[clang] 584cb67 - [IRSymTab] Set FB_used on llvm.compiler.used symbols
Fangrui Song via cfe-commits
cfe-commits at lists.llvm.org
Wed Mar 3 16:22:38 PST 2021
Author: Fangrui Song
Date: 2021-03-03T16:22:30-08:00
New Revision: 584cb67d2df32163efe61cfa58b3c798ee5b6b5c
URL: https://github.com/llvm/llvm-project/commit/584cb67d2df32163efe61cfa58b3c798ee5b6b5c
DIFF: https://github.com/llvm/llvm-project/commit/584cb67d2df32163efe61cfa58b3c798ee5b6b5c.diff
LOG: [IRSymTab] Set FB_used on llvm.compiler.used symbols
IR symbol table does not parse inline asm. A symbol only referenced by inline
asm is not in the IR symbol table, so LTO does not know that the definition (in
another translation unit) is referenced and may internalize it, even if that
definition has `__attribute__((used))` (which lowers to `llvm.compiler.used` on
ELF targets since D97446).
```
// cabac.c
__attribute__((used)) const uint8_t ff_h264_cabac_tables[...] = {...};
// h264_cabac.c
asm("lea ff_h264_cabac_tables(%rip), %0" : ...);
```
`__attribute__((used))` is the recommended way to tell the compiler there may
be inline asm references, so the usage is perfectly fine. This patch
conservatively sets the `FB_used` bit on `llvm.compiler.used` symbols to work
around the IR symbol table limitation. Note: before D97446, Clang never emitted
symbols in the `llvm.compiler.used` list, so this change does not punish any
Clang emitted global object.
Without the patch, `ff_h264_cabac_tables` may be assigned to a non-external
partition and get internalized. Then we will get a linker error because the
`cabac.c` definition is not exposed.
Differential Revision: https://reviews.llvm.org/D97755
Added:
clang/test/CodeGen/thinlto-inline-asm2.c
llvm/test/ThinLTO/X86/asm.ll
Modified:
llvm/lib/LTO/LTO.cpp
llvm/lib/Object/IRSymtab.cpp
llvm/test/tools/gold/X86/emit-llvm.ll
Removed:
################################################################################
diff --git a/clang/test/CodeGen/thinlto-inline-asm2.c b/clang/test/CodeGen/thinlto-inline-asm2.c
new file mode 100644
index 000000000000..30d8bfb04c5b
--- /dev/null
+++ b/clang/test/CodeGen/thinlto-inline-asm2.c
@@ -0,0 +1,29 @@
+// REQUIRES: x86-registered-target
+
+// RUN: split-file %s %t
+// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -flto=thin -emit-llvm-bc %t/a.c -o %t/a.bc
+// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -flto=thin -emit-llvm-bc %t/b.c -o %t/b.bc
+// RUN: llvm-nm %t/a.bc | FileCheck %s --check-prefix=NM
+
+// RUN: llvm-lto2 run %t/a.bc %t/b.bc -o %t/out -save-temps -r=%t/a.bc,ref,plx -r=%t/b.bc,ff_h264_cabac_tables,pl
+// RUN: llvm-dis < %t/out.2.2.internalize.bc | FileCheck %s
+
+//--- a.c
+/// IR symtab does not track inline asm symbols, so we don't know
+/// ff_h264_cabac_tables is undefined.
+// NM-NOT: {{.}}
+// NM: ---------------- T ref
+// NM-NOT: {{.}}
+const char *ref() {
+ const char *ret;
+ asm("lea ff_h264_cabac_tables(%%rip), %0" : "=r"(ret));
+ return ret;
+}
+
+//--- b.c
+/// ff_h264_cabac_tables has __attribute__((used)) in the source code, which means
+/// its definition must be retained because there can be references the compiler
+/// cannot see (inline asm reference). Test we don't internalize it.
+// CHECK: @ff_h264_cabac_tables = dso_local constant [1 x i8] c"\09"
+__attribute__((used))
+const char ff_h264_cabac_tables[1] = "\x09";
diff --git a/llvm/lib/LTO/LTO.cpp b/llvm/lib/LTO/LTO.cpp
index ea576269ec84..8bcb1600925d 100644
--- a/llvm/lib/LTO/LTO.cpp
+++ b/llvm/lib/LTO/LTO.cpp
@@ -563,8 +563,8 @@ void LTO::addModuleToGlobalRes(ArrayRef<InputFile::Symbol> Syms,
// Set the partition to external if we know it is re-defined by the linker
// with -defsym or -wrap options, used elsewhere, e.g. it is visible to a
- // regular object, is referenced from llvm.compiler_used, or was already
- // recorded as being referenced from a
diff erent partition.
+ // regular object, is referenced from llvm.compiler.used/llvm.used, or was
+ // already recorded as being referenced from a
diff erent partition.
if (Res.LinkerRedefined || Res.VisibleToRegularObj || Sym.isUsed() ||
(GlobalRes.Partition != GlobalResolution::Unknown &&
GlobalRes.Partition != Partition)) {
diff --git a/llvm/lib/Object/IRSymtab.cpp b/llvm/lib/Object/IRSymtab.cpp
index 16c6433a2061..4ff73f9356cb 100644
--- a/llvm/lib/Object/IRSymtab.cpp
+++ b/llvm/lib/Object/IRSymtab.cpp
@@ -119,8 +119,20 @@ Error Builder::addModule(Module *M) {
return make_error<StringError>("input module has no datalayout",
inconvertibleErrorCode());
+ // Symbols in the llvm.used list will get the FB_Used bit and will not be
+ // internalized. We do this for llvm.compiler.used as well:
+ //
+ // IR symbol table tracks module-level asm symbol references but not inline
+ // asm. A symbol only referenced by inline asm is not in the IR symbol table,
+ // so we may not know that the definition (in another translation unit) is
+ // referenced. That definition may have __attribute__((used)) (which lowers to
+ // llvm.compiler.used on ELF targets) to communicate to the compiler that it
+ // may be used by inline asm. The usage is perfectly fine, so we treat
+ // llvm.compiler.used conservatively as llvm.used to work around our own
+ // limitation.
SmallVector<GlobalValue *, 4> UsedV;
collectUsedGlobalVariables(*M, UsedV, /*CompilerUsed=*/false);
+ collectUsedGlobalVariables(*M, UsedV, /*CompilerUsed=*/true);
SmallPtrSet<GlobalValue *, 4> Used(UsedV.begin(), UsedV.end());
ModuleSymbolTable Msymtab;
diff --git a/llvm/test/ThinLTO/X86/asm.ll b/llvm/test/ThinLTO/X86/asm.ll
new file mode 100644
index 000000000000..0d4066e8f889
--- /dev/null
+++ b/llvm/test/ThinLTO/X86/asm.ll
@@ -0,0 +1,34 @@
+; RUN: split-file %s %t
+; RUN: opt -module-summary %t/a.s -o %t/a.bc
+; RUN: opt -module-summary %t/b.s -o %t/b.bc
+; RUN: llvm-nm %t/a.bc | FileCheck %s --check-prefix=NM
+
+; RUN: llvm-lto2 run %t/a.bc %t/b.bc -o %t/out -save-temps -r=%t/a.bc,ref,plx -r=%t/b.bc,ff_h264_cabac_tables,pl
+; RUN: llvm-dis < %t/out.2.2.internalize.bc | FileCheck %s
+
+;--- a.s
+;; IR symtab does not track inline asm symbols, so we don't know
+;; ff_h264_cabac_tables is undefined.
+; NM-NOT: {{.}}
+; NM: ---------------- T ref
+; NM-NOT: {{.}}
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define i8* @ref() {
+entry:
+ %0 = tail call i8* asm sideeffect "lea ff_h264_cabac_tables(%rip), $0", "=&r,~{dirflag},~{fpsr},~{flags}"()
+ ret i8* %0
+}
+
+;--- b.s
+target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+;; ff_h264_cabac_tables has __attribute__((used)) in the source code, which means
+;; its definition must be retained because there can be references the compiler
+;; cannot see (inline asm reference). Test we don't internalize it.
+; CHECK: @ff_h264_cabac_tables = dso_local constant [1 x i8] c"\09"
+ at ff_h264_cabac_tables = dso_local constant [1 x i8] c"\09"
+
+ at llvm.compiler.used = appending global [1 x i8*] [i8* bitcast ([1 x i8]* @ff_h264_cabac_tables to i8*)], section "llvm.metadata"
diff --git a/llvm/test/tools/gold/X86/emit-llvm.ll b/llvm/test/tools/gold/X86/emit-llvm.ll
index f1f6b22a9c3c..28a81071fdbb 100644
--- a/llvm/test/tools/gold/X86/emit-llvm.ll
+++ b/llvm/test/tools/gold/X86/emit-llvm.ll
@@ -54,6 +54,7 @@ define hidden void @f1() {
ret void
}
+;; f2 is not internalized because it is in the llvm.used list.
; CHECK-DAG: define hidden void @f2()
; OPT-DAG: define hidden void @f2()
define hidden void @f2() {
@@ -97,6 +98,15 @@ define i32* @f8() {
ret i32* @g8
}
+;; f9 is not internalized because it is in the llvm.compiler.used list.
+; CHECK-DAG: define hidden void @f9()
+; OPT-DAG: define hidden void @f9()
+define hidden void @f9() {
+ ret void
+}
+
+ at llvm.compiler.used = appending global [1 x i8*] [ i8* bitcast (void ()* @f9 to i8*)]
+
; RES: .o,f1,pl{{$}}
; RES: .o,f2,pl{{$}}
; RES: .o,f3,px{{$}}
@@ -105,6 +115,7 @@ define i32* @f8() {
; RES: .o,f6,p{{$}}
; RES: .o,f7,px{{$}}
; RES: .o,f8,px{{$}}
+; RES: .o,f9,pl{{$}}
; RES: .o,g1,px{{$}}
; RES: .o,g2,p{{$}}
; RES: .o,g3,p{{$}}
More information about the cfe-commits
mailing list