Skip to content

Limit the instructions that enable BFloat16ArithmeticINTEL capability #3561

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

Merged
vmaksimo merged 2 commits into from
Feb 6, 2026
Merged

Limit the instructions that enable BFloat16ArithmeticINTEL capability #3561

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
93aa217
Limit the instructions that enable `BFloat16ArithmeticINTEL` capability
vmaksimo Feb 2, 2026
dc8b84b
Merge remote-tracking branch 'origin/main' into HEAD
vmaksimo Feb 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion lib/SPIRV/libSPIRV/SPIRVModule.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1756,7 +1756,8 @@ SPIRVInstruction *
SPIRVModuleImpl::addUnaryInst(Op TheOpCode, SPIRVType *TheType, SPIRVValue *Op,
SPIRVBasicBlock *BB,
SPIRVInstruction *InsertBefore) {
if (TheType->isTypeFloat(16, FPEncodingBFloat16KHR) && TheOpCode != OpDot)
if (TheType->isTypeFloat(16, FPEncodingBFloat16KHR) &&
(TheOpCode == OpFNegate || isUnaryPredicateOpCode(TheOpCode)))
addCapability(internal::CapabilityBFloat16ArithmeticINTEL);
return addInstruction(
SPIRVInstTemplateBase::create(TheOpCode, TheType, getId(),
Expand Down
56 changes: 56 additions & 0 deletions test/extensions/KHR/SPV_KHR_bfloat16/conversions.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,56 @@
; RUN: llvm-as %s -o %t.bc
; RUN: llvm-spirv %t.bc --spirv-ext=+SPV_KHR_bfloat16 -o %t.spv
; RUN: llvm-spirv %t.spv -to-text -o - | FileCheck %s --check-prefix=CHECK-SPIRV

; RUN: llvm-spirv -r %t.spv -o %t.rev.bc
; RUN: llvm-dis %t.rev.bc -o - | FileCheck %s --check-prefix=CHECK-LLVM

; RUN: not llvm-spirv %t.bc 2>&1 >/dev/null | FileCheck %s --check-prefix=CHECK-ERROR
; RUN: not llvm-spirv %t.bc --spirv-ext=+SPV_INTEL_bfloat16_arithmetic 2>&1 >/dev/null | FileCheck %s --check-prefix=CHECK-ERROR
; CHECK-ERROR: RequiresExtension: Feature requires the following SPIR-V extension:

target datalayout = "e-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-n8:16:32:64"
target triple = "spirv64-unknown-unknown"

; CHECK-SPIRV: Capability BFloat16TypeKHR
; CHECK-SPIRV-NOT: Capability BFloat16ArithmeticINTEL
; CHECK-SPIRV-NOT: Extension "SPV_INTEL_bfloat16_arithmetic"
; CHECK-SPIRV: Extension "SPV_KHR_bfloat16"
; CHECK-SPIRV: TypeFloat [[BFLOAT:[0-9]+]] 16 0
; CHECK-SPIRV: Variable [[#]] [[ADDR1:[0-9]+]]
; CHECK-SPIRV: Load [[BFLOAT]] [[DATA1:[0-9]+]] [[ADDR1]]
; CHECK-SPIRV: ConvertFToU [[#]] [[#]] [[DATA1]]
; CHECK-SPIRV: ConvertFToS [[#]] [[#]] [[DATA1]]
; CHECK-SPIRV: ConvertSToF [[BFLOAT]] [[#]] [[#]]
; CHECK-SPIRV: ConvertUToF [[BFLOAT]] [[#]] [[#]]


; CHECK-LLVM: [[ADDR1:[%a-z0-9]+]] = alloca bfloat
; CHECK-LLVM: [[DATA1:[%a-z0-9]+]] = load bfloat, ptr [[ADDR1]]
; CHECK-LLVM: %OpConvertFToU = fptoui bfloat [[DATA1]] to i32
; CHECK-LLVM: %OpConvertFToS = fptosi bfloat [[DATA1]] to i32
; CHECK-LLVM: %OpConvertSToF = sitofp i32 0 to bfloat
; CHECK-LLVM: %OpConvertUToF = uitofp i32 0 to bfloat

define spir_kernel void @testConversions() {
entry:
%addr1 = alloca bfloat
%data1 = load bfloat, ptr %addr1
%OpConvertFToU = fptoui bfloat %data1 to i32
%OpConvertFToS = fptosi bfloat %data1 to i32
%OpConvertSToF = sitofp i32 0 to bfloat
%OpConvertUToF = uitofp i32 0 to bfloat
ret void
}

!opencl.enable.FP_CONTRACT = !{}
!opencl.spir.version = !{!0}
!opencl.ocl.version = !{!1}
!opencl.used.extensions = !{!2}
!opencl.used.optional.core.features = !{!3}
!opencl.compiler.options = !{!3}

!0 = !{i32 1, i32 2}
!1 = !{i32 2, i32 0}
!2 = !{!"cl_khr_fp16"}
!3 = !{}