[llvm-commits] proposed patch: shift when shift amount >= bit width

[Chris Lattner]

On Jun 13, 2007, at 11:34 AM, Lauro Ramos Venancio wrote:

>>
>> This is not enough to make a compiler handle shifts by greater  
>> than the
>> word size consistently; in both gcc and llvm-gcc, for example, the
>> result of
>>
>>     int y=32;
>>     x>>y
>>
>> depends on optimization level.
>
> I know. On x86, the result of
> int n = 32;
> 1<<n
>
> is 1 with "gcc -O0" and 0 with "gcc -O3". But some important programs
> (like ffmpeg) use these shifts expecting the gcc behavior. The ffmpeg,
> for example, works well with "gcc -O0" and "gcc -O3". So, for
> compatibility purpose, I think we should implement the gcc behavior
> and do not create another behavior.

I don't think this is appropriate Lauro.  The program needs to be  
fixed, it's not just a codegen or optimizer issue, everything in the  
compiler assumes that these shifts are undefined.  Programs cannot  
reliably expect a compiler to do the right thing here, and your patch  
isn't a fix.

-Chris