- * [iproute2 net-next v2 1/3] netlink: Add flag to suppress print of nlmsg error
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
@ 2017-03-24  2:51 ` David Ahern
  2017-03-24  2:51 ` [iproute2 net-next v2 2/3] ip netconf: Show all address families by default in dumps David Ahern
                   ` (4 subsequent siblings)
  5 siblings, 0 replies; 9+ messages in thread
From: David Ahern @ 2017-03-24  2:51 UTC (permalink / raw)
  To: netdev, stephen; +Cc: nicolas.dichtel, David Ahern
Allow callers of the dump API to handle nlmsg errors (e.g., an
unsupported feature). Setting RTNL_HANDLE_F_SUPPRESS_NLERR in the
rtnl_handle avoids unnecessary messages to the users in some case.
For example,
  RTNETLINK answers: Operation not supported
when probing for support of a new feature.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 include/libnetlink.h | 1 +
 lib/libnetlink.c     | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/include/libnetlink.h b/include/libnetlink.h
index bd0267dfcc02..c43ab0a2d9d9 100644
--- a/include/libnetlink.h
+++ b/include/libnetlink.h
@@ -21,6 +21,7 @@ struct rtnl_handle {
 	int			proto;
 	FILE		       *dump_fp;
 #define RTNL_HANDLE_F_LISTEN_ALL_NSID		0x01
+#define RTNL_HANDLE_F_SUPPRESS_NLERR		0x02
 	int			flags;
 };
 
diff --git a/lib/libnetlink.c b/lib/libnetlink.c
index 9303b6686e2c..5b75b2db4e0b 100644
--- a/lib/libnetlink.c
+++ b/lib/libnetlink.c
@@ -299,7 +299,8 @@ static void rtnl_dump_error(const struct rtnl_handle *rth,
 		     errno == EOPNOTSUPP))
 			return;
 
-		perror("RTNETLINK answers");
+		if (!(rth->flags & RTNL_HANDLE_F_SUPPRESS_NLERR))
+			perror("RTNETLINK answers");
 	}
 }
 
-- 
2.1.4
^ permalink raw reply related	[flat|nested] 9+ messages in thread
- * [iproute2 net-next v2 2/3] ip netconf: Show all address families by default in dumps
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
  2017-03-24  2:51 ` [iproute2 net-next v2 1/3] netlink: Add flag to suppress print of nlmsg error David Ahern
@ 2017-03-24  2:51 ` David Ahern
  2017-03-24  2:51 ` [iproute2 net-next v2 3/3] ip netconf: show all families on dev request David Ahern
                   ` (3 subsequent siblings)
  5 siblings, 0 replies; 9+ messages in thread
From: David Ahern @ 2017-03-24  2:51 UTC (permalink / raw)
  To: netdev, stephen; +Cc: nicolas.dichtel, David Ahern
Currently, 'ip netconf' only shows ipv4 and ipv6 netconf settings. If IPv6
is not enabled, the dump ends with
    RTNETLINK answers: Operation not supported
when IPv6 request is attempted. Further, if the mpls_router module is also
loaded a separate request is needed to get MPLS settings.
To make this better going forward, use the new PF_UNSPEC dump all option
if the kernel supports it. If the kernel does not, it sets NLMSG_ERROR and
returns EOPNOTSUPP which is trapped and we fall back to the existing output
to maintain compatibility with existing kernels.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 ip/ipnetconf.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/ip/ipnetconf.c b/ip/ipnetconf.c
index af539f5e945c..dc0851025223 100644
--- a/ip/ipnetconf.c
+++ b/ip/ipnetconf.c
@@ -19,6 +19,7 @@
 #include <sys/time.h>
 #include <sys/socket.h>
 #include <netinet/in.h>
+#include <errno.h>
 
 #include "rt_names.h"
 #include "utils.h"
@@ -197,16 +198,26 @@ static int do_show(int argc, char **argv)
 		}
 		rtnl_listen(&rth, print_netconf, stdout);
 	} else {
+		rth.flags = RTNL_HANDLE_F_SUPPRESS_NLERR;
 dump:
 		if (rtnl_wilddump_request(&rth, filter.family, RTM_GETNETCONF) < 0) {
 			perror("Cannot send dump request");
 			exit(1);
 		}
 		if (rtnl_dump_filter(&rth, print_netconf2, stdout) < 0) {
+			/* kernel does not support netconf dump on AF_UNSPEC;
+			 * fall back to requesting by family
+			 */
+			if (errno == EOPNOTSUPP &&
+			    filter.family == AF_UNSPEC) {
+				filter.family = AF_INET;
+				goto dump;
+			}
+			perror("RTNETLINK answers");
 			fprintf(stderr, "Dump terminated\n");
 			exit(1);
 		}
-		if (preferred_family == AF_UNSPEC) {
+		if (preferred_family == AF_UNSPEC && filter.family == AF_INET) {
 			preferred_family = AF_INET6;
 			filter.family = AF_INET6;
 			goto dump;
-- 
2.1.4
^ permalink raw reply related	[flat|nested] 9+ messages in thread
- * [iproute2 net-next v2 3/3] ip netconf: show all families on dev request
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
  2017-03-24  2:51 ` [iproute2 net-next v2 1/3] netlink: Add flag to suppress print of nlmsg error David Ahern
  2017-03-24  2:51 ` [iproute2 net-next v2 2/3] ip netconf: Show all address families by default in dumps David Ahern
@ 2017-03-24  2:51 ` David Ahern
  2017-03-24  8:37 ` [iproute2 net-next v2 0/3] ip netconf improvements Nicolas Dichtel
                   ` (2 subsequent siblings)
  5 siblings, 0 replies; 9+ messages in thread
From: David Ahern @ 2017-03-24  2:51 UTC (permalink / raw)
  To: netdev, stephen; +Cc: nicolas.dichtel, David Ahern
Currently specifying a device to ip netconf and it dumps only values
for IPv4. Change this to dump data for all families unless a specific
family is given.
Signed-off-by: David Ahern <dsa@cumulusnetworks.com>
---
 ip/ipnetconf.c | 23 +++++++++++++----------
 1 file changed, 13 insertions(+), 10 deletions(-)
diff --git a/ip/ipnetconf.c b/ip/ipnetconf.c
index dc0851025223..696e3dd51a1c 100644
--- a/ip/ipnetconf.c
+++ b/ip/ipnetconf.c
@@ -56,6 +56,7 @@ int print_netconf(const struct sockaddr_nl *who, struct rtnl_ctrl_data *ctrl,
 	struct netconfmsg *ncm = NLMSG_DATA(n);
 	int len = n->nlmsg_len;
 	struct rtattr *tb[NETCONFA_MAX+1];
+	int ifindex = 0;
 
 	if (n->nlmsg_type == NLMSG_ERROR)
 		return -1;
@@ -77,6 +78,12 @@ int print_netconf(const struct sockaddr_nl *who, struct rtnl_ctrl_data *ctrl,
 	parse_rtattr(tb, NETCONFA_MAX, netconf_rta(ncm),
 		     NLMSG_PAYLOAD(n, sizeof(*ncm)));
 
+	if (tb[NETCONFA_IFINDEX])
+		ifindex = rta_getattr_u32(tb[NETCONFA_IFINDEX]);
+
+	if (filter.ifindex && filter.ifindex != ifindex)
+		return 0;
+
 	switch (ncm->ncm_family) {
 	case AF_INET:
 		fprintf(fp, "ipv4 ");
@@ -93,9 +100,7 @@ int print_netconf(const struct sockaddr_nl *who, struct rtnl_ctrl_data *ctrl,
 	}
 
 	if (tb[NETCONFA_IFINDEX]) {
-		int *ifindex = (int *)rta_getattr_str(tb[NETCONFA_IFINDEX]);
-
-		switch (*ifindex) {
+		switch (ifindex) {
 		case NETCONFA_IFINDEX_ALL:
 			fprintf(fp, "all ");
 			break;
@@ -103,7 +108,7 @@ int print_netconf(const struct sockaddr_nl *who, struct rtnl_ctrl_data *ctrl,
 			fprintf(fp, "default ");
 			break;
 		default:
-			fprintf(fp, "dev %s ", ll_index_to_name(*ifindex));
+			fprintf(fp, "dev %s ", ll_index_to_name(ifindex));
 			break;
 		}
 	}
@@ -169,8 +174,6 @@ static int do_show(int argc, char **argv)
 
 	ipnetconf_reset_filter(0);
 	filter.family = preferred_family;
-	if (filter.family == AF_UNSPEC)
-		filter.family = AF_INET;
 
 	while (argc > 0) {
 		if (strcmp(*argv, "dev") == 0) {
@@ -186,11 +189,11 @@ static int do_show(int argc, char **argv)
 	}
 
 	ll_init_map(&rth);
-	if (filter.ifindex) {
+
+	if (filter.ifindex && filter.family != AF_UNSPEC) {
 		req.ncm.ncm_family = filter.family;
-		if (filter.ifindex)
-			addattr_l(&req.n, sizeof(req), NETCONFA_IFINDEX,
-				  &filter.ifindex, sizeof(filter.ifindex));
+		addattr_l(&req.n, sizeof(req), NETCONFA_IFINDEX,
+			  &filter.ifindex, sizeof(filter.ifindex));
 
 		if (rtnl_send(&rth, &req.n, req.n.nlmsg_len) < 0) {
 			perror("Can not send request");
-- 
2.1.4
^ permalink raw reply related	[flat|nested] 9+ messages in thread
- * Re: [iproute2 net-next v2 0/3] ip netconf improvements
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
                   ` (2 preceding siblings ...)
  2017-03-24  2:51 ` [iproute2 net-next v2 3/3] ip netconf: show all families on dev request David Ahern
@ 2017-03-24  8:37 ` Nicolas Dichtel
  2017-04-04 21:07 ` David Ahern
  2017-04-14 23:02 ` Stephen Hemminger
  5 siblings, 0 replies; 9+ messages in thread
From: Nicolas Dichtel @ 2017-03-24  8:37 UTC (permalink / raw)
  To: David Ahern, netdev, stephen
Le 24/03/2017 à 03:51, David Ahern a écrit :
> Currently, ip netconf only shows data for ipv4 and ipv6 for dumps
> and just ipv4 for device requests. Improve the user experience by
> using the new kernel patch to dump all address families that have
> registered. For example, if mpls_router module is loaded then mpls
> values are displayed along with ipv4 and ipv6.
> 
> If the new feature is not supported (new iproute2 on older kernel)
> the kernel returns the nlmsg error EOPNOTSUPP which can be trapped
> and fallback to existing behavior.
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
^ permalink raw reply	[flat|nested] 9+ messages in thread
- * Re: [iproute2 net-next v2 0/3] ip netconf improvements
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
                   ` (3 preceding siblings ...)
  2017-03-24  8:37 ` [iproute2 net-next v2 0/3] ip netconf improvements Nicolas Dichtel
@ 2017-04-04 21:07 ` David Ahern
  2017-04-04 21:39   ` Stephen Hemminger
  2017-04-14 23:02 ` Stephen Hemminger
  5 siblings, 1 reply; 9+ messages in thread
From: David Ahern @ 2017-04-04 21:07 UTC (permalink / raw)
  To: netdev, stephen; +Cc: nicolas.dichtel
On 3/23/17 10:51 PM, David Ahern wrote:
> Currently, ip netconf only shows data for ipv4 and ipv6 for dumps
> and just ipv4 for device requests. Improve the user experience by
> using the new kernel patch to dump all address families that have
> registered. For example, if mpls_router module is loaded then mpls
> values are displayed along with ipv4 and ipv6.
> 
> If the new feature is not supported (new iproute2 on older kernel)
> the kernel returns the nlmsg error EOPNOTSUPP which can be trapped
> and fallback to existing behavior.
> 
> v2
> - fixed index conversion in patch 3 per nicholas' comment
> 
> David Ahern (3):
>   netlink: Add flag to suppress print of nlmsg error
>   ip netconf: Show all address families by default in dumps
>   ip netconf: show all families on dev request
> 
>  include/libnetlink.h |  1 +
>  ip/ipnetconf.c       | 36 +++++++++++++++++++++++++-----------
>  lib/libnetlink.c     |  3 ++-
>  3 files changed, 28 insertions(+), 12 deletions(-)
> 
Hi Stephen: any comments? are you ok with this change?
^ permalink raw reply	[flat|nested] 9+ messages in thread
- * Re: [iproute2 net-next v2 0/3] ip netconf improvements
  2017-04-04 21:07 ` David Ahern
@ 2017-04-04 21:39   ` Stephen Hemminger
  2017-04-05 12:18     ` David Ahern
  0 siblings, 1 reply; 9+ messages in thread
From: Stephen Hemminger @ 2017-04-04 21:39 UTC (permalink / raw)
  To: David Ahern; +Cc: netdev, nicolas.dichtel
On Tue, 4 Apr 2017 17:07:31 -0400
David Ahern <dsa@cumulusnetworks.com> wrote:
> On 3/23/17 10:51 PM, David Ahern wrote:
> > Currently, ip netconf only shows data for ipv4 and ipv6 for dumps
> > and just ipv4 for device requests. Improve the user experience by
> > using the new kernel patch to dump all address families that have
> > registered. For example, if mpls_router module is loaded then mpls
> > values are displayed along with ipv4 and ipv6.
> > 
> > If the new feature is not supported (new iproute2 on older kernel)
> > the kernel returns the nlmsg error EOPNOTSUPP which can be trapped
> > and fallback to existing behavior.
> > 
> > v2
> > - fixed index conversion in patch 3 per nicholas' comment
> > 
> > David Ahern (3):
> >   netlink: Add flag to suppress print of nlmsg error
> >   ip netconf: Show all address families by default in dumps
> >   ip netconf: show all families on dev request
> > 
> >  include/libnetlink.h |  1 +
> >  ip/ipnetconf.c       | 36 +++++++++++++++++++++++++-----------
> >  lib/libnetlink.c     |  3 ++-
> >  3 files changed, 28 insertions(+), 12 deletions(-)
> >   
> 
> Hi Stephen: any comments? are you ok with this change?
I was holding off until all the upstream commits went through. Other than
that fine.
^ permalink raw reply	[flat|nested] 9+ messages in thread 
- * Re: [iproute2 net-next v2 0/3] ip netconf improvements
  2017-04-04 21:39   ` Stephen Hemminger
@ 2017-04-05 12:18     ` David Ahern
  0 siblings, 0 replies; 9+ messages in thread
From: David Ahern @ 2017-04-05 12:18 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: netdev, nicolas.dichtel
On 4/4/17 5:39 PM, Stephen Hemminger wrote:
> On Tue, 4 Apr 2017 17:07:31 -0400
> David Ahern <dsa@cumulusnetworks.com> wrote:
> 
>> On 3/23/17 10:51 PM, David Ahern wrote:
>>> Currently, ip netconf only shows data for ipv4 and ipv6 for dumps
>>> and just ipv4 for device requests. Improve the user experience by
>>> using the new kernel patch to dump all address families that have
>>> registered. For example, if mpls_router module is loaded then mpls
>>> values are displayed along with ipv4 and ipv6.
>>>
>>> If the new feature is not supported (new iproute2 on older kernel)
>>> the kernel returns the nlmsg error EOPNOTSUPP which can be trapped
>>> and fallback to existing behavior.
>>>
>>> v2
>>> - fixed index conversion in patch 3 per nicholas' comment
>>>
>>> David Ahern (3):
>>>   netlink: Add flag to suppress print of nlmsg error
>>>   ip netconf: Show all address families by default in dumps
>>>   ip netconf: show all families on dev request
>>>
>>>  include/libnetlink.h |  1 +
>>>  ip/ipnetconf.c       | 36 +++++++++++++++++++++++++-----------
>>>  lib/libnetlink.c     |  3 ++-
>>>  3 files changed, 28 insertions(+), 12 deletions(-)
>>>   
>>
>> Hi Stephen: any comments? are you ok with this change?
> 
> I was holding off until all the upstream commits went through. Other than
> that fine.
> 
I'm not aware of any kernel commits that not in net-next, so I think we
good to go on the kernel side.
^ permalink raw reply	[flat|nested] 9+ messages in thread 
 
 
- * Re: [iproute2 net-next v2 0/3] ip netconf improvements
  2017-03-24  2:51 [iproute2 net-next v2 0/3] ip netconf improvements David Ahern
                   ` (4 preceding siblings ...)
  2017-04-04 21:07 ` David Ahern
@ 2017-04-14 23:02 ` Stephen Hemminger
  5 siblings, 0 replies; 9+ messages in thread
From: Stephen Hemminger @ 2017-04-14 23:02 UTC (permalink / raw)
  To: David Ahern; +Cc: netdev, nicolas.dichtel
On Thu, 23 Mar 2017 19:51:19 -0700
David Ahern <dsa@cumulusnetworks.com> wrote:
> Currently, ip netconf only shows data for ipv4 and ipv6 for dumps
> and just ipv4 for device requests. Improve the user experience by
> using the new kernel patch to dump all address families that have
> registered. For example, if mpls_router module is loaded then mpls
> values are displayed along with ipv4 and ipv6.
> 
> If the new feature is not supported (new iproute2 on older kernel)
> the kernel returns the nlmsg error EOPNOTSUPP which can be trapped
> and fallback to existing behavior.
> 
> v2
> - fixed index conversion in patch 3 per nicholas' comment
> 
> David Ahern (3):
>   netlink: Add flag to suppress print of nlmsg error
>   ip netconf: Show all address families by default in dumps
>   ip netconf: show all families on dev request
> 
>  include/libnetlink.h |  1 +
>  ip/ipnetconf.c       | 36 +++++++++++++++++++++++++-----------
>  lib/libnetlink.c     |  3 ++-
>  3 files changed, 28 insertions(+), 12 deletions(-)
> 
Sure applied, just wanted to make sure there was nothing interrelated
with recent spate of changes to netlink notifications.
^ permalink raw reply	[flat|nested] 9+ messages in thread