|
On Fri, Jan 17, 2014 at 11:03:43PM +0200, Julian Anastasov wrote:
>
> Hello,
>
> On Fri, 17 Jan 2014, Ryan O'Hara wrote:
>
> > The ipvs_get_service function declares an ipvs_service_t type variable
> > and initializes some of the values, but should really start by
> > initializing the entire structure.
> >
> > Signed-off-by: Ryan O'Hara <rohara@xxxxxxxxxx>
> > ---
> > libipvs/libipvs.c | 1 +
> > 1 file changed, 1 insertion(+)
> >
> > diff --git a/libipvs/libipvs.c b/libipvs/libipvs.c
> > index d2fec49..8baafed 100644
> > --- a/libipvs/libipvs.c
> > +++ b/libipvs/libipvs.c
> > @@ -942,6 +942,7 @@ ipvs_get_service(__u32 fwmark, __u16 af, __u16
> > protocol, union nf_inet_addr addr
> > if (!svc)
> > return NULL;
> >
> > + memset(&tsvc, 0, sizeof(tsvc));
> > tsvc.fwmark = fwmark;
> > tsvc.af = af;
> > tsvc.protocol= protocol;
> > --
> > 1.8.1.4
>
> Yes, it is a good idea. Another variant is to
> change malloc to calloc. May be ipvs_nl_fill_service_attr()
> can not crash in all cases when reading svc->pe_name[].
> As for the kernel part, ip_vs_genl_parse_service() uses just
> the initialized fields when full_entry=0 for IPVS_CMD_GET_SERVICE
> but it is better to avoid problems in the future.
Right. This 'bug' was found while running the code through a static
analyzer, which finds all sorts of nit-picks. Since the variable in
question here is on the stack, a simple memset seems sufficient. Sorry
for trivial patches, just trying to get rid of the complaints from the
analyzer.
Ryan
> Regards
>
> --
> Julian Anastasov <ja@xxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe lvs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
|
