comment #2 from Ken Sharp kennybobs@o2.co.uk 2009-06-01 04:24:23 --- Do you mean screenshots where you just hit the Print Screen key?
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #3 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-01 04:25:58 --- (In reply to comment #2)
Do you mean screenshots where you just hit the Print Screen key?
I mean the in-game screenshot system. External software will screenshot just fine.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #4 from Ken Sharp kennybobs@o2.co.uk 2009-06-01 04:29:53 --- Ah right, what is the command for that?
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #5 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-01 04:32:24 --- (In reply to comment #4)
Ah right, what is the command for that?
By default, it is bound to Print Screen. But if your external screenshot software is bound to that key as well, the in-game screenshot won't work. Try Shift-Print Screen.
http://bugs.winehq.org/show_bug.cgi?id=18728
Adys adys.wh+winehqdotorg@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|-unknown |directx-d3d
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #6 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-05 06:07:51 --- A patch was sent. http://www.winehq.org/pipermail/wine-patches/2009-June/073856.html
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #7 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-05 17:04:02 --- (In reply to comment #6)
A patch was sent. http://www.winehq.org/pipermail/wine-patches/2009-June/073856.html
Still happening in wine-1.1.23 with the patch applied.
fixme:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 6660
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #8 from Henri Verbeet hverbeet@gmail.com 2009-06-06 04:06:47 --- Do you know what kind of surfaces are involved? Alternatively, a +d3d9,+d3d,+d3d_surface log might be useful.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #9 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-06 05:27:51 --- Created an attachment (id=21596) --> (http://bugs.winehq.org/attachment.cgi?id=21596) +d3d9,+d3d,+d3d_surface wine wow -d3d
Started the game, took a screenshot and exited promptly.
Warning: 42MB uncompressed.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #10 from Henri Verbeet hverbeet@gmail.com 2009-06-06 16:49:58 --- Does it work if you disable multisampling?
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #11 from Adys adys.wh+winehqdotorg@gmail.com 2009-06-06 17:18:44 --- (In reply to comment #10)
Does it work if you disable multisampling?
Yes.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #12 from Henri Verbeet hverbeet@gmail.com 2009-06-07 15:31:19 --- I suspect this is a driver bug, but I'm not completely sure. I'll see what I can do to get this fixed, either in Wine or in the driver.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #13 from Jerome Leclanche adys.wh@gmail.com 2009-09-01 20:10:16 --- Still an issue in wine-1.1.28.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #14 from Jerome Leclanche adys.wh@gmail.com 2009-10-31 16:52:36 --- Still an issue in wine-1.1.32-260-gf222a16.
http://bugs.winehq.org/show_bug.cgi?id=18728
Nephyrin zey Nephyrin@nephyrin.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |Nephyrin@nephyrin.net
--- Comment #15 from Nephyrin zey Nephyrin@nephyrin.net 2009-11-29 23:35:22 --- Still an issue in 1.1.33
I suspect that this has something to do with games like Aion and TF2 being all black when FBO + Multisampling is enabled. Using Backbuffer + MS or FBO with no MS fixes it in all three for me.
(You can test this in Aion as well on the login screen without a account, but its a 10gig download)
http://bugs.winehq.org/show_bug.cgi?id=18728
jmartinot contrebasse@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #16 from jmartinot contrebasse@gmail.com 2009-12-20 09:08:43 --- *** This bug has been confirmed by popular vote. ***
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #17 from Jerome Leclanche adys.wh@gmail.com 2010-04-17 03:48:41 --- Still an issue in wine-1.1.43
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #18 from Jerome Leclanche adys.wh@gmail.com 2010-05-25 23:00:58 --- Ok, here's a list of the issues this causes in d3d which are fixed by changing the OSRM to backbuffer: * Screenshots are black * Death effect is black (ffxDeath = 1) * Fullscreen special effects are black with rendered borders (ffxSpecial = 1) * Fullscreen glow permanently renders the screen black (ffxGlow = 1) * Cataclysm underwater is black (liquidDetail = 3)
Henri, were you able to get more info on this?
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|Black screenshots in wow |Black screenshots and |d3d with |ffxEffects in wow d3d with |OffScreenRendering=fbo |OffScreenRendering=fbo
http://bugs.winehq.org/show_bug.cgi?id=18728
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |pandaroux@gmail.com
--- Comment #19 from Austin English austinenglish@gmail.com 2010-05-31 14:56:58 --- *** Bug 22970 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=18728
Ruslan b7.10110111@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |b7.10110111@gmail.com
--- Comment #20 from Ruslan b7.10110111@gmail.com 2010-05-31 15:07:53 --- *** Bug 20108 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #21 from pandaroux@gmail.com 2010-06-01 02:08:11 --- I have OffscreenRenderingMode to pbuffer and this error still exists in Wine 1.2rc2:
err:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 5844
Is this a NVidia specific bug ? ATI players have this error ?
http://bugs.winehq.org/show_bug.cgi?id=18728
Roderick Colenbrander thunderbird2k@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mischenko.mm@gmail.com
--- Comment #22 from Roderick Colenbrander thunderbird2k@gmail.com 2010-06-01 09:51:53 --- *** Bug 21937 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=18728
Roderick Colenbrander thunderbird2k@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |lubosz@gmail.com
--- Comment #23 from Roderick Colenbrander thunderbird2k@gmail.com 2010-06-01 09:52:39 --- *** Bug 20235 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=18728
white.phoenix@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |white.phoenix@gmail.com
--- Comment #24 from white.phoenix@gmail.com 2010-06-17 09:17:35 --- err:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 5843
disabling "Full-Screen Glow Effect" fixes the black login screen and black graphics. (tested on ubuntu 10.04 wine 1.2-rc3)
not sure about the screenshots, but this bug was marked a dupe of the other black screen bugs, so I'm reporting here instead.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #25 from white.phoenix@gmail.com 2010-06-17 09:19:45 --- so something is going wrong with the shaders that do the glow effect. but with it off, the GL_INVALID_OPERATION is gone. (also forgot to mention this was tested on nvidia)
if there's any other additional info i can provide that may help get the glow effect bug fixed, let me know
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #26 from Jerome Leclanche adys.wh@gmail.com 2010-06-17 09:30:15 --- (In reply to comment #25)
so something is going wrong with the shaders that do the glow effect. but with it off, the GL_INVALID_OPERATION is gone. (also forgot to mention this was tested on nvidia)
if there's any other additional info i can provide that may help get the glow effect bug fixed, let me know
Repeating a previous comment:
Ok, here's a list of the issues this causes in d3d which are fixed by changing the OSRM to backbuffer:
- Screenshots are black
- Death effect is black (ffxDeath = 1)
- Fullscreen special effects are black with rendered borders (ffxSpecial = 1)
- Fullscreen glow permanently renders the screen black (ffxGlow = 1)
- Cataclysm underwater is black (liquidDetail = 3)
Henri, were you able to get more info on this?
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #27 from Henri Verbeet hverbeet@gmail.com 2010-06-17 09:33:13 --- Still inclined to blame nvidia for this, but I haven't written a proper test case to prove it yet.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #28 from Henri Verbeet hverbeet@gmail.com 2010-06-18 10:29:42 --- Created an attachment (id=28934) --> (http://bugs.winehq.org/attachment.cgi?id=28934) test case
Driver bug, see attached test case. Apparently it doesn't like "stretching" from a multi-sampled back buffer.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #29 from Ruslan b7.10110111@gmail.com 2010-06-20 07:32:55 --- (In reply to comment #28)
Driver bug, see attached test case. Apparently it doesn't like "stretching" from a multi-sampled back buffer.
Have you reported it to nvidia?
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #30 from Jerome Leclanche adys.wh@gmail.com 2010-06-25 18:53:58 --- (In reply to comment #28)
Created an attachment (id=28934)
--> (http://bugs.winehq.org/attachment.cgi?id=28934) [details]
test case
Driver bug, see attached test case. Apparently it doesn't like "stretching" from a multi-sampled back buffer.
Still not fixed in 256.x
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |testcase Component|directx-d3d |-unknown
http://bugs.winehq.org/show_bug.cgi?id=18728
Roderick Colenbrander thunderbird2k@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |thunderbird2k@gmail.com
--- Comment #31 from Roderick Colenbrander thunderbird2k@gmail.com 2010-07-05 13:49:40 --- A month ago I submitted a bug for a related issue to Nvidia (a test case is here http://pastebin.com/5AAKHpW6). It was about blitting from a multisampled FBO to a non-multisampled window/fbo. The issue appears to be the same (well in the Nvidia drivers it triggers the same code path). If I disable multisampling in my test and enable multisampling in nvidia-settings, you can see the same bug as here but blitting also fails without any flipping.
Nvidia is working on fixing the issues but some parts seem trivial but some of the issues require some more changes in their driver. I have no idea when we will see a fix but I guess it will take a few months before it enters a stable release.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #32 from Henri Verbeet hverbeet@gmail.com 2010-07-05 14:16:32 --- No, that's different. In your case there's actually a multisampled framebuffer object involved. The system framebuffer object is always supposed to behave as a single-sampled buffer, regardless of any multisampling settings. I.e., if this test case did actual stretching it would still be supposed to work.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #33 from Roderick Colenbrander thunderbird2k@gmail.com 2010-07-05 14:30:40 --- The issue is only similar if you disable multisampling in my test and force multisampling in nvidia-settings else only the error is similar but in the drivers the code path is likely the same.
After the fixes the Nvidia guy is implementing, the Nvidia drivers will still be buggy in some cases. He said he won't handle cases like the ones you point to like blitting between a multisampled fbo (lets say at 4x) and a system framebuffer (lets say at 2x) in case you will still get a GL_INVALID_OPERATION error. Also the nvidia-settings override are not hidden to the application anyway. If you would query GL_SAMPLES_ARB it would show 16x if you select that.
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|-unknown |directx-d3d Summary|Black screenshots and |[NVIDIA BUG] WoW d3d: Black |ffxEffects in wow d3d with |screenshots/ffxEffects with |OffScreenRendering=fbo |OffScreenRendering=fbo
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #34 from Jerome Leclanche adys.wh@gmail.com 2010-09-10 01:02:06 CDT --- Unfortunately still an issue in 260.x
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #35 from Jerome Leclanche adys.wh@gmail.com 2010-09-15 09:20:02 CDT --- (In reply to comment #34)
Unfortunately still an issue in 260.x
However, it's not an issue in the 4.x WoW client.
http://bugs.winehq.org/show_bug.cgi?id=18728
Thomas Spear Speeddymon@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |Speeddymon@gmail.com
--- Comment #36 from Thomas Spear Speeddymon@gmail.com 2010-09-15 12:08:51 CDT --- Sometimes I wonder if Blizzard watches our bugs and secretly adds in fixes for some issues at the major release points.. ;-)
http://bugs.winehq.org/show_bug.cgi?id=18728
Christian Weeks cpw@weeksfamily.ca changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |cpw@weeksfamily.ca
--- Comment #37 from Christian Weeks cpw@weeksfamily.ca 2010-10-13 10:48:45 CDT --- Just wish to add that the 4.0.1 WoW release does NOT fix this issue, at least as released on live yesterday (Oct 12).
err:d3d:stretch_rect_fbo >>>>>>>>>>>>>>>>> GL_INVALID_OPERATION (0x502) from glBlitFramebuffer() @ device.c / 5843 (many hundreds of times, repeated)
I get completely black screen with just the login widgets rendered over the top. Running with the 195.36.24 NVIDIA driver, Quadro FX 2800M chipset. This is repeatable on my desktop, with a GTX260, same driver version (or 256 I can't remember).
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #38 from Jerome Leclanche adys.wh@gmail.com 2010-10-17 11:13:06 CDT --- Affects the new water in 4.x.
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #39 from Nephyrin zey Nephyrin@nephyrin.net 2010-11-11 20:19:17 CST --- Breaks water, FSAA in WoW without backbuffer mode (which breaks other stuff)
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #40 from Roderick Colenbrander thunderbird2k@gmail.com 2011-02-06 17:40:51 CST --- Nvidia fixed this bug in the 270 version of their drivers. My test samples also work correctly now. Though your mileage with enabling multisampling in Wine may vary as it depends heavily on how games do their rendering and there are likely more bugs too :)
http://bugs.winehq.org/show_bug.cgi?id=18728
--- Comment #41 from Jerome Leclanche adys.wh@gmail.com 2011-02-06 18:01:51 CST --- (In reply to comment #40)
Nvidia fixed this bug in the 270 version of their drivers. My test samples also work correctly now. Though your mileage with enabling multisampling in Wine may vary as it depends heavily on how games do their rendering and there are likely more bugs too :)
Oh the love. Thanks for the heads up!I guess we can close this invalid now?
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID
--- Comment #42 from Jerome Leclanche adys.wh@gmail.com 2011-03-27 13:39:37 CDT --- Alright this should be enough time.
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #43 from Jerome Leclanche adys.wh@gmail.com 2011-03-27 13:39:49 CDT --- Closing invalid.
http://bugs.winehq.org/show_bug.cgi?id=18728
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|INVALID |UPSTREAM
--- Comment #44 from Jerome Leclanche adys.wh@gmail.com 2012-03-05 13:31:33 CST --- Upstream bug -> Resolving UPSTREAM
-
wine-bugs＠winehq.org