netdev - Re: [iproute PATCH v2 1/7] nstat: Avoid passing negative fd to fdopen()

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20170822104808.GC27194@orbyte.nwl.cc>
Date:   Tue, 22 Aug 2017 12:48:08 +0200
From:   Phil Sutter <phil@....cc>
To:     Stephen Hemminger <stephen@...workplumber.org>
Cc:     netdev@...r.kernel.org
Subject: Re: [iproute PATCH v2 1/7] nstat: Avoid passing negative fd to
 fdopen()

On Mon, Aug 21, 2017 at 05:23:23PM -0700, Stephen Hemminger wrote:
> On Mon, 21 Aug 2017 19:08:07 +0200
> Phil Sutter <phil@....cc> wrote:
> 
> > Introduce a wrapper which does the sanity checking and returns NULL
> > in case fd is invalid.
> > 
> > Signed-off-by: Phil Sutter <phil@....cc>
> > ---
> >  misc/nstat.c | 15 +++++++++++----
> >  1 file changed, 11 insertions(+), 4 deletions(-)
> > 
> > diff --git a/misc/nstat.c b/misc/nstat.c
> > index 1212b1f2c8128..7cdde75a56e4e 100644
> > --- a/misc/nstat.c
> > +++ b/misc/nstat.c
> > @@ -252,9 +252,16 @@ static void load_ugly_table(FILE *fp)
> >  	}
> >  }
> >  
> > +static FILE *fdopen_null(int fd, const char *mode)
> > +{
> > +	if (fd < 0)
> > +		return NULL;
> > +	return fdopen(fd, mode);
> > +}
> > +
> >  static void load_sctp_snmp(void)
> >  {
> > -	FILE *fp = fdopen(net_sctp_snmp_open(), "r");
> > +	FILE *fp = fdopen_null(net_sctp_snmp_open(), "r");
> >  
> >  	if (fp) {
> >  		load_good_table(fp);
[...]
> 
> Why not just fix it at the source of the open.
> I.e 
> static FILE *generic_proc_open(condt char * env, const char *name)
> {
> ...
> 	return fopen(p, "r");
> }

What should it do? All the load_*() functions are supposed to be
fault-tolerant, so calling abort() or something is not an option.

Thanks, Phil