[clang] f514b16 - [X86] Support -mnop-mcount and -mrecord-mcount on x86 elf targets (#184641)
via cfe-commits
cfe-commits at lists.llvm.org
Thu Oct 22 00:27:24 PDT 2026
Author: Paul Murphy
Date: 2026-10-22T09:27:19+02:00
New Revision: f514b16972f785fdd3fad1d19195aad5c9324fb5
URL: https://github.com/llvm/llvm-project/commit/f514b16972f785fdd3fad1d19195aad5c9324fb5
DIFF: https://github.com/llvm/llvm-project/commit/f514b16972f785fdd3fad1d19195aad5c9324fb5.diff
LOG: [X86] Support -mnop-mcount and -mrecord-mcount on x86 elf targets (#184641)
Support mnop-mcount and mrecord-mcount attributes when used with the
fentry-call=true attribute.
Similar to SystemZ, use the mnop-mcount attribute to place a nop instead
of a call to __fentry__.
Likewise, use the mrecord-mcount attribute to record the location of the
call to fentry (or nop if used with the above) to a section named
__mcount_loc.
This enables the use of -mnop-mcount -mrecord-mcount for x86 in Clang.
Added:
llvm/test/CodeGen/X86/mnop-mcount-01.ll
llvm/test/CodeGen/X86/mrecord-mcount-01.ll
llvm/test/Verifier/mnop-mcount.ll
llvm/test/Verifier/mrecord-mcount.ll
Modified:
clang/lib/Driver/ToolChains/Clang.cpp
clang/test/CodeGen/mnop-mcount.c
clang/test/CodeGen/mrecord-mcount.c
clang/test/Driver/mcount.c
llvm/lib/IR/Verifier.cpp
llvm/lib/Target/SystemZ/SystemZISelDAGToDAG.cpp
llvm/lib/Target/X86/X86MCInstLower.cpp
Removed:
llvm/test/CodeGen/SystemZ/mnop-mcount-02.ll
llvm/test/CodeGen/SystemZ/mrecord-mcount-02.ll
################################################################################
diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp
index ffb225be471800..454673dd20f9ad 100644
--- a/clang/lib/Driver/ToolChains/Clang.cpp
+++ b/clang/lib/Driver/ToolChains/Clang.cpp
@@ -7326,14 +7326,16 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA,
<< A->getAsString(Args) << TripleStr;
}
if (Arg *A = Args.getLastArg(options::OPT_mnop_mcount)) {
- if (Arch == llvm::Triple::systemz)
+ if (Arch == llvm::Triple::systemz ||
+ (TC.getTriple().isX86() && TC.getTriple().isOSBinFormatELF()))
A->render(Args, CmdArgs);
else
D.Diag(diag::err_drv_unsupported_opt_for_target)
<< A->getAsString(Args) << TripleStr;
}
if (Arg *A = Args.getLastArg(options::OPT_mrecord_mcount)) {
- if (Arch == llvm::Triple::systemz)
+ if (Arch == llvm::Triple::systemz ||
+ (TC.getTriple().isX86() && TC.getTriple().isOSBinFormatELF()))
A->render(Args, CmdArgs);
else
D.Diag(diag::err_drv_unsupported_opt_for_target)
diff --git a/clang/test/CodeGen/mnop-mcount.c b/clang/test/CodeGen/mnop-mcount.c
index ebb21cc560e370..9455043b0d4e33 100644
--- a/clang/test/CodeGen/mnop-mcount.c
+++ b/clang/test/CodeGen/mnop-mcount.c
@@ -6,6 +6,22 @@
// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
// RUN: %clang_cc1 -mnop-mcount -triple s390x-ibm-linux -emit-llvm -o - %s \
// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -pg -mfentry -mnop-mcount -triple x86_64-unknown-linux -emit-llvm \
+// RUN: -o - %s 2>&1 | FileCheck %s
+// RUN: not %clang_cc1 -pg -mnop-mcount -triple x86_64-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOMFENTRY %s
+// RUN: %clang_cc1 -mfentry -mnop-mcount -triple x86_64-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -mnop-mcount -triple x86_64-unknown-linux -emit-llvm -o - %s \
+// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -pg -mfentry -mnop-mcount -triple i686-unknown-linux -emit-llvm \
+// RUN: -o - %s 2>&1 | FileCheck %s
+// RUN: not %clang_cc1 -pg -mnop-mcount -triple i686-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOMFENTRY %s
+// RUN: %clang_cc1 -mfentry -mnop-mcount -triple i686-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -mnop-mcount -triple i686-unknown-linux -emit-llvm -o - %s \
+// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
int foo(void) {
return 0;
diff --git a/clang/test/CodeGen/mrecord-mcount.c b/clang/test/CodeGen/mrecord-mcount.c
index 292a807073cefa..f00b4594f5d8ec 100644
--- a/clang/test/CodeGen/mrecord-mcount.c
+++ b/clang/test/CodeGen/mrecord-mcount.c
@@ -6,6 +6,22 @@
// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
// RUN: %clang_cc1 -mrecord-mcount -triple s390x-ibm-linux -emit-llvm -o - %s \
// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -pg -mfentry -mrecord-mcount -triple x86_64-unknown-linux -emit-llvm \
+// RUN: -o - %s 2>&1 | FileCheck %s
+// RUN: not %clang_cc1 -pg -mrecord-mcount -triple x86_64-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOMFENTRY %s
+// RUN: %clang_cc1 -mfentry -mrecord-mcount -triple x86_64-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -mrecord-mcount -triple x86_64-unknown-linux -emit-llvm -o - %s \
+// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -pg -mfentry -mrecord-mcount -triple i686-unknown-linux -emit-llvm \
+// RUN: -o - %s 2>&1 | FileCheck %s
+// RUN: not %clang_cc1 -pg -mrecord-mcount -triple i686-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOMFENTRY %s
+// RUN: %clang_cc1 -mfentry -mrecord-mcount -triple i686-unknown-linux -emit-llvm -o - \
+// RUN: %s 2>&1 | FileCheck -check-prefix=NOPG %s
+// RUN: %clang_cc1 -mrecord-mcount -triple i686-unknown-linux -emit-llvm -o - %s \
+// RUN: 2>&1 | FileCheck -check-prefix=NOPG %s
int foo(void) {
return 0;
diff --git a/clang/test/Driver/mcount.c b/clang/test/Driver/mcount.c
index 70f554745b9c30..d50055f90344e0 100644
--- a/clang/test/Driver/mcount.c
+++ b/clang/test/Driver/mcount.c
@@ -1,12 +1,11 @@
// RUN: %clang --target=s390x -c -### %s -mnop-mcount -mrecord-mcount 2>&1 | FileCheck %s
+// RUN: %clang --target=x86_64 -c -### %s -mnop-mcount -mrecord-mcount 2>&1 | FileCheck %s
+// RUN: %clang --target=i686 -c -### %s -mnop-mcount -mrecord-mcount 2>&1 | FileCheck %s
// CHECK: "-mnop-mcount"
// CHECK: "-mrecord-mcount"
-// RUN: not %clang --target=x86_64 -c -### %s -mnop-mcount -mrecord-mcount 2>&1 | FileCheck --check-prefix=ERR1 %s
// RUN: not %clang --target=aarch64 -c -### %s -mnop-mcount -mrecord-mcount 2>&1 | FileCheck --check-prefix=ERR2 %s
-// ERR1: error: unsupported option '-mnop-mcount' for target 'x86_64'
-// ERR1: error: unsupported option '-mrecord-mcount' for target 'x86_64'
// ERR2: error: unsupported option '-mnop-mcount' for target 'aarch64'
// ERR2: error: unsupported option '-mrecord-mcount' for target 'aarch64'
diff --git a/llvm/lib/IR/Verifier.cpp b/llvm/lib/IR/Verifier.cpp
index e3d9cc8d7e6fdf..b78e1453f1dba4 100644
--- a/llvm/lib/IR/Verifier.cpp
+++ b/llvm/lib/IR/Verifier.cpp
@@ -2832,6 +2832,18 @@ void Verifier::verifyFunctionAttrs(FunctionType *FT, AttributeList Attrs,
Check(!(Attrs.hasFnAttr(Attribute::Hot) && Attrs.hasFnAttr(Attribute::Cold)),
"Attributes 'hot' and 'cold' are incompatible");
+
+ if (Attrs.hasFnAttr("mnop-mcount")) {
+ auto A = Attrs.getFnAttr("fentry-call");
+ Check(A.isValid() && A.getValueAsString() == "true",
+ "mnop-mcount only supported with fentry-call");
+ }
+
+ if (Attrs.hasFnAttr("mrecord-mcount")) {
+ auto A = Attrs.getFnAttr("fentry-call");
+ Check(A.isValid() && A.getValueAsString() == "true",
+ "mrecord-mcount only supported with fentry-call");
+ }
}
void Verifier::verifyUnknownProfileMetadata(MDNode *MD) {
diff --git a/llvm/lib/Target/SystemZ/SystemZISelDAGToDAG.cpp b/llvm/lib/Target/SystemZ/SystemZISelDAGToDAG.cpp
index 9ab0a4a7f23689..86fa6985a53ce6 100644
--- a/llvm/lib/Target/SystemZ/SystemZISelDAGToDAG.cpp
+++ b/llvm/lib/Target/SystemZ/SystemZISelDAGToDAG.cpp
@@ -373,12 +373,6 @@ class SystemZDAGToDAGISel : public SelectionDAGISel {
bool runOnMachineFunction(MachineFunction &MF) override {
const Function &F = MF.getFunction();
- if (F.getFnAttribute("fentry-call").getValueAsString() != "true") {
- if (F.hasFnAttribute("mnop-mcount"))
- report_fatal_error("mnop-mcount only supported with fentry-call");
- if (F.hasFnAttribute("mrecord-mcount"))
- report_fatal_error("mrecord-mcount only supported with fentry-call");
- }
if (F.getParent()->getStackProtectorGuard() != "global") {
if (F.getParent()->hasStackProtectorGuardRecord())
report_fatal_error("mstack-protector-guard-record only supported with "
diff --git a/llvm/lib/Target/X86/X86MCInstLower.cpp b/llvm/lib/Target/X86/X86MCInstLower.cpp
index 7a71b2746e7980..54fb24e7a06f33 100644
--- a/llvm/lib/Target/X86/X86MCInstLower.cpp
+++ b/llvm/lib/Target/X86/X86MCInstLower.cpp
@@ -26,6 +26,7 @@
#include "llvm/ADT/STLExtras.h"
#include "llvm/ADT/SmallString.h"
#include "llvm/ADT/StringExtras.h"
+#include "llvm/BinaryFormat/ELF.h"
#include "llvm/CodeGen/MachineBranchProbabilityInfo.h"
#include "llvm/CodeGen/MachineConstantPool.h"
#include "llvm/CodeGen/MachineFunction.h"
@@ -44,6 +45,7 @@
#include "llvm/MC/MCInst.h"
#include "llvm/MC/MCInstBuilder.h"
#include "llvm/MC/MCSection.h"
+#include "llvm/MC/MCSectionELF.h"
#include "llvm/MC/MCStreamer.h"
#include "llvm/MC/MCSymbol.h"
#include "llvm/MC/TargetRegistry.h"
@@ -661,9 +663,8 @@ static unsigned emitNop(MCStreamer &OS, unsigned NumBytes,
// target cpu. 15-bytes is the longest single NOP instruction, but some
// platforms can't decode the longest forms efficiently.
unsigned MaxNopLength = 1;
+ unsigned BaseReg = X86::RAX;
if (Subtarget->is64Bit()) {
- // FIXME: We can use NOOPL on 32-bit targets with FeatureNOPL, but the
- // IndexReg/BaseReg below need to be updated.
if (Subtarget->hasFeature(X86::TuningFast7ByteNOP))
MaxNopLength = 7;
else if (Subtarget->hasFeature(X86::TuningFast15ByteNOP))
@@ -672,16 +673,18 @@ static unsigned emitNop(MCStreamer &OS, unsigned NumBytes,
MaxNopLength = 11;
else
MaxNopLength = 10;
- } if (Subtarget->is32Bit())
+ } else if (Subtarget->is32Bit() && Subtarget->hasFeature(X86::FeatureNOPL)) {
+ BaseReg = X86::EAX;
+ MaxNopLength = 10;
+ } else if (Subtarget->is32Bit())
MaxNopLength = 2;
// Cap a single nop emission at the profitable value for the target
NumBytes = std::min(NumBytes, MaxNopLength);
unsigned NopSize;
- unsigned Opc, BaseReg, ScaleVal, IndexReg, Displacement, SegmentReg;
+ unsigned Opc, ScaleVal, IndexReg, Displacement, SegmentReg;
IndexReg = Displacement = SegmentReg = 0;
- BaseReg = X86::RAX;
ScaleVal = 1;
switch (NumBytes) {
case 0:
@@ -708,13 +711,13 @@ static unsigned emitNop(MCStreamer &OS, unsigned NumBytes,
NopSize = 5;
Opc = X86::NOOPL;
Displacement = 8;
- IndexReg = X86::RAX;
+ IndexReg = BaseReg;
break;
case 6:
NopSize = 6;
Opc = X86::NOOPW;
Displacement = 8;
- IndexReg = X86::RAX;
+ IndexReg = BaseReg;
break;
case 7:
NopSize = 7;
@@ -725,19 +728,19 @@ static unsigned emitNop(MCStreamer &OS, unsigned NumBytes,
NopSize = 8;
Opc = X86::NOOPL;
Displacement = 512;
- IndexReg = X86::RAX;
+ IndexReg = BaseReg;
break;
case 9:
NopSize = 9;
Opc = X86::NOOPW;
Displacement = 512;
- IndexReg = X86::RAX;
+ IndexReg = BaseReg;
break;
default:
NopSize = 10;
Opc = X86::NOOPW;
Displacement = 512;
- IndexReg = X86::RAX;
+ IndexReg = BaseReg;
SegmentReg = X86::CS;
break;
}
@@ -884,6 +887,24 @@ void X86AsmPrinter::LowerFENTRY_CALL(const MachineInstr &MI,
X86MCInstLower &MCIL) {
bool Is64Bits = Subtarget->is64Bit();
MCContext &Ctx = OutStreamer->getContext();
+
+ if (MF->getFunction().hasFnAttribute("mrecord-mcount")) {
+ MCSymbol *DotSym = OutContext.createTempSymbol();
+ OutStreamer->pushSection();
+ OutStreamer->switchSection(
+ Ctx.getELFSection("__mcount_loc", ELF::SHT_PROGBITS, ELF::SHF_ALLOC));
+ OutStreamer->emitSymbolValue(DotSym, Is64Bits ? 8 : 4);
+ OutStreamer->popSection();
+ OutStreamer->emitLabel(DotSym);
+ }
+
+ if (MF->getFunction().hasFnAttribute("mnop-mcount")) {
+ // Insert a NOP stream which can be rewritten into a 5-byte CALL
+ // instruction at runtime.
+ emitX86Nops(*OutStreamer, 5, Subtarget);
+ return;
+ }
+
MCSymbol *fentry = Ctx.getOrCreateSymbol("__fentry__");
const MCSymbolRefExpr *Op = MCSymbolRefExpr::create(fentry, Ctx);
diff --git a/llvm/test/CodeGen/SystemZ/mnop-mcount-02.ll b/llvm/test/CodeGen/SystemZ/mnop-mcount-02.ll
deleted file mode 100644
index d6bb1ae7150aae..00000000000000
--- a/llvm/test/CodeGen/SystemZ/mnop-mcount-02.ll
+++ /dev/null
@@ -1,10 +0,0 @@
-; RUN: not --crash llc %s -mtriple=s390x-linux-gnu -o - 2>&1 | FileCheck %s
-;
-; CHECK: LLVM ERROR: mnop-mcount only supported with fentry-call
-
-define void @test1() #0 {
-entry:
- ret void
-}
-
-attributes #0 = { "instrument-function-entry-inlined"="mcount" "mnop-mcount" }
diff --git a/llvm/test/CodeGen/SystemZ/mrecord-mcount-02.ll b/llvm/test/CodeGen/SystemZ/mrecord-mcount-02.ll
deleted file mode 100644
index e1f4a3748b76e2..00000000000000
--- a/llvm/test/CodeGen/SystemZ/mrecord-mcount-02.ll
+++ /dev/null
@@ -1,10 +0,0 @@
-; RUN: not --crash llc %s -mtriple=s390x-linux-gnu -o - 2>&1 | FileCheck %s
-;
-; CHECK: LLVM ERROR: mrecord-mcount only supported with fentry-call
-
-define void @test1() #0 {
-entry:
- ret void
-}
-
-attributes #0 = { "instrument-function-entry-inlined"="mcount" "mrecord-mcount" }
diff --git a/llvm/test/CodeGen/X86/mnop-mcount-01.ll b/llvm/test/CodeGen/X86/mnop-mcount-01.ll
new file mode 100644
index 00000000000000..a6e25e74c7a023
--- /dev/null
+++ b/llvm/test/CodeGen/X86/mnop-mcount-01.ll
@@ -0,0 +1,50 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
+; RUN: llc %s -mtriple=x86_64-linux-gnu -o - | FileCheck --check-prefixes=X64 %s
+; RUN: llc %s -mtriple=i686-linux-gnu -o - | FileCheck --check-prefix=X86 %s
+; RUN: llc %s -mtriple=i686-linux-gnu -mcpu=pentium_pro -o - | FileCheck --check-prefix=PPRO %s
+
+define void @test1() #0 {
+; X64-LABEL: test1:
+; X64: # %bb.0: # %entry
+; X64-NEXT: callq __fentry__
+; X64-NEXT: retq
+;
+; X86-LABEL: test1:
+; X86: # %bb.0: # %entry
+; X86-NEXT: calll __fentry__
+; X86-NEXT: retl
+;
+; PPRO-LABEL: test1:
+; PPRO: # %bb.0: # %entry
+; PPRO-NEXT: calll __fentry__
+; PPRO-NEXT: retl
+entry:
+ ret void
+
+}
+
+define void @test2() #1 {
+; X64-LABEL: test2:
+; X64: # %bb.0: # %entry
+; X64-NEXT: nopl 8(%rax,%rax)
+; X64-NEXT: retq
+;
+; X86-LABEL: test2:
+; X86: # %bb.0: # %entry
+; X86-NEXT: xchgw %ax, %ax
+; X86-NEXT: xchgw %ax, %ax
+; X86-NEXT: nop
+; X86-NEXT: retl
+;
+; PPRO-LABEL: test2:
+; PPRO: # %bb.0: # %entry
+; PPRO-NEXT: nopl 8(%eax,%eax)
+; PPRO-NEXT: retl
+entry:
+ ret void
+
+}
+
+attributes #0 = { "fentry-call"="true" }
+attributes #1 = { "fentry-call"="true" "mnop-mcount" }
+
diff --git a/llvm/test/CodeGen/X86/mrecord-mcount-01.ll b/llvm/test/CodeGen/X86/mrecord-mcount-01.ll
new file mode 100644
index 00000000000000..90dd75f217a43f
--- /dev/null
+++ b/llvm/test/CodeGen/X86/mrecord-mcount-01.ll
@@ -0,0 +1,65 @@
+; RUN: llc %s -mtriple=x86_64-linux-gnu -o - | FileCheck --check-prefix=X64 %s
+; RUN: llc %s -mtriple=i686-linux-gnu -o - | FileCheck --check-prefix=X86 %s
+; RUN: llc %s -mtriple=i686-linux-gnu -mcpu=pentium_pro -o - | FileCheck --check-prefix=PPRO %s
+
+define void @test1() #0 {
+entry:
+ ret void
+
+; X64-LABEL: test1:
+; X64: .section __mcount_loc,"a", at progbits
+; X64: .quad .Ltmp0
+; X64: .text
+; X64: .Ltmp0:
+; X64: callq __fentry__
+; X64: retq
+; X86-LABEL: test1:
+; X86: .section __mcount_loc,"a", at progbits
+; X86: .long .Ltmp0
+; X86: .text
+; X86: .Ltmp0:
+; X86: calll __fentry__
+; X86: retl
+; PPRO-LABEL: test1:
+; PPRO: .section __mcount_loc,"a", at progbits
+; PPRO: .long .Ltmp0
+; PPRO: .text
+; PPRO: .Ltmp0:
+; PPRO: calll __fentry__
+; PPRO: retl
+}
+
+define void @test2() #1 {
+entry:
+ ret void
+
+; X64-LABEL: test2:
+; X64: .section __mcount_loc,"a", at progbits
+; X64: .quad .Ltmp1
+; X64: .text
+; X64: .Ltmp1:
+; X64: nopl 8(%rax,%rax)
+; X64-NOT: callq __fentry__
+; X64: retq
+; X86-LABEL: test2:
+; X86: .section __mcount_loc,"a", at progbits
+; X86: .long .Ltmp1
+; X86: .text
+; X86: .Ltmp1:
+; X86: xchgw %ax, %ax
+; X86: xchgw %ax, %ax
+; X86: nop
+; X86-NOT: calll __fentry__
+; X86: retl
+; PPRO-LABEL: test2:
+; PPRO: .section __mcount_loc,"a", at progbits
+; PPRO: .long .Ltmp1
+; PPRO: .text
+; PPRO: .Ltmp1:
+; PPRO: nopl 8(%eax,%eax)
+; PPRO-NOT: calll __fentry__
+; PPRO: retl
+}
+
+attributes #0 = { "fentry-call"="true" "mrecord-mcount" }
+attributes #1 = { "fentry-call"="true" "mnop-mcount" "mrecord-mcount" }
diff --git a/llvm/test/Verifier/mnop-mcount.ll b/llvm/test/Verifier/mnop-mcount.ll
new file mode 100644
index 00000000000000..222a8896fa0636
--- /dev/null
+++ b/llvm/test/Verifier/mnop-mcount.ll
@@ -0,0 +1,10 @@
+; RUN: not llvm-as %s -o /dev/null 2>&1 | FileCheck %s
+;
+; CHECK: mnop-mcount only supported with fentry-call
+
+define void @test1() #0 {
+entry:
+ ret void
+}
+
+attributes #0 = { "instrument-function-entry-inlined"="mcount" "mnop-mcount" }
diff --git a/llvm/test/Verifier/mrecord-mcount.ll b/llvm/test/Verifier/mrecord-mcount.ll
new file mode 100644
index 00000000000000..abb69f9427dac3
--- /dev/null
+++ b/llvm/test/Verifier/mrecord-mcount.ll
@@ -0,0 +1,10 @@
+; RUN: not llvm-as %s -o /dev/null 2>&1 | FileCheck %s
+;
+; CHECK: mrecord-mcount only supported with fentry-call
+
+define void @test1() #0 {
+entry:
+ ret void
+}
+
+attributes #0 = { "instrument-function-entry-inlined"="mcount" "mrecord-mcount" }
More information about the cfe-commits
mailing list