[PGO][AIX] Improve dummy var retention and allow -bcdtors:csect linking.

1) Use a static array of pointer to retain the dummy vars.
2) Associate liveness of the array with that of the runtime hook variable
   __llvm_profile_runtime.
3) Perform the runtime initialization through the runtime hook variable.
4) Preserve the runtime hook variable using the -u linker flag.

Reviewed By: hubert.reinterpretcast

Differential Revision: https://reviews.llvm.org/D136192
This commit is contained in:
Wael Yehia 2022-10-20 16:07:35 +00:00
parent 4fd2b49115
commit 461a1836d3
9 changed files with 69 additions and 33 deletions

View File

@ -13,6 +13,7 @@
#include "clang/Driver/Options.h" #include "clang/Driver/Options.h"
#include "clang/Driver/SanitizerArgs.h" #include "clang/Driver/SanitizerArgs.h"
#include "llvm/Option/ArgList.h" #include "llvm/Option/ArgList.h"
#include "llvm/ProfileData/InstrProf.h"
#include "llvm/Support/Path.h" #include "llvm/Support/Path.h"
using AIX = clang::driver::toolchains::AIX; using AIX = clang::driver::toolchains::AIX;
@ -348,6 +349,16 @@ void AIX::AddCXXStdlibLibArgs(const llvm::opt::ArgList &Args,
llvm_unreachable("Unexpected C++ library type; only libc++ is supported."); llvm_unreachable("Unexpected C++ library type; only libc++ is supported.");
} }
void AIX::addProfileRTLibs(const llvm::opt::ArgList &Args,
llvm::opt::ArgStringList &CmdArgs) const {
// Add linker option -u__llvm_profile_runtime to cause runtime
// initialization to occur.
if (needsProfileRT(Args))
CmdArgs.push_back(Args.MakeArgString(
Twine("-u", llvm::getInstrProfRuntimeHookVarName())));
ToolChain::addProfileRTLibs(Args, CmdArgs);
}
ToolChain::CXXStdlibType AIX::GetDefaultCXXStdlibType() const { ToolChain::CXXStdlibType AIX::GetDefaultCXXStdlibType() const {
return ToolChain::CST_Libcxx; return ToolChain::CST_Libcxx;
} }

View File

@ -80,6 +80,9 @@ public:
void AddCXXStdlibLibArgs(const llvm::opt::ArgList &Args, void AddCXXStdlibLibArgs(const llvm::opt::ArgList &Args,
llvm::opt::ArgStringList &CmdArgs) const override; llvm::opt::ArgStringList &CmdArgs) const override;
void addProfileRTLibs(const llvm::opt::ArgList &Args,
llvm::opt::ArgStringList &CmdArgs) const override;
CXXStdlibType GetDefaultCXXStdlibType() const override; CXXStdlibType GetDefaultCXXStdlibType() const override;
RuntimeLibType GetDefaultRuntimeLibType() const override; RuntimeLibType GetDefaultRuntimeLibType() const override;

View File

@ -250,23 +250,21 @@ void __llvm_profile_register_names_function(void *NamesStart,
// section exists. So for the scenario where the user objects have no such // section exists. So for the scenario where the user objects have no such
// section (i.e. when they are compiled with -fno-profile-generate), we always // section (i.e. when they are compiled with -fno-profile-generate), we always
// define these zero length variables in each of the above 4 sections. // define these zero length variables in each of the above 4 sections.
COMPILER_RT_VISIBILITY int dummy_cnts[0] COMPILER_RT_SECTION( static int dummy_cnts[0] COMPILER_RT_SECTION(
COMPILER_RT_SEG INSTR_PROF_CNTS_SECT_NAME); COMPILER_RT_SEG INSTR_PROF_CNTS_SECT_NAME);
COMPILER_RT_VISIBILITY int dummy_data[0] COMPILER_RT_SECTION( static int dummy_data[0] COMPILER_RT_SECTION(
COMPILER_RT_SEG INSTR_PROF_DATA_SECT_NAME); COMPILER_RT_SEG INSTR_PROF_DATA_SECT_NAME);
COMPILER_RT_VISIBILITY const int dummy_name[0] COMPILER_RT_SECTION( static const int dummy_name[0] COMPILER_RT_SECTION(
COMPILER_RT_SEG INSTR_PROF_NAME_SECT_NAME); COMPILER_RT_SEG INSTR_PROF_NAME_SECT_NAME);
COMPILER_RT_VISIBILITY int dummy_vnds[0] COMPILER_RT_SECTION( static int dummy_vnds[0] COMPILER_RT_SECTION(
COMPILER_RT_SEG INSTR_PROF_VNODES_SECT_NAME); COMPILER_RT_SEG INSTR_PROF_VNODES_SECT_NAME);
// Create a fake reference to avoid GC'ing of the dummy variables by the linker. // To avoid GC'ing of the dummy variables by the linker, reference them in an
// Ideally, we create a ".ref" of each variable inside the function // array and reference the array in the runtime registration code
// __llvm_profile_begin_counters(), but there's no source level construct // (InstrProfilingRuntime.cpp)
// that allows us to generate that. COMPILER_RT_VISIBILITY
__attribute__((destructor)) void keep() { void *__llvm_profile_keep[] = {(void *)&dummy_cnts, (void *)&dummy_data,
int volatile use = &dummy_cnts < &dummy_data && &dummy_name < &dummy_vnds; (void *)&dummy_name, (void *)&dummy_vnds};
(void)use;
}
#endif #endif
#endif #endif

View File

@ -10,19 +10,23 @@ extern "C" {
#include "InstrProfiling.h" #include "InstrProfiling.h"
static int RegisterRuntime() {
__llvm_profile_initialize();
return 0;
}
#ifndef _AIX
/* int __llvm_profile_runtime */ /* int __llvm_profile_runtime */
COMPILER_RT_VISIBILITY int INSTR_PROF_PROFILE_RUNTIME_VAR; COMPILER_RT_VISIBILITY int INSTR_PROF_PROFILE_RUNTIME_VAR;
}
static int Registration = RegisterRuntime();
namespace { #else
extern COMPILER_RT_VISIBILITY void *__llvm_profile_keep[];
class RegisterRuntime { /* On AIX, when linking with -bcdtors:csect, the variable whose constructor does
public: * the registration needs to be explicitly kept, hence we reuse the runtime hook
RegisterRuntime() { * variable to do the registration since it'll be kept via the -u linker flag.
__llvm_profile_initialize(); * Create a volatile reference to __llvm_profile_keep to keep the array alive.*/
} COMPILER_RT_VISIBILITY int INSTR_PROF_PROFILE_RUNTIME_VAR =
}; ((void)*(void *volatile *)__llvm_profile_keep, RegisterRuntime());
#endif
RegisterRuntime Registration;
} }

View File

@ -0,0 +1,9 @@
def getRoot(config):
if not config.parent:
return config
return getRoot(config.parent)
root = getRoot(config)
if root.host_os not in ['AIX']:
config.unsupported = True

View File

@ -0,0 +1,14 @@
// RUN: split-file %s %t
// RUN: cd %t
//
// RUN: %clang_pgogen foo.c -c -o foo.o
// RUN: %clang_pgogen -shared foo.o -o libfoo.so -bexpall
// RUN: %clang_pgogen -L%t user.c libfoo.so -o user1
// RUN: ./user1
//--- foo.c
void foo() {}
//--- user.c
void foo();
int main() { foo(); }

View File

@ -1199,7 +1199,7 @@ void InstrProfiling::emitRegistration() {
bool InstrProfiling::emitRuntimeHook() { bool InstrProfiling::emitRuntimeHook() {
// We expect the linker to be invoked with -u<hook_var> flag for Linux // We expect the linker to be invoked with -u<hook_var> flag for Linux
// in which case there is no need to emit the external variable. // in which case there is no need to emit the external variable.
if (TT.isOSLinux()) if (TT.isOSLinux() || TT.isOSAIX())
return false; return false;
// If the module's provided its own runtime, we don't need to do anything. // If the module's provided its own runtime, we don't need to do anything.

View File

@ -2,7 +2,7 @@
; RUN: opt < %s -mtriple=x86_64-apple-macosx10.10.0 -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,DARWIN ; RUN: opt < %s -mtriple=x86_64-apple-macosx10.10.0 -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,DARWIN
; RUN: opt < %s -mtriple=x86_64-linux-unknown -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,LINUX ; RUN: opt < %s -mtriple=x86_64-linux-unknown -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,LINUX
; RUN: opt < %s -mtriple=powerpc64-ibm-aix-xcoff -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,DARWIN ; RUN: opt < %s -mtriple=powerpc64-ibm-aix-xcoff -passes=instrprof -S | FileCheck %s -check-prefixes=ALL,LINUX
; ALL-NOT: @__profc ; ALL-NOT: @__profc
; ALL-NOT: @__profd ; ALL-NOT: @__profd
; DARWIN: @__llvm_profile_runtime ; DARWIN: @__llvm_profile_runtime

View File

@ -13,7 +13,7 @@
; MACHO: @__llvm_profile_runtime = external hidden global i32 ; MACHO: @__llvm_profile_runtime = external hidden global i32
; ELF_GENERIC: @__llvm_profile_runtime = external hidden global i32 ; ELF_GENERIC: @__llvm_profile_runtime = external hidden global i32
; ELF-NOT: @__llvm_profile_runtime = external global i32 ; ELF-NOT: @__llvm_profile_runtime = external global i32
; XCOFF: @__llvm_profile_runtime = external hidden global i32 ; XCOFF-NOT: @__llvm_profile_runtime = external hidden global i32
; COFF: @__llvm_profile_runtime = external hidden global i32 ; COFF: @__llvm_profile_runtime = external hidden global i32
; ELF: $__profc_foo = comdat nodeduplicate ; ELF: $__profc_foo = comdat nodeduplicate
@ -101,7 +101,7 @@ declare void @llvm.instrprof.increment(ptr, i64, i32, i32)
; ELF_GENERIC: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime, ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern] ; ELF_GENERIC: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime, ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern]
; MACHO: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime_user, ptr @__profd_foo, {{.*}} ; MACHO: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime_user, ptr @__profd_foo, {{.*}}
; COFF: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime_user, ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern] ; COFF: @llvm.compiler.used = appending global [6 x ptr] [ptr @__llvm_profile_runtime_user, ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern]
; XCOFF: @llvm.used = appending global [7 x ptr] [ptr @__llvm_profile_runtime_user, ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern, ptr @__llvm_prf_nm] ; XCOFF: @llvm.used = appending global [6 x ptr] [ptr @__profd_foo, ptr @__profd_foo_weak, ptr @"__profd_linkage.ll:foo_internal", ptr @__profd_foo_inline, ptr @__profd_foo_extern, ptr @__llvm_prf_nm]
; MACHO: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} { ; MACHO: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} {
; MACHO: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime ; MACHO: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime
@ -112,10 +112,7 @@ declare void @llvm.instrprof.increment(ptr, i64, i32, i32)
; ELFRT-NOT: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime ; ELFRT-NOT: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime
; PS: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} { ; PS: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} {
; PS: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime ; PS: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime
; XCOFF: define linkonce_odr hidden i32 @__llvm_profile_runtime_user() {{.*}} { ; XCOFF-NOT: define .* __llvm_profile_runtime_user
; XCOFF: %[[REG:.*]] = load i32, ptr @__llvm_profile_runtime
; XCOFF: ret i32 %[[REG]]
; XCOFF: }
; ELF_GENERIC: define internal void @__llvm_profile_register_functions() unnamed_addr { ; ELF_GENERIC: define internal void @__llvm_profile_register_functions() unnamed_addr {
; ELF_GENERIC-NEXT: call void @__llvm_profile_register_function(ptr @__llvm_profile_runtime) ; ELF_GENERIC-NEXT: call void @__llvm_profile_register_function(ptr @__llvm_profile_runtime)