On Fri, 2024-06-28 at 13:33 +0200, Peter Hilber wrote:

quoted

        /*
         * What time is exposed in the time_sec/time_frac_sec fields?
         */
        uint8_t time_type;
#define VMCLOCK_TIME_UNKNOWN            0       /* Invalid / no time exposed */
#define VMCLOCK_TIME_UTC                1       /* Since 1970-01-01 00:00:00z */
#define VMCLOCK_TIME_TAI                2       /* Since 1970-01-01 00:00:00z */
#define VMCLOCK_TIME_MONOTONIC          3       /* Since undefined epoch */

        /* Bit shift for counter_period_frac_sec and its error rate */
        uint8_t counter_period_shift;

        /*
         * Unlike in NTP, this can indicate a leap second in the past. This
         * is needed to allow guests to derive an imprecise clock with
         * smeared leap seconds for themselves, as some modes of smearing
         * need the adjustments to continue even after the moment at which
         * the leap second should have occurred.
         */
        int8_t leapsecond_direction;
        uint64_t leapsecond_tai_sec; /* Since 1970-01-01 00:00:00z */

        /*
         * Paired values of counter and UTC at a given point in time.
         */
        uint64_t counter_value;
        uint64_t time_sec; /* Since 1970-01-01 00:00:00z */

Nitpick: The comment is not valid any more for TIME_MONOTONIC.

Ah yes, I "moved" that comment up to the UTC/TAI time_type values, but
neglected to actually delete it from here. Fixed; thanks.