Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

RFC: Fix intel libxess on intel GPUs #215

Open
wants to merge 2,984 commits into
base: bleeding-edge
Choose a base branch
from
Open

RFC: Fix intel libxess on intel GPUs #215

wants to merge 2,984 commits into from

Conversation

Etaash-mathamsetty
Copy link
Contributor

this is not a completely finished PR, a mechanism needs to be found to load igdext64 on the initialization of dxgi. (probably in dxvk). This is more of a question on whether something like this will be accepted into proton.

What I currently know:

  • this disables the xess fallback entirely
  • force_vendor_id = -1 is no longer required on intel
  • first patch is probably fine as-is

What I do not know:

  • number of games this could break

Paul Gofman and others added 30 commits December 19, 2023 18:59
@rbernon
Revert "mfplat/buffer: Validate buffer size in image copying functions."
baefbda 
This reverts commit eefc50a.

CW-Bug-Id: #22084
@rbernon
Revert "mfplat/buffer: Prepare image copying functions for error chec…
6d86380 
…king."

This reverts commit 3b621e5.

CW-Bug-Id: #22084
@ziqingh @rbernon
mfplat: Support YVYU, NV11, MEDIASUBTYPE_RGB* media types.
0704245 
(cherry picked from commit df26eae)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat: Fix locking flags usage for D3D9 buffers.
c4d9d33 
(cherry picked from commit 91c0eda)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat: Do not mark a DXGI buffer as locked if surface mapping fails.
0ca6d7e 
Analogous to 44c9ea5.

(cherry picked from commit df660e8)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat: Fix locking flags usage for DXGI buffers.
8054421 
(cherry picked from commit 071eb50)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat: Only download surface data from GPU for DXGI buffers when rea…
f9339fa 
…ding.

(cherry picked from commit 49b8e55)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat: Only upload surface data to GPU for DXGI buffers when writing.
c6a3b8e 
(cherry picked from commit d073d3e)

CW-Bug-Id: #22084
@rbernon
mfplat: Fix returned buffer length in dxgi_surface_buffer_lock().
132d396 
(cherry picked from commit 5a162f2)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat/buffer: Use the appropriate image copy function for NV11.
7aaa1a1 
(cherry picked from commit 6d1fc90)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat/buffer: Support YV12, I420 and IYUV image formats.
e4758fa 
(cherry picked from commit 4805546)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat/buffer: Implement IMF2DBuffer::ContiguousCopyFrom().
29d1bb4 
This is a generic implementation, which is probably fine for buffers
backed by system memory. The implementation for buffers backed by
GPU memory can probably be optimized.

(cherry picked from commit da1885c)

CW-Bug-Id: #22084
@giomasce @rbernon
mfplat/buffer: Implement IMF2DBuffer::ContiguousCopyTo().
fbfc957 
(cherry picked from commit 885d8df)

CW-Bug-Id: #22084
@rbernon
mfplat/buffer: Use absolute pitch in memory_1d_2d_buffer_[Un]Lock().
1e13d48 
CW-Bug-Id: #22084
@rbernon
mfplat/tests: Add more tests for copying 2d buffers.
47bdf47 
CW-Bug-Id: #22084
@rbernon
mfplat/buffer: Do not flip in memory_2d_buffer_ContiguousCopy{From|To…
f15a020 
…}().

CW-Bug-Id: #22084
@rbernon
mfplat/sample: Refactor sample_CopyToBuffer().
999d13e 
CW-Bug-Id: #22084
@rbernon
mfplat/sample: Optimize copying to 2d buffer.
63bbee2 
CW-Bug-Id: #22084
@rbernon
Revert "HACK: msmpeg2vdec: Refuse to load DLL for some games."
a1b00ad 
This reverts commit 7267a0e.

CW-Bug-Id: #22084
@ivyl @rbernon
Revert "nsiproxy.sys: Detect wireless interface type on Linux."
99a37ed 
This reverts commit e7fdbda.

This commit is already included as a cherry-pick. Upon experimental
rebase only whitespace removal was left - we can remove that.
@rbernon
dinput: Assume that clipping the cursor succeeds.
df85dd2 
Fixes a regression (in Deathloop) where it falls back to SetCursorPos
as we now sometimes shrink the clipping rectangle.

CW-Bug-Id: #21707
@rbernon
winex11: Always ignore MotionNotify event after SetCursorPos.
d728d00 
Trying to workaround spurious Wayland mouse motion.

CW-Bug-Id: #22650
@rbernon
dinput: Only call SetCursorPos if clipping fails.
37f9bee 
Fixes mouse sometimes jumping around in Mechwarrior Online on
KDE/Wayland.

CW-Bug-Id: #22650
@rbernon
mf/tests: Add tests for AAC decoder with different input number of ch…
5151c81 
…annels.

CW-Bug-Id: #22717
@rbernon
winegstreamer: Handle missing or zero channel count in _GetOutputAvai…
427799a 
…lableType in AAC decoder.

CW-Bug-Id: #22717
@rbernon
winegstreamer: Correct output available types attrs in AAC decoder fo…
a559dc9 
…r channel count > 2.

CW-Bug-Id: #22717
@rbernon
winegstreamer: Validate maximum channel count in _SetInputType in AAC…
b40d10d 
… decoder.

CW-Bug-Id: #22717
@rbernon
winegstreamer: Also return output with 2 channels for multichannel in…
a048c58 
…puts from AAC decoder.

CW-Bug-Id: #22717
@rbernon
Revert "include: Don't align the stack for PE builds."
f016c65 
This reverts commit 87fbf33.

We saw hangs in Phasmophobia with the commit, and then discovered this commit is
not strictly necessary for ntlea, so we revert it.
@rbernon
mf/session: Use local variables to access transform node streams.
3aba72f 
(cherry picked from commit b3e4a52)

CW-Bug-Id: #21713
rbernon and others added 19 commits January 11, 2024 11:55
@rbernon
HACK: windows.gaming.input: Emulate Steam Input native hooks.
222ada6 
CW-Bug-Id: #23185
@rbernon
HACK: xinput: Emulate Steam Input native hooks.
fc02b27 
CW-Bug-Id: #23185
@rbernon
HACK: winebus: Expose virtual controller for Steam Input hooks emulat…
2e778c6 
…ion.

Guarded behind a PROTON_EXPOSE_STEAM_CONTROLLER=1 variable.

CW-Bug-Id: #23185
@rbernon
HACK: win32u: Report underlying VID/PID in Steam Input hooks emulation.
f9a2920 
CW-Bug-Id: #23185
@rbernon
HACK: windows.gaming.input: Report underlying VID/PID in Steam Input …
20af009 
…hooks emulation.

CW-Bug-Id: #23185
@rbernon
fixup! HACK: winebus: Expose virtual controller for Steam Input hooks…
b4d1730 
… emulation.
@rbernon
fixup! fixup! HACK: winebus: Expose virtual controller for Steam Inpu…
fd607b5 
…t hooks emulation.
@rbernon
winebus: Disable SDL2.30 new behavior for the SDL unix backend.
8ad4687 
CW-Bug-Id: #23185
@rbernon
fixup! HACK: ntdll: Emulate Steam Input native hooks.
a56aa3a
@julliard
ntdll/tests: Simplify the KiUserExceptionDispatcher tests a little.
d9ebf11 
(cherry picked from commit 8164c70)

CW-Bug-Id: #23245
@julliard
ntdll: Call Wow64PrepareForException directly from KiUserExceptionDis…
bfa4856 
…patcher.

This requires moving the stack switch to Wow64PrepareForException.

Based on a patch by Paul Gofman.

(cherry picked from commit 334f54c)

CW-Bug-Id: #23245
@julliard
ntdll: Allow unwind rules to override the default stack pointer assig…
a179a88 
…nment.

(cherry picked from commit 93a3a6c)

CW-Bug-Id: #23245
@julliard
ntdll: Add a machine frame to the KiUserExceptionDispatcher stack on …
fe3d35d 
…x86-64.

(cherry picked from commit 061c612)

CW-Bug-Id: #23245
@julliard
ntdll: Add a machine frame to the KiUserApcDispatcher stack on x86-64.
50780ff 
(cherry picked from commit 8b24139)
@rbernon
winebus.sys: Ignore events on unsupported axes.
67e62ab 
Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55660
(cherry picked from commit 7d74938)
@rbernon
winebus.sys: Fix incorrect hid_device_set index check.
221a234 
Instead of checking it in set_report_from_joystick_event.

(cherry picked from commit bbdec47)
@rbernon
evr: Use D3DCREATE_MULTITHREADED device creation flag.
78ca70b 
CW-Bug-Id: #23231
@Etaash-mathamsetty
cfgmgr32: improve device id stubs.
b3537a5
@Etaash-mathamsetty
igdext64: add stub dll.
815073d
@Etaash-mathamsetty Etaash-mathamsetty changed the title RFC: Fix intel xess loader RFC: Fix intel libxess on intel GPUs Jan 19, 2024
@Plagman Plagman force-pushed the bleeding-edge branch 4 times, most recently from 03bcca0 to 24d3d22 Compare July 19, 2024 17:11
@Plagman Plagman force-pushed the bleeding-edge branch 3 times, most recently from 97aa0e7 to 4a22a2a Compare July 30, 2024 18:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.