[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 15 Jan 2016 13:57:18 -0500
From: Rich Felker <dalias@...c.org>
To: musl@...ts.openwall.com
Subject: Re: [PATCH] fix if_nametoindex return value when socket open
fails
On Fri, Jan 15, 2016 at 09:39:44AM +0000, Ron Yorston wrote:
> The return value of if_nametoindex is unsigned; it should return 0
> on error.
> ---
> src/network/if_nametoindex.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/network/if_nametoindex.c b/src/network/if_nametoindex.c
> index cb6ec05..331413c 100644
> --- a/src/network/if_nametoindex.c
> +++ b/src/network/if_nametoindex.c
> @@ -10,7 +10,7 @@ unsigned if_nametoindex(const char *name)
> struct ifreq ifr;
> int fd, r;
>
> - if ((fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC, 0)) < 0) return -1;
> + if ((fd = socket(AF_UNIX, SOCK_DGRAM|SOCK_CLOEXEC, 0)) < 0) return 0;
> strncpy(ifr.ifr_name, name, sizeof ifr.ifr_name);
> r = ioctl(fd, SIOCGIFINDEX, &ifr);
> __syscall(SYS_close, fd);
> --
> 2.5.0
Looks good. Actually it's unclear from the spec what should happen,
since "No errors are defined":
http://pubs.opengroup.org/onlinepubs/9699919799/functions/if_nametoindex.html
But 0 is specified as the return value for names which do not exist.
It seems unfortunate that applications can't distinguish between
errors and "no such name", but I don't think we can fix this on our
end. Someone should file a bug in POSIX and the resolution should
probably be that if_nametoindex leaves errno unchanged and returns 0
when the name does not exist, and sets errno and returns 0 when an
internal error occurs.
Rich
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
