This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[Bug c++/13959] un-init local variable detection enhancement request


------- Additional Comments From d dot binderman at virgin dot net  2004-02-02 10:51 -------
 
>We try very hard to avoid false positives. This is the primary reason why 
>that 'might' is there.

And in doing so, you remove the power of the warning. 

"might" isn't something most users react to.

>Have you looked at the thread Andrew mentioned in the other PR (11182)? 

Yes. Sad how a reasonable enhancement request is still being asked for two
years later.

>If you're still of the 
>same opinion after reading that, then perhaps you can open a new enhancement 
>request 
>specifically for the issue of "certainty".

Sigh. That's exactly what THIS bug report is doing. Making the certain
case more clear to the user. I don't care about the "might" case.

>And not to sound pessimistic, but we have enough code-generation bugs at this 
>point, that 
>enhancements tend to be a function of somebody being interested enough to 
>write the code...

Fair enough.

I thought I made a pretty strong case for fixing this. Dozens of bugs are going 
begging on this. Lots of C newbies think automatic variables are initialised
by default.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13959


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]