[Bug 21165] New: Borderlands is ~3x slower under Wine (GeForce Go 7900 GS).
http://bugs.winehq.org/show_bug.cgi?id=21165 Summary: Borderlands is ~3x slower under Wine (GeForce Go 7900 GS). Product: Wine Version: unspecified Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: directx-d3d AssignedTo: wine-bugs(a)winehq.org ReportedBy: default_357-line(a)yahoo.de Borderlands runs about three times slower under Wine than under Windows (from 30+FPS to 10FPS). This is with all settings as low as they go. Wine prints no output during execution. My graphics card is a GeForce Go 7900 GS, in a 1.6x2 Core2 T5500 laptop. Both Windows and Linux use the nVidia beta driver (v190). Wine version is Git head (1.1.35+). Graphics are identical, except that Z ordering of light decals seems broken (this may be linked to slowdowns). Adaptive clocking correctly switches to full-speed mode; Speedstep is off. perf record/report lists the following distribution: 40% in libGLcore.so, 30% in Borderlands.exe itself, 12% in the kernel, 5% in wineserver, 4% in ntdll, 3% in libc, 2% in wined3d (the rest is negligible). Any ideas? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Manuel Soukup <linuxuser-sky(a)gmx.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |linuxuser-sky(a)gmx.de --- Comment #1 from Manuel Soukup <linuxuser-sky(a)gmx.de> 2009-12-26 13:11:33 --- Hi i noticed this as well but thought it could be cause my ati grafics card is too slow. My system: T400 Thinkpad Intel(R) Core(TM)2 Duo CPU P8400 @ 2.26GHz 4 Gig Ram HD3470 ATI Mobility -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Jeff Zaroyko <jeffz(a)jeffz.name> changed: What |Removed |Added ---------------------------------------------------------------------------- Version|unspecified |1.1.35 -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Vitaliy Margolen <vitaliy(a)kievinfo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|normal |trivial -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #2 from default_357-line(a)yahoo.de 2009-12-29 21:31:51 --- This does throw up an interesting problem. How much speed loss from Wine is acceptable before it counts as a "serious" bug? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #3 from Dmitry Timoshkov <dmitry(a)codeweavers.com> 2009-12-29 21:44:43 --- http://bugs.winehq.org/page.cgi?id=fields.html#bug_severity Crashes are 'normal' in priority, usually performance bugs are considered lower. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #4 from Manuel Soukup <linuxuser-sky(a)gmx.de> 2009-12-30 05:41:04 --- But why is it slow and how can we/devs speed it up? Do you need any logs or something to research the problem? Happy to help -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #5 from Manuel Soukup <linuxuser-sky(a)gmx.de> 2009-12-30 05:45:42 --- Hi again just read the bug_severity link and trivial would be: Trivial For a UI glitch that doesn't affect running of a program But its very slow and it affects the program, for ex you cant aim properly on moving targets cause it is so slow. Is this still the right priority? I personally would suggest minor but there is sadly no woraround so the description does not fit. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Dmitry Timoshkov <dmitry(a)codeweavers.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|trivial |minor --- Comment #6 from Dmitry Timoshkov <dmitry(a)codeweavers.com> 2009-12-30 06:22:34 --- In reality the severity doesn't change much, the bug still needs a volunteer to work on it. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 default_357-line(a)yahoo.de changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|minor |trivial --- Comment #7 from default_357-line(a)yahoo.de 2009-12-30 07:20:02 --- Part of the bug is that there's _no_ interesting log infos that would point towards the cause of the slowdown, well, aside from fixme:d3d_surface:IWineD3DSurfaceImpl_LoadLocation 0x32dc10f8: Downloading rgb texture to reload it as srgb, but that happens maybe two times per frame and not often enough to be responsible for the slowdown (and if I bypass the upload in question, it's still as slow; so it's probably unrelated) - the textures in question are tiny anyway, just the control surfaces of the machines and the item highlights. ANYWAY, the lack of FIXMEs does point towards a systemic error - some information being misreported to the game, causing a slower codepath to be used. Is there any way to compare system call responses between Windows and Linux runs? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #8 from Dmitry Timoshkov <dmitry(a)codeweavers.com> 2009-12-30 07:27:49 --- (In reply to comment #7)
Is there any way to compare system call responses between Windows and Linux runs?
That won't help. You may want to try oprofile to try to figure out where most of the slowness is. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #9 from default_357-line(a)yahoo.de 2009-12-31 06:34:57 --- Tried that. Either oprofile doesn't like my system, or it doesn't like the Borderlands binary, but I got more samples from konqueror's javascript library than from the game (which was running at close to 100 cpu). -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Erich Hoover <ehoover(a)mines.edu> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |ehoover(a)mines.edu --- Comment #10 from Erich Hoover <ehoover(a)mines.edu> 2010-01-24 11:46:49 --- (In reply to comment #8)
(In reply to comment #7)
Is there any way to compare system call responses between Windows and Linux runs?
That won't help. You may want to try oprofile to try to figure out where most of the slowness is.
Are there any good examples of using oprofile with Wine? I tried some examples from googling and I'm either not reading the results correctly or the vast majority of the time is being spent outside of the Wine code. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Stefan Reich <winehq(a)halbling.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |winehq(a)halbling.de --- Comment #11 from Stefan Reich <winehq(a)halbling.de> 2010-02-16 13:39:55 --- I can confirm this on a Q6700 (4x2.66 GHz) and Nvidia GTS 250. I did some testing with oprofile (even built wine with debug-information) and this is what I ended up with: - Windows (XP) uses all four cores up to 80% while playing, gives me 65+ fps (including dynamic shadows) - wine (1.1.35 gentoo) uses exactly ONE core at 100% and gives about 12 fps (no dynamic shadows) - from what I could read, most of the time is spent mutex-locking in kernel and nvidia drivers. No idea why borderlands doesn't use multicore in wine :-/ -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #12 from Erich Hoover <ehoover(a)mines.edu> 2010-02-16 14:04:23 --- (In reply to comment #11)
... No idea why borderlands doesn't use multicore in wine :-/
It might be that Wine doesn't support SetThreadIdealProcessor yet. I've seen that show up in the log for a lot of games, though I cannot remember if Borderlands is one of them. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Stefan Reich <winehq(a)halbling.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 --- Comment #13 from Stefan Reich <winehq(a)halbling.de> 2010-02-17 08:46:35 --- *** This bug has been confirmed by popular vote. *** -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #14 from Stefan Reich <winehq(a)halbling.de> 2010-02-17 13:23:45 --- (in reply to comment #12)
(In reply to comment #11)
... No idea why borderlands doesn't use multicore in wine :-/
It might be that Wine doesn't support SetThreadIdealProcessor yet. I've seen that show up in the log for a lot of games, though I cannot remember if Borderlands is one of them.
Well, I did some more code reading, testing and adding debug output... Borderlands doesn't seem to call SetThreadIdealProcessor, but it does something (incredibly stupid?) with SetThreadAffinityMask:
fixme:thread:SetThreadAffinityMask (0xfffffffe,0x1): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x2): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x4): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x8): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x1): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x1): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x1): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0x1): stub fixme:thread:SetThreadAffinityMask (0xfffffffe,0xffffffff): stub fixme:thread:SetThreadAffinityMask (0x1bc,0xffffffff): stub fixme:thread:SetThreadAffinityMask (0x458,0xffffffff): stub fixme:thread:SetThreadAffinityMask (0x464,0xffffffff): stub ...[more 0xff sets]
As can be deduced from the debug output, I simply 'deactivated' the affinity mask for the time being...and what happen? --> Got 40/70/40/40 CPU usage and 30-40 fps, which is actually quite playable. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #15 from Erich Hoover <ehoover(a)mines.edu> 2010-02-17 13:48:46 --- (In reply to comment #14)
... As can be deduced from the debug output, I simply 'deactivated' the affinity mask for the time being...and what happen? --> Got 40/70/40/40 CPU usage and 30-40 fps, which is actually quite playable.
Did you add custom debug messages to see this? I don't see any debug messages for SetThreadAffinityMask in git. It looks like the game sets the thread to use specific processors and then later tries to set the thread to "ALL" processors. According to MSDN, and as seen in Wine, this should cause an ERROR_INVALID_PARAMETER. However, it's entirely possible that what really happens on Windows is that ((DWORD_PTR)-1) is a legit "ALL" request. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #16 from Stefan Reich <winehq(a)halbling.de> 2010-02-17 13:54:38 --- (in reply to comment #15)
Did you add custom debug messages to see this? I don't see any debug messages for SetThreadAffinityMask in git.
Yes, I manually added those FIXME's (working with 1.1.38 sources now). -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #17 from Erich Hoover <ehoover(a)mines.edu> 2010-02-17 14:12:42 --- (In reply to comment #15)
... However, it's entirely possible that what really happens on Windows is that ((DWORD_PTR)-1) is a legit "ALL" request.
Hmm, this is a result that I was not expecting (yes AND no): https://winetestbot.geldorp.nl/JobDetails.pl?Key=747 Line 806 of test: ok(SetThreadAffinityMask(curthread,(-1))==0, "SetThreadAffinityMask passed for an illegal processor\n"); -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #18 from Manuel Soukup <linuxuser-sky(a)gmx.de> 2010-02-17 14:32:03 --- How can i deactivate the affinity mask? Could someone make a patch? If this patch does not break any other things could it be in wine 1139 ? Many thanks Manuel -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 joaopa <jeremielapuree(a)yahoo.fr> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jeremielapuree(a)yahoo.fr --- Comment #19 from joaopa <jeremielapuree(a)yahoo.fr> 2010-02-17 14:58:32 --- It would be interesting in knowing the behaviour of the game in a real Vista or Seven box, since the test fails for these box. Erich, so you could be right for the behaviour of SetThreadAffinityMask for pre-Vista box, -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #20 from Stefan Reich <winehq(a)halbling.de> 2010-02-17 15:01:36 --- (in reply to comment #19)
It would be interesting in knowing the behaviour of the game in a real Vista or Seven box, since the test fails for these box.
Doesn't exhibit thread/processor problems in vista or seven, as far as me or my friends are concerned. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #21 from Erich Hoover <ehoover(a)mines.edu> 2010-02-17 15:13:02 --- (In reply to comment #19)
It would be interesting in knowing the behaviour of the game in a real Vista or Seven box, since the test fails for these box.
Erich, so you could be right for the behaviour of SetThreadAffinityMask for pre-Vista box,
That's backwards, ok() prints the error message (fails) when the parameter is true - so the problem would occur on XP. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #22 from Erich Hoover <ehoover(a)mines.edu> 2010-02-17 15:44:31 --- (In reply to comment #21)
(In reply to comment #19)
It would be interesting in knowing the behaviour of the game in a real Vista or Seven box, since the test fails for these box.
Erich, so you could be right for the behaviour of SetThreadAffinityMask for pre-Vista box,
That's backwards, ok() prints the error message (fails) when the parameter is true - so the problem would occur on XP.
Please excuse my mistake here, I wrote this too hastily. ok() prints the error message when the first parameter is false, which in this case is when SetThreadAffinityMask(thread,-1) succeeds (ie. you would see the problem on XP). However, it is possible that on XP a failed SetThreadAffinityMask(thread,<any>) actually resets the affinity mask. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #23 from default_357-line(a)yahoo.de 2010-02-18 17:03:46 --- Sadly doesn't make a difference in my case, speed-wise. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #24 from Erich Hoover <ehoover(a)mines.edu> 2010-02-18 20:59:05 --- (In reply to comment #23)
Sadly doesn't make a difference in my case, speed-wise.
I believe the processor affinity is a red herring. On my machine I only see ~10 FPS and my highest used processor is at approximately 70% according to htop. I'm thinking this is a driver issue. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #25 from Stefan Reich <winehq(a)halbling.de> 2010-02-19 14:36:09 --- I'll have to believe you there as windows XP still gives me 2-3 times the performance in BL. However, the affinity setting is definitely a problem, too...do we spawn a separate bug report for that? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #26 from Erich Hoover <ehoover(a)mines.edu> 2010-02-19 14:53:38 --- (In reply to comment #25)
I'll have to believe you there as windows XP still gives me 2-3 times the performance in BL.
However, the affinity setting is definitely a problem, too...do we spawn a separate bug report for that?
You can, though it's probably not worth it - I'm working on getting a patch submitted and it's almost ready. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Luke Bratch <l_bratch(a)yahoo.co.uk> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |l_bratch(a)yahoo.co.uk -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #27 from Erich Hoover <ehoover(a)mines.edu> 2010-02-22 10:11:00 --- (In reply to comment #26)
(In reply to comment #25)
I'll have to believe you there as windows XP still gives me 2-3 times the performance in BL.
However, the affinity setting is definitely a problem, too...do we spawn a separate bug report for that?
You can, though it's probably not worth it - I'm working on getting a patch submitted and it's almost ready.
The "all processors" patch (http://www.winehq.org/pipermail/wine-cvs/2010-February/064547.html) is included in the latest git. If you still find you're having multi-processor problems with Borderlands after loading the latest git then I suggest opening a new bug. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #28 from Erich Hoover <ehoover(a)mines.edu> 2010-02-23 17:16:37 --- This may just be my own perception*, but I adjusted the Quality/Performance slider in nvidia-settings all the way to High Performance (and restarted) and it looks like the game is running about 3x faster. * An artifact of running my PC for a long time without rebooting. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #29 from default_357-line(a)yahoo.de 2010-02-24 08:34:56 ---
From what I can tell, that slider has always been on high-performance for me.
StV == Sync to VBlank Video Texture Adapter: StV off Video Blitter Adapter Settings: StV off Cursor Shadow: off :) Performance: StV off, Allow Flipping on, Image Settings: High Performance Antialiasing/Anisotropic: Application Default Texture Sharpening: on (?) Overclocking disabled. Could it be that you'd installed some kernel or driver update before the reboot? 195.22 driver here. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #30 from Erich Hoover <ehoover(a)mines.edu> 2010-02-24 10:08:03 --- (In reply to comment #29)
... Could it be that you'd installed some kernel or driver update before the reboot? 195.22 driver here.
Entirely possible, I'm running the 185.18.36 driver - I don't believe Ubuntu ships any newer versions for Karmic. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 atomic.quark(a)gmail.com changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |atomic.quark(a)gmail.com --- Comment #31 from atomic.quark(a)gmail.com 2010-08-24 09:17:46 --- Just confirming that this bug still exists in wine 1.3.0. I'm using an NVIDIA GEFORCE GTS 360M w/ 1GB GDDR5 with the proprietary driver, version: 195.36.24 I have 4GB 1333 RAM and an i7 Q720. I get somewhere between 5 and 30 fps, though mostly on the lower end. While playable, it's just barely. Changing settings seems to have little to no effect (resolution, game/foliage/texture quality, etc) all seem to be independent of framerate, which makes me feel that there's a bottleneck not directly related to the rendering. Unfortunately, I know absolutely nothing about programming for Windows so even if I had the time I don't think I could do much, but I would love to help someone get this fixed. The slowdown seems to be the only problem for me (besides messed up outlines in cinematics), everything else appears to run fine. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #32 from Erich Hoover <ehoover(a)mines.edu> 2010-08-24 09:57:24 --- (In reply to comment #31)
... Changing settings seems to have little to no effect (resolution, game/foliage/texture quality, etc) all seem to be independent of framerate, which makes me feel that there's a bottleneck not directly related to the rendering. ...
Being cognizant of the performance issues we've all been experiencing I've been formulating a theory as to their source. After playing many times with my father, and watching certain types of artifacts occur with different settings, I am growing suspicious that somehow it is the processing of audio that is causing the slowdown. Evidence: 1) When loading games there is a black screen before the "wormhole effect" that consumes roughly the same time as the effect itself. 2) Only when playing using the ALSA driver do I see the "zoom crosshair doesn't show up" bug, and when I do see the crosshair under these conditions it is always timed with the end of an audio event (ie. hearing gunshots or footsteps). 3) When standing still the framerate appears to be much better than when walking, and it is even worse when two people are walking within earshot of each other. I don't exactly have the time to track this down right now (I have a backlog of patches I need to update), but I'm hoping this information will be useful to someone else. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #33 from atomic.quark(a)gmail.com 2010-08-24 10:22:47 --- (In reply to comment #32)
(In reply to comment #31)
... Changing settings seems to have little to no effect (resolution, game/foliage/texture quality, etc) all seem to be independent of framerate, which makes me feel that there's a bottleneck not directly related to the rendering. ...
Being cognizant of the performance issues we've all been experiencing I've been formulating a theory as to their source. After playing many times with my father, and watching certain types of artifacts occur with different settings, I am growing suspicious that somehow it is the processing of audio that is causing the slowdown.
Evidence: 1) When loading games there is a black screen before the "wormhole effect" that consumes roughly the same time as the effect itself. 2) Only when playing using the ALSA driver do I see the "zoom crosshair doesn't show up" bug, and when I do see the crosshair under these conditions it is always timed with the end of an audio event (ie. hearing gunshots or footsteps). 3) When standing still the framerate appears to be much better than when walking, and it is even worse when two people are walking within earshot of each other.
I don't exactly have the time to track this down right now (I have a backlog of patches I need to update), but I'm hoping this information will be useful to someone else. That's an interesting thought. I forgot to mention that in winecfg I was unable to select ALSA and had to select OSS (I believe this has to do with the fact that I'm using 64-bit Debian). However, after the wormhole effect I usually see a black screen as opposed to before it. I have noticed that when standing the framerate can be better, but I never linked this to the sounds. I tested the sound thing a little and found that at least some of it is linked with how the graphics are rendering, or so it seems. I tried starting a new character and without moving I noticed that where I looked had a somewhat significant impact on my framerate. If I faced the town like you start, I was getting 15fps at most, but if I turned around I could get far more by facing essentially nothing. So while sound may affect things, it seems doubtful that it's the primary thing slowing the game down. It'd be nice if there was a way to mute the game (so it did no sound processing at all) to see if this really was the problem.
-- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Xavier Vachon <xvachon(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |xvachon(a)gmail.com --- Comment #34 from Xavier Vachon <xvachon(a)gmail.com> 2011-04-09 13:04:29 CDT --- Is this still an issue in current wine (1.3.17)? -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 mardini <superppl(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |superppl(a)gmail.com --- Comment #35 from mardini <superppl(a)gmail.com> 2011-06-05 00:00:24 CDT --- I noticed that Borderlands runs much slower than other Unreal Engine 3 titles, like Mass Effect. This is one wine 1.3.21 The line: fixme:d3d:resource_check_usage Unhandled usage flags 0x8. Repeats nearly endlessly while the game is running. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #36 from Luis Alvarado <luisalvaradox(a)gmail.com> 2012-03-06 15:28:28 CST --- I tested with 1.4rc5 and 1.4rc6. Do not know when it was fixed but I can play with not much FPS lose. It might run about 3 to 5fps less than on windows but it runs perfect. I test with the following hardware: Test on 2 PCs: Core 2 Duo E8500 4GB RAM Nvidia 9500 GT Core i7 2600 16GB RAM Nvidia 440 GT -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 arthur.huillet(a)free.fr changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |arthur.huillet(a)free.fr -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 John P Sims <jsims2359(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jsims2359(a)gmail.com --- Comment #37 from John P Sims <jsims2359(a)gmail.com> 2012-06-29 00:16:29 CDT --- I think this bug is fixed. I play the Steam version of Borderlands for hours at a time and aside from occasional graphics stutter, the game is quite playable (and quite addicting) at 1920x1200 with max gfx settings. wine-1.5.7 NVIDIA GTX 480 w/ 295.40 drivers i7-920 w/ 6 Gigs of RAM -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Dan Kegel <dank(a)kegel.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED CC| |dank(a)kegel.com Resolution| |FIXED --- Comment #38 from Dan Kegel <dank(a)kegel.com> 2012-06-29 00:49:49 CDT --- Reported fixed twice. Please reopen if still a problem. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Alexandre Julliard <julliard(a)winehq.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED --- Comment #39 from Alexandre Julliard <julliard(a)winehq.org> 2012-07-03 14:15:16 CDT --- Closing bugs fixed in 1.5.8. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 --- Comment #40 from Mr.Weller <monsieurweller(a)gmail.com> 2013-02-02 09:09:23 CST --- Created attachment 43423 --> http://bugs.winehq.org/attachment.cgi?id=43423 Debug log Compressed (part of) debug log (WINEDEBUG=+d3d) -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Mr.Weller <monsieurweller(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |monsieurweller(a)gmail.com --- Comment #41 from Mr.Weller <monsieurweller(a)gmail.com> 2013-02-02 09:11:30 CST --- Actually the bug isn't fixed in wine 1.5.22. I've just attached a part of the debug log generated with WINEDEBUG=+d3d. Hope it helps. If further details of logs are needed, i will be happy to proceed. :) -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Zombie Ryushu <zombie_ryushu(a)yahoo.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |zombie_ryushu(a)yahoo.com --- Comment #42 from Zombie Ryushu <zombie_ryushu(a)yahoo.com> 2013-11-11 00:14:05 CST --- This bug has returned in wine 1.5.28 under Rosa 2012.1 -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Jerome Leclanche <adys.wh(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|CLOSED |REOPENED Resolution|FIXED | --- Comment #43 from Jerome Leclanche <adys.wh(a)gmail.com> 2013-11-11 00:20:54 CST --- Reopening. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Dmitry Timoshkov <dmitry(a)baikal.ru> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution| |FIXED --- Comment #44 from Dmitry Timoshkov <dmitry(a)baikal.ru> 2013-11-11 02:37:35 CST --- (In reply to comment #42)
This bug has returned in wine 1.5.28 under Rosa 2012.1
Please open a bew bug report with a proper regression testing info: http://wiki.winehq.org/RegressionTesting -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
http://bugs.winehq.org/show_bug.cgi?id=21165 Dmitry Timoshkov <dmitry(a)baikal.ru> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED --- Comment #45 from Dmitry Timoshkov <dmitry(a)baikal.ru> 2013-11-11 02:38:09 CST --- Closing. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.
participants (1)
-
wine-bugs＠winehq.org