fq_codel: generalise ce_threshold marking for subset of traffic

Commit e72aeb9e ("fq_codel: implement L4S style ce_threshold_ect1
marking") expanded the ce_threshold feature of FQ-CoDel so it can
be applied to a subset of the traffic, using the ECT(1) bit of the ECN
field as the classifier. However, hard-coding ECT(1) as the only
classifier for this feature seems limiting, so let's expand it to be more
general.

To this end, change the parameter from a ce_threshold_ect1 boolean, to a
one-byte selector/mask pair (ce_threshold_{selector,mask}) which is applied
to the whole diffserv/ECN field in the IP header. This makes it possible to
classify packets by any value in either the ECN field or the diffserv
field. In particular, setting a selector of INET_ECN_ECT_1 and a mask of
INET_ECN_MASK corresponds to the functionality before this patch, and a
mask of ~INET_ECN_MASK allows using the selector as a straight-forward
match against a diffserv code point:

 # apply ce_threshold to ECT(1) traffic
 tc qdisc replace dev eth0 root fq_codel ce_threshold 1ms ce_threshold_selector 0x1/0x3

 # apply ce_threshold to ECN-capable traffic marked as diffserv AF22
 tc qdisc replace dev eth0 root fq_codel ce_threshold 1ms ce_threshold_selector 0x50/0xfc

Regardless of the selector chosen, the normal rules for ECN-marking of
packets still apply, i.e., the flow must still declare itself ECN-capable
by setting one of the bits in the ECN field to get marked at all.

v2:
- Add tc usage examples to patch description
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20211019174709.69081-1-toke@redhat.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

include/net/codel.h

@@ -102,7 +102,9 @@ static inline u32 codel_time_to_us(codel_time_t val)
  * @interval:	width of moving time window
  * @mtu:	device mtu, or minimal queue backlog in bytes.
  * @ecn:	is Explicit Congestion Notification enabled
- * @ce_threshold_ect1: if ce_threshold only marks ECT(1) packets
+ * @ce_threshold_selector: apply ce_threshold to packets matching this value
+ *                         in the diffserv/ECN byte of the IP header
+ * @ce_threshold_mask: mask to apply to ce_threshold_selector comparison
  */
 struct codel_params {
 	codel_time_t	target;
@@ -110,7 +112,8 @@ struct codel_params {
 	codel_time_t	interval;
 	u32		mtu;
 	bool		ecn;
-	bool		ce_threshold_ect1;
+	u8		ce_threshold_selector;
+	u8		ce_threshold_mask;
 };
 
 /**

include/net/codel_impl.h

@@ -54,7 +54,8 @@ static void codel_params_init(struct codel_params *params)
 	params->interval = MS2TIME(100);
 	params->target = MS2TIME(5);
 	params->ce_threshold = CODEL_DISABLED_THRESHOLD;
-	params->ce_threshold_ect1 = false;
+	params->ce_threshold_mask = 0;
+	params->ce_threshold_selector = 0;
 	params->ecn = false;
 }
 
@@ -250,13 +251,12 @@ static struct sk_buff *codel_dequeue(void *ctx,
 	if (skb && codel_time_after(vars->ldelay, params->ce_threshold)) {
 		bool set_ce = true;
 
-		if (params->ce_threshold_ect1) {
-			/* Note: if skb_get_dsfield() returns -1, following
-			 * gives INET_ECN_MASK, which is != INET_ECN_ECT_1.
-			 */
-			u8 ecn = skb_get_dsfield(skb) & INET_ECN_MASK;
+		if (params->ce_threshold_mask) {
+			int dsfield = skb_get_dsfield(skb);
 
-			set_ce = (ecn == INET_ECN_ECT_1);
+			set_ce = (dsfield >= 0 &&
+				  (((u8)dsfield & params->ce_threshold_mask) ==
+				   params->ce_threshold_selector));
 		}
 		if (set_ce && INET_ECN_set_ce(skb))
 			stats->ce_mark++;

include/uapi/linux/pkt_sched.h

@@ -840,7 +840,8 @@ enum {
 	TCA_FQ_CODEL_CE_THRESHOLD,
 	TCA_FQ_CODEL_DROP_BATCH_SIZE,
 	TCA_FQ_CODEL_MEMORY_LIMIT,
-	TCA_FQ_CODEL_CE_THRESHOLD_ECT1,
+	TCA_FQ_CODEL_CE_THRESHOLD_SELECTOR,
+	TCA_FQ_CODEL_CE_THRESHOLD_MASK,
 	__TCA_FQ_CODEL_MAX
 };
 

net/mac80211/sta_info.c

@@ -513,7 +513,8 @@ struct sta_info *sta_info_alloc(struct ieee80211_sub_if_data *sdata,
 	sta->cparams.target = MS2TIME(20);
 	sta->cparams.interval = MS2TIME(100);
 	sta->cparams.ecn = true;
-	sta->cparams.ce_threshold_ect1 = false;
+	sta->cparams.ce_threshold_selector = 0;
+	sta->cparams.ce_threshold_mask = 0;
 
 	sta_dbg(sdata, "Allocated STA %pM\n", sta->sta.addr);
 

net/sched/sch_fq_codel.c

@@ -362,7 +362,8 @@ static const struct nla_policy fq_codel_policy[TCA_FQ_CODEL_MAX + 1] = {
 	[TCA_FQ_CODEL_CE_THRESHOLD] = { .type = NLA_U32 },
 	[TCA_FQ_CODEL_DROP_BATCH_SIZE] = { .type = NLA_U32 },
 	[TCA_FQ_CODEL_MEMORY_LIMIT] = { .type = NLA_U32 },
-	[TCA_FQ_CODEL_CE_THRESHOLD_ECT1] = { .type = NLA_U8 },
+	[TCA_FQ_CODEL_CE_THRESHOLD_SELECTOR] = { .type = NLA_U8 },
+	[TCA_FQ_CODEL_CE_THRESHOLD_MASK] = { .type = NLA_U8 },
 };
 
 static int fq_codel_change(struct Qdisc *sch, struct nlattr *opt,
@@ -409,8 +410,10 @@ static int fq_codel_change(struct Qdisc *sch, struct nlattr *opt,
 		q->cparams.ce_threshold = (val * NSEC_PER_USEC) >> CODEL_SHIFT;
 	}
 
-	if (tb[TCA_FQ_CODEL_CE_THRESHOLD_ECT1])
-		q->cparams.ce_threshold_ect1 = !!nla_get_u8(tb[TCA_FQ_CODEL_CE_THRESHOLD_ECT1]);
+	if (tb[TCA_FQ_CODEL_CE_THRESHOLD_SELECTOR])
+		q->cparams.ce_threshold_selector = nla_get_u8(tb[TCA_FQ_CODEL_CE_THRESHOLD_SELECTOR]);
+	if (tb[TCA_FQ_CODEL_CE_THRESHOLD_MASK])
+		q->cparams.ce_threshold_mask = nla_get_u8(tb[TCA_FQ_CODEL_CE_THRESHOLD_MASK]);
 
 	if (tb[TCA_FQ_CODEL_INTERVAL]) {
 		u64 interval = nla_get_u32(tb[TCA_FQ_CODEL_INTERVAL]);
@@ -552,7 +555,9 @@ static int fq_codel_dump(struct Qdisc *sch, struct sk_buff *skb)
 		if (nla_put_u32(skb, TCA_FQ_CODEL_CE_THRESHOLD,
 				codel_time_to_us(q->cparams.ce_threshold)))
 			goto nla_put_failure;
-		if (nla_put_u8(skb, TCA_FQ_CODEL_CE_THRESHOLD_ECT1, q->cparams.ce_threshold_ect1))
+		if (nla_put_u8(skb, TCA_FQ_CODEL_CE_THRESHOLD_SELECTOR, q->cparams.ce_threshold_selector))
+			goto nla_put_failure;
+		if (nla_put_u8(skb, TCA_FQ_CODEL_CE_THRESHOLD_MASK, q->cparams.ce_threshold_mask))
 			goto nla_put_failure;
 	}