diff --git a/lib/libc/x86/sys/__vdso_gettc.c b/lib/libc/x86/sys/__vdso_gettc.c
--- a/lib/libc/x86/sys/__vdso_gettc.c
+++ b/lib/libc/x86/sys/__vdso_gettc.c
@@ -53,57 +53,133 @@
 #include <x86/ifunc.h>
 #include "libc_private.h"
 
-static void
-rdtsc_mb_lfence(void)
+static inline u_int
+rdtsc_low(u_int shift)
 {
+	u_int rv;
 
-	lfence();
+	__asm __volatile("rdtsc; shrd %%cl, %%edx, %0"
+	    : "=a" (rv) : "c" (shift) : "edx");
+	return (rv);
 }
 
-static void
-rdtsc_mb_mfence(void)
+static inline u_int
+rdtscp_low(const struct vdso_timehands *th)
+{
+	u_int rv;
+
+	__asm __volatile("rdtscp; movl %%edi,%%ecx; shrd %%cl, %%edx, %0"
+	    : "=a" (rv) : "D" (th->th_x86_shift) : "ecx", "edx");
+	return (rv);
+}
+
+static u_int
+rdtsc_low_mb_lfence(const struct vdso_timehands *th)
 {
+	lfence();
+	return (rdtsc_low(th->th_x86_shift));
+}
 
+static u_int
+rdtsc_low_mb_mfence(const struct vdso_timehands *th)
+{
 	mfence();
+	return (rdtsc_low(th->th_x86_shift));
 }
 
-static void
-rdtsc_mb_none(void)
+static u_int
+rdtsc_low_mb_none(const struct vdso_timehands *th)
 {
+	return (rdtsc_low(th->th_x86_shift));
 }
 
-DEFINE_UIFUNC(static, void, rdtsc_mb, (void))
+DEFINE_UIFUNC(static, u_int, __vdso_gettc_rdtsc_low,
+    (const struct vdso_timehands *th))
 {
-	u_int p[4];
+	u_int amd_feature, cpu_exthigh, p[4];
 	/* Not a typo, string matches our do_cpuid() registers use. */
 	static const char intel_id[] = "GenuntelineI";
 
+	if (cpu_feature != 0) {
+		do_cpuid(0x80000000, p);
+		cpu_exthigh = p[0];
+	} else {
+		cpu_exthigh = 0;
+	}
+	if (cpu_exthigh >= 0x80000001) {
+		do_cpuid(0x80000001, p);
+		amd_feature = p[3];
+	} else {
+		amd_feature = 0;
+	}
+
+	if ((amd_feature & AMDID_RDTSCP) != 0)
+		return (rdtscp_low);
 	if ((cpu_feature & CPUID_SSE2) == 0)
-		return (rdtsc_mb_none);
+		return (rdtsc_low_mb_none);
 	do_cpuid(0, p);
 	return (memcmp(p + 1, intel_id, sizeof(intel_id) - 1) == 0 ?
-	    rdtsc_mb_lfence : rdtsc_mb_mfence);
+	    rdtsc_low_mb_lfence : rdtsc_low_mb_mfence);
 }
 
 static u_int
-__vdso_gettc_rdtsc_low(const struct vdso_timehands *th)
+rdtsc32_mb_lfence(void)
 {
-	u_int rv;
-
-	rdtsc_mb();
-	__asm __volatile("rdtsc; shrd %%cl, %%edx, %0"
-	    : "=a" (rv) : "c" (th->th_x86_shift) : "edx");
-	return (rv);
+	lfence();
+	return (rdtsc32());
 }
 
 static u_int
-__vdso_rdtsc32(void)
+rdtsc32_mb_mfence(void)
 {
+	mfence();
+	return (rdtsc32());
+}
 
-	rdtsc_mb();
+static u_int
+rdtsc32_mb_none(void)
+{
 	return (rdtsc32());
 }
 
+static u_int
+rdtscp32(void)
+{
+	uint32_t rv;
+
+	__asm __volatile("rdtscp"
+	    : "=a" (rv) :: "ecx", "edx");
+	return (rv);
+}
+
+DEFINE_UIFUNC(static, u_int, __vdso_gettc_rdtsc32, (void))
+{
+	u_int amd_feature, cpu_exthigh, p[4];
+	/* Not a typo, string matches our do_cpuid() registers use. */
+	static const char intel_id[] = "GenuntelineI";
+
+	if (cpu_feature != 0) {
+		do_cpuid(0x80000000, p);
+		cpu_exthigh = p[0];
+	} else {
+		cpu_exthigh = 0;
+	}
+	if (cpu_exthigh >= 0x80000001) {
+		do_cpuid(0x80000001, p);
+		amd_feature = p[3];
+	} else {
+		amd_feature = 0;
+	}
+
+	if ((amd_feature & AMDID_RDTSCP) != 0)
+		return (rdtscp32);
+	if ((cpu_feature & CPUID_SSE2) == 0)
+		return (rdtsc32_mb_none);
+	do_cpuid(0, p);
+	return (memcmp(p + 1, intel_id, sizeof(intel_id) - 1) == 0 ?
+	    rdtsc32_mb_lfence : rdtsc32_mb_mfence);
+}
+
 #define	HPET_DEV_MAP_MAX	10
 static volatile char *hpet_dev_map[HPET_DEV_MAP_MAX];
 
@@ -199,7 +275,7 @@
 		scale = tsc_ref->tsc_scale;
 		ofs = tsc_ref->tsc_ofs;
 
-		rdtsc_mb();
+		mfence();	/* XXXKIB */
 		tsc = rdtsc();
 
 		/* ret = ((tsc * scale) >> 64) + ofs */
@@ -231,7 +307,7 @@
 	switch (th->th_algo) {
 	case VDSO_TH_ALGO_X86_TSC:
 		*tc = th->th_x86_shift > 0 ? __vdso_gettc_rdtsc_low(th) :
-		    __vdso_rdtsc32();
+		    __vdso_gettc_rdtsc32();
 		return (0);
 	case VDSO_TH_ALGO_X86_HPET:
 		idx = th->th_x86_hpet_idx;
diff --git a/sys/x86/x86/tsc.c b/sys/x86/x86/tsc.c
--- a/sys/x86/x86/tsc.c
+++ b/sys/x86/x86/tsc.c
@@ -91,12 +91,14 @@
     int status);
 static void tsc_freq_changing(void *arg, const struct cf_level *level,
     int *status);
-static unsigned tsc_get_timecount(struct timecounter *tc);
-static inline unsigned tsc_get_timecount_low(struct timecounter *tc);
-static unsigned tsc_get_timecount_lfence(struct timecounter *tc);
-static unsigned tsc_get_timecount_low_lfence(struct timecounter *tc);
-static unsigned tsc_get_timecount_mfence(struct timecounter *tc);
-static unsigned tsc_get_timecount_low_mfence(struct timecounter *tc);
+static u_int tsc_get_timecount(struct timecounter *tc);
+static inline u_int tsc_get_timecount_low(struct timecounter *tc);
+static u_int tsc_get_timecount_lfence(struct timecounter *tc);
+static u_int tsc_get_timecount_low_lfence(struct timecounter *tc);
+static u_int tsc_get_timecount_mfence(struct timecounter *tc);
+static u_int tsc_get_timecount_low_mfence(struct timecounter *tc);
+static u_int tscp_get_timecount(struct timecounter *tc);
+static u_int tscp_get_timecount_low(struct timecounter *tc);
 static void tsc_levels_changed(void *arg, int unit);
 static uint32_t x86_tsc_vdso_timehands(struct vdso_timehands *vdso_th,
     struct timecounter *tc);
@@ -628,7 +630,10 @@
 init:
 	for (shift = 0; shift <= 31 && (tsc_freq >> shift) > max_freq; shift++)
 		;
-	if ((cpu_feature & CPUID_SSE2) != 0 && mp_ncpus > 1) {
+	if ((amd_feature & AMDID_RDTSCP) != 0) {
+		tsc_timecounter.tc_get_timecount = shift > 0 ?
+		    tscp_get_timecount_low : tscp_get_timecount;
+	} else if ((cpu_feature & CPUID_SSE2) != 0 && mp_ncpus > 1) {
 		if (cpu_vendor_id == CPU_VENDOR_AMD ||
 		    cpu_vendor_id == CPU_VENDOR_HYGON) {
 			tsc_timecounter.tc_get_timecount = shift > 0 ?
@@ -783,6 +788,16 @@
 	return (rdtsc32());
 }
 
+static u_int
+tscp_get_timecount(struct timecounter *tc __unused)
+{
+	uint32_t rv;
+
+	__asm __volatile("rdtscp"
+	    : "=a" (rv) :: "ecx", "edx");
+	return (rv);
+}
+
 static inline u_int
 tsc_get_timecount_low(struct timecounter *tc)
 {
@@ -793,6 +808,16 @@
 	return (rv);
 }
 
+static u_int
+tscp_get_timecount_low(struct timecounter *tc)
+{
+	uint32_t rv;
+
+	__asm __volatile("rdtscp; movl %1, %%ecx; shrd %%cl, %%edx, %0"
+	    : "=a" (rv) : "m" (tc->tc_priv) : "ecx", "edx");
+	return (rv);
+}
+
 static u_int
 tsc_get_timecount_lfence(struct timecounter *tc __unused)
 {