[Bug 54150] New: d3dcompiler_43:hlsl_d3d11 & d3dcompiler_43:hlsl_d3d11 - test_trig() fails on w11pro64_nv
https://bugs.winehq.org/show_bug.cgi?id=54150
Bug ID: 54150 Summary: d3dcompiler_43:hlsl_d3d11 & d3dcompiler_43:hlsl_d3d11 - test_trig() fails on w11pro64_nv Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: d3d Assignee: wine-bugs@winehq.org Reporter: fgouget@codeweavers.com
d3dcompiler_43:hlsl_d3d11 & d3dcompiler_43:hlsl_d3d11 - test_trig() fails on w11pro64_nv:
hlsl_d3d11.c:504: Test failed: Test 300: Got {-9.99755323e-001, -2.21209023e-002, 0.00000000e+000, 0.00000000e+000}, expected {-9.99755859e-001, -2.20966193e-002, 0.00000000e+000, 0.00000000e+000}. hlsl_d3d11.c:504: Test failed: Test 600: Got {4.42310758e-002, -9.99021292e-001, 0.00000000e+000, 0.00000000e+000}, expected {4.41824496e-002, -9.99023479e-001, 0.00000000e+000, 0.00000000e+000}.
See https://test.winehq.org/data/patterns.html#d3dcompiler_43:hlsl_d3d11 https://test.winehq.org/data/patterns.html#d3dcompiler_47:hlsl_d3d11
This test only fails on w11pro64_nv: there is no issue on the other test configurations, particularly on w11pro64_amd.
(-2.21209023e-002 - -2.20966193e-002) / -2.20966193e-002 = ~ 0.0011
Should the test should allow for a larger imprecision? But the first y result feels like it's off by a lot already.
https://bugs.winehq.org/show_bug.cgi?id=54150
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |z.figura12@gmail.com Keywords| |source, testcase
https://bugs.winehq.org/show_bug.cgi?id=54150
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Summary|d3dcompiler_43:hlsl_d3d11 & |d3dcompiler_43:hlsl_d3d11 & |d3dcompiler_43:hlsl_d3d11 - |d3dcompiler_47:hlsl_d3d11 - |test_trig() fails on |test_trig() fails on |w11pro64_nv |w11pro64_nv
https://bugs.winehq.org/show_bug.cgi?id=54150
--- Comment #1 from François Gouget fgouget@codeweavers.com --- Created attachment 73836 --> https://bugs.winehq.org/attachment.cgi?id=73836 Allow more imprecision in test_trig() for NVIDIA.
This may be a bit heavy handed.
https://bugs.winehq.org/show_bug.cgi?id=54150
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #73836|0 |1 is obsolete| |
--- Comment #2 from François Gouget fgouget@codeweavers.com --- Created attachment 73862 --> https://bugs.winehq.org/attachment.cgi?id=73862 Allow more imprecision in test_trig() for NVIDIA.
Attach the right patch this time!
https://bugs.winehq.org/show_bug.cgi?id=54150
Zeb Figura z.figura12@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|d3d |d3d-util
--- Comment #3 from Zeb Figura z.figura12@gmail.com --- It's a lot of error, but apparently legal. Microsoft says "The maximum absolute error is 0.0008 in the interval from -100*Pi to +100*Pi." Maybe we should be comparing float values instead of the mantissa. Either way, increasing the tolerance seems reasonable enough.
https://bugs.winehq.org/show_bug.cgi?id=54150
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Fixed by SHA1| |edf88aa9df420e20fb07bf5fa18 | |e7990593d2615 Status|NEW |RESOLVED
--- Comment #4 from François Gouget fgouget@codeweavers.com --- This is fixed by the commit below:
commit edf88aa9df420e20fb07bf5fa18e7990593d2615 Author: François Gouget fgouget@codeweavers.com AuthorDate: Thu Jan 5 11:45:45 2023 +0100
d3dcompiler/tests: Allow more imprecision in test_trig() for NVIDIA.
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=54150
https://bugs.winehq.org/show_bug.cgi?id=54150
Alexandre Julliard julliard@winehq.org changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #5 from Alexandre Julliard julliard@winehq.org --- Closing bugs fixed in 8.2.
-
WineHQ Bugzilla