On Fri, May 15, 2020 at 04:36:17PM +0300, Dan Carpenter wrote:
> On Thu, May 14, 2020 at 10:56:04PM +0200, Luc Van Oostenryck wrote:
> > Not sure if it's related to Dan's problem or not but with the
> > following code:
> >
> > static inline int foo(void)
> > {
> > return 1;
> > }
> >
> > extern int foo(void);
> >
> > int dummy(void)
> > {
> > return foo();
> > }
> >
> > the static definition of foo() and the extern declaration are
> > distinct symbols (in the sense that neither has its sym->same_symbol
> > pointing to the other). As far as I understand, this is correct
> > because they have a different 'scope'. The problem occurs later,
> > when doing the lookup in dummy(): which symbol should be returned?
>
> Yeah. That's it. When I see the call, I want to parse the statements
> so I need the symbol with the implementation.
There must something else too.
In the example here above I added 'extern' to the second declaration.
But in your first example no storage was given:
void nvme_put_ctrl(struct nvme_ctrl *ctrl);'
and in this case, Sparse give it the storage/linkage from the previous
declaration which was 'static'.
So in the case, the second occurent has its ->same_symbol set to the
previous static inline version and it's ->definition points to it too.
So, I think everything is correct here regarding Sparse (the question
of a warning is something else: IMO none should be give for a static
declaration/definition followed by a plain declaration (thus implicitly
static) but well if followed by an extern one. One is also when
a static follow an extern or a plain (implicitly extern).
Doesn't smatch uses ->same_symbol and more importantly ->definition?
Regards,
-- Luc
