LLVM ir如何导入其他模块？

Question

我正在通过这个 LangRef 学习 LLVM IR。

如本参考书所述：

LLVM programs are composed of Modules, each of which is a translation unit of the input programs. Each module consists of functions, global variables, and symbol table entries.

这表明一个程序可以由多个模块组成，但我没有找到组合多个模块的方法。

我找到的最可能的方法是 DICompileUnit or DIImportedEntry，但我不知道具体怎么做。

Answer 1

您可以像组合 .o 对象一样组合 LLVM 模块：通过 linking 它们。

例如拿这2个c文件：

a.c

int f(int);

int main(void) {
    return f(1);
}

b.c

int f(int x){
    return x*x;
}

将它们转换为 LLVM IR：

clang -c -emit-llvm a.c  # Produces a.bc
clang -c -emit-llvm b.c  # Produces b.bc

拆解它们：

llvm-dis a.bc  # Produces a.ll
llvm-dis b.bc  # Produces b.ll

现在 link 他们：

llvm-link a.bc b.c -o c.bc
llvm-dis c.bc  # Produces c.ll

原始IR文件为：

a.ll

; ModuleID = 'a.bc'
source_filename = "a.c"
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-pc-linux-gnu"

; Function Attrs: noinline nounwind optnone sspstrong uwtable
define dso_local i32 @main() #0 {
  %1 = alloca i32, align 4
  store i32 0, i32* %1, align 4
  %2 = call i32 @f(i32 1)
  ret i32 %2
}

declare i32 @f(i32) #1

attributes #0 = { noinline nounwind optnone sspstrong uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+cx8,+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }
attributes #1 = { "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+cx8,+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }

!llvm.module.flags = !{!0, !1, !2}
!llvm.ident = !{!3}

!0 = !{i32 1, !"wchar_size", i32 4}
!1 = !{i32 7, !"PIC Level", i32 2}
!2 = !{i32 7, !"PIE Level", i32 2}
!3 = !{!"clang version 10.0.0 "}

b.ll

; ModuleID = 'b.bc'
source_filename = "b.c"
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-pc-linux-gnu"

; Function Attrs: noinline nounwind optnone sspstrong uwtable
define dso_local i32 @f(i32 %0) #0 {
  %2 = alloca i32, align 4
  store i32 %0, i32* %2, align 4
  %3 = load i32, i32* %2, align 4
  %4 = load i32, i32* %2, align 4
  %5 = mul nsw i32 %3, %4
  ret i32 %5
}

attributes #0 = { noinline nounwind optnone sspstrong uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+cx8,+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }

!llvm.module.flags = !{!0, !1, !2}
!llvm.ident = !{!3}

!0 = !{i32 1, !"wchar_size", i32 4}
!1 = !{i32 7, !"PIC Level", i32 2}
!2 = !{i32 7, !"PIE Level", i32 2}
!3 = !{!"clang version 10.0.0 "}

生成的 linked 文件是：

c.ll

; ModuleID = 'c.bc'
source_filename = "llvm-link"
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-pc-linux-gnu"

; Function Attrs: noinline nounwind optnone sspstrong uwtable
define dso_local i32 @main() #0 {
  %1 = alloca i32, align 4
  store i32 0, i32* %1, align 4
  %2 = call i32 @f(i32 1)
  ret i32 %2
}

; Function Attrs: noinline nounwind optnone sspstrong uwtable
define dso_local i32 @f(i32 %0) #0 {
  %2 = alloca i32, align 4
  store i32 %0, i32* %2, align 4
  %3 = load i32, i32* %2, align 4
  %4 = load i32, i32* %2, align 4
  %5 = mul nsw i32 %3, %4
  ret i32 %5
}

attributes #0 = { noinline nounwind optnone sspstrong uwtable "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="all" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="x86-64" "target-features"="+cx8,+fxsr,+mmx,+sse,+sse2,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" }

!llvm.ident = !{!0, !0}
!llvm.module.flags = !{!1, !2, !3}

!0 = !{!"clang version 10.0.0 "}
!1 = !{i32 1, !"wchar_size", i32 4}
!2 = !{i32 7, !"PIC Level", i32 2}
!3 = !{i32 7, !"PIE Level", i32 2}

您现在可以编译 c.bc 和运行生成的可执行文件：

clang c.bc
./a.out  # Exit code 1 because 1*1=1

一个类似的替代方法是将每个位码文件编译成一个目标文件，然后link那些。

LLVM ir如何导入其他模块？

LLVM ir how to import other modules?

llvm

llvm-ir

a.c

b.c

a.ll

b.ll

c.ll