http://bugs.winehq.org/show_bug.cgi?id=29647
Bug #: 29647 Summary: majesty 2 crashes after a short time due to OOM Product: Wine Version: 1.3.37 Platform: x86-64 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-d3d AssignedTo: wine-bugs@winehq.org ReportedBy: alex.richardson@gmx.de Classification: Unclassified
Created attachment 38425 --> http://bugs.winehq.org/attachment.cgi?id=38425 the output on the console
I start Majesty 2, start the first mission. For the first few seconds everything runs fine, then it starts getting noticeably slower (probably due to it being harder to get free memory) and then after ~2 minutes the whole game crashes because there is no video memory left.
Should I run it again with some environment variable values so you can get better debug output?
http://bugs.winehq.org/show_bug.cgi?id=29647
GyB gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |gyebro69@gmail.com
--- Comment #1 from GyB gyebro69@gmail.com 2012-01-22 06:40:46 CST --- I can't reproduce the problem here, using a Nvidia 250 card with the proprietary driver. I have the following overrides set for Majesty 2: d3dx9_36, msvc*80. The game runs without issues here. My console output is the same as yours up to this point: ...
ALSA lib pcm.c:7343:(snd_pcm_recover) underrun occurred err:d3d:wined3d_occlusion_query_ops_issue >>>>>>>>>>>>>>>>> GL_OUT_OF_MEMORY >(0x505) from glBeginQuery() @ query.c / 494
...
What video card and driver are you using? How much ram is installed on the card? According to the system requirements the game needs at least a Nvidia 6800 GT with 256 MB Ram (and pixel shader 2.0 support).
wine: virtual memory exhausted
This looks suspicious, I've never seen such message before.
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #2 from Alex Richardson alex.richardson@gmx.de 2012-01-22 08:17:02 CST --- Ah yes, I guess I should have said I have a ATI Radeon 4850 with 512 MB RAM. I am running the fglrx driver from december 2011.
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #3 from Alex Richardson alex.richardson@gmx.de 2012-01-24 16:15:40 CST --- The game leaks 15-50 Megabytes of memory every 1-2 seconds while running. This happens both when using the wine dlls and when using overrides for mscp80 msvcr80 and d3dx9_36.
I tried to run the program using valgrind, but I didn't manage to get Wine and Valgrind to work together. It happened both when using wine 1.3.37 and when using 1.3.30, so the issue seems to have been there for a while.
Is there anything else I could do to locate the place this issue is coming from?
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #4 from Henri Verbeet hverbeet@gmail.com 2012-01-24 16:23:04 CST --- It sounds to me like it might be the graphics driver leaking memory / address space somewhere.
http://bugs.winehq.org/show_bug.cgi?id=29647
Graham libgradev@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |libgradev@gmail.com
--- Comment #5 from Graham libgradev@gmail.com 2012-02-23 15:19:45 CST --- Wine version: 1.4-rc4 Card: ATI Radeon HD6950 2048MB
I can confirm this is still an issue in Wine 1.4-rc4.
Opening/menu animations display fine with no issue. As soon as the game starts I can play for about 1min before a crash occurs:
err:d3d:wined3d_occlusion_query_ops_issue >>>>>>>>>>>>>>>>> GL_OUT_OF_MEMORY (0x505) from glBeginQuery() @ query.c / 494 wine: Unhandled page fault on read access to 0x00000000 at address 0x7a4451c3 (thread 0009), starting debugger...
Tested with high/low game graphics settings and versions of Wine (dedicated prefix) - always occurs. The game itself renders fine and is usable pre crash.
Swapping in an NVidia 8800GTX card resolves the issue so it does appear to be, as Henri said, a fglrx issue.
Is there anything else to try or should this be reported to AMD?
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #6 from Graham libgradev@gmail.com 2012-02-23 15:26:37 CST --- Created attachment 39033 --> http://bugs.winehq.org/attachment.cgi?id=39033 debugger output
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #7 from Graham libgradev@gmail.com 2012-02-23 15:30:52 CST --- Created attachment 39034 --> http://bugs.winehq.org/attachment.cgi?id=39034 console output
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #8 from Henri Verbeet hverbeet@gmail.com 2012-02-24 03:57:11 CST --- (In reply to comment #5)
Swapping in an NVidia 8800GTX card resolves the issue so it does appear to be, as Henri said, a fglrx issue.
Is there anything else to try or should this be reported to AMD?
For reporting the issue, it may help if you're able to reproduce it with a small GL application outside of Wine as well. But well, fglrx. It may also be worth giving the r600g driver a try if you haven't already.
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #9 from Graham libgradev@gmail.com 2012-03-06 04:34:35 CST --- In addition, the following is dumped to the kernel message buffer (post crash):
[fglrx:__create_mapping] *ERROR* Could not find the group private data [fglrx:__mc_heap_map_virtual_space] *ERROR* Failed to map the virtual space [fglrx:mc_heap_map_virtual_space] *ERROR* Can not get virtual address [fglrx:MCIL_GetVirtualAddressInDescriptor] *ERROR* Can not get the virtual address
I haven't run many native GL apps yet - only the Unigine Heaven bench and that runs flawlessly (x64 and x86). I will try some native games asap to see if I can reproduce it outside of Wine.
As Catalyst 12.2 is due imminently I will wait and see if that fixes (one can hope ;)) before reporting the issue to AMD.
http://bugs.winehq.org/show_bug.cgi?id=29647
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |adys.wh@gmail.com
--- Comment #10 from Jerome Leclanche adys.wh@gmail.com 2012-03-13 18:51:29 CDT --- (In reply to comment #9) They've been released. Could you retest?
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #11 from Alex Richardson alex.richardson@gmx.de 2012-03-14 09:49:10 CDT --- Still happening with fglrx 12.2.
Happens both with GLSL enabled and disabled. Are there any other registry settings I could try?
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #12 from Jerome Leclanche adys.wh@gmail.com 2012-03-14 09:51:50 CDT --- (In reply to comment #10) If it still happens in 12.2, FWIU this is upstream and needs to be reported upstream. Could someone do that?
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #13 from Alex Richardson alex.richardson@gmx.de 2012-05-27 11:01:00 CDT --- I managed to run it perfectly with r600g driver. Definitely an upstream issue then.
Sadly this will probably never be fixed for me, since AMD wants to drop r600/700 support very soon. I just hope open source drivers catch up performance-wise soon since other games are not really playable with them
http://bugs.winehq.org/show_bug.cgi?id=29647
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution| |UPSTREAM
--- Comment #14 from Jerome Leclanche adys.wh@gmail.com 2012-05-27 11:29:58 CDT --- Upstream.
http://bugs.winehq.org/show_bug.cgi?id=29647
--- Comment #15 from Graham libgradev@gmail.com 2012-06-02 08:49:36 CDT --- I have been playing for some time now on max settings I can confirm this is fixed in the Catalyst 12.6 beta drivers (Xorg 1.12).
http://bugs.winehq.org/show_bug.cgi?id=29647
Jerome Leclanche adys.wh@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #16 from Jerome Leclanche adys.wh@gmail.com 2012-06-02 08:50:16 CDT --- Closing then
-
wine-bugs＠winehq.org