On Mon, Oct 26, 2009 at 7:36 PM, Andrew Haley <aph@xxxxxxxxxx> wrote:
> Umashankar V.K. wrote:
>
>> I have few questions about inbuilt mechanism of gcc/g++ for
>> warning about uninitialized variable.
>>
>>
>> But then I came across the following failure case. This one fails
>> to warn when compiled with any level below 3.
>> With -O3, gcc is able to warn. Can someone tell me which is the
>> corresponding bug logged at bugzilla ?
>
> I don't understand why you think this is a bug in gcc. Inlining will
> always reveal more about the data flow of a program.
Of course, I know this is not a bug. But I didnt choose the term
here. The compiler developers prefer to
call them bugs and track them using bug-zilla.
>
> extern void foo (int *p);
>
> void bar
> {
> int a;
> foo (&a);
>
> printf("%d\n", a);
> }
>
> We could fix gcc so that in every case where we couldn't prove that a variable
> is initialized, we warned. You'd have to put up with a lot of spurious "may
> be used uninitialized" warnings.
>
> Andrew.
>
yes, i know that false positives is a concern for the compiler author.
But sometimes, a guy who is wondering
why the hell his board is crashing at random may not mind going
through a heap of compiler-warnings if he wants to rule
out that possibility. A code-misbehaviour due to uninitialised
variables can be as mysterious as one due to
stack corruption, isnt it ?
shankar
ps - I choose plain text for the messages I post. Still i see
strange characters when my post appears at
http://gcc.gnu.org/ml/gcc-help/
