[PATCH v4 2/2] net/af_xdp: add read_clock support to AF_XDP PMD

[Mark Blasko <blasko@google.com>]

Implement the ethdev read_clock operation in the AF_XDP Poll Mode
Driver. This allows DPDK applications to query the current time of the
NIC hardware clock.

At device start, the PMD queries ethtool for the interface's PTP
Hardware Clock index and opens the corresponding PTP device node. The
read_clock operation queries the current time via clock_gettime.

Signed-off-by: Mark Blasko <blasko@google.com>
Reviewed-by: Joshua Washington <joshwash@google.com>
---
v4:
  - Move ptp_fd into pmd_process_private for multi-process safety.
  - Decouple /dev/ptpX opening from RTE_ETH_RX_OFFLOAD_TIMESTAMP.
  - Move read_clock documentation out of "Options" into its own section
    in af_xdp.rst.
  - Add #ifndef guards around CLOCKFD and FD_TO_CLOCKID macro definitions.
  - Replace (caddr_t) cast with (void *).
v3:
  - Add feature description and PTP hardware clock prerequisites in
    doc/guides/nics/af_xdp.rst.
  - Add release notes entry in doc/guides/rel_notes/release_26_07.rst.
  - Add PTP file descriptor cleanup in eth_dev_start(), eth_dev_stop(),
    and eth_dev_close().
  - Return -errno on clock_gettime() failure.
  - Add strerror(errno) error logging on PTP device open failure.
v2:
  - New patch introduced in v2 to support read_clock ethdev operation.
---
 doc/guides/nics/af_xdp.rst             | 11 +++
 doc/guides/rel_notes/release_26_07.rst |  1 +
 drivers/net/af_xdp/rte_eth_af_xdp.c    | 99 ++++++++++++++++++++++++++
 3 files changed, 111 insertions(+)

diff --git a/doc/guides/nics/af_xdp.rst b/doc/guides/nics/af_xdp.rst
index b48f7fb9f0..c8495decc5 100644
--- a/doc/guides/nics/af_xdp.rst
+++ b/doc/guides/nics/af_xdp.rst
@@ -37,6 +37,9 @@ Prerequisites
    header is used to determine the kernel version at compile time.
 *  A kernel with version 5.4 or later is required for 32-bit OS.
 *  The busy polling feature requires kernel version >= v5.11.
+*  The ``read_clock`` feature requires a network interface with PTP
+   Hardware Clock support (capable of exposing a ``/dev/ptpX`` device via
+   ethtool ``ETHTOOL_GET_TS_INFO``).
 
 
 Options
@@ -247,6 +250,14 @@ netdev interface via ``SIOCSHWTSTAMP`` (which requires ``CAP_NET_ADMIN``). This
 hardware filter setting persists on the netdev interface after the DPDK application
 exits.
 
+read_clock
+----------
+
+The PMD supports querying the underlying PTP hardware clock time via
+``rte_eth_read_clock()``. During device start, the PMD automatically discovers
+the hardware PHC index via ethtool and opens the PTP character device
+(``/dev/ptpX``).
+
 Limitations
 -----------
 
diff --git a/doc/guides/rel_notes/release_26_07.rst b/doc/guides/rel_notes/release_26_07.rst
index 8af0b8d615..326c35243b 100644
--- a/doc/guides/rel_notes/release_26_07.rst
+++ b/doc/guides/rel_notes/release_26_07.rst
@@ -140,6 +140,7 @@ New Features
   * Added support for RX metadata hardware timestamping via vdev devargs
     ``xdp_meta_rx_ts_offset``, ``xdp_meta_valid_hint_offset``, and
     ``xdp_meta_rx_ts_valid_mask``.
+  * Added ``read_clock`` operation to query the PTP hardware clock.
 
 * **Updated Google gve driver.**
 
diff --git a/drivers/net/af_xdp/rte_eth_af_xdp.c b/drivers/net/af_xdp/rte_eth_af_xdp.c
index b75c87b641..c8f7a26716 100644
--- a/drivers/net/af_xdp/rte_eth_af_xdp.c
+++ b/drivers/net/af_xdp/rte_eth_af_xdp.c
@@ -5,6 +5,8 @@
 #include <errno.h>
 #include <stdlib.h>
 #include <string.h>
+#include <fcntl.h>
+#include <time.h>
 #include <netinet/in.h>
 #include <net/if.h>
 #include <sys/un.h>
@@ -29,6 +31,7 @@
 #include <dev_driver.h>
 #include <rte_eal.h>
 #include <rte_ether.h>
+#include <rte_time.h>
 #include <rte_lcore.h>
 #include <rte_log.h>
 #include <rte_memory.h>
@@ -198,6 +201,7 @@ struct pmd_internals {
 
 struct pmd_process_private {
 	int rxq_xsk_fds[RTE_MAX_QUEUES_PER_PORT];
+	int ptp_fd;
 };
 
 #define ETH_AF_XDP_IFACE_ARG			"iface"
@@ -834,10 +838,35 @@ eth_af_xdp_enable_hw_timestamping(const char *if_name)
 	return 0;
 }
 
+static int
+eth_af_xdp_get_ptp_index(const char *if_name)
+{
+	struct ethtool_ts_info info = {0};
+	struct ifreq ifr = {0};
+	int fd, ret;
+
+	fd = socket(AF_INET, SOCK_DGRAM, 0);
+	if (fd < 0)
+		return -1;
+
+	ifr.ifr_data = (void *)&info;
+	info.cmd = ETHTOOL_GET_TS_INFO;
+	strlcpy(ifr.ifr_name, if_name, IFNAMSIZ);
+
+	ret = ioctl(fd, SIOCETHTOOL, &ifr);
+	close(fd);
+
+	if (ret < 0)
+		return -1;
+
+	return info.phc_index;
+}
+
 static int
 eth_dev_start(struct rte_eth_dev *dev)
 {
 	struct pmd_internals *internals = dev->data->dev_private;
+	struct pmd_process_private *process_private = dev->process_private;
 	uint16_t i;
 
 	if (dev->data->dev_conf.rxmode.offloads &
@@ -870,6 +899,28 @@ eth_dev_start(struct rte_eth_dev *dev)
 		}
 	}
 
+	if (process_private != NULL) {
+		int phc_index = eth_af_xdp_get_ptp_index(internals->if_name);
+		if (phc_index >= 0) {
+			char ptp_dev[32];
+			snprintf(ptp_dev, sizeof(ptp_dev), "/dev/ptp%d", phc_index);
+			if (process_private->ptp_fd >= 0) {
+				close(process_private->ptp_fd);
+				process_private->ptp_fd = -1;
+			}
+			process_private->ptp_fd = open(ptp_dev, O_RDONLY);
+			if (process_private->ptp_fd >= 0) {
+				AF_XDP_LOG_LINE(INFO,
+					"Opened PTP device %s for read_clock",
+					ptp_dev);
+			} else {
+				AF_XDP_LOG_LINE(WARNING,
+					"Failed to open PTP device %s for read_clock: %s",
+					ptp_dev, strerror(errno));
+			}
+		}
+	}
+
 	dev->data->dev_link.link_status = RTE_ETH_LINK_UP;
 	for (i = 0; i < dev->data->nb_rx_queues; i++) {
 		dev->data->rx_queue_state[i] = RTE_ETH_QUEUE_STATE_STARTED;
@@ -883,6 +934,7 @@ eth_dev_start(struct rte_eth_dev *dev)
 static int
 eth_dev_stop(struct rte_eth_dev *dev)
 {
+	struct pmd_process_private *process_private = dev->process_private;
 	uint16_t i;
 
 	dev->data->dev_link.link_status = RTE_ETH_LINK_DOWN;
@@ -891,6 +943,11 @@ eth_dev_stop(struct rte_eth_dev *dev)
 		dev->data->tx_queue_state[i] = RTE_ETH_QUEUE_STATE_STOPPED;
 	}
 
+	if (process_private != NULL && process_private->ptp_fd >= 0) {
+		close(process_private->ptp_fd);
+		process_private->ptp_fd = -1;
+	}
+
 	return 0;
 }
 
@@ -1227,6 +1284,14 @@ eth_dev_close(struct rte_eth_dev *dev)
 		}
 	}
 
+	if (dev->process_private != NULL) {
+		struct pmd_process_private *process_private = dev->process_private;
+		if (process_private->ptp_fd >= 0) {
+			close(process_private->ptp_fd);
+			process_private->ptp_fd = -1;
+		}
+	}
+
 out:
 	rte_free(dev->process_private);
 
@@ -2097,6 +2162,35 @@ eth_dev_promiscuous_disable(struct rte_eth_dev *dev)
 	return eth_dev_change_flags(internals->if_name, 0, ~IFF_PROMISC);
 }
 
+/*
+ * In Linux, dynamic POSIX clock IDs from file descriptors (such as /dev/ptpX)
+ * are encoded with CLOCKFD (3) in the lower 3 bits and ~fd in the upper bits.
+ * As this is not defined in user-space UAPI headers, define the macro here.
+ */
+#ifndef CLOCKFD
+#define CLOCKFD 3
+#endif
+#ifndef FD_TO_CLOCKID
+#define FD_TO_CLOCKID(fd)	((clockid_t)(~(unsigned int)(fd) << 3 | CLOCKFD))
+#endif
+
+static int
+eth_af_xdp_read_clock(struct rte_eth_dev *dev, uint64_t *timestamp)
+{
+	struct pmd_process_private *process_private = dev->process_private;
+	struct timespec ts;
+
+	if (process_private == NULL || process_private->ptp_fd < 0)
+		return -ENOTSUP;
+
+	clockid_t clkid = FD_TO_CLOCKID(process_private->ptp_fd);
+	if (clock_gettime(clkid, &ts) < 0)
+		return -errno;
+
+	*timestamp = rte_timespec_to_ns(&ts);
+	return 0;
+}
+
 static const struct eth_dev_ops ops = {
 	.dev_start = eth_dev_start,
 	.dev_stop = eth_dev_stop,
@@ -2112,6 +2206,7 @@ static const struct eth_dev_ops ops = {
 	.stats_get = eth_stats_get,
 	.stats_reset = eth_stats_reset,
 	.get_monitor_addr = eth_get_monitor_addr,
+	.read_clock = eth_af_xdp_read_clock,
 };
 
 /* AF_XDP Device Plugin option works in unprivileged
@@ -2133,6 +2228,7 @@ static const struct eth_dev_ops ops_afxdp_dp = {
 	.stats_get = eth_stats_get,
 	.stats_reset = eth_stats_reset,
 	.get_monitor_addr = eth_get_monitor_addr,
+	.read_clock = eth_af_xdp_read_clock,
 };
 
 /** parse busy_budget argument */
@@ -2526,6 +2622,7 @@ init_internals(struct rte_vdev_device *dev, const char *if_name,
 	eth_dev->tx_pkt_burst = eth_af_xdp_tx;
 	eth_dev->process_private = process_private;
 
+	process_private->ptp_fd = -1;
 	for (i = 0; i < queue_cnt; i++)
 		process_private->rxq_xsk_fds[i] = -1;
 
@@ -2681,6 +2778,8 @@ rte_pmd_af_xdp_probe(struct rte_vdev_device *dev)
 			return -ENOMEM;
 		}
 
+		((struct pmd_process_private *)eth_dev->process_private)->ptp_fd = -1;
+
 		/* Obtain the xsk fds from the primary process. */
 		if (afxdp_mp_request_fds(name, eth_dev))
 			return -1;
-- 
2.55.0.508.g3f0d502094-goog