Internal/master #8053
Merged
Internal/master #8053
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add editor analytics for Render Graph Viewer. JIRA: https://jira.unity3d.com/browse/XPIPELINE-837 POI: https://jira.unity3d.com/browse/POI-817 Event data collection sheet: https://docs.google.com/spreadsheets/d/16jZa8H0o2LKQi-AiNHe8YgRmM4viAKLj40Csjpoljic/edit#gid=515574887 STAR Checklist: https://star-checklist.ds.unity3d.com/Checklists/995/overview schema: `com.unity3d.data.schemas.editor.analytics.uRenderGraphViewerLifetimeAnalytic_v1` taxonomy: `editor.analytics.uRenderGraphViewerLifetimeAnalytic.v1`
Added documentation on how to use the Screen Space Overlay render mode alongside the Graphics Compositor. We decided not to change the actual behavior of SupportedRenderingFeatures.active.rendersUIOverlay when the Graphics Compositor is in use, but instead, to document a way force the rendering of the UI in this context. When the user follows the new documentation steps, the UI is indeed well renderer in the GameView. Tested manually and locally on my Windows machine.
Update buginfo and codeowners for Environment Effect bug area
This PR fixes HDRP DXR Standalone tests (HDRP_DXR_Win_DX12_Standalone_mono_Linear) that have been red in trunk since early January because of missing shader variants in the shader variant collection. https://unity-ci.cds.internal.unity3d.com/project/3/branch/trunk/jobDefinition/.yamato%2Fsrp%2Fhdrp-dxr.yml%23hdrp_dxr_win_dx12_standalone_mono_linear?nav=jobDefinitions
Update codeowners and .buginfo files for the Postprocessing and UI Features area.
https://jira.unity3d.com/browse/UUM-58350 An internal operation (addComponent) was registering an editor Undo, marking the scene as dirty every time we entered the playmode.
Fix realloc of history buffer when XR is enabled
Support rendering water with custom pass, request from the forum
Fixes [UUM-62756](https://jira.unity3d.com/browse/UUM-62756). The public documentation states the shader compiler will ignore Unity-specific `#pragma` directives in files included through `#include`, however this behavior was broken as they were being reported in both `#include` and `#include_with_pragmas`. This change fixes that so they're not reported in `#include` files. Also, added a warning message when someone uses `#include_with_pragmas` within a file that was itself included with `#include`. In that particular case, `#include_with_pragmas` will be treated as `#include`, meaning all Unity-specific pragmas it might contain will be ignored.
Add public scripting API for URP TAA settings.
…ctions. Fixes https://jira.unity3d.com/browse/UUM-62216, in which artifacts appear (looks like uninitialized memory) when using the accurate G-buffer normals feature when targeting mobile. This feature uses a pair for function for performing octrahedral encoding and decoding of normals. The function misbehaves when targeting mobile, but not desktop. The main difference between mobile and desktop in shadercode is that desktop treats `half` as float, whereas on mobile `half` actually works. The functions for octahedral encoding take `float`'s as input, and are given `half`'s in the mobile path. In the function, we use `abs()` on the input and then negate it. This combination seems to cause a miscompilation of some sort, presumably because `abs` and negation are both input modifiers rather than real instructions - I'm not entirely sure, but it's some mix of those 3 things - half to float conversion, abs and negation. The most minimal fix that worked was this: ``` float3 UnpackNormalOctQuadEncode(float2 f) { - float3 n = float3(f.x, f.y, 1.0 - abs(f.x) - abs(f.y)); + float3 n = float3(f.x, f.y, 1.0 - abs(real(f.x)) - abs(real(f.y))); ``` But I instead decided to change the functions to use `real` instead of `float` throughout, as it it seemed cleaner. `real` is just an alias for `half` on mobile and `float` on desktop.
Fixes https://jira.unity3d.com/browse/UUM-54404, where screen space decals weren't taking into account light cookies. The code to sample cookies is already present in the relevant shader (a shader made with shadergraph decal template), but the keyword for light cookies, `_LIGHT_COOKIES` is not declared. The fix is simply to add the keyword to the relevant pass descriptor. The keyword will automatically be set earlier in the render loop, in `LightCookiesManager.Setup`.
…atrix operator - Fixes https://jira.unity3d.com/browse/UUM-57915 - Add a settings to select between row and column - Add symetrical operator SplitMatrix Another possible way to have done it would have been through expandable slots but it lacks the row/colum selection and would have been a bit hacky as not handled directly by our system. A proxy struct would have been needed to do it right but it seems more involved/risky. Here the solution chosen is consistent with how Shader Graph is doing it.
Fix for a distorted viewport when using no intermediate texture in XR rendering. When activating dynamic resolution and using a scale of < 1, the image becomes distorted. This fixes this issue.
We complete switching to the newest API in GraphicsSettings. GraphicsSettings.renderPipelineAsset API will be deprecated and it's replaced by defaultRenderPipeline. We should use GraphicsSettings.defaultRenderPipeline or GraphicsSettings.currentRenderPipeline. Incorrect API was used in some places. It supposed to rely on the current pipeline instead of default as we may not have a default but we could still have RP in the Quality Settings. 1. Scene Render mode window could have incorrect height 2. Avatar previews could be handled incorrectly 3. Light component have an issue in a multiple RP projects 4. ScreenSpaceAmbientOcclusion may not work when URP is deferred in a multiple RP projects 5. Enlighten Manager may not work if you don't set RP asset in the GraphicsSettings https://jira.unity3d.com/browse/UUM-66743
Add minimal coverage about not up to date template.
Improve missing shader message in custom post process, improve doc, and remove 's' in post process order
About UUM-62355, the Position Box has been migrated with this PR: https://github.cds.internal.unity3d.com/unity/unity/pull/40039 Since the normal/axis are aligned with the box, we can skip the scale if possible (aka the transform comes from a Translation/Rotation/Scale): https://media.github.cds.internal.unity3d.com/user/42/files/a979c270-ce82-4537-af32-d4ff6f0e5df3⚠️ If we are computing the box in other way, the scale won't be skipped and it still leads to invalid axis but it's an really rare case: https://media.github.cds.internal.unity3d.com/user/42/files/1aaaa3b8-d983-41fb-bb9c-5ce0222dfa1b About UUM-67233, the Position On SDF was incomplete: - The applyOrientation wasn't tested leading to compilation failure - The blending mode wasn't taken into account (Position Add/Mul/Blend, ...) - The axisX/Y/Z was never wrote, I made them consistent with the sphere orientation (see below) https://media.github.cds.internal.unity3d.com/user/42/files/43efeee4-0756-408f-9d46-5a3f0e6836f6
Fix unstable test: https://jira.unity3d.com/browse/UUM-59403 which was disabled earlier due to instability. The test was unstable because it sometimes took too long and timed out. The instability was actually introduced by this PR https://github.cds.internal.unity3d.com/unity/unity/pull/41268 that tracks timeouts correctly, meaning that the highly varying runtimes of this test (usually 100-300sec but up to 700 sec sometimes) were not being enforced before. After the UTF 1.4.2 release the timeouts started correctly failing the test, and the test became unstable as a consequence. This PR fixes the issue by refactoring it: it used to be a edit mode test that manually enters playmode (triggering domain reload) before running the test logic. It was originally done this way to more closely follow the repro steps where asset preview bugs would happen. But given the long runtime, it makes more sense to just implement it as a playmode test instead, so that we don't need to enter playmode separately which is time-consuming. As a result, the test is now passing reliably and only taking ~5sec. cc @alex-vazquez - There was a separate issue with URP_Foundation project where the playmode test assembly was being excluded from editor, meaning that we had tests that were not being run when they should have been. These tests have now been fixed and enabled too.
…ompiler JIRA: https://jira.unity3d.com/browse/UUM-65454 When merging subpasses together in a NativePass, we don't check if there will be more than 8 native render subpasses. This PR adds a check for this when verifying if a certain graph pass can be merged into a Native Pass, as well as a test. If merging the graph pass would result in too many native subpasses, we simply create a new Native Pass
Correctly handle HLSL includes in vfx custom operator. It was simply discarded before this change. 🎁 Bonus: - Remove unexpected warnings about `in` missing in operators  - Add errors when include refers to missing files https://media.github.cds.internal.unity3d.com/user/42/files/76139d70-9a35-472d-95d2-fa3176dd6da8 - Factorize error feedback for CustomHLSL at https://github.cds.internal.unity3d.com/unity/unity/pull/46642/commits/5a25691885a828b0af6d42313ed042b8362b52bb
… textures Fix https://jira.unity3d.com/browse/UUM-67987 Render Graph Viewer displays a "store reasoning message" for Raster Pass Attachments to help understand why the system selected a given store action (one of `Store`, `Resolve` , `StoreAndResolve` or `DontCare`). Because `StoreAndResolve` exists, it means that we can potentially keep both the resolved and non-resolved data from a MSAA texture, and the reasoning for both of is independently decided. Therefore for MSAA attachments, we need to display two separate store reason messages in Render Graph Viewer. This PR fixes a bug where the UI only picked the first, default store reason message. For MSAA textures, UI looks like this: - See separate store reason messages for "Resolved surface" and "MSAA surface" - For non-MSAA attachments, the MSAA-specific information is not displayed.  This PR also contains a separate UI fix for Render Graph Viewer that makes the highlight animation update smoothly. It's a simple workaround to [this issue](https://jira.unity3d.com/browse/UUM-68010) that has been flagged for UITK team.
Fix missing eye index in compute shader
…UUM-36359) Fix for [UUM-36359](https://jira.unity3d.com/browse/UUM-36359). Originally `RTHandle` manages external texture but doesn't have the same check for `RenderTexture` that generated as external resources.
Move water specific global shader variables to separate file per user request. https://forum.unity.com/threads/water-system-for-the-high-definition-render-pipeline.1203751/page-15#post-9716743
Jira: https://jira.unity3d.com/browse/UUM-57644 Fixed XROcclusionMesh.shader and XRMirrorView.shader are not stripped down when XR Plugin is not installed
The position SS wasn't passed to the APV evaluation function resulting in wrong jittering for TAA
Fixed an issue in recorder with output textures caused by a previous PR: https://github.cds.internal.unity3d.com/unity/unity/pull/46637
The lighting code for cinematic eye shader was using the directional light position instead of it's direction
update graphics compositor limitation for VR
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Please read the Contributing guide before making a PR.
Checklist for PR maker
need-backport-*label. After you backport the PR, the label changes tobackported-*.CHANGELOG.mdfile.Purpose of this PR
Why is this PR needed, what hard problem is it solving/fixing?
Testing status
Describe what manual/automated tests were performed for this PR
Comments to reviewers
Notes for the reviewers you have assigned.