Re: better -Wuninitialized (Re: Ada files now checked in)

[Daniel Berlin <dan at cgsoftware dot com>]

On Sunday, October 7, 2001, at 01:53  PM, Zack Weinberg wrote:

> On Sun, Oct 07, 2001 at 01:21:01PM -0400, Diego Novillo wrote:
>> On Sun, 07 Oct 2001, Zack Weinberg wrote:
>>> Maybe you could change the warning message to "...used uninitialized
>>> here" or "at this point" or something like that?  Just to be clear
>>> it's not the same old "somewhere in the function" message.
>>>
>> Sure.
>
> Thanks.
>
>>> I notice it says "_is_ used uninitialized".  Does that mean you've
>>> eliminated the old false positive problems?
>>>
>> That's the intent.  The new code still spews out false positives,
>> that's what I'm now working on.
>>
>> The analysis depends on the computation of reaching definitions.
>> Prior to computing the SSA form, we insert ghost definitions for
>> every symbol in the entry basic block.  After reaching
>> definitions, we traverse all the variable use references in the
>> function.  For every use:
>>
>> - if its only reaching definition is the ghost def, the variable
>>   *is* used uninitialized.
>>
>> - if one of its reaching definitions is the ghost def, the
>>   variable *may be* used uninitialized.
> ...
>
> I'm not too familiar with reaching definitions, do they take control
> dependencies into account?
>
> It would often be helpful if an uninitialized variable could be
> automatically set to a "poison" value by the compiler.  This would
> prevent one major cause of hard-to-find context-dependent bugs.  It
> sounds like this can easily be implemented by emitting real code for
> the ghost definitions; dead code elimination would then zap it in all
> cases where there isn't a problem.  Have you considered this?
>
>> Also, I'm about to add def-def chains to model non-killing
>> definitions like:
>>
>> 1: int a, b *p;
>> 2:
>> 3: a = 4;
>> 4: *p = 3;
>> 5: b = a + 1;
>>
>> The use of a at line 5 may be reached by the definitions of *p
>> and a at lines 4 and 3, respectively.  But this part is nowhere
>> near ready.
>
> Hmmm... since p itself is not initialized, it seems like you'd want to
> complain about it and then assume it doesn't alias anything.
>
>> - Compute the SSA form.  This involves computing immediate
>>   dominators and dominance frontiers.  I believe the algorithms
>>   we have in GCC are quite quick, but I haven't really looked.
>
> If I remember correctly we are using the state-of-the-art algorithm,
Not anymore, at least for phi placement.
Dominance calculation is as fast as it's gonna get without using *much* 
more complex datastructures (microtrees and whatnot).

> but its use of sbitmaps may cause problems.  (looking at ssa.c - dunno
> if the same code is used for trees).
>