aboutsummaryrefslogtreecommitdiff
path: root/net
diff options
context:
space:
mode:
authorGravatar Paolo Abeni <pabeni@redhat.com> 2023-02-16 13:30:06 +0100
committerGravatar Paolo Abeni <pabeni@redhat.com> 2023-02-16 13:30:06 +0100
commit40967f77dfa9fa728b7f36a5d2eb432f39de185c (patch)
treebb97f1a1c4c53e13d72df3aa40babf9ef28c2392 /net
parentnet: dsa: ocelot: fix selecting MFD_OCELOT (diff)
parentselftests: seg6: add selftest for PSP flavor in SRv6 End behavior (diff)
downloadlinux-40967f77dfa9fa728b7f36a5d2eb432f39de185c.tar.gz
linux-40967f77dfa9fa728b7f36a5d2eb432f39de185c.tar.bz2
linux-40967f77dfa9fa728b7f36a5d2eb432f39de185c.zip
Merge branch 'seg6-add-psp-flavor-support-for-srv6-end-behavior'
Andrea Mayer says: ==================== seg6: add PSP flavor support for SRv6 End behavior Segment Routing for IPv6 (SRv6 in short) [1] is the instantiation of the Segment Routing (SR) [2] architecture on the IPv6 dataplane. In SRv6, the segment identifiers (SID) are IPv6 addresses and the segment list (SID List) is carried in the Segment Routing Header (SRH). A segment may be bound to a specific packet processing operation called "behavior". The RFC8986 [3] defines and standardizes the most common/relevant behaviors for network operators, e.g., End, End.X and End.T and so on. The RFC8986 also introduces the "flavors" framework aiming to modify or extend the capabilities of SRv6 End, End.X and End.T behaviors. Specifically, these behaviors support the following flavors (either individually or in combinations): - Penultimate Segment Pop (PSP); - Ultimate Segment Pop (USP); - Ultimate Segment Decapsulation (USD). Such flavors enable an End/End.X/End.T behavior to pop the SRH on the penultimate/ultimate SR endpoint node listed in the SID List or to perform a full decapsulation. Currently, the Linux kernel supports a large subset of behaviors described in RFC8986, including the End, End.X and End.T. However, PSP, USP and USD flavors have not yet been implemented. In this patchset, we extend the SRv6 subsystem to implement the PSP flavor in the SRv6 End behavior. To accomplish this task, we leverage the flavor framework previously introduced by another patchset required for supporting the efficient representation of the SID List through the NEXT-C-SID mechanism [4]. In details, the patchset is made of: - patch 1/3: seg6: factor out End lookup nexthop processing to a dedicated function - patch 2/3: seg6: add PSP flavor support for SRv6 End behavior - patch 3/3: selftests: seg6: add selftest for PSP flavor in SRv6 End behavior From the user space perspective, we do not need to change the iproute2 code to support the PSP flavor. However, we provide the man page for the PSP flavor in a separate patch. Comments, improvements and suggestions are always appreciated. [1] - RFC8754: https://datatracker.ietf.org/doc/html/rfc8754 [2] - RFC8402: https://datatracker.ietf.org/doc/html/rfc8402 [3] - RFC8986: https://datatracker.ietf.org/doc/html/rfc8986 [4] - https://datatracker.ietf.org/doc/html/draft-ietf-spring-srv6-srh-compression ==================== Link: https://lore.kernel.org/r/20230215134659.7613-1-andrea.mayer@uniroma2.it Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Diffstat (limited to 'net')
-rw-r--r--net/ipv6/seg6_local.c352
1 files changed, 343 insertions, 9 deletions
diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c
index 487f8e98deaa..dd433cc265c8 100644
--- a/net/ipv6/seg6_local.c
+++ b/net/ipv6/seg6_local.c
@@ -109,8 +109,15 @@ struct bpf_lwt_prog {
#define next_csid_chk_lcnode_fn_bits(flen) \
next_csid_chk_lcblock_bits(flen)
+#define SEG6_F_LOCAL_FLV_OP(flvname) BIT(SEG6_LOCAL_FLV_OP_##flvname)
+#define SEG6_F_LOCAL_FLV_PSP SEG6_F_LOCAL_FLV_OP(PSP)
+
+/* Supported RFC8986 Flavor operations are reported in this bitmask */
+#define SEG6_LOCAL_FLV8986_SUPP_OPS SEG6_F_LOCAL_FLV_PSP
+
/* Supported Flavor operations are reported in this bitmask */
-#define SEG6_LOCAL_FLV_SUPP_OPS (BIT(SEG6_LOCAL_FLV_OP_NEXT_CSID))
+#define SEG6_LOCAL_FLV_SUPP_OPS (SEG6_F_LOCAL_FLV_OP(NEXT_CSID) | \
+ SEG6_LOCAL_FLV8986_SUPP_OPS)
struct seg6_flavors_info {
/* Flavor operations */
@@ -364,6 +371,14 @@ static void seg6_next_csid_advance_arg(struct in6_addr *addr,
memset(&addr->s6_addr[16 - fnc_octects], 0x00, fnc_octects);
}
+static int input_action_end_finish(struct sk_buff *skb,
+ struct seg6_local_lwt *slwt)
+{
+ seg6_lookup_nexthop(skb, NULL, 0);
+
+ return dst_input(skb);
+}
+
static int input_action_end_core(struct sk_buff *skb,
struct seg6_local_lwt *slwt)
{
@@ -375,9 +390,7 @@ static int input_action_end_core(struct sk_buff *skb,
advance_nextseg(srh, &ipv6_hdr(skb)->daddr);
- seg6_lookup_nexthop(skb, NULL, 0);
-
- return dst_input(skb);
+ return input_action_end_finish(skb, slwt);
drop:
kfree_skb(skb);
@@ -395,9 +408,7 @@ static int end_next_csid_core(struct sk_buff *skb, struct seg6_local_lwt *slwt)
/* update DA */
seg6_next_csid_advance_arg(daddr, finfo);
- seg6_lookup_nexthop(skb, NULL, 0);
-
- return dst_input(skb);
+ return input_action_end_finish(skb, slwt);
}
static bool seg6_next_csid_enabled(__u32 fops)
@@ -405,15 +416,331 @@ static bool seg6_next_csid_enabled(__u32 fops)
return fops & BIT(SEG6_LOCAL_FLV_OP_NEXT_CSID);
}
+/* We describe the packet state in relation to the absence/presence of the SRH
+ * and the Segment Left (SL) field.
+ * For our purposes, it is not necessary to record the exact value of the SL
+ * when the SID List consists of two or more segments.
+ */
+enum seg6_local_pktinfo {
+ /* the order really matters! */
+ SEG6_LOCAL_PKTINFO_NOHDR = 0,
+ SEG6_LOCAL_PKTINFO_SL_ZERO,
+ SEG6_LOCAL_PKTINFO_SL_ONE,
+ SEG6_LOCAL_PKTINFO_SL_MORE,
+ __SEG6_LOCAL_PKTINFO_MAX,
+};
+
+#define SEG6_LOCAL_PKTINFO_MAX (__SEG6_LOCAL_PKTINFO_MAX - 1)
+
+static enum seg6_local_pktinfo seg6_get_srh_pktinfo(struct ipv6_sr_hdr *srh)
+{
+ __u8 sgl;
+
+ if (!srh)
+ return SEG6_LOCAL_PKTINFO_NOHDR;
+
+ sgl = srh->segments_left;
+ if (sgl < 2)
+ return SEG6_LOCAL_PKTINFO_SL_ZERO + sgl;
+
+ return SEG6_LOCAL_PKTINFO_SL_MORE;
+}
+
+enum seg6_local_flv_action {
+ SEG6_LOCAL_FLV_ACT_UNSPEC = 0,
+ SEG6_LOCAL_FLV_ACT_END,
+ SEG6_LOCAL_FLV_ACT_PSP,
+ SEG6_LOCAL_FLV_ACT_USP,
+ SEG6_LOCAL_FLV_ACT_USD,
+ __SEG6_LOCAL_FLV_ACT_MAX
+};
+
+#define SEG6_LOCAL_FLV_ACT_MAX (__SEG6_LOCAL_FLV_ACT_MAX - 1)
+
+/* The action table for RFC8986 flavors (see the flv8986_act_tbl below)
+ * contains the actions (i.e. processing operations) to be applied on packets
+ * when flavors are configured for an End* behavior.
+ * By combining the pkinfo data and from the flavors mask, the macro
+ * computes the index used to access the elements (actions) stored in the
+ * action table. The index is structured as follows:
+ *
+ * index
+ * _______________/\________________
+ * / \
+ * +----------------+----------------+
+ * | pf | afm |
+ * +----------------+----------------+
+ * ph-1 ... p1 p0 fk-1 ... f1 f0
+ * MSB LSB
+ *
+ * where:
+ * - 'afm' (adjusted flavor mask) is the mask containing a combination of the
+ * RFC8986 flavors currently supported. 'afm' corresponds to the @fm
+ * argument of the macro whose value is righ-shifted by 1 bit. By doing so,
+ * we discard the SEG6_LOCAL_FLV_OP_UNSPEC flag (bit 0 in @fm) which is
+ * never used here;
+ * - 'pf' encodes the packet info (pktinfo) regarding the presence/absence of
+ * the SRH, SL = 0, etc. 'pf' is set with the value of @pf provided as
+ * argument to the macro.
+ */
+#define flv8986_act_tbl_idx(pf, fm) \
+ ((((pf) << bits_per(SEG6_LOCAL_FLV8986_SUPP_OPS)) | \
+ ((fm) & SEG6_LOCAL_FLV8986_SUPP_OPS)) >> SEG6_LOCAL_FLV_OP_PSP)
+
+/* We compute the size of the action table by considering the RFC8986 flavors
+ * actually supported by the kernel. In this way, the size is automatically
+ * adjusted when new flavors are supported.
+ */
+#define FLV8986_ACT_TBL_SIZE \
+ roundup_pow_of_two(flv8986_act_tbl_idx(SEG6_LOCAL_PKTINFO_MAX, \
+ SEG6_LOCAL_FLV8986_SUPP_OPS))
+
+/* tbl_cfg(act, pf, fm) macro is used to easily configure the action
+ * table; it accepts 3 arguments:
+ * i) @act, the suffix from SEG6_LOCAL_FLV_ACT_{act} representing
+ * the action that should be applied on the packet;
+ * ii) @pf, the suffix from SEG6_LOCAL_PKTINFO_{pf} reporting the packet
+ * info about the lack/presence of SRH, SRH with SL = 0, etc;
+ * iii) @fm, the mask of flavors.
+ */
+#define tbl_cfg(act, pf, fm) \
+ [flv8986_act_tbl_idx(SEG6_LOCAL_PKTINFO_##pf, \
+ (fm))] = SEG6_LOCAL_FLV_ACT_##act
+
+/* shorthand for improving readability */
+#define F_PSP SEG6_F_LOCAL_FLV_PSP
+
+/* The table contains, for each combination of the pktinfo data and
+ * flavors, the action that should be taken on a packet (e.g.
+ * "standard" Endpoint processing, Penultimate Segment Pop, etc).
+ *
+ * By default, table entries not explicitly configured are initialized with the
+ * SEG6_LOCAL_FLV_ACT_UNSPEC action, which generally has the effect of
+ * discarding the processed packet.
+ */
+static const u8 flv8986_act_tbl[FLV8986_ACT_TBL_SIZE] = {
+ /* PSP variant for packet where SRH with SL = 1 */
+ tbl_cfg(PSP, SL_ONE, F_PSP),
+ /* End for packet where the SRH with SL > 1*/
+ tbl_cfg(END, SL_MORE, F_PSP),
+};
+
+#undef F_PSP
+#undef tbl_cfg
+
+/* For each flavor defined in RFC8986 (or a combination of them) an action is
+ * performed on the packet. The specific action depends on:
+ * - info extracted from the packet (i.e. pktinfo data) regarding the
+ * lack/presence of the SRH, and if the SRH is available, on the value of
+ * Segment Left field;
+ * - the mask of flavors configured for the specific SRv6 End* behavior.
+ *
+ * The function combines both the pkinfo and the flavors mask to evaluate the
+ * corresponding action to be taken on the packet.
+ */
+static enum seg6_local_flv_action
+seg6_local_flv8986_act_lookup(enum seg6_local_pktinfo pinfo, __u32 flvmask)
+{
+ unsigned long index;
+
+ /* check if the provided mask of flavors is supported */
+ if (unlikely(flvmask & ~SEG6_LOCAL_FLV8986_SUPP_OPS))
+ return SEG6_LOCAL_FLV_ACT_UNSPEC;
+
+ index = flv8986_act_tbl_idx(pinfo, flvmask);
+ if (unlikely(index >= FLV8986_ACT_TBL_SIZE))
+ return SEG6_LOCAL_FLV_ACT_UNSPEC;
+
+ return flv8986_act_tbl[index];
+}
+
+/* skb->data must be aligned with skb->network_header */
+static bool seg6_pop_srh(struct sk_buff *skb, int srhoff)
+{
+ struct ipv6_sr_hdr *srh;
+ struct ipv6hdr *iph;
+ __u8 srh_nexthdr;
+ int thoff = -1;
+ int srhlen;
+ int nhlen;
+
+ if (unlikely(srhoff < sizeof(*iph) ||
+ !pskb_may_pull(skb, srhoff + sizeof(*srh))))
+ return false;
+
+ srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
+ srhlen = ipv6_optlen(srh);
+
+ /* we are about to mangle the pkt, let's check if we can write on it */
+ if (unlikely(skb_ensure_writable(skb, srhoff + srhlen)))
+ return false;
+
+ /* skb_ensure_writable() may change skb pointers; evaluate srh again */
+ srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
+ srh_nexthdr = srh->nexthdr;
+
+ if (unlikely(!skb_transport_header_was_set(skb)))
+ goto pull;
+
+ nhlen = skb_network_header_len(skb);
+ /* we have to deal with the transport header: it could be set before
+ * the SRH, after the SRH, or within it (which is considered wrong,
+ * however).
+ */
+ if (likely(nhlen <= srhoff))
+ thoff = nhlen;
+ else if (nhlen >= srhoff + srhlen)
+ /* transport_header is set after the SRH */
+ thoff = nhlen - srhlen;
+ else
+ /* transport_header falls inside the SRH; hence, we can't
+ * restore the transport_header pointer properly after
+ * SRH removing operation.
+ */
+ return false;
+pull:
+ /* we need to pop the SRH:
+ * 1) first of all, we pull out everything from IPv6 header up to SRH
+ * (included) evaluating also the rcsum;
+ * 2) we overwrite (and then remove) the SRH by properly moving the
+ * IPv6 along with any extension header that precedes the SRH;
+ * 3) At the end, we push back the pulled headers (except for SRH,
+ * obviously).
+ */
+ skb_pull_rcsum(skb, srhoff + srhlen);
+ memmove(skb_network_header(skb) + srhlen, skb_network_header(skb),
+ srhoff);
+ skb_push(skb, srhoff);
+
+ skb_reset_network_header(skb);
+ skb_mac_header_rebuild(skb);
+ if (likely(thoff >= 0))
+ skb_set_transport_header(skb, thoff);
+
+ iph = ipv6_hdr(skb);
+ if (iph->nexthdr == NEXTHDR_ROUTING) {
+ iph->nexthdr = srh_nexthdr;
+ } else {
+ /* we must look for the extension header (EXTH, for short) that
+ * immediately precedes the SRH we have just removed.
+ * Then, we update the value of the EXTH nexthdr with the one
+ * contained in the SRH nexthdr.
+ */
+ unsigned int off = sizeof(*iph);
+ struct ipv6_opt_hdr *hp, _hdr;
+ __u8 nexthdr = iph->nexthdr;
+
+ for (;;) {
+ if (unlikely(!ipv6_ext_hdr(nexthdr) ||
+ nexthdr == NEXTHDR_NONE))
+ return false;
+
+ hp = skb_header_pointer(skb, off, sizeof(_hdr), &_hdr);
+ if (unlikely(!hp))
+ return false;
+
+ if (hp->nexthdr == NEXTHDR_ROUTING) {
+ hp->nexthdr = srh_nexthdr;
+ break;
+ }
+
+ switch (nexthdr) {
+ case NEXTHDR_FRAGMENT:
+ fallthrough;
+ case NEXTHDR_AUTH:
+ /* we expect SRH before FRAG and AUTH */
+ return false;
+ default:
+ off += ipv6_optlen(hp);
+ break;
+ }
+
+ nexthdr = hp->nexthdr;
+ }
+ }
+
+ iph->payload_len = htons(skb->len - sizeof(struct ipv6hdr));
+
+ skb_postpush_rcsum(skb, iph, srhoff);
+
+ return true;
+}
+
+/* process the packet on the basis of the RFC8986 flavors set for the given
+ * SRv6 End behavior instance.
+ */
+static int end_flv8986_core(struct sk_buff *skb, struct seg6_local_lwt *slwt)
+{
+ const struct seg6_flavors_info *finfo = &slwt->flv_info;
+ enum seg6_local_flv_action action;
+ enum seg6_local_pktinfo pinfo;
+ struct ipv6_sr_hdr *srh;
+ __u32 flvmask;
+ int srhoff;
+
+ srh = seg6_get_srh(skb, 0);
+ srhoff = srh ? ((unsigned char *)srh - skb->data) : 0;
+ pinfo = seg6_get_srh_pktinfo(srh);
+#ifdef CONFIG_IPV6_SEG6_HMAC
+ if (srh && !seg6_hmac_validate_skb(skb))
+ goto drop;
+#endif
+ flvmask = finfo->flv_ops;
+ if (unlikely(flvmask & ~SEG6_LOCAL_FLV8986_SUPP_OPS)) {
+ pr_warn_once("seg6local: invalid RFC8986 flavors\n");
+ goto drop;
+ }
+
+ /* retrieve the action triggered by the combination of pktinfo data and
+ * the flavors mask.
+ */
+ action = seg6_local_flv8986_act_lookup(pinfo, flvmask);
+ switch (action) {
+ case SEG6_LOCAL_FLV_ACT_END:
+ /* process the packet as the "standard" End behavior */
+ advance_nextseg(srh, &ipv6_hdr(skb)->daddr);
+ break;
+ case SEG6_LOCAL_FLV_ACT_PSP:
+ advance_nextseg(srh, &ipv6_hdr(skb)->daddr);
+
+ if (unlikely(!seg6_pop_srh(skb, srhoff)))
+ goto drop;
+ break;
+ case SEG6_LOCAL_FLV_ACT_UNSPEC:
+ fallthrough;
+ default:
+ /* by default, we drop the packet since we could not find a
+ * suitable action.
+ */
+ goto drop;
+ }
+
+ return input_action_end_finish(skb, slwt);
+
+drop:
+ kfree_skb(skb);
+ return -EINVAL;
+}
+
/* regular endpoint function */
static int input_action_end(struct sk_buff *skb, struct seg6_local_lwt *slwt)
{
const struct seg6_flavors_info *finfo = &slwt->flv_info;
+ __u32 fops = finfo->flv_ops;
- if (seg6_next_csid_enabled(finfo->flv_ops))
+ if (!fops)
+ return input_action_end_core(skb, slwt);
+
+ /* check for the presence of NEXT-C-SID since it applies first */
+ if (seg6_next_csid_enabled(fops))
return end_next_csid_core(skb, slwt);
- return input_action_end_core(skb, slwt);
+ /* the specific processing function to be performed on the packet
+ * depends on the combination of flavors defined in RFC8986 and some
+ * information extracted from the packet, e.g. presence/absence of SRH,
+ * Segment Left = 0, etc.
+ */
+ return end_flv8986_core(skb, slwt);
}
/* regular endpoint, and forward to specified nexthop */
@@ -2300,6 +2627,13 @@ int __init seg6_local_init(void)
BUILD_BUG_ON(next_csid_chk_lcblock_bits(SEG6_LOCAL_LCBLOCK_DBITS));
BUILD_BUG_ON(next_csid_chk_lcnode_fn_bits(SEG6_LOCAL_LCNODE_FN_DBITS));
+ /* To be memory efficient, we use 'u8' to represent the different
+ * actions related to RFC8986 flavors. If the kernel build stops here,
+ * it means that it is not possible to correctly encode these actions
+ * with the data type chosen for the action table.
+ */
+ BUILD_BUG_ON(SEG6_LOCAL_FLV_ACT_MAX > (typeof(flv8986_act_tbl[0]))~0U);
+
return lwtunnel_encap_add_ops(&seg6_local_ops,
LWTUNNEL_ENCAP_SEG6_LOCAL);
}