llvm dev - Oct 2010 - [LLVMdev] lli : external functions and target datalayout

Hi All,

I have a C code:
//////////////////////////////
#include "stdio.h"

int main () {
  putchar('a');
  return 0;
}

llvm-gcc -emit-llvm, I got
////////////////////////////////////////
; ModuleID = 't1.bc'
target datalayout
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
target triple = "i386-pc-linux-gnu"

define i32 @main() nounwind {
entry:
  %retval = alloca i32                            ; <i32*> [#uses=2]
  %0 = alloca i32                                 ; <i32*> [#uses=2]
  %"alloca point" = bitcast i32 0 to i32          ; <i32>
[#uses=0]
  %1 = call i32 @putchar(i32 97) nounwind         ; <i32> [#uses=0]
  store i32 0, i32* %0, align 4
  %2 = load i32* %0, align 4                      ; <i32> [#uses=1]
  store i32 %2, i32* %retval, align 4
  br label %return

return:                                           ; preds = %entry
  %retval1 = load i32* %retval                    ; <i32> [#uses=1]
  ret i32 %retval1
}

declare i32 @putchar(i32)
//////////////////////////////////////////////

Is there a way to tell 'lli' where to load the external function
'putchar'? "lli -force-interpreter t1.bc" reports "LLVM
ERROR: Tried
to execute an unknown external function: i32 (i32)* putchar".

The other question is about
target datalayout
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"

If I use this string to creat a TargetData object directly, I got an
assertion error:

~/llvm-2.6/lib/Target/TargetData.cpp:246: void
llvm::TargetData::setAlignment(llvm::AlignTypeEnum, unsigned char,
unsigned char, uint32_t): Assertion `abi_align <= pref_align &&
"Preferred alignment worse than ABI!"' failed.
Aborted

This is because of the i64:32:64. It seems to be i64:64:64. 'lli' is
able to fix this problem automatically, but why does llvm-gcc output
i64:32:64 rather than i64:64:64?

Thanks
-- 
Jianzhou

Jianzhou Zhao <jianzhou at seas.upenn.edu> writes:

[snip]
> The other question is about
> target datalayout >
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
>
> If I use this string to creat a TargetData object directly, I got an
> assertion error:
>
> ~/llvm-2.6/lib/Target/TargetData.cpp:246: void
> llvm::TargetData::setAlignment(llvm::AlignTypeEnum, unsigned char,
> unsigned char, uint32_t): Assertion `abi_align <= pref_align &&
> "Preferred alignment worse than ABI!"' failed.
> Aborted
>
> This is because of the i64:32:64. It seems to be i64:64:64. 'lli'
is
> able to fix this problem automatically, but why does llvm-gcc output
> i64:32:64 rather than i64:64:64?
FYI: the X86 data layout is hard-coded in
lib/Target/X86/X86TargetMachine.cpp, so that assert is mostly bogus, and
the existence of "target datalayout" IR setting and
Module::setDataLayout are deceiving.

On Sun, Oct 24, 2010 at 1:48 PM, Óscar Fuentes <ofv at wanadoo.es>
wrote:
> Jianzhou Zhao <jianzhou at seas.upenn.edu> writes:
>
> [snip]
>
>> The other question is about
>> target datalayout >>
"e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:32:32"
>>
>> If I use this string to creat a TargetData object directly, I got an
>> assertion error:
>>
>> ~/llvm-2.6/lib/Target/TargetData.cpp:246: void
>> llvm::TargetData::setAlignment(llvm::AlignTypeEnum, unsigned char,
>> unsigned char, uint32_t): Assertion `abi_align <= pref_align
&&
>> "Preferred alignment worse than ABI!"' failed.
>> Aborted
>>
>> This is because of the i64:32:64. It seems to be i64:64:64.
'lli' is
>> able to fix this problem automatically, but why does llvm-gcc output
>> i64:32:64 rather than i64:64:64?
>
> FYI: the X86 data layout is hard-coded in
> lib/Target/X86/X86TargetMachine.cpp, so that assert is mostly bogus, and
Thanks. I should get the TargetData object from the ExecutionEngine class.
> the existence of "target datalayout" IR setting and
> Module::setDataLayout are deceiving.
>
Do you know if the latest release fixed this assertion? I actually
dont quite understand how this assertion affects the program.  Can I
use setDataLayout for other targets?

-- 
Jianzhou