RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-24 19:52:54 +01:00
Code Issues Projects Releases Wiki Activity
f6eab70376
llvm-mirror/test/LTO/no-undefined-puts-when-implemented.ll

41 lines
928 B
LLVM
Raw Normal View History

Module: Don't rename in getOrInsertFunction() During LTO, user-supplied definitions of C library functions often exist. -instcombine uses Module::getOrInsertFunction() to get a handle on library functions (e.g., @puts, when optimizing @printf). Previously, Module::getOrInsertFunction() would rename any matching functions with local linkage, and create a new declaration. In LTO, this is the opposite of desired behaviour, as it skips by the user-supplied version of the library function and creates a new undefined reference which the linker often cannot resolve. After some discussing with Rafael on the list, it looks like it's undesired behaviour. If a consumer actually *needs* this behaviour, we should add new API with a more explicit name. I added two testcases: one specifically for the -instcombine behaviour and one for the LTO flow. <rdar://problem/16165191> llvm-svn: 203513
2014-03-11 00:42:28 +01:00
; RUN: llvm-as <%s >%t1
; RUN: llvm-lto -exported-symbol=_uses_puts -exported-symbol=_uses_printf -o - %t1 | \
Change the default input for llvm-nm to be a.out instead of standard input to match llvm-size and other UNIX systems for their nm(1). Tweak test cases that used llvm-nm with standard input to add a "-" to indicate that and add a test case to check the default of a.out for llvm-nm. llvm-svn: 211529
2014-06-23 22:27:53 +02:00
; RUN: llvm-nm - | \
Module: Don't rename in getOrInsertFunction() During LTO, user-supplied definitions of C library functions often exist. -instcombine uses Module::getOrInsertFunction() to get a handle on library functions (e.g., @puts, when optimizing @printf). Previously, Module::getOrInsertFunction() would rename any matching functions with local linkage, and create a new declaration. In LTO, this is the opposite of desired behaviour, as it skips by the user-supplied version of the library function and creates a new undefined reference which the linker often cannot resolve. After some discussing with Rafael on the list, it looks like it's undesired behaviour. If a consumer actually *needs* this behaviour, we should add new API with a more explicit name. I added two testcases: one specifically for the -instcombine behaviour and one for the LTO flow. <rdar://problem/16165191> llvm-svn: 203513
2014-03-11 00:42:28 +01:00
; RUN: FileCheck %s
; rdar://problem/16165191
; runtime library implementations should not be renamed
target triple = "x86_64-apple-darwin11"
@str = private unnamed_addr constant [13 x i8] c"hello world\0A\00"
; CHECK-NOT: U _puts
; CHECK: T _uses_printf
; CHECK: T _uses_puts
define i32 @uses_puts(i32 %i) {
entry:
%s = call i8* @foo(i32 %i)
%ret = call i32 @puts(i8* %s)
ret i32 %ret
}
define i32 @uses_printf(i32 %i) {
entry:
%s = getelementptr [13 x i8]* @str, i64 0, i64 0
call i32 (i8*, ...)* @printf(i8* %s)
ret i32 0
}
define hidden i32 @printf(i8* readonly nocapture %fmt, ...) {
entry:
%ret = call i32 @bar(i8* %fmt)
ret i32 %ret
}
define hidden i32 @puts(i8* %s) {
entry:
%ret = call i32 @bar(i8* %s)
ret i32 %ret
}
declare i8* @foo(i32)
declare i32 @bar(i8*)
Reference in New Issue Copy Permalink