[Bug 59847] New: 3DMark Vantage almost works, crashes when trying to render the first scene
http://bugs.winehq.org/show_bug.cgi?id=59847 Bug ID: 59847 Summary: 3DMark Vantage almost works, crashes when trying to render the first scene Product: Wine Version: 11.10 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs@list.winehq.org Reporter: aros@gmx.com Distribution: --- Created attachment 81144 --> http://bugs.winehq.org/attachment.cgi?id=81144 backtrace Steps to reproduce: 1. Download https://benchmarks.ul.com/downloads/3dmark-vantage.exe sha256: 7a30ccb7c95f4315cbd92b01d41c6f138e6e2f7985b4730eac2cfe764d592c58 2. Download and install x86 libraries as overrides https://github.com/doitsujin/dxvk/releases/download/v2.7.1/dxvk-2.7.1.tar.gz sha256: d85ce7c79f57ecd765aaa1b9e7007cb875e6fde9f6d331df799bce73d513ce87 3. Get d3dx10_43.dll, e.g. using WineTricks winetricks d3dx10_43 4. Launch and use 3DMVA-S1R5T-HXXPD-PWFPT-LF3HV-TQ5AZ-K5T5J-W1UEC to register. 5. Launch the benchmark. -- 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=59847 Artem S. Tashkinov <aros@gmx.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Keywords| |download URL| |https://benchmarks.ul.com/l | |egacy-benchmarks -- 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=59847 --- Comment #1 from Artem S. Tashkinov <aros@gmx.com> --- Created attachment 81145 --> http://bugs.winehq.org/attachment.cgi?id=81145 Crashes with isfinite not defined The first CPU test sort of works except it's rendered incorrectly. Hooray! Sadly the test crashes soon after with "isfinite not defined" -- 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=59847 --- Comment #2 from Artem S. Tashkinov <aros@gmx.com> --- Created attachment 81146 --> http://bugs.winehq.org/attachment.cgi?id=81146 The first CPU test being wrongly rendered -- 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=59847 Stian Low <wineryyyyy@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |wineryyyyy@gmail.com --- Comment #3 from Stian Low <wineryyyyy@gmail.com> --- (In reply to Artem S. Tashkinov from comment #0)
2. Download and install x86 libraries as overrides https://github.com/doitsujin/dxvk/releases/download/v2.7.1/dxvk-2.7.1.tar.gz
Wine considers dxvk 3rd party and recommends tests without it to reduce false bug reports. Please retest without DXVK to use wine builtin renderers instead. If bugs persist then report may be considered higher priority. -- 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=59847 Artem S. Tashkinov <aros@gmx.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Status|UNCONFIRMED |RESOLVED --- Comment #4 from Artem S. Tashkinov <aros@gmx.com> --- Installing d3dcompiler_43 has made it work! Only the PhysX test is unusable/freezes. (In reply to Stian Low from comment #3)
Wine considers dxvk 3rd party and recommends tests without it to reduce false bug reports.
Please retest without DXVK to use wine builtin renderers instead.
If bugs persist then report may be considered higher priority.
Wine's own D3D10 implementation is extremely incomplete, so testing with no third-party overrides/libraries makes no sense. There's an already open bug report about that. -- 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=59847 Zeb Figura <z.figura12@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Resolution|FIXED |INVALID CC| |z.figura12@gmail.com --- Comment #5 from Zeb Figura <z.figura12@gmail.com> --- (In reply to Artem S. Tashkinov from comment #4)
Wine's own D3D10 implementation is extremely incomplete,
I'm curious where you're getting this impression. I don't think this has been true for many years. 3DMark Vantage needs a missing feature, stream output overflow queries. I'm struggling to think of another missing d3d10 *feature* per se, that I've ever seen an application ask for. Moreover, I have never seen a non-benchmark application ask for stream output overflow queries.
so testing with no third-party overrides/libraries makes no sense.
Even if that were true, this bug tracker is not for third-party libraries. They are not our responsibility.
There's an already open bug report about that.
It would be helpful to post the link... for reference, this is bug 46594. -- 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=59847 --- Comment #6 from Artem S. Tashkinov <aros@gmx.com> --- (In reply to Zeb Figura from comment #5) Thanks for the clarification. My impression came mostly from practical testing and from old unresolved bugs rather than from a precise inventory of missing D3D10 features. I agree that "extremely incomplete" was probably too broad a statement if the remaining WineD3D problem here is specifically stream output overflow queries. That said, from a user perspective, long-standing gaps tend to push people toward workarounds such as DXVK, especially when those alternatives appear to work better for many real-world applications. I understand that third-party libraries are not Wine's responsibility and that this tracker is for Wine itself. To be clear, I cannot confirm that the crash at the end of the first graphics test occurs with Wine's own D3D10 implementation, because without overrides the application does not get that far. With WineD3D, it appears to fail earlier due to the missing stream output overflow query support tracked in bug 46594. So the later crash may well be DXVK-specific and outside the scope of Wine's bug tracker. But the no-overrides WineD3D failure still seems to be a real missing-feature issue, not an invalid report. Is the intended conclusion that this should simply be tracked under bug 46594 instead? There is also the separate complication that 3DMark Vantage appears to require native d3dx10_43.dll and d3dcompiler_43 to get this far. That makes relying on Wine's built-in components alone difficult even for an application from the D3D10 era. -- 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=59847 Ken Sharp <imwellcushtymelike@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED --- Comment #7 from Ken Sharp <imwellcushtymelike@gmail.com> --- Closing. -- 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=59847 --- Comment #8 from Zeb Figura <z.figura12@gmail.com> --- Sorry I missed this reply. (In reply to Artem S. Tashkinov from comment #6)
(In reply to Zeb Figura from comment #5) That said, from a user perspective, long-standing gaps tend to push people toward workarounds such as DXVK, especially when those alternatives appear to work better for many real-world applications. I understand that third-party libraries are not Wine's responsibility and that this tracker is for Wine itself.
Believe me, I am well aware that wined3d lags behind DXVK. I do not need to be reminded day in and day out that our d3d implementation isn't good enough and that it's my fault people use DXVK. I do not ask people to use wined3d instead of DXVK in general. However, if people are going to file a bug in our bug tracker, it needs to be a bug against Wine. DXVK has very clearly communicated that they do not want to be associated with Wine. Bugs in DXVK are not our responsibility. Furthermore, we have only a fraction of the bandwidth that we need to respond to user reports in general, so if a bug looks like it *might* be DXVK's, we ask the user to test with upstream Wine instead, instead of potentially spending our time debugging DXVK's bug.
So the later crash may well be DXVK-specific and outside the scope of Wine's bug tracker. But the no-overrides WineD3D failure still seems to be a real missing-feature issue, not an invalid report. Is the intended conclusion that this should simply be tracked under bug 46594 instead?
Since we already have the bug, I think so.
There is also the separate complication that 3DMark Vantage appears to require native d3dx10_43.dll and d3dcompiler_43 to get this far. That makes relying on Wine's built-in components alone difficult even for an application from the D3D10 era.
There is some leeway when it comes to native components, or DXVK for that matter; I don't always ask a user to test only builtin components if it doesn't look like they matter. It's worth noting on the other hand that Wine developers generally can't debug with Microsoft DLLs in the prefix, lest we risk seeing what calls those DLLs make under the hood. -- 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)
-
WineHQ Bugzilla