From: Steve Dickson <steved@redhat.com>
To: NeilBrown <neilb@suse.de>
Cc: linux-nfs@vger.kernel.org, Petr Vorel <pvorel@suse.cz>
Subject: Re: [PATCH 2/3] Change local_rpcb() to take a targaddr pointer.
Date: Sat, 9 Mar 2024 09:27:10 -0500 [thread overview]
Message-ID: <52cc6a79-7c7c-47ec-b1f0-cc0de61fafab@redhat.com> (raw)
In-Reply-To: <20240225234337.19744-3-neilb@suse.de>
On 2/25/24 6:40 PM, NeilBrown wrote:
> Two callers of local_rpcb() want the target-addr, and local_rcpb() has
> easy access to it. So accept a pointer and fill it in if not NULL.
>
> Signed-off-by: NeilBrown <neilb@suse.de>
> ---
> src/rpcb_clnt.c | 35 +++++++++++------------------------
> 1 file changed, 11 insertions(+), 24 deletions(-)
>
> diff --git a/src/rpcb_clnt.c b/src/rpcb_clnt.c
> index 68fe69a320ff..f587580228ab 100644
> --- a/src/rpcb_clnt.c
> +++ b/src/rpcb_clnt.c
> @@ -89,7 +89,7 @@ static struct address_cache *copy_of_cached(const char *, char *);
> static void delete_cache(struct netbuf *);
> static void add_cache(const char *, const char *, struct netbuf *, char *);
> static CLIENT *getclnthandle(const char *, const struct netconfig *, char **);
> -static CLIENT *local_rpcb(void);
> +static CLIENT *local_rpcb(char **targaddr);
> #ifdef NOTUSED
> static struct netbuf *got_entry(rpcb_entry_list_ptr, const struct netconfig *);
> #endif
> @@ -430,19 +430,12 @@ getclnthandle(host, nconf, targaddr)
> nconf->nc_netid, si.si_af, si.si_proto, si.si_socktype));
>
> if (nconf->nc_protofmly != NULL && strcmp(nconf->nc_protofmly, NC_LOOPBACK) == 0) {
> - client = local_rpcb();
> + client = local_rpcb(targaddr);
> if (! client) {
> LIBTIRPC_DEBUG(1, ("getclnthandle: %s",
> clnt_spcreateerror("local_rpcb failed")));
> goto out_err;
> } else {
> - struct sockaddr_un sun;
> -
> - if (targaddr) {
> - *targaddr = malloc(sizeof(sun.sun_path));
> - strncpy(*targaddr, _PATH_RPCBINDSOCK,
> - sizeof(sun.sun_path));
> - }
> return (client);
> }
> } else {
> @@ -541,7 +534,8 @@ getpmaphandle(nconf, hostname, tgtaddr)
> * rpcbind. Returns NULL on error and free's everything.
> */
> static CLIENT *
> -local_rpcb()
> +local_rpcb(targaddr)
> + char **targaddr;
> {
> CLIENT *client;
> static struct netconfig *loopnconf;
> @@ -574,6 +568,8 @@ local_rpcb()
> if (client != NULL) {
> /* Mark the socket to be closed in destructor */
> (void) CLNT_CONTROL(client, CLSET_FD_CLOSE, NULL);
> + if (targaddr)
> + *targaddr = strdup(sun.sun_path);
> return client;
> }
>
> @@ -632,7 +628,7 @@ try_nconf:
> endnetconfig(nc_handle);
> }
> mutex_unlock(&loopnconf_lock);
> - client = getclnthandle(hostname, loopnconf, NULL);
> + client = getclnthandle(hostname, loopnconf, targaddr);
> return (client);
> }
>
> @@ -661,20 +657,11 @@ rpcb_set(program, version, nconf, address)
> rpc_createerr.cf_stat = RPC_UNKNOWNADDR;
> return (FALSE);
> }
> - client = local_rpcb();
> + client = local_rpcb(&parms.r_addr);
> if (! client) {
> return (FALSE);
> }
>
> - /* convert to universal */
> - /*LINTED const castaway*/
> - parms.r_addr = taddr2uaddr((struct netconfig *) nconf,
> - (struct netbuf *)address);
> - if (!parms.r_addr) {
> - CLNT_DESTROY(client);
> - rpc_createerr.cf_stat = RPC_N2AXLATEFAILURE;
> - return (FALSE); /* no universal address */
> - }
> parms.r_prog = program;
> parms.r_vers = version;
> parms.r_netid = nconf->nc_netid;
> @@ -712,7 +699,7 @@ rpcb_unset(program, version, nconf)
> RPCB parms;
> char uidbuf[32];
>
> - client = local_rpcb();
> + client = local_rpcb(NULL);
> if (! client) {
> return (FALSE);
> }
> @@ -1342,7 +1329,7 @@ rpcb_taddr2uaddr(nconf, taddr)
> rpc_createerr.cf_stat = RPC_UNKNOWNADDR;
> return (NULL);
> }
> - client = local_rpcb();
> + client = local_rpcb(NULL);
> if (! client) {
> return (NULL);
> }
> @@ -1376,7 +1363,7 @@ rpcb_uaddr2taddr(nconf, uaddr)
> rpc_createerr.cf_stat = RPC_UNKNOWNADDR;
> return (NULL);
> }
> - client = local_rpcb();
> + client = local_rpcb(NULL);
> if (! client) {
> return (NULL);
> }
It is not clear why... but this patch stop mountd from
registering with rpcbind (both the old and changed via
the latest patches), which means v3 mounts break.
Not good :-)
Turning debugging on... rpcbind is receiving the set prog
but not recording it, since port 0 is returned when
the client tries to do a v3 mount.
Are you guys seeing this??
I remove this patch and everything works!
steved.
next prev parent reply other threads:[~2024-03-09 14:27 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-02-25 23:40 [PATCH 0/3 libtirpc] Support abstract addresses for rpcbind in libtirpc NeilBrown
2024-02-25 23:40 ` [PATCH 1/3] Allow working with abstract AF_UNIX addresses NeilBrown
2024-02-25 23:40 ` [PATCH 2/3] Change local_rpcb() to take a targaddr pointer NeilBrown
2024-03-09 14:27 ` Steve Dickson [this message]
2024-03-11 1:17 ` NeilBrown
2024-03-11 9:35 ` Petr Vorel
2024-02-25 23:40 ` [PATCH 3/3] Try using a new abstract address when connecting rpcbind NeilBrown
-- strict thread matches above, loose matches on Subject: below --
2024-03-11 1:41 [PATCH 0/3 libtirpc v2] Support abstract addresses for rpcbind in libtirpc NeilBrown
2024-03-11 1:41 ` [PATCH 2/3] Change local_rpcb() to take a targaddr pointer NeilBrown
2023-05-10 22:20 [PATCH 0/3] Support abstract addresses for rpcbind in libtirpc NeilBrown
2023-05-10 22:20 ` [PATCH 2/3] Change local_rpcb() to take a targaddr pointer NeilBrown
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=52cc6a79-7c7c-47ec-b1f0-cc0de61fafab@redhat.com \
--to=steved@redhat.com \
--cc=linux-nfs@vger.kernel.org \
--cc=neilb@suse.de \
--cc=pvorel@suse.cz \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).