Merge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull perf updates from Ingo Molnar:
 "A number of fixes:

   - Fix segfault on perf trace -i perf.data, from Namhyung Kim.

   - Fix segfault with --no-mmap-pages, from David Ahern.

   - Don't force a refresh during progress update in the TUI, greatly
     reducing startup costs, fix from Patrick Palka.

   - Fix sw clock event period test wrt not checking if using >
     max_sample_freq.

   - Handle throttle events in 'object code reading' test, fix from
     Adrian Hunter.

   - Prevent condition that all sort keys are elided, fix from Namhyung
     Kim.

   - Round mmap pages to power 2, from David Ahern.

  And a number of late arrival changes:

   - Add summary only option to 'perf trace', suppressing the decoding
     of events, from David Ahern

   - 'perf trace --summary' formatting simplifications, from Pekka
     Enberg.

   - Beautify fifth argument of mmap() as fd, in 'perf trace', from
     Namhyung Kim.

   - Add direct access to dynamic arrays in libtraceevent, from Steven
     Rostedt.

   - Synthesize non-exec MMAP records when --data used, allowing the
     resolution of data addresses to symbols (global variables, etc), by
     Arnaldo Carvalho de Melo.

   - Code cleanups by David Ahern and Adrian Hunter"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (23 commits)
  tools lib traceevent: Add direct access to dynamic arrays
  perf target: Shorten perf_target__ to target__
  perf tests: Handle throttle events in 'object code reading' test
  perf evlist: Refactor mmap_pages parsing
  perf evlist: Round mmap pages to power 2 - v2
  perf record: Fix segfault with --no-mmap-pages
  perf trace: Add summary only option
  perf trace: Simplify '--summary' output
  perf trace: Change syscall summary duration order
  perf tests: Compensate lower sample freq with longer test loop
  perf trace: Fix segfault on perf trace -i perf.data
  perf trace: Separate tp syscall field caching into init routine to be reused
  perf trace: Beautify fifth argument of mmap() as fd
  perf tests: Use lower sample_freq in sw clock event period test
  perf tests: Check return of perf_evlist__open sw clock event period test
  perf record: Move existing write_output into helper function
  perf record: Use correct return type for write()
  perf tools: Prevent condition that all sort keys are elided
  perf machine: Simplify synthesize_threads method
  perf machine: Introduce synthesize_threads method out of open coded equivalent
  ...

31 files changed, 363 insertions, 264 deletions

diff --git a/tools/lib/traceevent/event-parse.c b/tools/lib/traceevent/event-parse.c
index 8f450adaa9c2..0362d575de7d 100644
--- a/tools/lib/traceevent/event-parse.c
+++ b/tools/lib/traceevent/event-parse.c
@@ -3435,6 +3435,19 @@ eval_num_arg(void *data, int size, struct event_format *event, struct print_arg
                         goto out_warning_op;
                 }
                 break;
+        case PRINT_DYNAMIC_ARRAY:
+                /* Without [], we pass the address to the dynamic data */
+                offset = pevent_read_number(pevent,
+                                            data + arg->dynarray.field->offset,
+                                            arg->dynarray.field->size);
+                /*
+                 * The actual length of the dynamic array is stored
+                 * in the top half of the field, and the offset
+                 * is in the bottom half of the 32 bit field.
+                 */
+                offset &= 0xffff;
+                val = (unsigned long long)(data + offset);
+                break;
         default: /* not sure what to do there */
                 return 0;
         }
diff --git a/tools/perf/Documentation/perf-trace.txt b/tools/perf/Documentation/perf-trace.txt
index 7b0497f95a75..fae38d9a44a4 100644
--- a/tools/perf/Documentation/perf-trace.txt
+++ b/tools/perf/Documentation/perf-trace.txt
@@ -93,9 +93,15 @@ the thread executes on the designated CPUs. Default is to monitor all CPUs.
 --comm::
         Show process COMM right beside its ID, on by default, disable with --no-comm.
 
+-s::
 --summary::
-        Show a summary of syscalls by thread with min, max, and average times (in
-    msec) and relative stddev.
+        Show only a summary of syscalls by thread with min, max, and average times
+    (in msec) and relative stddev.
+
+-S::
+--with-summary::
+        Show all syscalls followed by a summary by thread with min, max, and
+    average times (in msec) and relative stddev.
 
 --tool_stats::
         Show tool stats such as number of times fd->pathname was discovered thru
diff --git a/tools/perf/builtin-kvm.c b/tools/perf/builtin-kvm.c
index cd9f92078aba..f8bf5f244d77 100644
--- a/tools/perf/builtin-kvm.c
+++ b/tools/perf/builtin-kvm.c
@@ -1510,13 +1510,13 @@ static int kvm_events_live(struct perf_kvm_stat *kvm,
         /*
          * target related setups
          */
-        err = perf_target__validate(&kvm->opts.target);
+        err = target__validate(&kvm->opts.target);
         if (err) {
-                perf_target__strerror(&kvm->opts.target, err, errbuf, BUFSIZ);
+                target__strerror(&kvm->opts.target, err, errbuf, BUFSIZ);
                 ui__warning("%s", errbuf);
         }
 
-        if (perf_target__none(&kvm->opts.target))
+        if (target__none(&kvm->opts.target))
                 kvm->opts.target.system_wide = true;
 
 
@@ -1544,18 +1544,8 @@ static int kvm_events_live(struct perf_kvm_stat *kvm,
         }
         kvm->session->evlist = kvm->evlist;
         perf_session__set_id_hdr_size(kvm->session);
-
-
-        if (perf_target__has_task(&kvm->opts.target))
-                perf_event__synthesize_thread_map(&kvm->tool,
-                                                  kvm->evlist->threads,
-                                                  perf_event__process,
-                                                  &kvm->session->machines.host);
-        else
-                perf_event__synthesize_threads(&kvm->tool, perf_event__process,
-                                               &kvm->session->machines.host);
-
-
+        machine__synthesize_threads(&kvm->session->machines.host, &kvm->opts.target,
+                                    kvm->evlist->threads, false);
         err = kvm_live_open_events(kvm);
         if (err)
                 goto out;
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 15280b5e5574..4d644fe2d5b7 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -76,12 +76,12 @@ struct perf_record {
         long                    samples;
 };
 
-static int write_output(struct perf_record *rec, void *buf, size_t size)
+static int do_write_output(struct perf_record *rec, void *buf, size_t size)
 {
         struct perf_data_file *file = &rec->file;
 
         while (size) {
-                int ret = write(file->fd, buf, size);
+                ssize_t ret = write(file->fd, buf, size);
 
                 if (ret < 0) {
                         pr_err("failed to write perf data, error: %m\n");
@@ -97,6 +97,11 @@ static int write_output(struct perf_record *rec, void *buf, size_t size)
         return 0;
 }
 
+static int write_output(struct perf_record *rec, void *buf, size_t size)
+{
+        return do_write_output(rec, buf, size);
+}
+
 static int process_synthesized_event(struct perf_tool *tool,
                                      union perf_event *event,
                                      struct perf_sample *sample __maybe_unused,
@@ -480,16 +485,8 @@ static int __cmd_record(struct perf_record *rec, int argc, const char **argv)
                                          perf_event__synthesize_guest_os, tool);