[PATCH 14/28] perf callchain: Honour the ordering of PERF_CONTEXT_{USER,KERNEL,etc}

[PATCH 14/28] perf callchain: Honour the ordering of PERF_CONTEXT_{USER,KERNEL,etc}

[Arnaldo Carvalho de Melo]

From: "David S. Miller" <davem@davemloft.net>

When processing using 'perf report -g caller', which is the default, we
ended up reverting the callchain entries received from the kernel, but
simply reverting throws away the information that tells that from a
point onwards the addresses are for userspace, kernel, guest kernel,
guest user, hypervisor.

The idea is that if we are walking backwards, for each cluster of
non-cpumode entries we have to first scan backwards for the next one and
use that for the cluster.

This seems silly and more expensive than it needs to be but it is enough
for a initial fix.

The code here is really complicated because it is intimately intertwined
with the lbr and branch handling, as well as this callchain order,
further fixes will be needed to properly take into account the cpumode
in those cases.

Another problem with ORDER_CALLER is that the NULL "0" IP that is at the
end of most callchains shows up at the top of the histogram because
every callchain contains it and with ORDER_CALLER it is the first entry.

Signed-off-by: David S. Miller <davem@davemloft.net>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Souvik Banerjee <souvik1997@gmail.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: stable@vger.kernel.org # 4.19
Link: https://lkml.kernel.org/n/tip-2wt3ayp6j2y2f2xowixa8y6y@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/machine.c | 35 ++++++++++++++++++++++++++++++++++-
 1 file changed, 34 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
index 111ae858cbcb..8ee8ab39d8ac 100644
--- a/tools/perf/util/machine.c
+++ b/tools/perf/util/machine.c
@@ -2140,6 +2140,27 @@ static int resolve_lbr_callchain_sample(struct thread *thread,
 	return 0;
 }
 
+static int find_prev_cpumode(struct ip_callchain *chain, struct thread *thread,
+			     struct callchain_cursor *cursor,
+			     struct symbol **parent,
+			     struct addr_location *root_al,
+			     u8 *cpumode, int ent)
+{
+	int err = 0;
+
+	while (--ent >= 0) {
+		u64 ip = chain->ips[ent];
+
+		if (ip >= PERF_CONTEXT_MAX) {
+			err = add_callchain_ip(thread, cursor, parent,
+					       root_al, cpumode, ip,
+					       false, NULL, NULL, 0);
+			break;
+		}
+	}
+	return err;
+}
+
 static int thread__resolve_callchain_sample(struct thread *thread,
 					    struct callchain_cursor *cursor,
 					    struct perf_evsel *evsel,
@@ -2246,6 +2267,12 @@ static int thread__resolve_callchain_sample(struct thread *thread,
 	}
 
 check_calls:
+	if (callchain_param.order != ORDER_CALLEE) {
+		err = find_prev_cpumode(chain, thread, cursor, parent, root_al,
+					&cpumode, chain->nr - first_call);
+		if (err)
+			return (err < 0) ? err : 0;
+	}
 	for (i = first_call, nr_entries = 0;
 	     i < chain_nr && nr_entries < max_stack; i++) {
 		u64 ip;
@@ -2260,9 +2287,15 @@ static int thread__resolve_callchain_sample(struct thread *thread,
 			continue;
 #endif
 		ip = chain->ips[j];
-
 		if (ip < PERF_CONTEXT_MAX)
                        ++nr_entries;
+		else if (callchain_param.order != ORDER_CALLEE) {
+			err = find_prev_cpumode(chain, thread, cursor, parent,
+						root_al, &cpumode, j);
+			if (err)
+				return (err < 0) ? err : 0;
+			continue;
+		}
 
 		err = add_callchain_ip(thread, cursor, parent,
 				       root_al, &cpumode, ip,
-- 
2.14.4

[PATCH 27/28] perf intel-pt: Insert callchain context into synthesized callchains

[Arnaldo Carvalho de Melo]

From: Adrian Hunter <adrian.hunter@intel.com>

In the absence of a fallback, callchains must encode also the callchain
context. Do that now there is no fallback.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: stable@vger.kernel.org # 4.19
Link: http://lkml.kernel.org/r/100ea2ec-ed14-b56d-d810-e0a6d2f4b069@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/intel-pt.c     |  6 ++++--
 tools/perf/util/thread-stack.c | 44 +++++++++++++++++++++++++++++++++---------
 tools/perf/util/thread-stack.h |  2 +-
 3 files changed, 40 insertions(+), 12 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index ffa385a029b3..60732213d16a 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -759,7 +759,8 @@ static struct intel_pt_queue *intel_pt_alloc_queue(struct intel_pt *pt,
 	if (pt->synth_opts.callchain) {
 		size_t sz = sizeof(struct ip_callchain);
 
-		sz += pt->synth_opts.callchain_sz * sizeof(u64);
+		/* Add 1 to callchain_sz for callchain context */
+		sz += (pt->synth_opts.callchain_sz + 1) * sizeof(u64);
 		ptq->chain = zalloc(sz);
 		if (!ptq->chain)
 			goto out_free;
@@ -1160,7 +1161,8 @@ static void intel_pt_prep_sample(struct intel_pt *pt,
 
 	if (pt->synth_opts.callchain) {
 		thread_stack__sample(ptq->thread, ptq->chain,
-				     pt->synth_opts.callchain_sz, sample->ip);
+				     pt->synth_opts.callchain_sz + 1,
+				     sample->ip, pt->kernel_start);
 		sample->callchain = ptq->chain;
 	}
 
diff --git a/tools/perf/util/thread-stack.c b/tools/perf/util/thread-stack.c
index c091635bf7dc..61a4286a74dc 100644
--- a/tools/perf/util/thread-stack.c
+++ b/tools/perf/util/thread-stack.c
@@ -310,20 +310,46 @@ void thread_stack__free(struct thread *thread)
 	}
 }
 
+static inline u64 callchain_context(u64 ip, u64 kernel_start)
+{
+	return ip < kernel_start ? PERF_CONTEXT_USER : PERF_CONTEXT_KERNEL;
+}
+
 void thread_stack__sample(struct thread *thread, struct ip_callchain *chain,
-			  size_t sz, u64 ip)
+			  size_t sz, u64 ip, u64 kernel_start)
 {
-	size_t i;
+	u64 context = callchain_context(ip, kernel_start);
+	u64 last_context;
+	size_t i, j;
 
-	if (!thread || !thread->ts)
-		chain->nr = 1;
-	else
-		chain->nr = min(sz, thread->ts->cnt + 1);
+	if (sz < 2) {
+		chain->nr = 0;
+		return;
+	}
 
-	chain->ips[0] = ip;
+	chain->ips[0] = context;
+	chain->ips[1] = ip;
+
+	if (!thread || !thread->ts) {
+		chain->nr = 2;
+		return;
+	}
+
+	last_context = context;
+
+	for (i = 2, j = 1; i < sz && j <= thread->ts->cnt; i++, j++) {
+		ip = thread->ts->stack[thread->ts->cnt - j].ret_addr;
+		context = callchain_context(ip, kernel_start);
+		if (context != last_context) {
+			if (i >= sz - 1)
+				break;
+			chain->ips[i++] = context;
+			last_context = context;
+		}
+		chain->ips[i] = ip;
+	}
 
-	for (i = 1; i < chain->nr; i++)
-		chain->ips[i] = thread->ts->stack[thread->ts->cnt - i].ret_addr;
+	chain->nr = i;
 }
 
 struct call_return_processor *
diff --git a/tools/perf/util/thread-stack.h b/tools/perf/util/thread-stack.h
index b7e41c4ebfdd..f97c00a8c251 100644
--- a/tools/perf/util/thread-stack.h
+++ b/tools/perf/util/thread-stack.h
@@ -84,7 +84,7 @@ int thread_stack__event(struct thread *thread, u32 flags, u64 from_ip,
 			u64 to_ip, u16 insn_len, u64 trace_nr);
 void thread_stack__set_trace_nr(struct thread *thread, u64 trace_nr);
 void thread_stack__sample(struct thread *thread, struct ip_callchain *chain,
-			  size_t sz, u64 ip);
+			  size_t sz, u64 ip, u64 kernel_start);
 int thread_stack__flush(struct thread *thread);
 void thread_stack__free(struct thread *thread);
 size_t thread_stack__depth(struct thread *thread);
-- 
2.14.4

[PATCH 28/28] perf intel-pt/bts: Calculate cpumode for synthesized samples

[Arnaldo Carvalho de Melo]

From: Adrian Hunter <adrian.hunter@intel.com>

In the absence of a fallback, samples must provide a correct cpumode for
the 'ip'. Do that now there is no fallback.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Reviewed-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: stable@vger.kernel.org # 4.19
Link: http://lkml.kernel.org/r/20181031091043.23465-6-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/intel-bts.c | 17 ++++++++++++-----
 tools/perf/util/intel-pt.c  | 22 +++++++++++++---------
 2 files changed, 25 insertions(+), 14 deletions(-)

diff --git a/tools/perf/util/intel-bts.c b/tools/perf/util/intel-bts.c
index 3b3a3d55dca1..7b27d77306c2 100644
--- a/tools/perf/util/intel-bts.c
+++ b/tools/perf/util/intel-bts.c
@@ -269,6 +269,13 @@ static int intel_bts_do_fix_overlap(struct auxtrace_queue *queue,
 	return 0;
 }
 
+static inline u8 intel_bts_cpumode(struct intel_bts *bts, uint64_t ip)
+{
+	return machine__kernel_ip(bts->machine, ip) ?
+	       PERF_RECORD_MISC_KERNEL :
+	       PERF_RECORD_MISC_USER;
+}
+
 static int intel_bts_synth_branch_sample(struct intel_bts_queue *btsq,
 					 struct branch *branch)
 {
@@ -281,12 +288,8 @@ static int intel_bts_synth_branch_sample(struct intel_bts_queue *btsq,
 	    bts->num_events++ <= bts->synth_opts.initial_skip)
 		return 0;
 
-	event.sample.header.type = PERF_RECORD_SAMPLE;
-	event.sample.header.misc = PERF_RECORD_MISC_USER;
-	event.sample.header.size = sizeof(struct perf_event_header);
-
-	sample.cpumode = PERF_RECORD_MISC_USER;
 	sample.ip = le64_to_cpu(branch->from);
+	sample.cpumode = intel_bts_cpumode(bts, sample.ip);
 	sample.pid = btsq->pid;
 	sample.tid = btsq->tid;
 	sample.addr = le64_to_cpu(branch->to);
@@ -298,6 +301,10 @@ static int intel_bts_synth_branch_sample(struct intel_bts_queue *btsq,
 	sample.insn_len = btsq->intel_pt_insn.length;
 	memcpy(sample.insn, btsq->intel_pt_insn.buf, INTEL_PT_INSN_BUF_SZ);
 
+	event.sample.header.type = PERF_RECORD_SAMPLE;
+	event.sample.header.misc = sample.cpumode;
+	event.sample.header.size = sizeof(struct perf_event_header);
+
 	if (bts->synth_opts.inject) {
 		event.sample.header.size = bts->branches_event_size;
 		ret = perf_event__synthesize_sample(&event,
diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 60732213d16a..86cc9a64e982 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -407,6 +407,13 @@ intel_pt_cache_lookup(struct dso *dso, struct machine *machine, u64 offset)
 	return auxtrace_cache__lookup(dso->auxtrace_cache, offset);
 }
 
+static inline u8 intel_pt_cpumode(struct intel_pt *pt, uint64_t ip)
+{
+	return ip >= pt->kernel_start ?
+	       PERF_RECORD_MISC_KERNEL :
+	       PERF_RECORD_MISC_USER;
+}
+
 static int intel_pt_walk_next_insn(struct intel_pt_insn *intel_pt_insn,
 				   uint64_t *insn_cnt_ptr, uint64_t *ip,
 				   uint64_t to_ip, uint64_t max_insn_cnt,
@@ -429,10 +436,7 @@ static int intel_pt_walk_next_insn(struct intel_pt_insn *intel_pt_insn,
 	if (to_ip && *ip == to_ip)
 		goto out_no_cache;
 
-	if (*ip >= ptq->pt->kernel_start)
-		cpumode = PERF_RECORD_MISC_KERNEL;
-	else
-		cpumode = PERF_RECORD_MISC_USER;
+	cpumode = intel_pt_cpumode(ptq->pt, *ip);
 
 	thread = ptq->thread;
 	if (!thread) {
@@ -1059,15 +1063,11 @@ static void intel_pt_prep_b_sample(struct intel_pt *pt,
 				   union perf_event *event,
 				   struct perf_sample *sample)
 {
-	event->sample.header.type = PERF_RECORD_SAMPLE;
-	event->sample.header.misc = PERF_RECORD_MISC_USER;
-	event->sample.header.size = sizeof(struct perf_event_header);
-
 	if (!pt->timeless_decoding)
 		sample->time = tsc_to_perf_time(ptq->timestamp, &pt->tc);
 
-	sample->cpumode = PERF_RECORD_MISC_USER;
 	sample->ip = ptq->state->from_ip;
+	sample->cpumode = intel_pt_cpumode(pt, sample->ip);
 	sample->pid = ptq->pid;
 	sample->tid = ptq->tid;
 	sample->addr = ptq->state->to_ip;
@@ -1076,6 +1076,10 @@ static void intel_pt_prep_b_sample(struct intel_pt *pt,
 	sample->flags = ptq->flags;
 	sample->insn_len = ptq->insn_len;
 	memcpy(sample->insn, ptq->insn, INTEL_PT_INSN_BUF_SZ);
+
+	event->sample.header.type = PERF_RECORD_SAMPLE;
+	event->sample.header.misc = sample->cpumode;
+	event->sample.header.size = sizeof(struct perf_event_header);
 }
 
 static int intel_pt_inject_event(union perf_event *event,
-- 
2.14.4