April 2024 - Wine-gitlab - list.winehq.org

[PATCH v3 0/2] MR2724: wined3d: Submit command buffers that are built for longer than 2 ms.
by Zebediah Figura (＠zfigura) 28 May '24

28 May '24

This improves performance for the game "Grounded", on a AMD Radeon RX 6700 XT, with radv from Mesa 22.3.6. Testing was done with the "cb_access_map_w" option enabled, which also improves performance with the game by itself. >From my testing, it's possible to raise the threshold from 2 ms up to 5 ms or so, before the driver or GPU seems to reclock back to the lower power level. However, this measurement is questionable for several reasons. It seems to vary depending on the scene being rendered, and of course this will be specific to the game and driver and GPU in question anyway. The game also has a weird approach to vsync that seems to involve it presenting stale frames (and hence artificially inflating the FPS), which I'm not fully sure I accounted for while measuring. And of course, it's hard to be sure that 5 ms is actually the threshold for how long the driver will go before powering down the GPU. In any case, it seems better to err on the side of submitting more often, to make sure the fix affects more drivers. While submission isn't cheap, it seems to me that submitting every 2 ms is unlikely to cause a bottleneck [consider that this is at most 8 (more) submissions per frame]. The maximum of 4 concurrent periodically submitted buffers was chosen arbitrarily. Removing the maximum altogether does not measurably affect performance for this game either way. Credit goes to Philip Rebohle and his work on DXVK for helping me to notice that periodic submission might make a difference. -- v3: wined3d: Submit command buffers after 512 draw or dispatch commands. wined3d: Retrieve the VkCommandBuffer from wined3d_context_vk after executing RTV barriers. https://gitlab.winehq.org/wine/wine/-/merge_requests/2724

4 4

[PATCH v2 0/1] MR5070: win32u: Cache is_virtual_desktop.
by Gabriel Ivăncescu 28 May '24

28 May '24

It can be substantially faster in some cases. Notepad++ (with several plugins) start up goes down from around ~1.03 sec to ~0.61 sec on my machine (with wineprefix active). 9b7669592d6f8b40976b571b70f8543777d35167 is what introduced this performance regression. I don't know why exactly, but it made my Notepad++ noticeably slower when launched from a file manager (so it's not just at wineprefix startup overhead), which is very annoying. -- v2: win32u: Cache is_virtual_desktop. https://gitlab.winehq.org/wine/wine/-/merge_requests/5070

2 7

[PATCH v4 0/2] MR2120: user32: Post WM_INPUT_DEVICE_CHANGE when registering for notifications
by Evan Tang (＠etang-cw) 28 May '24

28 May '24

This seems to be relied on by some versions of [this Unreal Engine input plugin](https://www.unrealengine.com/marketplace/en-US/product/wm-input-man… Note: I'm not sure how to deal with `HID_USAGE_GENERIC_KEYPAD`, which (I think) would fall under `RIM_TYPEKEYBOARD`. Do we need to store extra info to differentiate these from `HID_USAGE_GENERIC_KEYBOARD` or is there something in the device info struct that can differentiate them? -- v4: user32: Post WM_INPUT_DEVICE_CHANGE when registering for notifications user32: Add tests for WM_INPUT_DEVICE_CHANGE messages https://gitlab.winehq.org/wine/wine/-/merge_requests/2120

3 4

[PATCH 0/3] MR5383: oleaut32: Load bitmap as DIB.
by Daniel Lehman (＠dlehman25) 28 May '24

28 May '24

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/5383

3 8

[PATCH v3 0/1] MR5333: vcruntime140_1: Pass original frame to catch block on stack.
by Daniel Lehman (＠dlehman25) 28 May '24

28 May '24

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=56451 local variables with alignas(32) or higher may cause the unwinding code to fetch the original frame from a different spot on the stack from where the catch block code places it -- v3: vcruntime140_1: Pass original frame to catch block on stack. https://gitlab.winehq.org/wine/wine/-/merge_requests/5333

3 3

[PATCH v3 0/5] MR5551: win32u: Get rid of most drivers vulkan surface wrappers.
by Rémi Bernon 28 May '24

28 May '24

-- v3: winex11: Remove now unnecessary surface wrapper struct. win32u: Move thread detach from winex11. win32u: Introduce a per-window vulkan surface list. winewayland: Get rid of the now unnecessary surface wrapper. win32u: Return the host surface directly from vulkan_surface_create. https://gitlab.winehq.org/wine/wine/-/merge_requests/5551

3 8

[PATCH 0/1] MR5559: include: Fix another 13 typoed GUIDs
by Alfred Agrell (＠Alcaro) 28 May '24

28 May '24

Same bug class as https://gitlab.winehq.org/wine/wine/-/merge_requests/5550. None of them are referenced by anything (not counting mfplat/main.c's debugstr_attr(), which names the two in mfapi.h - better debug output, yay I guess). -- https://gitlab.winehq.org/wine/wine/-/merge_requests/5559

4 5

[PATCH 0/1] MR5556: msvcr80/tests: Fix errno access in tests.
by Piotr Caban (＠piotr) 28 May '24

28 May '24

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/5556

2 3

[PATCH v4 0/2] MR5547: msvcrt: Implement _mbsncpy_s[_l]().
by Paul Gofman (＠gofman) 28 May '24

28 May '24

ucrtbase._mbsncpy_s is used by Marvel vs Capcom when trying to create multiplayer lobby. The functions are also present in msvcrt (unlike msvcr70, msvcr71) where I didn't add it because it behaves differently: there is at least one weirdness when it doubles the number of characters to copy ('n' parameter, not buffer size). I suppose we don't need to explore and deal with this specific until something needs those functions from msvcrt. -- v4: msvcrt: Implement _mbsncpy_s[_l](). https://gitlab.winehq.org/wine/wine/-/merge_requests/5547

4 5

[PATCH v2 0/3] MR5535: advapi32: CryptSetKeyParam() should accept NULL pbData.
by Dmitry Timoshkov (＠dmitry) 28 May '24

28 May '24

At least CryptSetKeyParam(KP_X) should accept NULL pbData. -- v2: advapi32: CryptSetKeyParam() should accept NULL pbData. rsaenh: Validate pbData in CPSetKeyParam(). dssenh: Add CPSetKeyParam() stub implementation. https://gitlab.winehq.org/wine/wine/-/merge_requests/5535

4 6