ptp: add PTP_SYS_OFFSET_EXTENDED ioctl

The PTP_SYS_OFFSET ioctl, which can be used to measure the offset
between a PHC and the system clock, includes the total time that the
driver needs to read the PHC timestamp.

This typically involves reading of multiple PCI registers (sometimes in
multiple iterations) and the register that contains the lowest bits of
the timestamp is not read in the middle between the two readings of the
system clock. This asymmetry causes the measured offset to have a
significant error.

Introduce a new ioctl, driver function, and helper functions, which
allow the reading of the lowest register to be isolated from the other
readings in order to reduce the asymmetry. The ioctl returns three
timestamps for each measurement:
- system time right before reading the lowest bits of the PHC timestamp
- PHC time
- system time immediately after reading the lowest bits of the PHC
  timestamp

Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Jacob Keller <jacob.e.keller@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Miroslav Lichvar <mlichvar@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/ptp/ptp_chardev.c

@@ -122,10 +122,12 @@ int ptp_open(struct posix_clock *pc, fmode_t fmode)
 long ptp_ioctl(struct posix_clock *pc, unsigned int cmd, unsigned long arg)
 {
 	struct ptp_clock *ptp = container_of(pc, struct ptp_clock, clock);
+	struct ptp_sys_offset_extended *extoff = NULL;
 	struct ptp_sys_offset_precise precise_offset;
 	struct system_device_crosststamp xtstamp;
 	struct ptp_clock_info *ops = ptp->info;
 	struct ptp_sys_offset *sysoff = NULL;
+	struct ptp_system_timestamp sts;
 	struct ptp_clock_request req;
 	struct ptp_clock_caps caps;
 	struct ptp_clock_time *pct;
@@ -211,6 +213,36 @@ long ptp_ioctl(struct posix_clock *pc, unsigned int cmd, unsigned long arg)
 			err = -EFAULT;
 		break;
 
+	case PTP_SYS_OFFSET_EXTENDED:
+		if (!ptp->info->gettimex64) {
+			err = -EOPNOTSUPP;
+			break;
+		}
+		extoff = memdup_user((void __user *)arg, sizeof(*extoff));
+		if (IS_ERR(extoff)) {
+			err = PTR_ERR(extoff);
+			extoff = NULL;
+			break;
+		}
+		if (extoff->n_samples > PTP_MAX_SAMPLES) {
+			err = -EINVAL;
+			break;
+		}
+		for (i = 0; i < extoff->n_samples; i++) {
+			err = ptp->info->gettimex64(ptp->info, &ts, &sts);
+			if (err)
+				goto out;
+			extoff->ts[i][0].sec = sts.pre_ts.tv_sec;
+			extoff->ts[i][0].nsec = sts.pre_ts.tv_nsec;
+			extoff->ts[i][1].sec = ts.tv_sec;
+			extoff->ts[i][1].nsec = ts.tv_nsec;
+			extoff->ts[i][2].sec = sts.post_ts.tv_sec;
+			extoff->ts[i][2].nsec = sts.post_ts.tv_nsec;
+		}
+		if (copy_to_user((void __user *)arg, extoff, sizeof(*extoff)))
+			err = -EFAULT;
+		break;
+
 	case PTP_SYS_OFFSET:
 		sysoff = memdup_user((void __user *)arg, sizeof(*sysoff));
 		if (IS_ERR(sysoff)) {
@@ -284,6 +316,7 @@ long ptp_ioctl(struct posix_clock *pc, unsigned int cmd, unsigned long arg)
 	}
 
 out:
+	kfree(extoff);
 	kfree(sysoff);
 	return err;
 }

include/linux/ptp_clock_kernel.h

@@ -39,6 +39,15 @@ struct ptp_clock_request {
 };
 
 struct system_device_crosststamp;
+
+/**
+ * struct ptp_system_timestamp - system time corresponding to a PHC timestamp
+ */
+struct ptp_system_timestamp {
+	struct timespec64 pre_ts;
+	struct timespec64 post_ts;
+};
+
 /**
  * struct ptp_clock_info - decribes a PTP hardware clock
  *
@@ -75,6 +84,14 @@ struct system_device_crosststamp;
  * @gettime64:  Reads the current time from the hardware clock.
  *              parameter ts: Holds the result.
  *
+ * @gettimex64:  Reads the current time from the hardware clock and optionally
+ *               also the system clock.
+ *               parameter ts: Holds the PHC timestamp.
+ *               parameter sts: If not NULL, it holds a pair of timestamps from
+ *               the system clock. The first reading is made right before
+ *               reading the lowest bits of the PHC timestamp and the second
+ *               reading immediately follows that.
+ *
  * @getcrosststamp:  Reads the current time from the hardware clock and
  *                   system clock simultaneously.
  *                   parameter cts: Contains timestamp (device,system) pair,
@@ -124,6 +141,8 @@ struct ptp_clock_info {
 	int (*adjfreq)(struct ptp_clock_info *ptp, s32 delta);
 	int (*adjtime)(struct ptp_clock_info *ptp, s64 delta);
 	int (*gettime64)(struct ptp_clock_info *ptp, struct timespec64 *ts);
+	int (*gettimex64)(struct ptp_clock_info *ptp, struct timespec64 *ts,
+			  struct ptp_system_timestamp *sts);
 	int (*getcrosststamp)(struct ptp_clock_info *ptp,
 			      struct system_device_crosststamp *cts);
 	int (*settime64)(struct ptp_clock_info *p, const struct timespec64 *ts);
@@ -247,4 +266,16 @@ static inline int ptp_schedule_worker(struct ptp_clock *ptp,
 
 #endif
 
+static inline void ptp_read_system_prets(struct ptp_system_timestamp *sts)
+{
+	if (sts)
+		ktime_get_real_ts64(&sts->pre_ts);
+}
+
+static inline void ptp_read_system_postts(struct ptp_system_timestamp *sts)
+{
+	if (sts)
+		ktime_get_real_ts64(&sts->post_ts);
+}
+
 #endif

include/uapi/linux/ptp_clock.h

@@ -84,6 +84,16 @@ struct ptp_sys_offset {
 	struct ptp_clock_time ts[2 * PTP_MAX_SAMPLES + 1];
 };
 
+struct ptp_sys_offset_extended {
+	unsigned int n_samples; /* Desired number of measurements. */
+	unsigned int rsv[3];    /* Reserved for future use. */
+	/*
+	 * Array of [system, phc, system] time stamps. The kernel will provide
+	 * 3*n_samples time stamps.
+	 */
+	struct ptp_clock_time ts[PTP_MAX_SAMPLES][3];
+};
+
 struct ptp_sys_offset_precise {
 	struct ptp_clock_time device;
 	struct ptp_clock_time sys_realtime;
@@ -136,6 +146,8 @@ struct ptp_pin_desc {
 #define PTP_PIN_SETFUNC    _IOW(PTP_CLK_MAGIC, 7, struct ptp_pin_desc)
 #define PTP_SYS_OFFSET_PRECISE \
 	_IOWR(PTP_CLK_MAGIC, 8, struct ptp_sys_offset_precise)
+#define PTP_SYS_OFFSET_EXTENDED \
+	_IOW(PTP_CLK_MAGIC, 9, struct ptp_sys_offset_extended)
 
 struct ptp_extts_event {
 	struct ptp_clock_time t; /* Time event occured. */