[llvm-dev] opt -instcombine interesting behavior

[Stephan Plöger via llvm-dev]

Hi all,

I've been playing around with the llvm optimizer and stumbled upon an
interesting behavior with respect
to the "instcombine" option. I tried to optimize the following small
program (see also attachment small.ll)
(only with "instcombine"):

clang -v
clang version 4.0.0 (trunk 288238)
Target: x86_64-unknown-linux-gnu
Thread model: posix
InstalledDir: /usr/local/bin
Found candidate GCC installation: /usr/lib/gcc/i686-linux-gnu/5
Found candidate GCC installation: /usr/lib/gcc/i686-linux-gnu/5.4.0
Found candidate GCC installation: /usr/lib/gcc/i686-linux-gnu/6
Found candidate GCC installation: /usr/lib/gcc/i686-linux-gnu/6.0.0
Found candidate GCC installation: /usr/lib/gcc/x86_64-linux-gnu/5
Found candidate GCC installation: /usr/lib/gcc/x86_64-linux-gnu/5.4.0
Found candidate GCC installation: /usr/lib/gcc/x86_64-linux-gnu/6
Found candidate GCC installation: /usr/lib/gcc/x86_64-linux-gnu/6.0.0
Selected GCC installation: /usr/lib/gcc/x86_64-linux-gnu/5.4.0
Candidate multilib: .;@m64
Candidate multilib: 32;@m32
Candidate multilib: x32;@mx32
Selected multilib: .;@m64


llvm-as small.ll -o small.bc
opt -instcombine small.bc -o smallOpt.bc


define i32 @main() {
  %reg = alloca i32, i32 0
  store i32 1, i32* %reg
  %1 = getelementptr i32, i32* %reg, i32 0
  %2 = bitcast i32* %1 to [4 x i8]*
  %3 = getelementptr [4 x i8], [4 x i8]* %2, i32 0, i32 1
  store i8 1, i8* %3
  %4 = load i32, i32* %reg
  ret i32 %4
}

The program in pseudo code would/should simplified look something like this:

int main(){
    reg = 1;
    [reg+1] = 1; //Second byte of reg is set to 1
    return reg;
}

I assumed it would be optimized to - again pseudo code -:

int main(){
    return 257;
}

Instead, the program is optimized to(see also attachment
smallOpt.ll/smallOpt.bc):

define i32 @main() {
  ret i32 1
}

So i verified my assumption by printing '%4' of the original program
(smallPrint.ll/smallPrint.bc) and the output was 257.
I also optimized smallPrint.ll/smallPrint.bc , again only with
instcombine, but the output is changed to 1 (smallPrintOpt.bc).

I also tested it with clang 3.8.0.

What am i missing?


Best regards,
Stephan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smallPrintOpt.bc
Type: application/octet-stream
Size: 868 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20161219/421fcb23/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smallPrint.bc
Type: application/octet-stream
Size: 912 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20161219/421fcb23/attachment-0001.obj>
-------------- next part --------------
@.str = private unnamed_addr constant [4 x i8] c"%d\0A\00"



declare i32 @printf(i8*, ...) #1


define i32 @main() {
  %reg = alloca i32, i32 0
  store i32 1, i32* %reg
  %1 = getelementptr i32, i32* %reg, i32 0
  %2 = bitcast i32* %1 to [4 x i8]*
  %3 = getelementptr [4 x i8], [4 x i8]* %2, i32 0, i32 1
  store i8 1, i8* %3
  %4 = load i32, i32* %reg
  %call = call i32 (i8*, ...) @printf(i8* getelementptr inbounds ([4 x i8], [4 x i8]* @.str, i32 0, i32 0), i32 %4)
  ret i32 %4
}
-------------- next part --------------
; ModuleID = 'smallOpt.bc'
source_filename = "small.ll"

define i32 @main() {
  ret i32 1
}
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smallOpt.bc
Type: application/octet-stream
Size: 736 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20161219/421fcb23/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: small.bc
Type: application/octet-stream
Size: 796 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20161219/421fcb23/attachment-0003.obj>
-------------- next part --------------
define i32 @main() {
  %reg = alloca i32, i32 0
  store i32 1, i32* %reg
  %1 = getelementptr i32, i32* %reg, i32 0
  %2 = bitcast i32* %1 to [4 x i8]*
  %3 = getelementptr [4 x i8], [4 x i8]* %2, i32 0, i32 1
  store i8 1, i8* %3
  %4 = load i32, i32* %reg
  ret i32 %4
}