From: Chuck Lever <chuck.lever@oracle.com>
To: trond.myklebust@netapp.com, bfields@citi.umich.edu
Cc: linux-nfs@vger.kernel.org
Subject: [PATCH 3/8] SUNRPC: Split portmap unregister API into separate function
Date: Mon, 30 Jun 2008 18:45:45 -0400 [thread overview]
Message-ID: <20080630224545.24887.61618.stgit@ellison.1015granger.net> (raw)
In-Reply-To: <20080630224147.24887.18730.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
Create a separate server-level interface for unregistering RPC services.
The mechanics of and the API for registering and unregistering RPC
services will diverge further as support for IPv6 is added.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
net/sunrpc/svc.c | 71 +++++++++++++++++++++++++++++++++++++++++++++---------
1 files changed, 59 insertions(+), 12 deletions(-)
diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
index d0e7865..a41b163 100644
--- a/net/sunrpc/svc.c
+++ b/net/sunrpc/svc.c
@@ -27,6 +27,8 @@
#define RPCDBG_FACILITY RPCDBG_SVCDSP
+static void svc_unregister(const struct svc_serv *serv);
+
#define svc_serv_is_pooled(serv) ((serv)->sv_function)
/*
@@ -426,9 +428,8 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
spin_lock_init(&pool->sp_lock);
}
-
/* Remove any stale portmap registrations */
- svc_register(serv, 0, 0);
+ svc_unregister(serv);
return serv;
}
@@ -496,8 +497,7 @@ svc_destroy(struct svc_serv *serv)
if (svc_serv_is_pooled(serv))
svc_pool_map_put();
- /* Unregister service with the portmapper */
- svc_register(serv, 0, 0);
+ svc_unregister(serv);
kfree(serv->sv_pools);
kfree(serv);
}
@@ -758,12 +758,10 @@ int
svc_register(struct svc_serv *serv, int proto, unsigned short port)
{
struct svc_program *progp;
- unsigned long flags;
unsigned int i;
int error = 0, dummy;
- if (!port)
- clear_thread_flag(TIF_SIGPENDING);
+ BUG_ON(proto == 0 && port == 0);
for (progp = serv->sv_program; progp; progp = progp->pg_next) {
for (i = 0; i < progp->pg_nvers; i++) {
@@ -791,13 +789,62 @@ svc_register(struct svc_serv *serv, int proto, unsigned short port)
}
}
- if (!port) {
- spin_lock_irqsave(¤t->sighand->siglock, flags);
- recalc_sigpending();
- spin_unlock_irqrestore(¤t->sighand->siglock, flags);
+ return error;
+}
+
+/*
+ * The local rpcbind daemon listens on either only IPv6 or only
+ * IPv4. The kernel can't tell how it's configured.
+ *
+ * However, AF_INET addresses are mapped to AF_INET6 in IPv6-only
+ * configurations, so even an unregistration request on AF_INET
+ * will get to a local rpcbind daemon listening only on AF_INET6.
+ *
+ * So we always unregister via AF_INET (the loopback address is
+ * fairly unambiguous anyway).
+ *
+ * At this point we don't need rpcbind version 4 for unregistration:
+ * A v2 UNSET request will clear all transports (netids), addresses,
+ * and address families for [program, version].
+ *
+ * This should allow automatic support for both an all-IPv4 and
+ * an all-IPv6 configuration.
+ */
+static void __svc_unregister(struct svc_program *program, u32 version)
+{
+ int error, boolean;
+
+ error = rpcb_register(program->pg_prog, version, 0, 0, &boolean);
+ dprintk("svc: svc_unregister(%sv%u), error %d, %s\n",
+ program->pg_name, version, error,
+ (boolean ? "succeeded" : "failed"));
+}
+
+/*
+ * All transport protocols and ports for this service are removed from
+ * the local rpcbind database. The result of unregistration is reported
+ * via dprintk for those who want verification of the result, but is
+ * otherwise not important.
+ */
+static void svc_unregister(const struct svc_serv *serv)
+{
+ struct svc_program *program;
+ unsigned long flags;
+ u32 version;
+
+ clear_thread_flag(TIF_SIGPENDING);
+
+ for (program = serv->sv_program; program; program = program->pg_next) {
+ for (version = 0; version < program->pg_nvers; version++) {
+ if (program->pg_vers[version] == NULL)
+ continue;
+ __svc_unregister(program, version);
+ }
}
- return error;
+ spin_lock_irqsave(¤t->sighand->siglock, flags);
+ recalc_sigpending();
+ spin_unlock_irqrestore(¤t->sighand->siglock, flags);
}
/*
next prev parent reply other threads:[~2008-06-30 22:46 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-06-30 22:45 [PATCH 0/8] rpcbind v4 support in net/sunrpc/svc* Chuck Lever
[not found] ` <20080630224147.24887.18730.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-06-30 22:45 ` [PATCH 1/8] SUNRPC: Add address family field to svc_serv data structure Chuck Lever
[not found] ` <20080630224529.24887.47412.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-03 21:14 ` J. Bruce Fields
2008-07-04 13:45 ` Chuck Lever
2008-06-30 22:45 ` [PATCH 2/8] SUNRPC: Use proper INADDR_ANY when setting up RPC services on IPv6 Chuck Lever
2008-06-30 22:45 ` Chuck Lever [this message]
[not found] ` <20080630224545.24887.61618.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-18 23:21 ` [PATCH 3/8] SUNRPC: Split portmap unregister API into separate function J. Bruce Fields
2008-07-21 3:17 ` Chuck Lever
[not found] ` <76bd70e30807202017hec9d1der1bbbf5c5dcedac45-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-07-23 16:58 ` J. Bruce Fields
2008-06-30 22:45 ` [PATCH 4/8] SUNRPC: Clean up svc_register Chuck Lever
[not found] ` <20080630224553.24887.73617.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-18 23:29 ` J. Bruce Fields
2008-07-21 19:24 ` Chuck Lever
2008-06-30 22:46 ` [PATCH 5/8] SUNRPC: Use new rpcb_v4_register() interface in svc_register() Chuck Lever
[not found] ` <20080630224601.24887.59241.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-18 23:40 ` J. Bruce Fields
2008-07-21 19:26 ` Chuck Lever
2008-06-30 22:46 ` [PATCH 6/8] SUNRPC: Add kernel build option to disable server-side use of rpcbind v3/v4 Chuck Lever
[not found] ` <20080630224609.24887.20585.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-18 23:42 ` J. Bruce Fields
2008-07-21 19:30 ` Chuck Lever
[not found] ` <76bd70e30807211230y4b7c2b21qa89d8cca05e08dab-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-07-30 16:18 ` J. Bruce Fields
2008-06-30 22:46 ` [PATCH 7/8] SUNRPC: Set V6ONLY socket option for RPC listener sockets Chuck Lever
[not found] ` <20080630224616.24887.13171.stgit-ewv44WTpT0t9HhUboXbp9zCvJB+x5qRC@public.gmane.org>
2008-07-19 1:05 ` J. Bruce Fields
2008-07-21 19:32 ` Chuck Lever
2008-06-30 22:46 ` [PATCH 8/8] NFS: Enable NFSv4 callback server to listen on AF_INET6 sockets Chuck Lever
2008-07-19 1:07 ` [PATCH 0/8] rpcbind v4 support in net/sunrpc/svc* J. Bruce Fields
2008-07-20 21:17 ` J. Bruce Fields
2008-07-21 19:07 ` Chuck Lever
[not found] ` <76bd70e30807211207q4fc509e0h4a1a560fe8097de7-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-07-23 21:53 ` J. Bruce Fields
2008-07-23 22:47 ` Chuck Lever
[not found] ` <76bd70e30807231547j19e9fd8dv7a14c2795226dcd6-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2008-07-23 23:05 ` Trond Myklebust
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=20080630224545.24887.61618.stgit@ellison.1015granger.net \
--to=chuck.lever@oracle.com \
--cc=bfields@citi.umich.edu \
--cc=linux-nfs@vger.kernel.org \
--cc=trond.myklebust@netapp.com \
/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