[llvm] [NVPTX] Remove `sm_1x` / non-ABI compilation support (PR #125977)
Justin Fargnoli via llvm-commits
llvm-commits at lists.llvm.org
Wed Feb 5 17:04:35 PST 2025
https://github.com/justinfargnoli created https://github.com/llvm/llvm-project/pull/125977
@Artem-B, I believe you said we don't intend to support `sm_1x`. Assuming that's correct, this PR will remove all remaining support I could find for `sm_1x`.
>From 882c6caacdfc90b09a08c9ea835ff3dd2761d0f9 Mon Sep 17 00:00:00 2001
From: Justin Fargnoli <jfargnoli at nvidia.com>
Date: Wed, 5 Feb 2025 13:25:45 -0800
Subject: [PATCH 1/3] Save work
---
llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp | 79 ++++++---------------
llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp | 20 ------
2 files changed, 23 insertions(+), 76 deletions(-)
diff --git a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
index cb756246b8d116c..79aeefed68a3c7f 100644
--- a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
+++ b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
@@ -343,61 +343,32 @@ void NVPTXAsmPrinter::printReturnValStr(const Function *F, raw_ostream &O) {
const auto *TLI = cast<NVPTXTargetLowering>(STI.getTargetLowering());
Type *Ty = F->getReturnType();
-
- bool isABI = (STI.getSmVersion() >= 20);
-
if (Ty->getTypeID() == Type::VoidTyID)
return;
O << " (";
- if (isABI) {
- if ((Ty->isFloatingPointTy() || Ty->isIntegerTy()) &&
- !ShouldPassAsArray(Ty)) {
- unsigned size = 0;
- if (auto *ITy = dyn_cast<IntegerType>(Ty)) {
- size = ITy->getBitWidth();
- } else {
- assert(Ty->isFloatingPointTy() && "Floating point type expected here");
- size = Ty->getPrimitiveSizeInBits();
- }
- size = promoteScalarArgumentSize(size);
- O << ".param .b" << size << " func_retval0";
- } else if (isa<PointerType>(Ty)) {
- O << ".param .b" << TLI->getPointerTy(DL).getSizeInBits()
- << " func_retval0";
- } else if (ShouldPassAsArray(Ty)) {
- unsigned totalsz = DL.getTypeAllocSize(Ty);
- Align RetAlignment = TLI->getFunctionArgumentAlignment(
- F, Ty, AttributeList::ReturnIndex, DL);
- O << ".param .align " << RetAlignment.value() << " .b8 func_retval0["
- << totalsz << "]";
- } else
- llvm_unreachable("Unknown return type");
- } else {
- SmallVector<EVT, 16> vtparts;
- ComputeValueVTs(*TLI, DL, Ty, vtparts);
- unsigned idx = 0;
- for (unsigned i = 0, e = vtparts.size(); i != e; ++i) {
- unsigned elems = 1;
- EVT elemtype = vtparts[i];
- if (vtparts[i].isVector()) {
- elems = vtparts[i].getVectorNumElements();
- elemtype = vtparts[i].getVectorElementType();
- }
-
- for (unsigned j = 0, je = elems; j != je; ++j) {
- unsigned sz = elemtype.getSizeInBits();
- if (elemtype.isInteger())
- sz = promoteScalarArgumentSize(sz);
- O << ".reg .b" << sz << " func_retval" << idx;
- if (j < je - 1)
- O << ", ";
- ++idx;
- }
- if (i < e - 1)
- O << ", ";
+ if ((Ty->isFloatingPointTy() || Ty->isIntegerTy()) &&
+ !ShouldPassAsArray(Ty)) {
+ unsigned size = 0;
+ if (auto *ITy = dyn_cast<IntegerType>(Ty)) {
+ size = ITy->getBitWidth();
+ } else {
+ assert(Ty->isFloatingPointTy() && "Floating point type expected here");
+ size = Ty->getPrimitiveSizeInBits();
}
- }
+ size = promoteScalarArgumentSize(size);
+ O << ".param .b" << size << " func_retval0";
+ } else if (isa<PointerType>(Ty)) {
+ O << ".param .b" << TLI->getPointerTy(DL).getSizeInBits()
+ << " func_retval0";
+ } else if (ShouldPassAsArray(Ty)) {
+ unsigned totalsz = DL.getTypeAllocSize(Ty);
+ Align RetAlignment = TLI->getFunctionArgumentAlignment(
+ F, Ty, AttributeList::ReturnIndex, DL);
+ O << ".param .align " << RetAlignment.value() << " .b8 func_retval0["
+ << totalsz << "]";
+ } else
+ llvm_unreachable("Unknown return type");
O << ") ";
}
@@ -1513,7 +1484,6 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
unsigned paramIndex = 0;
bool first = true;
bool isKernelFunc = isKernelFunction(*F);
- bool isABI = (STI.getSmVersion() >= 20);
if (F->arg_empty() && !F->isVarArg()) {
O << "()";
@@ -1646,10 +1616,7 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
sz = PTySizeInBits;
} else
sz = Ty->getPrimitiveSizeInBits();
- if (isABI)
- O << "\t.param .b" << sz << " ";
- else
- O << "\t.reg .b" << sz << " ";
+ O << "\t.param .b" << sz << " ";
O << TLI->getParamName(F, paramIndex);
continue;
}
@@ -1658,7 +1625,7 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
Type *ETy = PAL.getParamByValType(paramIndex);
assert(ETy && "Param should have byval type");
- if (isABI || isKernelFunc) {
+ if (isKernelFunc) {
// Just print .param .align <a> .b8 .param[size];
// <a> = optimal alignment for the element type; always multiple of
// PAL.getParamAlignment
diff --git a/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp b/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
index 9e7e1dbcea25d11..58ad92a8934a66d 100644
--- a/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
+++ b/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
@@ -1156,11 +1156,6 @@ std::string NVPTXTargetLowering::getPrototype(
const CallBase &CB, unsigned UniqueCallSite) const {
auto PtrVT = getPointerTy(DL);
- bool isABI = (STI.getSmVersion() >= 20);
- assert(isABI && "Non-ABI compilation is not supported");
- if (!isABI)
- return "";
-
std::string Prototype;
raw_string_ostream O(Prototype);
O << "prototype_" << UniqueCallSite << " : .callprototype ";
@@ -1429,11 +1424,6 @@ SDValue NVPTXTargetLowering::LowerCall(TargetLowering::CallLoweringInfo &CLI,
const CallBase *CB = CLI.CB;
const DataLayout &DL = DAG.getDataLayout();
- bool isABI = (STI.getSmVersion() >= 20);
- assert(isABI && "Non-ABI compilation is not supported");
- if (!isABI)
- return Chain;
-
// Variadic arguments.
//
// Normally, for each argument, we declare a param scalar or a param
@@ -3091,11 +3081,6 @@ SDValue NVPTXTargetLowering::LowerFormalArguments(
SDValue Root = DAG.getRoot();
std::vector<SDValue> OutChains;
- bool isABI = (STI.getSmVersion() >= 20);
- assert(isABI && "Non-ABI compilation is not supported");
- if (!isABI)
- return Chain;
-
std::vector<Type *> argTypes;
std::vector<const Argument *> theArgs;
for (const Argument &I : F->args()) {
@@ -3310,11 +3295,6 @@ NVPTXTargetLowering::LowerReturn(SDValue Chain, CallingConv::ID CallConv,
const Function &F = MF.getFunction();
Type *RetTy = MF.getFunction().getReturnType();
- bool isABI = (STI.getSmVersion() >= 20);
- assert(isABI && "Non-ABI compilation is not supported");
- if (!isABI)
- return Chain;
-
const DataLayout &DL = DAG.getDataLayout();
SmallVector<SDValue, 16> PromotedOutVals;
SmallVector<EVT, 16> VTs;
>From 831f82d6a22b2a8a6c6555fbbb68bf788f1d091d Mon Sep 17 00:00:00 2001
From: Justin Fargnoli <jfargnoli at nvidia.com>
Date: Wed, 5 Feb 2025 15:56:47 -0800
Subject: [PATCH 2/3] Fix asm printer + try writing a test
---
llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp | 61 +++++---------------
llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll | 9 +++
2 files changed, 23 insertions(+), 47 deletions(-)
create mode 100644 llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
diff --git a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
index 79aeefed68a3c7f..ad1433821036be6 100644
--- a/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
+++ b/llvm/lib/Target/NVPTX/NVPTXAsmPrinter.cpp
@@ -1625,53 +1625,20 @@ void NVPTXAsmPrinter::emitFunctionParamList(const Function *F, raw_ostream &O) {
Type *ETy = PAL.getParamByValType(paramIndex);
assert(ETy && "Param should have byval type");
- if (isKernelFunc) {
- // Just print .param .align <a> .b8 .param[size];
- // <a> = optimal alignment for the element type; always multiple of
- // PAL.getParamAlignment
- // size = typeallocsize of element type
- Align OptimalAlign =
- isKernelFunc
- ? getOptimalAlignForParam(ETy)
- : TLI->getFunctionByValParamAlign(
- F, ETy, PAL.getParamAlignment(paramIndex).valueOrOne(), DL);
-
- unsigned sz = DL.getTypeAllocSize(ETy);
- O << "\t.param .align " << OptimalAlign.value() << " .b8 ";
- O << TLI->getParamName(F, paramIndex);
- O << "[" << sz << "]";
- continue;
- } else {
- // Split the ETy into constituent parts and
- // print .param .b<size> <name> for each part.
- // Further, if a part is vector, print the above for
- // each vector element.
- SmallVector<EVT, 16> vtparts;
- ComputeValueVTs(*TLI, DL, ETy, vtparts);
- for (unsigned i = 0, e = vtparts.size(); i != e; ++i) {
- unsigned elems = 1;
- EVT elemtype = vtparts[i];
- if (vtparts[i].isVector()) {
- elems = vtparts[i].getVectorNumElements();
- elemtype = vtparts[i].getVectorElementType();
- }
-
- for (unsigned j = 0, je = elems; j != je; ++j) {
- unsigned sz = elemtype.getSizeInBits();
- if (elemtype.isInteger())
- sz = promoteScalarArgumentSize(sz);
- O << "\t.reg .b" << sz << " ";
- O << TLI->getParamName(F, paramIndex);
- if (j < je - 1)
- O << ",\n";
- ++paramIndex;
- }
- if (i < e - 1)
- O << ",\n";
- }
- --paramIndex;
- continue;
- }
+ // Print .param .align <a> .b8 .param[size];
+ // <a> = optimal alignment for the element type; always multiple of
+ // PAL.getParamAlignment
+ // size = typeallocsize of element type
+ Align OptimalAlign =
+ isKernelFunc
+ ? getOptimalAlignForParam(ETy)
+ : TLI->getFunctionByValParamAlign(
+ F, ETy, PAL.getParamAlignment(paramIndex).valueOrOne(), DL);
+
+ unsigned sz = DL.getTypeAllocSize(ETy);
+ O << "\t.param .align " << OptimalAlign.value() << " .b8 ";
+ O << TLI->getParamName(F, paramIndex);
+ O << "[" << sz << "]";
}
if (F->isVarArg()) {
diff --git a/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll b/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
new file mode 100644
index 000000000000000..a81a7f0d6e16fdb
--- /dev/null
+++ b/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
@@ -0,0 +1,9 @@
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_10 -debug-only=nvptx-subtarget -o /dev/null 2>&1 | FileCheck %s --check-prefix=SM10
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_11 -debug-only=nvptx-subtarget -o /dev/null 2>&1 | FileCheck %s --check-prefix=SM11
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_12 -debug-only=nvptx-subtarget -o /dev/null 2>&1 | FileCheck %s --check-prefix=SM12
+; RUN: llc < %s -mtriple=nvptx -mcpu=sm_13 -debug-only=nvptx-subtarget -o /dev/null 2>&1 | FileCheck %s --check-prefix=SM13
+
+; SM10: 'sm_10' is not a recognized processor for this target (ignoring processor)
+; SM11: 'sm_11' is not a recognized processor for this target (ignoring processor)
+; SM12: 'sm_12' is not a recognized processor for this target (ignoring processor)
+; SM13: 'sm_13' is not a recognized processor for this target (ignoring processor)
\ No newline at end of file
>From 5808054a54cca9c0895ce8a2873b826f2076fed6 Mon Sep 17 00:00:00 2001
From: Justin Fargnoli <jfargnoli at nvidia.com>
Date: Wed, 5 Feb 2025 15:57:36 -0800
Subject: [PATCH 3/3] Newline
---
llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll b/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
index a81a7f0d6e16fdb..9a1dc122915dded 100644
--- a/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
+++ b/llvm/test/CodeGen/NVPTX/unrecognized-sm1x.ll
@@ -6,4 +6,4 @@
; SM10: 'sm_10' is not a recognized processor for this target (ignoring processor)
; SM11: 'sm_11' is not a recognized processor for this target (ignoring processor)
; SM12: 'sm_12' is not a recognized processor for this target (ignoring processor)
-; SM13: 'sm_13' is not a recognized processor for this target (ignoring processor)
\ No newline at end of file
+; SM13: 'sm_13' is not a recognized processor for this target (ignoring processor)
More information about the llvm-commits
mailing list