Skip to contentSkip to conversation

perf test: Add bpf-output event

PATCH In review Sukadev Bhattiprolu 4 participants 6 messages Nov 18, 2015 – Nov 20, 2015 perf events

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 perf test: Add bpf-output event

Quoted 1 line

From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001

From: Sukadev Bhattiprolu <[email protected]>
Date: Wed, 18 Nov 2015 19:06:08 -0500
Subject: [PATCH] perf test: Add bpf-output event

The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is missing from the perf tool. Among other things, results in the 'roundtrip evsel->name check' test case of 'perf test' failing on Powerpc.

---
 tools/perf/util/evsel.c        | 1 +
 tools/perf/util/parse-events.c | 4 ++++
 tools/perf/util/parse-events.l | 1 +
 3 files changed, 6 insertions(+)
tools/perf/util/evsel.c +1 -0
index 397fb4e..2033632 100644
@@ -342,6 +342,7 @@ const char *perf_evsel__sw_names[PERF_COUNT_SW_MAX] = {
 	"alignment-faults",
 	"emulation-faults",
 	"dummy",
+	"bpf-output",
 };
 
 static const char *__perf_evsel__sw_name(u64 config)
tools/perf/util/parse-events.c +4 -0
index e48d9da..40ae92a 100644
@@ -124,6 +124,10 @@ struct event_symbol event_symbols_sw[PERF_COUNT_SW_MAX] = {
 		.symbol = "dummy",
 		.alias  = "",
 	},
+	[PERF_COUNT_SW_BPF_OUTPUT] = {
+		.symbol = "bpf-output",
+		.alias  = "",
+	},
 };
 
 #define __PERF_EVENT_FIELD(config, name) \
tools/perf/util/parse-events.l +1 -0
index 58c5831..380298d 100644
@@ -238,6 +238,7 @@ cpu-migrations|migrations			{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COU
 alignment-faults				{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_ALIGNMENT_FAULTS); }
 emulation-faults				{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_EMULATION_FAULTS); }
 dummy						{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_DUMMY); }
+bpf-output						{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_BPF_OUTPUT); }
 
 	/*
 	 * We have to handle the kernel PMU event cycles-ct/cycles-t/mem-loads/mem-stores separately.
Signature · 1 line
1.8.3.1
#1 lore raw

Em Wed, Nov 18, 2015 at 11:26:04AM -0800, Sukadev Bhattiprolu escreveu:

Quoted 8 lines
Quoted 1 line

From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001

From: Sukadev Bhattiprolu <[email protected]>
Date: Wed, 18 Nov 2015 19:06:08 -0500
Subject: [PATCH] perf test: Add bpf-output event

The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is missing from the perf tool. Among other things, results in the 'roundtrip evsel->name check' test case of 'perf test' failing on Powerpc.

Next time can you please state if this is for this merge window or for the next?

Will apply it for perf/core, for the next merge window.

- Arnaldo

Quoted 48 lines
---
 tools/perf/util/evsel.c        | 1 +
 tools/perf/util/parse-events.c | 4 ++++
 tools/perf/util/parse-events.l | 1 +
 3 files changed, 6 insertions(+)
tools/perf/util/evsel.c +1 -0
index 397fb4e..2033632 100644
@@ -342,6 +342,7 @@ const char *perf_evsel__sw_names[PERF_COUNT_SW_MAX] = {
 	"alignment-faults",
 	"emulation-faults",
 	"dummy",
+	"bpf-output",
 };
 
 static const char *__perf_evsel__sw_name(u64 config)
tools/perf/util/parse-events.c +4 -0
index e48d9da..40ae92a 100644
@@ -124,6 +124,10 @@ struct event_symbol event_symbols_sw[PERF_COUNT_SW_MAX] = {
 		.symbol = "dummy",
 		.alias  = "",
 	},
+	[PERF_COUNT_SW_BPF_OUTPUT] = {
+		.symbol = "bpf-output",
+		.alias  = "",
+	},
 };
 
 #define __PERF_EVENT_FIELD(config, name) \
tools/perf/util/parse-events.l +1 -0
index 58c5831..380298d 100644
@@ -238,6 +238,7 @@ cpu-migrations|migrations			{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COU
 alignment-faults				{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_ALIGNMENT_FAULTS); }
 emulation-faults				{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_EMULATION_FAULTS); }
 dummy						{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_DUMMY); }
+bpf-output						{ return sym(yyscanner, PERF_TYPE_SOFTWARE, PERF_COUNT_SW_BPF_OUTPUT); }
 
 	/*
 	 * We have to handle the kernel PMU event cycles-ct/cycles-t/mem-loads/mem-stores separately.
Signature · 1 line
1.8.3.1
#2 lore raw
Quoted 6 lines · Arnaldo Carvalho de Melo

Em Wed, Nov 18, 2015 at 11:26:04AM -0800, Sukadev Bhattiprolu escreveu:

Quoted 5 lines
Quoted 1 line

From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001

From: Sukadev Bhattiprolu <[email protected]>
Date: Wed, 18 Nov 2015 19:06:08 -0500
Subject: [PATCH] perf test: Add bpf-output event

Quoted 8 lines
Quoted 3 lines

The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is missing from the perf tool. Among other things, results in the 'roundtrip evsel->name check' test case of 'perf test' failing on Powerpc.

Next time can you please state if this is for this merge window or for the next?

Will apply it for perf/core, for the next merge window.

wait a sec, I believe Wang has posted an RFC a month ago that adds support for this properly instead of simply shutting up the error.
Or I missed something?

#3 lore raw
Quoted 10 lines · Alexei Starovoitov
Quoted 9 lines · Arnaldo Carvalho de Melo

Em Wed, Nov 18, 2015 at 11:26:04AM -0800, Sukadev Bhattiprolu escreveu:

Quoted 8 lines
Quoted 1 line

From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001

From: Sukadev Bhattiprolu <[email protected]>
Date: Wed, 18 Nov 2015 19:06:08 -0500
Subject: [PATCH] perf test: Add bpf-output event

The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is missing from the perf tool. Among other things, results in the 'roundtrip evsel->name check' test case of 'perf test' failing on Powerpc.

Quoted 6 lines
Quoted 4 lines

Next time can you please state if this is for this merge window or for the next?

Will apply it for perf/core, for the next merge window.

wait a sec, I believe Wang has posted an RFC a month ago that adds support for this properly instead of simply shutting up the error.

Yes. Please have a look at

http://lkml.kernel.org/g/[email protected]

and patch 23/31 - 31/31 in

http://lkml.kernel.org/g/[email protected]

We are working on a solution which truly utilizes bpf-output, including filling event array and extracting BPF output through perf data to CTF format. Now they are separated in multiple patchset, and I'd like to resend them to Arnaldo in this week.

Thank you.

[SNIP]

#4 lore raw

Wangnan (F) [[email protected]] wrote:

| 
| 
| On 2015/11/19 7:14, Alexei Starovoitov wrote:
| >On Wed, Nov 18, 2015 at 05:50:39PM -0300, Arnaldo Carvalho de Melo wrote:
| >>Em Wed, Nov 18, 2015 at 11:26:04AM -0800, Sukadev Bhattiprolu escreveu:
| >>>>From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001
| >>>From: Sukadev Bhattiprolu <[email protected]>
| >>>Date: Wed, 18 Nov 2015 19:06:08 -0500
| >>>Subject: [PATCH] perf test: Add bpf-output event
| >>>
| >>>The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is
| >>>missing from the perf tool. Among other things, results in the 'roundtrip
| >>>evsel->name check' test case of 'perf test' failing on Powerpc.
| >>Next time can you please state if this is for this merge window or for
| >>the next?
| >>
| >>Will apply it for perf/core, for the next merge window.
| >wait a sec, I believe Wang has posted an RFC a month ago that adds support
| >for this properly instead of simply shutting up the error.
| 
| Yes. Please have a look at
| 
| http://lkml.kernel.org/g/[email protected]
| 
| and patch 23/31 - 31/31 in
| 
| http://lkml.kernel.org/g/[email protected]
| 
| We are working on a solution which truly utilizes bpf-output, including
| filling event array and extracting BPF output through perf data to CTF
| format. Now they are separated in multiple patchset, and I'd like to
| resend them to Arnaldo in this week.

Ok. It looked like there was development going on around BPF so I put in the stubs to avoid false positives in the perf test. There is a BPF filter test case that is also getting Skipped. Would this patchset address that too?

Sukadev

#5 lore raw
Quoted 34 lines · Sukadev Bhattiprolu

Wangnan (F) [[email protected]] wrote:

|
|
| On 2015/11/19 7:14, Alexei Starovoitov wrote:
| >On Wed, Nov 18, 2015 at 05:50:39PM -0300, Arnaldo Carvalho de Melo wrote:
| >>Em Wed, Nov 18, 2015 at 11:26:04AM -0800, Sukadev Bhattiprolu escreveu:
| >>>>From 8f71d55dd3e27e6ca2138e3ed6dfeceb1c00a426 Mon Sep 17 00:00:00 2001
| >>>From: Sukadev Bhattiprolu <[email protected]>
| >>>Date: Wed, 18 Nov 2015 19:06:08 -0500
| >>>Subject: [PATCH] perf test: Add bpf-output event
| >>>
| >>>The kernel has added support for 'PERF_COUNT_SW_BPF_OUTPUT' but that is
| >>>missing from the perf tool. Among other things, results in the 'roundtrip
| >>>evsel->name check' test case of 'perf test' failing on Powerpc.
| >>Next time can you please state if this is for this merge window or for
| >>the next?
| >>
| >>Will apply it for perf/core, for the next merge window.
| >wait a sec, I believe Wang has posted an RFC a month ago that adds support
| >for this properly instead of simply shutting up the error.
|
| Yes. Please have a look at
|
| http://lkml.kernel.org/g/[email protected]
|
| and patch 23/31 - 31/31 in
|
| http://lkml.kernel.org/g/[email protected]
|
| We are working on a solution which truly utilizes bpf-output, including
| filling event array and extracting BPF output through perf data to CTF
| format. Now they are separated in multiple patchset, and I'd like to
| resend them to Arnaldo in this week.
Quoted 4 lines

Ok. It looked like there was development going on around BPF so I put in the stubs to avoid false positives in the perf test. There is a BPF filter test case that is also getting Skipped. Would this patchset address that too?

Do you mean this?

$ ./perf test bpf
37: Test BPF filter                                          :
37.1: Test basic BPF filtering                               : Skip

To make it pass you need to build BPF compiling environment first.

First of all you must be root:

$ ./perf test -v bpf
37: Test BPF filter                                          :
37.1: Test basic BPF filtering                               :
--- start ---
test child forked, pid 19359
Only root can run BPF test
test child finished with -2
---- end ----
Test BPF filter subtest 0: Skip

Then it would tell you how to setup your LLVM environment:

$ sudo -s
# ./perf test -v bpf
37: Test BPF filter                                          :
37.1: Test basic BPF filtering                               :
--- start ---
test child forked, pid 19463
ERROR:    unable to find clang.
Hint:    Try to install latest clang/llvm to support BPF. Check your $PATH
         and 'clang-path' option in [llvm] section of ~/.perfconfig.
         LLVM 3.7 or newer is required. Which can be found from 
http://llvm.org
         You may want to try git trunk:
             git clone http://llvm.org/git/llvm.git
                  and
             git clone http://llvm.org/git/clang.git

Or fetch the latest clang/llvm 3.7 from pre-built llvm
packages for
debian/ubuntu:
http://llvm.org/apt

If you are using old version of clang, change
'clang-bpf-cmd-template'
option in [llvm] section of ~/.perfconfig to:

"$CLANG_EXEC $CLANG_OPTIONS $KERNEL_INC_OPTIONS \
-working-directory $WORKING_DIR -c $CLANG_SOURCE \
-emit-llvm -o - | /path/to/llc -march=bpf -filetype=obj -o -"
(Replace /path/to/llc with path to your llc)

Failed to compile test case: 'Basic BPF llvm compiling test' Unable to get BPF object, fix 'perf test LLVM' first test child finished with -2
---- end ----
Test BPF filter subtest 0: Skip

Set your clang path to ~/.perfconfig

# cat ~/.perfconfig
[llvm]
    clang-path = "/opt/llvm-3.7.0/x86_64-oe-linux-clang"

Then try 'perf test LLVM', fix all failure:

# ./perf test LLVM
35: Test LLVM searching and compiling                        :
35.1: Basic BPF llvm compiling test                          : Ok
35.2: Test kbuild searching                                  : Ok
35.3: Compile source for BPF prologue generation test        : Ok

After that you can try perf test BPF:

# ./perf test BPF
37: Test BPF filter                                          :
37.1: Test basic BPF filtering                               : Ok
37.2: Test BPF prologue generation                           : Ok

Thank you.