Skip to contentSkip to conversation

perf iostat: Use system-wide mode if the target cpu_list is unspecified

PATCH In review Like Xu 3 participants 5 messages Sep 27, 2021 – Sep 28, 2021 perf events

lore
Share & tools
lore

Like Xu

author
#0 lore raw

PATCH 1/2 perf iostat: Use system-wide mode if the target cpu_list is unspecified

From: Like Xu <[email protected]>

An iostate use case like "perf iostat 0000:16,0000:97 -- ls" should be implemented to work in system-wide mode to ensure that the output from print_header() is consistent with the user documentation perf-iostat.txt, rather than incorrectly assuming that the kernel does not support it:

Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument) \
for event (uncore_iio_0/event=0x83,umask=0x04,ch_mask=0xF,fc_mask=0x07/). /bin/dmesg | grep -i perf may provide additional information.

This error is easily fixed by assigning system-wide mode by default for IOSTAT_RUN only when the target cpu_list is unspecified.

---
 tools/perf/builtin-stat.c | 2 ++
 1 file changed, 2 insertions(+)
tools/perf/builtin-stat.c +2 -0
index f6e87b7be5fa..f0ecfda34ece 100644
@@ -2408,6 +2408,8 @@ int cmd_stat(int argc, const char **argv)
 			goto out;
 		} else if (verbose)
 			iostat_list(evsel_list, &stat_config);
+		if (iostat_mode == IOSTAT_RUN && !target__has_cpu(&target))
+			target.system_wide = true;
 	}
 
 	if (add_default_attributes())
Signature · 1 line
2.32.0
author
#1 lore raw

PATCH 2/2 perf iostat: Fix Segmentation fault from NULL 'struct perf_counts_values *'

From: Like Xu <[email protected]>

If the perf-iostat user specifies two or more iio_root_ports and also specifies the cpu(s) by -C which is not *connected to all* the above iio ports, the iostat_print_metric() will run into trouble:

For example:

 $ perf iostat list
 S0-uncore_iio_0<0000:16>
 S1-uncore_iio_0<0000:97> # <--- CPU 1 is located in the socket S0
 $ perf iostat 0000:16,0000:97 -C 1 -- ls
 port 	Inbound Read(MB)	Inbound Write(MB)	Outbound Read(MB)	Outbound
 Write(MB) ../perf-iostat: line 12: 104418 Segmentation fault
 (core dumped) perf stat --iostat$DELIMITER$*

The core-dump stack says, in the above corner case, the returned (struct perf_counts_values *) count will be NULL, and the caller iostat_print_metric() apparently doesn't not handle this case.

433 struct perf_counts_values *count = perf_counts(evsel->counts, die, 0);
434
435 if (count->run && count->ena) {
(gdb) p count
$1 = (struct perf_counts_values *) 0x0

The deeper reason is that there are actually no statistics from the user specified pair "iostat 0000:X, -C (disconnected) Y ", but let's fix it with minimum cost by adding a NULL check in the user space.

---
 tools/perf/arch/x86/util/iostat.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/arch/x86/util/iostat.c +1 -1
index eeafe97b8105..792cd75ade33 100644
@@ -432,7 +432,7 @@ void iostat_print_metric(struct perf_stat_config *config, struct evsel *evsel,
 	u8 die = ((struct iio_root_port *)evsel->priv)->die;
 	struct perf_counts_values *count = perf_counts(evsel->counts, die, 0);
 
-	if (count->run && count->ena) {
+	if (count && count->run && count->ena) {
 		if (evsel->prev_raw_counts && !out->force_header) {
 			struct perf_counts_values *prev_count =
 				perf_counts(evsel->prev_raw_counts, die, 0);
Signature · 1 line
2.32.0
#2 lore raw

Re: [PATCH 2/2] perf iostat: Fix Segmentation fault from NULL 'struct perf_counts_values *'

Em Mon, Sep 27, 2021 at 04:11:15PM +0800, Like Xu escreveu:

Quoted 27 lines

From: Like Xu <[email protected]>

If the perf-iostat user specifies two or more iio_root_ports and also specifies the cpu(s) by -C which is not *connected to all* the above iio ports, the iostat_print_metric() will run into trouble:

For example:

 $ perf iostat list
 S0-uncore_iio_0<0000:16>
 S1-uncore_iio_0<0000:97> # <--- CPU 1 is located in the socket S0
 $ perf iostat 0000:16,0000:97 -C 1 -- ls
 port 	Inbound Read(MB)	Inbound Write(MB)	Outbound Read(MB)	Outbound
 Write(MB) ../perf-iostat: line 12: 104418 Segmentation fault
 (core dumped) perf stat --iostat$DELIMITER$*

The core-dump stack says, in the above corner case, the returned (struct perf_counts_values *) count will be NULL, and the caller iostat_print_metric() apparently doesn't not handle this case.

433 struct perf_counts_values *count = perf_counts(evsel->counts, die, 0);
434
435 if (count->run && count->ena) {
(gdb) p count
$1 = (struct perf_counts_values *) 0x0

Quoted 3 lines

The deeper reason is that there are actually no statistics from the user specified pair "iostat 0000:X, -C (disconnected) Y ", but let's fix it with minimum cost by adding a NULL check in the user space.

Added:

Fixes f9ed693e8bc0e7de ("perf stat: Enable iostat mode for x86 platforms")

Please do that next time,

- Arnaldo

Quoted 20 lines
---
 tools/perf/arch/x86/util/iostat.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/arch/x86/util/iostat.c +1 -1
index eeafe97b8105..792cd75ade33 100644
@@ -432,7 +432,7 @@ void iostat_print_metric(struct perf_stat_config *config, struct evsel *evsel,
 	u8 die = ((struct iio_root_port *)evsel->priv)->die;
 	struct perf_counts_values *count = perf_counts(evsel->counts, die, 0);
 
-	if (count->run && count->ena) {
+	if (count && count->run && count->ena) {
 		if (evsel->prev_raw_counts && !out->force_header) {
 			struct perf_counts_values *prev_count =
 				perf_counts(evsel->prev_raw_counts, die, 0);
Signature · 1 line
2.32.0
Signature · 1 line
- Arnaldo
#3 lore raw
In reply to Like Xu

Em Mon, Sep 27, 2021 at 04:11:14PM +0800, Like Xu escreveu:

Quoted 7 lines

From: Like Xu <[email protected]>

An iostate use case like "perf iostat 0000:16,0000:97 -- ls" should be implemented to work in system-wide mode to ensure that the output from print_header() is consistent with the user documentation perf-iostat.txt, rather than incorrectly assuming that the kernel does not support it:

Quoted 7 lines

Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument) \
for event (uncore_iio_0/event=0x83,umask=0x04,ch_mask=0xF,fc_mask=0x07/). /bin/dmesg | grep -i perf may provide additional information.

This error is easily fixed by assigning system-wide mode by default for IOSTAT_RUN only when the target cpu_list is unspecified.

Looks ok, added:

Fixes f07952b179697771 ("perf stat: Basic support for iostat in perf")

For [email protected] sake.

- Arnaldo

Quoted 20 lines
---
 tools/perf/builtin-stat.c | 2 ++
 1 file changed, 2 insertions(+)
tools/perf/builtin-stat.c +2 -0
index f6e87b7be5fa..f0ecfda34ece 100644
@@ -2408,6 +2408,8 @@ int cmd_stat(int argc, const char **argv)
 			goto out;
 		} else if (verbose)
 			iostat_list(evsel_list, &stat_config);
+		if (iostat_mode == IOSTAT_RUN && !target__has_cpu(&target))
+			target.system_wide = true;
 	}
 
 	if (add_default_attributes())
Signature · 1 line
2.32.0
Signature · 1 line
- Arnaldo
Tested-by on v1 1/2
#4 lore raw
In reply to Like Xu
Quoted 7 lines · Like Xu

From: Like Xu <[email protected]>

An iostate use case like "perf iostat 0000:16,0000:97 -- ls" should be implemented to work in system-wide mode to ensure that the output from print_header() is consistent with the user documentation perf-iostat.txt, rather than incorrectly assuming that the kernel does not support it:

Quoted 7 lines

Error:
The sys_perf_event_open() syscall returned with 22 (Invalid argument) \
for event (uncore_iio_0/event=0x83,umask=0x04,ch_mask=0xF,fc_mask=0x07/). /bin/dmesg | grep -i perf may provide additional information.

This error is easily fixed by assigning system-wide mode by default for IOSTAT_RUN only when the target cpu_list is unspecified.

Hello,
Thank you for your fixes.

Quoted 18 lines
---
  tools/perf/builtin-stat.c | 2 ++
  1 file changed, 2 insertions(+)
tools/perf/builtin-stat.c +2 -0
index f6e87b7be5fa..f0ecfda34ece 100644
@@ -2408,6 +2408,8 @@ int cmd_stat(int argc, const char **argv)
  			goto out;
  		} else if (verbose)
  			iostat_list(evsel_list, &stat_config);
+		if (iostat_mode == IOSTAT_RUN && !target__has_cpu(&target))
+			target.system_wide = true;
  	}
 
  	if (add_default_attributes())