diff mbox series

[bpf,v2,2/3] bpf_fib_lookup: optionally skip neighbour lookup

Message ID 160319106331.15822.2945713836148003890.stgit@toke.dk (mailing list archive)
State Superseded
Delegated to: BPF
Headers show
Series bpf: Rework bpf_redirect_neigh() to allow supplying nexthop from caller | expand

Commit Message

Toke Høiland-Jørgensen Oct. 20, 2020, 10:51 a.m. UTC 
From: Toke Høiland-Jørgensen <toke@redhat.com>

The bpf_fib_lookup() helper performs a neighbour lookup for the destination
IP and returns BPF_FIB_LKUP_NO_NEIGH if this fails, with the expectation
that the BPF program will deal with this condition, either by passing the
packet up the stack, or by using bpf_redirect_neigh().

The neighbour lookup is done via a hash table (through ___neigh_lookup_noref()),
which incurs some overhead. If the caller knows this is likely to fail
anyway, it may want to skip that and go unconditionally to
bpf_redirect_neigh(). For this use case, add a flag to bpf_fib_lookup()
that will make it skip the neighbour lookup and instead always return
BPF_FIB_LKUP_RET_NO_NEIGH (but still populate the gateway and target
ifindex).

Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
---
 include/uapi/linux/bpf.h       |   10 ++++++----
 net/core/filter.c              |   16 ++++++++++++++--
 tools/include/uapi/linux/bpf.h |   10 ++++++----
 3 files changed, 26 insertions(+), 10 deletions(-)

Comments

David Ahern Oct. 20, 2020, 1:49 p.m. UTC | #1 
On 10/20/20 4:51 AM, Toke Høiland-Jørgensen wrote:
> From: Toke Høiland-Jørgensen <toke@redhat.com>
> 
> The bpf_fib_lookup() helper performs a neighbour lookup for the destination
> IP and returns BPF_FIB_LKUP_NO_NEIGH if this fails, with the expectation
> that the BPF program will deal with this condition, either by passing the
> packet up the stack, or by using bpf_redirect_neigh().
> 
> The neighbour lookup is done via a hash table (through ___neigh_lookup_noref()),
> which incurs some overhead. If the caller knows this is likely to fail
> anyway, it may want to skip that and go unconditionally to
> bpf_redirect_neigh(). For this use case, add a flag to bpf_fib_lookup()
> that will make it skip the neighbour lookup and instead always return
> BPF_FIB_LKUP_RET_NO_NEIGH (but still populate the gateway and target
> ifindex).
> 
> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
> ---
>  include/uapi/linux/bpf.h       |   10 ++++++----
>  net/core/filter.c              |   16 ++++++++++++++--
>  tools/include/uapi/linux/bpf.h |   10 ++++++----
>  3 files changed, 26 insertions(+), 10 deletions(-)

Nack. Please don't.

As I mentioned in my reply to Daniel, I would prefer such logic be
pushed to the bpf programs. There is no reason for rare run time events
to warrant a new flag and new check in the existing FIB helpers. The bpf
programs can take the hit of the extra lookup.
Daniel Borkmann Oct. 20, 2020, 3:04 p.m. UTC | #2 
On 10/20/20 3:49 PM, David Ahern wrote:
> On 10/20/20 4:51 AM, Toke Høiland-Jørgensen wrote:
>> From: Toke Høiland-Jørgensen <toke@redhat.com>
>>
>> The bpf_fib_lookup() helper performs a neighbour lookup for the destination
>> IP and returns BPF_FIB_LKUP_NO_NEIGH if this fails, with the expectation
>> that the BPF program will deal with this condition, either by passing the
>> packet up the stack, or by using bpf_redirect_neigh().
>>
>> The neighbour lookup is done via a hash table (through ___neigh_lookup_noref()),
>> which incurs some overhead. If the caller knows this is likely to fail
>> anyway, it may want to skip that and go unconditionally to
>> bpf_redirect_neigh(). For this use case, add a flag to bpf_fib_lookup()
>> that will make it skip the neighbour lookup and instead always return
>> BPF_FIB_LKUP_RET_NO_NEIGH (but still populate the gateway and target
>> ifindex).
>>
>> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
>> ---
>>   include/uapi/linux/bpf.h       |   10 ++++++----
>>   net/core/filter.c              |   16 ++++++++++++++--
>>   tools/include/uapi/linux/bpf.h |   10 ++++++----
>>   3 files changed, 26 insertions(+), 10 deletions(-)
> 
> Nack. Please don't.
> 
> As I mentioned in my reply to Daniel, I would prefer such logic be
> pushed to the bpf programs. There is no reason for rare run time events
> to warrant a new flag and new check in the existing FIB helpers. The bpf
> programs can take the hit of the extra lookup.

Fair enough, lets push it to progs then.
Toke Høiland-Jørgensen Oct. 20, 2020, 6:10 p.m. UTC | #3 
Daniel Borkmann <daniel@iogearbox.net> writes:

> On 10/20/20 3:49 PM, David Ahern wrote:
>> On 10/20/20 4:51 AM, Toke Høiland-Jørgensen wrote:
>>> From: Toke Høiland-Jørgensen <toke@redhat.com>
>>>
>>> The bpf_fib_lookup() helper performs a neighbour lookup for the destination
>>> IP and returns BPF_FIB_LKUP_NO_NEIGH if this fails, with the expectation
>>> that the BPF program will deal with this condition, either by passing the
>>> packet up the stack, or by using bpf_redirect_neigh().
>>>
>>> The neighbour lookup is done via a hash table (through ___neigh_lookup_noref()),
>>> which incurs some overhead. If the caller knows this is likely to fail
>>> anyway, it may want to skip that and go unconditionally to
>>> bpf_redirect_neigh(). For this use case, add a flag to bpf_fib_lookup()
>>> that will make it skip the neighbour lookup and instead always return
>>> BPF_FIB_LKUP_RET_NO_NEIGH (but still populate the gateway and target
>>> ifindex).
>>>
>>> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
>>> ---
>>>   include/uapi/linux/bpf.h       |   10 ++++++----
>>>   net/core/filter.c              |   16 ++++++++++++++--
>>>   tools/include/uapi/linux/bpf.h |   10 ++++++----
>>>   3 files changed, 26 insertions(+), 10 deletions(-)
>> 
>> Nack. Please don't.
>> 
>> As I mentioned in my reply to Daniel, I would prefer such logic be
>> pushed to the bpf programs. There is no reason for rare run time events
>> to warrant a new flag and new check in the existing FIB helpers. The bpf
>> programs can take the hit of the extra lookup.
>
> Fair enough, lets push it to progs then.

OK, with this and the other comments, this goes back to v1 + the
compilation fix. Will send that as v3...

-Toke
diff mbox series

Patch

diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index 9668cde9d684..4bfd3c72dae6 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -4841,12 +4841,14 @@  struct bpf_raw_tracepoint_args {
 	__u64 args[0];
 };
 
-/* DIRECT:  Skip the FIB rules and go to FIB table associated with device
- * OUTPUT:  Do lookup from egress perspective; default is ingress
+/* DIRECT:      Skip the FIB rules and go to FIB table associated with device
+ * OUTPUT:      Do lookup from egress perspective; default is ingress
+ * SKIP_NEIGH:  Skip neighbour lookup and return BPF_FIB_LKUP_RET_NO_NEIGH on success
  */
 enum {
-	BPF_FIB_LOOKUP_DIRECT  = (1U << 0),
-	BPF_FIB_LOOKUP_OUTPUT  = (1U << 1),
+	BPF_FIB_LOOKUP_DIRECT	  = (1U << 0),
+	BPF_FIB_LOOKUP_OUTPUT	  = (1U << 1),
+	BPF_FIB_LOOKUP_SKIP_NEIGH = (1U << 2),
 };
 
 enum {
diff --git a/net/core/filter.c b/net/core/filter.c
index fa09b4f141ae..9791e6311afa 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -5382,6 +5382,9 @@  static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params,
 		if (nhc->nhc_gw_family)
 			params->ipv4_dst = nhc->nhc_gw.ipv4;
 
+		if (flags & BPF_FIB_LOOKUP_SKIP_NEIGH)
+			return BPF_FIB_LKUP_RET_NO_NEIGH;
+
 		neigh = __ipv4_neigh_lookup_noref(dev,
 						 (__force u32)params->ipv4_dst);
 	} else {
@@ -5389,6 +5392,10 @@  static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params,
 
 		params->family = AF_INET6;
 		*dst = nhc->nhc_gw.ipv6;
+
+		if (flags & BPF_FIB_LOOKUP_SKIP_NEIGH)
+			return BPF_FIB_LKUP_RET_NO_NEIGH;
+
 		neigh = __ipv6_neigh_lookup_noref_stub(dev, dst);
 	}
 
@@ -5501,6 +5508,9 @@  static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params,
 	params->rt_metric = res.f6i->fib6_metric;
 	params->ifindex = dev->ifindex;
 
+	if (flags & BPF_FIB_LOOKUP_SKIP_NEIGH)
+		return BPF_FIB_LKUP_RET_NO_NEIGH;
+
 	/* xdp and cls_bpf programs are run in RCU-bh so rcu_read_lock_bh is
 	 * not needed here.
 	 */
@@ -5518,7 +5528,8 @@  BPF_CALL_4(bpf_xdp_fib_lookup, struct xdp_buff *, ctx,
 	if (plen < sizeof(*params))
 		return -EINVAL;
 
-	if (flags & ~(BPF_FIB_LOOKUP_DIRECT | BPF_FIB_LOOKUP_OUTPUT))
+	if (flags & ~(BPF_FIB_LOOKUP_DIRECT | BPF_FIB_LOOKUP_OUTPUT |
+		      BPF_FIB_LOOKUP_SKIP_NEIGH))
 		return -EINVAL;
 
 	switch (params->family) {
@@ -5555,7 +5566,8 @@  BPF_CALL_4(bpf_skb_fib_lookup, struct sk_buff *, skb,
 	if (plen < sizeof(*params))
 		return -EINVAL;
 
-	if (flags & ~(BPF_FIB_LOOKUP_DIRECT | BPF_FIB_LOOKUP_OUTPUT))
+	if (flags & ~(BPF_FIB_LOOKUP_DIRECT | BPF_FIB_LOOKUP_OUTPUT |
+		      BPF_FIB_LOOKUP_SKIP_NEIGH))
 		return -EINVAL;
 
 	switch (params->family) {
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 9668cde9d684..4bfd3c72dae6 100644
--- a/tools/include/uapi/linux/bpf.h
+++ b/tools/include/uapi/linux/bpf.h
@@ -4841,12 +4841,14 @@  struct bpf_raw_tracepoint_args {
 	__u64 args[0];
 };
 
-/* DIRECT:  Skip the FIB rules and go to FIB table associated with device
- * OUTPUT:  Do lookup from egress perspective; default is ingress
+/* DIRECT:      Skip the FIB rules and go to FIB table associated with device
+ * OUTPUT:      Do lookup from egress perspective; default is ingress
+ * SKIP_NEIGH:  Skip neighbour lookup and return BPF_FIB_LKUP_RET_NO_NEIGH on success
  */
 enum {
-	BPF_FIB_LOOKUP_DIRECT  = (1U << 0),
-	BPF_FIB_LOOKUP_OUTPUT  = (1U << 1),
+	BPF_FIB_LOOKUP_DIRECT	  = (1U << 0),
+	BPF_FIB_LOOKUP_OUTPUT	  = (1U << 1),
+	BPF_FIB_LOOKUP_SKIP_NEIGH = (1U << 2),
 };
 
 enum {