Re: [PATCH] Add -Womitted-conditional-op warning

[Andi Kleen <ak at suse dot de>]

On Friday 02 February 2007 18:57, Joseph S. Myers wrote:
> On Fri, 2 Feb 2007, Andi Kleen wrote:
> 
> > I implemented a warning for this case. I guess it's relatively obscure, but will
> > not hurt. The warning is enabled by default, but can be disabled with
> > -Wno-omitted-conditional-op.
> 
> This extension seems like one that could be used in macros in system 
> headers and so maybe the warning should be disabled by __extension__ (for 

That would disable the warning completely for those headers then
because it can only happen in a __extension__ because it relies on a GNU
extension in the first place.  

> C, see disable_extension_diagnostics / restore_extension_diagnostics).  
> Or do you think that the particular cases you warn for are never going to 
> arise from macros (given those could be replaced by ... ? 1 : ...)?

If it arises from macros it was very likely a mistake the macro
writer did (IMNSHO) 

> > +@item -Wno-omitted-conditional-op
> > +@opindex Wno-omitted-conditional-op
> > +Don't warn for dangerous uses of the 
> > +?: with omitted middle operand GNU extension. When the condition
> > +in the ?: operator is a computed boolean the omitted value will
> > +be always 1. Often the user expects it to be a value computed
> > +inside the conditional expression instead. Gcc by default warns 
> > +for this, but this option disables it.
> 
> Two spaces after "." at end of sentence in Texinfo.  "GCC" not "Gcc".  Use 
> @code or @samp around "?:".  Include a cross-reference to the section of 
> the manual documenting this extension.

Thanks, will fix that.

-Andi