Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next

Pull networking updates from David Miller:

 1) Significantly shrink the core networking routing structures. Result
    of http://vger.kernel.org/~davem/seoul2017_netdev_keynote.pdf

 2) Add netdevsim driver for testing various offloads, from Jakub
    Kicinski.

 3) Support cross-chip FDB operations in DSA, from Vivien Didelot.

 4) Add a 2nd listener hash table for TCP, similar to what was done for
    UDP. From Martin KaFai Lau.

 5) Add eBPF based queue selection to tun, from Jason Wang.

 6) Lockless qdisc support, from John Fastabend.

 7) SCTP stream interleave support, from Xin Long.

 8) Smoother TCP receive autotuning, from Eric Dumazet.

 9) Lots of erspan tunneling enhancements, from William Tu.

10) Add true function call support to BPF, from Alexei Starovoitov.

11) Add explicit support for GRO HW offloading, from Michael Chan.

12) Support extack generation in more netlink subsystems. From Alexander
    Aring, Quentin Monnet, and Jakub Kicinski.

13) Add 1000BaseX, flow control, and EEE support to mvneta driver. From
    Russell King.

14) Add flow table abstraction to netfilter, from Pablo Neira Ayuso.

15) Many improvements and simplifications to the NFP driver bpf JIT,
    from Jakub Kicinski.

16) Support for ipv6 non-equal cost multipath routing, from Ido
    Schimmel.

17) Add resource abstration to devlink, from Arkadi Sharshevsky.

18) Packet scheduler classifier shared filter block support, from Jiri
    Pirko.

19) Avoid locking in act_csum, from Davide Caratti.

20) devinet_ioctl() simplifications from Al viro.

21) More TCP bpf improvements from Lawrence Brakmo.

22) Add support for onlink ipv6 route flag, similar to ipv4, from David
    Ahern.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next: (1925 commits)
  tls: Add support for encryption using async offload accelerator
  ip6mr: fix stale iterator
  net/sched: kconfig: Remove blank help texts
  openvswitch: meter: Use 64-bit arithmetic instead of 32-bit
  tcp_nv: fix potential integer overflow in tcpnv_acked
  r8169: fix RTL8168EP take too long to complete driver initialization.
  qmi_wwan: Add support for Quectel EP06
  rtnetlink: enable IFLA_IF_NETNSID for RTM_NEWLINK
  ipmr: Fix ptrdiff_t print formatting
  ibmvnic: Wait for device response when changing MAC
  qlcnic: fix deadlock bug
  tcp: release sk_frag.page in tcp_disconnect
  ipv4: Get the address of interface correctly.
  net_sched: gen_estimator: fix lockdep splat
  net: macb: Handle HRESP error
  net/mlx5e: IPoIB, Fix copy-paste bug in flow steering refactoring
  ipv6: addrconf: break critical section in addrconf_verify_rtnl()
  ipv6: change route cache aging logic
  i40e/i40evf: Update DESC_NEEDED value to reflect larger value
  bnxt_en: cleanup DIM work on device shutdown
  ...

1 files changed, 304 insertions, 66 deletions

diff --git a/drivers/net/ethernet/sfc/ptp.c b/drivers/net/ethernet/sfc/ptp.c
index caa89bf..f216615 100644
--- a/drivers/net/ethernet/sfc/ptp.c
+++ b/drivers/net/ethernet/sfc/ptp.c
@@ -149,18 +149,14 @@ enum ptp_packet_state {
 /* Maximum parts-per-billion adjustment that is acceptable */
 #define MAX_PPB			1000000
 
-/* Number of bits required to hold the above */
-#define	MAX_PPB_BITS		20
-
-/* Number of extra bits allowed when calculating fractional ns.
- * EXTRA_BITS + MC_CMD_PTP_IN_ADJUST_BITS + MAX_PPB_BITS should
- * be less than 63.
- */
-#define	PPB_EXTRA_BITS		2
-
 /* Precalculate scale word to avoid long long division at runtime */
-#define	PPB_SCALE_WORD	((1LL << (PPB_EXTRA_BITS + MC_CMD_PTP_IN_ADJUST_BITS +\
-			MAX_PPB_BITS)) / 1000000000LL)
+/* This is equivalent to 2^66 / 10^9. */
+#define PPB_SCALE_WORD  ((1LL << (57)) / 1953125LL)
+
+/* How much to shift down after scaling to convert to FP40 */
+#define PPB_SHIFT_FP40		26
+/* ... and FP44. */
+#define PPB_SHIFT_FP44		22
 
 #define PTP_SYNC_ATTEMPTS	4
 
@@ -218,8 +214,8 @@ struct efx_ptp_timeset {
  * @channel: The PTP channel (Siena only)
  * @rx_ts_inline: Flag for whether RX timestamps are inline (else they are
  *	separate events)
- * @rxq: Receive queue (awaiting timestamps)
- * @txq: Transmit queue
+ * @rxq: Receive SKB queue (awaiting timestamps)
+ * @txq: Transmit SKB queue
  * @evt_list: List of MC receive events awaiting packets
  * @evt_free_list: List of free events
  * @evt_lock: Lock for manipulating evt_list and evt_free_list
@@ -233,19 +229,36 @@ struct efx_ptp_timeset {
  * @config: Current timestamp configuration
  * @enabled: PTP operation enabled
  * @mode: Mode in which PTP operating (PTP version)
- * @time_format: Time format supported by this NIC
  * @ns_to_nic_time: Function to convert from scalar nanoseconds to NIC time
  * @nic_to_kernel_time: Function to convert from NIC to kernel time
+ * @nic_time.minor_max: Wrap point for NIC minor times
+ * @nic_time.sync_event_diff_min: Minimum acceptable difference between time
+ * in packet prefix and last MCDI time sync event i.e. how much earlier than
+ * the last sync event time a packet timestamp can be.
+ * @nic_time.sync_event_diff_max: Maximum acceptable difference between time
+ * in packet prefix and last MCDI time sync event i.e. how much later than
+ * the last sync event time a packet timestamp can be.
+ * @nic_time.sync_event_minor_shift: Shift required to make minor time from
+ * field in MCDI time sync event.
  * @min_synchronisation_ns: Minimum acceptable corrected sync window
- * @ts_corrections.tx: Required driver correction of transmit timestamps
- * @ts_corrections.rx: Required driver correction of receive timestamps
+ * @capabilities: Capabilities flags from the NIC
+ * @ts_corrections.ptp_tx: Required driver correction of PTP packet transmit
+ *                         timestamps
+ * @ts_corrections.ptp_rx: Required driver correction of PTP packet receive
+ *                         timestamps
  * @ts_corrections.pps_out: PPS output error (information only)
  * @ts_corrections.pps_in: Required driver correction of PPS input timestamps
+ * @ts_corrections.general_tx: Required driver correction of general packet
+ *                             transmit timestamps
+ * @ts_corrections.general_rx: Required driver correction of general packet
+ *                             receive timestamps
  * @evt_frags: Partly assembled PTP events
  * @evt_frag_idx: Current fragment number
  * @evt_code: Last event code
  * @start: Address at which MC indicates ready for synchronisation
  * @host_time_pps: Host time at last PPS
+ * @adjfreq_ppb_shift: Shift required to convert scaled parts-per-billion
+ * frequency adjustment into a fixed point fractional nanosecond format.
  * @current_adjfreq: Current ppb adjustment.
  * @phc_clock: Pointer to registered phc device (if primary function)
  * @phc_clock_info: Registration structure for phc device
@@ -264,6 +277,7 @@ struct efx_ptp_timeset {
  * @oversize_sync_windows: Number of corrected sync windows that are too large
  * @rx_no_timestamp: Number of packets received without a timestamp.
  * @timeset: Last set of synchronisation statistics.
+ * @xmit_skb: Transmit SKB function.
  */
 struct efx_ptp_data {
 	struct efx_nic *efx;
@@ -284,22 +298,31 @@ struct efx_ptp_data {
 	struct hwtstamp_config config;
 	bool enabled;
 	unsigned int mode;
-	unsigned int time_format;
 	void (*ns_to_nic_time)(s64 ns, u32 *nic_major, u32 *nic_minor);
 	ktime_t (*nic_to_kernel_time)(u32 nic_major, u32 nic_minor,
 				      s32 correction);
+	struct {
+		u32 minor_max;
+		u32 sync_event_diff_min;
+		u32 sync_event_diff_max;
+		unsigned int sync_event_minor_shift;
+	} nic_time;
 	unsigned int min_synchronisation_ns;
+	unsigned int capabilities;
 	struct {
-		s32 tx;
-		s32 rx;
+		s32 ptp_tx;
+		s32 ptp_rx;
 		s32 pps_out;
 		s32 pps_in;
+		s32 general_tx;
+		s32 general_rx;
 	} ts_corrections;
 	efx_qword_t evt_frags[MAX_EVENT_FRAGS];
 	int evt_frag_idx;
 	int evt_code;
 	struct efx_buffer start;
 	struct pps_event_time host_time_pps;
+	unsigned int adjfreq_ppb_shift;
 	s64 current_adjfreq;
 	struct ptp_clock *phc_clock;
 	struct ptp_clock_info phc_clock_info;
@@ -319,6 +342,7 @@ struct efx_ptp_data {
 	unsigned int rx_no_timestamp;
 	struct efx_ptp_timeset
 	timeset[MC_CMD_PTP_OUT_SYNCHRONIZE_TIMESET_MAXNUM];
+	void (*xmit_skb)(struct efx_nic *efx, struct sk_buff *skb);
 };
 
 static int efx_phc_adjfreq(struct ptp_clock_info *ptp, s32 delta);
@@ -329,6 +353,24 @@ static int efx_phc_settime(struct ptp_clock_info *ptp,
 static int efx_phc_enable(struct ptp_clock_info *ptp,
 			  struct ptp_clock_request *request, int on);
 
+bool efx_ptp_use_mac_tx_timestamps(struct efx_nic *efx)
+{
+	struct efx_ef10_nic_data *nic_data = efx->nic_data;
+
+	return ((efx_nic_rev(efx) >= EFX_REV_HUNT_A0) &&
+		(nic_data->datapath_caps2 &
+		 (1 << MC_CMD_GET_CAPABILITIES_V2_OUT_TX_MAC_TIMESTAMPING_LBN)
+		));
+}
+
+/* PTP 'extra' channel is still a traffic channel, but we only create TX queues
+ * if PTP uses MAC TX timestamps, not if PTP uses the MC directly to transmit.
+ */
+static bool efx_ptp_want_txqs(struct efx_channel *channel)
+{
+	return efx_ptp_use_mac_tx_timestamps(channel->efx);
+}
+
 #define PTP_SW_STAT(ext_name, field_name)				\
 	{ #ext_name, 0, offsetof(struct efx_ptp_data, field_name) }
 #define PTP_MC_STAT(ext_name, mcdi_name)				\
@@ -471,6 +513,89 @@ static ktime_t efx_ptp_s27_to_ktime_correction(u32 nic_major, u32 nic_minor,
 	return efx_ptp_s27_to_ktime(nic_major, nic_minor);
 }
 
+/* For Medford2 platforms the time is in seconds and quarter nanoseconds. */
+static void efx_ptp_ns_to_s_qns(s64 ns, u32 *nic_major, u32 *nic_minor)
+{
+	struct timespec64 ts = ns_to_timespec64(ns);
+
+	*nic_major = (u32)ts.tv_sec;
+	*nic_minor = ts.tv_nsec * 4;
+}
+
+static ktime_t efx_ptp_s_qns_to_ktime_correction(u32 nic_major, u32 nic_minor,
+						 s32 correction)
+{
+	ktime_t kt;
+
+	nic_minor = DIV_ROUND_CLOSEST(nic_minor, 4);
+	correction = DIV_ROUND_CLOSEST(correction, 4);
+
+	kt = ktime_set(nic_major, nic_minor);
+
+	if (correction >= 0)
+		kt = ktime_add_ns(kt, (u64)correction);
+	else
+		kt = ktime_sub_ns(kt, (u64)-correction);
+	return kt;
+}
+
+struct efx_channel *efx_ptp_channel(struct efx_nic *efx)
+{
+	return efx->ptp_data ? efx->ptp_data->channel : NULL;
+}
+
+static u32 last_sync_timestamp_major(struct efx_nic *efx)
+{
+	struct efx_channel *channel = efx_ptp_channel(efx);
+	u32 major = 0;
+
+	if (channel)
+		major = channel->sync_timestamp_major;
+	return major;
+}
+
+/* The 8000 series and later can provide the time from the MAC, which is only
+ * 48 bits long and provides meta-information in the top 2 bits.
+ */
+static ktime_t
+efx_ptp_mac_nic_to_ktime_correction(struct efx_nic *efx,
+				    struct efx_ptp_data *ptp,
+				    u32 nic_major, u32 nic_minor,
+				    s32 correction)
+{
+	ktime_t kt = { 0 };
+
+	if (!(nic_major & 0x80000000)) {
+		WARN_ON_ONCE(nic_major >> 16);
+		/* Use the top bits from the latest sync event. */
+		nic_major &= 0xffff;
+		nic_major |= (last_sync_timestamp_major(efx) & 0xffff0000);
+
+		kt = ptp->nic_to_kernel_time(nic_major, nic_minor,
+					     correction);
+	}
+	return kt;
+}
+
+ktime_t efx_ptp_nic_to_kernel_time(struct efx_tx_queue *tx_queue)
+{
+	struct efx_nic *efx = tx_queue->efx;
+	struct efx_ptp_data *ptp = efx->ptp_data;
+	ktime_t kt;
+
+	if (efx_ptp_use_mac_tx_timestamps(efx))
+		kt = efx_ptp_mac_nic_to_ktime_correction(efx, ptp,
+				tx_queue->completed_timestamp_major,
+				tx_queue->completed_timestamp_minor,
+				ptp->ts_corrections.general_tx);
+	else
+		kt = ptp->nic_to_kernel_time(
+				tx_queue->completed_timestamp_major,
+				tx_queue->completed_timestamp_minor,
+				ptp->ts_corrections.general_tx);
+	return kt;
+}
+
 /* Get PTP attributes and set up time conversions */
 static int efx_ptp_get_attributes(struct efx_nic *efx)
 {
@@ -502,31 +627,71 @@ static int efx_ptp_get_attributes(struct efx_nic *efx)
 		return rc;
 	}
 
-	if (fmt == MC_CMD_PTP_OUT_GET_ATTRIBUTES_SECONDS_27FRACTION) {
+	switch (fmt) {
+	case MC_CMD_PTP_OUT_GET_ATTRIBUTES_SECONDS_27FRACTION:
 		ptp->ns_to_nic_time = efx_ptp_ns_to_s27;
 		ptp->nic_to_kernel_time = efx_ptp_s27_to_ktime_correction;
-	} else if (fmt == MC_CMD_PTP_OUT_GET_ATTRIBUTES_SECONDS_NANOSECONDS) {
+		ptp->nic_time.minor_max = 1 << 27;
+		ptp->nic_time.sync_event_minor_shift = 19;
+		break;
+	case MC_CMD_PTP_OUT_GET_ATTRIBUTES_SECONDS_NANOSECONDS:
 		ptp->ns_to_nic_time = efx_ptp_ns_to_s_ns;
 		ptp->nic_to_kernel_time = efx_ptp_s_ns_to_ktime_correction;
-	} else {
+		ptp->nic_time.minor_max = 1000000000;
+		ptp->nic_time.sync_event_minor_shift = 22;
+		break;
+	case MC_CMD_PTP_OUT_GET_ATTRIBUTES_SECONDS_QTR_NANOSECONDS:
+		ptp->ns_to_nic_time = efx_ptp_ns_to_s_qns;
+		ptp->nic_to_kernel_time = efx_ptp_s_qns_to_ktime_correction;
+		ptp->nic_time.minor_max = 4000000000UL;
+		ptp->nic_time.sync_event_minor_shift = 24;
+		break;
+	default:
 		return -ERANGE;
 	}
 
-	ptp->time_format = fmt;
-
-	/* MC_CMD_PTP_OP_GET_ATTRIBUTES is an extended version of an older
-	 * operation MC_CMD_PTP_OP_GET_TIME_FORMAT that also returns a value
-	 * to use for the minimum acceptable corrected synchronization window.
+	/* Precalculate acceptable difference between the minor time in the
+	 * packet prefix and the last MCDI time sync event. We expect the
+	 * packet prefix timestamp to be after of sync event by up to one
+	 * sync event interval (0.25s) but we allow it to exceed this by a
+	 * fuzz factor of (0.1s)
+	 */
+	ptp->nic_time.sync_event_diff_min = ptp->nic_time.minor_max
+		- (ptp->nic_time.minor_max / 10);
+	ptp->nic_time.sync_event_diff_max = (ptp->nic_time.minor_max / 4)
+		+ (ptp->nic_time.minor_max / 10);
+
+	/* MC_CMD_PTP_OP_GET_ATTRIBUTES has been extended twice from an older
+	 * operation MC_CMD_PTP_OP_GET_TIME_FORMAT. The function now may return
+	 * a value to use for the minimum acceptable corrected synchronization
+	 * window and may return further capabilities.
 	 * If we have the extra information store it. For older firmware that
 	 * does not implement the extended command use the default value.
 	 */
-	if (rc == 0 && out_len >= MC_CMD_PTP_OUT_GET_ATTRIBUTES_LEN)
+	if (rc == 0 &&
+	    out_len >= MC_CMD_PTP_OUT_GET_ATTRIBUTES_CAPABILITIES_OFST)
 		ptp->min_synchronisation_ns =
 			MCDI_DWORD(outbuf,
 				   PTP_OUT_GET_ATTRIBUTES_SYNC_WINDOW_MIN);
 	else
 		ptp->min_synchronisation_ns = DEFAULT_MIN_SYNCHRONISATION_NS;
 
+	if (rc == 0 &&
+	    out_len >= MC_CMD_PTP_OUT_GET_ATTRIBUTES_LEN)
+		ptp->capabilities = MCDI_DWORD(outbuf,
+					PTP_OUT_GET_ATTRIBUTES_CAPABILITIES);
+	else
+		ptp->capabilities = 0;
+
+	/* Set up the shift for conversion between frequency
+	 * adjustments in parts-per-billion and the fixed-point
+	 * fractional ns format that the adapter uses.
+	 */
+	if (ptp->capabilities & (1 << MC_CMD_PTP_OUT_GET_ATTRIBUTES_FP44_FREQ_ADJ_LBN))
+		ptp->adjfreq_ppb_shift = PPB_SHIFT_FP44;
+	else
+		ptp->adjfreq_ppb_shift = PPB_SHIFT_FP40;
+
 	return 0;
 }
 
@@ -534,8 +699,9 @@ static int efx_ptp_get_attributes(struct efx_nic *efx)
 static int efx_ptp_get_timestamp_corrections(struct efx_nic *efx)
 {
 	MCDI_DECLARE_BUF(inbuf, MC_CMD_PTP_IN_GET_TIMESTAMP_CORRECTIONS_LEN);
-	MCDI_DECLARE_BUF(outbuf, MC_CMD_PTP_OUT_GET_TIMESTAMP_CORRECTIONS_LEN);
+	MCDI_DECLARE_BUF(outbuf, MC_CMD_PTP_OUT_GET_TIMESTAMP_CORRECTIONS_V2_LEN);
 	int rc;
+	size_t out_len;
 
 	/* Get the timestamp corrections from the NIC. If this operation is
 	 * not supported (older NICs) then no correction is required.
@@ -545,21 +711,37 @@ static int efx_ptp_get_timestamp_corrections(struct efx_nic *efx)
 	MCDI_SET_DWORD(inbuf, PTP_IN_PERIPH_ID, 0);
 
 	rc = efx_mcdi_rpc_quiet(efx, MC_CMD_PTP, inbuf, sizeof(inbuf),
-				outbuf, sizeof(outbuf), NULL);
+				outbuf, sizeof(outbuf), &out_len);
 	if (rc == 0) {
-		efx->ptp_data->ts_corrections.tx = MCDI_DWORD(outbuf,
+		efx->ptp_data->ts_corrections.ptp_tx = MCDI_DWORD(outbuf,
 			PTP_OUT_GET_TIMESTAMP_CORRECTIONS_TRANSMIT);
-		efx->ptp_data->ts_corrections.rx = MCDI_DWORD(outbuf,
+		efx->ptp_data->ts_corrections.ptp_rx = MCDI_DWORD(outbuf,
 			PTP_OUT_GET_TIMESTAMP_CORRECTIONS_RECEIVE);
 		efx->ptp_data->ts_corrections.pps_out = MCDI_DWORD(outbuf,
 			PTP_OUT_GET_TIMESTAMP_CORRECTIONS_PPS_OUT);
 		efx->ptp_data->ts_corrections.pps_in = MCDI_DWORD(outbuf,
 			PTP_OUT_GET_TIMESTAMP_CORRECTIONS_PPS_IN);
+
+		if (out_len >= MC_CMD_PTP_OUT_GET_TIMESTAMP_CORRECTIONS_V2_LEN) {
+			efx->ptp_data->ts_corrections.general_tx = MCDI_DWORD(
+				outbuf,
+				PTP_OUT_GET_TIMESTAMP_CORRECTIONS_V2_GENERAL_TX);
+			efx->ptp_data->ts_corrections.general_rx = MCDI_DWORD(
+				outbuf,
+				PTP_OUT_GET_TIMESTAMP_CORRECTIONS_V2_GENERAL_RX);
+		} else {
+			efx->ptp_data->ts_corrections.general_tx =
+				efx->ptp_data->ts_corrections.ptp_tx;
+			efx->ptp_data->ts_corrections.general_rx =
+				efx->ptp_data->ts_corrections.ptp_rx;
+		}
 	} else if (rc == -EINVAL) {
-		efx->ptp_data->ts_corrections.tx = 0;
-		efx->ptp_data->ts_corrections.rx = 0;
+		efx->ptp_data->ts_corrections.ptp_tx = 0;
+		efx->ptp_data->ts_corrections.ptp_rx = 0;
 		efx->ptp_data->ts_corrections.pps_out = 0;
 		efx->ptp_data->ts_corrections.pps_in = 0;
+		efx->ptp_data->ts_corrections.general_tx = 0;
+		efx->ptp_data->ts_corrections.general_rx = 0;
 	} else {
 		efx_mcdi_display_error(efx, MC_CMD_PTP, sizeof(inbuf), outbuf,
 				       sizeof(outbuf), rc);
@@ -873,8 +1055,24 @@ static int efx_ptp_synchronize(struct efx_nic *efx, unsigned int num_readings)
 	return rc;
 }
 
+/* Transmit a PTP packet via the dedicated hardware timestamped queue. */
+static void efx_ptp_xmit_skb_queue(struct efx_nic *efx, struct sk_buff *skb)
+{
+	struct efx_ptp_data *ptp_data = efx->ptp_data;
+	struct efx_tx_queue *tx_queue;
+	u8 type = skb->ip_summed == CHECKSUM_PARTIAL ? EFX_TXQ_TYPE_OFFLOAD : 0;
+
+	tx_queue = &ptp_data->channel->tx_queue[type];
+	if (tx_queue && tx_queue->timestamping) {
+		efx_enqueue_skb(tx_queue, skb);
+	} else {
+		WARN_ONCE(1, "PTP channel has no timestamped tx queue\n");
+		dev_kfree_skb_any(skb);
+	}
+}
+
 /* Transmit a PTP packet, via the MCDI interface, to the wire. */
-static int efx_ptp_xmit_skb(struct efx_nic *efx, struct sk_buff *skb)
+static void efx_ptp_xmit_skb_mc(struct efx_nic *efx, struct sk_buff *skb)
 {
 	struct efx_ptp_data *ptp_data = efx->ptp_data;
 	struct skb_shared_hwtstamps timestamps;
@@ -910,16 +1108,16 @@ static int efx_ptp_xmit_skb(struct efx_nic *efx, struct sk_buff *skb)
 	timestamps.hwtstamp = ptp_data->nic_to_kernel_time(
 		MCDI_DWORD(txtime, PTP_OUT_TRANSMIT_MAJOR),
 		MCDI_DWORD(txtime, PTP_OUT_TRANSMIT_MINOR),
-		ptp_data->ts_corrections.tx);
+		ptp_data->ts_corrections.ptp_tx);
 
 	skb_tstamp_tx(skb, &timestamps);
 
 	rc = 0;
 
 fail:
-	dev_kfree_skb(skb);
+	dev_kfree_skb_any(skb);
 
-	return rc;
+	return;
 }
 
 static void efx_ptp_drop_time_expired_events(struct efx_nic *efx)
@@ -1189,7 +1387,7 @@ static void efx_ptp_worker(struct work_struct *work)
 	efx_ptp_process_events(efx, &tempq);
 
 	while ((skb = skb_dequeue(&ptp_data->txq)))
-		efx_ptp_xmit_skb(efx, skb);
+		ptp_data->xmit_skb(efx, skb);
 
 	while ((skb = __skb_dequeue(&tempq)))
 		efx_ptp_process_rx(efx, skb);
@@ -1239,6 +1437,14 @@ int efx_ptp_probe(struct efx_nic *efx, struct efx_channel *channel)
 		goto fail2;
 	}
 
+	if (efx_ptp_use_mac_tx_timestamps(efx)) {
+		ptp->xmit_skb = efx_ptp_xmit_skb_queue;
+		/* Request sync events on this channel. */
+		channel->sync_events_state = SYNC_EVENTS_QUIESCENT;
+	} else {
+		ptp->xmit_skb = efx_ptp_xmit_skb_mc;
+	}
+
 	INIT_WORK(&ptp->work, efx_ptp_worker);
 	ptp->config.flags = 0;
 	ptp->config.tx_type = HWTSTAMP_TX_OFF;
@@ -1303,11 +1509,21 @@ fail1:
 static int efx_ptp_probe_channel(struct efx_channel *channel)
 {
 	struct efx_nic *efx = channel->efx;
+	int rc;
 
 	channel->irq_moderation_us = 0;
 	channel->rx_queue.core_index = 0;
 
-	return efx_ptp_probe(efx, channel);
+	rc = efx_ptp_probe(efx, channel);
+	/* Failure to probe PTP is not fatal; this channel will just not be
+	 * used for anything.
+	 * In the case of EPERM, efx_ptp_probe will print its own message (in
+	 * efx_ptp_get_attributes()), so we don't need to.
+	 */
+	if (rc && rc != -EPERM)
+		netif_warn(efx, drv, efx->net_dev,
+			   "Failed to probe PTP, rc=%d\n", rc);
+	return 0;
 }
 
 void efx_ptp_remove(struct efx_nic *efx)
@@ -1332,6 +1548,7 @@ void efx_ptp_remove(struct efx_nic *efx)
 
 	efx_nic_free_buffer(efx, &efx->ptp_data->start);
 	kfree(efx->ptp_data);
+	efx->ptp_data = NULL;
 }
 
 static void efx_ptp_remove_channel(struct efx_channel *channel)
@@ -1548,6 +1765,17 @@ void efx_ptp_get_ts_info(struct efx_nic *efx, struct ethtool_ts_info *ts_info)
 	ts_info->so_timestamping |= (SOF_TIMESTAMPING_TX_HARDWARE |
 				     SOF_TIMESTAMPING_RX_HARDWARE |
 				     SOF_TIMESTAMPING_RAW_HARDWARE);
+	/* Check licensed features.  If we don't have the license for TX
+	 * timestamps, the NIC will not support them.
+	 */
+	if (efx_ptp_use_mac_tx_timestamps(efx)) {
+		struct efx_ef10_nic_data *nic_data = efx->nic_data;
+
+		if (!(nic_data->licensed_features &
+		      (1 << LICENSED_V3_FEATURES_TX_TIMESTAMPS_LBN)))
+			ts_info->so_timestamping &=
+				~SOF_TIMESTAMPING_TX_HARDWARE;
+	}
 	if (primary && primary->ptp_data && primary->ptp_data->phc_clock)
 		ts_info->phc_index =
 			ptp_clock_index(primary->ptp_data->phc_clock);
@@ -1627,7 +1855,7 @@ static void ptp_event_rx(struct efx_nic *efx, struct efx_ptp_data *ptp)
 		evt->hwtimestamp = efx->ptp_data->nic_to_kernel_time(
 			EFX_QWORD_FIELD(ptp->evt_frags[0], MCDI_EVENT_DATA),
 			EFX_QWORD_FIELD(ptp->evt_frags[1], MCDI_EVENT_DATA),
-			ptp->ts_corrections.rx);
+			ptp->ts_corrections.ptp_rx);
 		evt->expiry = jiffies + msecs_to_jiffies(PKT_EVENT_LIFETIME_MS);
 		list_add_tail(&evt->link, &ptp->evt_list);
 
@@ -1662,9 +1890,11 @@ void efx_ptp_event(struct efx_nic *efx, efx_qword_t *ev)
 	int code = EFX_QWORD_FIELD(*ev, MCDI_EVENT_CODE);
 
 	if (!ptp) {
-		if (net_ratelimit())
+		if (!efx->ptp_warned) {
 			netif_warn(efx, drv, efx->net_dev,
 				   "Received PTP event but PTP not set up\n");
+			efx->ptp_warned = true;
+		}
 		return;
 	}
 
@@ -1707,9 +1937,20 @@ void efx_ptp_event(struct efx_nic *efx, efx_qword_t *ev)
 
 void efx_time_sync_event(struct efx_channel *channel, efx_qword_t *ev)
 {
+	struct efx_nic *efx = channel->efx;
+	struct efx_ptp_data *ptp = efx->ptp_data;
+
+	/* When extracting the sync timestamp minor value, we should discard
+	 * the least significant two bits. These are not required in order
+	 * to reconstruct full-range timestamps and they are optionally used
+	 * to report status depending on the options supplied when subscribing
+	 * for sync events.
+	 */
 	channel->sync_timestamp_major = MCDI_EVENT_FIELD(*ev, PTP_TIME_MAJOR);
 	channel->sync_timestamp_minor =
-		MCDI_EVENT_FIELD(*ev, PTP_TIME_MINOR_26_19) << 19;
+		(MCDI_EVENT_FIELD(*ev, PTP_TIME_MINOR_MS_8BITS) & 0xFC)
+			<< ptp->nic_time.sync_event_minor_shift;
+
 	/* if sync events have been disabled then we want to silently ignore
 	 * this event, so throw away result.
 	 */
@@ -1717,15 +1958,6 @@ void efx_time_sync_event(struct efx_channel *channel, efx_qword_t *ev)
 		       SYNC_EVENTS_VALID);
 }
 
-/* make some assumptions about the time representation rather than abstract it,
- * since we currently only support one type of inline timestamping and only on
- * EF10.
- */
-#define MINOR_TICKS_PER_SECOND 0x8000000
-/* Fuzz factor for sync events to be out of order with RX events */
-#define FUZZ (MINOR_TICKS_PER_SECOND / 10)
-#define EXPECTED_SYNC_EVENTS_PER_SECOND 4
-
 static inline u32 efx_rx_buf_timestamp_minor(struct efx_nic *efx, const u8 *eh)
 {
 #if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS)
@@ -1743,31 +1975,33 @@ void __efx_rx_skb_attach_timestamp(struct efx_channel *channel,
 				   struct sk_buff *skb)
 {
 	struct efx_nic *efx = channel->efx;
+	struct efx_ptp_data *ptp = efx->ptp_data;
 	u32 pkt_timestamp_major, pkt_timestamp_minor;
 	u32 diff, carry;
 	struct skb_shared_hwtstamps *timestamps;
 
-	pkt_timestamp_minor = (efx_rx_buf_timestamp_minor(efx,
-							  skb_mac_header(skb)) +
-			       (u32) efx->ptp_data->ts_corrections.rx) &
-			      (MINOR_TICKS_PER_SECOND - 1);
+	if (channel->sync_events_state != SYNC_EVENTS_VALID)
+		return;
+
+	pkt_timestamp_minor = efx_rx_buf_timestamp_minor(efx, skb_mac_header(skb));
 
 	/* get the difference between the packet and sync timestamps,
 	 * modulo one second
 	 */
-	diff = (pkt_timestamp_minor - channel->sync_timestamp_minor) &
-		(MINOR_TICKS_PER_SECOND - 1);
+	diff = pkt_timestamp_minor - channel->sync_timestamp_minor;
+	if (pkt_timestamp_minor < channel->sync_timestamp_minor)
+		diff += ptp->nic_time.minor_max;
+
 	/* do we roll over a second boundary and need to carry the one? */
-	carry = channel->sync_timestamp_minor + diff > MINOR_TICKS_PER_SECOND ?
+	carry = (channel->sync_timestamp_minor >= ptp->nic_time.minor_max - diff) ?
 		1 : 0;
 
-	if (diff <= MINOR_TICKS_PER_SECOND / EXPECTED_SYNC_EVENTS_PER_SECOND +
-		    FUZZ) {
+	if (diff <= ptp->nic_time.sync_event_diff_max) {
 		/* packet is ahead of the sync event by a quarter of a second or
 		 * less (allowing for fuzz)
 		 */
 		pkt_timestamp_major = channel->sync_timestamp_major + carry;
-	} else if (diff >= MINOR_TICKS_PER_SECOND - FUZZ) {
+	} else if (diff >= ptp->nic_time.sync_event_diff_min) {
 		/* packet is behind the sync event but within the fuzz factor.
 		 * This means the RX packet and sync event crossed as they were
 		 * placed on the event queue, which can sometimes happen.
@@ -1789,7 +2023,9 @@ void __efx_rx_skb_attach_timestamp(struct efx_channel *channel,
 	/* attach the timestamps to the skb */
 	timestamps = skb_hwtstamps(skb);
 	timestamps->hwtstamp =
-		efx_ptp_s27_to_ktime(pkt_timestamp_major, pkt_timestamp_minor);
+		ptp->nic_to_kernel_time(pkt_timestamp_major,
+					pkt_timestamp_minor,
+					ptp->ts_corrections.general_rx);
 }
 
 static int efx_phc_adjfreq(struct ptp_clock_info *ptp, s32 delta)
@@ -1807,9 +2043,10 @@ static int efx_phc_adjfreq(struct ptp_clock_info *ptp, s32 delta)
 	else if (delta < -MAX_PPB)
 		delta = -MAX_PPB;
 
-	/* Convert ppb to fixed point ns. */
-	adjustment_ns = (((s64)delta * PPB_SCALE_WORD) >>
-			 (PPB_EXTRA_BITS + MAX_PPB_BITS));
+	/* Convert ppb to fixed point ns taking care to round correctly. */
+	adjustment_ns = ((s64)delta * PPB_SCALE_WORD +
+			 (1 << (ptp_data->adjfreq_ppb_shift - 1))) >>
+			ptp_data->adjfreq_ppb_shift;
 
 	MCDI_SET_DWORD(inadj, PTP_IN_OP, MC_CMD_PTP_OP_ADJUST);
 	MCDI_SET_DWORD(inadj, PTP_IN_PERIPH_ID, 0);
@@ -1916,6 +2153,7 @@ static const struct efx_channel_type efx_ptp_channel_type = {
 	.get_name		= efx_ptp_get_channel_name,
 	/* no copy operation; there is no need to reallocate this channel */
 	.receive_skb		= efx_ptp_rx,
+	.want_txqs		= efx_ptp_want_txqs,
 	.keep_eventq		= false,
 };