* [PATCH 1/7] perf tools: Fix report -F abort for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 2/7] perf tools: Fix report -F in_tx " Jiri Olsa
` (5 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F abort
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 4906cd81cb56..8355260cbf3f 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -989,6 +989,9 @@ struct sort_entry sort_mem_dcacheline = {
static int64_t
sort__abort_cmp(struct hist_entry *left, struct hist_entry *right)
{
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
return left->branch_info->flags.abort !=
right->branch_info->flags.abort;
}
@@ -996,10 +999,15 @@ sort__abort_cmp(struct hist_entry *left, struct hist_entry *right)
static int hist_entry__abort_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- static const char *out = ".";
+ static const char *out = "N/A";
+
+ if (he->branch_info) {
+ if (he->branch_info->flags.abort)
+ out = "A";
+ else
+ out = ".";
+ }
- if (he->branch_info->flags.abort)
- out = "A";
return repsep_snprintf(bf, size, "%-*s", width, out);
}
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 2/7] perf tools: Fix report -F in_tx for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
2014-10-16 12:20 ` [PATCH 1/7] perf tools: Fix report -F abort for data without branch info Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 3/7] perf tools: Fix report -F mispredict " Jiri Olsa
` (4 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F in_tx
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 13 ++++++++++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 8355260cbf3f..923dbbd9f65a 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -1021,6 +1021,9 @@ struct sort_entry sort_abort = {
static int64_t
sort__in_tx_cmp(struct hist_entry *left, struct hist_entry *right)
{
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
return left->branch_info->flags.in_tx !=
right->branch_info->flags.in_tx;
}
@@ -1028,10 +1031,14 @@ sort__in_tx_cmp(struct hist_entry *left, struct hist_entry *right)
static int hist_entry__in_tx_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- static const char *out = ".";
+ static const char *out = "N/A";
- if (he->branch_info->flags.in_tx)
- out = "T";
+ if (he->branch_info) {
+ if (he->branch_info->flags.in_tx)
+ out = "T";
+ else
+ out = ".";
+ }
return repsep_snprintf(bf, size, "%-*s", width, out);
}
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 3/7] perf tools: Fix report -F mispredict for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
2014-10-16 12:20 ` [PATCH 1/7] perf tools: Fix report -F abort for data without branch info Jiri Olsa
2014-10-16 12:20 ` [PATCH 2/7] perf tools: Fix report -F in_tx " Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 4/7] perf tools: Fix report -F symbol_to " Jiri Olsa
` (3 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F mispredict
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 20 ++++++++++++--------
1 file changed, 12 insertions(+), 8 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 923dbbd9f65a..07d3ffb3ee2b 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -471,11 +471,13 @@ struct sort_entry sort_sym_to = {
static int64_t
sort__mispredict_cmp(struct hist_entry *left, struct hist_entry *right)
{
- const unsigned char mp = left->branch_info->flags.mispred !=
- right->branch_info->flags.mispred;
- const unsigned char p = left->branch_info->flags.predicted !=
- right->branch_info->flags.predicted;
+ unsigned char mp, p;
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
+ mp = left->branch_info->flags.mispred != right->branch_info->flags.mispred;
+ p = left->branch_info->flags.predicted != right->branch_info->flags.predicted;
return mp || p;
}
@@ -483,10 +485,12 @@ static int hist_entry__mispredict_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width){
static const char *out = "N/A";
- if (he->branch_info->flags.predicted)
- out = "N";
- else if (he->branch_info->flags.mispred)
- out = "Y";
+ if (he->branch_info) {
+ if (he->branch_info->flags.predicted)
+ out = "N";
+ else if (he->branch_info->flags.mispred)
+ out = "Y";
+ }
return repsep_snprintf(bf, size, "%-*.*s", width, width, out);
}
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 4/7] perf tools: Fix report -F symbol_to for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
` (2 preceding siblings ...)
2014-10-16 12:20 ` [PATCH 3/7] perf tools: Fix report -F mispredict " Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 5/7] perf tools: Fix report -F symbol_from " Jiri Olsa
` (2 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F symbol_to
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 19 ++++++++++++++-----
1 file changed, 14 insertions(+), 5 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 07d3ffb3ee2b..134633c474a0 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -413,8 +413,13 @@ sort__sym_from_cmp(struct hist_entry *left, struct hist_entry *right)
static int64_t
sort__sym_to_cmp(struct hist_entry *left, struct hist_entry *right)
{
- struct addr_map_symbol *to_l = &left->branch_info->to;
- struct addr_map_symbol *to_r = &right->branch_info->to;
+ struct addr_map_symbol *to_l, *to_r;
+
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
+ to_l = &left->branch_info->to;
+ to_r = &right->branch_info->to;
if (!to_l->sym && !to_r->sym)
return _sort__addr_cmp(to_l->addr, to_r->addr);
@@ -434,10 +439,14 @@ static int hist_entry__sym_from_snprintf(struct hist_entry *he, char *bf,
static int hist_entry__sym_to_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- struct addr_map_symbol *to = &he->branch_info->to;
- return _hist_entry__sym_snprintf(to->map, to->sym, to->addr,
- he->level, bf, size, width);
+ if (he->branch_info) {
+ struct addr_map_symbol *to = &he->branch_info->to;
+
+ return _hist_entry__sym_snprintf(to->map, to->sym, to->addr,
+ he->level, bf, size, width);
+ }
+ return repsep_snprintf(bf, size, "%-*.*s", width, width, "N/A");
}
struct sort_entry sort_dso_from = {
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 5/7] perf tools: Fix report -F symbol_from for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
` (3 preceding siblings ...)
2014-10-16 12:20 ` [PATCH 4/7] perf tools: Fix report -F symbol_to " Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 6/7] perf tools: Fix report -F dso_to " Jiri Olsa
2014-10-16 12:20 ` [PATCH 7/7] perf tools: Fix report -F dso_from " Jiri Olsa
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F symbol_from
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 134633c474a0..d64c020acac5 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -404,6 +404,12 @@ sort__sym_from_cmp(struct hist_entry *left, struct hist_entry *right)
struct addr_map_symbol *from_l = &left->branch_info->from;
struct addr_map_symbol *from_r = &right->branch_info->from;
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
+ from_l = &left->branch_info->from;
+ from_r = &right->branch_info->from;
+
if (!from_l->sym && !from_r->sym)
return _sort__addr_cmp(from_l->addr, from_r->addr);
@@ -430,10 +436,14 @@ sort__sym_to_cmp(struct hist_entry *left, struct hist_entry *right)
static int hist_entry__sym_from_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- struct addr_map_symbol *from = &he->branch_info->from;
- return _hist_entry__sym_snprintf(from->map, from->sym, from->addr,
- he->level, bf, size, width);
+ if (he->branch_info) {
+ struct addr_map_symbol *from = &he->branch_info->from;
+ return _hist_entry__sym_snprintf(from->map, from->sym, from->addr,
+ he->level, bf, size, width);
+ }
+
+ return repsep_snprintf(bf, size, "%-*.*s", width, width, "N/A");
}
static int hist_entry__sym_to_snprintf(struct hist_entry *he, char *bf,
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 6/7] perf tools: Fix report -F dso_to for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
` (4 preceding siblings ...)
2014-10-16 12:20 ` [PATCH 5/7] perf tools: Fix report -F symbol_from " Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
2014-10-16 12:20 ` [PATCH 7/7] perf tools: Fix report -F dso_from " Jiri Olsa
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F dso_to
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index d64c020acac5..1c303f8cc61d 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -387,6 +387,9 @@ static int hist_entry__dso_from_snprintf(struct hist_entry *he, char *bf,
static int64_t
sort__dso_to_cmp(struct hist_entry *left, struct hist_entry *right)
{
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
return _sort__dso_cmp(left->branch_info->to.map,
right->branch_info->to.map);
}
@@ -394,8 +397,11 @@ sort__dso_to_cmp(struct hist_entry *left, struct hist_entry *right)
static int hist_entry__dso_to_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- return _hist_entry__dso_snprintf(he->branch_info->to.map,
- bf, size, width);
+ if (he->branch_info)
+ return _hist_entry__dso_snprintf(he->branch_info->to.map,
+ bf, size, width);
+ else
+ return repsep_snprintf(bf, size, "%-*.*s", width, width, "N/A");
}
static int64_t
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH 7/7] perf tools: Fix report -F dso_from for data without branch info
2014-10-16 12:20 [PATCH 0/7] perf tools: Fix branch report segfaults Jiri Olsa
` (5 preceding siblings ...)
2014-10-16 12:20 ` [PATCH 6/7] perf tools: Fix report -F dso_to " Jiri Olsa
@ 2014-10-16 12:20 ` Jiri Olsa
6 siblings, 0 replies; 8+ messages in thread
From: Jiri Olsa @ 2014-10-16 12:20 UTC (permalink / raw)
To: linux-kernel
Cc: Jiri Olsa, Andi Kleen, Arnaldo Carvalho de Melo, Corey Ashford,
David Ahern, Frederic Weisbecker, Ingo Molnar, Namhyung Kim,
Paul Mackerras, Peter Zijlstra
The branch field sorting code assumes hist_entry::branch_info
is allocated, which is wrong and following perf session ends
up with report segfault.
$ perf record ls
$ perf report -F dso_from
perf: Segmentation fault
Checking that hist_entry::branch_info is valid and display
"N/A" string in snprint callback if it's not.
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
tools/perf/util/sort.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 1c303f8cc61d..6a8435a550d7 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -373,6 +373,9 @@ struct sort_entry sort_cpu = {
static int64_t
sort__dso_from_cmp(struct hist_entry *left, struct hist_entry *right)
{
+ if (!left->branch_info || !right->branch_info)
+ return 0;
+
return _sort__dso_cmp(left->branch_info->from.map,
right->branch_info->from.map);
}
@@ -380,8 +383,11 @@ sort__dso_from_cmp(struct hist_entry *left, struct hist_entry *right)
static int hist_entry__dso_from_snprintf(struct hist_entry *he, char *bf,
size_t size, unsigned int width)
{
- return _hist_entry__dso_snprintf(he->branch_info->from.map,
- bf, size, width);
+ if (he->branch_info)
+ return _hist_entry__dso_snprintf(he->branch_info->from.map,
+ bf, size, width);
+ else
+ return repsep_snprintf(bf, size, "%-*.*s", width, width, "N/A");
}
static int64_t
--
1.9.3
^ permalink raw reply related [flat|nested] 8+ messages in thread