Commit graph

6882 commits

Author SHA1 Message Date
ameerj
276de8a751 vk_blit_screen: Minor refactor of filter pipeline selection 2021-12-05 15:35:35 -05:00
ameerj
c701429fb4 Revert "Merge pull request #7395 from Morph1984/resolve-comments"
This reverts commit 0442db9ad6, reversing
changes made to 15212de85a.
2021-12-05 15:35:35 -05:00
Feng Chen
4a3c1192b8 Address feedback 2021-12-05 00:06:14 +08:00
Fernando Sahmkow
3b1694245f Texture Cache: Fix crashes on NVIDIA. 2021-12-04 11:26:58 +01:00
bunnei
94dddbc155 Merge pull request #7467 from liushuyu/fix-linux-decoding
video_core/codecs: more robust ffmpeg hwdecoder selection logic
2021-12-03 17:11:12 -08:00
liushuyu
18ddfcc0cf video_core/cmake: link against libva explicitly ...
... to fix build on Flatpak (and self-builds)
2021-12-02 22:35:30 -07:00
liushuyu
168882cb7a video_core/codecs: more fixes for VAAPI detection ...
* skip impersonated VAAPI implementaions ("imposter detection")
* place VAAPI priority below CUDA/NVDEC/CUVID
2021-12-02 21:31:51 -07:00
liushuyu
ef00c213e8 video_core/codec: address comments 2021-12-02 21:01:34 -07:00
liushuyu
a32139fdab video_core/codecs: more robust ffmpeg hwdecoder selection logic 2021-12-02 21:01:34 -07:00
Morph
2b9afa4d56 general: Replace high_resolution_clock with steady_clock
On some OSes, high_resolution_clock is an alias to system_clock and is not monotonic in nature. Replace this with steady_clock.
2021-12-02 14:20:43 -05:00
Feng Chen
10cc89bfdf Support multiple videos playing 2021-12-02 12:48:42 +08:00
Feng Chen
2e834bfa4a Add missing pixel format mapping 2021-11-29 12:39:37 +08:00
Fernando S
d15bbac985 Merge pull request #7396 from FernandoS27/blit-this-mf
TextureCache: Eliminate format deduction as full depth conversion has been supported.
2021-11-28 12:16:31 +01:00
Fernando Sahmkow
4e74fa32a4 Texture Cache: Secure insertions against deletions. 2021-11-28 10:53:27 +01:00
Fernando Sahmkow
fc91c0f51d Texture Cache: Redesigning the blitting system (again). 2021-11-27 11:22:16 +01:00
Fernando Sahmkow
dcce920306 Texture Cache: Further fix regressions. 2021-11-26 17:03:48 +01:00
liushuyu
1a5c1d70de video_core/codec: address comments 2021-11-24 18:06:38 -07:00
liushuyu
f91cc356fb video_core/codecs: fix multiple decoding issues on Linux ...
* when someone installed Intel video drivers on an AMD system, the
  decoder will select the Intel VA-API decoding driver and yuzu will
  crash due to incorrect driver selection; the fix will check if the
  currently about-to-use driver is loaded in the kernel
* when using NVIDIA driver on Linux with a ffmpeg that does not have
  CUDA capability enabled, the decoder will crash; the fix simply
  making the decoder prefers the VDPAU driver over CUDA on Linux
2021-11-24 17:23:57 -07:00
Fernando Sahmkow
81f52ea8fc Texture Cache: Fix issue with blitting 3D textures. 2021-11-22 06:07:21 +01:00
Fernando Sahmkow
dc87895c22 Texture Cache: Correct conversion shaders. 2021-11-22 00:21:42 +01:00
Fernando Sahmkow
c03917860b Texture Cache: Always copy on NVIDIA. 2021-11-22 00:06:56 +01:00
Fernando Sahmkow
09a236dbbf TextureCache: Simplify blitting of D24S8 formats and fix bugs. 2021-11-22 00:00:01 +01:00
Fernando Sahmkow
eae2158c12 VulkanTexturECache: Use reinterpret on D32_S8 formats. 2021-11-21 21:09:49 +01:00
Fernando Sahmkow
5b92772fe6 HostShaders: Fix D24S8 convertion shaders. 2021-11-21 21:04:04 +01:00
bunnei
0442db9ad6 Merge pull request #7395 from Morph1984/resolve-comments
general: Resolve comments in PR #7368
2021-11-21 02:42:24 -08:00
bunnei
15212de85a Merge pull request #7389 from ameerj/screenshot-1x
Fix screenshot dimensions when at 1x scale
2021-11-21 02:31:32 -08:00
Fernando Sahmkow
df62d93979 TextureCache: Eliminate format deduction as full depth conversion has been supported. 2021-11-21 05:37:01 +01:00
Morph
e0c4af1b7b vk_texture_cache: Mark VkBufferUsageFlags as static constexpr 2021-11-20 21:49:37 -05:00
Morph
feba9b7300 vk_blit_image: Consolidate CreatePipelineTargetEx functions 2021-11-20 21:18:37 -05:00
bunnei
9e4707b508 Merge pull request #7368 from FernandoS27/vulkan-conv
Fix ART Blit detection regression and add D24S8 <-> RGBA8 conv to Vulkan
2021-11-20 16:51:13 -08:00
ameerj
69759e55ee Fix screenshot dimensions when at 1x scale
This was regressed by ART.
Prior to ART, the screenshots were saved at the title's framebuffer resolution. A misunderstanding of the existing logic led to screenshot dimensions becoming dependent on the host render window size.

This changes the behavior to match how it was prior to ART at 1x, with screenshots now always being the title's framebuffer dimensions scaled by the resolution scaling factor.
2021-11-20 17:50:24 -05:00
Fernando Sahmkow
e6d7d48d78 TextureCache: Refactor and fix linux compiling. 2021-11-20 14:46:19 +01:00
Fernando Sahmkow
1bd224dda0 TextureCache: Assure full conversions on depth/stencil write shaders. 2021-11-20 06:17:01 +01:00
Fernando Sahmkow
e926797029 TextureCache: Implement buffer copies on Vulkan. 2021-11-20 06:15:29 +01:00
bunnei
cee5a7b762 Merge pull request #7294 from vonchenplus/fix_image_update_error_when_width_too_small
Fix image update/download error when width too small
2021-11-19 15:56:27 -08:00
Fernando Sahmkow
df1eabfc53 TextureCache: Add R16G16 to D24S8 converter. 2021-11-20 00:02:12 +01:00
Fernando Sahmkow
4241b9e26c TextureCache: Add B10G11R11 to D24S8 converter. 2021-11-19 23:22:44 +01:00
Fernando Sahmkow
efe634608e TextureCache: Further fixes on resolve algorithm. 2021-11-19 23:02:04 +01:00
Feng Chen
dbfa89c40b Implement convert legacy to generic 2021-11-19 22:53:58 +08:00
bunnei
2b5f99cebd Merge pull request #7357 from Morph1984/s8_uint
video_core: Implement S8_UINT format
2021-11-19 01:16:49 -08:00
Fernando Sahmkow
5e58fd30b6 TextureCache: Implement additional D24S8 convertions. 2021-11-19 06:27:44 +01:00
Fernando Sahmkow
57e1775bc8 TextureCache: force same image format when resolving an image. 2021-11-19 05:46:57 +01:00
Fernando Sahmkow
37459e821b TextureCache: Fix regression caused by ART and improve blit detection algorithm to be smarter. 2021-11-19 03:17:54 +01:00
Fernando Sahmkow
a21d1447ab Vulkan: implement D24S8 <-> RGBA8 convertions. 2021-11-19 03:17:02 +01:00
Morph
39a49776b6 renderer_vulkan: Implement S8_UINT stencil format
It should be noted that on Windows, only nvidia gpus support this format natively as of this commit.
2021-11-18 00:05:51 -05:00
ameerj
b31bfc9ca5 gl_texture_cache: Round format conversion PBO to next power of 2 2021-11-17 23:49:44 -05:00
Morph
be263c113c renderer_opengl: Implement S8_UINT stencil format 2021-11-17 15:05:07 -05:00
Morph
c7bdf8825f video_core: Add S8_UINT stencil format 2021-11-17 15:04:38 -05:00
Feng Chen
f4ee936f8b Fix image update/download error when width too small 2021-11-17 12:21:17 +08:00
ameerj
ba5778b0e3 texture_cache: Use pixel format conversion when supported by the runtime 2021-11-16 22:32:46 -05:00