[alsa-devel] Misuses of ** ? (was Re: [PATCH 1/1] ASoC: Intel: Atom: add a missing star in a memcpy call)

Joe Perches joe at perches.com
Sun Aug 28 23:54:44 CEST 2016


On Sun, 2016-08-28 at 23:40 +0200, Julia Lawall wrote:
> On Sun, 28 Aug 2016, Joe Perches wrote:
> > On Sun, 2016-08-28 at 21:38 +0200, Julia Lawall wrote:
> > > On Sun, 28 Aug 2016, Nicolas Iooss wrote:
> > > > On 28/08/16 19:50, Joe Perches wrote:
> > > > > On Sun, 2016-08-28 at 19:39 +0200, Nicolas Iooss wrote:
> > > > >> In sst_prepare_and_post_msg(), when a response is received in "block",
> > > > >> the following code gets executed:
> > > > >>
> > > > >>     *data = kzalloc(block->size, GFP_KERNEL);
> > > > >>     memcpy(data, (void *) block->data, block->size);
> > > > >
> > > > > Yuck, thanks.
> > > > >
> > > > > Julia, Dan, could cocci or smatch help find any other
> > > > > similar misuses here?
> > []
> > > I tried the following semantic patch, that is quite general, and the fixed
> > > issue was the only report.
> > >
> > > @@
> > > expression x,y,sz;
> > > identifier f,g;
> > > @@
> > >
> > > * *x = f(sz,...);
> > >   ...
> > > * g(x,y,sz);
> >
> > Hi Julia,
> >
> > This would find exactly the same form, but I think
> > the question is are there assignments of a **pp
> > that should have been *pp
> >
> > Something like:
> >
> > @@
> > type P;
> > P **pp;
> > @@
> >
> > * pp = \|\|(..., sizeof(P), ...)

> I didn't get anything for this.  Did you mean for the left hand side of
> the assignment to be pp or *pp?  Is the issue that the type is wrong?

Yes, the issue here is the type may be wrong.

A function passed a ** and assigned like:

    type function foo(type **bar)
    {
    	    ...
    	    bar = baz();
    	    ...
    }

bar is rarely correct and *bar is generally correct.

I suppose the example would have been clearer with something

-	pp = foo;
+	*pp = foo;

Also, any function that calls another function with
implicit casts to void * from a specific type **pp
after an assignment to *pp could be suspect.



More information about the Alsa-devel mailing list