diff --git a/bpf/process/bpf_execve_event.c b/bpf/process/bpf_execve_event.c
index 76b38ce911d..fd6849fa360 100644
--- a/bpf/process/bpf_execve_event.c
+++ b/bpf/process/bpf_execve_event.c
@@ -166,6 +166,84 @@ read_exe(struct task_struct *task, struct heap_exe *exe)
 }
 #endif
 
+/* Parameters used to convert the timespec values: */
+#define MSEC_PER_SEC  1000L
+#define USEC_PER_MSEC 1000L
+#define NSEC_PER_USEC 1000L
+#define NSEC_PER_MSEC 1000000L
+#define USEC_PER_SEC  1000000L
+#define NSEC_PER_SEC  1000000000L
+#define PSEC_PER_SEC  1000000000000LL
+#define FSEC_PER_SEC  1000000000000000LL
+
+/* Parameters used to convert the timespec values: */
+#define PSEC_PER_NSEC 1000L
+
+/* Located here for timespec[64]_valid_strict */
+#define TIME64_MAX ((s64) ~((u64)1 << 63))
+#define TIME64_MIN (-TIME64_MAX - 1)
+
+#define KTIME_MAX     ((s64) ~((u64)1 << 63))
+#define KTIME_MIN     (-KTIME_MAX - 1)
+#define KTIME_SEC_MAX (KTIME_MAX / NSEC_PER_SEC)
+#define KTIME_SEC_MIN (KTIME_MIN / NSEC_PER_SEC)
+
+#define USER_HZ 100 /* some user interfaces are */
+
+static __attribute__((always_inline)) inline s64 timespec64_to_ns(struct timespec64 *ts)
+{
+	__s64 tv_sec = BPF_CORE_READ(ts, tv_sec);
+	long tv_nsec = BPF_CORE_READ(ts, tv_nsec);
+
+	/* Prevent multiplication overflow / underflow */
+	if (tv_sec >= KTIME_SEC_MAX)
+		return KTIME_MAX;
+
+	if (tv_sec <= KTIME_SEC_MIN)
+		return KTIME_MIN;
+
+	return ((s64)tv_sec * NSEC_PER_SEC) + tv_nsec;
+}
+
+static __attribute__((always_inline)) inline u64 timens_add_boottime_ns(u64 nsec)
+{
+	struct task_struct *task = (struct task_struct *)get_current_task();
+	struct time_namespace *time_ns = BPF_CORE_READ(task, nsproxy, time_ns);
+	struct timens_offsets ns_offsets;
+
+	probe_read(&ns_offsets, sizeof(struct timens_offsets), _(&time_ns->offsets));
+
+	return nsec + timespec64_to_ns(&ns_offsets.boottime);
+}
+
+static __attribute__((always_inline)) inline u64 div_u64_rem(u64 dividend, u32 divisor, u32 *remainder)
+{
+	*remainder = dividend % divisor;
+	return dividend / divisor;
+}
+
+static __attribute__((always_inline)) inline u64 div_u64(u64 dividend, u32 divisor)
+{
+	u32 remainder;
+	return div_u64_rem(dividend, divisor, &remainder);
+}
+
+static __attribute__((always_inline)) inline u64 nsec_to_clock_t(u64 x)
+{
+#if (NSEC_PER_SEC % USER_HZ) == 0
+	return div_u64(x, NSEC_PER_SEC / USER_HZ);
+#elif (USER_HZ % 512) == 0
+	return div_u64(x * USER_HZ / 512, NSEC_PER_SEC / 512);
+#else
+	/*
+         * max relative error 5.7e-8 (1.8s per year) for USER_HZ <= 1024,
+         * overflow after 64.99 years.
+         * exact for HZ=60, 72, 90, 120, 144, 180, 300, 600, 900, ...
+         */
+	return div_u64(x * 9, (9ull * NSEC_PER_SEC + (USER_HZ / 2)) / USER_HZ);
+#endif
+}
+
 __attribute__((section("tracepoint/sys_execve"), used)) int
 event_execve(struct sched_execve_args *ctx)
 {
@@ -175,7 +253,7 @@ event_execve(struct sched_execve_args *ctx)
 	struct execve_map_value *parent;
 	struct msg_process *p;
 	__u32 zero = 0;
-	__u64 pid;
+	__u64 pid, start_boottime;
 
 	event = map_lookup_elem(&execve_msg_heap_map, &zero);
 	if (!event)
@@ -200,7 +278,8 @@ event_execve(struct sched_execve_args *ctx)
 	p->pid = pid >> 32;
 	p->tid = (__u32)pid;
 	p->nspid = get_task_pid_vnr();
-	p->ktime = ktime_get_ns();
+	start_boottime = BPF_CORE_READ(task, start_boottime);
+	p->ktime = nsec_to_clock_t(timens_add_boottime_ns(start_boottime)) * 10000000LLU; // similarly to what /proc/<PID>/stat provides (i.e. https://elixir.bootlin.com/linux/v6.1.75/source/fs/proc/array.c#L568)
 	p->size = offsetof(struct msg_process, args);
 	p->auid = get_auid();
 	p->uid = get_current_uid_gid();