Commit Graph

5168 Commits

Author SHA1 Message Date
Alvin Wong 97aa278edb Pass window exclusive and transient properties for subwindow creation
On Windows this allows to avoid having to change the owner of the window
after it has been created, which in rare circumstances may cause the
window to bug out.
2024-07-25 00:27:27 +08:00
Rémi Verschelde 62c88a4052
Merge pull request #94671 from Chaosus/shader_fix_varying
Fix shader crash when using a varying in separate func before it defined
2024-07-24 09:59:31 +02:00
Rémi Verschelde 139cf025c9
Merge pull request #94629 from RandomShaper/fix_undef_position
Fix position from vertex shader partially uninitialized
2024-07-24 09:59:24 +02:00
Feiyue Zhang 6f30df4b6a
Linearize color if HDR 2D is on 2024-07-24 09:49:50 +02:00
Yuri Rubinsky 965b2f6ca4 Fix shader crash when using a varying in separate func before it defined 2024-07-23 21:55:18 +03:00
Rémi Verschelde e4f7b69470
Merge pull request #94656 from BlueCube3310/cubemap-layers-retrieve-fix
RenderingDevice: Fix getting cubemap layer data
2024-07-23 15:55:53 +02:00
Rémi Verschelde 29e1bdee2b
Merge pull request #94653 from mihe/bind-physics-state-sync
Bind `PhysicsServer*D::body_set_state_sync_callback`
2024-07-23 15:55:46 +02:00
BlueCube3310 741bb152ae RenderingDevice: Fix getting cubemap data 2024-07-23 13:44:28 +02:00
Mikael Hermansson a559a1c6d8 Bind `PhysicsServer*D::body_set_state_sync_callback` 2024-07-23 13:30:44 +02:00
Pedro J. Estébanez 0064532c1e Fix position from vertex shader partially uninitialized 2024-07-23 09:04:54 +02:00
Yuri Rubinsky 574e61a542 Fix crash when assigning more textures than expected to texture array 2024-07-22 20:51:11 +03:00
Rémi Verschelde 587f1d0cb0
Merge pull request #92817 from Alex2782/fix_glsl_Mali-G
Fix glsl shader for Android Mali-GXXx GPUs and Vulkan API 1.3.xxx
2024-07-22 17:24:04 +02:00
Alexander Hartmann 78ede4f8b6 Fix glsl shader for Android Mali-GXXx GPUs and Vulkan API 1.3.xxx 2024-07-20 22:01:35 +02:00
Garteal 2606831d71 SDFGI: fix crash if update_data is nullptr 2024-07-19 18:27:32 +02:00
Rémi Verschelde 47e99d55b3
Merge pull request #94388 from BastiaanOlij/expose_more_methods_scene_buffers_rd
Expose more state in RenderSceneBuffersRD
2024-07-19 11:10:46 +02:00
Rémi Verschelde 1e81a946cc
Merge pull request #94412 from dsnopek/xr-always-render
Always render when XR is enabled, even if no OS windows can draw
2024-07-18 15:38:14 +02:00
Rémi Verschelde ceb1801db2
Merge pull request #94492 from jsjtxietian/fsr-1
Add warning when use FSR1 on renderer other than forward plus
2024-07-18 10:45:50 +02:00
Rémi Verschelde 0f1e2c38a8
Merge pull request #94069 from RandomShaper/fix_vk_vali_blocky
RenderingDevice: Leave handling of compressed block granularity to the driver
2024-07-18 10:45:27 +02:00
Rémi Verschelde 590628feda
Merge pull request #86516 from jsjtxietian/fix-camera-direction
Fix incorrect `CAMERA_DIRECTION_WORLD` calculation
2024-07-18 10:45:20 +02:00
jsjtxietian 46ed290fef Add warning when use FSR1 on renderer other than forward plus 2024-07-18 12:10:18 +08:00
Adam Scott 2f5f84b470
Fix audio samples not being able to be "finished" 2024-07-17 10:45:38 -04:00
David Snopek 7a5a8597eb Always render when XR is enabled, even if no OS windows can draw 2024-07-17 08:19:21 -05:00
Rémi Verschelde a963f111cb
Merge pull request #94462 from jsjtxietian/more-type
Make ShaderLanguage's parser recognize sampler passed in from array
2024-07-17 12:24:11 +02:00
Rémi Verschelde c5e5fa3d9f
Merge pull request #94289 from clayjohn/MOBILE-multimesh-spec-constant
Use a spec constant to control whether the MultiMesh branch is used in the vertex shader.
2024-07-17 12:24:01 +02:00
Rémi Verschelde 82b57201b3
Merge pull request #94279 from clayjohn/physics-sync-warn
Only print physics sync warning on dev builds to avoid warning spam from using `move_and_slide`
2024-07-17 11:43:47 +02:00
Rémi Verschelde c2375d0b12
Merge pull request #94184 from mertkasar/ssr-rotation-fix
Fix SSR orientation issues when using orthogonal camera
2024-07-17 11:43:25 +02:00
Rémi Verschelde 0268ceaa5c
Merge pull request #94064 from bruvzg/lcd_aa_modulate
Fix LCD font AA modulation in RendererRD.
2024-07-17 11:43:03 +02:00
jsjtxietian 3859817614 Make shaderLanguage's parser recognize sampler passed in from array 2024-07-17 16:52:44 +08:00
Bastiaan Olij 9442eb8194 Expose more state in RenderSceneBuffersRD 2024-07-15 22:04:10 +10:00
clayjohn 4cbc967f50 Use a spec constant to control whether the MultiMesh branch is used in the vertex shader.
This works around a bug on the Quest3 and slightly improves performance on all mobile devices at the cost of increased pipeline count.
2024-07-12 15:44:06 -07:00
Mert Kasar 56ed2cb6d1 Fix SSR orientation issues when using orthogonal camera
- Use negative clip space values to fix reversed rotations in reflections
- Use constant -z view vector when raymarching to fix perspective in reflections
2024-07-12 21:56:31 +03:00
clayjohn 8ce4f655f4 Only print physics sync warning on dev builds to avoid warning spam from using move_and_slide
The check should be restored once we implement move_and_slide in a way that doesn't sync the physics thread
2024-07-12 10:52:30 -07:00
Rémi Verschelde 198e55e402
Merge pull request #94153 from clayjohn/SSAO-linearize
Properly linearize depth buffer for SSAO when using orthogonal camera
2024-07-11 00:27:43 +02:00
Rémi Verschelde f5bb5db8ae
Merge pull request #94135 from Calinou/vulkan-swapchain-resize-error-fail-silently
Silence Vulkan "Unable to acquire framebuffer." swapchain error
2024-07-11 00:27:36 +02:00
Rémi Verschelde c5ca288771
Merge pull request #94121 from BastiaanOlij/remove_render_thread_guard
Remove our `ERR_ON_RENDER_THREAD` guard, it is not reliable
2024-07-11 00:27:21 +02:00
Rémi Verschelde 46e8377a5e
Merge pull request #93960 from BastiaanOlij/fix_mobile_compositor
Pre transparent compositor effects needs to run later
2024-07-11 00:27:06 +02:00
clayjohn e0f736ec10 Properly linearize depth buffer for SSAO when using orthogonal camera 2024-07-09 15:29:45 -07:00
Yuri Rubinsky af5fc8354b Fix a bunch of orphan StringName errors at ProjectSettings/Editor exit 2024-07-09 19:38:52 +03:00
Hugo Locurcio cf2d2e2dd8
Silence Vulkan "Unable to acquire framebuffer." swapchain error
This error is harmless and can be safely ignored.
2024-07-09 17:12:54 +02:00
Rémi Verschelde 40cb283dd7
Merge pull request #93866 from Chaosus/shader_fix_global_func_set
Improve code for setup of `global_func_set` in `ShaderLanguage`
2024-07-09 16:47:13 +02:00
Bastiaan Olij bf8c85e13b Remove our ERR_ON_RENDER_THREAD guard, it is not reliable 2024-07-09 18:48:00 +10:00
Rémi Verschelde 6f9c0aa40d
Merge pull request #93630 from BastiaanOlij/scene_data_projection_correction
Make RenderSceneData take projection correction into account
2024-07-09 00:03:14 +02:00
Pedro J. Estébanez 00e89229bf RenderingDevice: Leave handling of compressed block granularity to the driver 2024-07-08 11:58:51 +02:00
bruvzg 4790e12f66
Fix LCD font AA modulation in RendererRD. 2024-07-08 11:17:31 +03:00
Adam Scott a38f30fbd5 Fix Web samples finished missing signal 2024-07-07 14:47:54 -04:00
Rémi Verschelde 7197678df2
Merge pull request #91551 from m4rr5/configure_timestamp_query_elements
Add a project setting to configure the maximum number of timestamps.
2024-07-07 12:38:19 +02:00
Bastiaan Olij e42870b4d1 Pre transparent compositor effects needs to run later 2024-07-05 12:37:26 +10:00
Rémi Verschelde 099b9b2e85
Merge pull request #93916 from zeux/aabb-zfight
Fix AABB computation for position compression to not depend on vertex order
2024-07-04 23:27:13 +02:00
Arseny Kapoulkine 4e9e35b58a Fix AABB computation for position compression to not depend on vertex order
The previous computation was dependent on the vertex order in two ways:

- If the first vertex was on the AABB boundary, the AABB would be
  increased by the epsilon due to size clamping
- Every time the AABB would get expanded, we would recompute end from
  size and reconstruct size again, which resulted in slow floating point
  drift.

In isolation this may not seem like it matters, but it means that the
same mesh with a different vertex order may result in a slightly different
AABB. This can be a significant issue due to shadow meshes and their use in
depth prepass: shadow meshes reorder vertex data as part of the
deduplication process, as they append one unique position at a time and
as such remove the duplicate positions; this can result in a different
AABB which would result in a different reconstructed vertex position
during a depth pre-pass, causing mesh self-occlusion.
2024-07-03 23:53:24 -07:00
Yuri Rubinsky 4f657f5c6d Improve code for setup of `global_func_set` in `ShaderLanguage` 2024-07-02 18:16:04 +03:00