Compile DXR shaders using the same profile when targeting PCD3D_SM5 and PCD3D_SM6 (i.e. lib_6_6 or lib_6_5 based on USE_SHADER_MODEL_6_6 define).
#rb christopher.waters
#preflight 63376cc6466fb43669315f62
[CL 22286556 by yuriy odonnell in ue5-main branch]
When changing SM5 raytracing shaders to use lib_6_5 from lib_6_3, the __SHADER_TARGET_MINOR value wasn't updated to 5.
Bumping SM6 and SM5 shader versions.
#jira none
#rb yuriy.odonnell
#preflight 6335b2abf76de2d4d559891a
[CL 22263071 by christopher waters in ue5-main branch]
When D3DCOMPILE_SKIP_OPTIMIZATION is set, none of the other D3DCOMPILE_OPTIMIZATION_LEVEL flags are set. Given that D3DCOMPILE_OPTIMIZATION_LEVEL1 is "0", this always forced debug shader builds to be compiled with "/Od /O1".
The fix is to ignore the other flags when D3DCOMPILE_SKIP_OPTIMIZATION is set.
#jira none
#rb massimo.tristano, mihnea.balta
#preflight 630e1989e54ec9d581800a92
[CL 21701349 by christopher waters in ue5-main branch]
Since the cbuffer uses the same name as the struct, we have to rename the cbuffer using a consistent prefix. The reflection handling also needs to reinterpret the renamed cbuffer as the correct UniformBuffer.
#jira none
#rb mihnea.balta
#preflight 626c1762ce3959ce8fe7766e
[CL 20026049 by christopher waters in ue5-main branch]
We can't share DDPI infrastructure with PC D3D ES3.1 because of conflicts from sharing. We need a separate shader platform.
For now, we'll just 'intrusively' plumb the new Hololens shader platform. In the future, we want to move as much code out into
Hololens-specific platform code to reduce conflict with shared paths that don't need to know about Hololens.
#rb steve.smith, christopher.waters
#jira UE-133375
#preflight 61f9ca302251c6fb596e583e
#ROBOMERGE-AUTHOR: robert.srinivasiah
#ROBOMERGE-SOURCE: CL 18818524 in //UE5/Release-5.0/... via CL 18818546 via CL 18822881
#ROBOMERGE-BOT: UE5 (Release-Engine-Test -> Main) (v910-18824042)
[CL 18824735 by robert srinivasiah in ue5-main branch]
- SM6.6 not enabled just yet.
- Adding our own defines of __SHADER_TARGET_MAJOR and __SHADER_TARGET_MINOR to mimic the DXC defines. Since we preprocess, we can't rely on DXC to define them for us.
- Adding support for 64bit atomic operations
- Changing part of Nanite PC atomic permutation selection to be based on DataDrivenPlatformInfo
- Fixing DXC shader debugging. The PDB section was still being stripped from the final DXIL.
#jira none
#rb rune.stubbe
#preflight 61548928f5e4fd0001d7d0b3
[CL 17665881 by christopher waters in ue5-main branch]
- Removed some unnecessary code from D3D shader compilers.
- Removing shader VendorExtensions runtime data from platforms that don't need them.
- Turning FShaderCodeFeatures into a set of flags to reduce its size. Adding a few new flags for future use.
- Moved the common D3D12 shader data and creation code into shared locations to cut down on code duplication.
- Removed the unnecessary copying of header data from all D3D12 shader types.
- Removed FD3D12ShaderBytecode since its values can be inferred from other shader data.
#jira none
#rb rolando.caloca, arciel.rekman
#preflight 6140c2df3aa82d000149f13b
[CL 17509617 by christopher waters in ue5-main branch]
- Splitting symbol generation and shader "debug info" generation into two different cvars. This should now allow us to generate symbols post-packaging without generating different shader variants. This should also enable us to always generate symbols and still have the final deduplicated shader sizes.
- r.Shaders.Symbols is now the primary way to enable shader symbols for debugging.
- r.Shaders.GenerateSymbols can be enabled to always generate symbols without writing them to disk.
- r.Shaders.WriteSymbols can be used to force write symbols if they were previously generated.
- r.Shaders.ExtraData can be used to generate shader names for engine runtime systems to use.
- NotifyShaderCooked replaced with NotifyShaderCompiled, prep for upcoming non-cook symbol writing.
- PC DXC shader compiles should only generate debug info when requested, not all the time.
- PC FXC shader compiles should generate debug info when requested, not just when r.Shaders.Optimize is 0.
#jira none
#rb arciel.rekman, lukas.hermanns
#preflight 611ad9035e737200015992db
[CL 17186155 by christopher waters in ue5-main branch]
Disabled by default (guarded by USE_SHADER_MODEL_6_6).
Includes work done by Yuriy.
#rb yuriy.odonnell
#fyi brian.karis, christopher.waters, lukas.hermanns
[CL 15914077 by graham wihlidal in ue5-main branch]
#rnx
#rb none
#ROBOMERGE-SOURCE: CL 10869240 via CL 10869516 via CL 10869902
#ROBOMERGE-BOT: (v613-10869866)
[CL 10870584 by ryan durand in Main branch]
- This fixes various issues around depth divergence that can be see in the Niagara Mesh Vertex Factory as well as doing certain operations inside world position offset
#rb Uriel.Doyon,Rolando.Caloca
#jira UE-81271
[CL 9660243 by Stu McKenna in 4.24 branch]
