Skip to content

[DirectX] Add support for vector type checking in DXIL Shader Flag Analysis #150532

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Merged
merged 2 commits into from
Jul 25, 2025
Merged

[DirectX] Add support for vector type checking in DXIL Shader Flag Analysis #150532

merged 2 commits into from
Jul 25, 2025
+125 −7

Conversation

Icohedron
Copy link
Contributor

Fixes #150482 by replacing all Instruction.getType() in DXILShaderFlags.cpp with Instruction.getType()->getScalarType() to account for vectors types as suggested by @bogner

@llvmbot
Copy link
Member

llvmbot commented Jul 24, 2025

@llvm/pr-subscribers-backend-directx

Author: Deric C. (Icohedron)

Changes

Fixes #150482 by replacing all Instruction.getType() in DXILShaderFlags.cpp with Instruction.getType()->getScalarType() to account for vectors types as suggested by @bogner

Full diff: https://github.com/llvm/llvm-project/pull/150532.diff

2 Files Affected:

  • (modified) llvm/lib/Target/DirectX/DXILShaderFlags.cpp (+8-7)
  • (added) llvm/test/CodeGen/DirectX/ShaderFlags/rawbuffer-low-precision.ll (+44)
diff --git a/llvm/lib/Target/DirectX/DXILShaderFlags.cpp b/llvm/lib/Target/DirectX/DXILShaderFlags.cpp
index eb4adfea5aed6..e7e7f2ce66ae8 100644
--- a/llvm/lib/Target/DirectX/DXILShaderFlags.cpp
+++ b/llvm/lib/Target/DirectX/DXILShaderFlags.cpp
@@ -106,11 +106,11 @@ void ModuleShaderFlags::updateFunctionFlags(ComputedShaderFlags &CSF,
                                             DXILResourceTypeMap &DRTM,
                                             const ModuleMetadataInfo &MMDI) {
   if (!CSF.Doubles)
-    CSF.Doubles = I.getType()->isDoubleTy();
+    CSF.Doubles = I.getType()->getScalarType()->isDoubleTy();
 
   if (!CSF.Doubles) {
     for (const Value *Op : I.operands()) {
-      if (Op->getType()->isDoubleTy()) {
+      if (Op->getType()->getScalarType()->isDoubleTy()) {
         CSF.Doubles = true;
         break;
       }
@@ -130,12 +130,13 @@ void ModuleShaderFlags::updateFunctionFlags(ComputedShaderFlags &CSF,
   }
 
   if (!CSF.LowPrecisionPresent)
-    CSF.LowPrecisionPresent =
-        I.getType()->isIntegerTy(16) || I.getType()->isHalfTy();
+    CSF.LowPrecisionPresent = I.getType()->getScalarType()->isIntegerTy(16) ||
+                              I.getType()->getScalarType()->isHalfTy();
 
   if (!CSF.LowPrecisionPresent) {
     for (const Value *Op : I.operands()) {
-      if (Op->getType()->isIntegerTy(16) || Op->getType()->isHalfTy()) {
+      if (Op->getType()->getScalarType()->isIntegerTy(16) ||
+          Op->getType()->getScalarType()->isHalfTy()) {
         CSF.LowPrecisionPresent = true;
         break;
       }
@@ -150,11 +151,11 @@ void ModuleShaderFlags::updateFunctionFlags(ComputedShaderFlags &CSF,
   }
 
   if (!CSF.Int64Ops)
-    CSF.Int64Ops = I.getType()->isIntegerTy(64);
+    CSF.Int64Ops = I.getType()->getScalarType()->isIntegerTy(64);
 
   if (!CSF.Int64Ops && !isa<LifetimeIntrinsic>(&I)) {
     for (const Value *Op : I.operands()) {
-      if (Op->getType()->isIntegerTy(64)) {
+      if (Op->getType()->getScalarType()->isIntegerTy(64)) {
         CSF.Int64Ops = true;
         break;
       }
diff --git a/llvm/test/CodeGen/DirectX/ShaderFlags/rawbuffer-low-precision.ll b/llvm/test/CodeGen/DirectX/ShaderFlags/rawbuffer-low-precision.ll
new file mode 100644
index 0000000000000..ec4b87fd81799
--- /dev/null
+++ b/llvm/test/CodeGen/DirectX/ShaderFlags/rawbuffer-low-precision.ll
@@ -0,0 +1,44 @@
+; RUN: opt -S --passes="print-dx-shader-flags" 2>&1 %s | FileCheck %s
+; RUN: llc %s --filetype=obj -o - | obj2yaml | FileCheck %s --check-prefix=DXC
+
+target triple = "dxil-pc-shadermodel6.7-library"
+
+; CHECK: ; Combined Shader Flags for Module
+; CHECK-NEXT: ; Shader Flags Value: 0x00800030
+; CHECK-NEXT: ;
+; CHECK-NEXT: ; Note: shader requires additional functionality:
+; CHECK-NEXT: ;       Native low-precision data types
+; CHECK-NEXT: ; Note: extra DXIL module flags:
+; CHECK-NEXT: ;       Raw and structured buffers
+; CHECK-NEXT: ;       Low-precision data types present
+; CHECK-NEXT: ;       Enable native low-precision data types
+; CHECK-NEXT: ;
+; CHECK-NEXT: ; Shader Flags for Module Functions
+
+; CHECK: Function rawbuf : 0x00800030
+define void @rawbuf() "hlsl.export" {
+  %halfrb = tail call target("dx.RawBuffer", <4 x half>, 0, 0)
+    @llvm.dx.resource.handlefrombinding.tdx.RawBuffer_v4f16_0_0t(i32 0, i32 0, i32 1, i32 0, i1 false, ptr null)
+  %i16rb = tail call target("dx.RawBuffer", <4 x i16>, 1, 0)
+    @llvm.dx.resource.handlefrombinding.tdx.RawBuffer_v4i16_1_0t(i32 0, i32 1, i32 1, i32 0, i1 false, ptr null)
+  %loadhalfrb = call { <4 x i16>, i1 }
+    @llvm.dx.resource.load.rawbuffer.v4i16.tdx.RawBuffer_v4f16_0_0t(target("dx.RawBuffer", <4 x half>, 0, 0) %halfrb, i32 0, i32 0)
+  %storei16rb = extractvalue { <4 x i16>, i1 } %loadhalfrb, 0
+  call void @llvm.dx.resource.store.rawbuffer.tdx.RawBuffer_v4i16_1_0t.v4i16(target("dx.RawBuffer", <4 x i16>, 1, 0) %i16rb, i32 0, i32 0, <4 x i16> %storei16rb)
+  ret void
+}
+
+; Metadata to avoid adding flags not currently of interest to this test, and
+; enable native low precision data types
+!dx.valver = !{!0}
+!0 = !{i32 1, i32 8}
+!llvm.module.flags = !{!1, !2}
+!1 = !{i32 1, !"dx.nativelowprec", i32 1}
+!2 = !{i32 1, !"dx.resmayalias", i32 1}
+
+; DXC: - Name:            SFI0
+; DXC-NEXT:     Size:            8
+; DXC-NEXT:     Flags:
+; DXC:      MinimumPrecision: false
+; DXC:      NativeLowPrecision: true
+; DXC: ...

bob80905
bob80905 approved these changes Jul 24, 2025
View reviewed changes
Copy link
Contributor

@bob80905 bob80905 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

spall
spall approved these changes Jul 24, 2025
View reviewed changes
Copy link
Contributor

@spall spall left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm

@Icohedron Icohedron merged commit 4ae2678 into llvm:main Jul 25, 2025
10 checks passed
@llvm-ci
Copy link
Collaborator

llvm-ci commented Jul 25, 2025

LLVM Buildbot has detected a new failure on builder llvm-clang-x86_64-expensive-checks-debian running on gribozavr4 while building llvm at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/16/builds/23219

Here is the relevant piece of the build log for the reference 
Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LLVM :: TableGen/RuntimeLibcallEmitter.td' FAILED ********************
Exit Code: 1

Command Output (stderr):
--
/b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llvm-tblgen -gen-runtime-libcalls -I /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/../../include /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td | /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td # RUN: at line 1
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/FileCheck /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td
+ /b/1/llvm-clang-x86_64-expensive-checks-debian/build/bin/llvm-tblgen -gen-runtime-libcalls -I /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/../../include /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td
/b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td:98:16: error: CHECK-NEXT: expected string not found in input
// CHECK-NEXT: sqrtl_f80 = 7, // sqrtl
               ^
<stdin>:32:23: note: scanning from here
 calloc = 6, // calloc
                      ^
<stdin>:34:2: note: possible intended match here
 sqrtl_f80 = 8, // sqrtl
 ^

Input file: <stdin>
Check file: /b/1/llvm-clang-x86_64-expensive-checks-debian/llvm-project/llvm/test/TableGen/RuntimeLibcallEmitter.td

-dump-input=help explains the following input dump.

Input was:
<<<<<<
           .
           .
           .
          27:  ___memcpy = 1, // ___memcpy 
          28:  ___memset = 2, // ___memset 
          29:  __ashlsi3 = 3, // __ashlsi3 
          30:  __lshrdi3 = 4, // __lshrdi3 
          31:  bzero = 5, // bzero 
          32:  calloc = 6, // calloc 
next:98'0                           X error: no match found
          33:  sqrtl_f128 = 7, // sqrtl 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~
          34:  sqrtl_f80 = 8, // sqrtl 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
next:98'1      ?                        possible intended match
          35:  NumLibcallImpls = 9 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~
          36: }; 
next:98'0     ~~~
          37: } // End namespace RTLIB 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
          38: } // End namespace llvm 
next:98'0     ~~~~~~~~~~~~~~~~~~~~~~~~
          39: #endif 
next:98'0     ~~~~~~~
...

mahesh-attarde pushed a commit to mahesh-attarde/llvm-project that referenced this pull request Jul 28, 2025
@Icohedron @mahesh-attarde
[DirectX] Add support for vector type checking in DXIL Shader Flag An…
f4604dc 
…alysis (llvm#150532)

Fixes llvm#150482 by replacing all `Instruction.getType()` in
DXILShaderFlags.cpp with `Instruction.getType()->getScalarType()` to
account for vectors types as suggested by @bogner
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@spall spall spall approved these changes

@bob80905 bob80905 bob80905 approved these changes

Assignees
No one assigned
Labels
backend:DirectX
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

[DirectX] Shader flag analysis does not take into account vector types
5 participants
@Icohedron @llvmbot @llvm-ci @spall @bob80905