Commit Graph

321 Commits

Author SHA1 Message Date
Rémi Verschelde 7ebc866418
Merge pull request #91545 from clayjohn/RD-shadow-atlas
Properly set size of shadow atlas quadrant when subdivision is 8 or higher.
2024-05-04 11:52:45 +02:00
clayjohn caeef3473b Properly set size of shadow atlas quadrant when subdivision is 8 or higher.
Also fix renderpass rect validation
2024-05-03 17:11:05 -07:00
Bastiaan Olij 9042ddf19f Improvements to VRS/Foveated rendering 2024-05-03 17:20:30 +10:00
A Thousand Ships 308dbb8c63
[Core] Add scalar versions of `Vector*` `min/max/clamp/snap(ped)`
Convenience for a number of cases operating on single values
2024-05-02 10:31:13 +02:00
Chris Clyne 44364faee5 Ensure global shader sampler parameters are initialized when loading the editor 2024-05-01 18:12:46 +01:00
A Thousand Ships 31e7ee63f2
Fix unsafe uses of `Callable.is_null()`
`Callable.is_null()` is not equivalent to `!Callable.is_valid()` and
doesn't guarantee the call is valid.
2024-04-27 16:22:57 +02:00
clayjohn adfc9ccc44 Use a src rect for copying from screen with CanvasGroup in the mobile backend 2024-04-17 15:19:20 -07:00
Juan Linietsky 65686dedf9 Use WorkerThreadPool for Server threads
* Servers now use WorkerThreadPool for background computation.
* This helps keep the number of threads used fixed at all times.
* It also ensures everything works on HTML5 with threads.
* And makes it easier to support disabling threads for also HTML5.

CommandQueueMT now syncs with the servers via the WorkerThreadPool
yielding mechanism, which makes its classic main sync semaphore
superfluous.

Also, some warnings about calls that kill performance when using
threaded rendering are removed because there's a mechanism that
warns about that in a more general fashion.

Co-authored-by: Pedro J. Estébanez <pedrojrulez@gmail.com>
2024-04-10 18:47:42 +02:00
Rémi Verschelde 5c75fa260f
Merge pull request #90458 from clayjohn/RD-weight-buffer
Clear RIDs of weight buffers when freeing to avoid double free
2024-04-10 14:22:37 +02:00
Rémi Verschelde c59f493620
Merge pull request #89729 from jitspoe/master.shadow_distance_fade_optimization
Shadow fade for omni lights actually stops the shadow from updating while faded out to improve performance.
2024-04-10 14:21:56 +02:00
clayjohn bafbde9f4b Clear RIDs of weight buffers when freeing to avoid double free 2024-04-09 17:52:05 -07:00
Rémi Verschelde 64c4a53a2c
Merge pull request #89849 from EIREXE/cute_and_funny_decals
Fix decal modulate being passed as srgb instead of as linear color.
2024-04-04 14:31:53 +02:00
Rémi Verschelde a28be933ee
Merge pull request #89398 from clayjohn/visual-debugger-coverage
Increase coverage of timestamps for visual profiler
2024-04-04 14:31:19 +02:00
Khasehemwy d950f5f838
Use Reverse Z for the depth buffer 2024-04-04 13:54:15 +02:00
clayjohn ac6c648645 Increase coverage of timestamps for visual profiler 2024-04-03 19:29:02 -07:00
jitspoe a0969a0931 Shadow fade for omni lights actually stops the shadow from updating while faded out to improve performance. 2024-03-25 21:25:56 -04:00
Álex Román Núñez 2f0ba1ed3f Fix decal modulate being passed as srgb instead of as linear color. 2024-03-24 14:59:43 +01:00
Rémi Verschelde 70c05f63da
Merge pull request #89531 from sullyscience/fix-resource-leaks
Fix mobile renderer RID leaks
2024-03-24 01:16:48 +01:00
A Thousand Ships 79ba22a73f
Use `Vector*` component-wise `min/max/clamp` functions where applicable 2024-03-20 13:47:42 +01:00
David Sullivan dfa326ef55 Fix: Address mobile resource leaks in render
Update render_forward_mobile destructor to include instance buffers.
Update render_scene_buffers_rd cleanup to include weight buffers / blur
textures.
2024-03-15 14:48:12 -04:00
Rémi Verschelde cebaf9dac5
Merge pull request #89042 from permelin/fix-particle-trail-activation
Fix early activation of particle trail sections
2024-03-08 13:09:25 +01:00
Rémi Verschelde a52e575652
Merge pull request #89134 from BastiaanOlij/fix_recursive_reflection_probes
Fix never ending loop with overlapping probes
2024-03-05 09:56:05 +01:00
Bastiaan Olij a5d3d23db4 Fix never ending loop with overlapping probes 2024-03-05 10:07:56 +11:00
Per Melin 853935a5c9 Fix error in AABB calculation for particles with USERDATA
Selecting "Generate AABB" on a 3D particle node in the editor would not work
and printed an error about incorrect buffer size if the particle shader used
one or more of the USERDATA build-ins.
2024-03-02 14:26:09 +01:00
Per Melin 9b2d77a238 Initialize particle trail history frame numbers
Fixes #88712

All particle trail sections were activated at once on the first cycle, instead
of being spread out over the trail's lifetime.
2024-03-01 19:32:43 +01:00
Robert Yevdokimov 13e82094ee Remove word duplicates in comments and strings, and fix casing and punctuation 2024-02-23 17:28:28 -05:00
Rémi Verschelde 292f4c7f97
Merge pull request #80214 from BastiaanOlij/rendering_effect
Implement hooks into renderer
2024-02-20 23:59:27 +01:00
Arman Elgudzhyan 533eab89ab Fixed reversed custom AABB null check when recalculating multimesh AABBs 2024-02-18 12:07:23 -08:00
Bastiaan Olij 5a98845655 Implement hooks into renderer 2024-02-18 21:54:21 +11:00
Eidolon 08f4560e69 Add optional depth fog 2024-02-17 22:39:34 -03:00
Rémi Verschelde 59643bdb86
Merge pull request #88325 from akien-mga/texture-formats
Export: Unify settings for PC texture formats, removed obsoleted ETC feature
2024-02-17 00:23:45 +01:00
Arman Elgudzhyan 7ac8365e11 Support custom AABB within MultiMesh resources
- Supporting custom AABB on the MultiMesh resource itself allows us to prevent costly runtime AABB recalculations.
- Should also help improve CPU Particle performance.
2024-02-15 22:37:07 -08:00
Rémi Verschelde e457f41f9b
Remove code relative to obsoleted ETC texture format
Co-authored-by: BlueCube3310 <53150244+BlueCube3310@users.noreply.github.com>
2024-02-14 17:16:41 +01:00
Rémi Verschelde 5e58bcd793
Merge pull request #88205 from 0x0ACB/thread_safe_texture_rid
Make `RID_Owner<Texture>` threadsafe in `TextureStorage` for GLES3
2024-02-13 23:43:40 +01:00
Erik Johnson eab95698cd
Fix shadows for billboarded Sprite3D's 2024-02-13 10:36:38 +01:00
ACB 09d2c09434 Make `RID_Owner<Texture>` threadsafe in `TextureStorage` for gles3 2024-02-11 17:23:00 +01:00
A Thousand Ships 684752e75b
Replace error checks against `size` with `is_empty` 2024-02-09 12:50:15 +01:00
Yuri Sizov 2ce450f8bc Merge pull request #84894 from clayjohn/tangent-error-message
Store ArrayMesh path in RenderingServer for use in error messages
2024-01-24 14:07:47 +01:00
Yuri Sizov e8105adb0e Merge pull request #87155 from griffinkh/build_warning_fix
Fix build warning with memset value being too large
2024-01-22 20:38:56 +01:00
Yuri Sizov e95456b7f3 Merge pull request #86073 from m4rr5/add_reflection_mask
Add a reflection mask to the reflection probes
2024-01-22 20:38:37 +01:00
Marcel Offermans 25f573ca63 Add a reflection mask to the reflection probes.
Co-authored-by: Bastiaan Olij <mux213@gmail.com>
2024-01-20 08:07:31 +01:00
Rémi Verschelde 0bcc0e92b3
Merge pull request #86730 from reduz/64-bit-cowdata
Promote CowData to 64 bits
2024-01-19 18:46:11 +01:00
Juan Linietsky 24fe0715b5 Promote CowData to 64 bits
Fixes a lot of bugs, please help me fill the list.
2024-01-19 17:41:10 +01:00
Griffin Kroah-Hartman 21ae69a1de fix build warning with memset value being too large
same resolution for issue #83342 for texture_storage.cpp
2024-01-13 19:05:38 +01:00
clayjohn 7b355a128c Default to 32-bit depth map on Forward+ renderer and 24-bit only on Mobile renderer 2024-01-10 12:23:59 -08:00
Dario cc4d39b0c1 Acyclic Command Graph for RenderingDevice.
Adds a new system to automatically reorder commands, perform layout transitions and insert synchronization barriers based on the commands issued to RenderingDevice.
2024-01-08 14:54:56 -03:00
Rémi Verschelde 798632422d
Merge pull request #86245 from ecmjohnson/fix-skeleton-culling
Check skeleton RID when using cached AABB
2024-01-08 11:52:41 +01:00
Muller-Castro 96a95cb974 Add const lvalue ref to container parameters 2024-01-05 14:49:57 -03:00
EddieBreeg 8747c67d9e
Fix potential integer underflow in rounded up divisions
A new `Math::division_round_up()` function was added, allowing for easy
and correct computation of integer divisions when the result needs to
be rounded up.

Fixes #80358.

Co-authored-by: Rémi Verschelde <rverschelde@gmail.com>
2024-01-02 14:14:47 +01:00
Pedro J. Estébanez 12a519bae2 Split RenderingDevice into API-agnostic and RenderingDeviceDriver parts
Credit and thanks to @bruzvg for multiple build fixes, update of 3rd-party items and MinGW support.

Co-authored-by: bruvzg <7645683+bruvzg@users.noreply.github.com>
2023-12-20 19:18:08 +01:00