[LLVMdev] gmake check failures

Nate Begeman natebegeman at mac.com
Sat Jul 30 20:37:29 PDT 2005


I'll attempt to deal with this tomorrow.  Probably just a missed case  
in the new LSR implementation.

Nate

On Jul 30, 2005, at 2:48 PM, Jeff Cohen wrote:

> The only non-empty output file I can find is ops_after_indvar.ll.out:
>
> Instruction does not dominate all uses!
>        %idx = cast int %idx to uint            ; <uint> [#uses=1]
>        %tmp. = mul uint %idx, 4                ; <uint> [#uses=1]
> Instruction does not dominate all uses!
>        %tmp. = mul uint %idx, 4                ; <uint> [#uses=1]
>        %tmp.1 = add uint %P, %tmp.             ; <uint> [#uses=1]
> Instruction does not dominate all uses!
>        %tmp.1 = add uint %P, %tmp.             ; <uint> [#uses=1]
>        %tmp.1 = cast uint %tmp.1 to int*               ; <int*>  
> [#uses=1]
> Broken module found, compilation aborted!
> llvm-dis: Standard Input empty!
> Abort trap (core dumped)
>
> No out file at all was generated for the two failing tests.  The other  
> out files in Regression/Transforms/LoopStrengthReduce/Output are  
> empty.
>
> Reid Spencer wrote:
>
>> No error messages with these failures? Can you look into the output
>> files?
>>
>> Reid.
>>
>> On Sat, 2005-07-30 at 14:04 -0700, Jeff Cohen wrote:
>>
>>> FAIL:  
>>> /usr/home/jeffc/llvm/obj/../test/Regression/Transforms/ 
>>> LoopStrengthReduce/dont_insert_redundant_ops.ll:
>>>
>>> FAIL:  
>>> /usr/home/jeffc/llvm/obj/../test/Regression/Transforms/ 
>>> LoopStrengthReduce/dont_reduce_bytes.ll:
>>>
>>>
>>> _______________________________________________
>>> LLVM Developers mailing list
>>> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
>>>
>
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev




More information about the llvm-dev mailing list