Merge remote-tracking branch 'torvalds/master' into perf/core

To pick up fixes.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

46 files changed, 992 insertions, 682 deletions

diff --git a/tools/bootconfig/scripts/bconf2ftrace.sh b/tools/bootconfig/scripts/bconf2ftrace.sh
index 595e164dc352..feb30c2c7881 100755
--- a/tools/bootconfig/scripts/bconf2ftrace.sh
+++ b/tools/bootconfig/scripts/bconf2ftrace.sh
@@ -152,6 +152,7 @@ setup_instance() { # [instance]
 	set_array_of ${instance}.options ${instancedir}/trace_options
 	set_value_of ${instance}.trace_clock ${instancedir}/trace_clock
 	set_value_of ${instance}.cpumask ${instancedir}/tracing_cpumask
+	set_value_of ${instance}.tracing_on ${instancedir}/tracing_on
 	set_value_of ${instance}.tracer ${instancedir}/current_tracer
 	set_array_of ${instance}.ftrace.filters \
 		${instancedir}/set_ftrace_filter
diff --git a/tools/bootconfig/scripts/ftrace2bconf.sh b/tools/bootconfig/scripts/ftrace2bconf.sh
index 6c0d4b61e0c2..a0c3bcc6da4f 100755
--- a/tools/bootconfig/scripts/ftrace2bconf.sh
+++ b/tools/bootconfig/scripts/ftrace2bconf.sh
@@ -221,6 +221,10 @@ instance_options() { # [instance-name]
 	if [ `echo $val | sed -e s/f//g`x != x ]; then
 		emit_kv $PREFIX.cpumask = $val
 	fi
+	val=`cat $INSTANCE/tracing_on`
+	if [ `echo $val | sed -e s/f//g`x != x ]; then
+		emit_kv $PREFIX.tracing_on = $val
+	fi
 
 	val=
 	for i in `cat $INSTANCE/set_event`; do
diff --git a/tools/bpf/bpftool/net.c b/tools/bpf/bpftool/net.c
index 3fae61ef6339..ff3aa0cf3997 100644
--- a/tools/bpf/bpftool/net.c
+++ b/tools/bpf/bpftool/net.c
@@ -11,7 +11,6 @@
 #include <bpf/bpf.h>
 #include <bpf/libbpf.h>
 #include <net/if.h>
-#include <linux/if.h>
 #include <linux/rtnetlink.h>
 #include <linux/socket.h>
 #include <linux/tc_act/tc_bpf.h>
diff --git a/tools/bpf/resolve_btfids/main.c b/tools/bpf/resolve_btfids/main.c
index e3ea569ee125..7409d7860aa6 100644
--- a/tools/bpf/resolve_btfids/main.c
+++ b/tools/bpf/resolve_btfids/main.c
@@ -139,6 +139,8 @@ int eprintf(int level, int var, const char *fmt, ...)
 #define pr_debug2(fmt, ...) pr_debugN(2, pr_fmt(fmt), ##__VA_ARGS__)
 #define pr_err(fmt, ...) \
 	eprintf(0, verbose, pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_info(fmt, ...) \
+	eprintf(0, verbose, pr_fmt(fmt), ##__VA_ARGS__)
 
 static bool is_btf_id(const char *name)
 {
@@ -472,7 +474,7 @@ static int symbols_resolve(struct object *obj)
 	int nr_funcs    = obj->nr_funcs;
 	int err, type_id;
 	struct btf *btf;
-	__u32 nr;
+	__u32 nr_types;
 
 	btf = btf__parse(obj->btf ?: obj->path, NULL);
 	err = libbpf_get_error(btf);
@@ -483,12 +485,12 @@ static int symbols_resolve(struct object *obj)
 	}
 
 	err = -1;
-	nr  = btf__get_nr_types(btf);
+	nr_types = btf__get_nr_types(btf);
 
 	/*
 	 * Iterate all the BTF types and search for collected symbol IDs.
 	 */
-	for (type_id = 1; type_id <= nr; type_id++) {
+	for (type_id = 1; type_id <= nr_types; type_id++) {
 		const struct btf_type *type;
 		struct rb_root *root;
 		struct btf_id *id;
@@ -526,8 +528,13 @@ static int symbols_resolve(struct object *obj)
 
 		id = btf_id__find(root, str);
 		if (id) {
-			id->id = type_id;
-			(*nr)--;
+			if (id->id) {
+				pr_info("WARN: multiple IDs found for '%s': %d, %d - using %d\n",
+					str, id->id, type_id, id->id);
+			} else {
+				id->id = type_id;
+				(*nr)--;
+			}
 		}
 	}
 
diff --git a/tools/include/linux/build_bug.h b/tools/include/linux/build_bug.h
index ce365d212768..cc7070c7439b 100644
--- a/tools/include/linux/build_bug.h
+++ b/tools/include/linux/build_bug.h
@@ -79,9 +79,4 @@
 #define __static_assert(expr, msg, ...) _Static_assert(expr, msg)
 #endif // static_assert
 
-#ifdef __GENKSYMS__
-/* genksyms gets confused by _Static_assert */
-#define _Static_assert(expr, ...)
-#endif
-
 #endif	/* _LINUX_BUILD_BUG_H */
diff --git a/tools/include/uapi/linux/kvm.h b/tools/include/uapi/linux/kvm.h
index 886802b8ffba..374c67875cdb 100644
--- a/tools/include/uapi/linux/kvm.h
+++ b/tools/include/uapi/linux/kvm.h
@@ -251,6 +251,7 @@ struct kvm_hyperv_exit {
 #define KVM_EXIT_X86_RDMSR        29
 #define KVM_EXIT_X86_WRMSR        30
 #define KVM_EXIT_DIRTY_RING_FULL  31
+#define KVM_EXIT_AP_RESET_HOLD    32
 
 /* For KVM_EXIT_INTERNAL_ERROR */
 /* Emulate instruction failed. */
@@ -573,6 +574,7 @@ struct kvm_vapic_addr {
 #define KVM_MP_STATE_CHECK_STOP        6
 #define KVM_MP_STATE_OPERATING         7
 #define KVM_MP_STATE_LOAD              8
+#define KVM_MP_STATE_AP_RESET_HOLD     9
 
 struct kvm_mp_state {
 	__u32 mp_state;
diff --git a/tools/lib/perf/tests/test-cpumap.c b/tools/lib/perf/tests/test-cpumap.c
index c8d45091e7c2..c70e9e03af3e 100644
--- a/tools/lib/perf/tests/test-cpumap.c
+++ b/tools/lib/perf/tests/test-cpumap.c
@@ -27,5 +27,5 @@ int main(int argc, char **argv)
 	perf_cpu_map__put(cpus);
 
 	__T_END;
-	return 0;
+	return tests_failed == 0 ? 0 : -1;
 }
diff --git a/tools/lib/perf/tests/test-evlist.c b/tools/lib/perf/tests/test-evlist.c
index 6d8ebe0c2504..e2ac0b7f432e 100644
--- a/tools/lib/perf/tests/test-evlist.c
+++ b/tools/lib/perf/tests/test-evlist.c
@@ -208,13 +208,13 @@ static int test_mmap_thread(void)
 	char path[PATH_MAX];
 	int id, err, pid, go_pipe[2];
 	union perf_event *event;
-	char bf;
 	int count = 0;
 
 	snprintf(path, PATH_MAX, "%s/kernel/debug/tracing/events/syscalls/sys_enter_prctl/id",
 		 sysfs__mountpoint());
 
 	if (filename__read_int(path, &id)) {
+		tests_failed++;
 		fprintf(stderr, "error: failed to get tracepoint id: %s\n", path);
 		return -1;
 	}
@@ -229,6 +229,7 @@ static int test_mmap_thread(void)
 	pid = fork();
 	if (!pid) {
 		int i;
+		char bf;
 
 		read(go_pipe[0], &bf, 1);
 
@@ -266,7 +267,7 @@ static int test_mmap_thread(void)
 	perf_evlist__enable(evlist);
 
 	/* kick the child and wait for it to finish */
-	write(go_pipe[1], &bf, 1);
+	write(go_pipe[1], "A", 1);
 	waitpid(pid, NULL, 0);
 
 	/*
@@ -409,5 +410,5 @@ int main(int argc, char **argv)
 	test_mmap_cpus();
 
 	__T_END;
-	return 0;
+	return tests_failed == 0 ? 0 : -1;
 }
diff --git a/tools/lib/perf/tests/test-evsel.c b/tools/lib/perf/tests/test-evsel.c
index 135722ac965b..0ad82d7a2a51 100644
--- a/tools/lib/perf/tests/test-evsel.c
+++ b/tools/lib/perf/tests/test-evsel.c
@@ -131,5 +131,5 @@ int main(int argc, char **argv)
 	test_stat_thread_enable();
 
 	__T_END;
-	return 0;
+	return tests_failed == 0 ? 0 : -1;
 }
diff --git a/tools/lib/perf/tests/test-threadmap.c b/tools/lib/perf/tests/test-threadmap.c
index 7dc4d6fbedde..384471441b48 100644
--- a/tools/lib/perf/tests/test-threadmap.c
+++ b/tools/lib/perf/tests/test-threadmap.c
@@ -27,5 +27,5 @@ int main(int argc, char **argv)
 	perf_thread_map__put(threads);
 
 	__T_END;
-	return 0;
+	return tests_failed == 0 ? 0 : -1;
 }
diff --git a/tools/perf/examples/bpf/5sec.c b/tools/perf/examples/bpf/5sec.c
index 65c4ff6892d9..e6b6181c6dc6 100644
--- a/tools/perf/examples/bpf/5sec.c
+++ b/tools/perf/examples/bpf/5sec.c
@@ -39,7 +39,7 @@
    Copyright (C) 2018 Red Hat, Inc., Arnaldo Carvalho de Melo <acme@redhat.com>
 */
 
-#include <bpf/bpf.h>
+#include <bpf.h>
 
 #define NSEC_PER_SEC	1000000000L
 
diff --git a/tools/perf/tests/shell/stat+shadow_stat.sh b/tools/perf/tests/shell/stat+shadow_stat.sh
index 249dfe48cf6a..ebebd3596cf9 100755
--- a/tools/perf/tests/shell/stat+shadow_stat.sh
+++ b/tools/perf/tests/shell/stat+shadow_stat.sh
@@ -9,31 +9,29 @@ perf stat -a true > /dev/null 2>&1 || exit 2
 
 test_global_aggr()
 {
-	local cyc
-
 	perf stat -a --no-big-num -e cycles,instructions sleep 1  2>&1 | \
 	grep -e cycles -e instructions | \
 	while read num evt hash ipc rest
 	do
 		# skip not counted events
-		if [[ $num == "<not" ]]; then
+		if [ "$num" = "<not" ]; then
 			continue
 		fi
 
 		# save cycles count
-		if [[ $evt == "cycles" ]]; then
+		if [ "$evt" = "cycles" ]; then
 			cyc=$num
 			continue
 		fi
 
 		# skip if no cycles
-		if [[ -z $cyc ]]; then
+		if [ -z "$cyc" ]; then
 			continue
 		fi
 
 		# use printf for rounding and a leading zero
-		local res=`printf "%.2f" $(echo "scale=6; $num / $cyc" | bc -q)`
-		if [[ $ipc != $res ]]; then
+		res=`printf "%.2f" $(echo "scale=6; $num / $cyc" | bc -q)`
+		if [ "$ipc" != "$res" ]; then
 			echo "IPC is different: $res != $ipc  ($num / $cyc)"
 			exit 1
 		fi
@@ -42,32 +40,32 @@ test_global_aggr()
 
 test_no_aggr()
 {
-	declare -A results
-
 	perf stat -a -A --no-big-num -e cycles,instructions sleep 1  2>&1 | \
 	grep ^CPU | \
 	while read cpu num evt hash ipc rest
 	do
 		# skip not counted events
-		if [[ $num == "<not" ]]; then
+		if [ "$num" = "<not" ]; then
 			continue
 		fi
 
 		# save cycles count
-		if [[ $evt == "cycles" ]]; then
-			results[$cpu]=$num
+		if [ "$evt" = "cycles" ]; then
+			results="$results $cpu:$num"
 			continue
 		fi
 
+		cyc=${results##* $cpu:}
+		cyc=${cyc%% *}
+
 		# skip if no cycles
-		local cyc=${results[$cpu]}
-		if [[ -z $cyc ]]; then
+		if [ -z "$cyc" ]; then
 			continue
 		fi
 
 		# use printf for rounding and a leading zero
-		local res=`printf "%.2f" $(echo "scale=6; $num / $cyc" | bc -q)`
-		if [[ $ipc != $res ]]; then
+		res=`printf "%.2f" $(echo "scale=6; $num / $cyc" | bc -q)`
+		if [ "$ipc" != "$res" ]; then
 			echo "IPC is different for $cpu: $res != $ipc  ($num / $cyc)"
 			exit 1
 		fi
diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 062383e225a3..c4ed3dc2c8f4 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -3323,6 +3323,14 @@ int perf_session__write_header(struct perf_session *session,
 	attr_offset = lseek(ff.fd, 0, SEEK_CUR);
 
 	evlist__for_each_entry(evlist, evsel) {
+		if (evsel->core.attr.size < sizeof(evsel->core.attr)) {
+			/*
+			 * We are likely in "perf inject" and have read
+			 * from an older file. Update attr size so that
+			 * reader gets the right offset to the ids.
+			 */
+			evsel->core.attr.size = sizeof(evsel->core.attr);
+		}
 		f_attr = (struct perf_file_attr){
 			.attr = evsel->core.attr,
 			.ids  = {
diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 522ea3236bcc..ab8a6b3e801d 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -2994,7 +2994,7 @@ int machines__for_each_thread(struct machines *machines,
 
 pid_t machine__get_current_tid(struct machine *machine, int cpu)
 {
-	int nr_cpus = min(machine->env->nr_cpus_online, MAX_NR_CPUS);
+	int nr_cpus = min(machine->env->nr_cpus_avail, MAX_NR_CPUS);
 
 	if (cpu < 0 || cpu >= nr_cpus || !machine->current_tid)
 		return -1;
@@ -3006,7 +3006,7 @@ int machine__set_current_tid(struct machine *machine, int cpu, pid_t pid,
 			     pid_t tid)
 {
 	struct thread *thread;
-	int nr_cpus = min(machine->env->nr_cpus_online, MAX_NR_CPUS);
+	int nr_cpus = min(machine->env->nr_cpus_avail, MAX_NR_CPUS);
 
 	if (cpu < 0)
 		return -EINVAL;
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 492c994c948a..5b1a31cae047 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -2410,7 +2410,7 @@ int perf_session__cpu_bitmap(struct perf_session *session,
 {
 	int i, err = -1;
 	struct perf_cpu_map *map;
-	int nr_cpus = min(session->header.env.nr_cpus_online, MAX_NR_CPUS);
+	int nr_cpus = min(session->header.env.nr_cpus_avail, MAX_NR_CPUS);
 
 	for (i = 0; i < PERF_TYPE_MAX; ++i) {
 		struct evsel *evsel;
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index 901265127e36..12eafd12a693 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -8,6 +8,7 @@
 #include "evlist.h"
 #include "expr.h"
 #include "metricgroup.h"
+#include "cgroup.h"
 #include <linux/zalloc.h>
 
 /*
@@ -28,6 +29,7 @@ struct saved_value {
 	enum stat_type type;
 	int ctx;
 	int cpu;
+	struct cgroup *cgrp;
 	struct runtime_stat *stat;
 	struct stats stats;
 	u64 metric_total;
@@ -57,6 +59,9 @@ static int saved_value_cmp(struct rb_node *rb_node, const void *entry)
 	if (a->ctx != b->ctx)
 		return a->ctx - b->ctx;
 
+	if (a->cgrp != b->cgrp)
+		return (char *)a->cgrp < (char *)b->cgrp ? -1 : +1;
+
 	if (a->evsel == NULL && b->evsel == NULL) {
 		if (a->stat == b->stat)
 			return 0;
@@ -100,7 +105,8 @@ static struct saved_value *saved_value_lookup(struct evsel *evsel,
 					      bool create,
 					      enum stat_type type,
 					      int ctx,
-					      struct runtime_stat *st)
+					      struct runtime_stat *st,
+					      struct cgroup *cgrp)
 {
 	struct rblist *rblist;
 	struct rb_node *nd;
@@ -110,10 +116,15 @@ static struct saved_value *saved_value_lookup(struct evsel *evsel,
 		.type = type,
 		.ctx = ctx,
 		.stat = st,
+		.cgrp = cgrp,
 	};
 
 	rblist = &st->value_list;
 
+	/* don't use context info for clock events */
+	if (type == STAT_NSECS)
+		dm.ctx = 0;
+
 	nd = rblist__find(rblist, &dm);
 	if (nd)
 		return container_of(nd, struct saved_value, rb_node);
@@ -191,12 +202,18 @@ void perf_stat__reset_shadow_per_stat(struct runtime_stat *st)
 	reset_stat(st);
 }
 
+struct runtime_stat_data {
+	int ctx;
+	struct cgroup *cgrp;
+};
+
 static void update_runtime_stat(struct runtime_stat *st,
 				enum stat_type type,
-				int ctx, int cpu, u64 count)
+				int cpu, u64 count,
+				struct runtime_stat_data *rsd)
 {
-	struct saved_value *v = saved_value_lookup(NULL, cpu, true,
-						   type, ctx, st);
+	struct saved_value *v = saved_value_lookup(NULL, cpu, true, type,
+						   rsd->ctx, st, rsd->cgrp);
 
 	if (v)
 		update_stats(&v->stats, count);
@@ -210,82 +227,86 @@ static void update_runtime_stat(struct runtime_stat *st,
 void perf_stat__update_shadow_stats(struct evsel *counter, u64 count,
 				    int cpu, struct runtime_stat *st)
 {
-	int ctx = evsel_context(counter);
 	u64 count_ns = count;
 	struct saved_value *v;
+	struct runtime_stat_data rsd = {
+		.ctx = evsel_context(counter),
+		.cgrp = counter->cgrp,
+	};
 
 	count *= counter->scale;
 
 	if (evsel__is_clock(counter))
-		update_runtime_stat(st, STAT_NSECS, 0, cpu, count_ns);
+		update_runtime_stat(st, STAT_NSECS, cpu, count_ns, &rsd);
 	else if (evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
-		update_runtime_stat(st, STAT_CYCLES, ctx, cpu, count);
+		update_runtime_stat(st, STAT_CYCLES, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, CYCLES_IN_TX))
-		update_runtime_stat(st, STAT_CYCLES_IN_TX, ctx, cpu, count);
+		update_runtime_stat(st, STAT_CYCLES_IN_TX, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TRANSACTION_START))
-		update_runtime_stat(st, STAT_TRANSACTION, ctx, cpu, count);
+		update_runtime_stat(st, STAT_TRANSACTION, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, ELISION_START))
-		update_runtime_stat(st, STAT_ELISION, ctx, cpu, count);
+		update_runtime_stat(st, STAT_ELISION, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_TOTAL_SLOTS))
 		update_runtime_stat(st, STAT_TOPDOWN_TOTAL_SLOTS,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_SLOTS_ISSUED))
 		update_runtime_stat(st, STAT_TOPDOWN_SLOTS_ISSUED,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_SLOTS_RETIRED))
 		update_runtime_stat(st, STAT_TOPDOWN_SLOTS_RETIRED,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_FETCH_BUBBLES))
 		update_runtime_stat(st, STAT_TOPDOWN_FETCH_BUBBLES,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_RECOVERY_BUBBLES))
 		update_runtime_stat(st, STAT_TOPDOWN_RECOVERY_BUBBLES,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_RETIRING))
 		update_runtime_stat(st, STAT_TOPDOWN_RETIRING,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_BAD_SPEC))
 		update_runtime_stat(st, STAT_TOPDOWN_BAD_SPEC,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_FE_BOUND))
 		update_runtime_stat(st, STAT_TOPDOWN_FE_BOUND,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, TOPDOWN_BE_BOUND))
 		update_runtime_stat(st, STAT_TOPDOWN_BE_BOUND,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND))
 		update_runtime_stat(st, STAT_STALLED_CYCLES_FRONT,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND))
 		update_runtime_stat(st, STAT_STALLED_CYCLES_BACK,
-				    ctx, cpu, count);
+				    cpu, count, &rsd);
 	else if (evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
-		update_runtime_stat(st, STAT_BRANCHES, ctx, cpu, count);
+		update_runtime_stat(st, STAT_BRANCHES, cpu, count, &rsd);
 	else if (evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES))
-		update_runtime_stat(st, STAT_CACHEREFS, ctx, cpu, count);
+		update_runtime_stat(st, STAT_CACHEREFS, cpu, count, &rsd);
 	else if (evsel__match(counter, HW_CACHE, HW_CACHE_L1D))
-		update_runtime_stat(st, STAT_L1_DCACHE, ctx, cpu, count);
+		update_runtime_stat(st, STAT_L1_DCACHE, cpu, count, &rsd);
 	else if (evsel__match(counter, HW_CACHE, HW_CACHE_L1I))
-		update_runtime_stat(st, STAT_L1_ICACHE, ctx, cpu, count);
+		update_runtime_stat(st, STAT_L1_ICACHE, cpu, count, &rsd);
 	else if (evsel__match(counter, HW_CACHE, HW_CACHE_LL))
-		update_runtime_stat(st, STAT_LL_CACHE, ctx, cpu, count);
+		update_runtime_stat(st, STAT_LL_CACHE, cpu, count, &rsd);
 	else if (evsel__match(counter, HW_CACHE, HW_CACHE_DTLB))
-		update_runtime_stat(st, STAT_DTLB_CACHE, ctx, cpu, count);
+		update_runtime_stat(st, STAT_DTLB_CACHE, cpu, count, &rsd);
 	else if (evsel__match(counter, HW_CACHE, HW_CACHE_ITLB))
-		update_runtime_stat(st, STAT_ITLB_CACHE, ctx, cpu, count);
+		update_runtime_stat(st, STAT_ITLB_CACHE, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, SMI_NUM))
-		update_runtime_stat(st, STAT_SMI_NUM, ctx, cpu, count);
+		update_runtime_stat(st, STAT_SMI_NUM, cpu, count, &rsd);
 	else if (perf_stat_evsel__is(counter, APERF))
-		update_runtime_stat(st, STAT_APERF, ctx, cpu, count);
+		update_runtime_stat(st, STAT_APERF, cpu, count, &rsd);
 
 	if (counter->collect_stat) {
-		v = saved_value_lookup(counter, cpu, true, STAT_NONE, 0, st);
+		v = saved_value_lookup(counter, cpu, true, STAT_NONE, 0, st,
+				       rsd.cgrp);
 		update_stats(&v->stats, count);
 		if (counter->metric_leader)
 			v->metric_total += count;
 	} else if (counter->metric_leader) {
 		v = saved_value_lookup(counter->metric_leader,
-				       cpu, true, STAT_NONE, 0, st);
+				       cpu, true, STAT_NONE, 0, st, rsd.cgrp);
 		v->metric_total += count;
 		v->metric_other++;
 	}
@@ -422,11 +443,12 @@ void perf_stat__collect_metric_expr(struct evlist *evsel_list)
 }
 
 static double runtime_stat_avg(struct runtime_stat *st,
-			       enum stat_type type, int ctx, int cpu)
+			       enum stat_type type, int cpu,
+			       struct runtime_stat_data *rsd)
 {
 	struct saved_value *v;
 
-	v = saved_value_lookup(NULL, cpu, false, type, ctx, st);
+	v = saved_value_lookup(NULL, cpu, false, type, rsd->ctx, st, rsd->cgrp);
 	if (!v)
 		return 0.0;
 
@@ -434,11 +456,12 @@ static double runtime_stat_avg(struct runtime_stat *st,
 }
 
 static double runtime_stat_n(struct runtime_stat *st,
-			     enum stat_type type, int ctx, int cpu)
+			     enum stat_type type, int cpu,
+			     struct runtime_stat_data *rsd)
 {
 	struct saved_value *v;
 
-	v = saved_value_lookup(NULL, cpu, false, type, ctx, st);
+	v = saved_value_lookup(NULL, cpu, false, type, rsd->ctx, st, rsd->cgrp);
 	if (!v)
 		return 0.0;
 
@@ -446,16 +469,15 @@ static double runtime_stat_n(struct runtime_stat *st,
 }
 
 static void print_stalled_cycles_frontend(struct perf_stat_config *config,
-					  int cpu,
-					  struct evsel *evsel, double avg,
+					  int cpu, double avg,
 					  struct perf_stat_output_ctx *out,
-					  struct runtime_stat *st)
+					  struct runtime_stat *st,
+					  struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_CYCLES, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -470,16 +492,15 @@ static void print_stalled_cycles_frontend(struct perf_stat_config *config,
 }
 
 static void print_stalled_cycles_backend(struct perf_stat_config *config,
-					 int cpu,
-					 struct evsel *evsel, double avg,
+					 int cpu, double avg,
 					 struct perf_stat_output_ctx *out,
-					 struct runtime_stat *st)
+					 struct runtime_stat *st,
+					 struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_CYCLES, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -490,17 +511,15 @@ static void print_stalled_cycles_backend(struct perf_stat_config *config,
 }
 
 static void print_branch_misses(struct perf_stat_config *config,
-				int cpu,
-				struct evsel *evsel,
-				double avg,
+				int cpu, double avg,
 				struct perf_stat_output_ctx *out,
-				struct runtime_stat *st)
+				struct runtime_stat *st,
+				struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_BRANCHES, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_BRANCHES, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -511,18 +530,15 @@ static void print_branch_misses(struct perf_stat_config *config,
 }
 
 static void print_l1_dcache_misses(struct perf_stat_config *config,
-				   int cpu,
-				   struct evsel *evsel,
-				   double avg,
+				   int cpu, double avg,
 				   struct perf_stat_output_ctx *out,
-				   struct runtime_stat *st)
-
+				   struct runtime_stat *st,
+				   struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_L1_DCACHE, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_L1_DCACHE, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -533,18 +549,15 @@ static void print_l1_dcache_misses(struct perf_stat_config *config,
 }
 
 static void print_l1_icache_misses(struct perf_stat_config *config,
-				   int cpu,
-				   struct evsel *evsel,
-				   double avg,
+				   int cpu, double avg,
 				   struct perf_stat_output_ctx *out,
-				   struct runtime_stat *st)
-
+				   struct runtime_stat *st,
+				   struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_L1_ICACHE, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_L1_ICACHE, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -554,17 +567,15 @@ static void print_l1_icache_misses(struct perf_stat_config *config,
 }
 
 static void print_dtlb_cache_misses(struct perf_stat_config *config,
-				    int cpu,
-				    struct evsel *evsel,
-				    double avg,
+				    int cpu, double avg,
 				    struct perf_stat_output_ctx *out,
-				    struct runtime_stat *st)
+				    struct runtime_stat *st,
+				    struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_DTLB_CACHE, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_DTLB_CACHE, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -574,17 +585,15 @@ static void print_dtlb_cache_misses(struct perf_stat_config *config,
 }
 
 static void print_itlb_cache_misses(struct perf_stat_config *config,
-				    int cpu,
-				    struct evsel *evsel,
-				    double avg,
+				    int cpu, double avg,
 				    struct perf_stat_output_ctx *out,
-				    struct runtime_stat *st)
+				    struct runtime_stat *st,
+				    struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_ITLB_CACHE, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_ITLB_CACHE, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -594,17 +603,15 @@ static void print_itlb_cache_misses(struct perf_stat_config *config,
 }
 
 static void print_ll_cache_misses(struct perf_stat_config *config,
-				  int cpu,
-				  struct evsel *evsel,
-				  double avg,
+				  int cpu, double avg,
 				  struct perf_stat_output_ctx *out,
-				  struct runtime_stat *st)
+				  struct runtime_stat *st,
+				  struct runtime_stat_data *rsd)
 {
 	double total, ratio = 0.0;
 	const char *color;
-	int ctx = evsel_context(evsel);
 
-	total = runtime_stat_avg(st, STAT_LL_CACHE, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_LL_CACHE, cpu, rsd);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -662,56 +669,61 @@ static double sanitize_val(double x)
 	return x;
 }
 
-static double td_total_slots(int ctx, int cpu, struct runtime_stat *st)
+static double td_total_slots(int cpu, struct runtime_stat *st,
+			     struct runtime_stat_data *rsd)
 {
-	return runtime_stat_avg(st, STAT_TOPDOWN_TOTAL_SLOTS, ctx, cpu);
+	return runtime_stat_avg(st, STAT_TOPDOWN_TOTAL_SLOTS, cpu, rsd);
 }
 
-static double td_bad_spec(int ctx, int cpu, struct runtime_stat *st)
+static double td_bad_spec(int cpu, struct runtime_stat *st,
+			  struct runtime_stat_data *rsd)
 {
 	double bad_spec = 0;
 	double total_slots;
 	double total;
 
-	total = runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_ISSUED, ctx, cpu) -
-		runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_RETIRED, ctx, cpu) +
-		runtime_stat_avg(st, STAT_TOPDOWN_RECOVERY_BUBBLES, ctx, cpu);
+	total = runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_ISSUED, cpu, rsd) -
+		runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_RETIRED, cpu, rsd) +
+		runtime_stat_avg(st, STAT_TOPDOWN_RECOVERY_BUBBLES, cpu, rsd);
 
-	total_slots = td_total_slots(ctx, cpu, st);
+	total_slots = td_total_slots(cpu, st, rsd);
 	if (total_slots)
 		bad_spec = total / total_slots;
 	return sanitize_val(bad_spec);
 }
 
-static double td_retiring(int ctx, int cpu, struct runtime_stat *st)
+static double td_retiring(int cpu, struct runtime_stat *st,
+			  struct runtime_stat_data *rsd)
 {
 	double retiring = 0;
-	double total_slots = td_total_slots(ctx, cpu, st);
+	double total_slots = td_total_slots(cpu, st, rsd);
 	double ret_slots = runtime_stat_avg(st, STAT_TOPDOWN_SLOTS_RETIRED,
-					    ctx, cpu);
+					    cpu, rsd);
 
 	if (total_slots)
 		retiring = ret_slots / total_slots;
 	return retiring;
 }
 
-static double td_fe_bound(int ctx, int cpu, struct runtime_stat *st)
+static double td_fe_bound(int cpu, struct runtime_stat *st,
+			  struct runtime_stat_data *rsd)
 {
 	double fe_bound = 0;
-	double total_slots = td_total_slots(ctx, cpu, st);
+	double total_slots = td_total_slots(cpu, st, rsd);
 	double fetch_bub = runtime_stat_avg(st, STAT_TOPDOWN_FETCH_BUBBLES,
-					    ctx, cpu);
+					    cpu, rsd);
 
 	if (total_slots)
 		fe_bound = fetch_bub / total_slots;
 	return fe_bound;
 }
 
-static double td_be_bound(int ctx, int cpu, struct runtime_stat *st)
+static double td_be_bound(int cpu, struct runtime_stat *st,
+			  struct runtime_stat_data *rsd)
 {
-	double sum = (td_fe_bound(ctx, cpu, st) +
-		      td_bad_spec(ctx, cpu, st) +
-		      td_retiring(ctx, cpu, st));
+	double sum = (td_fe_bound(cpu, st, rsd) +
+		      td_bad_spec(cpu, st, rsd) +
+		      td_retiring(cpu, st, rsd));
 	if (sum == 0)
 		return 0;
 	return sanitize_val(1.0 - sum);
@@ -722,15 +734,15 @@ static double td_be_bound(int ctx, int cpu, struct runtime_stat *st)
  * the ratios we need to recreate the sum.
  */
 
-static double td_metric_ratio(int ctx, int cpu,
-			      enum stat_type type,
-			      struct runtime_stat *stat)
+static double td_metric_ratio(int cpu, enum stat_type type,
+			      struct runtime_stat *stat,
+			      struct runtime_stat_data *rsd)
 {
-	double sum = runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, ctx, cpu) +
-		runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, ctx, cpu) +
-		runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, ctx, cpu) +
-		runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, ctx, cpu);
-	double d = runtime_stat_avg(stat, type, ctx, cpu);
+	double sum = runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, cpu, rsd) +
+		runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, cpu, rsd) +
+		runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, cpu, rsd) +
+		runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, cpu, rsd);
+	double d = runtime_stat_avg(stat, type, cpu, rsd);
 
 	if (sum)
 		return d / sum;
@@ -742,34 +754,33 @@ static double td_metric_ratio(int ctx, int cpu,
  * We allow two missing.
  */
 
-static bool full_td(int ctx, int cpu,
-		    struct runtime_stat *stat)
+static bool full_td(int cpu, struct runtime_stat *stat,
+		    struct runtime_stat_data *rsd)
 {
 	int c = 0;
 
-	if (runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, ctx, cpu) > 0)
+	if (runtime_stat_avg(stat, STAT_TOPDOWN_RETIRING, cpu, rsd) > 0)
 		c++;
-	if (runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, ctx, cpu) > 0)
+	if (runtime_stat_avg(stat, STAT_TOPDOWN_BE_BOUND, cpu, rsd) > 0)
 		c++;
-	if (runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, ctx, cpu) > 0)
+	if (runtime_stat_avg(stat, STAT_TOPDOWN_FE_BOUND, cpu, rsd) > 0)
 		c++;
-	if (runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, ctx, cpu) > 0)
+	if (runtime_stat_avg(stat, STAT_TOPDOWN_BAD_SPEC, cpu, rsd) > 0)
 		c++;
 	return c >= 2;
 }
 
-static void print_smi_cost(struct perf_stat_config *config,
-			   int cpu, struct evsel *evsel,
+static void print_smi_cost(struct perf_stat_config *config, int cpu,
 			   struct perf_stat_output_ctx *out,
-			   struct runtime_stat *st)
+			   struct runtime_stat *st,
+			   struct runtime_stat_data *rsd)
 {
 	double smi_num, aperf, cycles, cost = 0.0;
-	int ctx = evsel_context(evsel);
 	const char *color = NULL;
 
-	smi_num = runtime_stat_avg(st, STAT_SMI_NUM, ctx, cpu);
-	aperf = runtime_stat_avg(st, STAT_APERF, ctx, cpu);
-	cycles = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
+	smi_num = runtime_stat_avg(st, STAT_SMI_NUM, cpu, rsd);
+	aperf = runtime_stat_avg(st, STAT_APERF, cpu, rsd);
+	cycles = runtime_stat_avg(st, STAT_CYCLES, cpu, rsd);
 
 	if ((cycles == 0) || (aperf == 0))
 		return;
@@ -804,7 +815,8 @@ static int prepare_metric(struct evsel **metric_events,
 			scale = 1e-9;
 		} else {
 			v = saved_value_lookup(metric_events[i], cpu, false,
-					       STAT_NONE, 0, st);
+					       STAT_NONE, 0, st,
+					       metric_events[i]->cgrp);
 			if (!v)
 				break;
 			stats = &v->stats;
@@ -930,12 +942,15 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 	print_metric_t print_metric = out->print_metric;
 	double total, ratio = 0.0, total2;
 	const char *color = NULL;
-	int ctx = evsel_context(evsel);
+	struct runtime_stat_data rsd = {
+		.ctx = evsel_context(evsel),
+		.cgrp = evsel->cgrp,
+	};
 	struct metric_event *me;
 	int num = 1;
 
 	if (evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) {
-		total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CYCLES, cpu, &rsd);
 
 		if (total) {
 			ratio = avg / total;
@@ -945,12 +960,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 			print_metric(config, ctxp, NULL, NULL, "insn per cycle", 0);
 		}
 
-		total = runtime_stat_avg(st, STAT_STALLED_CYCLES_FRONT,
-					 ctx, cpu);
+		total = runtime_stat_avg(st, STAT_STALLED_CYCLES_FRONT, cpu, &rsd);
 
 		total = max(total, runtime_stat_avg(st,
 						    STAT_STALLED_CYCLES_BACK,
-						    ctx, cpu));
+						    cpu, &rsd));
 
 		if (total && avg) {
 			out->new_line(config, ctxp);
@@ -960,8 +974,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					ratio);
 		}
 	} else if (evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES)) {
-		if (runtime_stat_n(st, STAT_BRANCHES, ctx, cpu) != 0)
-			print_branch_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_BRANCHES, cpu, &rsd) != 0)
+			print_branch_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all branches", 0);
 	} else if (
@@ -970,8 +984,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					 ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) {
 
-		if (runtime_stat_n(st, STAT_L1_DCACHE, ctx, cpu) != 0)
-			print_l1_dcache_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_L1_DCACHE, cpu, &rsd) != 0)
+			print_l1_dcache_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all L1-dcache accesses", 0);
 	} else if (
@@ -980,8 +994,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					 ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) {
 
-		if (runtime_stat_n(st, STAT_L1_ICACHE, ctx, cpu) != 0)
-			print_l1_icache_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_L1_ICACHE, cpu, &rsd) != 0)
+			print_l1_icache_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all L1-icache accesses", 0);
 	} else if (
@@ -990,8 +1004,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					 ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) {
 
-		if (runtime_stat_n(st, STAT_DTLB_CACHE, ctx, cpu) != 0)
-			print_dtlb_cache_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_DTLB_CACHE, cpu, &rsd) != 0)
+			print_dtlb_cache_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all dTLB cache accesses", 0);
 	} else if (
@@ -1000,8 +1014,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					 ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) {
 
-		if (runtime_stat_n(st, STAT_ITLB_CACHE, ctx, cpu) != 0)
-			print_itlb_cache_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_ITLB_CACHE, cpu, &rsd) != 0)
+			print_itlb_cache_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all iTLB cache accesses", 0);
 	} else if (
@@ -1010,27 +1024,27 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					 ((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16))) {
 
-		if (runtime_stat_n(st, STAT_LL_CACHE, ctx, cpu) != 0)
-			print_ll_cache_misses(config, cpu, evsel, avg, out, st);
+		if (runtime_stat_n(st, STAT_LL_CACHE, cpu, &rsd) != 0)
+			print_ll_cache_misses(config, cpu, avg, out, st, &rsd);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all LL-cache accesses", 0);
 	} else if (evsel__match(evsel, HARDWARE, HW_CACHE_MISSES)) {
-		total = runtime_stat_avg(st, STAT_CACHEREFS, ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CACHEREFS, cpu, &rsd);
 
 		if (total)
 			ratio = avg * 100 / total;
 
-		if (runtime_stat_n(st, STAT_CACHEREFS, ctx, cpu) != 0)
+		if (runtime_stat_n(st, STAT_CACHEREFS, cpu, &rsd) != 0)
 			print_metric(config, ctxp, NULL, "%8.3f %%",
 				     "of all cache refs", ratio);
 		else
 			print_metric(config, ctxp, NULL, NULL, "of all cache refs", 0);
 	} else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_FRONTEND)) {
-		print_stalled_cycles_frontend(config, cpu, evsel, avg, out, st);
+		print_stalled_cycles_frontend(config, cpu, avg, out, st, &rsd);
 	} else if (evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_BACKEND)) {
-		print_stalled_cycles_backend(config, cpu, evsel, avg, out, st);
+		print_stalled_cycles_backend(config, cpu, avg, out, st, &rsd);
 	} else if (evsel__match(evsel, HARDWARE, HW_CPU_CYCLES)) {
-		total = runtime_stat_avg(st, STAT_NSECS, 0, cpu);
+		total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
 
 		if (total) {
 			ratio = avg / total;
@@ -1039,7 +1053,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 			print_metric(config, ctxp, NULL, NULL, "Ghz", 0);
 		}
 	} else if (perf_stat_evsel__is(evsel, CYCLES_IN_TX)) {
-		total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CYCLES, cpu, &rsd);
 
 		if (total)
 			print_metric(config, ctxp, NULL,
@@ -1049,8 +1063,8 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 			print_metric(config, ctxp, NULL, NULL, "transactional cycles",
 				     0);
 	} else if (perf_stat_evsel__is(evsel, CYCLES_IN_TX_CP)) {
-		total = runtime_stat_avg(st, STAT_CYCLES, ctx, cpu);
-		total2 = runtime_stat_avg(st, STAT_CYCLES_IN_TX, ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CYCLES, cpu, &rsd);
+		total2 = runtime_stat_avg(st, STAT_CYCLES_IN_TX, cpu, &rsd);
 
 		if (total2 < avg)
 			total2 = avg;
@@ -1060,21 +1074,19 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 		else
 			print_metric(config, ctxp, NULL, NULL, "aborted cycles", 0);
 	} else if (perf_stat_evsel__is(evsel, TRANSACTION_START)) {
-		total = runtime_stat_avg(st, STAT_CYCLES_IN_TX,
-					 ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CYCLES_IN_TX, cpu, &rsd);
 
 		if (avg)
 			ratio = total / avg;
 
-		if (runtime_stat_n(st, STAT_CYCLES_IN_TX, ctx, cpu) != 0)
+		if (runtime_stat_n(st, STAT_CYCLES_IN_TX, cpu, &rsd) != 0)
 			print_metric(config, ctxp, NULL, "%8.0f",
 				     "cycles / transaction", ratio);
 		else
 			print_metric(config, ctxp, NULL, NULL, "cycles / transaction",
 				      0);
 	} else if (perf_stat_evsel__is(evsel, ELISION_START)) {
-		total = runtime_stat_avg(st, STAT_CYCLES_IN_TX,
-					 ctx, cpu);
+		total = runtime_stat_avg(st, STAT_CYCLES_IN_TX, cpu, &rsd);
 
 		if (avg)
 			ratio = total / avg;
@@ -1087,28 +1099,28 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 		else
 			print_metric(config, ctxp, NULL, NULL, "CPUs utilized", 0);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_FETCH_BUBBLES)) {
-		double fe_bound = td_fe_bound(ctx, cpu, st);
+		double fe_bound = td_fe_bound(cpu, st, &rsd);
 
 		if (fe_bound > 0.2)
 			color = PERF_COLOR_RED;
 		print_metric(config, ctxp, color, "%8.1f%%", "frontend bound",
 				fe_bound * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_SLOTS_RETIRED)) {
-		double retiring = td_retiring(ctx, cpu, st);
+		double retiring = td_retiring(cpu, st, &rsd);
 
 		if (retiring > 0.7)
 			color = PERF_COLOR_GREEN;
 		print_metric(config, ctxp, color, "%8.1f%%", "retiring",
 				retiring * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_RECOVERY_BUBBLES)) {
-		double bad_spec = td_bad_spec(ctx, cpu, st);
+		double bad_spec = td_bad_spec(cpu, st, &rsd);
 
 		if (bad_spec > 0.1)
 			color = PERF_COLOR_RED;
 		print_metric(config, ctxp, color, "%8.1f%%", "bad speculation",
 				bad_spec * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_SLOTS_ISSUED)) {
-		double be_bound = td_be_bound(ctx, cpu, st);
+		double be_bound = td_be_bound(cpu, st, &rsd);
 		const char *name = "backend bound";
 		static int have_recovery_bubbles = -1;
 
@@ -1121,43 +1133,43 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 
 		if (be_bound > 0.2)
 			color = PERF_COLOR_RED;
-		if (td_total_slots(ctx, cpu, st) > 0)
+		if (td_total_slots(cpu, st, &rsd) > 0)
 			print_metric(config, ctxp, color, "%8.1f%%", name,
 					be_bound * 100.);
 		else
 			print_metric(config, ctxp, NULL, NULL, name, 0);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_RETIRING) &&
-			full_td(ctx, cpu, st)) {
-		double retiring = td_metric_ratio(ctx, cpu,
-						  STAT_TOPDOWN_RETIRING, st);
-
+		   full_td(cpu, st, &rsd)) {
+		double retiring = td_metric_ratio(cpu,
+						  STAT_TOPDOWN_RETIRING, st,
+						  &rsd);
 		if (retiring > 0.7)
 			color = PERF_COLOR_GREEN;
 		print_metric(config, ctxp, color, "%8.1f%%", "retiring",
 				retiring * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_FE_BOUND) &&
-			full_td(ctx, cpu, st)) {
-		double fe_bound = td_metric_ratio(ctx, cpu,
-						  STAT_TOPDOWN_FE_BOUND, st);
-
+		   full_td(cpu, st, &rsd)) {
+		double fe_bound = td_metric_ratio(cpu,
+						  STAT_TOPDOWN_FE_BOUND, st,
+						  &rsd);
 		if (fe_bound > 0.2)
 			color = PERF_COLOR_RED;
 		print_metric(config, ctxp, color, "%8.1f%%", "frontend bound",
 				fe_bound * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_BE_BOUND) &&
-			full_td(ctx, cpu, st)) {
-		double be_bound = td_metric_ratio(ctx, cpu,
-						  STAT_TOPDOWN_BE_BOUND, st);
-
+		   full_td(cpu, st, &rsd)) {
+		double be_bound = td_metric_ratio(cpu,
+						  STAT_TOPDOWN_BE_BOUND, st,
+						  &rsd);
 		if (be_bound > 0.2)
 			color = PERF_COLOR_RED;
 		print_metric(config, ctxp, color, "%8.1f%%", "backend bound",
 				be_bound * 100.);
 	} else if (perf_stat_evsel__is(evsel, TOPDOWN_BAD_SPEC) &&
-			full_td(ctx, cpu, st)) {
-		double bad_spec = td_metric_ratio(ctx, cpu,
-						  STAT_TOPDOWN_BAD_SPEC, st);
-
+		   full_td(cpu, st, &rsd)) {
+		double bad_spec = td_metric_ratio(cpu,
+						  STAT_TOPDOWN_BAD_SPEC, st,
+						  &rsd);
 		if (bad_spec > 0.1)
 			color = PERF_COLOR_RED;
 		print_metric(config, ctxp, color, "%8.1f%%", "bad speculation",
@@ -1165,11 +1177,11 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 	} else if (evsel->metric_expr) {
 		generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
 				evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
-	} else if (runtime_stat_n(st, STAT_NSECS, 0, cpu) != 0) {
+	} else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
 		char unit = 'M';
 		char unit_buf[10];
 
-		total = runtime_stat_avg(st, STAT_NSECS, 0, cpu);
+		total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
 
 		if (total)
 			ratio = 1000.0 * avg / total;
@@ -1180,7 +1192,7 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
 		snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
 		print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
 	} else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
-		print_smi_cost(config, cpu, evsel, out, st);
+		print_smi_cost(config, cpu, out, st, &rsd);
 	} else {
 		num = 0;
 	}
diff --git a/tools/testing/kunit/kunit_kernel.py b/tools/testing/kunit/kunit_kernel.py
index 57c1724b7e5d..698358c9c0d6 100644
--- a/tools/testing/kunit/kunit_kernel.py
+++ b/tools/testing/kunit/kunit_kernel.py
@@ -198,7 +198,7 @@ class LinuxSourceTree(object):
 		return self.validate_config(build_dir)
 
 	def run_kernel(self, args=[], build_dir='', timeout=None):
-		args.extend(['mem=1G'])
+		args.extend(['mem=1G', 'console=tty'])
 		self._ops.linux_bin(args, timeout, build_dir)
 		outfile = get_outfile_path(build_dir)
 		subprocess.call(['stty', 'sane'])
diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile
index afbab4aeef3c..8a917cb4426a 100644
--- a/tools/testing/selftests/Makefile
+++ b/tools/testing/selftests/Makefile
@@ -77,8 +77,10 @@ TARGETS += zram
 TARGETS_HOTPLUG = cpu-hotplug
 TARGETS_HOTPLUG += memory-hotplug
 
-# User can optionally provide a TARGETS skiplist.
-SKIP_TARGETS ?=
+# User can optionally provide a TARGETS skiplist.  By default we skip
+# BPF since it has cutting edge build time dependencies which require
+# more effort to install.
+SKIP_TARGETS ?= bpf
 ifneq ($(SKIP_TARGETS),)
 	TMP := $(filter-out $(SKIP_TARGETS), $(TARGETS))
 	override TARGETS := $(TMP)
diff --git a/tools/testing/selftests/arm64/fp/fpsimd-test.S b/tools/testing/selftests/arm64/fp/fpsimd-test.S
index 1c5556bdd11d..0dbd594c2747 100644
--- a/tools/testing/selftests/arm64/fp/fpsimd-test.S
+++ b/tools/testing/selftests/arm64/fp/fpsimd-test.S
@@ -457,7 +457,7 @@ function barf
 	mov	x11, x1	// actual data
 	mov	x12, x2	// data size
 
-	puts	"Mistatch: PID="
+	puts	"Mismatch: PID="
 	mov	x0, x20
 	bl	putdec
 	puts	", iteration="
diff --git a/tools/testing/selftests/arm64/fp/sve-test.S b/tools/testing/selftests/arm64/fp/sve-test.S
index f95074c9b48b..9210691aa998 100644
--- a/tools/testing/selftests/arm64/fp/sve-test.S
+++ b/tools/testing/selftests/arm64/fp/sve-test.S
@@ -625,7 +625,7 @@ function barf
 	mov	x11, x1	// actual data
 	mov	x12, x2	// data size
 
-	puts	"Mistatch: PID="
+	puts	"Mismatch: PID="
 	mov	x0, x20
 	bl	putdec
 	puts	", iteration="
diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index 8c33e999319a..c51df6b91bef 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -121,6 +121,9 @@ VMLINUX_BTF_PATHS ?= $(if $(O),$(O)/vmlinux)				\
 		     /sys/kernel/btf/vmlinux				\
 		     /boot/vmlinux-$(shell uname -r)
 VMLINUX_BTF ?= $(abspath $(firstword $(wildcard $(VMLINUX_BTF_PATHS))))
+ifeq ($(VMLINUX_BTF),)
+$(error Cannot find a vmlinux for VMLINUX_BTF at any of "$(VMLINUX_BTF_PATHS)")
+endif
 
 # Define simple and short `make test_progs`, `make test_sysctl`, etc targets
 # to build individual tests.
diff --git a/tools/testing/selftests/bpf/progs/bprm_opts.c b/tools/testing/selftests/bpf/progs/bprm_opts.c
index 5bfef2887e70..418d9c6d4952 100644
--- a/tools/testing/selftests/bpf/progs/bprm_opts.c
+++ b/tools/testing/selftests/bpf/progs/bprm_opts.c
@@ -4,7 +4,7 @@
  * Copyright 2020 Google LLC.
  */
 
-#include "vmlinux.h"
+#include <linux/bpf.h>
 #include <errno.h>
 #include <bpf/bpf_helpers.h>
 #include <bpf/bpf_tracing.h>
diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c
index 0ad3e6305ff0..51adc42b2b40 100644
--- a/tools/testing/selftests/bpf/test_maps.c
+++ b/tools/testing/selftests/bpf/test_maps.c
@@ -1312,22 +1312,58 @@ static void test_map_stress(void)
 #define DO_UPDATE 1
 #define DO_DELETE 0
 
+#define MAP_RETRIES 20
+
+static int map_update_retriable(int map_fd, const void *key, const void *value,
+				int flags, int attempts)
+{
+	while (bpf_map_update_elem(map_fd, key, value, flags)) {
+		if (!attempts || (errno != EAGAIN && errno != EBUSY))
+			return -errno;
+
+		usleep(1);
+		attempts--;
+	}
+
+	return 0;
+}
+
+static int map_delete_retriable(int map_fd, const void *key, int attempts)
+{
+	while (bpf_map_delete_elem(map_fd, key)) {
+		if (!attempts || (errno != EAGAIN && errno != EBUSY))
+			return -errno;
+
+		usleep(1);
+		attempts--;
+	}
+
+	return 0;
+}
+
 static void test_update_delete(unsigned int fn, void *data)
 {
 	int do_update = ((int *)data)[1];
 	int fd = ((int *)data)[0];
-	int i, key, value;
+	int i, key, value, err;
 
 	for (i = fn; i < MAP_SIZE; i += TASKS) {
 		key = value = i;
 
 		if (do_update) {
-			assert(bpf_map_update_elem(fd, &key, &value,
-						   BPF_NOEXIST) == 0);
-			assert(bpf_map_update_elem(fd, &key, &value,
-						   BPF_EXIST) == 0);
+			err = map_update_retriable(fd, &key, &value, BPF_NOEXIST, MAP_RETRIES);
+			if (err)
+				printf("error %d %d\n", err, errno);
+			assert(err == 0);
+			err = map_update_retriable(fd, &key, &value, BPF_EXIST, MAP_RETRIES);
+			if (err)
+				printf("error %d %d\n", err, errno);
+			assert(err == 0);
 		} else {
-			assert(bpf_map_delete_elem(fd, &key) == 0);
+			err = map_delete_retriable(fd, &key, MAP_RETRIES);
+			if (err)
+				printf("error %d %d\n", err, errno);
+			assert(err == 0);
 		}
 	}
 }
diff --git a/tools/testing/selftests/bpf/xdpxceiver.c b/tools/testing/selftests/bpf/xdpxceiver.c
index 014dedaa4dd2..1e722ee76b1f 100644
--- a/tools/testing/selftests/bpf/xdpxceiver.c
+++ b/tools/testing/selftests/bpf/xdpxceiver.c
@@ -715,7 +715,7 @@ static void worker_pkt_dump(void)
 		int payload = *((uint32_t *)(pkt_buf[iter]->payload + PKT_HDR_SIZE));
 
 		if (payload == EOT) {
-			ksft_print_msg("End-of-tranmission frame received\n");
+			ksft_print_msg("End-of-transmission frame received\n");
 			fprintf(stdout, "---------------------------------------\n");
 			break;
 		}
@@ -747,7 +747,7 @@ static void worker_pkt_validate(void)
 			}
 
 			if (payloadseqnum == EOT) {
-				ksft_print_msg("End-of-tranmission frame received: PASS\n");
+				ksft_print_msg("End-of-transmission frame received: PASS\n");
 				sigvar = 1;
 				break;
 			}
diff --git a/tools/testing/selftests/drivers/net/mlxsw/qos_pfc.sh b/tools/testing/selftests/drivers/net/mlxsw/qos_pfc.sh
index 4d900bc1f76c..5c7700212f75 100755
--- a/tools/testing/selftests/drivers/net/mlxsw/qos_pfc.sh
+++ b/tools/testing/selftests/drivers/net/mlxsw/qos_pfc.sh
@@ -230,7 +230,7 @@ switch_create()
 	__mlnx_qos -i $swp4 --pfc=0,1,0,0,0,0,0,0 >/dev/null
 	# PG0 will get autoconfigured to Xoff, give PG1 arbitrarily 100K, which
 	# is (-2*MTU) about 80K of delay provision.
-	__mlnx_qos -i $swp3 --buffer_size=0,$_100KB,0,0,0,0,0,0 >/dev/null
+	__mlnx_qos -i $swp4 --buffer_size=0,$_100KB,0,0,0,0,0,0 >/dev/null
 
 	# bridges
 	# -------
diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index c7ca4faba272..fe41c6a0fa67 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -33,7 +33,7 @@ ifeq ($(ARCH),s390)
 	UNAME_M := s390x
 endif
 
-LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c lib/test_util.c
+LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c lib/test_util.c lib/guest_modes.c lib/perf_test_util.c
 LIBKVM_x86_64 = lib/x86_64/processor.c lib/x86_64/vmx.c lib/x86_64/svm.c lib/x86_64/ucall.c lib/x86_64/handlers.S
 LIBKVM_aarch64 = lib/aarch64/processor.c lib/aarch64/ucall.c
 LIBKVM_s390x = lib/s390x/processor.c lib/s390x/ucall.c lib/s390x/diag318_test_handler.c
diff --git a/tools/testing/selftests/kvm/demand_paging_test.c b/tools/testing/selftests/kvm/demand_paging_test.c
index 3d96a7bfaff3..cdad1eca72f7 100644
--- a/tools/testing/selftests/kvm/demand_paging_test.c
+++ b/tools/testing/selftests/kvm/demand_paging_test.c
@@ -7,23 +7,20 @@
  * Copyright (C) 2019, Google, Inc.
  */
 
-#define _GNU_SOURCE /* for program_invocation_name */
+#define _GNU_SOURCE /* for pipe2 */
 
 #include <stdio.h>
 #include <stdlib.h>
-#include <sys/syscall.h>
-#include <unistd.h>
-#include <asm/unistd.h>
 #include <time.h>
 #include <poll.h>
 #include <pthread.h>
-#include <linux/bitmap.h>
-#include <linux/bitops.h>
 #include <linux/userfaultfd.h>
+#include <sys/syscall.h>
 
-#include "perf_test_util.h"
-#include "processor.h"
+#include "kvm_util.h"
 #include "test_util.h"
+#include "perf_test_util.h"
+#include "guest_modes.h"
 
 #ifdef __NR_userfaultfd
 
@@ -39,12 +36,14 @@
 #define PER_VCPU_DEBUG(...) _no_printf(__VA_ARGS__)
 #endif
 
+static int nr_vcpus = 1;
+static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
 static char *guest_data_prototype;
 
 static void *vcpu_worker(void *data)
 {
 	int ret;
-	struct vcpu_args *vcpu_args = (struct vcpu_args *)data;
+	struct perf_test_vcpu_args *vcpu_args = (struct perf_test_vcpu_args *)data;
 	int vcpu_id = vcpu_args->vcpu_id;
 	struct kvm_vm *vm = perf_test_args.vm;
 	struct kvm_run *run;
@@ -248,9 +247,14 @@ static int setup_demand_paging(struct kvm_vm *vm,
 	return 0;
 }
 
-static void run_test(enum vm_guest_mode mode, bool use_uffd,
-		     useconds_t uffd_delay)
+struct test_params {
+	bool use_uffd;
+	useconds_t uffd_delay;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	pthread_t *uffd_handler_threads = NULL;
 	struct uffd_handler_args *uffd_args = NULL;
@@ -261,7 +265,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 	int vcpu_id;
 	int r;
 
-	vm = create_vm(mode, nr_vcpus, guest_percpu_mem_size);
+	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size);
 
 	perf_test_args.wr_fract = 1;
 
@@ -273,9 +277,9 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 	vcpu_threads = malloc(nr_vcpus * sizeof(*vcpu_threads));
 	TEST_ASSERT(vcpu_threads, "Memory allocation failed");
 
-	add_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
+	perf_test_setup_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
 
-	if (use_uffd) {
+	if (p->use_uffd) {
 		uffd_handler_threads =
 			malloc(nr_vcpus * sizeof(*uffd_handler_threads));
 		TEST_ASSERT(uffd_handler_threads, "Memory allocation failed");
@@ -308,7 +312,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 			r = setup_demand_paging(vm,
 						&uffd_handler_threads[vcpu_id],
 						pipefds[vcpu_id * 2],
-						uffd_delay, &uffd_args[vcpu_id],
+						p->uffd_delay, &uffd_args[vcpu_id],
 						vcpu_hva, guest_percpu_mem_size);
 			if (r < 0)
 				exit(-r);
@@ -339,7 +343,7 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 
 	pr_info("All vCPU threads joined\n");
 
-	if (use_uffd) {
+	if (p->use_uffd) {
 		char c;
 
 		/* Tell the user fault fd handler threads to quit */
@@ -357,43 +361,23 @@ static void run_test(enum vm_guest_mode mode, bool use_uffd,
 		perf_test_args.vcpu_args[0].pages * nr_vcpus /
 		((double)ts_diff.tv_sec + (double)ts_diff.tv_nsec / 100000000.0));
 
-	ucall_uninit(vm);
-	kvm_vm_free(vm);
+	perf_test_destroy_vm(vm);
 
 	free(guest_data_prototype);
 	free(vcpu_threads);
-	if (use_uffd) {
+	if (p->use_uffd) {
 		free(uffd_handler_threads);
 		free(uffd_args);
 		free(pipefds);
 	}
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-m mode] [-u] [-d uffd_delay_usec]\n"
 	       "          [-b memory] [-v vcpus]\n", name);
-	printf(" -m: specify the guest mode ID to test\n"
-	       "     (default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
+	guest_modes_help();
 	printf(" -u: use User Fault FD to handle vCPU page\n"
 	       "     faults.\n");
 	printf(" -d: add a delay in usec to the User Fault\n"
@@ -410,53 +394,22 @@ static void help(char *name)
 int main(int argc, char *argv[])
 {
 	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
-	bool mode_selected = false;
-	unsigned int mode;
-	int opt, i;
-	bool use_uffd = false;
-	useconds_t uffd_delay = 0;
-
-#ifdef __x86_64__
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
+	struct test_params p = {};
+	int opt;
+
+	guest_modes_append_default();
 
 	while ((opt = getopt(argc, argv, "hm:ud:b:v:")) != -1) {
 		switch (opt) {
 		case 'm':
-			if (!mode_selected) {
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
+			guest_modes_cmdline(optarg);
 			break;
 		case 'u':
-			use_uffd = true;
+			p.use_uffd = true;
 			break;
 		case 'd':
-			uffd_delay = strtoul(optarg, NULL, 0);
-			TEST_ASSERT(uffd_delay >= 0,
-				    "A negative UFFD delay is not supported.");
+			p.uffd_delay = strtoul(optarg, NULL, 0);
+			TEST_ASSERT(p.uffd_delay >= 0, "A negative UFFD delay is not supported.");
 			break;
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
@@ -473,14 +426,7 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		if (!guest_modes[i].enabled)
-			continue;
-		TEST_ASSERT(guest_modes[i].supported,
-			    "Guest mode ID %d (%s) not supported.",
-			    i, vm_guest_mode_string(i));
-		run_test(i, use_uffd, uffd_delay);
-	}
+	for_each_guest_mode(run_test, &p);
 
 	return 0;
 }
diff --git a/tools/testing/selftests/kvm/dirty_log_perf_test.c b/tools/testing/selftests/kvm/dirty_log_perf_test.c
index 9c6a7be31e03..2283a0ec74a9 100644
--- a/tools/testing/selftests/kvm/dirty_log_perf_test.c
+++ b/tools/testing/selftests/kvm/dirty_log_perf_test.c
@@ -8,29 +8,28 @@
  * Copyright (C) 2020, Google, Inc.
  */
 
-#define _GNU_SOURCE /* for program_invocation_name */
-
 #include <stdio.h>
 #include <stdlib.h>
-#include <unistd.h>
 #include <time.h>
 #include <pthread.h>
 #include <linux/bitmap.h>
-#include <linux/bitops.h>
 
 #include "kvm_util.h"
-#include "perf_test_util.h"
-#include "processor.h"
 #include "test_util.h"
+#include "perf_test_util.h"
+#include "guest_modes.h"
 
 /* How many host loops to run by default (one KVM_GET_DIRTY_LOG for each loop)*/
 #define TEST_HOST_LOOP_N		2UL
 
+static int nr_vcpus = 1;
+static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
+
 /* Host variables */
 static u64 dirty_log_manual_caps;
 static bool host_quit;
 static uint64_t iteration;
-static uint64_t vcpu_last_completed_iteration[MAX_VCPUS];
+static uint64_t vcpu_last_completed_iteration[KVM_MAX_VCPUS];
 
 static void *vcpu_worker(void *data)
 {
@@ -42,7 +41,7 @@ static void *vcpu_worker(void *data)
 	struct timespec ts_diff;
 	struct timespec total = (struct timespec){0};
 	struct timespec avg;
-	struct vcpu_args *vcpu_args = (struct vcpu_args *)data;
+	struct perf_test_vcpu_args *vcpu_args = (struct perf_test_vcpu_args *)data;
 	int vcpu_id = vcpu_args->vcpu_id;
 
 	vcpu_args_set(vm, vcpu_id, 1, vcpu_id);
@@ -89,9 +88,15 @@ static void *vcpu_worker(void *data)
 	return NULL;
 }
 
-static void run_test(enum vm_guest_mode mode, unsigned long iterations,
-		     uint64_t phys_offset, int wr_fract)
+struct test_params {
+	unsigned long iterations;
+	uint64_t phys_offset;
+	int wr_fract;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	pthread_t *vcpu_threads;
 	struct kvm_vm *vm;
 	unsigned long *bmap;
@@ -106,9 +111,9 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	struct kvm_enable_cap cap = {};
 	struct timespec clear_dirty_log_total = (struct timespec){0};
 
-	vm = create_vm(mode, nr_vcpus, guest_percpu_mem_size);
+	vm = perf_test_create_vm(mode, nr_vcpus, guest_percpu_mem_size);
 
-	perf_test_args.wr_fract = wr_fract;
+	perf_test_args.wr_fract = p->wr_fract;
 
 	guest_num_pages = (nr_vcpus * guest_percpu_mem_size) >> vm_get_page_shift(vm);
 	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
@@ -124,7 +129,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	vcpu_threads = malloc(nr_vcpus * sizeof(*vcpu_threads));
 	TEST_ASSERT(vcpu_threads, "Memory allocation failed");
 
-	add_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
+	perf_test_setup_vcpus(vm, nr_vcpus, guest_percpu_mem_size);
 
 	sync_global_to_guest(vm, perf_test_args);
 
@@ -150,13 +155,13 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	/* Enable dirty logging */
 	clock_gettime(CLOCK_MONOTONIC, &start);
-	vm_mem_region_set_flags(vm, TEST_MEM_SLOT_INDEX,
+	vm_mem_region_set_flags(vm, PERF_TEST_MEM_SLOT_INDEX,
 				KVM_MEM_LOG_DIRTY_PAGES);
 	ts_diff = timespec_diff_now(start);
 	pr_info("Enabling dirty logging time: %ld.%.9lds\n\n",
 		ts_diff.tv_sec, ts_diff.tv_nsec);
 
-	while (iteration < iterations) {
+	while (iteration < p->iterations) {
 		/*
 		 * Incrementing the iteration number will start the vCPUs
 		 * dirtying memory again.
@@ -177,7 +182,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 			iteration, ts_diff.tv_sec, ts_diff.tv_nsec);
 
 		clock_gettime(CLOCK_MONOTONIC, &start);
-		kvm_vm_get_dirty_log(vm, TEST_MEM_SLOT_INDEX, bmap);
+		kvm_vm_get_dirty_log(vm, PERF_TEST_MEM_SLOT_INDEX, bmap);
 
 		ts_diff = timespec_diff_now(start);
 		get_dirty_log_total = timespec_add(get_dirty_log_total,
@@ -187,7 +192,7 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 		if (dirty_log_manual_caps) {
 			clock_gettime(CLOCK_MONOTONIC, &start);
-			kvm_vm_clear_dirty_log(vm, TEST_MEM_SLOT_INDEX, bmap, 0,
+			kvm_vm_clear_dirty_log(vm, PERF_TEST_MEM_SLOT_INDEX, bmap, 0,
 					       host_num_pages);
 
 			ts_diff = timespec_diff_now(start);
@@ -205,43 +210,30 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	/* Disable dirty logging */
 	clock_gettime(CLOCK_MONOTONIC, &start);
-	vm_mem_region_set_flags(vm, TEST_MEM_SLOT_INDEX, 0);
+	vm_mem_region_set_flags(vm, PERF_TEST_MEM_SLOT_INDEX, 0);
 	ts_diff = timespec_diff_now(start);
 	pr_info("Disabling dirty logging time: %ld.%.9lds\n",
 		ts_diff.tv_sec, ts_diff.tv_nsec);
 
-	avg = timespec_div(get_dirty_log_total, iterations);
+	avg = timespec_div(get_dirty_log_total, p->iterations);
 	pr_info("Get dirty log over %lu iterations took %ld.%.9lds. (Avg %ld.%.9lds/iteration)\n",
-		iterations, get_dirty_log_total.tv_sec,
+		p->iterations, get_dirty_log_total.tv_sec,
 		get_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 
 	if (dirty_log_manual_caps) {
-		avg = timespec_div(clear_dirty_log_total, iterations);
+		avg = timespec_div(clear_dirty_log_total, p->iterations);
 		pr_info("Clear dirty log over %lu iterations took %ld.%.9lds. (Avg %ld.%.9lds/iteration)\n",
-			iterations, clear_dirty_log_total.tv_sec,
+			p->iterations, clear_dirty_log_total.tv_sec,
 			clear_dirty_log_total.tv_nsec, avg.tv_sec, avg.tv_nsec);
 	}
 
 	free(bmap);
 	free(vcpu_threads);
-	ucall_uninit(vm);
-	kvm_vm_free(vm);
+	perf_test_destroy_vm(vm);
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-i iterations] [-p offset] "
 	       "[-m mode] [-b vcpu bytes] [-v vcpus]\n", name);
@@ -250,14 +242,7 @@ static void help(char *name)
 	       TEST_HOST_LOOP_N);
 	printf(" -p: specify guest physical test memory offset\n"
 	       "     Warning: a low offset can conflict with the loaded test code.\n");
-	printf(" -m: specify the guest mode ID to test "
-	       "(default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
+	guest_modes_help();
 	printf(" -b: specify the size of the memory region which should be\n"
 	       "     dirtied by each vCPU. e.g. 10M or 3G.\n"
 	       "     (default: 1G)\n");
@@ -272,74 +257,43 @@ static void help(char *name)
 
 int main(int argc, char *argv[])
 {
-	unsigned long iterations = TEST_HOST_LOOP_N;
-	bool mode_selected = false;
-	uint64_t phys_offset = 0;
-	unsigned int mode;
-	int opt, i;
-	int wr_fract = 1;
+	int max_vcpus = kvm_check_cap(KVM_CAP_MAX_VCPUS);
+	struct test_params p = {
+		.iterations = TEST_HOST_LOOP_N,
+		.wr_fract = 1,
+	};
+	int opt;
 
 	dirty_log_manual_caps =
 		kvm_check_cap(KVM_CAP_MANUAL_DIRTY_LOG_PROTECT2);
 	dirty_log_manual_caps &= (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE |
 				  KVM_DIRTY_LOG_INITIALLY_SET);
 
-#ifdef __x86_64__
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
+	guest_modes_append_default();
 
 	while ((opt = getopt(argc, argv, "hi:p:m:b:f:v:")) != -1) {
 		switch (opt) {
 		case 'i':
-			iterations = strtol(optarg, NULL, 10);
+			p.iterations = strtol(optarg, NULL, 10);
 			break;
 		case 'p':
-			phys_offset = strtoull(optarg, NULL, 0);
+			p.phys_offset = strtoull(optarg, NULL, 0);
 			break;
 		case 'm':
-			if (!mode_selected) {
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
+			guest_modes_cmdline(optarg);
 			break;
 		case 'b':
 			guest_percpu_mem_size = parse_size(optarg);
 			break;
 		case 'f':
-			wr_fract = atoi(optarg);
-			TEST_ASSERT(wr_fract >= 1,
+			p.wr_fract = atoi(optarg);
+			TEST_ASSERT(p.wr_fract >= 1,
 				    "Write fraction cannot be less than one");
 			break;
 		case 'v':
 			nr_vcpus = atoi(optarg);
-			TEST_ASSERT(nr_vcpus > 0,
-				    "Must have a positive number of vCPUs");
-			TEST_ASSERT(nr_vcpus <= MAX_VCPUS,
-				    "This test does not currently support\n"
-				    "more than %d vCPUs.", MAX_VCPUS);
+			TEST_ASSERT(nr_vcpus > 0 && nr_vcpus <= max_vcpus,
+				    "Invalid number of vcpus, must be between 1 and %d", max_vcpus);
 			break;
 		case 'h':
 		default:
@@ -348,18 +302,11 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	TEST_ASSERT(iterations >= 2, "The test should have at least two iterations");
+	TEST_ASSERT(p.iterations >= 2, "The test should have at least two iterations");
 
-	pr_info("Test iterations: %"PRIu64"\n",	iterations);
+	pr_info("Test iterations: %"PRIu64"\n",	p.iterations);
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		if (!guest_modes[i].enabled)
-			continue;
-		TEST_ASSERT(guest_modes[i].supported,
-			    "Guest mode ID %d (%s) not supported.",
-			    i, vm_guest_mode_string(i));
-		run_test(i, iterations, phys_offset, wr_fract);
-	}
+	for_each_guest_mode(run_test, &p);
 
 	return 0;
 }
diff --git a/tools/testing/selftests/kvm/dirty_log_test.c b/tools/testing/selftests/kvm/dirty_log_test.c
index 471baecb7772..bb2752d78fe3 100644
--- a/tools/testing/selftests/kvm/dirty_log_test.c
+++ b/tools/testing/selftests/kvm/dirty_log_test.c
@@ -9,8 +9,6 @@
 
 #include <stdio.h>
 #include <stdlib.h>
-#include <unistd.h>
-#include <time.h>
 #include <pthread.h>
 #include <semaphore.h>
 #include <sys/types.h>
@@ -20,8 +18,9 @@
 #include <linux/bitops.h>
 #include <asm/barrier.h>
 
-#include "test_util.h"
 #include "kvm_util.h"
+#include "test_util.h"
+#include "guest_modes.h"
 #include "processor.h"
 
 #define VCPU_ID				1
@@ -673,9 +672,15 @@ static struct kvm_vm *create_vm(enum vm_guest_mode mode, uint32_t vcpuid,
 #define DIRTY_MEM_BITS 30 /* 1G */
 #define PAGE_SHIFT_4K  12
 
-static void run_test(enum vm_guest_mode mode, unsigned long iterations,
-		     unsigned long interval, uint64_t phys_offset)
+struct test_params {
+	unsigned long iterations;
+	unsigned long interval;
+	uint64_t phys_offset;
+};
+
+static void run_test(enum vm_guest_mode mode, void *arg)
 {
+	struct test_params *p = arg;
 	struct kvm_vm *vm;
 	unsigned long *bmap;
 
@@ -709,12 +714,12 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	host_page_size = getpagesize();
 	host_num_pages = vm_num_host_pages(mode, guest_num_pages);
 
-	if (!phys_offset) {
+	if (!p->phys_offset) {
 		guest_test_phys_mem = (vm_get_max_gfn(vm) -
 				       guest_num_pages) * guest_page_size;
 		guest_test_phys_mem &= ~(host_page_size - 1);
 	} else {
-		guest_test_phys_mem = phys_offset;
+		guest_test_phys_mem = p->phys_offset;
 	}
 
 #ifdef __s390x__
@@ -758,9 +763,9 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 
 	pthread_create(&vcpu_thread, NULL, vcpu_worker, vm);
 
-	while (iteration < iterations) {
+	while (iteration < p->iterations) {
 		/* Give the vcpu thread some time to dirty some pages */
-		usleep(interval * 1000);
+		usleep(p->interval * 1000);
 		log_mode_collect_dirty_pages(vm, TEST_MEM_SLOT_INDEX,
 					     bmap, host_num_pages);
 		vm_dirty_log_verify(mode, bmap);
@@ -783,20 +788,8 @@ static void run_test(enum vm_guest_mode mode, unsigned long iterations,
 	kvm_vm_free(vm);
 }
 
-struct guest_mode {
-	bool supported;
-	bool enabled;
-};
-static struct guest_mode guest_modes[NUM_VM_MODES];
-
-#define guest_mode_init(mode, supported, enabled) ({ \
-	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
-})
-
 static void help(char *name)
 {
-	int i;
-
 	puts("");
 	printf("usage: %s [-h] [-i iterations] [-I interval] "
 	       "[-p offset] [-m mode]\n", name);
@@ -813,51 +806,23 @@ static void help(char *name)
 	printf(" -M: specify the host logging mode "
 	       "(default: run all log modes).  Supported modes: \n\t");
 	log_modes_dump();
-	printf(" -m: specify the guest mode ID to test "
-	       "(default: test all supported modes)\n"
-	       "     This option may be used multiple times.\n"
-	       "     Guest mode IDs:\n");
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
-		       guest_modes[i].supported ? " (supported)" : "");
-	}
+	guest_modes_help();
 	puts("");
 	exit(0);
 }
 
 int main(int argc, char *argv[])
 {
-	unsigned long iterations = TEST_HOST_LOOP_N;
-	unsigned long interval = TEST_HOST_LOOP_INTERVAL;
-	bool mode_selected = false;
-	uint64_t phys_offset = 0;
-	unsigned int mode;
-	int opt, i, j;
+	struct test_params p = {
+		.iterations = TEST_HOST_LOOP_N,
+		.interval = TEST_HOST_LOOP_INTERVAL,
+	};
+	int opt, i;
 
 	sem_init(&dirty_ring_vcpu_stop, 0, 0);
 	sem_init(&dirty_ring_vcpu_cont, 0, 0);
 
-#ifdef __x86_64__
-	guest_mode_init(VM_MODE_PXXV48_4K, true, true);
-#endif
-#ifdef __aarch64__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-	guest_mode_init(VM_MODE_P40V48_64K, true, true);
-
-	{
-		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
-
-		if (limit >= 52)
-			guest_mode_init(VM_MODE_P52V48_64K, true, true);
-		if (limit >= 48) {
-			guest_mode_init(VM_MODE_P48V48_4K, true, true);
-			guest_mode_init(VM_MODE_P48V48_64K, true, true);
-		}
-	}
-#endif
-#ifdef __s390x__
-	guest_mode_init(VM_MODE_P40V48_4K, true, true);
-#endif
+	guest_modes_append_default();
 
 	while ((opt = getopt(argc, argv, "c:hi:I:p:m:M:")) != -1) {
 		switch (opt) {
@@ -865,24 +830,16 @@ int main(int argc, char *argv[])
 			test_dirty_ring_count = strtol(optarg, NULL, 10);
 			break;
 		case 'i':
-			iterations = strtol(optarg, NULL, 10);
+			p.iterations = strtol(optarg, NULL, 10);
 			break;
 		case 'I':
-			interval = strtol(optarg, NULL, 10);
+			p.interval = strtol(optarg, NULL, 10);
 			break;
 		case 'p':
-			phys_offset = strtoull(optarg, NULL, 0);
+			p.phys_offset = strtoull(optarg, NULL, 0);
 			break;
 		case 'm':
-			if (!mode_selected) {
-				for (i = 0; i < NUM_VM_MODES; ++i)
-					guest_modes[i].enabled = false;
-				mode_selected = true;
-			}
-			mode = strtoul(optarg, NULL, 10);
-			TEST_ASSERT(mode < NUM_VM_MODES,
-				    "Guest mode ID %d too big", mode);
-			guest_modes[mode].enabled = true;
+			guest_modes_cmdline(optarg);
 			break;
 		case 'M':
 			if (!strcmp(optarg, "all")) {
@@ -911,32 +868,24 @@ int main(int argc, char *argv[])
 		}
 	}
 
-	TEST_ASSERT(iterations > 2, "Iterations must be greater than two");
-	TEST_ASSERT(interval > 0, "Interval must be greater than zero");
+	TEST_ASSERT(p.iterations > 2, "Iterations must be greater than two");
+	TEST_ASSERT(p.interval > 0, "Interval must be greater than zero");
 
 	pr_info("Test iterations: %"PRIu64", interval: %"PRIu64" (ms)\n",
-		iterations, interval);
+		p.iterations, p.interval);
 
 	srandom(time(0));
 
-	for (i = 0; i < NUM_VM_MODES; ++i) {
-		if (!guest_modes[i].enabled)
-			continue;
-		TEST_ASSERT(guest_modes[i].supported,
-			    "Guest mode ID %d (%s) not supported.",
-			    i, vm_guest_mode_string(i));
-		if (host_log_mode_option == LOG_MODE_ALL) {
-			/* Run each log mode */
-			for (j = 0; j < LOG_MODE_NUM; j++) {
-				pr_info("Testing Log Mode '%s'\n",
-					log_modes[j].name);
-				host_log_mode = j;
-				run_test(i, iterations, interval, phys_offset);
-			}
-		} else {
-			host_log_mode = host_log_mode_option;
-			run_test(i, iterations, interval, phys_offset);
+	if (host_log_mode_option == LOG_MODE_ALL) {
+		/* Run each log mode */
+		for (i = 0; i < LOG_MODE_NUM; i++) {
+			pr_info("Testing Log Mode '%s'\n", log_modes[i].name);
+			host_log_mode = i;
+			for_each_guest_mode(run_test, &p);
 		}
+	} else {
+		host_log_mode = host_log_mode_option;
+		for_each_guest_mode(run_test, &p);
 	}
 
 	return 0;
diff --git a/tools/testing/selftests/kvm/include/guest_modes.h b/tools/testing/selftests/kvm/include/guest_modes.h
new file mode 100644
index 000000000000..b691df33e64e
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/guest_modes.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2020, Red Hat, Inc.
+ */
+#include "kvm_util.h"
+
+struct guest_mode {
+	bool supported;
+	bool enabled;
+};
+
+extern struct guest_mode guest_modes[NUM_VM_MODES];
+
+#define guest_mode_append(mode, supported, enabled) ({ \
+	guest_modes[mode] = (struct guest_mode){ supported, enabled }; \
+})
+
+void guest_modes_append_default(void);
+void for_each_guest_mode(void (*func)(enum vm_guest_mode, void *), void *arg);
+void guest_modes_help(void);
+void guest_modes_cmdline(const char *arg);
diff --git a/tools/testing/selftests/kvm/include/kvm_util.h b/tools/testing/selftests/kvm/include/kvm_util.h
index dfa9d369e8fc..5cbb861525ed 100644
--- a/tools/testing/selftests/kvm/include/kvm_util.h
+++ b/tools/testing/selftests/kvm/include/kvm_util.h
@@ -16,6 +16,7 @@
 
 #include "sparsebit.h"
 
+#define KVM_MAX_VCPUS 512
 
 /*
  * Callers of kvm_util only have an incomplete/opaque description of the
@@ -70,6 +71,14 @@ enum vm_guest_mode {
 #define vm_guest_mode_string(m) vm_guest_mode_string[m]
 extern const char * const vm_guest_mode_string[];
 
+struct vm_guest_mode_params {
+	unsigned int pa_bits;
+	unsigned int va_bits;
+	unsigned int page_size;
+	unsigned int page_shift;
+};
+extern const struct vm_guest_mode_params vm_guest_mode_params[];
+
 enum vm_mem_backing_src_type {
 	VM_MEM_SRC_ANONYMOUS,
 	VM_MEM_SRC_ANONYMOUS_THP,
diff --git a/tools/testing/selftests/kvm/include/perf_test_util.h b/tools/testing/selftests/kvm/include/perf_test_util.h
index 239421e4f6b8..b1188823c31b 100644
--- a/tools/testing/selftests/kvm/include/perf_test_util.h
+++ b/tools/testing/selftests/kvm/include/perf_test_util.h
@@ -9,38 +9,15 @@
 #define SELFTEST_KVM_PERF_TEST_UTIL_H
 
 #include "kvm_util.h"
-#include "processor.h"
-
-#define MAX_VCPUS 512
-
-#define PAGE_SHIFT_4K  12
-#define PTES_PER_4K_PT 512
-
-#define TEST_MEM_SLOT_INDEX		1
 
 /* Default guest test virtual memory offset */
 #define DEFAULT_GUEST_TEST_MEM		0xc0000000
 
 #define DEFAULT_PER_VCPU_MEM_SIZE	(1 << 30) /* 1G */
 
-/*
- * Guest physical memory offset of the testing memory slot.
- * This will be set to the topmost valid physical address minus
- * the test memory size.
- */
-static uint64_t guest_test_phys_mem;
-
-/*
- * Guest virtual memory offset of the testing memory slot.
- * Must not conflict with identity mapped test code.
- */
-static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
-static uint64_t guest_percpu_mem_size = DEFAULT_PER_VCPU_MEM_SIZE;
-
-/* Number of VCPUs for the test */
-static int nr_vcpus = 1;
+#define PERF_TEST_MEM_SLOT_INDEX	1
 
-struct vcpu_args {
+struct perf_test_vcpu_args {
 	uint64_t gva;
 	uint64_t pages;
 
@@ -54,141 +31,21 @@ struct perf_test_args {
 	uint64_t guest_page_size;
 	int wr_fract;
 
-	struct vcpu_args vcpu_args[MAX_VCPUS];
+	struct perf_test_vcpu_args vcpu_args[KVM_MAX_VCPUS];
 };
 
-static struct perf_test_args perf_test_args;
+extern struct perf_test_args perf_test_args;
 
 /*
- * Continuously write to the first 8 bytes of each page in the
- * specified region.
+ * Guest physical memory offset of the testing memory slot.
+ * This will be set to the topmost valid physical address minus
+ * the test memory size.
  */
-static void guest_code(uint32_t vcpu_id)
-{
-	struct vcpu_args *vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
-	uint64_t gva;
-	uint64_t pages;
-	int i;
-
-	/* Make sure vCPU args data structure is not corrupt. */
-	GUEST_ASSERT(vcpu_args->vcpu_id == vcpu_id);
-
-	gva = vcpu_args->gva;
-	pages = vcpu_args->pages;
-
-	while (true) {
-		for (i = 0; i < pages; i++) {
-			uint64_t addr = gva + (i * perf_test_args.guest_page_size);
-
-			if (i % perf_test_args.wr_fract == 0)
-				*(uint64_t *)addr = 0x0123456789ABCDEF;
-			else
-				READ_ONCE(*(uint64_t *)addr);
-		}
-
-		GUEST_SYNC(1);
-	}
-}
-
-static struct kvm_vm *create_vm(enum vm_guest_mode mode, int vcpus,
-				uint64_t vcpu_memory_bytes)
-{
-	struct kvm_vm *vm;
-	uint64_t pages = DEFAULT_GUEST_PHY_PAGES;
-	uint64_t guest_num_pages;
-
-	/* Account for a few pages per-vCPU for stacks */
-	pages += DEFAULT_STACK_PGS * vcpus;
-
-	/*
-	 * Reserve twice the ammount of memory needed to map the test region and
-	 * the page table / stacks region, at 4k, for page tables. Do the
-	 * calculation with 4K page size: the smallest of all archs. (e.g., 64K
-	 * page size guest will need even less memory for page tables).
-	 */
-	pages += (2 * pages) / PTES_PER_4K_PT;
-	pages += ((2 * vcpus * vcpu_memory_bytes) >> PAGE_SHIFT_4K) /
-		 PTES_PER_4K_PT;
-	pages = vm_adjust_num_guest_pages(mode, pages);
-
-	pr_info("Testing guest mode: %s\n", vm_guest_mode_string(mode));
-
-	vm = vm_create(mode, pages, O_RDWR);
-	kvm_vm_elf_load(vm, program_invocation_name, 0, 0);
-#ifdef __x86_64__
-	vm_create_irqchip(vm);
-#endif
-
-	perf_test_args.vm = vm;
-	perf_test_args.guest_page_size = vm_get_page_size(vm);
-	perf_test_args.host_page_size = getpagesize();
-
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.guest_page_size == 0,
-		    "Guest memory size is not guest page size aligned.");
-
-	guest_num_pages = (vcpus * vcpu_memory_bytes) /
-			  perf_test_args.guest_page_size;
-	guest_num_pages = vm_adjust_num_guest_pages(mode, guest_num_pages);
-
-	/*
-	 * If there should be more memory in the guest test region than there
-	 * can be pages in the guest, it will definitely cause problems.
-	 */
-	TEST_ASSERT(guest_num_pages < vm_get_max_gfn(vm),
-		    "Requested more guest memory than address space allows.\n"
-		    "    guest pages: %lx max gfn: %x vcpus: %d wss: %lx]\n",
-		    guest_num_pages, vm_get_max_gfn(vm), vcpus,
-		    vcpu_memory_bytes);
-
-	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.host_page_size == 0,
-		    "Guest memory size is not host page size aligned.");
-
-	guest_test_phys_mem = (vm_get_max_gfn(vm) - guest_num_pages) *
-			      perf_test_args.guest_page_size;
-	guest_test_phys_mem &= ~(perf_test_args.host_page_size - 1);
-
-#ifdef __s390x__
-	/* Align to 1M (segment size) */
-	guest_test_phys_mem &= ~((1 << 20) - 1);
-#endif
-
-	pr_info("guest physical test memory offset: 0x%lx\n", guest_test_phys_mem);
-
-	/* Add an extra memory slot for testing */
-	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
-				    guest_test_phys_mem,
-				    TEST_MEM_SLOT_INDEX,
-				    guest_num_pages, 0);
-
-	/* Do mapping for the demand paging memory slot */
-	virt_map(vm, guest_test_virt_mem, guest_test_phys_mem, guest_num_pages, 0);
-
-	ucall_init(vm, NULL);
-
-	return vm;
-}
-
-static void add_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes)
-{
-	vm_paddr_t vcpu_gpa;
-	struct vcpu_args *vcpu_args;
-	int vcpu_id;
-
-	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
-		vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
-
-		vm_vcpu_add_default(vm, vcpu_id, guest_code);
-
-		vcpu_args->vcpu_id = vcpu_id;
-		vcpu_args->gva = guest_test_virt_mem +
-				 (vcpu_id * vcpu_memory_bytes);
-		vcpu_args->pages = vcpu_memory_bytes /
-				   perf_test_args.guest_page_size;
+extern uint64_t guest_test_phys_mem;
 
-		vcpu_gpa = guest_test_phys_mem + (vcpu_id * vcpu_memory_bytes);
-		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
-			 vcpu_id, vcpu_gpa, vcpu_gpa + vcpu_memory_bytes);
-	}
-}
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
+				uint64_t vcpu_memory_bytes);
+void perf_test_destroy_vm(struct kvm_vm *vm);
+void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes);
 
 #endif /* SELFTEST_KVM_PERF_TEST_UTIL_H */
diff --git a/tools/testing/selftests/kvm/lib/guest_modes.c b/tools/testing/selftests/kvm/lib/guest_modes.c
new file mode 100644
index 000000000000..25bff307c71f
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/guest_modes.c
@@ -0,0 +1,70 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020, Red Hat, Inc.
+ */
+#include "guest_modes.h"
+
+struct guest_mode guest_modes[NUM_VM_MODES];
+
+void guest_modes_append_default(void)
+{
+	guest_mode_append(VM_MODE_DEFAULT, true, true);
+
+#ifdef __aarch64__
+	guest_mode_append(VM_MODE_P40V48_64K, true, true);
+	{
+		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
+		if (limit >= 52)
+			guest_mode_append(VM_MODE_P52V48_64K, true, true);
+		if (limit >= 48) {
+			guest_mode_append(VM_MODE_P48V48_4K, true, true);
+			guest_mode_append(VM_MODE_P48V48_64K, true, true);
+		}
+	}
+#endif
+}
+
+void for_each_guest_mode(void (*func)(enum vm_guest_mode, void *), void *arg)
+{
+	int i;
+
+	for (i = 0; i < NUM_VM_MODES; ++i) {
+		if (!guest_modes[i].enabled)
+			continue;
+		TEST_ASSERT(guest_modes[i].supported,
+			    "Guest mode ID %d (%s) not supported.",
+			    i, vm_guest_mode_string(i));
+		func(i, arg);
+	}
+}
+
+void guest_modes_help(void)
+{
+	int i;
+
+	printf(" -m: specify the guest mode ID to test\n"
+	       "     (default: test all supported modes)\n"
+	       "     This option may be used multiple times.\n"
+	       "     Guest mode IDs:\n");
+	for (i = 0; i < NUM_VM_MODES; ++i) {
+		printf("         %d:    %s%s\n", i, vm_guest_mode_string(i),
+		       guest_modes[i].supported ? " (supported)" : "");
+	}
+}
+
+void guest_modes_cmdline(const char *arg)
+{
+	static bool mode_selected;
+	unsigned int mode;
+	int i;
+
+	if (!mode_selected) {
+		for (i = 0; i < NUM_VM_MODES; ++i)
+			guest_modes[i].enabled = false;
+		mode_selected = true;
+	}
+
+	mode = strtoul(optarg, NULL, 10);
+	TEST_ASSERT(mode < NUM_VM_MODES, "Guest mode ID %d too big", mode);
+	guest_modes[mode].enabled = true;
+}
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index 88ef7067f1e6..fa5a90e6c6f0 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -153,14 +153,7 @@ const char * const vm_guest_mode_string[] = {
 _Static_assert(sizeof(vm_guest_mode_string)/sizeof(char *) == NUM_VM_MODES,
 	       "Missing new mode strings?");
 
-struct vm_guest_mode_params {
-	unsigned int pa_bits;
-	unsigned int va_bits;
-	unsigned int page_size;
-	unsigned int page_shift;
-};
-
-static const struct vm_guest_mode_params vm_guest_mode_params[] = {
+const struct vm_guest_mode_params vm_guest_mode_params[] = {
 	{ 52, 48,  0x1000, 12 },
 	{ 52, 48, 0x10000, 16 },
 	{ 48, 48,  0x1000, 12 },
diff --git a/tools/testing/selftests/kvm/lib/perf_test_util.c b/tools/testing/selftests/kvm/lib/perf_test_util.c
new file mode 100644
index 000000000000..9be1944c2d1c
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/perf_test_util.c
@@ -0,0 +1,134 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2020, Google LLC.
+ */
+
+#include "kvm_util.h"
+#include "perf_test_util.h"
+#include "processor.h"
+
+struct perf_test_args perf_test_args;
+
+uint64_t guest_test_phys_mem;
+
+/*
+ * Guest virtual memory offset of the testing memory slot.
+ * Must not conflict with identity mapped test code.
+ */
+static uint64_t guest_test_virt_mem = DEFAULT_GUEST_TEST_MEM;
+
+/*
+ * Continuously write to the first 8 bytes of each page in the
+ * specified region.
+ */
+static void guest_code(uint32_t vcpu_id)
+{
+	struct perf_test_vcpu_args *vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+	uint64_t gva;
+	uint64_t pages;
+	int i;
+
+	/* Make sure vCPU args data structure is not corrupt. */
+	GUEST_ASSERT(vcpu_args->vcpu_id == vcpu_id);
+
+	gva = vcpu_args->gva;
+	pages = vcpu_args->pages;
+
+	while (true) {
+		for (i = 0; i < pages; i++) {
+			uint64_t addr = gva + (i * perf_test_args.guest_page_size);
+
+			if (i % perf_test_args.wr_fract == 0)
+				*(uint64_t *)addr = 0x0123456789ABCDEF;
+			else
+				READ_ONCE(*(uint64_t *)addr);
+		}
+
+		GUEST_SYNC(1);
+	}
+}
+
+struct kvm_vm *perf_test_create_vm(enum vm_guest_mode mode, int vcpus,
+				   uint64_t vcpu_memory_bytes)
+{
+	struct kvm_vm *vm;
+	uint64_t guest_num_pages;
+
+	pr_info("Testing guest mode: %s\n", vm_guest_mode_string(mode));
+
+	perf_test_args.host_page_size = getpagesize();
+	perf_test_args.guest_page_size = vm_guest_mode_params[mode].page_size;
+
+	guest_num_pages = vm_adjust_num_guest_pages(mode,
+				(vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size);
+
+	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.host_page_size == 0,
+		    "Guest memory size is not host page size aligned.");
+	TEST_ASSERT(vcpu_memory_bytes % perf_test_args.guest_page_size == 0,
+		    "Guest memory size is not guest page size aligned.");
+
+	vm = vm_create_with_vcpus(mode, vcpus,
+				  (vcpus * vcpu_memory_bytes) / perf_test_args.guest_page_size,
+				  0, guest_code, NULL);
+
+	perf_test_args.vm = vm;
+
+	/*
+	 * If there should be more memory in the guest test region than there
+	 * can be pages in the guest, it will definitely cause problems.
+	 */
+	TEST_ASSERT(guest_num_pages < vm_get_max_gfn(vm),
+		    "Requested more guest memory than address space allows.\n"
+		    "    guest pages: %lx max gfn: %x vcpus: %d wss: %lx]\n",
+		    guest_num_pages, vm_get_max_gfn(vm), vcpus,
+		    vcpu_memory_bytes);
+
+	guest_test_phys_mem = (vm_get_max_gfn(vm) - guest_num_pages) *
+			      perf_test_args.guest_page_size;
+	guest_test_phys_mem &= ~(perf_test_args.host_page_size - 1);
+#ifdef __s390x__
+	/* Align to 1M (segment size) */
+	guest_test_phys_mem &= ~((1 << 20) - 1);
+#endif
+	pr_info("guest physical test memory offset: 0x%lx\n", guest_test_phys_mem);
+
+	/* Add an extra memory slot for testing */
+	vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
+				    guest_test_phys_mem,
+				    PERF_TEST_MEM_SLOT_INDEX,
+				    guest_num_pages, 0);
+
+	/* Do mapping for the demand paging memory slot */
+	virt_map(vm, guest_test_virt_mem, guest_test_phys_mem, guest_num_pages, 0);
+
+	ucall_init(vm, NULL);
+
+	return vm;
+}
+
+void perf_test_destroy_vm(struct kvm_vm *vm)
+{
+	ucall_uninit(vm);
+	kvm_vm_free(vm);
+}
+
+void perf_test_setup_vcpus(struct kvm_vm *vm, int vcpus, uint64_t vcpu_memory_bytes)
+{
+	vm_paddr_t vcpu_gpa;
+	struct perf_test_vcpu_args *vcpu_args;
+	int vcpu_id;
+
+	for (vcpu_id = 0; vcpu_id < vcpus; vcpu_id++) {
+		vcpu_args = &perf_test_args.vcpu_args[vcpu_id];
+
+		vcpu_args->vcpu_id = vcpu_id;
+		vcpu_args->gva = guest_test_virt_mem +
+				 (vcpu_id * vcpu_memory_bytes);
+		vcpu_args->pages = vcpu_memory_bytes /
+				   perf_test_args.guest_page_size;
+
+		vcpu_gpa = guest_test_phys_mem + (vcpu_id * vcpu_memory_bytes);
+		pr_debug("Added VCPU %d with test mem gpa [%lx, %lx)\n",
+			 vcpu_id, vcpu_gpa, vcpu_gpa + vcpu_memory_bytes);
+	}
+}
diff --git a/tools/testing/selftests/net/fib_nexthops.sh b/tools/testing/selftests/net/fib_nexthops.sh
index eb693a3b7b4a..4c7d33618437 100755
--- a/tools/testing/selftests/net/fib_nexthops.sh
+++ b/tools/testing/selftests/net/fib_nexthops.sh
@@ -869,7 +869,7 @@ ipv6_torture()
 	pid3=$!
 	ip netns exec me ping -f 2001:db8:101::2 >/dev/null 2>&1 &
 	pid4=$!
-	ip netns exec me mausezahn veth1 -B 2001:db8:101::2 -A 2001:db8:91::1 -c 0 -t tcp "dp=1-1023, flags=syn" >/dev/null 2>&1 &
+	ip netns exec me mausezahn -6 veth1 -B 2001:db8:101::2 -A 2001:db8:91::1 -c 0 -t tcp "dp=1-1023, flags=syn" >/dev/null 2>&1 &
 	pid5=$!
 
 	sleep 300
diff --git a/tools/testing/selftests/net/pmtu.sh b/tools/testing/selftests/net/pmtu.sh
index 464e31eabc73..64cd2e23c568 100755
--- a/tools/testing/selftests/net/pmtu.sh
+++ b/tools/testing/selftests/net/pmtu.sh
@@ -162,7 +162,15 @@
 # - list_flush_ipv6_exception
 #	Using the same topology as in pmtu_ipv6, create exceptions, and check
 #	they are shown when listing exception caches, gone after flushing them
-
+#
+# - pmtu_ipv4_route_change
+#	Use the same topology as in pmtu_ipv4, but issue a route replacement
+#	command and delete the corresponding device afterward. This tests for
+#	proper cleanup of the PMTU exceptions by the route replacement path.
+#	Device unregistration should complete successfully
+#
+# - pmtu_ipv6_route_change
+#	Same as above but with IPv6
 
 # Kselftest framework requirement - SKIP code is 4.
 ksft_skip=4
@@ -224,7 +232,9 @@ tests="
 	cleanup_ipv4_exception		ipv4: cleanup of cached exceptions	1
 	cleanup_ipv6_exception		ipv6: cleanup of cached exceptions	1
 	list_flush_ipv4_exception	ipv4: list and flush cached exceptions	1
-	list_flush_ipv6_exception	ipv6: list and flush cached exceptions	1"
+	list_flush_ipv6_exception	ipv6: list and flush cached exceptions	1
+	pmtu_ipv4_route_change		ipv4: PMTU exception w/route replace	1
+	pmtu_ipv6_route_change		ipv6: PMTU exception w/route replace	1"
 
 NS_A="ns-A"
 NS_B="ns-B"
@@ -1782,6 +1792,63 @@ test_list_flush_ipv6_exception() {
 	return ${fail}
 }
 
+test_pmtu_ipvX_route_change() {
+	family=${1}
+
+	setup namespaces routing || return 2
+	trace "${ns_a}"  veth_A-R1    "${ns_r1}" veth_R1-A \
+	      "${ns_r1}" veth_R1-B    "${ns_b}"  veth_B-R1 \
+	      "${ns_a}"  veth_A-R2    "${ns_r2}" veth_R2-A \
+	      "${ns_r2}" veth_R2-B    "${ns_b}"  veth_B-R2
+
+	if [ ${family} -eq 4 ]; then
+		ping=ping
+		dst1="${prefix4}.${b_r1}.1"
+		dst2="${prefix4}.${b_r2}.1"
+		gw="${prefix4}.${a_r1}.2"
+	else
+		ping=${ping6}
+		dst1="${prefix6}:${b_r1}::1"
+		dst2="${prefix6}:${b_r2}::1"
+		gw="${prefix6}:${a_r1}::2"
+	fi
+
+	# Set up initial MTU values
+	mtu "${ns_a}"  veth_A-R1 2000
+	mtu "${ns_r1}" veth_R1-A 2000
+	mtu "${ns_r1}" veth_R1-B 1400
+	mtu "${ns_b}"  veth_B-R1 1400
+
+	mtu "${ns_a}"  veth_A-R2 2000
+	mtu "${ns_r2}" veth_R2-A 2000
+	mtu "${ns_r2}" veth_R2-B 1500
+	mtu "${ns_b}"  veth_B-R2 1500
+
+	# Create route exceptions
+	run_cmd ${ns_a} ${ping} -q -M want -i 0.1 -w 1 -s 1800 ${dst1}
+	run_cmd ${ns_a} ${ping} -q -M want -i 0.1 -w 1 -s 1800 ${dst2}
+
+	# Check that exceptions have been created with the correct PMTU
+	pmtu_1="$(route_get_dst_pmtu_from_exception "${ns_a}" ${dst1})"
+	check_pmtu_value "1400" "${pmtu_1}" "exceeding MTU" || return 1
+	pmtu_2="$(route_get_dst_pmtu_from_exception "${ns_a}" ${dst2})"
+	check_pmtu_value "1500" "${pmtu_2}" "exceeding MTU" || return 1
+
+	# Replace the route from A to R1
+	run_cmd ${ns_a} ip route change default via ${gw}
+
+	# Delete the device in A
+	run_cmd ${ns_a} ip link del "veth_A-R1"
+}
+
+test_pmtu_ipv4_route_change() {
+	test_pmtu_ipvX_route_change 4
+}
+
+test_pmtu_ipv6_route_change() {
+	test_pmtu_ipvX_route_change 6
+}
+
 usage() {
 	echo
 	echo "$0 [OPTIONS] [TEST]..."
diff --git a/tools/testing/selftests/net/tls.c b/tools/testing/selftests/net/tls.c
index cb0d1890a860..e0088c2d38a5 100644
--- a/tools/testing/selftests/net/tls.c
+++ b/tools/testing/selftests/net/tls.c
@@ -103,8 +103,8 @@ FIXTURE(tls)
 
 FIXTURE_VARIANT(tls)
 {
-	u16 tls_version;
-	u16 cipher_type;
+	uint16_t tls_version;
+	uint16_t cipher_type;
 };
 
 FIXTURE_VARIANT_ADD(tls, 12_gcm)
diff --git a/tools/testing/selftests/net/udpgro.sh b/tools/testing/selftests/net/udpgro.sh
index ac2a30be9b32..f8a19f548ae9 100755
--- a/tools/testing/selftests/net/udpgro.sh
+++ b/tools/testing/selftests/net/udpgro.sh
@@ -5,6 +5,14 @@
 
 readonly PEER_NS="ns-peer-$(mktemp -u XXXXXX)"
 
+# set global exit status, but never reset nonzero one.
+check_err()
+{
+	if [ $ret -eq 0 ]; then
+		ret=$1
+	fi
+}
+
 cleanup() {
 	local -r jobs="$(jobs -p)"
 	local -r ns="$(ip netns list|grep $PEER_NS)"
@@ -44,7 +52,9 @@ run_one() {
 	# Hack: let bg programs complete the startup
 	sleep 0.1
 	./udpgso_bench_tx ${tx_args}
+	ret=$?
 	wait $(jobs -p)
+	return $ret
 }
 
 run_test() {
@@ -87,8 +97,10 @@ run_one_nat() {
 
 	sleep 0.1
 	./udpgso_bench_tx ${tx_args}
+	ret=$?
 	kill -INT $pid
 	wait $(jobs -p)
+	return $ret
 }
 
 run_one_2sock() {
@@ -110,7 +122,9 @@ run_one_2sock() {
 	sleep 0.1
 	# first UDP GSO socket should be closed at this point
 	./udpgso_bench_tx ${tx_args}
+	ret=$?
 	wait $(jobs -p)
+	return $ret
 }
 
 run_nat_test() {
@@ -131,36 +145,54 @@ run_all() {
 	local -r core_args="-l 4"
 	local -r ipv4_args="${core_args} -4 -D 192.168.1.1"
 	local -r ipv6_args="${core_args} -6 -D 2001:db8::1"
+	ret=0
 
 	echo "ipv4"
 	run_test "no GRO" "${ipv4_args} -M 10 -s 1400" "-4 -n 10 -l 1400"
+	check_err $?
 
 	# explicitly check we are not receiving UDP_SEGMENT cmsg (-S -1)
 	# when GRO does not take place
 	run_test "no GRO chk cmsg" "${ipv4_args} -M 10 -s 1400" "-4 -n 10 -l 1400 -S -1"
+	check_err $?
 
 	# the GSO packets are aggregated because:
 	# * veth schedule napi after each xmit
 	# * segmentation happens in BH context, veth napi poll is delayed after
 	#   the transmission of the last segment
 	run_test "GRO" "${ipv4_args} -M 1 -s 14720 -S 0 " "-4 -n 1 -l 14720"
+	check_err $?
 	run_test "GRO chk cmsg" "${ipv4_args} -M 1 -s 14720 -S 0 " "-4 -n 1 -l 14720 -S 1472"
+	check_err $?
 	run_test "GRO with custom segment size" "${ipv4_args} -M 1 -s 14720 -S 500 " "-4 -n 1 -l 14720"
+	check_err $?
 	run_test "GRO with custom segment size cmsg" "${ipv4_args} -M 1 -s 14720 -S 500 " "-4 -n 1 -l 14720 -S 500"
+	check_err $?
 
 	run_nat_test "bad GRO lookup" "${ipv4_args} -M 1 -s 14720 -S 0" "-n 10 -l 1472"
+	check_err $?
 	run_2sock_test "multiple GRO socks" "${ipv4_args} -M 1 -s 14720 -S 0 " "-4 -n 1 -l 14720 -S 1472"
+	check_err $?
 
 	echo "ipv6"
 	run_test "no GRO" "${ipv6_args} -M 10 -s 1400" "-n 10 -l 1400"
+	check_err $?
 	run_test "no GRO chk cmsg" "${ipv6_args} -M 10 -s 1400" "-n 10 -l 1400 -S -1"
+	check_err $?
 	run_test "GRO" "${ipv6_args} -M 1 -s 14520 -S 0" "-n 1 -l 14520"
+	check_err $?
 	run_test "GRO chk cmsg" "${ipv6_args} -M 1 -s 14520 -S 0" "-n 1 -l 14520 -S 1452"
+	check_err $?
 	run_test "GRO with custom segment size" "${ipv6_args} -M 1 -s 14520 -S 500" "-n 1 -l 14520"
+	check_err $?
 	run_test "GRO with custom segment size cmsg" "${ipv6_args} -M 1 -s 14520 -S 500" "-n 1 -l 14520 -S 500"
+	check_err $?
 
 	run_nat_test "bad GRO lookup" "${ipv6_args} -M 1 -s 14520 -S 0" "-n 10 -l 1452"
+	check_err $?
 	run_2sock_test "multiple GRO socks" "${ipv6_args} -M 1 -s 14520 -S 0 " "-n 1 -l 14520 -S 1452"
+	check_err $?
+	return $ret
 }
 
 if [ ! -f ../bpf/xdp_dummy.o ]; then
@@ -180,3 +212,5 @@ elif [[ $1 == "__subprocess_2sock" ]]; then
 	shift
 	run_one_2sock $@
 fi
+
+exit $?
diff --git a/tools/testing/selftests/netfilter/Makefile b/tools/testing/selftests/netfilter/Makefile
index a374e10ef506..3006a8e5b41a 100644
--- a/tools/testing/selftests/netfilter/Makefile
+++ b/tools/testing/selftests/netfilter/Makefile
@@ -4,7 +4,8 @@
 TEST_PROGS := nft_trans_stress.sh nft_nat.sh bridge_brouter.sh \
 	conntrack_icmp_related.sh nft_flowtable.sh ipvs.sh \
 	nft_concat_range.sh nft_conntrack_helper.sh \
-	nft_queue.sh nft_meta.sh
+	nft_queue.sh nft_meta.sh \
+	ipip-conntrack-mtu.sh
 
 LDLIBS = -lmnl
 TEST_GEN_FILES =  nf-queue
diff --git a/tools/testing/selftests/netfilter/ipip-conntrack-mtu.sh b/tools/testing/selftests/netfilter/ipip-conntrack-mtu.sh
new file mode 100755
index 000000000000..4a6f5c3b3215
--- /dev/null
+++ b/tools/testing/selftests/netfilter/ipip-conntrack-mtu.sh
@@ -0,0 +1,206 @@
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+# Kselftest framework requirement - SKIP code is 4.
+ksft_skip=4
+
+# Conntrack needs to reassemble fragments in order to have complete
+# packets for rule matching.  Reassembly can lead to packet loss.
+
+# Consider the following setup:
+#            +--------+       +---------+       +--------+
+#            |Router A|-------|Wanrouter|-------|Router B|
+#            |        |.IPIP..|         |..IPIP.|        |
+#            +--------+       +---------+       +--------+
+#           /                  mtu 1400                   \
+#          /                                               \
+#+--------+                                                 +--------+
+#|Client A|                                                 |Client B|
+#|        |                                                 |        |
+#+--------+                                                 +--------+
+
+# Router A and Router B use IPIP tunnel interfaces to tunnel traffic
+# between Client A and Client B over WAN. Wanrouter has MTU 1400 set
+# on its interfaces.
+
+rnd=$(mktemp -u XXXXXXXX)
+rx=$(mktemp)
+
+r_a="ns-ra-$rnd"
+r_b="ns-rb-$rnd"
+r_w="ns-rw-$rnd"
+c_a="ns-ca-$rnd"
+c_b="ns-cb-$rnd"
+
+checktool (){
+	if ! $1 > /dev/null 2>&1; then
+		echo "SKIP: Could not $2"
+		exit $ksft_skip
+	fi
+}
+
+checktool "iptables --version" "run test without iptables"
+checktool "ip -Version" "run test without ip tool"
+checktool "which nc" "run test without nc (netcat)"
+checktool "ip netns add ${r_a}" "create net namespace"
+
+for n in ${r_b} ${r_w} ${c_a} ${c_b};do
+	ip netns add ${n}
+done
+
+cleanup() {
+	for n in ${r_a} ${r_b} ${r_w} ${c_a} ${c_b};do
+		ip netns del ${n}
+	done
+	rm -f ${rx}
+}
+
+trap cleanup EXIT
+
+test_path() {
+	msg="$1"
+
+	ip netns exec ${c_b} nc -n -w 3 -q 3 -u -l -p 5000 > ${rx} < /dev/null &
+
+	sleep 1
+	for i in 1 2 3; do
+		head -c1400 /dev/zero | tr "\000" "a" | ip netns exec ${c_a} nc -n -w 1 -u 192.168.20.2 5000
+	done
+
+	wait
+
+	bytes=$(wc -c < ${rx})
+
+	if [ $bytes -eq 1400 ];then
+		echo "OK: PMTU $msg connection tracking"
+	else
+		echo "FAIL: PMTU $msg connection tracking: got $bytes, expected 1400"
+		exit 1
+	fi
+}
+
+# Detailed setup for Router A
+# ---------------------------
+# Interfaces:
+# eth0: 10.2.2.1/24
+# eth1: 192.168.10.1/24
+# ipip0: No IP address, local 10.2.2.1 remote 10.4.4.1
+# Routes:
+# 192.168.20.0/24 dev ipip0    (192.168.20.0/24 is subnet of Client B)
+# 10.4.4.1 via 10.2.2.254      (Router B via Wanrouter)
+# No iptables rules at all.
+
+ip link add veth0 netns ${r_a} type veth peer name veth0 netns ${r_w}
+ip link add veth1 netns ${r_a} type veth peer name veth0 netns ${c_a}
+
+l_addr="10.2.2.1"
+r_addr="10.4.4.1"
+ip netns exec ${r_a} ip link add ipip0 type ipip local ${l_addr} remote ${r_addr} mode ipip || exit $ksft_skip
+
+for dev in lo veth0 veth1 ipip0; do
+    ip -net ${r_a} link set $dev up
+done
+
+ip -net ${r_a} addr add 10.2.2.1/24 dev veth0
+ip -net ${r_a} addr add 192.168.10.1/24 dev veth1
+
+ip -net ${r_a} route add 192.168.20.0/24 dev ipip0
+ip -net ${r_a} route add 10.4.4.0/24 via 10.2.2.254
+
+ip netns exec ${r_a} sysctl -q net.ipv4.conf.all.forwarding=1 > /dev/null
+
+# Detailed setup for Router B
+# ---------------------------
+# Interfaces:
+# eth0: 10.4.4.1/24
+# eth1: 192.168.20.1/24
+# ipip0: No IP address, local 10.4.4.1 remote 10.2.2.1
+# Routes:
+# 192.168.10.0/24 dev ipip0    (192.168.10.0/24 is subnet of Client A)
+# 10.2.2.1 via 10.4.4.254      (Router A via Wanrouter)
+# No iptables rules at all.
+
+ip link add veth0 netns ${r_b} type veth peer name veth1 netns ${r_w}
+ip link add veth1 netns ${r_b} type veth peer name veth0 netns ${c_b}
+
+l_addr="10.4.4.1"
+r_addr="10.2.2.1"
+
+ip netns exec ${r_b} ip link add ipip0 type ipip local ${l_addr} remote ${r_addr} mode ipip || exit $ksft_skip
+
+for dev in lo veth0 veth1 ipip0; do
+	ip -net ${r_b} link set $dev up
+done
+
+ip -net ${r_b} addr add 10.4.4.1/24 dev veth0
+ip -net ${r_b} addr add 192.168.20.1/24 dev veth1
+
+ip -net ${r_b} route add 192.168.10.0/24 dev ipip0
+ip -net ${r_b} route add 10.2.2.0/24 via 10.4.4.254
+ip netns exec ${r_b} sysctl -q net.ipv4.conf.all.forwarding=1 > /dev/null
+
+# Client A
+ip -net ${c_a} addr add 192.168.10.2/24 dev veth0
+ip -net ${c_a} link set dev lo up
+ip -net ${c_a} link set dev veth0 up
+ip -net ${c_a} route add default via 192.168.10.1
+
+# Client A
+ip -net ${c_b} addr add 192.168.20.2/24 dev veth0
+ip -net ${c_b} link set dev veth0 up
+ip -net ${c_b} link set dev lo up
+ip -net ${c_b} route add default via 192.168.20.1
+
+# Wan
+ip -net ${r_w} addr add 10.2.2.254/24 dev veth0
+ip -net ${r_w} addr add 10.4.4.254/24 dev veth1
+
+ip -net ${r_w} link set dev lo up
+ip -net ${r_w} link set dev veth0 up mtu 1400
+ip -net ${r_w} link set dev veth1 up mtu 1400
+
+ip -net ${r_a} link set dev veth0 mtu 1400
+ip -net ${r_b} link set dev veth0 mtu 1400
+
+ip netns exec ${r_w} sysctl -q net.ipv4.conf.all.forwarding=1 > /dev/null
+
+# Path MTU discovery
+# ------------------
+# Running tracepath from Client A to Client B shows PMTU discovery is working
+# as expected:
+#
+# clienta:~# tracepath 192.168.20.2
+# 1?: [LOCALHOST]                      pmtu 1500
+# 1:  192.168.10.1                                          0.867ms
+# 1:  192.168.10.1                                          0.302ms
+# 2:  192.168.10.1                                          0.312ms pmtu 1480
+# 2:  no reply
+# 3:  192.168.10.1                                          0.510ms pmtu 1380
+# 3:  192.168.20.2                                          2.320ms reached
+# Resume: pmtu 1380 hops 3 back 3
+
+# ip netns exec ${c_a} traceroute --mtu 192.168.20.2
+
+# Router A has learned PMTU (1400) to Router B from Wanrouter.
+# Client A has learned PMTU (1400 - IPIP overhead = 1380) to Client B
+# from Router A.
+
+#Send large UDP packet
+#---------------------
+#Now we send a 1400 bytes UDP packet from Client A to Client B:
+
+# clienta:~# head -c1400 /dev/zero | tr "\000" "a" | nc -u 192.168.20.2 5000
+test_path "without"
+
+# The IPv4 stack on Client A already knows the PMTU to Client B, so the
+# UDP packet is sent as two fragments (1380 + 20). Router A forwards the
+# fragments between eth1 and ipip0. The fragments fit into the tunnel and
+# reach their destination.
+
+#When sending the large UDP packet again, Router A now reassembles the
+#fragments before routing the packet over ipip0. The resulting IPIP
+#packet is too big (1400) for the tunnel PMTU (1380) to Router B, it is
+#dropped on Router A before sending.
+
+ip netns exec ${r_a} iptables -A FORWARD -m conntrack --ctstate NEW
+test_path "with"
diff --git a/tools/testing/selftests/netfilter/nft_conntrack_helper.sh b/tools/testing/selftests/netfilter/nft_conntrack_helper.sh
index edf0a48da6bf..bf6b9626c7dd 100755
--- a/tools/testing/selftests/netfilter/nft_conntrack_helper.sh
+++ b/tools/testing/selftests/netfilter/nft_conntrack_helper.sh
@@ -94,7 +94,13 @@ check_for_helper()
 	local message=$2
 	local port=$3
 
-	ip netns exec ${netns} conntrack -L -p tcp --dport $port 2> /dev/null |grep -q 'helper=ftp'
+	if echo $message |grep -q 'ipv6';then
+		local family="ipv6"
+	else
+		local family="ipv4"
+	fi
+
+	ip netns exec ${netns} conntrack -L -f $family -p tcp --dport $port 2> /dev/null |grep -q 'helper=ftp'
 	if [ $? -ne 0 ] ; then
 		echo "FAIL: ${netns} did not show attached helper $message" 1>&2
 		ret=1
@@ -111,8 +117,8 @@ test_helper()
 
 	sleep 3 | ip netns exec ${ns2} nc -w 2 -l -p $port > /dev/null &
 
-	sleep 1
 	sleep 1 | ip netns exec ${ns1} nc -w 2 10.0.1.2 $port > /dev/null &
+	sleep 1
 
 	check_for_helper "$ns1" "ip $msg" $port
 	check_for_helper "$ns2" "ip $msg" $port
@@ -128,8 +134,8 @@ test_helper()
 
 	sleep 3 | ip netns exec ${ns2} nc -w 2 -6 -l -p $port > /dev/null &
 
-	sleep 1
 	sleep 1 | ip netns exec ${ns1} nc -w 2 -6 dead:1::2 $port > /dev/null &
+	sleep 1
 
 	check_for_helper "$ns1" "ipv6 $msg" $port
 	check_for_helper "$ns2" "ipv6 $msg" $port
diff --git a/tools/testing/selftests/vDSO/.gitignore b/tools/testing/selftests/vDSO/.gitignore
index 5eb64d41e541..a8dc51af5a9c 100644
--- a/tools/testing/selftests/vDSO/.gitignore
+++ b/tools/testing/selftests/vDSO/.gitignore
@@ -1,5 +1,8 @@
 # SPDX-License-Identifier: GPL-2.0-only
 vdso_test
+vdso_test_abi
+vdso_test_clock_getres
+vdso_test_correctness
 vdso_test_gettimeofday
 vdso_test_getcpu
 vdso_standalone_test_x86
diff --git a/tools/testing/selftests/vDSO/vdso_test_correctness.c b/tools/testing/selftests/vDSO/vdso_test_correctness.c
index 5029ef9b228c..c4aea794725a 100644
--- a/tools/testing/selftests/vDSO/vdso_test_correctness.c
+++ b/tools/testing/selftests/vDSO/vdso_test_correctness.c
@@ -349,7 +349,7 @@ static void test_one_clock_gettime64(int clock, const char *name)
 		return;
 	}
 
-	printf("\t%llu.%09ld %llu.%09ld %llu.%09ld\n",
+	printf("\t%llu.%09lld %llu.%09lld %llu.%09lld\n",
 	       (unsigned long long)start.tv_sec, start.tv_nsec,
 	       (unsigned long long)vdso.tv_sec, vdso.tv_nsec,
 	       (unsigned long long)end.tv_sec, end.tv_nsec);
diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile
index 9a25307f6115..d42115e4284d 100644
--- a/tools/testing/selftests/vm/Makefile
+++ b/tools/testing/selftests/vm/Makefile
@@ -4,7 +4,7 @@
 include local_config.mk
 
 uname_M := $(shell uname -m 2>/dev/null || echo not)
-MACHINE ?= $(shell echo $(uname_M) | sed -e 's/aarch64.*/arm64/')
+MACHINE ?= $(shell echo $(uname_M) | sed -e 's/aarch64.*/arm64/' -e 's/ppc64.*/ppc64/')
 
 # Without this, failed build products remain, with up-to-date timestamps,
 # thus tricking Make (and you!) into believing that All Is Well, in subsequent
@@ -43,7 +43,7 @@ TEST_GEN_FILES += thuge-gen
 TEST_GEN_FILES += transhuge-stress
 TEST_GEN_FILES += userfaultfd
 
-ifeq ($(ARCH),x86_64)
+ifeq ($(MACHINE),x86_64)
 CAN_BUILD_I386 := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_32bit_program.c -m32)
 CAN_BUILD_X86_64 := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_64bit_program.c)
 CAN_BUILD_WITH_NOPIE := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_program.c -no-pie)
@@ -65,13 +65,13 @@ TEST_GEN_FILES += $(BINARIES_64)
 endif
 else
 
-ifneq (,$(findstring $(ARCH),powerpc))
+ifneq (,$(findstring $(MACHINE),ppc64))
 TEST_GEN_FILES += protection_keys
 endif
 
 endif
 
-ifneq (,$(filter $(MACHINE),arm64 ia64 mips64 parisc64 ppc64 ppc64le riscv64 s390x sh64 sparc64 x86_64))
+ifneq (,$(filter $(MACHINE),arm64 ia64 mips64 parisc64 ppc64 riscv64 s390x sh64 sparc64 x86_64))
 TEST_GEN_FILES += va_128TBswitch
 TEST_GEN_FILES += virtual_address_range
 TEST_GEN_FILES += write_to_hugetlbfs
@@ -84,7 +84,7 @@ TEST_FILES := test_vmalloc.sh
 KSFT_KHDR_INSTALL := 1
 include ../lib.mk
 
-ifeq ($(ARCH),x86_64)
+ifeq ($(MACHINE),x86_64)
 BINARIES_32 := $(patsubst %,$(OUTPUT)/%,$(BINARIES_32))
 BINARIES_64 := $(patsubst %,$(OUTPUT)/%,$(BINARIES_64))
 
diff --git a/tools/testing/selftests/wireguard/qemu/debug.config b/tools/testing/selftests/wireguard/qemu/debug.config
index b50c2085c1ac..fe07d97df9fa 100644
--- a/tools/testing/selftests/wireguard/qemu/debug.config
+++ b/tools/testing/selftests/wireguard/qemu/debug.config
@@ -1,5 +1,4 @@
 CONFIG_LOCALVERSION="-debug"
-CONFIG_ENABLE_MUST_CHECK=y
 CONFIG_FRAME_POINTER=y
 CONFIG_STACK_VALIDATION=y
 CONFIG_DEBUG_KERNEL=y