MaskRay updated this revision to Diff 361467.
MaskRay marked 2 inline comments as done.
MaskRay added a comment.
comments. add a test/CodeGen test. add HelpText.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D106701/new/
https://reviews.llvm.org/D106701
Files:
clang/include/clang/Basic/CodeGenOptions.def
clang/include/clang/Basic/DiagnosticDriverKinds.td
clang/include/clang/Basic/LangOptions.def
clang/include/clang/Driver/Options.td
clang/lib/CodeGen/BackendUtil.cpp
clang/lib/Driver/ToolChains/Clang.cpp
clang/test/CodeGen/align-loops.c
clang/test/Driver/falign-loops.c
llvm/include/llvm/CodeGen/CommandFlags.h
llvm/include/llvm/CodeGen/TargetLowering.h
llvm/include/llvm/Target/TargetOptions.h
llvm/lib/CodeGen/CommandFlags.cpp
llvm/lib/CodeGen/TargetLoweringBase.cpp
llvm/test/CodeGen/RISCV/align-loops.ll
llvm/test/CodeGen/X86/innermost-loop-alignment.ll
Index: llvm/test/CodeGen/X86/innermost-loop-alignment.ll
===================================================================
--- llvm/test/CodeGen/X86/innermost-loop-alignment.ll
+++ llvm/test/CodeGen/X86/innermost-loop-alignment.ll
@@ -1,7 +1,7 @@
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=i686-pc-linux-gnu | FileCheck %s -check-prefix=DEFAULT
; RUN: llc < %s -mtriple=i686-pc-linux-gnu -x86-experimental-pref-innermost-loop-alignment=5 | FileCheck %s -check-prefix=ALIGN32
-; RUN: llc < %s -mtriple=i686-pc-linux-gnu -x86-experimental-pref-loop-alignment=5 -x86-experimental-pref-innermost-loop-alignment=6 | FileCheck %s -check-prefix=ALIGN64
+; RUN: llc < %s -mtriple=i686-pc-linux-gnu -align-loops=32 -x86-experimental-pref-innermost-loop-alignment=6 | FileCheck %s -check-prefix=ALIGN64
declare void @foo()
Index: llvm/test/CodeGen/RISCV/align-loops.ll
===================================================================
--- /dev/null
+++ llvm/test/CodeGen/RISCV/align-loops.ll
@@ -0,0 +1,44 @@
+; RUN: llc < %s -mtriple=riscv64 | FileCheck %s
+; RUN: llc < %s -mtriple=riscv64 -align-loops=16 | FileCheck %s -check-prefix=ALIGN_16
+; RUN: llc < %s -mtriple=riscv64 -align-loops=32 | FileCheck %s -check-prefix=ALIGN_32
+
+declare void @foo()
+
+define void @test(i32 %n, i32 %m) nounwind {
+; CHECK-LABEL: test:
+; CHECK-NOT: .p2align
+; CHECK: ret
+
+; ALIGN_16-LABEL: test:
+; ALIGN_16: .p2align 4{{$}}
+; ALIGN_16-NEXT: .LBB0_1: # %outer
+; ALIGN_16: .p2align 4{{$}}
+; ALIGN_16-NEXT: .LBB0_2: # %inner
+
+; ALIGN_32-LABEL: test:
+; ALIGN_32: .p2align 5{{$}}
+; ALIGN_32-NEXT: .LBB0_1: # %outer
+; ALIGN_32: .p2align 5{{$}}
+; ALIGN_32-NEXT: .LBB0_2: # %inner
+entry:
+ br label %outer
+
+outer:
+ %outer.iv = phi i32 [0, %entry], [%outer.iv.next, %outer_bb]
+ br label %inner
+
+inner:
+ %inner.iv = phi i32 [0, %outer], [%inner.iv.next, %inner]
+ call void @foo()
+ %inner.iv.next = add i32 %inner.iv, 1
+ %inner.cond = icmp ne i32 %inner.iv.next, %m
+ br i1 %inner.cond, label %inner, label %outer_bb
+
+outer_bb:
+ %outer.iv.next = add i32 %outer.iv, 1
+ %outer.cond = icmp ne i32 %outer.iv.next, %n
+ br i1 %outer.cond, label %outer, label %exit
+
+exit:
+ ret void
+}
Index: llvm/lib/CodeGen/TargetLoweringBase.cpp
===================================================================
--- llvm/lib/CodeGen/TargetLoweringBase.cpp
+++ llvm/lib/CodeGen/TargetLoweringBase.cpp
@@ -52,6 +52,7 @@
#include "llvm/Support/MachineValueType.h"
#include "llvm/Support/MathExtras.h"
#include "llvm/Target/TargetMachine.h"
+#include "llvm/Target/TargetOptions.h"
#include "llvm/Transforms/Utils/SizeOpts.h"
#include <algorithm>
#include <cassert>
@@ -2020,6 +2021,12 @@
return getTargetMachine().isPositionIndependent();
}
+Align TargetLoweringBase::getPrefLoopAlignment(MachineLoop *ML) const {
+ if (TM.Options.LoopAlignment)
+ return Align(TM.Options.LoopAlignment);
+ return PrefLoopAlignment;
+}
+
//===----------------------------------------------------------------------===//
// Reciprocal Estimates
//===----------------------------------------------------------------------===//
Index: llvm/lib/CodeGen/CommandFlags.cpp
===================================================================
--- llvm/lib/CodeGen/CommandFlags.cpp
+++ llvm/lib/CodeGen/CommandFlags.cpp
@@ -94,6 +94,7 @@
CGOPT(bool, ForceDwarfFrameSection)
CGOPT(bool, XRayOmitFunctionIndex)
CGOPT(bool, DebugStrictDwarf)
+CGOPT(unsigned, AlignLoops)
codegen::RegisterCodeGenFlags::RegisterCodeGenFlags() {
#define CGBINDOPT(NAME) \
@@ -452,6 +453,10 @@
"strict-dwarf", cl::desc("use strict dwarf"), cl::init(false));
CGBINDOPT(DebugStrictDwarf);
+ static cl::opt<unsigned> AlignLoops("align-loops",
+ cl::desc("Default alignment for loops"));
+ CGBINDOPT(AlignLoops);
+
#undef CGBINDOPT
mc::RegisterMCTargetOptionsFlags();
@@ -527,6 +532,7 @@
Options.ForceDwarfFrameSection = getForceDwarfFrameSection();
Options.XRayOmitFunctionIndex = getXRayOmitFunctionIndex();
Options.DebugStrictDwarf = getDebugStrictDwarf();
+ Options.LoopAlignment = getAlignLoops();
Options.MCOptions = mc::InitMCTargetOptionsFromFlags();
Index: llvm/include/llvm/Target/TargetOptions.h
===================================================================
--- llvm/include/llvm/Target/TargetOptions.h
+++ llvm/include/llvm/Target/TargetOptions.h
@@ -328,6 +328,9 @@
/// passed on the command line.
std::string StackUsageOutput;
+ /// If greater than 0, override TargetLoweringBase::PrefLoopAlignment.
+ unsigned LoopAlignment = 0;
+
/// FloatABIType - This setting is set by -float-abi=xxx option is specfied
/// on the command line. This setting may either be Default, Soft, or Hard.
/// Default selects the target's default behavior. Soft selects the ABI for
Index: llvm/include/llvm/CodeGen/TargetLowering.h
===================================================================
--- llvm/include/llvm/CodeGen/TargetLowering.h
+++ llvm/include/llvm/CodeGen/TargetLowering.h
@@ -1763,9 +1763,7 @@
Align getPrefFunctionAlignment() const { return PrefFunctionAlignment; }
/// Return the preferred loop alignment.
- virtual Align getPrefLoopAlignment(MachineLoop *ML = nullptr) const {
- return PrefLoopAlignment;
- }
+ virtual Align getPrefLoopAlignment(MachineLoop *ML = nullptr) const;
/// Should loops be aligned even when the function is marked OptSize (but not
/// MinSize).
Index: llvm/include/llvm/CodeGen/CommandFlags.h
===================================================================
--- llvm/include/llvm/CodeGen/CommandFlags.h
+++ llvm/include/llvm/CodeGen/CommandFlags.h
@@ -138,6 +138,8 @@
bool getDebugStrictDwarf();
+unsigned getAlignLoops();
+
/// Create this object with static storage to register codegen-related command
/// line options.
struct RegisterCodeGenFlags {
Index: clang/test/Driver/falign-loops.c
===================================================================
--- /dev/null
+++ clang/test/Driver/falign-loops.c
@@ -0,0 +1,17 @@
+/// Treat -falign-loops=0 as not specifying the option.
+// RUN: %clang -### -falign-loops=0 %s 2>&1 | FileCheck %s --check-prefix=CHECK_NO
+// RUN: %clang -### -falign-loops=1 %s 2>&1 | FileCheck %s --check-prefix=CHECK_1
+// RUN: %clang -### -falign-loops=4 %s 2>&1 | FileCheck %s --check-prefix=CHECK_4
+/// Only powers of 2 are supported for now.
+// RUN: %clang -### -falign-loops=5 %s 2>&1 | FileCheck %s --check-prefix=CHECK_5
+// RUN: %clang -### -falign-loops=65536 %s 2>&1 | FileCheck %s --check-prefix=CHECK_65536
+// RUN: %clang -### -falign-loops=65537 %s 2>&1 | FileCheck %s --check-prefix=CHECK_65537
+// RUN: %clang -### -falign-loops=a %s 2>&1 | FileCheck %s --check-prefix=CHECK_ERR_A
+
+// CHECK_NO-NOT: "-falign-loops=
+// CHECK_1: "-falign-loops=1"
+// CHECK_4: "-falign-loops=4"
+// CHECK_5: error: alignment is not a power of 2 in '-falign-loops=5'
+// CHECK_65536: "-falign-loops=65536"
+// CHECK_65537: error: invalid integral value '65537' in '-falign-loops=65537'
+// CHECK_ERR_A: error: invalid integral value 'a' in '-falign-loops=a'
Index: clang/test/CodeGen/align-loops.c
===================================================================
--- /dev/null
+++ clang/test/CodeGen/align-loops.c
@@ -0,0 +1,15 @@
+// REQUIRES: x86-registered-target
+/// Check asm because we use llvm::TargetOptions.
+
+// RUN: %clang_cc1 -triple=x86_64 -S %s -falign-loops=8 -O -o - | FileCheck %s --check-prefixes=CHECK,CHECK_8
+// RUN: %clang_cc1 -triple=x86_64 -S %s -falign-loops=32 -O -o - | FileCheck %s --check-prefixes=CHECK,CHECK_32
+
+// CHECK-LABEL: foo:
+// CHECK_8: .p2align 3, 0x90
+// CHECK_32: .p2align 5, 0x90
+
+void bar();
+void foo() {
+ for (int i = 0; i < 64; ++i)
+ bar();
+}
Index: clang/lib/Driver/ToolChains/Clang.cpp
===================================================================
--- clang/lib/Driver/ToolChains/Clang.cpp
+++ clang/lib/Driver/ToolChains/Clang.cpp
@@ -4720,6 +4720,22 @@
CmdArgs.push_back(Args.MakeArgString(std::to_string(FunctionAlignment)));
}
+ // We support -falign-loops=N where N is a power of 2. GCC supports more
+ // forms.
+ if (const Arg *A = Args.getLastArg(options::OPT_falign_loops_EQ)) {
+ unsigned Value = 0;
+ if (StringRef(A->getValue()).getAsInteger(10, Value) || Value > 65536)
+ TC.getDriver().Diag(diag::err_drv_invalid_int_value)
+ << A->getAsString(Args) << A->getValue();
+ else if (Value & Value - 1)
+ TC.getDriver().Diag(diag::err_drv_alignment_not_power_of_two)
+ << A->getAsString(Args) << A->getValue();
+ // Treat =0 as unspecified (use the target preference).
+ if (Value)
+ CmdArgs.push_back(Args.MakeArgString("-falign-loops=" +
+ Twine(std::min(Value, 65536u))));
+ }
+
llvm::Reloc::Model RelocationModel;
unsigned PICLevel;
bool IsPIE;
Index: clang/lib/CodeGen/BackendUtil.cpp
===================================================================
--- clang/lib/CodeGen/BackendUtil.cpp
+++ clang/lib/CodeGen/BackendUtil.cpp
@@ -580,6 +580,7 @@
Options.ValueTrackingVariableLocations =
CodeGenOpts.ValueTrackingVariableLocations;
Options.XRayOmitFunctionIndex = CodeGenOpts.XRayOmitFunctionIndex;
+ Options.LoopAlignment = CodeGenOpts.LoopAlignment;
Options.MCOptions.SplitDwarfFile = CodeGenOpts.SplitDwarfFile;
Options.MCOptions.MCRelaxAll = CodeGenOpts.RelaxAll;
Index: clang/include/clang/Driver/Options.td
===================================================================
--- clang/include/clang/Driver/Options.td
+++ clang/include/clang/Driver/Options.td
@@ -1064,6 +1064,9 @@
PosFlag<SetTrue>>;
def falign_functions : Flag<["-"], "falign-functions">, Group<f_Group>;
def falign_functions_EQ : Joined<["-"], "falign-functions=">, Group<f_Group>;
+def falign_loops_EQ : Joined<["-"], "falign-loops=">, Group<f_Group>, Flags<[CC1Option]>, MetaVarName<"<N>">,
+ HelpText<"N must be a power of two. Align loops to the boundary">,
+ MarshallingInfoInt<CodeGenOpts<"LoopAlignment">>;
def fno_align_functions: Flag<["-"], "fno-align-functions">, Group<f_Group>;
defm allow_editor_placeholders : BoolFOption<"allow-editor-placeholders",
LangOpts<"AllowEditorPlaceholders">, DefaultFalse,
@@ -4293,7 +4296,6 @@
defm align_labels : BooleanFFlag<"align-labels">, Group<clang_ignored_gcc_optimization_f_Group>;
def falign_labels_EQ : Joined<["-"], "falign-labels=">, Group<clang_ignored_gcc_optimization_f_Group>;
defm align_loops : BooleanFFlag<"align-loops">, Group<clang_ignored_gcc_optimization_f_Group>;
-def falign_loops_EQ : Joined<["-"], "falign-loops=">, Group<clang_ignored_gcc_optimization_f_Group>;
defm align_jumps : BooleanFFlag<"align-jumps">, Group<clang_ignored_gcc_optimization_f_Group>;
def falign_jumps_EQ : Joined<["-"], "falign-jumps=">, Group<clang_ignored_gcc_optimization_f_Group>;
Index: clang/include/clang/Basic/LangOptions.def
===================================================================
--- clang/include/clang/Basic/LangOptions.def
+++ clang/include/clang/Basic/LangOptions.def
@@ -396,6 +396,7 @@
"with")
COMPATIBLE_VALUE_LANGOPT(FunctionAlignment, 5, 0, "Default alignment for functions")
+COMPATIBLE_VALUE_LANGOPT(LoopAlignment, 32, 0, "Default alignment for loops")
LANGOPT(FixedPoint, 1, 0, "fixed point types")
LANGOPT(PaddingOnUnsignedFixedPoint, 1, 0,
Index: clang/include/clang/Basic/DiagnosticDriverKinds.td
===================================================================
--- clang/include/clang/Basic/DiagnosticDriverKinds.td
+++ clang/include/clang/Basic/DiagnosticDriverKinds.td
@@ -218,6 +218,7 @@
def err_drv_invalid_int_value : Error<"invalid integral value '%1' in '%0'">;
def err_drv_invalid_value_with_suggestion : Error<
"invalid value '%1' in '%0', expected one of: %2">;
+def err_drv_alignment_not_power_of_two : Error<"alignment is not a power of 2 in '%0'">;
def err_drv_invalid_remap_file : Error<
"invalid option '%0' not of the form <from-file>;<to-file>">;
def err_drv_invalid_gcc_output_type : Error<
Index: clang/include/clang/Basic/CodeGenOptions.def
===================================================================
--- clang/include/clang/Basic/CodeGenOptions.def
+++ clang/include/clang/Basic/CodeGenOptions.def
@@ -293,6 +293,8 @@
///< realignment.
CODEGENOPT(UseInitArray , 1, 0) ///< Control whether to use .init_array or
///< .ctors.
+VALUE_CODEGENOPT(LoopAlignment , 32, 0) ///< Overrides default loop
+ ///< alignment, if not 0.
VALUE_CODEGENOPT(StackAlignment , 32, 0) ///< Overrides default stack
///< alignment, if not 0.
VALUE_CODEGENOPT(StackProbeSize , 32, 4096) ///< Overrides default stack
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
