1 files changed, 373 insertions, 111 deletions

diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 92a1bd695e8a..c51049087e4e 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -23,6 +23,7 @@
 #include "strlist.h"
 #include "strbuf.h"
 #include "mem-events.h"
+#include "mem-info.h"
 #include "annotate.h"
 #include "annotate-data.h"
 #include "event.h"
@@ -34,7 +35,7 @@
 #include <linux/string.h>
 
 #ifdef HAVE_LIBTRACEEVENT
-#include <traceevent/event-parse.h>
+#include <event-parse.h>
 #endif
 
 regex_t		parent_regex;
@@ -239,11 +240,11 @@ static int64_t _sort__dso_cmp(struct map *map_l, struct map *map_r)
 		return cmp_null(dso_r, dso_l);
 
 	if (verbose > 0) {
-		dso_name_l = dso_l->long_name;
-		dso_name_r = dso_r->long_name;
+		dso_name_l = dso__long_name(dso_l);
+		dso_name_r = dso__long_name(dso_r);
 	} else {
-		dso_name_l = dso_l->short_name;
-		dso_name_r = dso_r->short_name;
+		dso_name_l = dso__short_name(dso_l);
+		dso_name_r = dso__short_name(dso_r);
 	}
 
 	return strcmp(dso_name_l, dso_name_r);
@@ -262,7 +263,7 @@ static int _hist_entry__dso_snprintf(struct map *map, char *bf,
 	const char *dso_name = "[unknown]";
 
 	if (dso)
-		dso_name = verbose > 0 ? dso->long_name : dso->short_name;
+		dso_name = verbose > 0 ? dso__long_name(dso) : dso__short_name(dso);
 
 	return repsep_snprintf(bf, size, "%-*.*s", width, width, dso_name);
 }
@@ -333,7 +334,7 @@ sort__sym_cmp(struct hist_entry *left, struct hist_entry *right)
 	 * comparing symbol address alone is not enough since it's a
 	 * relative address within a dso.
 	 */
-	if (!hists__has(left->hists, dso) || hists__has(right->hists, dso)) {
+	if (!hists__has(left->hists, dso)) {
 		ret = sort__dso_cmp(left, right);
 		if (ret != 0)
 			return ret;
@@ -364,7 +365,7 @@ static int _hist_entry__sym_snprintf(struct map_symbol *ms,
 		char o = dso ? dso__symtab_origin(dso) : '!';
 		u64 rip = ip;
 
-		if (dso && dso->kernel && dso->adjust_symbols)
+		if (dso && dso__kernel(dso) && dso__adjust_symbols(dso))
 			rip = map__unmap_ip(map, ip);
 
 		ret += repsep_snprintf(bf, size, "%-#*llx %c ",
@@ -676,6 +677,102 @@ struct sort_entry sort_sym_ipc_null = {
 	.se_width_idx	= HISTC_SYMBOL_IPC,
 };
 
+/* --sort callchain_branch_predicted */
+
+static int64_t
+sort__callchain_branch_predicted_cmp(struct hist_entry *left __maybe_unused,
+				     struct hist_entry *right __maybe_unused)
+{
+	return 0;
+}
+
+static int hist_entry__callchain_branch_predicted_snprintf(
+	struct hist_entry *he, char *bf, size_t size, unsigned int width)
+{
+	u64 branch_count, predicted_count;
+	double percent = 0.0;
+	char str[32];
+
+	callchain_branch_counts(he->callchain, &branch_count,
+				&predicted_count, NULL, NULL);
+
+	if (branch_count)
+		percent = predicted_count * 100.0 / branch_count;
+
+	snprintf(str, sizeof(str), "%.1f%%", percent);
+	return repsep_snprintf(bf, size, "%-*.*s", width, width, str);
+}
+
+struct sort_entry sort_callchain_branch_predicted = {
+	.se_header	= "Predicted",
+	.se_cmp		= sort__callchain_branch_predicted_cmp,
+	.se_snprintf	= hist_entry__callchain_branch_predicted_snprintf,
+	.se_width_idx	= HISTC_CALLCHAIN_BRANCH_PREDICTED,
+};
+
+/* --sort callchain_branch_abort */
+
+static int64_t
+sort__callchain_branch_abort_cmp(struct hist_entry *left __maybe_unused,
+				 struct hist_entry *right __maybe_unused)
+{
+	return 0;
+}
+
+static int hist_entry__callchain_branch_abort_snprintf(struct hist_entry *he,
+						       char *bf, size_t size,
+						       unsigned int width)
+{
+	u64 branch_count, abort_count;
+	char str[32];
+
+	callchain_branch_counts(he->callchain, &branch_count,
+				NULL, &abort_count, NULL);
+
+	snprintf(str, sizeof(str), "%" PRId64, abort_count);
+	return repsep_snprintf(bf, size, "%-*.*s", width, width, str);
+}
+
+struct sort_entry sort_callchain_branch_abort = {
+	.se_header	= "Abort",
+	.se_cmp		= sort__callchain_branch_abort_cmp,
+	.se_snprintf	= hist_entry__callchain_branch_abort_snprintf,
+	.se_width_idx	= HISTC_CALLCHAIN_BRANCH_ABORT,
+};
+
+/* --sort callchain_branch_cycles */
+
+static int64_t
+sort__callchain_branch_cycles_cmp(struct hist_entry *left __maybe_unused,
+				  struct hist_entry *right __maybe_unused)
+{
+	return 0;
+}
+
+static int hist_entry__callchain_branch_cycles_snprintf(struct hist_entry *he,
+							char *bf, size_t size,
+							unsigned int width)
+{
+	u64 branch_count, cycles_count, cycles = 0;
+	char str[32];
+
+	callchain_branch_counts(he->callchain, &branch_count,
+				NULL, NULL, &cycles_count);
+
+	if (branch_count)
+		cycles = cycles_count / branch_count;
+
+	snprintf(str, sizeof(str), "%" PRId64 "", cycles);
+	return repsep_snprintf(bf, size, "%-*.*s", width, width, str);
+}
+
+struct sort_entry sort_callchain_branch_cycles = {
+	.se_header	= "Cycles",
+	.se_cmp		= sort__callchain_branch_cycles_cmp,
+	.se_snprintf	= hist_entry__callchain_branch_cycles_snprintf,
+	.se_width_idx	= HISTC_CALLCHAIN_BRANCH_CYCLES,
+};
+
 /* --sort srcfile */
 
 static char no_srcfile[1];
@@ -795,6 +892,38 @@ struct sort_entry sort_cpu = {
 	.se_width_idx	= HISTC_CPU,
 };
 
+/* --sort parallelism */
+
+static int64_t
+sort__parallelism_cmp(struct hist_entry *left, struct hist_entry *right)
+{
+	return right->parallelism - left->parallelism;
+}
+
+static int hist_entry__parallelism_filter(struct hist_entry *he, int type, const void *arg)
+{
+	const unsigned long *parallelism_filter = arg;
+
+	if (type != HIST_FILTER__PARALLELISM)
+		return -1;
+
+	return test_bit(he->parallelism, parallelism_filter);
+}
+
+static int hist_entry__parallelism_snprintf(struct hist_entry *he, char *bf,
+				    size_t size, unsigned int width)
+{
+	return repsep_snprintf(bf, size, "%*d", width, he->parallelism);
+}
+
+struct sort_entry sort_parallelism = {
+	.se_header      = "Parallelism",
+	.se_cmp	        = sort__parallelism_cmp,
+	.se_filter	= hist_entry__parallelism_filter,
+	.se_snprintf    = hist_entry__parallelism_snprintf,
+	.se_width_idx	= HISTC_PARALLELISM,
+};
+
 /* --sort cgroup_id */
 
 static int64_t _sort__cgroup_dev_cmp(u64 left_dev, u64 right_dev)
@@ -941,17 +1070,19 @@ static char *get_trace_output(struct hist_entry *he)
 		.data = he->raw_data,
 		.size = he->raw_size,
 	};
+	struct tep_event *tp_format;
 
 	evsel = hists_to_evsel(he->hists);
 
 	trace_seq_init(&seq);
-	if (symbol_conf.raw_trace) {
-		tep_print_fields(&seq, he->raw_data, he->raw_size,
-				 evsel->tp_format);
-	} else {
-		tep_print_event(evsel->tp_format->tep,
-				&seq, &rec, "%s", TEP_PRINT_INFO);
+	tp_format = evsel__tp_format(evsel);
+	if (tp_format) {
+		if (symbol_conf.raw_trace)
+			tep_print_fields(&seq, he->raw_data, he->raw_size, tp_format);
+		else
+			tep_print_event(tp_format->tep, &seq, &rec, "%s", TEP_PRINT_INFO);
 	}
+
 	/*
 	 * Trim the buffer, it starts at 4KB and we're not going to
 	 * add anything more to this buffer.
@@ -1364,9 +1495,9 @@ sort__daddr_cmp(struct hist_entry *left, struct hist_entry *right)
 	uint64_t l = 0, r = 0;
 
 	if (left->mem_info)
-		l = left->mem_info->daddr.addr;
+		l = mem_info__daddr(left->mem_info)->addr;
 	if (right->mem_info)
-		r = right->mem_info->daddr.addr;
+		r = mem_info__daddr(right->mem_info)->addr;
 
 	return (int64_t)(r - l);
 }
@@ -1378,8 +1509,8 @@ static int hist_entry__daddr_snprintf(struct hist_entry *he, char *bf,
 	struct map_symbol *ms = NULL;
 
 	if (he->mem_info) {
-		addr = he->mem_info->daddr.addr;
-		ms = &he->mem_info->daddr.ms;
+		addr = mem_info__daddr(he->mem_info)->addr;
+		ms = &mem_info__daddr(he->mem_info)->ms;
 	}
 	return _hist_entry__sym_snprintf(ms, addr, he->level, bf, size, width);
 }
@@ -1390,9 +1521,9 @@ sort__iaddr_cmp(struct hist_entry *left, struct hist_entry *right)
 	uint64_t l = 0, r = 0;
 
 	if (left->mem_info)
-		l = left->mem_info->iaddr.addr;
+		l = mem_info__iaddr(left->mem_info)->addr;
 	if (right->mem_info)
-		r = right->mem_info->iaddr.addr;
+		r = mem_info__iaddr(right->mem_info)->addr;
 
 	return (int64_t)(r - l);
 }
@@ -1404,8 +1535,8 @@ static int hist_entry__iaddr_snprintf(struct hist_entry *he, char *bf,
 	struct map_symbol *ms = NULL;
 
 	if (he->mem_info) {
-		addr = he->mem_info->iaddr.addr;
-		ms   = &he->mem_info->iaddr.ms;
+		addr = mem_info__iaddr(he->mem_info)->addr;
+		ms   = &mem_info__iaddr(he->mem_info)->ms;
 	}
 	return _hist_entry__sym_snprintf(ms, addr, he->level, bf, size, width);
 }
@@ -1417,9 +1548,9 @@ sort__dso_daddr_cmp(struct hist_entry *left, struct hist_entry *right)
 	struct map *map_r = NULL;
 
 	if (left->mem_info)
-		map_l = left->mem_info->daddr.ms.map;
+		map_l = mem_info__daddr(left->mem_info)->ms.map;
 	if (right->mem_info)
-		map_r = right->mem_info->daddr.ms.map;
+		map_r = mem_info__daddr(right->mem_info)->ms.map;
 
 	return _sort__dso_cmp(map_l, map_r);
 }
@@ -1430,7 +1561,7 @@ static int hist_entry__dso_daddr_snprintf(struct hist_entry *he, char *bf,
 	struct map *map = NULL;
 
 	if (he->mem_info)
-		map = he->mem_info->daddr.ms.map;
+		map = mem_info__daddr(he->mem_info)->ms.map;
 
 	return _hist_entry__dso_snprintf(map, bf, size, width);
 }
@@ -1442,12 +1573,12 @@ sort__locked_cmp(struct hist_entry *left, struct hist_entry *right)
 	union perf_mem_data_src data_src_r;
 
 	if (left->mem_info)
-		data_src_l = left->mem_info->data_src;
+		data_src_l = *mem_info__data_src(left->mem_info);
 	else
 		data_src_l.mem_lock = PERF_MEM_LOCK_NA;
 
 	if (right->mem_info)
-		data_src_r = right->mem_info->data_src;
+		data_src_r = *mem_info__data_src(right->mem_info);
 	else
 		data_src_r.mem_lock = PERF_MEM_LOCK_NA;
 
@@ -1470,12 +1601,12 @@ sort__tlb_cmp(struct hist_entry *left, struct hist_entry *right)
 	union perf_mem_data_src data_src_r;
 
 	if (left->mem_info)
-		data_src_l = left->mem_info->data_src;
+		data_src_l = *mem_info__data_src(left->mem_info);
 	else
 		data_src_l.mem_dtlb = PERF_MEM_TLB_NA;
 
 	if (right->mem_info)
-		data_src_r = right->mem_info->data_src;
+		data_src_r = *mem_info__data_src(right->mem_info);
 	else
 		data_src_r.mem_dtlb = PERF_MEM_TLB_NA;
 
@@ -1498,12 +1629,12 @@ sort__lvl_cmp(struct hist_entry *left, struct hist_entry *right)
 	union perf_mem_data_src data_src_r;
 
 	if (left->mem_info)
-		data_src_l = left->mem_info->data_src;
+		data_src_l = *mem_info__data_src(left->mem_info);
 	else
 		data_src_l.mem_lvl = PERF_MEM_LVL_NA;
 
 	if (right->mem_info)
-		data_src_r = right->mem_info->data_src;
+		data_src_r = *mem_info__data_src(right->mem_info);
 	else
 		data_src_r.mem_lvl = PERF_MEM_LVL_NA;
 
@@ -1526,12 +1657,12 @@ sort__snoop_cmp(struct hist_entry *left, struct hist_entry *right)
 	union perf_mem_data_src data_src_r;
 
 	if (left->mem_info)
-		data_src_l = left->mem_info->data_src;
+		data_src_l = *mem_info__data_src(left->mem_info);
 	else
 		data_src_l.mem_snoop = PERF_MEM_SNOOP_NA;
 
 	if (right->mem_info)
-		data_src_r = right->mem_info->data_src;
+		data_src_r = *mem_info__data_src(right->mem_info);
 	else
 		data_src_r.mem_snoop = PERF_MEM_SNOOP_NA;
 
@@ -1562,8 +1693,8 @@ sort__dcacheline_cmp(struct hist_entry *left, struct hist_entry *right)
 	if (left->cpumode > right->cpumode) return -1;
 	if (left->cpumode < right->cpumode) return 1;
 
-	l_map = left->mem_info->daddr.ms.map;
-	r_map = right->mem_info->daddr.ms.map;
+	l_map = mem_info__daddr(left->mem_info)->ms.map;
+	r_map = mem_info__daddr(right->mem_info)->ms.map;
 
 	/* if both are NULL, jump to sort on al_addr instead */
 	if (!l_map && !r_map)
@@ -1586,8 +1717,8 @@ sort__dcacheline_cmp(struct hist_entry *left, struct hist_entry *right)
 	 */
 
 	if ((left->cpumode != PERF_RECORD_MISC_KERNEL) &&
-	    (!(map__flags(l_map) & MAP_SHARED)) && !l_dso->id.maj && !l_dso->id.min &&
-	    !l_dso->id.ino && !l_dso->id.ino_generation) {
+	    (!(map__flags(l_map) & MAP_SHARED)) && !dso__id(l_dso)->maj && !dso__id(l_dso)->min &&
+	     !dso__id(l_dso)->ino && !dso__id(l_dso)->ino_generation) {
 		/* userspace anonymous */
 
 		if (thread__pid(left->thread) > thread__pid(right->thread))
@@ -1598,8 +1729,8 @@ sort__dcacheline_cmp(struct hist_entry *left, struct hist_entry *right)
 
 addr:
 	/* al_addr does all the right addr - start + offset calculations */
-	l = cl_address(left->mem_info->daddr.al_addr, chk_double_cl);
-	r = cl_address(right->mem_info->daddr.al_addr, chk_double_cl);
+	l = cl_address(mem_info__daddr(left->mem_info)->al_addr, chk_double_cl);
+	r = cl_address(mem_info__daddr(right->mem_info)->al_addr, chk_double_cl);
 
 	if (l > r) return -1;
 	if (l < r) return 1;
@@ -1616,17 +1747,18 @@ static int hist_entry__dcacheline_snprintf(struct hist_entry *he, char *bf,
 	char level = he->level;
 
 	if (he->mem_info) {
-		struct map *map = he->mem_info->daddr.ms.map;
+		struct map *map = mem_info__daddr(he->mem_info)->ms.map;
 		struct dso *dso = map ? map__dso(map) : NULL;
 
-		addr = cl_address(he->mem_info->daddr.al_addr, chk_double_cl);
-		ms = &he->mem_info->daddr.ms;
+		addr = cl_address(mem_info__daddr(he->mem_info)->al_addr, chk_double_cl);
+		ms = &mem_info__daddr(he->mem_info)->ms;
 
 		/* print [s] for shared data mmaps */
 		if ((he->cpumode != PERF_RECORD_MISC_KERNEL) &&
 		     map && !(map__prot(map) & PROT_EXEC) &&
 		     (map__flags(map) & MAP_SHARED) &&
-		    (dso->id.maj || dso->id.min || dso->id.ino || dso->id.ino_generation))
+		     (dso__id(dso)->maj || dso__id(dso)->min || dso__id(dso)->ino ||
+		      dso__id(dso)->ino_generation))
 			level = 's';
 		else if (!map)
 			level = 'X';
@@ -1804,12 +1936,12 @@ sort__blocked_cmp(struct hist_entry *left, struct hist_entry *right)
 	union perf_mem_data_src data_src_r;
 
 	if (left->mem_info)
-		data_src_l = left->mem_info->data_src;
+		data_src_l = *mem_info__data_src(left->mem_info);
 	else
 		data_src_l.mem_blk = PERF_MEM_BLK_NA;
 
 	if (right->mem_info)
-		data_src_r = right->mem_info->data_src;
+		data_src_r = *mem_info__data_src(right->mem_info);
 	else
 		data_src_r.mem_blk = PERF_MEM_BLK_NA;
 
@@ -1838,9 +1970,9 @@ sort__phys_daddr_cmp(struct hist_entry *left, struct hist_entry *right)
 	uint64_t l = 0, r = 0;
 
 	if (left->mem_info)
-		l = left->mem_info->daddr.phys_addr;
+		l = mem_info__daddr(left->mem_info)->phys_addr;
 	if (right->mem_info)
-		r = right->mem_info->daddr.phys_addr;
+		r = mem_info__daddr(right->mem_info)->phys_addr;
 
 	return (int64_t)(r - l);
 }
@@ -1852,7 +1984,7 @@ static int hist_entry__phys_daddr_snprintf(struct hist_entry *he, char *bf,
 	size_t ret = 0;
 	size_t len = BITS_PER_LONG / 4;
 
-	addr = he->mem_info->daddr.phys_addr;
+	addr = mem_info__daddr(he->mem_info)->phys_addr;
 
 	ret += repsep_snprintf(bf + ret, size - ret, "[%c] ", he->level);
 
@@ -1879,9 +2011,9 @@ sort__data_page_size_cmp(struct hist_entry *left, struct hist_entry *right)
 	uint64_t l = 0, r = 0;
 
 	if (left->mem_info)
-		l = left->mem_info->daddr.data_page_size;
+		l = mem_info__daddr(left->mem_info)->data_page_size;
 	if (right->mem_info)
-		r = right->mem_info->daddr.data_page_size;
+		r = mem_info__daddr(right->mem_info)->data_page_size;
 
 	return (int64_t)(r - l);
 }
@@ -1892,7 +2024,7 @@ static int hist_entry__data_page_size_snprintf(struct hist_entry *he, char *bf,
 	char str[PAGE_SIZE_NAME_LEN];
 
 	return repsep_snprintf(bf, size, "%-*s", width,
-			       get_page_size_name(he->mem_info->daddr.data_page_size, str));
+			get_page_size_name(mem_info__daddr(he->mem_info)->data_page_size, str));
 }
 
 struct sort_entry sort_mem_data_page_size = {
@@ -2271,46 +2403,21 @@ sort__typeoff_sort(struct hist_entry *left, struct hist_entry *right)
 	return left->mem_type_off - right->mem_type_off;
 }
 
-static void fill_member_name(char *buf, size_t sz, struct annotated_member *m,
-			     int offset, bool first)
-{
-	struct annotated_member *child;
-
-	if (list_empty(&m->children))
-		return;
-
-	list_for_each_entry(child, &m->children, node) {
-		if (child->offset <= offset && offset < child->offset + child->size) {
-			int len = 0;
-
-			/* It can have anonymous struct/union members */
-			if (child->var_name) {
-				len = scnprintf(buf, sz, "%s%s",
-						first ? "" : ".", child->var_name);
-				first = false;
-			}
-
-			fill_member_name(buf + len, sz - len, child, offset, first);
-			return;
-		}
-	}
-}
-
 static int hist_entry__typeoff_snprintf(struct hist_entry *he, char *bf,
 				     size_t size, unsigned int width __maybe_unused)
 {
 	struct annotated_data_type *he_type = he->mem_type;
 	char buf[4096];
 
-	buf[0] = '\0';
-	if (list_empty(&he_type->self.children))
-		snprintf(buf, sizeof(buf), "no field");
-	else
-		fill_member_name(buf, sizeof(buf), &he_type->self,
-				 he->mem_type_off, true);
-	buf[4095] = '\0';
+	if (he_type == &unknown_type || he_type == &stackop_type ||
+	    he_type == &canary_type)
+		return repsep_snprintf(bf, size, "%s", he_type->self.type_name);
 
-	return repsep_snprintf(bf, size, "%s %+d (%s)", he_type->self.type_name,
+	if (!annotated_data_type__get_member_name(he_type, buf, sizeof(buf),
+						  he->mem_type_off))
+		scnprintf(buf, sizeof(buf), "no field");
+
+	return repsep_snprintf(bf, size, "%s +%#x (%s)", he_type->self.type_name,
 			       he->mem_type_off, buf);
 }
 
@@ -2324,6 +2431,57 @@ struct sort_entry sort_type_offset = {
 	.se_width_idx	= HISTC_TYPE_OFFSET,
 };
 
+/* --sort typecln */
+
+/* TODO: use actual value in the system */
+#define TYPE_CACHELINE_SIZE  64
+
+static int64_t
+sort__typecln_sort(struct hist_entry *left, struct hist_entry *right)
+{
+	struct annotated_data_type *left_type = left->mem_type;
+	struct annotated_data_type *right_type = right->mem_type;
+	int64_t left_cln, right_cln;
+	int64_t ret;
+
+	if (!left_type) {
+		sort__type_init(left);
+		left_type = left->mem_type;
+	}
+
+	if (!right_type) {
+		sort__type_init(right);
+		right_type = right->mem_type;
+	}
+
+	ret = strcmp(left_type->self.type_name, right_type->self.type_name);
+	if (ret)
+		return ret;
+
+	left_cln = left->mem_type_off / TYPE_CACHELINE_SIZE;
+	right_cln = right->mem_type_off / TYPE_CACHELINE_SIZE;
+	return left_cln - right_cln;
+}
+
+static int hist_entry__typecln_snprintf(struct hist_entry *he, char *bf,
+				     size_t size, unsigned int width __maybe_unused)
+{
+	struct annotated_data_type *he_type = he->mem_type;
+
+	return repsep_snprintf(bf, size, "%s: cache-line %d", he_type->self.type_name,
+			       he->mem_type_off / TYPE_CACHELINE_SIZE);
+}
+
+struct sort_entry sort_type_cacheline = {
+	.se_header	= "Data Type Cacheline",
+	.se_cmp		= sort__type_cmp,
+	.se_collapse	= sort__typecln_sort,
+	.se_sort	= sort__typecln_sort,
+	.se_init	= sort__type_init,
+	.se_snprintf	= hist_entry__typecln_snprintf,
+	.se_width_idx	= HISTC_TYPE_CACHELINE,
+};
+
 
 struct sort_dimension {
 	const char		*name;
@@ -2382,6 +2540,8 @@ static struct sort_dimension common_sort_dimensions[] = {
 	DIM(SORT_ANNOTATE_DATA_TYPE, "type", sort_type),
 	DIM(SORT_ANNOTATE_DATA_TYPE_OFFSET, "typeoff", sort_type_offset),
 	DIM(SORT_SYM_OFFSET, "symoff", sort_sym_offset),
+	DIM(SORT_ANNOTATE_DATA_TYPE_CACHELINE, "typecln", sort_type_cacheline),
+	DIM(SORT_PARALLELISM, "parallelism", sort_parallelism),
 };
 
 #undef DIM
@@ -2402,6 +2562,15 @@ static struct sort_dimension bstack_sort_dimensions[] = {
 	DIM(SORT_SYM_IPC, "ipc_lbr", sort_sym_ipc),
 	DIM(SORT_ADDR_FROM, "addr_from", sort_addr_from),
 	DIM(SORT_ADDR_TO, "addr_to", sort_addr_to),
+	DIM(SORT_CALLCHAIN_BRANCH_PREDICTED,
+		"callchain_branch_predicted",
+		sort_callchain_branch_predicted),
+	DIM(SORT_CALLCHAIN_BRANCH_ABORT,
+		"callchain_branch_abort",
+		sort_callchain_branch_abort),
+	DIM(SORT_CALLCHAIN_BRANCH_CYCLES,
+		"callchain_branch_cycles",
+		sort_callchain_branch_cycles)
 };
 
 #undef DIM
@@ -2428,19 +2597,29 @@ struct hpp_dimension {
 	const char		*name;
 	struct perf_hpp_fmt	*fmt;
 	int			taken;
+	int			was_taken;
 };
 
 #define DIM(d, n) { .name = n, .fmt = &perf_hpp__format[d], }
 
 static struct hpp_dimension hpp_sort_dimensions[] = {
 	DIM(PERF_HPP__OVERHEAD, "overhead"),
+	DIM(PERF_HPP__LATENCY, "latency"),
 	DIM(PERF_HPP__OVERHEAD_SYS, "overhead_sys"),
 	DIM(PERF_HPP__OVERHEAD_US, "overhead_us"),
 	DIM(PERF_HPP__OVERHEAD_GUEST_SYS, "overhead_guest_sys"),
 	DIM(PERF_HPP__OVERHEAD_GUEST_US, "overhead_guest_us"),
 	DIM(PERF_HPP__OVERHEAD_ACC, "overhead_children"),
+	DIM(PERF_HPP__LATENCY_ACC, "latency_children"),
 	DIM(PERF_HPP__SAMPLES, "sample"),
 	DIM(PERF_HPP__PERIOD, "period"),
+	DIM(PERF_HPP__WEIGHT1, "weight1"),
+	DIM(PERF_HPP__WEIGHT2, "weight2"),
+	DIM(PERF_HPP__WEIGHT3, "weight3"),
+	/* aliases for weight_struct */
+	DIM(PERF_HPP__WEIGHT2, "ins_lat"),
+	DIM(PERF_HPP__WEIGHT3, "retire_lat"),
+	DIM(PERF_HPP__WEIGHT3, "p_stage_cyc"),
 };
 
 #undef DIM
@@ -2567,6 +2746,7 @@ MK_SORT_ENTRY_CHK(thread)
 MK_SORT_ENTRY_CHK(comm)
 MK_SORT_ENTRY_CHK(dso)
 MK_SORT_ENTRY_CHK(sym)
+MK_SORT_ENTRY_CHK(parallelism)
 
 
 static bool __sort__hpp_equal(struct perf_hpp_fmt *a, struct perf_hpp_fmt *b)
@@ -3127,9 +3307,8 @@ static int __dynamic_dimension__add(struct evsel *evsel,
 static int add_evsel_fields(struct evsel *evsel, bool raw_trace, int level)
 {
 	int ret;
-	struct tep_format_field *field;
-
-	field = evsel->tp_format->format.fields;
+	struct tep_event *tp_format = evsel__tp_format(evsel);
+	struct tep_format_field *field = tp_format ? tp_format->format.fields : NULL;
 	while (field) {
 		ret = __dynamic_dimension__add(evsel, field, raw_trace, level);
 		if (ret < 0)
@@ -3162,13 +3341,19 @@ static int add_all_matching_fields(struct evlist *evlist,
 {
 	int ret = -ESRCH;
 	struct evsel *evsel;
-	struct tep_format_field *field;
 
 	evlist__for_each_entry(evlist, evsel) {
+		struct tep_event *tp_format;
+		struct tep_format_field *field;
+
 		if (evsel->core.attr.type != PERF_TYPE_TRACEPOINT)
 			continue;
 
-		field = tep_find_any_field(evsel->tp_format, field_name);
+		tp_format = evsel__tp_format(evsel);
+		if (tp_format == NULL)
+			continue;
+
+		field = tep_find_any_field(tp_format, field_name);
 		if (field == NULL)
 			continue;
 
@@ -3250,7 +3435,9 @@ static int add_dynamic_entry(struct evlist *evlist, const char *tok,
 	if (!strcmp(field_name, "*")) {
 		ret = add_evsel_fields(evsel, raw_trace, level);
 	} else {
-		struct tep_format_field *field = tep_find_any_field(evsel->tp_format, field_name);
+		struct tep_event *tp_format = evsel__tp_format(evsel);
+		struct tep_format_field *field =
+			tp_format ? tep_find_any_field(tp_format, field_name) : NULL;
 
 		if (field == NULL) {
 			pr_debug("Cannot find event field for %s.%s\n",
@@ -3302,6 +3489,7 @@ static int __hpp_dimension__add(struct hpp_dimension *hd,
 		return -1;
 
 	hd->taken = 1;
+	hd->was_taken = 1;
 	perf_hpp_list__register_sort_field(list, fmt);
 	return 0;
 }
@@ -3336,10 +3524,15 @@ static int __hpp_dimension__add_output(struct perf_hpp_list *list,
 	return 0;
 }
 
-int hpp_dimension__add_output(unsigned col)
+int hpp_dimension__add_output(unsigned col, bool implicit)
 {
+	struct hpp_dimension *hd;
+
 	BUG_ON(col >= PERF_HPP__MAX_INDEX);
-	return __hpp_dimension__add_output(&perf_hpp_list, &hpp_sort_dimensions[col]);
+	hd = &hpp_sort_dimensions[col];
+	if (implicit && !hd->was_taken)
+		return 0;
+	return __hpp_dimension__add_output(&perf_hpp_list, hd);
 }
 
 int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
@@ -3423,7 +3616,13 @@ int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
 		if (!sd->name || strncasecmp(tok, sd->name, strlen(tok)))
 			continue;
 
-		if (sort__mode != SORT_MODE__BRANCH)
+		if ((sort__mode != SORT_MODE__BRANCH) &&
+			strncasecmp(tok, "callchain_branch_predicted",
+				    strlen(tok)) &&
+			strncasecmp(tok, "callchain_branch_abort",
+				    strlen(tok)) &&
+			strncasecmp(tok, "callchain_branch_cycles",
+				    strlen(tok)))
 			return -EINVAL;
 
 		if (sd->entry == &sort_sym_from || sd->entry == &sort_sym_to)
@@ -3458,6 +3657,34 @@ int sort_dimension__add(struct perf_hpp_list *list, const char *tok,
 	return -ESRCH;
 }
 
+/* This should match with sort_dimension__add() above */
+static bool is_hpp_sort_key(const char *key)
+{
+	unsigned i;
+
+	for (i = 0; i < ARRAY_SIZE(arch_specific_sort_keys); i++) {
+		if (!strcmp(arch_specific_sort_keys[i], key) &&
+		    !arch_support_sort_key(key)) {
+			return false;
+		}
+	}
+
+	for (i = 0; i < ARRAY_SIZE(common_sort_dimensions); i++) {
+		struct sort_dimension *sd = &common_sort_dimensions[i];
+
+		if (sd->name && !strncasecmp(key, sd->name, strlen(key)))
+			return false;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(hpp_sort_dimensions); i++) {
+		struct hpp_dimension *hd = &hpp_sort_dimensions[i];
+
+		if (!strncasecmp(key, hd->name, strlen(key)))
+			return true;
+	}
+	return false;
+}
+
 static int setup_sort_list(struct perf_hpp_list *list, char *str,
 			   struct evlist *evlist)
 {
@@ -3465,7 +3692,9 @@ static int setup_sort_list(struct perf_hpp_list *list, char *str,
 	int ret = 0;
 	int level = 0;
 	int next_level = 1;
+	int prev_level = 0;
 	bool in_group = false;
+	bool prev_was_hpp = false;
 
 	do {
 		tok = str;
@@ -3486,6 +3715,19 @@ static int setup_sort_list(struct perf_hpp_list *list, char *str,
 		}
 
 		if (*tok) {
+			if (is_hpp_sort_key(tok)) {
+				/* keep output (hpp) sort keys in the same level */
+				if (prev_was_hpp) {
+					bool next_same = (level == next_level);
+
+					level = prev_level;
+					next_level = next_same ? level : level+1;
+				}
+				prev_was_hpp = true;
+			} else {
+				prev_was_hpp = false;
+			}
+
 			ret = sort_dimension__add(list, tok, evlist, level);
 			if (ret == -EINVAL) {
 				if (!cacheline_size() && !strncasecmp(tok, "dcacheline", strlen(tok)))
@@ -3497,6 +3739,7 @@ static int setup_sort_list(struct perf_hpp_list *list, char *str,
 				ui__error("Unknown --sort key: `%s'", tok);
 				break;
 			}
+			prev_level = level;
 		}
 
 		level = next_level;
@@ -3592,10 +3835,24 @@ static char *setup_overhead(char *keys)
 	if (sort__mode == SORT_MODE__DIFF)
 		return keys;
 
-	keys = prefix_if_not_in("overhead", keys);
-
-	if (symbol_conf.cumulate_callchain)
-		keys = prefix_if_not_in("overhead_children", keys);
+	if (symbol_conf.prefer_latency) {
+		keys = prefix_if_not_in("overhead", keys);
+		keys = prefix_if_not_in("latency", keys);
+		if (symbol_conf.cumulate_callchain) {
+			keys = prefix_if_not_in("overhead_children", keys);
+			keys = prefix_if_not_in("latency_children", keys);
+		}
+	} else if (!keys || (!strstr(keys, "overhead") &&
+			!strstr(keys, "latency"))) {
+		if (symbol_conf.enable_latency)
+			keys = prefix_if_not_in("latency", keys);
+		keys = prefix_if_not_in("overhead", keys);
+		if (symbol_conf.cumulate_callchain) {
+			if (symbol_conf.enable_latency)
+				keys = prefix_if_not_in("latency_children", keys);
+			keys = prefix_if_not_in("overhead_children", keys);
+		}
+	}
 
 	return keys;
 }
@@ -3743,26 +4000,29 @@ void sort__setup_elide(FILE *output)
 	}
 }
 
-int output_field_add(struct perf_hpp_list *list, char *tok)
+int output_field_add(struct perf_hpp_list *list, const char *tok)
 {
 	unsigned int i;
 
-	for (i = 0; i < ARRAY_SIZE(common_sort_dimensions); i++) {
-		struct sort_dimension *sd = &common_sort_dimensions[i];
+	for (i = 0; i < ARRAY_SIZE(hpp_sort_dimensions); i++) {
+		struct hpp_dimension *hd = &hpp_sort_dimensions[i];
 
-		if (!sd->name || strncasecmp(tok, sd->name, strlen(tok)))
+		if (strncasecmp(tok, hd->name, strlen(tok)))
 			continue;
 
-		return __sort_dimension__add_output(list, sd);
+		if (!strcasecmp(tok, "weight"))
+			ui__warning("--fields weight shows the average value unlike in the --sort key.\n");
+
+		return __hpp_dimension__add_output(list, hd);
 	}
 
-	for (i = 0; i < ARRAY_SIZE(hpp_sort_dimensions); i++) {
-		struct hpp_dimension *hd = &hpp_sort_dimensions[i];
+	for (i = 0; i < ARRAY_SIZE(common_sort_dimensions); i++) {
+		struct sort_dimension *sd = &common_sort_dimensions[i];
 
-		if (strncasecmp(tok, hd->name, strlen(tok)))
+		if (!sd->name || strncasecmp(tok, sd->name, strlen(tok)))
 			continue;
 
-		return __hpp_dimension__add_output(list, hd);
+		return __sort_dimension__add_output(list, sd);
 	}
 
 	for (i = 0; i < ARRAY_SIZE(bstack_sort_dimensions); i++) {
@@ -3948,7 +4208,7 @@ static void add_hpp_sort_string(struct strbuf *sb, struct hpp_dimension *s, int
 		add_key(sb, s[i].name, llen);
 }
 
-char *sort_help(const char *prefix)
+char *sort_help(const char *prefix, enum sort_mode mode)
 {
 	struct strbuf sb;
 	char *s;
@@ -3960,10 +4220,12 @@ char *sort_help(const char *prefix)
 			    ARRAY_SIZE(hpp_sort_dimensions), &len);
 	add_sort_string(&sb, common_sort_dimensions,
 			    ARRAY_SIZE(common_sort_dimensions), &len);
-	add_sort_string(&sb, bstack_sort_dimensions,
-			    ARRAY_SIZE(bstack_sort_dimensions), &len);
-	add_sort_string(&sb, memory_sort_dimensions,
-			    ARRAY_SIZE(memory_sort_dimensions), &len);
+	if (mode == SORT_MODE__NORMAL || mode == SORT_MODE__BRANCH)
+		add_sort_string(&sb, bstack_sort_dimensions,
+				ARRAY_SIZE(bstack_sort_dimensions), &len);
+	if (mode == SORT_MODE__NORMAL || mode == SORT_MODE__MEMORY)
+		add_sort_string(&sb, memory_sort_dimensions,
+				ARRAY_SIZE(memory_sort_dimensions), &len);
 	s = strbuf_detach(&sb, NULL);
 	strbuf_release(&sb);
 	return s;