| Message ID | 20210309101321.2138655-3-liuhangbin@gmail.com |
|---|---|
| State | Superseded |
| Headers | show |
| Series | xdp: extend xdp_redirect_map with broadcast support | expand |
Hangbin Liu <liuhangbin@gmail.com> writes: > This patch add two flags BPF_F_BROADCAST and BPF_F_EXCLUDE_INGRESS to extend > xdp_redirect_map for broadcast support. > > Keep the general data path in net/core/filter.c and the native data > path in kernel/bpf/devmap.c so we can use direct calls to get better > performace. > > Here is the performance result by using xdp_redirect_{map, map_multi} in > sample/bpf and send pkts via pktgen cmd: > ./pktgen_sample03_burst_single_flow.sh -i eno1 -d $dst_ip -m $dst_mac -t 10 -s 64 > > There are some drop back as we need to loop the map and get each > interface. > > Version | Test | Generic | Native > 5.11 | redirect_map i40e->i40e | 1.9M | 9.3M > 5.11 | redirect_map i40e->veth | 1.5M | 11.2M > 5.11 + patch | redirect_map i40e->i40e | 1.9M | 9.6M > 5.11 + patch | redirect_map i40e->veth | 1.5M | 11.9M > 5.11 + patch | redirect_map_multi i40e->i40e | 1.5M | 7.7M > 5.11 + patch | redirect_map_multi i40e->veth | 1.2M | 9.1M > 5.11 + patch | redirect_map_multi i40e->mlx4+veth | 0.9M | 3.2M > > v2: fix flag renaming issue in v1 > > Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> FYI, this no longer applies to bpf-next due to Björn's refactor in commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") Also, two small nits below: > --- > include/linux/bpf.h | 16 +++++ > include/net/xdp.h | 1 + > include/uapi/linux/bpf.h | 17 ++++- > kernel/bpf/devmap.c | 119 +++++++++++++++++++++++++++++++++ > net/core/filter.c | 74 ++++++++++++++++++-- > net/core/xdp.c | 29 ++++++++ > tools/include/uapi/linux/bpf.h | 17 ++++- > 7 files changed, 262 insertions(+), 11 deletions(-) > > diff --git a/include/linux/bpf.h b/include/linux/bpf.h > index c931bc97019d..bb07ccd170f2 100644 > --- a/include/linux/bpf.h > +++ b/include/linux/bpf.h > @@ -1458,6 +1458,9 @@ int dev_xdp_enqueue(struct net_device *dev, struct xdp_buff *xdp, > struct net_device *dev_rx); > int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, > struct net_device *dev_rx); > +bool dst_dev_is_ingress(struct bpf_dtab_netdev *obj, int ifindex); > +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, > + struct bpf_map *map, bool exclude_ingress); > int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, struct sk_buff *skb, > struct bpf_prog *xdp_prog); > bool dev_map_can_have_prog(struct bpf_map *map); > @@ -1630,6 +1633,19 @@ int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, > return 0; > } > > +static inline > +bool dst_dev_is_ingress(struct bpf_dtab_netdev *obj, int ifindex) > +{ > + return false; > +} > + > +static inline > +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, > + struct bpf_map *map, bool exclude_ingress) > +{ > + return 0; > +} > + > struct sk_buff; > > static inline int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, > diff --git a/include/net/xdp.h b/include/net/xdp.h > index a5bc214a49d9..5533f0ab2afc 100644 > --- a/include/net/xdp.h > +++ b/include/net/xdp.h > @@ -170,6 +170,7 @@ struct sk_buff *__xdp_build_skb_from_frame(struct xdp_frame *xdpf, > struct sk_buff *xdp_build_skb_from_frame(struct xdp_frame *xdpf, > struct net_device *dev); > int xdp_alloc_skb_bulk(void **skbs, int n_skb, gfp_t gfp); > +struct xdp_frame *xdpf_clone(struct xdp_frame *xdpf); > > static inline > void xdp_convert_frame_to_buff(struct xdp_frame *frame, struct xdp_buff *xdp) > diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h > index 2d3036e292a9..5982ceb217dc 100644 > --- a/include/uapi/linux/bpf.h > +++ b/include/uapi/linux/bpf.h > @@ -2508,8 +2508,12 @@ union bpf_attr { > * The lower two bits of *flags* are used as the return code if > * the map lookup fails. This is so that the return value can be > * one of the XDP program return codes up to **XDP_TX**, as chosen > - * by the caller. Any higher bits in the *flags* argument must be > - * unset. > + * by the caller. The higher bits of *flags* can be set to > + * BPF_F_BROADCAST or BPF_F_EXCLUDE_INGRESS as defined below. > + * > + * With BPF_F_BROADCAST the packet will be broadcasted to all the > + * interfaces in the map. with BPF_F_EXCLUDE_INGRESS the ingress > + * interface will be excluded when do broadcasting. > * > * See also **bpf_redirect**\ (), which only supports redirecting > * to an ifindex, but doesn't require a map to do so. > @@ -5004,6 +5008,15 @@ enum { > BPF_F_BPRM_SECUREEXEC = (1ULL << 0), > }; > > +/* Flags for bpf_redirect_map helper */ > +enum { > + BPF_F_BROADCAST = (1ULL << 3), > + BPF_F_EXCLUDE_INGRESS = (1ULL << 4), > +}; > + > +#define BPF_F_ACTION_MASK (XDP_ABORTED | XDP_DROP | XDP_PASS | XDP_TX) > +#define BPF_F_REDIR_MASK (BPF_F_BROADCAST | BPF_F_EXCLUDE_INGRESS) > + > #define __bpf_md_ptr(type, name) \ > union { \ > type name; \ > diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c > index f80cf5036d39..ad616a043d2a 100644 > --- a/kernel/bpf/devmap.c > +++ b/kernel/bpf/devmap.c > @@ -519,6 +519,125 @@ int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, > return __xdp_enqueue(dev, xdp, dev_rx, dst->xdp_prog); > } > > +/* Use direct call in fast path instead of map->ops->map_get_next_key() */ > +static int devmap_get_next_key(struct bpf_map *map, void *key, void *next_key) > +{ > + switch (map->map_type) { > + case BPF_MAP_TYPE_DEVMAP: > + return dev_map_get_next_key(map, key, next_key); > + case BPF_MAP_TYPE_DEVMAP_HASH: > + return dev_map_hash_get_next_key(map, key, next_key); > + default: > + break; > + } > + > + return -ENOENT; > +} > + > +bool dst_dev_is_ingress(struct bpf_dtab_netdev *dst, int ifindex) > +{ > + return dst->dev->ifindex == ifindex; > +} > + > +static struct bpf_dtab_netdev *devmap_get_next_obj(struct xdp_buff *xdp, > + struct bpf_map *map, > + u32 *key, u32 *next_key, > + int ex_ifindex) > +{ > + struct bpf_dtab_netdev *obj; > + struct net_device *dev; > + u32 *tmp_key = key; why is tmp_key needed? you're not using key for anything else, so you could just substitute that for all of the uses of tmp_key below? > + u32 index; > + int err; > + > + err = devmap_get_next_key(map, tmp_key, next_key); > + if (err) > + return NULL; > + > + /* When using dev map hash, we could restart the hashtab traversal > + * in case the key has been updated/removed in the mean time. > + * So we may end up potentially looping due to traversal restarts > + * from first elem. > + * > + * Let's use map's max_entries to limit the loop number. > + */ > + for (index = 0; index < map->max_entries; index++) { > + switch (map->map_type) { > + case BPF_MAP_TYPE_DEVMAP: > + obj = __dev_map_lookup_elem(map, *next_key); > + break; > + case BPF_MAP_TYPE_DEVMAP_HASH: > + obj = __dev_map_hash_lookup_elem(map, *next_key); > + break; > + default: > + break; > + } > + > + if (!obj || dst_dev_is_ingress(obj, ex_ifindex)) > + goto find_next; > + > + dev = obj->dev; > + > + if (!dev->netdev_ops->ndo_xdp_xmit) > + goto find_next; > + > + err = xdp_ok_fwd_dev(dev, xdp->data_end - xdp->data); > + if (unlikely(err)) > + goto find_next; > + > + return obj; > + > +find_next: > + tmp_key = next_key; > + err = devmap_get_next_key(map, tmp_key, next_key); > + if (err) > + break; > + } > + > + return NULL; > +} > + > +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, > + struct bpf_map *map, bool exclude_ingress) > +{ > + struct bpf_dtab_netdev *obj = NULL, *next_obj = NULL; > + struct xdp_frame *xdpf, *nxdpf; > + int ex_ifindex; > + u32 key, next_key; Out of reverse-xmas-tree order... -Toke
On Wed, Mar 17, 2021 at 01:03:02PM +0100, Toke Høiland-Jørgensen wrote: > FYI, this no longer applies to bpf-next due to Björn's refactor in > commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") Thanks Toke, I need to see how to get the map via map_id, does bpf_map_get_curr_or_next() works? Should I call bpf_map_put() after using? The ri->flags = flags also need to be add back as we need to use the flags value. This looks like an opposite of Björn's restructure... And I have expected another rebase after Lorenzo's "bpf: devmap: move drop error path to devmap for XDP_REDIRECT" > > Also, two small nits below: Thanks, I will fix them if there is a way to do the rebase. Hangbin
On Thu, 18 Mar 2021 at 04:52, Hangbin Liu <liuhangbin@gmail.com> wrote: > > On Wed, Mar 17, 2021 at 01:03:02PM +0100, Toke Høiland-Jørgensen wrote: > > FYI, this no longer applies to bpf-next due to Björn's refactor in > > commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") > > Thanks Toke, I need to see how to get the map via map_id, does > bpf_map_get_curr_or_next() works? Should I call bpf_map_put() after using? > > The ri->flags = flags also need to be add back as we need to use the flags > value. > Hmm, I was under the impression that ri->flags was only to be used by the new bpf_redirect_map_multi(), but now I see that you're planning to use the bpf_redirect_map instead. Well... I guess the flag is back then. Björn
Hangbin Liu <liuhangbin@gmail.com> writes: > On Wed, Mar 17, 2021 at 01:03:02PM +0100, Toke Høiland-Jørgensen wrote: >> FYI, this no longer applies to bpf-next due to Björn's refactor in >> commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") > > Thanks Toke, I need to see how to get the map via map_id, does > bpf_map_get_curr_or_next() works? Should I call bpf_map_put() after > using? I would expect that to be terrible for performance; I think it would be better to just add back the map pointer into struct bpf_redirect_info. If you only set the map pointer when the multicast flag is set, you can just check that pointer to disambiguate between when you need to call dev_map_enqueue() and dev_map_enqueue_multi(), in which case you don't need to add back the flags member... -Toke
On Thu, Mar 18, 2021 at 03:19:47PM +0100, Toke Høiland-Jørgensen wrote: > Hangbin Liu <liuhangbin@gmail.com> writes: > > > On Wed, Mar 17, 2021 at 01:03:02PM +0100, Toke Høiland-Jørgensen wrote: > >> FYI, this no longer applies to bpf-next due to Björn's refactor in > >> commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") > > > > Thanks Toke, I need to see how to get the map via map_id, does > > bpf_map_get_curr_or_next() works? Should I call bpf_map_put() after > > using? > > I would expect that to be terrible for performance; I think it would be > better to just add back the map pointer into struct bpf_redirect_info. > If you only set the map pointer when the multicast flag is set, you can > just check that pointer to disambiguate between when you need to call > dev_map_enqueue() and dev_map_enqueue_multi(), in which case you don't > need to add back the flags member... There are 2 flags, BROADCAST and EXCLUDE_INGRESS. There is no way to only check the map pointer and ignore flags.. Thanks Hangbin
Hangbin Liu <liuhangbin@gmail.com> writes: > On Thu, Mar 18, 2021 at 03:19:47PM +0100, Toke Høiland-Jørgensen wrote: >> Hangbin Liu <liuhangbin@gmail.com> writes: >> >> > On Wed, Mar 17, 2021 at 01:03:02PM +0100, Toke Høiland-Jørgensen wrote: >> >> FYI, this no longer applies to bpf-next due to Björn's refactor in >> >> commit: ee75aef23afe ("bpf, xdp: Restructure redirect actions") >> > >> > Thanks Toke, I need to see how to get the map via map_id, does >> > bpf_map_get_curr_or_next() works? Should I call bpf_map_put() after >> > using? >> >> I would expect that to be terrible for performance; I think it would be >> better to just add back the map pointer into struct bpf_redirect_info. >> If you only set the map pointer when the multicast flag is set, you can >> just check that pointer to disambiguate between when you need to call >> dev_map_enqueue() and dev_map_enqueue_multi(), in which case you don't >> need to add back the flags member... > > There are 2 flags, BROADCAST and EXCLUDE_INGRESS. There is no way > to only check the map pointer and ignore flags.. Ah, right, of course, my bad :) Well, in that case adding both members back is probably the right thing to do... -Toke
diff --git a/include/linux/bpf.h b/include/linux/bpf.h index c931bc97019d..bb07ccd170f2 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -1458,6 +1458,9 @@ int dev_xdp_enqueue(struct net_device *dev, struct xdp_buff *xdp, struct net_device *dev_rx); int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, struct net_device *dev_rx); +bool dst_dev_is_ingress(struct bpf_dtab_netdev *obj, int ifindex); +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, + struct bpf_map *map, bool exclude_ingress); int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, struct sk_buff *skb, struct bpf_prog *xdp_prog); bool dev_map_can_have_prog(struct bpf_map *map); @@ -1630,6 +1633,19 @@ int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, return 0; } +static inline +bool dst_dev_is_ingress(struct bpf_dtab_netdev *obj, int ifindex) +{ + return false; +} + +static inline +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, + struct bpf_map *map, bool exclude_ingress) +{ + return 0; +} + struct sk_buff; static inline int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, diff --git a/include/net/xdp.h b/include/net/xdp.h index a5bc214a49d9..5533f0ab2afc 100644 --- a/include/net/xdp.h +++ b/include/net/xdp.h @@ -170,6 +170,7 @@ struct sk_buff *__xdp_build_skb_from_frame(struct xdp_frame *xdpf, struct sk_buff *xdp_build_skb_from_frame(struct xdp_frame *xdpf, struct net_device *dev); int xdp_alloc_skb_bulk(void **skbs, int n_skb, gfp_t gfp); +struct xdp_frame *xdpf_clone(struct xdp_frame *xdpf); static inline void xdp_convert_frame_to_buff(struct xdp_frame *frame, struct xdp_buff *xdp) diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 2d3036e292a9..5982ceb217dc 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2508,8 +2508,12 @@ union bpf_attr { * The lower two bits of *flags* are used as the return code if * the map lookup fails. This is so that the return value can be * one of the XDP program return codes up to **XDP_TX**, as chosen - * by the caller. Any higher bits in the *flags* argument must be - * unset. + * by the caller. The higher bits of *flags* can be set to + * BPF_F_BROADCAST or BPF_F_EXCLUDE_INGRESS as defined below. + * + * With BPF_F_BROADCAST the packet will be broadcasted to all the + * interfaces in the map. with BPF_F_EXCLUDE_INGRESS the ingress + * interface will be excluded when do broadcasting. * * See also **bpf_redirect**\ (), which only supports redirecting * to an ifindex, but doesn't require a map to do so. @@ -5004,6 +5008,15 @@ enum { BPF_F_BPRM_SECUREEXEC = (1ULL << 0), }; +/* Flags for bpf_redirect_map helper */ +enum { + BPF_F_BROADCAST = (1ULL << 3), + BPF_F_EXCLUDE_INGRESS = (1ULL << 4), +}; + +#define BPF_F_ACTION_MASK (XDP_ABORTED | XDP_DROP | XDP_PASS | XDP_TX) +#define BPF_F_REDIR_MASK (BPF_F_BROADCAST | BPF_F_EXCLUDE_INGRESS) + #define __bpf_md_ptr(type, name) \ union { \ type name; \ diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c index f80cf5036d39..ad616a043d2a 100644 --- a/kernel/bpf/devmap.c +++ b/kernel/bpf/devmap.c @@ -519,6 +519,125 @@ int dev_map_enqueue(struct bpf_dtab_netdev *dst, struct xdp_buff *xdp, return __xdp_enqueue(dev, xdp, dev_rx, dst->xdp_prog); } +/* Use direct call in fast path instead of map->ops->map_get_next_key() */ +static int devmap_get_next_key(struct bpf_map *map, void *key, void *next_key) +{ + switch (map->map_type) { + case BPF_MAP_TYPE_DEVMAP: + return dev_map_get_next_key(map, key, next_key); + case BPF_MAP_TYPE_DEVMAP_HASH: + return dev_map_hash_get_next_key(map, key, next_key); + default: + break; + } + + return -ENOENT; +} + +bool dst_dev_is_ingress(struct bpf_dtab_netdev *dst, int ifindex) +{ + return dst->dev->ifindex == ifindex; +} + +static struct bpf_dtab_netdev *devmap_get_next_obj(struct xdp_buff *xdp, + struct bpf_map *map, + u32 *key, u32 *next_key, + int ex_ifindex) +{ + struct bpf_dtab_netdev *obj; + struct net_device *dev; + u32 *tmp_key = key; + u32 index; + int err; + + err = devmap_get_next_key(map, tmp_key, next_key); + if (err) + return NULL; + + /* When using dev map hash, we could restart the hashtab traversal + * in case the key has been updated/removed in the mean time. + * So we may end up potentially looping due to traversal restarts + * from first elem. + * + * Let's use map's max_entries to limit the loop number. + */ + for (index = 0; index < map->max_entries; index++) { + switch (map->map_type) { + case BPF_MAP_TYPE_DEVMAP: + obj = __dev_map_lookup_elem(map, *next_key); + break; + case BPF_MAP_TYPE_DEVMAP_HASH: + obj = __dev_map_hash_lookup_elem(map, *next_key); + break; + default: + break; + } + + if (!obj || dst_dev_is_ingress(obj, ex_ifindex)) + goto find_next; + + dev = obj->dev; + + if (!dev->netdev_ops->ndo_xdp_xmit) + goto find_next; + + err = xdp_ok_fwd_dev(dev, xdp->data_end - xdp->data); + if (unlikely(err)) + goto find_next; + + return obj; + +find_next: + tmp_key = next_key; + err = devmap_get_next_key(map, tmp_key, next_key); + if (err) + break; + } + + return NULL; +} + +int dev_map_enqueue_multi(struct xdp_buff *xdp, struct net_device *dev_rx, + struct bpf_map *map, bool exclude_ingress) +{ + struct bpf_dtab_netdev *obj = NULL, *next_obj = NULL; + struct xdp_frame *xdpf, *nxdpf; + int ex_ifindex; + u32 key, next_key; + + ex_ifindex = exclude_ingress ? dev_rx->ifindex : 0; + + /* Find first available obj */ + obj = devmap_get_next_obj(xdp, map, NULL, &key, ex_ifindex); + if (!obj) + return -ENOENT; + + xdpf = xdp_convert_buff_to_frame(xdp); + if (unlikely(!xdpf)) + return -EOVERFLOW; + + for (;;) { + /* Check if we still have one more available obj */ + next_obj = devmap_get_next_obj(xdp, map, &key, &next_key, ex_ifindex); + if (!next_obj) { + bq_enqueue(obj->dev, xdpf, dev_rx, obj->xdp_prog); + return 0; + } + + nxdpf = xdpf_clone(xdpf); + if (unlikely(!nxdpf)) { + xdp_return_frame_rx_napi(xdpf); + return -ENOMEM; + } + + bq_enqueue(obj->dev, nxdpf, dev_rx, obj->xdp_prog); + + /* Deal with next obj */ + obj = next_obj; + key = next_key; + } +} + int dev_map_generic_redirect(struct bpf_dtab_netdev *dst, struct sk_buff *skb, struct bpf_prog *xdp_prog) { diff --git a/net/core/filter.c b/net/core/filter.c index 588b19ba0da8..3736a9116be9 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3919,12 +3919,17 @@ static const struct bpf_func_proto bpf_xdp_adjust_meta_proto = { }; static int __bpf_tx_xdp_map(struct net_device *dev_rx, void *fwd, - struct bpf_map *map, struct xdp_buff *xdp) + struct bpf_map *map, struct xdp_buff *xdp, + u32 flags) { switch (map->map_type) { case BPF_MAP_TYPE_DEVMAP: case BPF_MAP_TYPE_DEVMAP_HASH: - return dev_map_enqueue(fwd, xdp, dev_rx); + if (flags & BPF_F_BROADCAST) + return dev_map_enqueue_multi(xdp, dev_rx, map, + flags & BPF_F_EXCLUDE_INGRESS); + else + return dev_map_enqueue(fwd, xdp, dev_rx); case BPF_MAP_TYPE_CPUMAP: return cpu_map_enqueue(fwd, xdp, dev_rx); case BPF_MAP_TYPE_XSKMAP: @@ -3998,7 +4003,7 @@ int xdp_do_redirect(struct net_device *dev, struct xdp_buff *xdp, err = dev_xdp_enqueue(fwd, xdp, dev); } else { - err = __bpf_tx_xdp_map(dev, fwd, map, xdp); + err = __bpf_tx_xdp_map(dev, fwd, map, xdp, ri->flags); } if (unlikely(err)) @@ -4012,6 +4017,57 @@ int xdp_do_redirect(struct net_device *dev, struct xdp_buff *xdp, } EXPORT_SYMBOL_GPL(xdp_do_redirect); +static int dev_map_redirect_multi(struct net_device *dev, struct sk_buff *skb, + struct bpf_prog *xdp_prog, struct bpf_map *map, + bool exclude_ingress) +{ + struct bpf_dtab_netdev *dst; + u32 key, next_key, index; + struct sk_buff *nskb; + void *fwd; + int err; + + err = map->ops->map_get_next_key(map, NULL, &key); + if (err) + return err; + + /* When using dev map hash, we could restart the hashtab traversal + * in case the key has been updated/removed in the mean time. + * So we may end up potentially looping due to traversal restarts + * from first elem. + * + * Let's use map's max_entries to limit the loop number. + */ + + for (index = 0; index < map->max_entries; index++) { + fwd = __xdp_map_lookup_elem(map, key); + if (fwd) { + dst = (struct bpf_dtab_netdev *)fwd; + if (dst_dev_is_ingress(dst, exclude_ingress ? dev->ifindex : 0)) + goto find_next; + + nskb = skb_clone(skb, GFP_ATOMIC); + if (!nskb) + return -ENOMEM; + + /* Try forword next one no mater the current forward + * succeed or not. + */ + dev_map_generic_redirect(dst, nskb, xdp_prog); + } + +find_next: + err = map->ops->map_get_next_key(map, &key, &next_key); + if (err) + break; + + key = next_key; + } + + consume_skb(skb); + return 0; +} + static int xdp_do_generic_redirect_map(struct net_device *dev, struct sk_buff *skb, struct xdp_buff *xdp, @@ -4031,7 +4087,11 @@ static int xdp_do_generic_redirect_map(struct net_device *dev, map->map_type == BPF_MAP_TYPE_DEVMAP_HASH) { struct bpf_dtab_netdev *dst = fwd; - err = dev_map_generic_redirect(dst, skb, xdp_prog); + if (ri->flags & BPF_F_BROADCAST) + err = dev_map_redirect_multi(dev, skb, xdp_prog, map, + ri->flags & BPF_F_EXCLUDE_INGRESS); + else + err = dev_map_generic_redirect(dst, skb, xdp_prog); if (unlikely(err)) goto err; } else if (map->map_type == BPF_MAP_TYPE_XSKMAP) { @@ -4115,18 +4175,18 @@ BPF_CALL_3(bpf_xdp_redirect_map, struct bpf_map *, map, u32, ifindex, struct bpf_redirect_info *ri = this_cpu_ptr(&bpf_redirect_info); /* Lower bits of the flags are used as return code on lookup failure */ - if (unlikely(flags > XDP_TX)) + if (unlikely(flags & ~(BPF_F_ACTION_MASK | BPF_F_REDIR_MASK))) return XDP_ABORTED; ri->tgt_value = __xdp_map_lookup_elem(map, ifindex); - if (unlikely(!ri->tgt_value)) { + if (unlikely(!ri->tgt_value) && !(flags & BPF_F_BROADCAST)) { /* If the lookup fails we want to clear out the state in the * redirect_info struct completely, so that if an eBPF program * performs multiple lookups, the last one always takes * precedence. */ WRITE_ONCE(ri->map, NULL); - return flags; + return flags & BPF_F_ACTION_MASK; } ri->flags = flags; diff --git a/net/core/xdp.c b/net/core/xdp.c index 05354976c1fc..aba84d04642b 100644 --- a/net/core/xdp.c +++ b/net/core/xdp.c @@ -583,3 +583,32 @@ struct sk_buff *xdp_build_skb_from_frame(struct xdp_frame *xdpf, return __xdp_build_skb_from_frame(xdpf, skb, dev); } EXPORT_SYMBOL_GPL(xdp_build_skb_from_frame); + +struct xdp_frame *xdpf_clone(struct xdp_frame *xdpf) +{ + unsigned int headroom, totalsize; + struct xdp_frame *nxdpf; + struct page *page; + void *addr; + + headroom = xdpf->headroom + sizeof(*xdpf); + totalsize = headroom + xdpf->len; + + if (unlikely(totalsize > PAGE_SIZE)) + return NULL; + page = dev_alloc_page(); + if (!page) + return NULL; + addr = page_to_virt(page); + + memcpy(addr, xdpf, totalsize); + + nxdpf = addr; + nxdpf->data = addr + headroom; + nxdpf->frame_sz = PAGE_SIZE; + nxdpf->mem.type = MEM_TYPE_PAGE_ORDER0; + nxdpf->mem.id = 0; + + return nxdpf; +} +EXPORT_SYMBOL_GPL(xdpf_clone); diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 2d3036e292a9..5982ceb217dc 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -2508,8 +2508,12 @@ union bpf_attr { * The lower two bits of *flags* are used as the return code if * the map lookup fails. This is so that the return value can be * one of the XDP program return codes up to **XDP_TX**, as chosen - * by the caller. Any higher bits in the *flags* argument must be - * unset. + * by the caller. The higher bits of *flags* can be set to + * BPF_F_BROADCAST or BPF_F_EXCLUDE_INGRESS as defined below. + * + * With BPF_F_BROADCAST the packet will be broadcasted to all the + * interfaces in the map. with BPF_F_EXCLUDE_INGRESS the ingress + * interface will be excluded when do broadcasting. * * See also **bpf_redirect**\ (), which only supports redirecting * to an ifindex, but doesn't require a map to do so. @@ -5004,6 +5008,15 @@ enum { BPF_F_BPRM_SECUREEXEC = (1ULL << 0), }; +/* Flags for bpf_redirect_map helper */ +enum { + BPF_F_BROADCAST = (1ULL << 3), + BPF_F_EXCLUDE_INGRESS = (1ULL << 4), +}; + +#define BPF_F_ACTION_MASK (XDP_ABORTED | XDP_DROP | XDP_PASS | XDP_TX) +#define BPF_F_REDIR_MASK (BPF_F_BROADCAST | BPF_F_EXCLUDE_INGRESS) + #define __bpf_md_ptr(type, name) \ union { \ type name; \
This patch add two flags BPF_F_BROADCAST and BPF_F_EXCLUDE_INGRESS to extend xdp_redirect_map for broadcast support. Keep the general data path in net/core/filter.c and the native data path in kernel/bpf/devmap.c so we can use direct calls to get better performace. Here is the performance result by using xdp_redirect_{map, map_multi} in sample/bpf and send pkts via pktgen cmd: ./pktgen_sample03_burst_single_flow.sh -i eno1 -d $dst_ip -m $dst_mac -t 10 -s 64 There are some drop back as we need to loop the map and get each interface. Version | Test | Generic | Native 5.11 | redirect_map i40e->i40e | 1.9M | 9.3M 5.11 | redirect_map i40e->veth | 1.5M | 11.2M 5.11 + patch | redirect_map i40e->i40e | 1.9M | 9.6M 5.11 + patch | redirect_map i40e->veth | 1.5M | 11.9M 5.11 + patch | redirect_map_multi i40e->i40e | 1.5M | 7.7M 5.11 + patch | redirect_map_multi i40e->veth | 1.2M | 9.1M 5.11 + patch | redirect_map_multi i40e->mlx4+veth | 0.9M | 3.2M v2: fix flag renaming issue in v1 Signed-off-by: Hangbin Liu <liuhangbin@gmail.com> --- include/linux/bpf.h | 16 +++++ include/net/xdp.h | 1 + include/uapi/linux/bpf.h | 17 ++++- kernel/bpf/devmap.c | 119 +++++++++++++++++++++++++++++++++ net/core/filter.c | 74 ++++++++++++++++++-- net/core/xdp.c | 29 ++++++++ tools/include/uapi/linux/bpf.h | 17 ++++- 7 files changed, 262 insertions(+), 11 deletions(-)