http://bugs.winehq.org/show_bug.cgi?id=30557
Bug #: 30557 Summary: winegstreamer broken with >=glib-2.32.0 Product: Wine Version: 1.3.24 Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: winegstreamer AssignedTo: wine-bugs@winehq.org ReportedBy: galtgendo@o2.pl Classification: Unclassified
As of glib 2.32.0, a call to g_thread_init(foo), where foo!=NULL is no longer effective, it just triggers a warning: GThread-WARNING **: GThread system no longer supports custom thread implementations. While I'm not sure this is the reason, a game using gstreamer sounds like a stuck vinyl record, it does seem this way. This sort of invalidates most of dlls/winegstreamer/glibthread.c content.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #1 from Vitaliy Margolen vitaliy-bugzilla@kievinfo.com 2012-04-30 23:10:22 CDT --- These changes will pretty much break gstreamer usability for Wine. Considering that it is mostly broken in 64-bit distros it's not a huge loss.
Regardless, Wine absolutely requires vtable functionality. Without it gstreamer can not be used.
http://bugs.winehq.org/show_bug.cgi?id=30557
Rafał Mużyło galtgendo@o2.pl changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|1.3.24 |unspecified
http://bugs.winehq.org/show_bug.cgi?id=30557
Ruslan b7.10110111@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |b7.10110111@gmail.com
--- Comment #2 from Ruslan b7.10110111@gmail.com 2012-05-18 08:39:13 CDT ---
Considering that it is mostly broken in 64-bit distros it's not a huge loss.
Somehow it's also broken on my 32-bit LFS.
http://bugs.winehq.org/show_bug.cgi?id=30557
Ralf Jung ralfjung-e@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |ralfjung-e@gmx.de
--- Comment #3 from Ralf Jung ralfjung-e@gmx.de 2012-05-21 09:56:27 CDT --- What is gstreamer used for in wine anyway? I only recently got it working on my 64bit system (on Debian, using multiarch) only to discover the error message mentioned above.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #4 from Ruslan b7.10110111@gmail.com 2012-05-21 10:59:02 CDT --- (In reply to comment #3)
What is gstreamer used for in wine anyway? I only recently got it working on my 64bit system (on Debian, using multiarch) only to discover the error message mentioned above.
It's used to at least play some videos in games, e.g. intro videos in GTA 3,VC,SA.
http://bugs.winehq.org/show_bug.cgi?id=30557
Dan Kegel dank@kegel.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |dank@kegel.com
--- Comment #5 from Dan Kegel dank@kegel.com 2012-05-21 11:15:09 CDT --- Looks like this change was made some time ago: http://mail.gnome.org/archives/commits-list/2011-September/msg03773.html and was discussed here: http://mail.gnome.org/archives/gtk-devel-list/2011-August/msg00059.html http://mail.gnome.org/archives/gtk-devel-list/2011-September/msg00003.html It looks like they weren't aware that Wine used custom threads?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #6 from Rafał Mużyło galtgendo@o2.pl 2012-05-21 11:26:15 CDT --- Actually, it's more like they were of opinion, that it's not working correctly in some of the cases and as that it couldn't be fixed, so they simply removed the option to break things, when they've made glib initialize threads unconditionally.
TBH, this implementation, wine tries to use, seems a bizarre mix of pthreads and Windows threading - IIRC it was already (that is: before glib 2.32) not working correctly some of the times.
http://bugs.winehq.org/show_bug.cgi?id=30557
Andrew Eikum aeikum@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |aeikum@codeweavers.com
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download CC| |gyebro69@gmail.com
--- Comment #7 from GyB gyebro69@gmail.com 2012-06-01 10:42:27 CDT --- I bumped into this bug after upgrading to Fedora 17 (32 bit) which is shipped with Glib-2.32.1. All those games which use winegstreamer to play music/videos either hang or crash with a segmentation fault on Fedora 17 (Gstreamer 0.10.36 installed).
Echoes (from Binary Zoo) is a small (11 MB), freeware asteroids clone game. It uses winegstreamer to play music. The game hangs on Fedora 17 shortly after starting, sometimes I can get to the main menu and the game crashes there. http://binaryzoo.com/games/echoes/echoes.zip
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1
--- Comment #8 from GyB gyebro69@gmail.com 2012-06-01 10:56:23 CDT --- confirming
http://bugs.winehq.org/show_bug.cgi?id=30557
jamesonpc@vcu.edu changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jamesonpc@vcu.edu
http://bugs.winehq.org/show_bug.cgi?id=30557
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Severity|normal |major
--- Comment #9 from Rosanne DiMesio dimesio@earthlink.net 2012-06-10 08:21:20 CDT --- Affects a wide range of apps; setting severity to major.
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |liesdiedatei@gmail.com
--- Comment #10 from GyB gyebro69@gmail.com 2012-06-10 10:57:44 CDT --- *** Bug 30886 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |e826081@rtrtr.com
--- Comment #11 from GyB gyebro69@gmail.com 2012-06-27 12:54:06 CDT --- *** Bug 31044 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Michael Stefaniuc mstefani@redhat.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |austinenglish@gmail.com
--- Comment #12 from Michael Stefaniuc mstefani@redhat.com 2012-07-04 09:43:47 CDT --- *** Bug 31106 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Kostas hunter18@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |hunter18@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #13 from Ruslan b7.10110111@gmail.com 2012-08-05 05:12:12 CDT --- As a workaround for crashes because of winegstreamer not working, one can disable winegstreamer.dll in winecfg. Then instead of crash the games will (as it seems) think that the media file is playing without end. This at least lets play GTA{3,VC,SA}.
http://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |22921
http://bugs.winehq.org/show_bug.cgi?id=30557
ax 34noff otaku@rambler.ru changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |otaku@rambler.ru
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #14 from ax 34noff otaku@rambler.ru 2012-09-28 13:49:10 CDT --- Looks like applying this workaround to Fallout 3 leads game crash (see attached file)
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #15 from ax 34noff otaku@rambler.ru 2012-09-28 13:50:29 CDT --- Created attachment 41870 --> http://bugs.winehq.org/attachment.cgi?id=41870 Crash of Fallout 3
etc.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #16 from ax 34noff otaku@rambler.ru 2012-09-28 13:51:59 CDT --- (game works near an hour and then crashes)(In reply to comment #15)
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #17 from ax 34noff otaku@rambler.ru 2012-09-28 13:54:00 CDT --- Comment on attachment 41870 --> http://bugs.winehq.org/attachment.cgi?id=41870 Crash of Fallout 3
fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. fixme:d3d:resource_check_usage Unhandled usage flags 0x8. //........working good and then:
err:ole:COMPOBJ_DllList_Add couldn't load in-process dll L"C:\windows\system32\winegstreamer.dll" err:ole:create_server class {f9d8d64e-a144-47dc-8ee0-f53498372c29} not registered fixme:ole:CoGetClassObject CLSCTX_REMOTE_SERVER not supported err:ole:CoGetClassObject no class object {f9d8d64e-a144-47dc-8ee0-f53498372c29} could be created for context 0x17 wine: Unhandled page fault on write access to 0x00000000 at address 0x1781863f (thread 002a), starting debugger... Unhandled exception: page fault on write access to 0x00000000 in 32-bit code (0x1781863f). Register dump: CS:0073 SS:007b DS:007b ES:007b FS:0033 GS:003b EIP:1781863f ESP:0033ed6c EBP:0033eda4 EFLAGS:00210287( R- -- I S - -P-C) EAX:ffff8000 EBX:ffff8000 ECX:00000004 EDX:00000000 ESI:00000000 EDI:00000000 Stack dump: 0x0033ed6c: ffff8000 00000004 00000000 17819273 0x0033ed7c: 00000004 17872a30 00000001 00000001 0x0033ed8c: 00000004 00000000 00000000 17803760 0x0033ed9c: 644312a8 00000000 0033edc4 17802740 0x0033edac: 00000050 00000000 0284ce18 00000000 0x0033edbc: 0284cebc 0284ce18 00000000 0ff83a25 Backtrace: =>0 0x1781863f schro_tables_init+0xbf() [z:/workspace/codecs/schroedinger.git/msys/schroedinger/../../schroedinger/schrotablesinit.c:37] in libschroedinger-1.0-0 (0x0033eda4) 1 0x17802740 schro_decoder_new+0x1f() [z:/workspace/codecs/schroedinger.git/msys/schroedinger/../../schroedinger/schrodecoder.c:162] in libschroedinger-1.0-0 (0x0033edc4) 0x1781863f schro_tables_init+0xbf [z:/workspace/codecs/schroedinger.git/msys/schroedinger/../../schroedinger/schrotablesinit.c:37] in libschroedinger-1.0-0: movw %ax,0x0(%edi,%esi,2) Unable to access file 'z:/workspace/codecs/schroedinger.git/msys/schroedinger/../../schroedinger/schrotablesinit.c' Modules: Module Address Debug info Name (181 modules) PE 340000- 356000 Deferred xinput1_3 PE 400000- 136b000 Deferred fallout3 PE 1370000- 2221000 Deferred xlive PE 2230000- 25f8000 Deferred d3dx9_38 PE 2600000- 2733000 Deferred libvorbis PE ff80000- ffab000 Export diracsplitter-libschroedinger.axC:\Program Files\Dirac\DiracSplitter-libschroedinger.ax PE 10000000-1001e000 Deferred libvorbisfile PE 15e30000-15f8b000 Deferred liboil-0.3-0 PE 17800000-179e2000 Stabs libschroedinger-1.0-0 PE 179f0000-17d8d000 Deferred mfc90 PE 18000000-18068000 Deferred binkw32 PE 27500000-2760d000 Deferred msidcrl40 PE 35500000-35708000 Deferred quartz PE 5d360000-5d36e000 Deferred mfc90rus PE 78480000-7850e000 Deferred msvcp90 PE 78520000-785c3000 Deferred msvcr90 ELF 79311000-79403000 Deferred libasound.so.2 ELF 79bca000-7b800000 Deferred libnvidia-glcore.so.295.40 ELF 7b800000-7ba2d000 Deferred kernel32<elf> -PE 7b810000-7ba2d000 \ kernel32 ELF 7bc00000-7bcc6000 Deferred ntdll<elf> -PE 7bc10000-7bcc6000 \ ntdll ELF 7bdb3000-7bf00000 Deferred libxml2.so.2 ELF 7bf00000-7bf04000 Deferred <wine-loader> ELF 7bf7e000-7c003000 Deferred urlmon<elf> -PE 7bf90000-7c003000 \ urlmon ELF 7c448000-7c45e000 Deferred dwmapi<elf> -PE 7c450000-7c45e000 \ dwmapi ELF 7c45e000-7c500000 Deferred msxml3<elf> -PE 7c470000-7c500000 \ msxml3 ELF 7c607000-7c61d000 Deferred midimap<elf> -PE 7c610000-7c61d000 \ midimap ELF 7c61d000-7c636000 Deferred msacm32<elf> -PE 7c620000-7c636000 \ msacm32 ELF 7c636000-7c672000 Deferred libxslt.so.1 ELF 7c6c8000-7c6d0000 Deferred libogg.so.0 ELF 7c6d0000-7c6fb000 Deferred libvorbis.so.0 ELF 7c6fb000-7c873000 Deferred libvorbisenc.so.2 ELF 7c873000-7c8c1000 Deferred libflac.so.8 ELF 7c8c1000-7c933000 Deferred libsndfile.so.1 ELF 7c933000-7c998000 Deferred libpulsecommon-1.1.so ELF 7c998000-7c9e6000 Deferred libpulse.so.0 ELF 7c9e8000-7ca15000 Deferred winealsa<elf> -PE 7c9f0000-7ca15000 \ winealsa ELF 7ca15000-7ca2a000 Deferred avicap32<elf> -PE 7ca20000-7ca2a000 \ avicap32 ELF 7ce6c000-7ce8a000 Deferred libgcc_s.so.1 ELF 7cef6000-7cf0a000 Deferred msimg32<elf> -PE 7cf00000-7cf0a000 \ msimg32 ELF 7d1bc000-7d1c3000 Deferred libasyncns.so.0 ELF 7d1c3000-7d1cd000 Deferred libwrap.so.0 ELF 7d1d1000-7d1f5000 Deferred devenum<elf> -PE 7d1e0000-7d1f5000 \ devenum ELF 7d1f5000-7d1fd000 Deferred libjson.so.0 ELF 7d21f000-7d245000 Deferred winepulse<elf> -PE 7d230000-7d245000 \ winepulse ELF 7d245000-7d268000 Deferred mmdevapi<elf> -PE 7d250000-7d268000 \ mmdevapi ELF 7d268000-7d2ac000 Deferred dinput<elf> -PE 7d270000-7d2ac000 \ dinput ELF 7d32d000-7d405000 Deferred libgl.so.1 ELF 7d427000-7d4f8000 Deferred opengl32<elf> -PE 7d440000-7d4f8000 \ opengl32 ELF 7d4f8000-7d501000 Deferred librt.so.1 ELF 7d501000-7d506000 Deferred libgpg-error.so.0 ELF 7d506000-7d51e000 Deferred libresolv.so.2 ELF 7d51e000-7d522000 Deferred libkeyutils.so.1 ELF 7d522000-7d56b000 Deferred libdbus-1.so.3 ELF 7d56b000-7d57d000 Deferred libp11-kit.so.0 ELF 7d57d000-7d602000 Deferred libgcrypt.so.11 ELF 7d602000-7d614000 Deferred libtasn1.so.3 ELF 7d614000-7d61d000 Deferred libkrb5support.so.0 ELF 7d61d000-7d645000 Deferred libk5crypto.so.3 ELF 7d645000-7d714000 Deferred libkrb5.so.3 ELF 7d714000-7d726000 Deferred libavahi-client.so.3 ELF 7d726000-7d734000 Deferred libavahi-common.so.3 ELF 7d734000-7d7f8000 Deferred libgnutls.so.26 ELF 7d7f8000-7d836000 Deferred libgssapi_krb5.so.2 ELF 7d836000-7d889000 Deferred libcups.so.2 ELF 7d894000-7d898000 Deferred libnvidia-tls.so.295.40 ELF 7d898000-7d8ab000 Deferred gnome-keyring-pkcs11.so ELF 7d8c1000-7d8f5000 Deferred uxtheme<elf> -PE 7d8d0000-7d8f5000 \ uxtheme ELF 7d8f5000-7d8fb000 Deferred libxfixes.so.3 ELF 7d8fb000-7d906000 Deferred libxcursor.so.1 ELF 7d909000-7d90e000 Deferred libcom_err.so.2 ELF 7d9b6000-7d9e0000 Deferred libexpat.so.1 ELF 7d9e0000-7da14000 Deferred libfontconfig.so.1 ELF 7da14000-7da24000 Deferred libxi.so.6 ELF 7da24000-7da28000 Deferred libxcomposite.so.1 ELF 7da28000-7da31000 Deferred libxrandr.so.2 ELF 7da31000-7da3b000 Deferred libxrender.so.1 ELF 7da3b000-7da41000 Deferred libxxf86vm.so.1 ELF 7da41000-7da45000 Deferred libxinerama.so.1 ELF 7da45000-7da67000 Deferred imm32<elf> -PE 7da50000-7da67000 \ imm32 ELF 7da67000-7da6e000 Deferred libxdmcp.so.6 ELF 7da6e000-7da72000 Deferred libxau.so.6 ELF 7da72000-7da93000 Deferred libxcb.so.1 ELF 7da93000-7daad000 Deferred libice.so.6 ELF 7daad000-7dbe1000 Deferred libx11.so.6 ELF 7dbe1000-7dbf3000 Deferred libxext.so.6 ELF 7dbf3000-7dc7c000 Deferred winex11<elf> -PE 7dc00000-7dc7c000 \ winex11 ELF 7dc7c000-7dd16000 Deferred libfreetype.so.6 ELF 7dd38000-7dd54000 Deferred dinput8<elf> -PE 7dd40000-7dd54000 \ dinput8 ELF 7dd54000-7dd6f000 Deferred wsock32<elf> -PE 7dd60000-7dd6f000 \ wsock32 ELF 7dd6f000-7ddfd000 Deferred msvcrt<elf> -PE 7dd80000-7ddfd000 \ msvcrt ELF 7ddfd000-7de26000 Deferred msacm32<elf> -PE 7de00000-7de26000 \ msacm32 ELF 7de26000-7ded4000 Deferred winmm<elf> -PE 7de30000-7ded4000 \ winmm ELF 7ded4000-7df19000 Deferred dsound<elf> -PE 7dee0000-7df19000 \ dsound ELF 7df19000-7df4c000 Deferred wintrust<elf> -PE 7df20000-7df4c000 \ wintrust ELF 7df4c000-7df60000 Deferred psapi<elf> -PE 7df50000-7df60000 \ psapi ELF 7df60000-7df74000 Deferred sensapi<elf> -PE 7df70000-7df74000 \ sensapi ELF 7df74000-7df9a000 Deferred mpr<elf> -PE 7df80000-7df9a000 \ mpr ELF 7df9a000-7dfb0000 Deferred libz.so.1 ELF 7dfb2000-7dfd2000 Deferred oleacc<elf> -PE 7dfc0000-7dfd2000 \ oleacc ELF 7dfd2000-7e043000 Deferred wininet<elf> -PE 7dfe0000-7e043000 \ wininet ELF 7e043000-7e0fd000 Deferred crypt32<elf> -PE 7e050000-7e0fd000 \ crypt32 ELF 7e0fd000-7e20c000 Deferred oleaut32<elf> -PE 7e110000-7e20c000 \ oleaut32 ELF 7e20c000-7e314000 Deferred ole32<elf> -PE 7e220000-7e314000 \ ole32 ELF 7e314000-7e37e000 Deferred shlwapi<elf> -PE 7e320000-7e37e000 \ shlwapi ELF 7e37e000-7e592000 Deferred shell32<elf> -PE 7e390000-7e592000 \ shell32 ELF 7e592000-7e5c4000 Deferred ws2_32<elf> -PE 7e5a0000-7e5c4000 \ ws2_32 ELF 7e5c4000-7e5e7000 Deferred iphlpapi<elf> -PE 7e5d0000-7e5e7000 \ iphlpapi ELF 7e5e7000-7e612000 Deferred netapi32<elf> -PE 7e5f0000-7e612000 \ netapi32 ELF 7e612000-7e640000 Deferred secur32<elf> -PE 7e620000-7e640000 \ secur32 ELF 7e640000-7e6b6000 Deferred rpcrt4<elf> -PE 7e650000-7e6b6000 \ rpcrt4 ELF 7e6b6000-7e6f3000 Deferred winspool<elf> -PE 7e6c0000-7e6f3000 \ winspool ELF 7e6f3000-7e75a000 Deferred setupapi<elf> -PE 7e700000-7e75a000 \ setupapi ELF 7e788000-7e8b0000 Deferred wined3d<elf> -PE 7e7a0000-7e8b0000 \ wined3d ELF 7e8b0000-7e8e7000 Deferred d3d9<elf> -PE 7e8c0000-7e8e7000 \ d3d9 ELF 7e8e7000-7e948000 Deferred advapi32<elf> -PE 7e8f0000-7e948000 \ advapi32 ELF 7e948000-7ea0e000 Deferred gdi32<elf> -PE 7e950000-7ea0e000 \ gdi32 ELF 7ea0e000-7eb4f000 Deferred user32<elf> -PE 7ea20000-7eb4f000 \ user32 ELF 7eb4f000-7ec48000 Deferred comctl32<elf> -PE 7eb60000-7ec48000 \ comctl32 ELF 7ec48000-7ec55000 Deferred libnss_files.so.2 ELF 7ec55000-7ec5e000 Deferred libnss_compat.so.2 ELF 7ec5e000-7ec67000 Deferred libsm.so.6 ELF 7ec67000-7ec80000 Deferred version<elf> -PE 7ec70000-7ec80000 \ version ELF 7efb2000-7efde000 Deferred libm.so.6 ELF 7efe6000-7f000000 Deferred libnsl.so.1 ELF b7464000-b7469000 Deferred libdl.so.2 ELF b7469000-b760e000 Deferred libc.so.6 ELF b760f000-b762a000 Deferred libpthread.so.0 ELF b762a000-b7630000 Deferred libuuid.so.1 ELF b7640000-b764c000 Deferred libnss_nis.so.2 ELF b764c000-b778e000 Dwarf libwine.so.1 ELF b7790000-b77b2000 Deferred ld-linux.so.2 ELF b77b2000-b77b3000 Deferred [vdso].so Threads: process tid prio (all id:s are in hex) 0000000e services.exe 0000001f 0 0000001e 0 00000015 0 00000010 0 0000000f 0 00000012 winedevice.exe 0000001c 0 00000018 0 00000014 0 00000013 0 0000001a plugplay.exe 00000020 0 0000001d 0 0000001b 0 00000021 explorer.exe 00000022 0 00000029 (D) Z:\home\redneck\games\Fallout3\Fallout3.exe 00000027 0 00000044 0 00000043 15 00000042 0 00000041 0 0000003f 1 00000039 0 00000036 15 00000035 0 00000034 0 00000033 15 00000032 0 00000030 0 0000002f 0 0000002e 0 0000002d 0 0000002c 0 0000002b 0 0000002a 0 <==
http://bugs.winehq.org/show_bug.cgi?id=30557
don bright hugh.m.bright@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |hugh.m.bright@gmail.com
--- Comment #18 from don bright hugh.m.bright@gmail.com 2012-11-03 08:59:33 CDT --- workaround: per another thread on a similar bug - download the source code for wine, and "./configure" using "--without-gstreamer" (on ubuntu i had to add --with-alsa and install several development library packages as well).
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #19 from Dan Kegel dank@kegel.com 2012-11-18 10:44:42 CST --- Affects Lego Digital Designer, http://ldd.lego.com/en-us/download/ Game starts fine, but after placing a brick, crashes with
fixme:gstreamer:event_sink 0x74c14290 stub tag wine client error:9: read: Bad address wine client error:9: read: Bad file descriptor Segmentation fault (core dumped)
Either building --without-gstreamer or WINEDLLOVERRIDES=winegstreamer=d works around the problem.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #20 from Dan Kegel dank@kegel.com 2012-11-18 12:47:37 CST --- Created attachment 42537 --> http://bugs.winehq.org/attachment.cgi?id=42537 Patch to disable winegstreamer if glib too new
This bug is annoying enough that we should probably just disable winegstreamer when faced with glib-2.32 or newer. Here's a patch to do this.
(Note that it won't delete any existing winegstreamer dlls for you, so if you try this on an already compiled tree, winegstreamer will still be there until you clean it.)
http://bugs.winehq.org/show_bug.cgi?id=30557
Jay jaynobyl@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jaynobyl@gmx.de
http://bugs.winehq.org/show_bug.cgi?id=30557
Brandon Corujo haku08879@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |haku08879@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
Adam Bolte boltronics@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |boltronics@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |andrey.goosev@gmail.com
--- Comment #21 from GyB gyebro69@gmail.com 2013-01-19 22:48:26 CST --- *** Bug 32759 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #22 from ax 34noff otaku@rambler.ru 2013-01-20 04:17:59 CST --- Can anybody delete my unintentional Comment 17 (and then this comment too)?
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jitsumi@gmail.com
--- Comment #23 from GyB gyebro69@gmail.com 2013-01-25 11:50:15 CST --- *** Bug 32813 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- See Also| |https://bugzilla.gnome.org/ | |show_bug.cgi?id=692608
--- Comment #24 from Austin English austinenglish@gmail.com 2013-01-26 18:28:09 CST --- No one seems to have filed a bug for this, and I couldn't find one on Gnome's bugzilla, so I did: https://bugzilla.gnome.org/show_bug.cgi?id=692608
http://bugs.winehq.org/show_bug.cgi?id=30557
James Eder jimportal@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jimportal@gmail.com
--- Comment #25 from James Eder jimportal@gmail.com 2013-02-08 17:24:02 CST --- Would moving to GStreamer 1.0 (Bug 31836) help Wine in this regard?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #26 from Dan Kegel dank@kegel.com 2013-02-08 17:25:15 CST --- Dunno. I doubt it. Running Windows gstreamer would probably help, though. We could install that like we install mono and gecko. The downside is not being able to use system codecs.
http://bugs.winehq.org/show_bug.cgi?id=30557
Lucas Fialho Zawacki lfzawacki@yahoo.com.br changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |lfzawacki@yahoo.com.br
http://bugs.winehq.org/show_bug.cgi?id=30557
Lara Maia lara@craft.net.br changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |lara@craft.net.br
--- Comment #27 from Lara Maia lara@craft.net.br 2013-02-17 00:03:10 CST --- This problem is not being reproduced in wine 1.5.24
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #28 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-02-19 02:54:01 CST --- I still reproduce it with GTA:VC intro videos in wine-1.5.24-35-gc449427
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #29 from jitsumi@gmail.com 2013-02-19 07:41:47 CST --- Yeah, still have the problem wit worms revolution and wine 1.5.24.
http://bugs.winehq.org/show_bug.cgi?id=30557
Mike Goldberg gdimike@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |gdimike@yahoo.com
--- Comment #30 from Mike Goldberg gdimike@yahoo.com 2013-02-20 14:35:56 CST --- This is still a problem with winSPMBT, too. This program used to run fairly well.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #31 from Mike Goldberg gdimike@yahoo.com 2013-02-20 14:38:19 CST --- Created attachment 43656 --> http://bugs.winehq.org/attachment.cgi?id=43656 backtrace for winSPMBT.exe
http://bugs.winehq.org/show_bug.cgi?id=30557
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |moses-winehq@mozai.com
--- Comment #32 from GyB gyebro69@gmail.com 2013-03-11 11:52:46 CDT --- *** Bug 33181 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Leffe leffeman@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |leffeman@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |adys.wh@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
Bruno Jesus 00cpxxx@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |ubuntu@ozzyfrank.com
--- Comment #33 from Bruno Jesus 00cpxxx@gmail.com 2013-04-22 14:19:41 CDT --- *** Bug 30817 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |aankoopdenissen@skynet.be
--- Comment #34 from Rosanne DiMesio dimesio@earthlink.net 2013-05-03 11:05:16 CDT --- *** Bug 33517 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
Sylvain Petreolle spetreolle@yahoo.fr changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |spetreolle@yahoo.fr
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #35 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-06-28 05:59:45 CDT --- Created attachment 45014 --> http://bugs.winehq.org/attachment.cgi?id=45014 override pthread_create
Should fix things again..
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #36 from Ralf Jung post+wine@ralfj.de 2013-06-28 07:46:09 CDT --- Thanks for working on this issue :)
With the patch applied on top of 1.6-rc3, wine doesn't compile anymore here (Debian testing amd64, compiling in an i386 chroot):
thread.c: In function 'RtlCreateUserThread': thread.c:610:9: error: 'native_thread' undeclared (first use in this function) thread.c:610:9: note: each undeclared identifier is reported only once for each function it appears in
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #37 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 08:29:37 CDT --- (In reply to comment #35)
Should fix things again..
Fails as said in comment #36, and after I did 's/((native_thread)/(info->\1/', it compiled successfully, and I got this output with GTAVC: http://pastebin.com/raw.php?i=0PPg41S0
Couldn't get a winedbg backtrace though.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #38 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 08:42:39 CDT --- BTW, older glib isn't affected by the patch - works as before.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #39 from Ralf Jung post+wine@ralfj.de 2013-06-28 09:27:58 CDT --- Created attachment 45020 --> http://bugs.winehq.org/attachment.cgi?id=45020 Log of Warcraft III trying to play a video, gstreamer disabled (--without-gstreamer), WINEDEBUG unset
After fixing the compilation as indicated above, the behaviour of Warcraft III when it tries to play a video changed. Previously, it would blank the screen for a short time and return to the game UI quickly. Now it blanks the screen and changes the cursor to my KDE theme (instead of the one the game uses), and then it doesn't do anything anymore, not even Alt-Tab or the shortcut I set up to minimise a window. After getting it killed, I saw that it crashed.
I will attach both the log of a compilation with gstreamer disabled (--without-gstreamer), WINEDEBUG unset and with gstreamer enabled and the patch applied, WINEDEBUG set to warn+all.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #40 from Ralf Jung post+wine@ralfj.de 2013-06-28 09:29:00 CDT --- Created attachment 45021 --> http://bugs.winehq.org/attachment.cgi?id=45021 Log of Warcraft III trying to play a video, gstreamer enabled and the patch applied, WINEDEBUG set to warn+all
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #41 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 09:30:54 CDT --- Well, the patch in fact does fix things (for GTAVC at least) on Ubuntu Precise. Though didn't on my LFS. I'll have to dig more. But to apply the patch you still need to add "info->" before native_thread in if expression.
(In reply to comment #39)
Log of Warcraft III trying to play a video, gstreamer disabled (--without-gstreamer), WINEDEBUG unset
What's the point in disabling gstreamer at this stage? The patch should make gstreamer work.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #42 from Ralf Jung post+wine@ralfj.de 2013-06-28 09:32:27 CDT --- That's for comparison, the "base case". Feel free to mark it obsolete if you think it adds no information.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #43 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 09:43:09 CDT --- (In reply to comment #37)
..., and I got this output with GTAVC:
OK, this was my mistake — either I forgot to install compiled wine, or wineserver was still running, after retry it works. So, now for all three games I tested — GTA{VC,3,SA} the patch fixes this bug.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #44 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 09:49:04 CDT --- Created attachment 45024 --> http://bugs.winehq.org/attachment.cgi?id=45024 [patch] override pthread_create — fixed compilation
Here's updated version of Maarten's patch, so the testers don't need to edit source manually.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #45 from Ralf Jung post+wine@ralfj.de 2013-06-28 09:53:42 CDT --- (In reply to comment #43)
So, now for all three games I tested — GTA{VC,3,SA} the patch fixes this bug.
So it actually plays the video for you? Still no luck here, I just installed the "good" gstreamer plugins for i386, but wine still crashes, and then it spits out tons of errors and warnings while it tries to produce debug information.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #46 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 09:57:26 CDT --- (In reply to comment #45)
So it actually plays the video for you?
Yes, I can watch intro or skip and then proceed to playing the game. Maybe other games still need something more to be fixed.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #47 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-06-28 10:19:44 CDT --- (In reply to comment #45)
I just installed the "good" gstreamer plugins for i386
You might in fact need ugly or bad plugins, MPEG2 is among them afair.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #48 from Ralf Jung post+wine@ralfj.de 2013-06-28 10:44:23 CDT --- (In reply to comment #47)
You might in fact need ugly or bad plugins, MPEG2 is among them afair.
That would be unfortunate, those plugins are not foreign-arch-installable in Debian. I will try running wine inside the i386 chroot I used to build it.
Of course, wine should not crash just because gstreamer does not support the given media, but that's another bug (if it is the reason for my troubles).
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #49 from James Eder jimportal@gmail.com 2013-06-28 11:32:04 CDT --- +#ifdef __linux__
Would it be better to check for affected pthread implementations in configure or is this really something that will always be a Linux specific thing?
http://bugs.winehq.org/show_bug.cgi?id=30557
Maarten Lankhorst m.b.lankhorst@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |m.b.lankhorst@gmail.com
--- Comment #50 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-07-04 12:55:10 CDT --- Sorry, after some more testing it looks like the patch is incomplete. The crash is a direct result from wine running pthread_join on a thread it doesn't own, causing any calls to pthread_*join* or pthread_detach to run into freed memory and crashing the application.
I don't have a fix yet, it will probably involve hooking both functions, keeping a list of threads and their id's, without introducing a race condition.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #51 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-07-04 15:24:55 CDT --- Created attachment 45104 --> http://bugs.winehq.org/attachment.cgi?id=45104 try2
hook pthread_join and detach too, using robust mutexes to ensure the thread is really dead without waiting on it. :P
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #52 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-07-04 16:03:37 CDT --- (In reply to comment #51)
Created attachment 45104 [details]
This time it crashes immediately, doesn't work at all. Terminal output: http://pastebin.com/ebPHtK4m
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #53 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-07-04 16:29:04 CDT --- Created attachment 45106 --> http://bugs.winehq.org/attachment.cgi?id=45106 try3
oh, dumb thinko.. fixed version
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #54 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-07-04 16:37:14 CDT --- (In reply to comment #53)
Created attachment 45106 [details] try3
oh, dumb thinko.. fixed version
Still crashes with the same output.
http://bugs.winehq.org/show_bug.cgi?id=30557
Maarten Lankhorst m.b.lankhorst@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #45104|0 |1 is obsolete| | Attachment #45106|0 |1 is obsolete| |
--- Comment #55 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-07-05 03:27:58 CDT --- Created attachment 45108 --> http://bugs.winehq.org/attachment.cgi?id=45108 try4
Oops, I didn't notice the compiler trying to warn me. :-)
Fixed exit_thread_common now to use NtCurrentTeb() correctly..
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #56 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-07-05 03:53:59 CDT --- (In reply to comment #55)
Created attachment 45108 [details] try4
Yes, now it works.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #57 from Rosanne DiMesio dimesio@earthlink.net 2013-07-05 08:25:50 CDT --- (In reply to comment #55)
Created attachment 45108 [details] try4
Oops, I didn't notice the compiler trying to warn me. :-)
Fixed exit_thread_common now to use NtCurrentTeb() correctly..
Doesn't build on openSUSE 12.3:
ccache gcc -m32 -fno-omit-frame-pointer -fno-asynchronous-unwind-tables -m32 -o wine -Wl,--export-dynamic -Wl,--section-start,.interp=0x7bf00400 main.o -L../libs/wine -lwine ../libs/port/libwine_port.a -lpthread -Wl,--rpath,$ORIGIN/../libs/wine /usr/lib64/gcc/x86_64-suse-linux/4.7/../../../../x86_64-suse-linux/bin/ld: main.o: undefined reference to symbol 'dlvsym@@GLIBC_2.1' /usr/lib64/gcc/x86_64-suse-linux/4.7/../../../../x86_64-suse-linux/bin/ld: note: 'dlvsym@@GLIBC_2.1' is defined in DSO /lib/libdl.so.2 so try adding it to the linker command line /lib/libdl.so.2: could not read symbols: Invalid operation collect2: error: ld returned 1 exit status make[1]: *** [wine] Error 1 make[1]: Leaving directory `/home/dimesio/wine-git/loader' make: *** [loader] Error 2
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #58 from Maarten Lankhorst m.b.lankhorst@gmail.com 2013-07-07 14:21:38 CDT --- add -ldl somewhere..
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #59 from Rosanne DiMesio dimesio@earthlink.net 2013-07-07 14:40:47 CDT --- (In reply to comment #58)
add -ldl somewhere..
Can you be a little more specific?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #60 from Rafał Mużyło galtgendo@o2.pl 2013-07-07 14:42:10 CDT --- (In reply to comment #58)
add -ldl somewhere..
As we're being lazy here, appending '@LIBDL@' to 'LIBPTHREAD =' line in loader/Makefile.in looks as if it should work.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #61 from Bjoern Bidar theodorstormgrade@googlemail.com 2013-07-13 09:13:42 CDT --- Created attachment 45227 --> http://bugs.winehq.org/attachment.cgi?id=45227 Warcraft 3 error msg
http://bugs.winehq.org/show_bug.cgi?id=30557
Bjoern Bidar theodorstormgrade@googlemail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |theodorstormgrade@googlemai | |l.com
--- Comment #62 from Bjoern Bidar theodorstormgrade@googlemail.com 2013-07-13 09:14:24 CDT --- I tried the patch with Warcraft 3 in playing a campaign video but: it fixes the custom thead error but I still get: (wine:11527): GLib-ERROR **: gmem.c:421: overflow allocating 1701866823*4 bytes
and the exception error from war3 (see attached screen).
http://bugs.winehq.org/show_bug.cgi?id=30557
Maurizio Oliveri 6tsukiyami9@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |6tsukiyami9@gmail.com
--- Comment #63 from Maurizio Oliveri 6tsukiyami9@gmail.com 2013-07-27 16:53:41 CDT --- Right, mmmh... The "try4" patch doesn't work for me, as it still gives this error whenever I try to build wine: /usr/bin/ld: main.o: undefined reference to symbol 'dlvsym@@GLIBC_2.1' /lib/i386-linux-gnu/libdl.so.2: error adding symbols: DSO missing from command line collect2: error: ld returned 1 exit status
The same goes for the "override pthread_create" patch. I've tried this with the latest git, under Debian Sid with glib version 2.36.3-3
http://bugs.winehq.org/show_bug.cgi?id=30557
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |30687
http://bugs.winehq.org/show_bug.cgi?id=30557
flamingmo@online.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |flamingmo@online.de
http://bugs.winehq.org/show_bug.cgi?id=30557
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |29920
http://bugs.winehq.org/show_bug.cgi?id=30557
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |27437
http://bugs.winehq.org/show_bug.cgi?id=30557
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |systemofdown@gmail.com
--- Comment #64 from Jerome Leclanche adys.wh@gmail.com 2013-08-18 22:40:14 CDT --- *** Bug 32465 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #65 from Maurizio Oliveri 6tsukiyami9@gmail.com 2013-09-03 09:27:15 CDT --- With latest git, wine compiles flawlessly and works properly (as far as I've tested). I've been away for a while though, so it may have been fixed before as well...
Tested on Debian Sid amd64 with libglib2.0-dev version 2.36.4
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #66 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-09-03 11:56:48 CDT --- (In reply to comment #65)
With latest git, wine compiles flawlessly and works properly (as far as I've tested). I've been away for a while though, so it may have been fixed before as well...
Tested on Debian Sid amd64 with libglib2.0-dev version 2.36.4
Can't confirm on Ubuntu Precise 32 bit. Currently quartz only correctly fails, i.e. works as it did before gstreamer support was implemented, so I can skip non-working videos in GTAVC. As for older glib, I can still watch the videos on my LFS with old glib&gst versions.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #67 from Maurizio Oliveri 6tsukiyami9@gmail.com 2013-09-04 10:46:53 CDT --- (In reply to comment #66)
(In reply to comment #65)
With latest git, wine compiles flawlessly and works properly (as far as I've tested). I've been away for a while though, so it may have been fixed before as well...
Tested on Debian Sid amd64 with libglib2.0-dev version 2.36.4
Can't confirm on Ubuntu Precise 32 bit. Currently quartz only correctly fails, i.e. works as it did before gstreamer support was implemented, so I can skip non-working videos in GTAVC. As for older glib, I can still watch the videos on my LFS with old glib&gst versions.
Just wondering, what version of glib are you using? And during compilation did you get any error at all?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #68 from Ruslan Kabatsayev b7.10110111@gmail.com 2013-09-04 10:54:08 CDT --- (In reply to comment #67)
Just wondering, what version of glib are you using? And during compilation did you get any error at all?
2.32.3 and 2.30.3. Don't remember any errors while compiling.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #69 from Maurizio Oliveri 6tsukiyami9@gmail.com 2013-09-04 11:18:08 CDT --- (In reply to comment #68)
2.32.3 and 2.30.3. Don't remember any errors while compiling.
I can say it didn't compile at all a month ago with glib 2.36.4 and this lib hasn't been updated since then, so I guess there must've been some improvements on Wine's side... Maybe the fix works with 2.34 but not with 2.32, that's possible as well
http://bugs.winehq.org/show_bug.cgi?id=30557
Hector Martin marcan@marcansoft.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |marcan@marcansoft.com
--- Comment #70 from Hector Martin marcan@marcansoft.com 2013-09-14 10:36:31 CDT --- I can confirm that the try4 patch (plus adding @LIBDL@ to LIBPTHREAD= in loader/Makefile.in ) works great, both with wine git and 1.7.0 (on Gentoo Linux). This is the only way to progress through the opening sequence in some games that *require* working video support.
Hooking pthread_create is ugly, but at least it's future-proof (I had the "pleasure" of debugging the same exact problem in winepulse a couple years ago, which this solution would've also worked around) and it works. I can't think of another way, short of getting the glib folks to revert the removal of the custom threading hooks or switching to something other than gstreamer.
Is there a chance this patch will get merged?
Gentoo users: # mkdir -p /etc/portage/patches/app-emulation/wine # wget http://www.marcansoft.com/paste/weH812MX.txt -O /etc/portage/patches/app-emulation/wine/gstreamer-threading-fix.patch # emerge wine
http://bugs.winehq.org/show_bug.cgi?id=30557
Roc Vallès Domènech vallesroc@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |vallesroc@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
Mikael Åkersund mikael.akersund@bigfoot.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mikael.akersund@bigfoot.com
--- Comment #71 from Mikael Åkersund mikael.akersund@bigfoot.com 2013-09-22 12:15:25 CDT --- (In reply to comment #60)
(In reply to comment #58)
add -ldl somewhere..
As we're being lazy here, appending '@LIBDL@' to 'LIBPTHREAD =' line in loader/Makefile.in looks as if it should work.
There is a slight problem with just setting "LIBPTHREAD = @LIBPTHREAD@ @LIBDL@" in 'loader/Makefile.in'.
It breaks the ability to run wine directly from the build directory without first doing a "make install" as the "wine" wrapper script in the build directory will look for stuff in the install directory instead of the build directory.
If you do:
<quick edit of source file> ./configure --prefix=/usr/local/wine-hacks-1.7.02-local-00 make -j7 ./wine WindowsExecutable.exe
You'll only get errors because wine can't find '/usr/local/wine-hacks-1.7.02-local-00/bin/wineserver' instead of using the freshly compiled '<sourcedir>/server/wineserver'.
Doing a 'make install' fixes that but makes regression-testing a bit more annoying
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #72 from Mikael Åkersund mikael.akersund@bigfoot.com 2013-09-22 12:26:59 CDT --- (In reply to comment #63)
Right, mmmh... The "try4" patch doesn't work for me, as it still gives this error whenever I try to build wine: /usr/bin/ld: main.o: undefined reference to symbol 'dlvsym@@GLIBC_2.1' /lib/i386-linux-gnu/libdl.so.2: error adding symbols: DSO missing from command line collect2: error: ld returned 1 exit status
The same goes for the "override pthread_create" patch. I've tried this with the latest git, under Debian Sid with glib version 2.36.3-3
(In reply to comment #69)
(In reply to comment #68)
2.32.3 and 2.30.3. Don't remember any errors while compiling.
I can say it didn't compile at all a month ago with glib 2.36.4 and this lib hasn't been updated since then, so I guess there must've been some improvements on Wine's side... Maybe the fix works with 2.34 but not with 2.32, that's possible as well
The try4 patch does compile for me on Debian Testing with glib 2.36.4-1 with a modified loader/Makefile.in. I do have to compile in a i368 chroot for wine/configure to detect and enable gstreamer correctly but that is another bug. You probably forgot to set "LIBPTHREAD = @LIBPTHREAD@ @LIBDL@" in 'loader/Makefile.in' as they said in: (In reply to comment #60)
(In reply to comment #58)
add -ldl somewhere..
As we're being lazy here, appending '@LIBDL@' to 'LIBPTHREAD =' line in loader/Makefile.in looks as if it should work.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #73 from Rafał Mużyło galtgendo@o2.pl 2013-09-22 12:55:26 CDT --- (In reply to comment #71)
(In reply to comment #60)
(In reply to comment #58)
add -ldl somewhere..
As we're being lazy here, appending '@LIBDL@' to 'LIBPTHREAD =' line in loader/Makefile.in looks as if it should work.
There is a slight problem with just setting "LIBPTHREAD = @LIBPTHREAD@ @LIBDL@" in 'loader/Makefile.in'.
It breaks the ability to run wine directly from the build directory without first doing a "make install" as the "wine" wrapper script in the build directory will look for stuff in the install directory instead of the build directory.
If you do:
<quick edit of source file> ./configure --prefix=/usr/local/wine-hacks-1.7.02-local-00 make -j7 ./wine WindowsExecutable.exe
You'll only get errors because wine can't find '/usr/local/wine-hacks-1.7.02-local-00/bin/wineserver' instead of using the freshly compiled '<sourcedir>/server/wineserver'.
Doing a 'make install' fixes that but makes regression-testing a bit more annoying
Are you testing by running via 'wine' symlink in $(topbuildir) or otherwise, cause the second might not work.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #74 from Mikael Åkersund mikael.akersund@bigfoot.com 2013-09-22 13:43:06 CDT --- (In reply to comment #73)
(In reply to comment #71)
It breaks the ability to run wine directly from the build directory without first doing a "make install" as the "wine" wrapper script in the build directory will look for stuff in the install directory instead of the build directory.
Are you testing by running via 'wine' symlink in $(topbuildir) or otherwise, cause the second might not work.
Through the 'wine' symlink in $(topbuildir) ./wine winecfg wine: failed to initialize: /usr/local/wine-hacks-1.7.02-local-01/lib/wine/ntdll.dll.so: cannot open shared object file: No such file or directory
http://bugs.winehq.org/show_bug.cgi?id=30557
Semenov Alexey sov.info@mail.ru changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sov.info@mail.ru
--- Comment #75 from Semenov Alexey sov.info@mail.ru 2013-10-12 08:18:53 CDT --- (In reply to comment #70)
I can confirm that the try4 patch (plus adding @LIBDL@ to LIBPTHREAD= in loader/Makefile.in ) works great, both with wine git and 1.7.0 (on Gentoo Linux). This is the only way to progress through the opening sequence in some games that *require* working video support.
Hooking pthread_create is ugly, but at least it's future-proof (I had the "pleasure" of debugging the same exact problem in winepulse a couple years ago, which this solution would've also worked around) and it works. I can't think of another way, short of getting the glib folks to revert the removal of the custom threading hooks or switching to something other than gstreamer.
Is there a chance this patch will get merged?
Gentoo users: # mkdir -p /etc/portage/patches/app-emulation/wine # wget http://www.marcansoft.com/paste/weH812MX.txt -O /etc/portage/patches/app-emulation/wine/gstreamer-threading-fix.patch # emerge wine
isshin@kurosaki:~$ cd wine-1.7.4 isshin@kurosaki:~/wine-1.7.4$ patch < /home/isshin/Desktop/0.diff -p 1 patching file dlls/ntdll/ntdll_misc.h patching file dlls/ntdll/thread.c patching file dlls/winegstreamer/glibthread.c patching file libs/wine/config.c Hunk #1 succeeded at 353 (offset 14 lines). patching file libs/wine/loader.c Hunk #1 succeeded at 69 (offset 4 lines). patching file libs/wine/wine.map patching file loader/Makefile.in Hunk #1 FAILED at 33. 1 out of 1 hunk FAILED -- saving rejects to file loader/Makefile.in.rej patching file loader/main.c isshin@kurosaki:~/wine-1.7.4$
It seems that this patch doesn't work with wine 1.7.4. I tried to manually edit loader/Makefile.in file, so it looks that:
@MAKE_RULES@
LIBPTHREAD = @LIBPTHREAD@ @LIBDL@ LDEXECFLAGS = @LDEXECFLAGS@
but after that compilation simply don't work.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #76 from Rafał Mużyło galtgendo@o2.pl 2013-10-12 08:44:25 CDT --- @comment 75: Did you, after modifying loader/Makefile.in, had run configure/config.status to recreate loader/Makefile ? Or are you talking about a different type of failure ?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #77 from Semenov Alexey sov.info@mail.ru 2013-10-12 08:47:43 CDT --- (In reply to comment #76)
@comment 75: Did you, after modifying loader/Makefile.in, had run configure/config.status to recreate loader/Makefile ? Or are you talking about a different type of failure ?
yes, i run after that ./configure && make It fails 100%, no doubt.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #78 from Rafał Mużyło galtgendo@o2.pl 2013-10-12 09:27:13 CDT --- (In reply to comment #77)
(In reply to comment #76)
@comment 75: Did you, after modifying loader/Makefile.in, had run configure/config.status to recreate loader/Makefile ? Or are you talking about a different type of failure ?
yes, i run after that ./configure && make It fails 100%, no doubt.
:roll: sorry for not being psychic. It fails exactly how ?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #79 from Semenov Alexey sov.info@mail.ru 2013-10-12 09:50:12 CDT --- (In reply to comment #78)
(In reply to comment #77)
(In reply to comment #76)
@comment 75: Did you, after modifying loader/Makefile.in, had run configure/config.status to recreate loader/Makefile ? Or are you talking about a different type of failure ?
yes, i run after that ./configure && make It fails 100%, no doubt.
:roll: sorry for not being psychic. It fails exactly how ?
It fails, as i can see, with a lot of warnings and error. See attachment "wine output.tar.gz"
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #80 from Semenov Alexey sov.info@mail.ru 2013-10-12 09:54:52 CDT --- Created attachment 46268 --> http://bugs.winehq.org/attachment.cgi?id=46268 wine output.tar.gz
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #81 from Rafał Mużyło galtgendo@o2.pl 2013-10-12 12:34:15 CDT --- (In reply to comment #80)
Created attachment 46268 [details] wine output.tar.gz
Uhm...
patching file loader/Makefile.in Hunk #1 FAILED at 33. 1 out of 1 hunk FAILED -- saving rejects to file loader/Makefile.in.rej
That means that unless you've manually modified the file, it didn't get patched (and the later error suggests you didn't).
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #82 from Mikael Åkersund mikael.akersund@bigfoot.com 2013-10-12 13:18:33 CDT --- (In reply to comment #75) (In reply to comment #81)
(In reply to comment #80)
Created attachment 46268 [details] wine output.tar.gz
Uhm...
patching file loader/Makefile.in Hunk #1 FAILED at 33. 1 out of 1 hunk FAILED -- saving rejects to file loader/Makefile.in.rej
That means that unless you've manually modified the file, it didn't get patched (and the later error suggests you didn't).
loader/Makefile.in changed in 1.7.4 so you need to add -ldl or @LIBDL@ to EXTRALIBS instead.
Change the start of loader/Makefile.in to MODCFLAGS = @BUILTINFLAG@ EXTRALIBS = @LIBPTHREAD@ -ldl
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #83 from Semenov Alexey sov.info@mail.ru 2013-10-12 15:11:36 CDT --- (In reply to comment #82)
(In reply to comment #75) (In reply to comment #81)
(In reply to comment #80)
Change the start of loader/Makefile.in to MODCFLAGS = @BUILTINFLAG@ EXTRALIBS = @LIBPTHREAD@ -ldl
Thank you, now it compiling well.
-But, after all it didn't make anything in Silent Hill 3, just give another output.-
http://bugs.winehq.org/show_bug.cgi?id=30557
Alexander Dorofeyev alexd4@inbox.lv changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |alexd4@inbox.lv
--- Comment #84 from Alexander Dorofeyev alexd4@inbox.lv 2013-11-09 23:37:46 CST --- Question: what's the criteria for telling duplicates of this bug from something different? Is there supposed to be something similar in the backtraces of the crash? Or is it, basically, like - if you see the "GThread system no longer supports custom thread implementations." message then any kind of crash can be expected when the app tries to play a video?
The old looking glass games like Thief & System Shock 2 (they have similar engine), are crashing in winegstreamer as well. I reproduced in this demo http://www.thiefmissions.com/demos/thief2demo.exe and full SS2 game.
But backtraces don't seem to have any similarity with what I see here. What I also don't understand is why winegstreamer even gets called for these games - they used to work perfectly without it and do, in fact, work just fine when this library is disabled. Can it be this same bug or it's a separate issue?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #85 from ax 34noff otaku@rambler.ru 2013-11-10 03:58:18 CST ---
when this library is disabled
Yes, it's a workaround for the bug. Maybe disable winegstreamer by default in Wine 1.6.1 and even in 1.8 ?
http://bugs.winehq.org/show_bug.cgi?id=30557
aankoopdenissen@skynet.be changed:
What |Removed |Added ---------------------------------------------------------------------------- CC|aankoopdenissen@skynet.be |
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #86 from Hector Martin marcan@marcansoft.com 2013-11-10 06:32:47 CST --- If you get the "GThread system no longer supports custom thread implementations." message then you can expect any kind of weird crash when gstreamer is used. This results in mixing pthreads and Windows threads inside callbacks to Wine and all kinds of strange failure modes are to be expected.
http://bugs.winehq.org/show_bug.cgi?id=30557
Dr 4asm ignas2526@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |ignas2526@gmail.com
--- Comment #87 from Dr 4asm ignas2526@gmail.com 2013-11-24 04:45:16 CST --- This is also a problem on a Mac OS X. I'm using wine 1.7.5, after starting GTA3, screen turns white and game freezes. I tried disabling winegstreamer, but it doesn't help. There are errors regarding GThread, winegstreamer and so on.
http://bugs.winehq.org/show_bug.cgi?id=30557
Czikus czikus@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |czikus@gmail.com
--- Comment #88 from Czikus czikus@gmail.com 2013-12-01 16:19:18 CST --- This is also a problem with Office 2010 and videos in Powerpoint. After installing ffdshow and directX or trying with native quartz.dll, the following problem occurs WARNING **: GThread system no longer supports custom thread implementations with various gstreamer errors after that depending on the details of the setup. Some more details in this post: http://www.codeweavers.com/support/forums/general/?t=26;msg=157134
If the patch is actually working, could someone post a patched version of 32-bit binaries? I would like to run it on Ubuntu 12.10 64-bit. I managed to compile wine in 64 bits, but the 32-bit would require too many i386 packages on my system to actually make it work. Is there something else I can do to help debug the problem?
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #89 from Silens dr.silens@gmail.com --- Created attachment 47005 --> http://bugs.winehq.org/attachment.cgi?id=47005 Fallout New Vegas crash
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #90 from Andrey Gusev andrey.goosev@gmail.com --- Dark Souls - Prepare to Die Edition also affected. Wine 1.7.10
http://bugs.winehq.org/show_bug.cgi?id=30557
Felix Yan felixonmars@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |felixonmars@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
Malina Salina lakmilis@hotmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |lakmilis@hotmail.com
--- Comment #91 from Malina Salina lakmilis@hotmail.com --- Heya... This is my first post.. I am not sure if this is the right bug.. but there seems to be a regression here or so?
I notice that I get a black screen with dragon age origins on arch, with latest wine (1.7.11/12) but by chance, I noticed on mint 16 it works!
I found this bug to be a possible culprit, so I am trying to patch the wine git with the 'try 4' as someone said.
The issue is, the loader/Makefile.in has changed quite a bit, so I have tried a few bits and bobs (sorry, I am a poor C patcher and have little clue on Makefiles).... so it fails.
The try4 works in it's entirety to patch on current git bar the makefile.
Also, I assume I should try with that path, to *not* disable gstreamer in configure.
The current Makefile defines: EXTRALIBS = $(PTHREAD_LIBS)
I can't test this bug before I get that compiled... yes, I feel.. silly... although hopefully someone might know what the difference between the ppa in mint 16 and the newest wine is in this regard.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #92 from Malina Salina lakmilis@hotmail.com --- Sorry, I mistook my own attempts and did compile.. I just added -ldl in the new Makefile.in (somehow that failed first time, but it wasn't the actual edit, but the patch format).
Anyway, it didn't seem to work on the issue with dragon age, but that was not in a clean prefix, which I will need to get aorund to test and feedback if need be.
https://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |26146
http://bugs.winehq.org/show_bug.cgi?id=30557
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |28987
http://bugs.winehq.org/show_bug.cgi?id=30557
Pavel Ondračka pavel.ondracka@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |pavel.ondracka@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=30557
~J~ funtimeschaos@rocketmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |funtimeschaos@rocketmail.co | |m
--- Comment #93 from ~J~ funtimeschaos@rocketmail.com --- Created attachment 47862 --> http://bugs.winehq.org/attachment.cgi?id=47862 Lego Designer console output with Try4 patch
I applied the try4 patch to a wine build from git, and tested the Lego Designer, reported by an earlier comment to crash when a brick is placed.
I am able to place bricks no problem after the patch, no problem, no crashing, but the console is being rather verbose. A log has been attached.
I'd test it with something else, but the lego thing was the only free download I found among the comments.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #94 from Jerome Leclanche adys.wh@gmail.com --- No progress on this? :(
http://bugs.winehq.org/show_bug.cgi?id=30557
Alexandre Rostovtsev tetromino@gentoo.org changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |tetromino@gentoo.org
http://bugs.winehq.org/show_bug.cgi?id=30557
David Heidelberger (okias) david.heidelberger@ixit.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- Depends on| |36772
http://bugs.winehq.org/show_bug.cgi?id=30557
Czikus czikus@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC|czikus@gmail.com |
https://bugs.winehq.org/show_bug.cgi?id=30557
linuxthefishsecret@dispostable.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |linuxthefishsecret@disposta | |ble.com
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #95 from linuxthefishsecret@dispostable.com --- Still broken as of latest version and workarounds do not work,
http://pastebin.com/raw.php?i=mdrnP5rr
http://bugs.winehq.org/show_bug.cgi?id=30557
Zombie Ryushu zombie_ryushu@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |zombie_ryushu@yahoo.com
--- Comment #96 from Zombie Ryushu zombie_ryushu@yahoo.com --- Is there a way we could have Wine build an independent version of Gstreamer that only Wine Links too, that meets Wine's needs, then link Wine against the build. as a configure Arguement.
http://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #97 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Zombie Ryushu from comment #96)
You would first need at least an independent version of glib, only then gstreamer with all its plugins. Looks quite much of library copying.
https://bugs.winehq.org/show_bug.cgi?id=30557
David Heidelberger (okias) david.heidelberger@ixit.cz changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |david.heidelberger@ixit.cz
--- Comment #98 from David Heidelberger (okias) david.heidelberger@ixit.cz --- (In reply to linuxthefishsecret from comment #95)
Still broken as of latest version and workarounds do not work,
try update OpenAL version to development git. In my case it helped.
https://bugs.winehq.org/show_bug.cgi?id=30557
Nick Bowler nbowler@draconx.ca changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |nbowler@draconx.ca
https://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |tirili_@fastmail.fm
--- Comment #99 from Austin English austinenglish@gmail.com --- *** Bug 30953 has been marked as a duplicate of this bug. ***
https://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |andydunne71@eircom.net
--- Comment #100 from Austin English austinenglish@gmail.com --- *** Bug 37647 has been marked as a duplicate of this bug. ***
https://bugs.winehq.org/show_bug.cgi?id=30557
mrdeathjr28@yahoo.es changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |mrdeathjr28@yahoo.es
https://bugs.winehq.org/show_bug.cgi?id=30557
Tim-Philipp Müller tim@centricular.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |tim@centricular.net
https://bugs.winehq.org/show_bug.cgi?id=30557
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Blocks| |17448
https://bugs.winehq.org/show_bug.cgi?id=30557
Rihards richlv@nakts.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |richlv@nakts.net
--- Comment #101 from Rihards richlv@nakts.net --- for the record, this seems to affect Crusader Kings + Deus Vult as well (wine 1.7.28)
https://bugs.winehq.org/show_bug.cgi?id=30557
Julian Rüger jr98@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |jr98@gmx.net
https://bugs.winehq.org/show_bug.cgi?id=30557
Anastasius Focht focht@gmx.net changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |giorgio.ventavoli@gmail.com
--- Comment #102 from Anastasius Focht focht@gmx.net --- *** Bug 33223 has been marked as a duplicate of this bug. ***
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #103 from Andrew Eikum aeikum@codeweavers.com --- Created attachment 52882 --> https://bugs.winehq.org/attachment.cgi?id=52882 winegstreamer: Use gstreamer threadpool API to create new Wine threads
This patch, applied on top of Wine 1.8-rc1 (actually 1d19eb15d4a) fixes Echoes and Lego Designer for me.
https://bugs.winehq.org/show_bug.cgi?id=30557
Andrew Eikum aeikum@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #52882|0 |1 is obsolete| |
--- Comment #104 from Andrew Eikum aeikum@codeweavers.com --- Created attachment 52884 --> https://bugs.winehq.org/attachment.cgi?id=52884 winegstreamer: Use gstreamer threadpool API to create new Wine threads
Improved version, which now uses the kernel32 threadpool API.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #105 from Rafał Mużyło galtgendo@o2.pl --- ...not to diss you, but till bug 31836 gets fixed, this is a bit of a swan song - major distros disable winegstreamer due to that.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #106 from Rosanne DiMesio dimesio@earthlink.net --- (In reply to Andrew Eikum from comment #104)
Created attachment 52884 [details] winegstreamer: Use gstreamer threadpool API to create new Wine threads
Improved version, which now uses the kernel32 threadpool API.
The patch fixes Oblivion, which formerly would hang on start unless I disabled winegstreamer.
It also enables me to insert the two videos mentioned in bug 17448#c14 into Powerpoint 2007, though they do not play. Without your patch, they can't even be inserted.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #107 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Andrew Eikum from comment #104)
Created attachment 52884 [details] winegstreamer: Use gstreamer threadpool API to create new Wine threads
Improved version, which now uses the kernel32 threadpool API.
Doesn't help with GTA:VC, and with older glib (which worked without patch) just crashes:
=>0 0x00000000 (0x0033f118) 1 0x7cac6b28 _gst_debug_category_new+0xc1(name="taskpool", color=0, description="Thread pool") [/home/ruslan/src/gstreamer-0.10.30/gst/gstinfo.c:1339] in libgstreamer-0.10.so.0 (0x0033f148) 2 0x7cb078c2 gst_task_pool_get_type+0xa9() [/home/ruslan/src/gstreamer-0.10.30/gst/gsttaskpool.c:50] in libgstreamer-0.10.so.0 (0x0033f1a8) 3 0x7cbe1a9d wine_gst_pool_get_type+0x68() [/home/ruslan/src/wine-git/dlls/winegstreamer/threadpool.c:40] in winegstreamer (0x0033f208) 4 0x7cbe1d8d wine_gst_pool_new+0x16() [/home/ruslan/src/wine-git/dlls/winegstreamer/threadpool.c:131] in winegstreamer (0x0033f228) 5 0x7cbe1406 Gstreamer_init+0xbe() [/home/ruslan/src/wine-git/dlls/winegstreamer/main.c:259] in winegstreamer (0x0033f2a8) 6 0x7cbdb859 Gstreamer_Splitter_create+0x2b(punkout=<couldn't compute location>, phr=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/winegstreamer/gstdemux.c:1066] in winegstreamer (0x0033f2f8) 7 0x7cbe28e2 DSCF_CreateInstance+0xfc(iface=<couldn't compute location>, pOuter=<couldn't compute location>, riid=<couldn't compute location>, ppobj=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/strmbase/dllfunc.c:296] in winegstreamer (0x0033f368) 8 0x7e5fd497 CoCreateInstanceEx+0x354(rclsid=<couldn't compute location>, pUnkOuter=<couldn't compute location>, dwClsContext=<couldn't compute location>, pServerInfo=<couldn't compute location>, cmq=<couldn't compute location>, pResults=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/ole32/../../include/unknwn.h:236] in ole32 (0x0033f438) 9 0x7e5fcfd8 CoCreateInstance+0xf4(rclsid=<couldn't compute location>, pUnkOuter=<couldn't compute location>, dwClsContext=<couldn't compute location>, iid=<couldn't compute location>, ppv=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/ole32/compobj.c:3200] in ole32 (0x0033f4b8) 10 0x7cd82fbf DEVENUM_IMediaCatMoniker_BindToObject+0x1f6(iface=<couldn't compute location>, pbc=<couldn't compute location>, pmkToLeft=<couldn't compute location>, riidResult=<couldn't compute location>, ppvResult=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/devenum/mediacatenum.c:429] in devenum (0x0033f598) 11 0x7cffa605 FilterGraph2_Render+0xbe1(iface=<couldn't compute location>, ppinOut=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/quartz/../../include/objidl.h:3769] in quartz (0x0033f988) 12 0x7cffb048 FilterGraph2_RenderFile+0x1dc(iface=<couldn't compute location>, lpcwstrFile=<couldn't compute location>, lpcwstrPlayList=<couldn't compute location>) [/home/ruslan/src/wine-git/dlls/quartz/../../include/strmif.h:8652] in quartz (0x0033fa68) 13 0x00600871 in gta-vc (+0x200870) (0x006de5ac)
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #108 from Andrew Eikum aeikum@codeweavers.com --- (In reply to Rafał Mużyło from comment #105)
...not to diss you, but till bug 31836 gets fixed, this is a bit of a swan song - major distros disable winegstreamer due to that.
Yes, but I think the 1.0 API will have the same problem (right?), so getting our existing code working with the new glib is the first step.
(In reply to Ruslan Kabatsayev from comment #107)
(In reply to Andrew Eikum from comment #104)
Created attachment 52884 [details] winegstreamer: Use gstreamer threadpool API to create new Wine threads
Improved version, which now uses the kernel32 threadpool API.
Doesn't help with GTA:VC, and with older glib (which worked without patch) just crashes:
Thanks, I'll test that game and Rosanne's videos, and double-check that I'm using the glib class APIs correctly.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #109 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Andrew Eikum from comment #108)
Thanks, I'll test that game and Rosanne's videos, and double-check that I'm using the glib class APIs correctly.
For GTAVC, beware of bug 37716.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #110 from Andrew Eikum aeikum@codeweavers.com --- Unfortunately, the approach in my patch isn't sufficient. gstreamer has other threads which can call Wine code that are not able to be overridden. Additionally, plugins may make their own threads which gstreamer has no way of overriding.
Unless glib gives us the custom threading API back, I'm afraid overriding pthreads is our only option.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #111 from Andrew Eikum aeikum@codeweavers.com --- (In reply to Andrew Eikum from comment #110)
Unless glib gives us the custom threading API back, I'm afraid overriding pthreads is our only option.
Sebastian on IRC suggested a command-stream approach, where callback requests are queued from pthreads threads and handled by Wine threads.
https://bugs.winehq.org/show_bug.cgi?id=30557
Andrew Eikum aeikum@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #52884|0 |1 is obsolete| |
--- Comment #112 from Andrew Eikum aeikum@codeweavers.com --- Created attachment 52985 --> https://bugs.winehq.org/attachment.cgi?id=52985 Always run gstreamer callbacks on a Wine thread
Here's an implementation of the command stream approach. It seems to fix Echoes and GTA:VC for me.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #113 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Andrew Eikum from comment #112)
Here's an implementation of the command stream approach. It seems to fix Echoes and GTA:VC for me.
Seems to work with GTAVC indeed.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #114 from Bjoern Bidar theodorstormgrade@googlemail.com --- (In reply to Andrew Eikum from comment #112)
Created attachment 52985 [details] Always run gstreamer callbacks on a Wine thread
Here's an implementation of the command stream approach. It seems to fix Echoes and GTA:VC for me.
What about supplying your patch to wine-staging?
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #115 from Rosanne DiMesio dimesio@earthlink.net --- (In reply to Andrew Eikum from comment #112)
Created attachment 52985 [details] Always run gstreamer callbacks on a Wine thread
Here's an implementation of the command stream approach. It seems to fix Echoes and GTA:VC for me.
Building wine64 fails here with the patches:
../../../wine-git/dlls/ntdll/ntdll.spec:1477: function 'wine_get_current_teb' not defined winegcc: ../../tools/winebuild/winebuild failed make[1]: *** [ntdll.dll.so] Error 2 make[1]: Leaving directory `/wine/wine64/dlls/ntdll' make: *** [dlls/ntdll] Error 2
32 bit Wine builds fine. The two mpeg-2 videos from bug 17448 can be inserted into Powerpoint and they do play, but at the wrong size--too large and very pixellated, with only the top left part of the video actually showing.
https://bugs.winehq.org/show_bug.cgi?id=30557
Sebastian Lackner sebastian@fds-team.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |sebastian@fds-team.de
--- Comment #116 from Sebastian Lackner sebastian@fds-team.de --- (In reply to Bjoern Bidar from comment #114)
What about supplying your patch to wine-staging?
Don't worry, Wine Staging people (like me) are already watching this bug report. If this solution really fixes all issues I will consider it for inclusion.
(In reply to Rosanne DiMesio from comment #115)
Building wine64 fails here with the patches:
Yes, thats because of my hacky patch 1/2. It only modifies signal_i386.c but other architectures would need similar changes.
@Andrew: If you need help with improving patch 1/2, feel free to ask. Thanks for your work btw! :)
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #117 from Sebastian Lackner sebastian@fds-team.de --- Created attachment 53015 --> https://bugs.winehq.org/attachment.cgi?id=53015 ntdll: Add Wine export to determine if we're on a Wine thread (v2)
I've attached an improved version of patch 1/2 which is x86_64 compatible.
While reviewing patch 2/2, I've found a couple of things which probably should be fixed. I am aware that its a very early proof of concept version, but decided to share my feedback nevertheless:
- If wine_get_current_teb() returns != NULL, pthread lock/condition variable functions are executed on uninitialized memory. - Instead of CreateThreadpoolWork, I would suggest to use TrySubmitThreadpoolCallback(). It basically does the same in your case in a single command, instead of three. - "Small" functions should probably be executed directly in the dispatcher thread, to avoid the additional overhead. Also, it probably should be used as a fallback if creating a threadpool task fails. - The thread handle returned by CreateThread is currently leaked. It would also be useful to implement a mechanism to terminate the dispatcher thread when its no longer required. - Functions and global variables in gst_cbs.h should be marked as DECLSPEC_HIDDEN.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #118 from Andrew Eikum aeikum@codeweavers.com --- (In reply to Sebastian Lackner from comment #117)
While reviewing patch 2/2, I've found a couple of things which probably should be fixed. I am aware that its a very early proof of concept version, but decided to share my feedback nevertheless:
Thanks very much for the review. It needs a little more thought and cleanup, but otherwise I think it's in OK shape for review.
- If wine_get_current_teb() returns != NULL, pthread lock/condition variable
functions are executed on uninitialized memory.
Yup, thanks.
- Instead of CreateThreadpoolWork, I would suggest to use
TrySubmitThreadpoolCallback(). It basically does the same in your case in a single command, instead of three.
OK, I'll look at that.
- "Small" functions should probably be executed directly in the dispatcher
thread, to avoid the additional overhead. Also, it probably should be used as a fallback if creating a threadpool task fails.
This one I'm not sure about. I like the consistency of ALL callbacks going through call_cb, so we only need to avoid Wine code in gst_cbs.c and call_cb. It may be important if any of the cbs are on performance-critical paths, though. I'll do some analysis.
- The thread handle returned by CreateThread is currently leaked. It would
also be useful to implement a mechanism to terminate the dispatcher thread when its no longer required.
I thought about this. One thing is that winegstreamer never unloads (see Gstreamer_init), so having one extra thread idling in the background doesn't sound too terrible to me. Otherwise we have to determine when all objects with callbacks have been destroyed. I'll take another look at it, anyway.
- Functions and global variables in gst_cbs.h should be marked as
DECLSPEC_HIDDEN.
I wanted to avoid including Windows headers in gst_cbs.{h,c}, so any Windows code calls would fail to compile. Unfortunately DECLSPEC_HIDDEN lives in a Windows header. Probably I can just include a Windows header, but I liked the cleanliness of avoiding it :)
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #119 from Zombie Ryushu zombie_ryushu@yahoo.com --- static HRESULT WINAPI IAMMultiMediaStreamImpl_AddMediaStream() is the function that Shogun total War is trying to pass to wine-gstreamer to get it to play video.
https://bugs.winehq.org/show_bug.cgi?id=30557
Andrew Eikum aeikum@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #52985|0 |1 is obsolete| |
--- Comment #120 from Andrew Eikum aeikum@codeweavers.com --- Created attachment 53264 --> https://bugs.winehq.org/attachment.cgi?id=53264 Patch sequence imlementing callback wrapper approach
Here's a 3-patch sequence update for my previous attempt. I believe I addressed all of Sebastian's comments. None of the callbacks looked "small" enough to bother executing them in the main thread, so all callbacks continue to get wrapped. Please give it a test and let me know if things are not working as before.
Rosanne, I am able to reproduce the issue with those videos in Powerpoint. I haven't looked into it yet, but I will before submitting.
https://bugs.winehq.org/show_bug.cgi?id=30557
Andrew Eikum aeikum@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #53264|0 |1 is obsolete| |
--- Comment #121 from Andrew Eikum aeikum@codeweavers.com --- Created attachment 53326 --> https://bugs.winehq.org/attachment.cgi?id=53326 Another marshalling approach without new ntdll export
Here's another take, this time moving the pthread specific stuff into winegstreamer. This means we don't need to add another ntdll export, which I think will be difficult to get accepted upstream. Unless I get review feedback or hear about something broken, this is probably what I will submit upstream soon.
https://bugs.winehq.org/show_bug.cgi?id=30557
Johan Gardhage johan.gardhage@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |johan.gardhage@gmail.com
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #122 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Andrew Eikum from comment #121)
Created attachment 53326 [details] Another marshalling approach without new ntdll export
Seems to work as expected with GTA III, VC and SA with both glib 2.30.3 and 2.32.4.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #123 from Sebastian Lackner sebastian@fds-team.de --- (In reply to Andrew Eikum from comment #121)
Created attachment 53326 [details] Another marshalling approach without new ntdll export
Here's another take, this time moving the pthread specific stuff into winegstreamer. This means we don't need to add another ntdll export, which I think will be difficult to get accepted upstream. Unless I get review feedback or hear about something broken, this is probably what I will submit upstream soon.
No objections here. With this new approach its more difficult to ensure that mark_wine_thread() is called on each thread, but it should still be fine. Also, depending on the OS there might be more ways to optimize it further (for example, using a futex instead of lock/condvar), but its good to start with a general implementation. If necessary, such optimizations can still be added later.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #124 from Andrew Eikum aeikum@codeweavers.com --- This should be fixed, please retest to confirm if you can. I'm planning to send a patch to use the gstreamer-1.0 API, so don't rush out and build 32-bit gstreamer-0.10 if you don't already have it.
commit 56b6523842f8fd9125f6d3c1651d836163dfef00 Author: Andrew Eikum aeikum@codeweavers.com Date: Tue Jan 12 13:35:04 2016 -0600
winegstreamer: Always run gstreamer callbacks on a Wine thread.
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #125 from Ruslan Kabatsayev b7.10110111@gmail.com --- (In reply to Andrew Eikum from comment #124)
This should be fixed, please retest to confirm if you can.
Works for GTA 3, VC and SA.
https://bugs.winehq.org/show_bug.cgi?id=30557
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Fixed by SHA1| |56b6523842f8fd9125f6d3c1651 | |d836163dfef00 Status|NEW |RESOLVED Resolution|--- |FIXED
--- Comment #126 from Austin English austinenglish@gmail.com --- (In reply to Ruslan Kabatsayev from comment #125)
(In reply to Andrew Eikum from comment #124)
This should be fixed, please retest to confirm if you can.
Works for GTA 3, VC and SA.
Reported fixed
https://bugs.winehq.org/show_bug.cgi?id=30557
--- Comment #127 from Zombie Ryushu zombie_ryushu@yahoo.com --- Has anyone tried Shogun Total War?
https://bugs.winehq.org/show_bug.cgi?id=30557
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #128 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 1.9.2.
https://bugs.winehq.org/show_bug.cgi?id=30557
Jay jaynobyl@gmx.de changed:
What |Removed |Added ---------------------------------------------------------------------------- CC|jaynobyl@gmx.de |
-
wine-bugs＠winehq.org