| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 27 Apr 2017 23:59:50 +0200
From: Milian Wolff <milian.wolff@...b.com>
To: Linux-kernel@...r.kernel.org
Cc: linux-perf-users@...r.kernel.org,
Milian Wolff <milian.wolff@...b.com>,
Arnaldo Carvalho de Melo <acme@...hat.com>,
David Ahern <dsahern@...il.com>,
Namhyung Kim <namhyung@...nel.org>,
Peter Zijlstra <a.p.zijlstra@...llo.nl>,
Yao Jin <yao.jin@...ux.intel.com>
Subject: [PATCH] perf report: distinguish between inliners in the same function
When different functions get inlined into the same function, we
want to show them individually in the reports. But when we group by
function, we would aggregate all IPs and would only keep the first
one in that function. E.g. for C++ code like the following:
~~~~~
#include <cmath>
#include <random>
#include <iostream>
using namespace std;
int main()
{
//--> slide
uniform_real_distribution<double> uniform(-1E5, 1E5);
default_random_engine engine;
double s = 0;
for (int i = 0; i < 10000000; ++i) {
s += uniform(engine);
}
//<-- slide
cout << "random sum: " << s << '\n';
return 0;
}
~~~~~
Building it with `g++ -O2 -g` and recording some samples with
`perf record --call-graph dwarf` yields for me:
~~~~~
$ perf report --stdio --inline --no-children
Failed to open [ext4], continuing without symbols
# To display the perf.data header info, please use --header/--header-only
options.
#
#
# Total Lost Samples: 0
#
# Samples: 499 of event 'cycles'
# Event count (approx.): 329354953
#
# Overhead Command Shared Object Symbol
# ........ ......... ................. ................................
#
96.70% ex_random ex_random [.] main
|
---main
__libc_start_main
_start
...
~~~~~
Note how no inlined frames are actually shown, because the first
sample in main points to an IP that does not correspond to any
inlined frames.
With this patch applied, we instead get the following, much more
meaningful, report:
~~~~~
# Overhead Command Shared Object Symbol
# ........ ......... ................. ................................
#
96.70% ex_random ex_random [.] main
|
|--47.19%--main
| std::__detail::_Adaptor<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul>, double>::operator() (inline)
| std::uniform_real_distribution<double>::operator()<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
| std::uniform_real_distribution<double>::operator()<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
| main (inline)
| __libc_start_main
| _start
|
|--32.61%--main
| std::__detail::__mod<unsigned long, 2147483647ul, 16807ul, 0ul> (inline)
| std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul>::operator() (inline)
| std::generate_canonical<double, 53ul, std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
| std::__detail::_Adaptor<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul>, double>::operator() (inline)
| std::uniform_real_distribution<double>::operator()<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
| std::uniform_real_distribution<double>::operator()<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
| main (inline)
| __libc_start_main
| _start
|
|--15.07%--main
| __libc_start_main
| _start
|
--1.84%--main
std::uniform_real_distribution<double>::operator()<std::linear_congruential_engine<unsigned long, 16807ul, 0ul, 2147483647ul> > (inline)
main (inline)
__libc_start_main
_start
...
~~~~~
Cc: Arnaldo Carvalho de Melo <acme@...hat.com>
Cc: David Ahern <dsahern@...il.com>
Cc: Namhyung Kim <namhyung@...nel.org>
Cc: Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc: Yao Jin <yao.jin@...ux.intel.com>
Signed-off-by: Milian Wolff <milian.wolff@...b.com>
---
tools/perf/util/callchain.c | 71 +++++++++++++++++++++++++++++++++++++--------
1 file changed, 59 insertions(+), 12 deletions(-)
diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
index 81fc29ac798f..9984dbda3e61 100644
--- a/tools/perf/util/callchain.c
+++ b/tools/perf/util/callchain.c
@@ -618,15 +618,9 @@ enum match_result {
MATCH_GT,
};
-static enum match_result match_chain_srcline(struct callchain_cursor_node *node,
- struct callchain_list *cnode)
+static enum match_result match_chain_strings(const char *left,
+ const char *right)
{
- char *left = get_srcline(cnode->ms.map->dso,
- map__rip_2objdump(cnode->ms.map, cnode->ip),
- cnode->ms.sym, true, false);
- char *right = get_srcline(node->map->dso,
- map__rip_2objdump(node->map, node->ip),
- node->sym, true, false);
enum match_result ret = MATCH_EQ;
int cmp;
@@ -636,19 +630,66 @@ static enum match_result match_chain_srcline(struct callchain_cursor_node *node,
cmp = 1;
else if (left && !right)
cmp = -1;
- else if (cnode->ip == node->ip)
- cmp = 0;
else
- cmp = (cnode->ip < node->ip) ? -1 : 1;
+ return MATCH_ERROR;
if (cmp != 0)
ret = cmp < 0 ? MATCH_LT : MATCH_GT;
+ return ret;
+}
+
+static enum match_result match_chain_srcline(struct callchain_cursor_node *node,
+ struct callchain_list *cnode)
+{
+ char *left = get_srcline(cnode->ms.map->dso,
+ map__rip_2objdump(cnode->ms.map, cnode->ip),
+ cnode->ms.sym, true, false);
+ char *right = get_srcline(node->map->dso,
+ map__rip_2objdump(node->map, node->ip),
+ node->sym, true, false);
+ enum match_result ret = match_chain_strings(left, right);
+
free_srcline(left);
free_srcline(right);
return ret;
}
+static const char *first_inlined_funcname(struct inline_node *node)
+{
+ struct inline_list *entry = NULL;
+
+ if (node)
+ entry = list_first_entry(&node->val, struct inline_list, list);
+ return entry ? entry->funcname : NULL;
+}
+
+static enum match_result match_chain_inliner(struct callchain_cursor_node *node,
+ struct callchain_list *cnode)
+{
+ u64 left_ip = map__rip_2objdump(cnode->ms.map, cnode->ip);
+ u64 right_ip = map__rip_2objdump(node->map, node->ip);
+ struct inline_node *left_node = NULL;
+ struct inline_node *right_node = NULL;
+ const char *left_func = NULL;
+ const char *right_func = NULL;
+ enum match_result ret = MATCH_EQ;
+
+ left_node = dso__parse_addr_inlines(cnode->ms.map->dso, left_ip);
+ left_func = first_inlined_funcname(left_node);
+
+ right_node = dso__parse_addr_inlines(node->map->dso, right_ip);
+ right_func = first_inlined_funcname(right_node);
+
+ ret = match_chain_strings(left_func, right_func);
+
+ if (left_node)
+ inline_node__delete(left_node);
+ if (right_node)
+ inline_node__delete(right_node);
+ return ret;
+}
+
static enum match_result match_chain(struct callchain_cursor_node *node,
struct callchain_list *cnode)
{
@@ -671,7 +712,13 @@ static enum match_result match_chain(struct callchain_cursor_node *node,
}
if (left == right) {
- if (node->branch) {
+ if (symbol_conf.inline_name && cnode->ip != node->ip) {
+ enum match_result match = match_chain_inliner(node,
+ cnode);
+
+ if (match != MATCH_ERROR)
+ return match;
+ } else if (node->branch) {
cnode->branch_count++;
if (node->branch_flags.predicted)
--
2.12.2
Powered by blists - more mailing lists