Skip to contentSkip to conversation

ipv6: Move ipv6 stubs to a separate header file

PATCH Applied David Ahern 5 participants 16 messages Mar 22, 2019 – Mar 29, 2019

Subsystems (12)
lore
Share & tools
lore

David Ahern

author maintainer
#0 lore raw

PATCH 1/1 ipv6: Move ipv6 stubs to a separate header file

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Code move only; no functional change intended.

---
 drivers/infiniband/core/addr.c                  |  2 +-
 drivers/net/ethernet/mellanox/mlx5/core/en_tc.c |  1 +
 drivers/net/geneve.c                            |  1 +
 drivers/net/usb/cdc_mbim.c                      |  1 +
 drivers/net/vxlan.c                             |  1 +
 include/net/addrconf.h                          | 47 ------------------
 include/net/ipv6_stubs.h                        | 63 +++++++++++++++++++++++++
 include/net/udp_tunnel.h                        |  2 +-
 net/bridge/br_arp_nd_proxy.c                    |  1 +
 net/core/filter.c                               |  1 +
 net/core/lwt_bpf.c                              |  1 +
 net/ipv6/addrconf_core.c                        |  2 +-
 net/ipv6/af_inet6.c                             |  1 +
 net/mpls/af_mpls.c                              |  2 +-
 net/tipc/udp_media.c                            |  2 +-
 15 files changed, 76 insertions(+), 52 deletions(-)
 create mode 100644 include/net/ipv6_stubs.h
drivers/infiniband/core/addr.c +1 -1
index 0dce94e3c495..2649e0f2ff65 100644
@@ -42,7 +42,7 @@
 #include <net/neighbour.h>
 #include <net/route.h>
 #include <net/netevent.h>
-#include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #include <net/ip6_route.h>
 #include <rdma/ib_addr.h>
 #include <rdma/ib_sa.h>
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +1 -0
index b4967a0ff8c7..6f2852c7365c 100644
@@ -44,6 +44,7 @@
 #include <net/tc_act/tc_pedit.h>
 #include <net/tc_act/tc_csum.h>
 #include <net/arp.h>
+#include <net/ipv6_stubs.h>
 #include "en.h"
 #include "en_rep.h"
 #include "en_tc.h"
drivers/net/geneve.c +1 -0
index 5583d993480d..384d2dbca324 100644
@@ -14,6 +14,7 @@
 #include <linux/module.h>
 #include <linux/etherdevice.h>
 #include <linux/hash.h>
+#include <net/ipv6_stubs.h>
 #include <net/dst_metadata.h>
 #include <net/gro_cells.h>
 #include <net/rtnetlink.h>
drivers/net/usb/cdc_mbim.c +1 -0
index 0362acd5cdca..28321aca48fe 100644
@@ -23,6 +23,7 @@
 #include <linux/usb/cdc_ncm.h>
 #include <net/ipv6.h>
 #include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 
 /* alternative VLAN for IP session 0 if not untagged */
 #define MBIM_IPS0_VID	4094
drivers/net/vxlan.c +1 -0
index d76dfed8d9bb..5994d5415a03 100644
@@ -20,6 +20,7 @@
 #include <linux/ethtool.h>
 #include <net/arp.h>
 #include <net/ndisc.h>
+#include <net/ipv6_stubs.h>
 #include <net/ip.h>
 #include <net/icmp.h>
 #include <net/rtnetlink.h>
include/net/addrconf.h +0 -47

53 lines not on this page · raw message

include/net/ipv6_stubs.h new file +63 -0

63 lines not on this page · raw message

include/net/udp_tunnel.h +1 -1
index b8137953fea3..4b1f95e08307 100644
@@ -7,7 +7,7 @@
 
 #if IS_ENABLED(CONFIG_IPV6)
 #include <net/ipv6.h>
-#include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #endif
 
 struct udp_port_cfg {
net/bridge/br_arp_nd_proxy.c +1 -0
index 6b78e6351719..724b474ade54 100644
@@ -21,6 +21,7 @@
 #include <linux/if_vlan.h>
 #include <linux/inetdevice.h>
 #include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #if IS_ENABLED(CONFIG_IPV6)
 #include <net/ip6_checksum.h>
 #endif
net/core/filter.c +1 -0
index 647c63a7b25b..841502f894bc 100644
@@ -74,6 +74,7 @@
 #include <net/seg6.h>
 #include <net/seg6_local.h>
 #include <net/lwtunnel.h>
+#include <net/ipv6_stubs.h>
 
 /**
  *	sk_filter_trim_cap - run a packet through a socket filter
net/core/lwt_bpf.c +1 -0
index 126d31ff5ee3..3c5c24a5d9f5 100644
@@ -18,6 +18,7 @@
 #include <net/lwtunnel.h>
 #include <net/gre.h>
 #include <net/ip6_route.h>
+#include <net/ipv6_stubs.h>
 
 struct bpf_lwt_prog {
 	struct bpf_prog *prog;
net/ipv6/addrconf_core.c +1 -1
index 6c79af056d9b..945b66e3008f 100644
@@ -5,7 +5,7 @@
 
 #include <linux/export.h>
 #include <net/ipv6.h>
-#include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #include <net/ip.h>
 
 /* if ipv6 module registers this function is used by xfrm to force all
net/ipv6/af_inet6.c +1 -0
index fa6b404cbd10..1789bf99c419 100644
@@ -56,6 +56,7 @@
 #include <net/transp_v6.h>
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #include <net/ndisc.h>
 #ifdef CONFIG_IPV6_TUNNEL
 #include <net/ip6_tunnel.h>
net/mpls/af_mpls.c +1 -1
index f7c544592ec8..8120e04f15e4 100644
@@ -22,7 +22,7 @@
 #if IS_ENABLED(CONFIG_IPV6)
 #include <net/ipv6.h>
 #endif
-#include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #include <net/nexthop.h>
 #include "internal.h"
 
net/tipc/udp_media.c +1 -1
index 4d85d71f16e2..6f166fbbfff1 100644
@@ -44,7 +44,7 @@
 #include <net/sock.h>
 #include <net/ip.h>
 #include <net/udp_tunnel.h>
-#include <net/addrconf.h>
+#include <net/ipv6_stubs.h>
 #include <linux/tipc_netlink.h>
 #include "core.h"
 #include "addr.h"
Signature · 1 line
2.11.0
#1 lore raw
Quoted 9 lines · David Ahern

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Code move only; no functional change intended.

nack.
why this code churn?
I very much prefer to disallow ipv6 as a module.

#2 lore raw
Quoted 7 lines · Alexei Starovoitov
Quoted 6 lines · David Ahern

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Quoted 6 lines
Quoted 3 lines

Code move only; no functional change intended.

nack.
why this code churn?

This is a prep patch for nexthops as separate objects. I will be adding 4 more functions to the stubs, and I hit a bit of a header nightmare with it in addrconf.h.

And this has no impact on ipv6 built in or as a module; it has only to do with a proper header file for a definition.

#3 lore raw
Quoted 10 lines · David Ahern
Quoted 9 lines · Alexei Starovoitov
Quoted 8 lines · David Ahern

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Code move only; no functional change intended.

Quoted 8 lines
Quoted 4 lines
Quoted 1 line

nack.
why this code churn?

This is a prep patch for nexthops as separate objects. I will be adding 4 more functions to the stubs, and I hit a bit of a header nightmare with it in addrconf.h.

please do not.
I see no value in splitting nexthops either.

maintainer
#4 lore raw
In reply to David Ahern

From: David Ahern <[email protected]>
Date: Fri, 22 Mar 2019 06:06:09 -0700

Quoted 9 lines

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Code move only; no functional change intended.

Eric, I fully support David's overall plan to make separate nexthop objects as it will significantly empower the stack to do more sensible things when links flap etc.

If you don't like this road he is going down to achieve that, you must suggest an alternative.

#5 lore raw
Quoted 9 lines · David Miller

From: David Ahern <[email protected]>
Date: Fri, 22 Mar 2019 06:06:09 -0700

Quoted 6 lines

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Quoted 7 lines
Quoted 3 lines

Code move only; no functional change intended.

Eric, I fully support David's overall plan to make separate nexthop objects as it will significantly empower the stack to do more sensible things when links flap etc.

let's agree to disagree.
'link flaps' were not mentioned in the cover letter for:
"net: Improve route scalability via support for nexthop objects"

The _only_ value of 86 patches is to align linux kernel routing with switch ASICs, because cumulus is trying to reuse iproute2
to manage them.
It was broken model to begin with and it keeps complicating routing when linux is used as a host while not achieving the goal of iproute2 for switches.
Can anyone use off the shelf linux to manage trident/tomahawk switches? Nope. brcm sdk is still necessary.
nexthop objects are essential to configure enterprise switches. Clearly cumulus customers don't like iproute2 style because it's missing this feature, so David's proposal is to add that to the kernel. Even after kernel and iproute2 understand nexthop id the kernel is still not going to be competitive with switching os. The linux kernel is an OS to run on the host cpu and to run on a control plane cpu of a switch. That is all great, but the reasons to push routing into the kernel of control plane cpu were weak. It's not using these routes.
Such architecture allowed temporary reuse of bgp daemons, but it fails to scale. No need to push route to the kernel when kernel won't use them. Hence an alternative proposal:
- introduce hooks at netlink layer and steal back and forth messages from your favorite daemon without populating the kernel
- same for iproute2 netlink interaction

#6 lore raw
Quoted 36 lines · Alexei Starovoitov
Quoted 9 lines · David Miller

From: David Ahern <[email protected]>
Date: Fri, 22 Mar 2019 06:06:09 -0700

Quoted 6 lines

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Quoted 7 lines
Quoted 3 lines

Code move only; no functional change intended.

Eric, I fully support David's overall plan to make separate nexthop objects as it will significantly empower the stack to do more sensible things when links flap etc.

let's agree to disagree.
'link flaps' were not mentioned in the cover letter for:
"net: Improve route scalability via support for nexthop objects"

The _only_ value of 86 patches is to align linux kernel routing with switch ASICs, because cumulus is trying to reuse iproute2
to manage them.
It was broken model to begin with and it keeps complicating routing when linux is used as a host while not achieving the goal of iproute2 for switches.
Can anyone use off the shelf linux to manage trident/tomahawk switches? Nope. brcm sdk is still necessary.
nexthop objects are essential to configure enterprise switches. Clearly cumulus customers don't like iproute2 style because it's missing this feature, so David's proposal is to add that to the kernel. Even after kernel and iproute2 understand nexthop id the kernel is still not going to be competitive with switching os. The linux kernel is an OS to run on the host cpu and to run on a control plane cpu of a switch.

Quoted 9 lines

That is all great, but the reasons to push routing into the kernel of control plane cpu were weak. It's not using these routes.
Such architecture allowed temporary reuse of bgp daemons, but it fails to scale. No need to push route to the kernel when kernel won't use them. Hence an alternative proposal:
- introduce hooks at netlink layer and steal back and forth messages from your favorite daemon without populating the kernel
- same for iproute2 netlink interaction

The use case here is not just Cumulus or switchdev, but ANY OS using the Linux API and the kernel to configure and manage its networking state
[1] and that includes XDP based use cases [2] and routing on the host.[3] This is not about iproute2 driving networking deployments. This is about continuing to remove the 'fails to scale' notion which *forces* a NOS architecture away from the kernel databases as the single source of truth and the kernel's IPC/notification mechanisms, and the subsequent impacts of that choice which negates the Linux ecosystem forcing a customization of all of the software running in the control plane to work in some vendor's custom environment. You should read the paper I wrote last summer [1].

This current patch set is not just about link flaps, but improving the overall scaling properties of managing the FIB. This is about leveraging existing ideas about network models and their scalability properties and bringing that efficiency to Linux. With nexthops, the time to insert routes is near constant regardless of the number of nexthops in the route. So the time to insert a single path route and the time to insert a route with 2, 4, 8, 16, 32, … nexthops is the same. That is a HUGE scalability improvement from a simple idea. The “near” constant is because of the need to expand nexthop definitions in the route notifications to userspace to enable legacy applications to work with this new API. In time, a lever can be added to not expand the definitions and let the RTA_NHID alone point to it, allowing companies who know that there are no legacy apps that need the nexthop expansion to gain the full scaling improvements.

This change also enables many other key features:
1. IPv4 multipath routes are not evicted just because 1 hop goes down.
2. IPv6 multipath routes with device only nexthops (e.g., tunnels).
3. IPv6 nexthop with IPv4 route (aka, RFC 5549) which enables a more natural BGP unnumbered.
4. Lower memory consumption for IPv6 FIB entries which has no sharing at all like IPv4 does.
5. Allows atomic update of nexthop definitions with a single replace command as opposed to replacing the N-routes using it.

The list goes on, but 2-5 of the above were in the cover letter I sent on March 14.

I have spent a lot of time over the last few years not just working on features like VRF and MPLS, but improving the scaling properties of Linux and removing this 'fails to scale' notion you and others hold. This current patch set is just another step in that path.

[1] https://www.files.netdevconf.org/d/f982086fdd6946d9b596/
[2] http://vger.kernel.org/lpc_net2018_talks/dsa-xdp-kernel-tables-paper.pdf
[3]
https://netdevconf.org/1.2/slides/oct7/01_ahern_microservice_net_vrf_on_host.pdf

#7 lore raw
Quoted 10 lines · David Ahern

This change also enables many other key features:
1. IPv4 multipath routes are not evicted just because 1 hop goes down.
2. IPv6 multipath routes with device only nexthops (e.g., tunnels).
3. IPv6 nexthop with IPv4 route (aka, RFC 5549) which enables a more natural BGP unnumbered.
4. Lower memory consumption for IPv6 FIB entries which has no sharing at all like IPv4 does.
5. Allows atomic update of nexthop definitions with a single replace command as opposed to replacing the N-routes using it.

Does kernel work as data plane or control plane in any of the above features ?
Sadly the patches allow it to do both, but cumulus doesn't use it for data path. The kernel on control plane cpu is merely a database. And today it doesn't scale when used as a database.
The kernel has to be fast as a dataplane but these extra features will slow down the routing by making kernel-as-database scale a bit better. Hence my suggestion in the previous email: use proper database to store routes, nexthops and whatever else necessary to program the asic. The kernel doesn't need to hold this information.

#8 lore raw

Just a comment: on the system that I work on, the kernel closely matches the hardware data plane, so the kernel has both roles as data and control plane.
Whatever routes are installed on the hardware are installed also in the kernel.

Quoted 24 lines · Alexei Starovoitov
Quoted 10 lines · David Ahern

This change also enables many other key features:
1. IPv4 multipath routes are not evicted just because 1 hop goes down.
2. IPv6 multipath routes with device only nexthops (e.g., tunnels).
3. IPv6 nexthop with IPv4 route (aka, RFC 5549) which enables a more natural BGP unnumbered.
4. Lower memory consumption for IPv6 FIB entries which has no sharing at all like IPv4 does.
5. Allows atomic update of nexthop definitions with a single replace command as opposed to replacing the N-routes using it.

Does kernel work as data plane or control plane in any of the above features ?
Sadly the patches allow it to do both, but cumulus doesn't use it for data path. The kernel on control plane cpu is merely a database. And today it doesn't scale when used as a database.
The kernel has to be fast as a dataplane but these extra features will slow down the routing by making kernel-as-database scale a bit better. Hence my suggestion in the previous email: use proper database to store routes, nexthops and whatever else necessary to program the asic. The kernel doesn't need to hold this information.

#9 lore raw
In reply to Alexei Starovoitov
Quoted 12 lines · Alexei Starovoitov
Quoted 10 lines · David Ahern

This change also enables many other key features:
1. IPv4 multipath routes are not evicted just because 1 hop goes down.
2. IPv6 multipath routes with device only nexthops (e.g., tunnels).
3. IPv6 nexthop with IPv4 route (aka, RFC 5549) which enables a more natural BGP unnumbered.
4. Lower memory consumption for IPv6 FIB entries which has no sharing at all like IPv4 does.
5. Allows atomic update of nexthop definitions with a single replace command as opposed to replacing the N-routes using it.

Quoted 11 lines

Does kernel work as data plane or control plane in any of the above features ?
Sadly the patches allow it to do both, but cumulus doesn't use it for data path. The kernel on control plane cpu is merely a database. And today it doesn't scale when used as a database.
The kernel has to be fast as a dataplane but these extra features will slow down the routing by making kernel-as-database scale a bit better. Hence my suggestion in the previous email: use proper database to store routes, nexthops and whatever else necessary to program the asic. The kernel doesn't need to hold this information.

The first 40 patches align fib_nh and fib6_nh providing more consistency and alignment between IPv4 and IPv6 and allowing more code re-use between the protocols. The end result is the ability to have IPv6 gateways with IPv4 routes, a much needed control plane feature other companies have been harassing me about as well as the internal need for Cumulus. In the refactoring I have been very careful about changes to data structure layout and cacheline hits as well as adverse changes to memory use. I believe at the end of this change set there is no impact to existing performance - control plane or data plane.

That is followed by refactoring IPv6 again in a direction that makes IPv4 and IPv6 more consistent and enables changes (outside of the nexthop sets) that will improve IPv6 for a number of cases by removing the need to always generate a dst_entry.

After that are a few patches exporting functions for use by nexthop code and then diving into the refactoring enabling separate nexthop objects. Again, impacts to performance have been top of mind, and I have done what I can to minimize any overhead in the datapath - to the point of a few ‘if (nh)’ checks wrapped in an unlikely. And with the nexthop code in place it gives users an alternative to a broken IPv6 multipath API as one example.

As far as scalability goes, I can already inject a million routes into the kernel FIB. This allows me to it more efficiently and to manage the FIBs more efficiently in the face of changes such as a link going down as we move to higher end systems - such as spectrum2.

As for routes in the kernel, they need to be there for any control plane processes to properly function. One example is ping and traceroute to troubleshoot data path problems, and another is for bgp (or any other service) to connect to a peer through the data plane (do not assume a peer is on a directly connected route). Further, the routes need to go through the kernel to get to the switchdev driver. The routes need to be there for XDP forwarding and routing on the host. Pawel has already expressed interest in using XDP for fast path forwarding with FRR managing the route table.

You keep trying to make this about Cumulus. This is about bringing next level features to Linux and in the process bringing more consistency and code sharing between IPv4 and IPv6. This is about 1-API for the data center be it servers, hosts, switches or routers regardless of datapath (hardware offload, XDP, or kernel forwarding), and maintaining consistency in configuring, monitoring and troubleshooting across those systems. That is the common theme of both the netdev talk last summer and the talk at LPC in November.

Again, I have tried to be very careful with the intrusion of checks into the datapath with the goal of no measurable impact to performance. I am invested to seeing that through and will continue looking for ways to improve it for all use cases.

#14 lore raw
In reply to Alexei Starovoitov
Quoted 2 lines · Alexei Starovoitov

The kernel has to be fast as a dataplane but these extra features will slow down the routing by making kernel-as-database scale a bit better.

I took Vincent Bernat's fib lookup microbenchmark and added input and output route lookups to it. Through this commit:

commit 4d5edf447c88e59f5c5716180c1001cde9bf4473

Author: David Ahern <[email protected]>
Date:   Mon Mar 11 18:34:48 2019 -0700

ipv4: Allow ipv6 gateway with ipv4 routes

There is no real measurable difference for IPv4 fib lookups (fib_lookup), input routes (ip_route_input_rcu) and output routes (ip_route_output_key_hash_rcu). This is using a bare metal server with a L5640 Xeon CPU.

The above commit is the target for the first 2 rounds of patches. I will of course be checking performance each step along the way to nexthop objects.

#10 lore raw
In reply to David Ahern
Quoted 5 lines · David Ahern

That is followed by refactoring IPv6 again in a direction that makes IPv4 and IPv6 more consistent and enables changes (outside of the nexthop sets) that will improve IPv6 for a number of cases by removing the need to always generate a dst_entry.

That would be a nice feature to have.

Quoted 7 lines

After that are a few patches exporting functions for use by nexthop code and then diving into the refactoring enabling separate nexthop objects. Again, impacts to performance have been top of mind, and I have done what I can to minimize any overhead in the datapath - to the point of a few ‘if (nh)’ checks wrapped in an unlikely. And with the nexthop code in place it gives users an alternative to a broken IPv6 multipath API as one example.

...

Quoted 4 lines

Again, I have tried to be very careful with the intrusion of checks into the datapath with the goal of no measurable impact to performance. I am invested to seeing that through and will continue looking for ways to improve it for all use cases.

Great to hear.
Can you split up your set into reviewable chunks?
Just this patch alone is too small to see the road ahead
and 80+ patches are too much to review properly.
I still have reservations regarding nexthop id concept, but sounds like the first 20 or so patches should clean things up.
Especially if you can get rid of dst alloc/free back and forth in ipv6 case.

#11 lore raw
Quoted 2 lines · Alexei Starovoitov

Just this patch alone is too small to see the road ahead and 80+ patches are too much to review properly.

I will be sending in reviewable chunks. I think I have established a history of evolving the code in small, singly focused patches moving the code from where it is to where I want it to be. Any time there is a 'large' patch I try to keep it to something mechanical - like a name change or a change to a function signature.

The email I sent ("net: Improve route scalability via support for nexthop objects") outlined the targeted steps and the intermediate prep work needed. To be able to reach certain clear steps and still maintain the 20'ish patches in a set requirement, I have sent a few patches which are standalone and basically noise - like this one - to avoid distractions on the real change. I pushed the entire set to github so anyone can see the end goal as well as the individual patches.

I am not sure what else I can do to be open about this change and convey the what and the why. I discussed the feature at netconf in Seoul in October 2017. I sent RFC patches last August on the UAPI with a working implementation as well as more details about what this change is going to allow. I have given talks on what I believe is the best architecture and implementation for Linux which is the motivation for improving the kernel. And the previous email about improving scalability is almost a novelette in length.

#12 lore raw
Quoted 4 lines · David Ahern

work needed. To be able to reach certain clear steps and still maintain the 20'ish patches in a set requirement, I have sent a few patches which are standalone and basically noise - like this one - to avoid distractions on the real change.

This patch is far from noise. In the thread you said that you'll be adding 4 more indirect calls and that rings performance alarms. retpoline is expensive. If you want to add more indirect calls to fib lookup it's no go. Benchmarks with retpoline=off won't show the difference.

#13 lore raw
Quoted 11 lines · Alexei Starovoitov
Quoted 4 lines · David Ahern

work needed. To be able to reach certain clear steps and still maintain the 20'ish patches in a set requirement, I have sent a few patches which are standalone and basically noise - like this one - to avoid distractions on the real change.

This patch is far from noise. In the thread you said that you'll be adding 4 more indirect calls and that rings performance alarms. retpoline is expensive. If you want to add more indirect calls to fib lookup it's no go. Benchmarks with retpoline=off won't show the difference.

As I stated in the email I sent ("net: Improve route scalability via support for nexthop objects"):

"adding hooks to the ipv6 stubs (bump sernum, send route notifications and delete routes based on nexthop updates)."

Those plus initialization and release of fib6_nh are all done in the control path.

maintainer Applied
#15 lore raw
In reply to David Ahern

From: David Ahern <[email protected]>
Date: Fri, 22 Mar 2019 06:06:09 -0700

Quoted 9 lines

From: David Ahern <[email protected]>

The number of stubs is growing and has nothing to do with addrconf. Move the definition of the stubs to a separate header file and update users. In the move, drop the vxlan specific comment before ipv6_stub.

Code move only; no functional change intended.

Applied, thanks David.