http://bugs.winehq.org/show_bug.cgi?id=31754
Bug #: 31754 Summary: Full-page Silverlight 5.x on Firefox fails to render on top of browser elements Product: Wine Version: 1.5.13 Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: ehoover@mines.edu Depends on: 31406 Classification: Unclassified Regression SHA1: 514eb695846b45c89c0fc0d56653d187f1840eec
Created attachment 41757 --> http://bugs.winehq.org/attachment.cgi?id=41757 Hack to work around rendering issue
Silverlight 5.x appears to use Direct3D for rendering, and for some reason the full-page windows no longer render on top of the browser window. This is likely not a regression per se, but since commit 514eb695846b45c89c0fc0d56653d187f1840eec the Silverlight 5.x window can no longer be seen if it's a full-page application. Examples of this are applications like Netflix.
Regression test results: === 514eb695846b45c89c0fc0d56653d187f1840eec is the first bad commit commit 514eb695846b45c89c0fc0d56653d187f1840eec Author: Alexandre Julliard julliard@winehq.org Date: Tue Sep 4 15:03:04 2012 +0200
winex11: Store the rectangle of the GL drawable and use it when updating the window size.
:040000 040000 c02b6dc1c8940fc75d22753888c3fba415848f38 a5a07991ea9141a1f14275d875f27de77ae54c47 M dlls ===
The attached hack can be used to somewhat work around the issue (the window can still be occluded, but mousing over elements will bring things back).
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #1 from Erich Hoover ehoover@mines.edu 2012-09-20 15:56:25 CDT --- Created attachment 41758 --> http://bugs.winehq.org/attachment.cgi?id=41758 Hack to work around SIO_ADDRESS_LIST_CHANGE bug
This issue can also be seen on this Silverlight test page (which does not require a Netflix subscription): http://playready.directtaps.net/smoothstreaming/TTLSS720VC1/Default.html
However, the page hits a bug in winsock (use the attached hack to work around it).
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #2 from Bruno Jesus 00cpxxx@gmail.com 2012-09-20 18:26:39 CDT --- (In reply to comment #1)
However, the page hits a bug in winsock (use the attached hack to work around it).
As far as I can see this winsock bug will only get fixed when your patches to store stuff about each socket get in.
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #3 from Erich Hoover ehoover@mines.edu 2012-09-20 23:17:19 CDT --- (In reply to comment #2)
(In reply to comment #1)
However, the page hits a bug in winsock (use the attached hack to work around it).
As far as I can see this winsock bug will only get fixed when your patches to store stuff about each socket get in.
I'm not certain that that's necessary, from the description on MSDN it sounds like I should be able to spawn an APC with NtQueueApcThread and have that check for network interface changes and then, on change, alert the event on the OVERLAPPED handle. Is there something else that I'm missing here?
http://bugs.winehq.org/show_bug.cgi?id=31754
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download, regression
http://bugs.winehq.org/show_bug.cgi?id=31754
Bug 31754 depends on bug 31406, which changed state.
Bug 31406 Summary: Borderlands freezes in game when items details are displayed http://bugs.winehq.org/show_bug.cgi?id=31406
What |Old Value |New Value ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |UPSTREAM
http://bugs.winehq.org/show_bug.cgi?id=31754
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |53806917a3fc32460201a40a9ed | |d6d8d24983c25 Status|NEW |RESOLVED Resolution| |FIXED
--- Comment #4 from Alexandre Julliard julliard@winehq.org 2012-10-02 14:11:53 CDT --- Fixed by 53806917a3fc32460201a40a9edd6d8d24983c25.
http://bugs.winehq.org/show_bug.cgi?id=31754
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #5 from Alexandre Julliard julliard@winehq.org 2012-10-12 13:34:53 CDT --- Closing bugs fixed in 1.5.15.
http://bugs.winehq.org/show_bug.cgi?id=31754
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|CLOSED |REOPENED Component|-unknown |directx-d3d Resolution|FIXED | Regression SHA1|514eb695846b45c89c0fc0d5665 | |3d187f1840eec |
--- Comment #6 from Alexandre Julliard julliard@winehq.org 2012-10-23 13:41:58 CDT --- Setting the pixel format of HWND_MESSAGE windows is supposed to work (cf. f950f41bfe821879b9974f204c1f0e7ad26ebf89). Wined3d will have to be made to cope with this.
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #7 from Henri Verbeet hverbeet@gmail.com 2013-06-12 10:04:20 CDT --- (In reply to comment #6)
Setting the pixel format of HWND_MESSAGE windows is supposed to work (cf. f950f41bfe821879b9974f204c1f0e7ad26ebf89). Wined3d will have to be made to cope with this.
I didn't follow this bug very closely, what exactly is the issue in wined3d?
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #8 from Alexandre Julliard julliard@winehq.org 2013-06-12 10:24:20 CDT --- IIRC the app is passing an HWND_MESSAGE as focus_window, and wined3d was relying on SetPixelFormat to fail on that window to trigger the backup code path.
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #9 from Henri Verbeet hverbeet@gmail.com 2013-06-12 10:39:04 CDT --- (In reply to comment #8)
IIRC the app is passing an HWND_MESSAGE as focus_window, and wined3d was relying on SetPixelFormat to fail on that window to trigger the backup code path.
Assuming you mean swapchain_get_backup_dc() in context_set_gl_context(), that makes some sense, but I don't see how that would result in visible rendering.
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #10 from Alexandre Julliard julliard@winehq.org 2013-06-12 11:01:32 CDT --- Actually the d3d9 device creation fails, so the fallback path may be in the app.
http://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #11 from Henri Verbeet hverbeet@gmail.com 2013-06-12 11:06:15 CDT --- Makes sense, probably mostly just needs some tests then.
https://bugs.winehq.org/show_bug.cgi?id=31754
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords|regression |
--- Comment #12 from Austin English austinenglish@gmail.com --- Erich, is this still an issue?
https://bugs.winehq.org/show_bug.cgi?id=31754
joaopa jeremielapuree@yahoo.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jeremielapuree@yahoo.fr
--- Comment #13 from joaopa jeremielapuree@yahoo.fr --- Does the bug still occur with wine-4.7?
https://bugs.winehq.org/show_bug.cgi?id=31754
--- Comment #14 from joaopa jeremielapuree@yahoo.fr --- Does the bug still occur with wine-5.12?
https://bugs.winehq.org/show_bug.cgi?id=31754
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- URL| |https://web.archive.org/web | |/20140529102802/http://silv | |erlight.dlservice.microsoft | |.com/download/B/A/9/BA94BEC | |9-5DBC-4B50-BC2B-046A423990 | |67/30214.00/Silverlight.exe CC| |focht@gmx.net