Re: [PATCHv2 RFC] RTEXT_FILTER_SKIP_STATS support to avoid dumping inet/inet6 stats

[Sowmini Varadhan]

On (09/12/15 00:22), Raghavendra K T wrote:
> 
> Sowmini, Thanks for the patch which is more cleaner way without
> breaking current behaviour.
> 
> [ Though RTEXT_FILTER_NEED_STATS flag with reverse effect  would have
> helped immediately :)]

Agree, but existing legacy usage will not set this flag, so I had 
few choices here.

> /me waits for the RTEXT_FILTER_SKIP_STATS to be supported in
> gccgo/golang, so that it can be used in docker newNetlinkRequest() to
> exploit this.

yes, I'm working on lining up the glibc bit as well (thus the cc to Jose..)
I'll send out the non-rfc version in a bit..

thanks for the feedback!

--Sowmini

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCHv2 RFC] RTEXT_FILTER_SKIP_STATS support to avoid dumping inet/inet6 stats

[Raghavendra K T]

On 09/11/2015 03:04 AM, Sowmini Varadhan wrote:


Many commonly used functions like getifaddrs() invoke RTM_GETLINK
to dump the interface information, and do not need the
the AF_INET6 statististics that are always returned by default
from rtnl_fill_ifinfo().

Computing the statistics can be an expensive operation that impacts
scaling, so it is desirable to avoid this if the information is
not needed.

This patch adds a the RTEXT_FILTER_SKIP_STATS extended info flag that
can be passed with netlink_request() to avoid statistics computation
for the ifinfo path.

Signed-off-by: Sowmini Varadhan 
---
v2: David Miller comments: pass u32 ext_filter_mask down.

  include/net/rtnetlink.h|3 ++-
  include/uapi/linux/rtnetlink.h |1 +
  net/core/rtnetlink.c   |2 +-
  net/ipv4/devinet.c |3 ++-
  net/ipv6/addrconf.c|   13 +
  5 files changed, 15 insertions(+), 7 deletions(-)


Sowmini, Thanks for the patch which is more cleaner way without
breaking current behaviour.

[ Though RTEXT_FILTER_NEED_STATS flag with reverse effect  would have
helped immediately :)]

/me waits for the RTEXT_FILTER_SKIP_STATS to be supported in
gccgo/golang, so that it can be used in docker newNetlinkRequest() to
exploit this.



--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCHv2 RFC] RTEXT_FILTER_SKIP_STATS support to avoid dumping inet/inet6 stats

[Sowmini Varadhan]

Many commonly used functions like getifaddrs() invoke RTM_GETLINK
to dump the interface information, and do not need the
the AF_INET6 statististics that are always returned by default
from rtnl_fill_ifinfo().

Computing the statistics can be an expensive operation that impacts
scaling, so it is desirable to avoid this if the information is
not needed.

This patch adds a the RTEXT_FILTER_SKIP_STATS extended info flag that
can be passed with netlink_request() to avoid statistics computation
for the ifinfo path.

Signed-off-by: Sowmini Varadhan 
---
v2: David Miller comments: pass u32 ext_filter_mask down.

 include/net/rtnetlink.h|3 ++-
 include/uapi/linux/rtnetlink.h |1 +
 net/core/rtnetlink.c   |2 +-
 net/ipv4/devinet.c |3 ++-
 net/ipv6/addrconf.c|   13 +
 5 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/include/net/rtnetlink.h b/include/net/rtnetlink.h
index 18fdb98..aff6ceb 100644
--- a/include/net/rtnetlink.h
+++ b/include/net/rtnetlink.h
@@ -122,7 +122,8 @@ struct rtnl_af_ops {
int family;
 
int (*fill_link_af)(struct sk_buff *skb,
-   const struct net_device *dev);
+   const struct net_device *dev,
+   u32 ext_filter_mask);
size_t  (*get_link_af_size)(const struct net_device 
*dev);
 
int (*validate_link_af)(const struct net_device 
*dev,
diff --git a/include/uapi/linux/rtnetlink.h b/include/uapi/linux/rtnetlink.h
index 7020247..434227f 100644
--- a/include/uapi/linux/rtnetlink.h
+++ b/include/uapi/linux/rtnetlink.h
@@ -666,6 +666,7 @@ struct tcamsg {
 #define RTEXT_FILTER_VF(1 << 0)
 #define RTEXT_FILTER_BRVLAN(1 << 1)
 #define RTEXT_FILTER_BRVLAN_COMPRESSED (1 << 2)
+#defineRTEXT_FILTER_SKIP_STATS (1 << 3)
 
 /* End of information exported to user level */
 
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index a466821..e545229 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -1272,7 +1272,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct 
net_device *dev,
if (!(af = nla_nest_start(skb, af_ops->family)))
goto nla_put_failure;
 
-   err = af_ops->fill_link_af(skb, dev);
+   err = af_ops->fill_link_af(skb, dev, ext_filter_mask);
 
/*
 * Caller may return ENODATA to indicate that there
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index 2d9cb17..7350084 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -1654,7 +1654,8 @@ static size_t inet_get_link_af_size(const struct 
net_device *dev)
return nla_total_size(IPV4_DEVCONF_MAX * 4); /* IFLA_INET_CONF */
 }
 
-static int inet_fill_link_af(struct sk_buff *skb, const struct net_device *dev)
+static int inet_fill_link_af(struct sk_buff *skb, const struct net_device *dev,
+u32 ext_filter_mask)
 {
struct in_device *in_dev = rcu_dereference_rtnl(dev->ip_ptr);
struct nlattr *nla;
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 99c0f2b..9acbb09 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -4760,7 +4760,8 @@ static void snmp6_fill_stats(u64 *stats, struct inet6_dev 
*idev, int attrtype,
}
 }
 
-static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev)
+static int inet6_fill_ifla6_attrs(struct sk_buff *skb, struct inet6_dev *idev,
+ u32 ext_filter_mask)
 {
struct nlattr *nla;
struct ifla_cacheinfo ci;
@@ -4780,6 +4781,9 @@ static int inet6_fill_ifla6_attrs(struct sk_buff *skb, 
struct inet6_dev *idev)
 
/* XXX - MC not implemented */
 
+   if (!!(ext_filter_mask & RTEXT_FILTER_SKIP_STATS))
+   return 0;
+
nla = nla_reserve(skb, IFLA_INET6_STATS, IPSTATS_MIB_MAX * sizeof(u64));
if (!nla)
goto nla_put_failure;
@@ -4815,14 +4819,15 @@ static size_t inet6_get_link_af_size(const struct 
net_device *dev)
return inet6_ifla6_size();
 }
 
-static int inet6_fill_link_af(struct sk_buff *skb, const struct net_device 
*dev)
+static int inet6_fill_link_af(struct sk_buff *skb, const struct net_device 
*dev,
+ u32 ext_filter_mask)
 {
struct inet6_dev *idev = __in6_dev_get(dev);
 
if (!idev)
return -ENODATA;
 
-   if (inet6_fill_ifla6_attrs(skb, idev) < 0)
+   if (inet6_fill_ifla6_attrs(skb, idev, ext_filter_mask) < 0)
return -EMSGSIZE;
 
return 0;
@@ -4977,7 +4982,7 @@ static int inet6_fill_ifinfo(struct sk_buff *skb, struct 
inet6_dev *idev,
if (!protoinfo)
goto nla_put_failure;