which streams are matched.
Signed-off-by: Jin Yao
---
v7:
- Rename functions with 'stream__' prefix.
v6:
- Rebase to perf/core
v5:
- Remove enum stream_type
v4:
- New patch in v4.
tools/perf/util/stream.c | 40
tools/perf/uti
main div.c:44
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v7:
- Use new struct evlist_streams in data__file.
- Free the streams in data__free.
v6:
- Rebase to perf/core
v5:
e,
cycles: 2, hits: 4.08%
--
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v7:
- Rename functions with 'evsel_streams__' prefix
v6:
- Rebase to perf/core
v5:
- Rebase to perf/core
v4:
- Remove "Hot chains in
factor the code in a generic way.
v3:
---
v2 has 14 patches, it's hard to review.
v3 is only 7 patches for basic stream comparison.
Jin Yao (7):
perf util: Create streams
perf util: Get the evsel_streams by evsel_idx
perf util: Compare two streams
perf util: Link stream pair
per
We have used callchain_node->hit to measure the hot level of one
stream. This patch calculates the sum of hits of total streams.
Thus in next patch, we can use following formula to report hot
percent for one stream.
hot percent = callchain_node->hit / sum of total hits
Signed-off-by: J
In previous patch, we have created evsel_streams array
This patch returns the specified evsel_streams according to the
evsel_idx.
Signed-off-by: Jin Yao
---
v7:
- Rename functions:
evsel_streams__entry
v6:
- Rebase to perf/core
v5:
- Rebase to perf/core
v4:
- Rename
x27;t consider the case that
source code is changed).
The matching logic is, compare the chain string first. If it's not
matched, fallback to dso address comparison.
Signed-off-by: Jin Yao
---
v7:
- No change
v6:
- Rebase to perf/core
v5:
- Remove enum stream_type
- Rebase to perf
ams array per event, and saves
the top N hottest streams in a stream array.
So now we can get the per-event top N hottest streams.
Signed-off-by: Jin Yao
---
v7:
- Create a new struct evlist_streams.
- Rename functions with appropriate prefix
v6:
- Rebase to perf/core
Hi Arnaldo,
On 9/18/2020 4:26 AM, Arnaldo Carvalho de Melo wrote:
Em Fri, Sep 11, 2020 at 04:03:53PM +0800, Jin Yao escreveu:
This patch enables perf-diff with "--stream" option.
"--stream": Enable hot streams comparison
Now let's see examples.
perf record -b ...
Hi Arnaldo,
On 9/18/2020 4:26 AM, Arnaldo Carvalho de Melo wrote:
Em Fri, Sep 11, 2020 at 04:03:53PM +0800, Jin Yao escreveu:
This patch enables perf-diff with "--stream" option.
"--stream": Enable hot streams comparison
Now let's see examples.
perf record -b ...
Hi Arnaldo,
On 9/18/2020 4:13 AM, Arnaldo Carvalho de Melo wrote:
Em Thu, Sep 17, 2020 at 03:05:56PM +0200, Jiri Olsa escreveu:
On Fri, Sep 11, 2020 at 04:03:46PM +0800, Jin Yao wrote:
SNIP
main div.c:40
main div.c:40
main
Hi Ian,
On 9/18/2020 9:36 AM, Ian Rogers wrote:
On Thu, Sep 17, 2020, 6:21 PM Jin, Yao wrote:
Hi Arnaldo,
On 9/7/2020 4:01 PM, Jin, Yao wrote:
Hi Arnaldo,
On 7/17/2020 2:06 PM, Jin, Yao wrote:
Hi,
On 6/17/2020 3:38 AM, Arnaldo Carvalho de Melo wrote:
Em Tue, Jun 16, 2020 at 02:27:40PM
Hi Arnaldo,
On 9/7/2020 4:01 PM, Jin, Yao wrote:
Hi Arnaldo,
On 7/17/2020 2:06 PM, Jin, Yao wrote:
Hi,
On 6/17/2020 3:38 AM, Arnaldo Carvalho de Melo wrote:
Em Tue, Jun 16, 2020 at 02:27:40PM +0800, Jin, Yao escreveu:
On 6/16/2020 2:16 PM, Ian Rogers wrote:
On Mon, Jun 15, 2020 at 6:00 PM
In previous patch, we have created evsel_streams array
This patch returns the specified evsel_streams according to the
evsel_idx.
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Rebase to perf/core
v4:
- Rename the patch from 'perf util: Return per-event call
urce line based comparison. Now we
only supports the basic functionality of stream comparison.
2. Refactor the code in a generic way.
v3:
---
v2 has 14 patches, it's hard to review.
v3 is only 7 patches for basic stream comparison.
Jin Yao (7):
perf util: Create streams
perf util: Get
ams array per event, and saves
the top N hottest streams in a stream array.
So now we can get the per-event top N hottest streams.
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Remove enum stram_type
- Rebase to perf/core
v4:
- Refactor the code
- Rename patch n
We have used callchain_node->hit to measure the hot level of one
stream. This patch calculates the sum of hits of total streams.
Thus in next patch, we can use following formula to report hot
percent for one stream.
hot percent = callchain_node->hit / sum of total hits
Signed-off-by: J
x27;t consider the case that
source code is changed).
The matching logic is, compare the chain string first. If it's not
matched, fallback to dso address comparison.
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Remove enum stream_type
- Rebase to perf/core
v4:
- Remov
main div.c:44
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Remove enum stream_type
- Rebase to perf/core
v4:
- Remove the "--before" a
e,
cycles: 2, hits: 4.08%
--
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Rebase to perf/core
v4:
- Remove "Hot chains in old perf data but source line changed
in new perf data&q
which streams are matched.
Signed-off-by: Jin Yao
---
v6:
- Rebase to perf/core
v5:
- Remove enum stream_type
v4:
- New patch in v4.
tools/perf/util/stream.c | 40
tools/perf/util/stream.h | 4
2 files changed, 44 insertions(+)
diff
Hi Jiri,
On 9/10/2020 5:22 PM, Jiri Olsa wrote:
On Mon, Sep 07, 2020 at 12:15:59PM +0800, Jin Yao wrote:
SNIP
main div.c:39
[ Hot streams in new perf data only ]
hot chain 1:
cycles: 4, hits: 4.54
Hi Arnaldo,
On 7/17/2020 2:06 PM, Jin, Yao wrote:
Hi,
On 6/17/2020 3:38 AM, Arnaldo Carvalho de Melo wrote:
Em Tue, Jun 16, 2020 at 02:27:40PM +0800, Jin, Yao escreveu:
On 6/16/2020 2:16 PM, Ian Rogers wrote:
On Mon, Jun 15, 2020 at 6:00 PM Jin, Yao wrote:
Can I get an ACK for this
x27;t consider the case that
source code is changed).
The matching logic is, compare the chain string first. If it's not
matched, fallback to dso address comparison.
Signed-off-by: Jin Yao
---
v5:
- Remove enum stream_type
- Rebase to perf/core
v4:
- Remove original source line compa
main div.c:44
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v5:
- Remove enum stream_type
- Rebase to perf/core
v4:
- Remove the "--before" and "--after" opti
ams array per event, and saves
the top N hottest streams in a stream array.
So now we can get the per-event top N hottest streams.
Signed-off-by: Jin Yao
---
v5:
- Remove enum stram_type
- Rebase to perf/core
v4:
- Refactor the code
- Rename patch name from 'perf util: Create
e,
cycles: 2, hits: 4.08%
--
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v5:
- Rebase to perf/core
v4:
- Remove "Hot chains in old perf data but source line changed
in new perf data"
tools/perf/util/callchain.c
We have used callchain_node->hit to measure the hot level of one
stream. This patch calculates the sum of hits of total streams.
Thus in next patch, we can use following formula to report hot
percent for one stream.
hot percent = callchain_node->hit / sum of total hits
Signed-off-by: J
only supports the basic functionality of stream comparison.
2. Refactor the code in a generic way.
v3:
---
v2 has 14 patches, it's hard to review.
v3 is only 7 patches for basic stream comparison.
Jin Yao (7):
perf util: Create streams
perf util: Get the evsel_streams by evsel_idx
which streams are matched.
Signed-off-by: Jin Yao
---
v5:
- Remove enum stream_type
v4:
- New patch in v4.
tools/perf/util/stream.c | 40
tools/perf/util/stream.h | 4
2 files changed, 44 insertions(+)
diff --git a/tools/perf/util/stream.c
In previous patch, we have created evsel_streams array
This patch returns the specified evsel_streams according to the
evsel_idx.
Signed-off-by: Jin Yao
---
v5:
- Rebase to perf/core
v4:
- Rename the patch from 'perf util: Return per-event callchain
streams' to 'per
6124542 seconds time elapsed
Fixes: c7e5b328a8d4 ("perf stat: Report summary for interval mode")
Signed-off-by: Jin Yao
---
tools/perf/Documentation/perf-stat.txt | 3 +++
tools/perf/builtin-stat.c | 8 +---
tools/perf/util/stat.h | 1 +
3 files changed, 9 inse
Hi Jiri,
On 9/2/2020 4:09 AM, Jiri Olsa wrote:
On Tue, Sep 01, 2020 at 10:26:25AM +0800, Jin, Yao wrote:
Hi Jiri,
On 8/31/2020 9:56 PM, Jiri Olsa wrote:
On Tue, Aug 25, 2020 at 07:35:07AM +0800, Jin Yao wrote:
SNIP
+ int nr_streams_max
Hi Jiri,
On 8/31/2020 9:56 PM, Jiri Olsa wrote:
On Tue, Aug 25, 2020 at 07:35:07AM +0800, Jin Yao wrote:
SNIP
+ int nr_streams_max,
+ enum stream_type type)
+{
+ struct evsel_streams *es
Hi Arnaldo,
On 8/19/2020 11:05 AM, Like Xu wrote:
Hi Arnaldo,
On 2020/8/17 10:32, Jin, Yao wrote:
Hi Arnaldo,
On 8/14/2020 8:47 PM, Arnaldo Carvalho de Melo wrote:
Em Fri, Aug 14, 2020 at 09:21:20AM +0800, Jin Yao escreveu:
Currently if we run 'perf record -e cycles:u', exclu
We have used callchain_node->hit to measure the hot level of one
stream. This patch calculates the sum of hits of total streams.
Thus in next patch, we can use following formula to report hot
percent for one stream.
hot percent = callchain_node->hit / sum of total hits
Signed-off-by: J
main div.c:44
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v4:
- Remove the "--before" and "--after" options since they are for
source line based comparison. In t
which streams are matched.
Signed-off-by: Jin Yao
---
v4:
- New patch in v4.
tools/perf/util/stream.c | 53
tools/perf/util/stream.h | 4 +++
2 files changed, 57 insertions(+)
diff --git a/tools/perf/util/stream.c b/tools/perf/util/stream.c
index
x27;t consider the case that
source code is changed).
The matching logic is, compare the chain string first. If it's not
matched, fallback to dso address comparison.
Signed-off-by: Jin Yao
---
v4:
- Remove original source line comparison code.
tools/perf/util/callch
ams array per event, and saves
the top N hottest streams in a stream array.
So now we can get the per-event top N hottest streams.
Signed-off-by: Jin Yao
---
v4:
- Refactor the code
- Rename patch name from 'perf util: Create streams for managing
top N hottest callchains'
e,
cycles: 2, hits: 4.08%
--
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
v4:
- Remove "Hot chains in old perf data but source line changed
in new perf data"
tools/perf/util/callchain.c | 13
tools/perf/util/callc
2. Refactor the code in a generic way.
v3:
---
v2 has 14 patches, it's hard to review.
v3 is only 7 patches for basic stream comparison.
Jin Yao (7):
perf util: Create streams
perf util: Get the evsel_streams by evsel_idx
perf util: Compare two streams
perf util: Link stream pair
In previous patch, we have created evsel_streams array
This patch returns the specified evsel_streams according to the
evsel_idx.
Signed-off-by: Jin Yao
---
v4:
- Rename the patch from 'perf util: Return per-event callchain
streams' to 'perf util: Get the evsel_strea
Hi Arnaldo,
On 8/14/2020 8:47 PM, Arnaldo Carvalho de Melo wrote:
Em Fri, Aug 14, 2020 at 09:21:20AM +0800, Jin Yao escreveu:
Currently if we run 'perf record -e cycles:u', exclude_guest=0.
But it doesn't make sense in most cases that we request for
user-space counting but
1
For Before/After, exclude_guest are both 0 for perf kvm usage.
perf test 6
6: Parse event definition strings : Ok
Signed-off-by: Jin Yao
---
v3:
---
For perf kvm, if we have 'exclude_guest=1', we can't get guest
events. So we don't set 'exclude_gue
Signed-off-by: Jin Yao
---
v2:
---
Fix the 'perf test 6' failure.
tools/perf/tests/parse-events.c | 4 ++--
tools/perf/util/parse-events.c | 2 ++
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/tools/perf/tests/parse-events.c b/tools/perf/tests/parse-events.c
index 7
On 8/13/2020 2:57 PM, Like Xu wrote:
Hi Yao,
On 2020/8/13 11:11, Jin, Yao wrote:
Hi Like,
On 8/12/2020 9:02 PM, Like Xu wrote:
On 2020/8/12 20:15, Arnaldo Carvalho de Melo wrote:
Em Wed, Aug 12, 2020 at 02:59:53PM +0800, Jin Yao escreveu:
Currently if we run 'perf record -e cyc
Hi Arnaldo,
On 8/12/2020 8:55 PM, Arnaldo Carvalho de Melo wrote:
Em Wed, Aug 12, 2020 at 09:15:04AM -0300, Arnaldo Carvalho de Melo escreveu:
Em Wed, Aug 12, 2020 at 02:59:53PM +0800, Jin Yao escreveu:
Currently if we run 'perf record -e cycles:u', exclude_guest is 0.
But it do
Hi Like,
On 8/12/2020 9:02 PM, Like Xu wrote:
On 2020/8/12 20:15, Arnaldo Carvalho de Melo wrote:
Em Wed, Aug 12, 2020 at 02:59:53PM +0800, Jin Yao escreveu:
Currently if we run 'perf record -e cycles:u', exclude_guest is 0.
But it doesn't make sense that we request for user
Currently if we run 'perf record -e cycles:u', exclude_guest is 0.
But it doesn't make sense that we request for user-space counting
but we also get the guest report.
To keep perf semantics consistent and clear, this patch sets
exclude_guest for user-space counting.
Signed-
Hi Peter,
On 8/11/2020 4:45 PM, Peter Zijlstra wrote:
On Tue, Aug 11, 2020 at 04:31:10PM +0800, Jin, Yao wrote:
Hi Peter,
On 8/11/2020 3:59 PM, Peter Zijlstra wrote:
On Tue, Aug 11, 2020 at 03:50:43PM +0800, Jin, Yao wrote:
Could I post v2 which basically refers to your patch but removes
Hi Peter,
On 8/11/2020 3:59 PM, Peter Zijlstra wrote:
On Tue, Aug 11, 2020 at 03:50:43PM +0800, Jin, Yao wrote:
Could I post v2 which basically refers to your patch but removes some
conditions since I see some issues in test if we use these conditions.
1. Remove '!event->attr.ex
Hi Peter,
On 8/6/2020 10:26 AM, Jin, Yao wrote:
Hi Peter,
On 8/5/2020 8:44 PM, pet...@infradead.org wrote:
On Wed, Aug 05, 2020 at 10:15:26AM +0800, Jin, Yao wrote:
Hi Peter,
On 8/4/2020 7:49 PM, pet...@infradead.org wrote:
On Fri, Jul 31, 2020 at 10:56:17AM +0800, Jin Yao wrote
Hi Peter,
On 8/7/2020 5:02 PM, pet...@infradead.org wrote:
On Fri, Aug 07, 2020 at 02:24:30PM +0800, Jin, Yao wrote:
Hi Peter,
On 8/6/2020 7:00 PM, pet...@infradead.org wrote:
On Thu, Aug 06, 2020 at 11:18:27AM +0200, pet...@infradead.org wrote:
Suppose we have nested virt:
L0-hv
escreveu:
Em Thu, Aug 06, 2020 at 09:43:57PM +0200, Jiri Olsa escreveu:
On Wed, Aug 05, 2020 at 10:29:37AM +0800, Jin Yao wrote:
We received an error report that perf-record caused 'Segmentation fault'
on a newly system (e.g. on the new installed ubuntu).
(gdb) backtrace
#0 __read_once_
Hi Jiri,
On 8/7/2020 3:43 AM, Jiri Olsa wrote:
On Wed, Aug 05, 2020 at 10:29:37AM +0800, Jin Yao wrote:
We received an error report that perf-record caused 'Segmentation fault'
on a newly system (e.g. on the new installed ubuntu).
(gdb) backtrace
#0 __read_once_size (size=4, re
__set_bit(idx, (unsigned long *)&cpuc->intel_ctrl_guest_mask);
How can we know it's guest or host even if exclude_host is set in guest?
Thanks
Jin Yao
Also, ARM64 seems to also implement this stuff, Mark, do you have any
insight on how all this is 'supposed' to work?
Hi Peter,
On 8/6/2020 5:24 PM, pet...@infradead.org wrote:
On Thu, Aug 06, 2020 at 11:18:27AM +0200, pet...@infradead.org wrote:
On Thu, Aug 06, 2020 at 10:26:29AM +0800, Jin, Yao wrote:
+static struct pt_regs *sanitize_sample_regs(struct perf_event *event, struct
pt_regs *regs
Hi Peter,
On 8/6/2020 5:18 PM, pet...@infradead.org wrote:
On Thu, Aug 06, 2020 at 10:26:29AM +0800, Jin, Yao wrote:
+static struct pt_regs *sanitize_sample_regs(struct perf_event *event, struct
pt_regs *regs)
+{
+ struct pt_regs *sample_regs = regs;
+
+ /* user only
Hi Peter,
On 8/5/2020 8:44 PM, pet...@infradead.org wrote:
On Wed, Aug 05, 2020 at 10:15:26AM +0800, Jin, Yao wrote:
Hi Peter,
On 8/4/2020 7:49 PM, pet...@infradead.org wrote:
On Fri, Jul 31, 2020 at 10:56:17AM +0800, Jin Yao wrote:
@@ -6973,7 +6973,8 @@ static struct perf_callchain_entry
o skip the side-band event setup if HAVE_LIBBPF_SUPPORT
is not set.
Signed-off-by: Jin Yao
---
tools/perf/builtin-record.c | 7 +++
1 file changed, 7 insertions(+)
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index b6bdccd875bc..ae97f98e2753 100644
--- a/tools/pe
Hi Peter,
On 8/4/2020 7:49 PM, pet...@infradead.org wrote:
On Fri, Jul 31, 2020 at 10:56:17AM +0800, Jin Yao wrote:
@@ -6973,7 +6973,8 @@ static struct perf_callchain_entry __empty_callchain = {
.nr = 0, };
struct perf_callchain_entry *
perf_callchain(struct perf_event *event, struct
Hi Peter,
On 8/4/2020 7:31 PM, pet...@infradead.org wrote:
On Fri, Jul 31, 2020 at 10:56:16AM +0800, Jin Yao wrote:
There is a potential security issue that perf kernel samples
may be leaked even though kernel sampling is disabled. For fixing
the potential leakage, the idea is to use
3.97% div div[.] rand@plt
0.00% div ld-2.27.so [.] __GI___tunables_init
0.00% div ld-2.27.so [.] _start
Now there is no kernel address leaked.
Inspired-by: Mark Rutland
Signed-off-by: Jin Yao
---
kernel/events/core.c | 48 ++
on some architectures.
Define instruction_pointer_set for these architectures.
Signed-off-by: Jin Yao
---
arch/alpha/include/asm/ptrace.h | 6 ++
arch/arc/include/asm/ptrace.h| 6 ++
arch/nds32/include/asm/ptrace.h | 7 +++
arch/xtensa/include/asm/ptrace.h | 6 ++
4 files
Hi Adrian,
Could you help to check if following condition will break PT?
"(opts->sample_intr_regs && !evsel->no_aux_samples &&
!evsel__is_dummy_event(evsel))"
Thanks
Jin Yao
On 7/23/2020 9:01 AM, Jin, Yao wrote:
Hi Jiri, Adrian,
On 7/22/2020 7:08 PM, Jiri O
On 7/27/2020 8:21 AM, Jin, Yao wrote:
On 7/26/2020 3:52 PM, Jiri Olsa wrote:
Jin Yao reported issue with possible conflict between raw
events and term values in pmu event syntax.
Currently following syntax is resolved as raw event with
0xead value:
uncore_imc_free_running/read
On 7/26/2020 3:52 PM, Jiri Olsa wrote:
Jin Yao reported issue with possible conflict between raw
events and term values in pmu event syntax.
Currently following syntax is resolved as raw event with
0xead value:
uncore_imc_free_running/read/
instead of using 'read'
Hi Jiri, Adrian,
On 7/22/2020 7:08 PM, Jiri Olsa wrote:
On Wed, Jul 22, 2020 at 01:00:03PM +0800, Jin, Yao wrote:
SNIP
If we use -IXMM0, the attr>sample_regs_intr will be set with
PERF_REG_EXTENDED_MASK bit.
It doesn't make sense to set attr->sample_regs_intr for a
software
Hi Jiri,
On 7/20/2020 5:17 PM, Jiri Olsa wrote:
On Mon, Jul 20, 2020 at 09:00:13AM +0800, Jin Yao wrote:
Since commit 0a892c1c9472 ("perf record: Add dummy event during system wide
synthesis"),
a dummy event is added to capture mmaps.
But if we run perf-record as,
# perf
p 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.413 MB perf.data (45 samples) ]
v2:
---
Rebase to perf/core
Fixes: 0a892c1c9472 ("perf record: Add dummy event during system wide
synthesis")
Signed-off-by: Jin Yao
---
tools/perf/util/evsel.c
On 7/17/2020 4:24 PM, Jiri Olsa wrote:
On Fri, Jul 17, 2020 at 11:33:46AM +0800, Jin, Yao wrote:
Hi,
On 7/6/2020 8:55 AM, Jin, Yao wrote:
Hi Ian,
On 7/6/2020 8:47 AM, Ian Rogers wrote:
On Fri, Jul 3, 2020 at 5:31 PM Jin, Yao wrote:
Hi Jiri,
On 7/3/2020 7:00 PM, Jiri Olsa wrote:
On
Hi,
On 6/17/2020 3:38 AM, Arnaldo Carvalho de Melo wrote:
Em Tue, Jun 16, 2020 at 02:27:40PM +0800, Jin, Yao escreveu:
On 6/16/2020 2:16 PM, Ian Rogers wrote:
On Mon, Jun 15, 2020 at 6:00 PM Jin, Yao wrote:
Can I get an ACK for this patchset?
On 6/3/2020 10:18 AM, Jin Yao wrote:
This
Hi,
On 7/6/2020 8:55 AM, Jin, Yao wrote:
Hi Ian,
On 7/6/2020 8:47 AM, Ian Rogers wrote:
On Fri, Jul 3, 2020 at 5:31 PM Jin, Yao wrote:
Hi Jiri,
On 7/3/2020 7:00 PM, Jiri Olsa wrote:
On Fri, Jul 03, 2020 at 08:42:15AM +0800, Jin Yao wrote:
Since commit 0a892c1c9472 ("perf record
Hi Ian,
On 7/6/2020 8:47 AM, Ian Rogers wrote:
On Fri, Jul 3, 2020 at 5:31 PM Jin, Yao wrote:
Hi Jiri,
On 7/3/2020 7:00 PM, Jiri Olsa wrote:
On Fri, Jul 03, 2020 at 08:42:15AM +0800, Jin Yao wrote:
Since commit 0a892c1c9472 ("perf record: Add dummy event during system wide
synthesis
Hi Jiri,
On 7/3/2020 7:00 PM, Jiri Olsa wrote:
On Fri, Jul 03, 2020 at 08:42:15AM +0800, Jin Yao wrote:
Since commit 0a892c1c9472 ("perf record: Add dummy event during system wide
synthesis"),
a dummy event is added to capture mmaps.
But if we run perf-record as,
# perf record -
rd: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.413 MB perf.data (45 samples) ]
Fixes: 0a892c1c9472 ("perf record: Add dummy event during system wide
synthesis")
Signed-off-by: Jin Yao
---
tools/perf/util/evsel.c | 4 ++--
1 file changed, 2 insertions(+), 2
Hi Ian, Arnaldo,
On 6/16/2020 2:16 PM, Ian Rogers wrote:
On Mon, Jun 15, 2020 at 6:00 PM Jin, Yao wrote:
Hi Andi,
Can I get an ACK for this patchset?
Thanks
Jin Yao
On 6/3/2020 10:18 AM, Jin Yao wrote:
This patchset updates CascadelakeX events to v1.08 and
updates SkylakeX events to
Hi Andi,
Can I get an ACK for this patchset?
Thanks
Jin Yao
On 6/3/2020 10:18 AM, Jin Yao wrote:
This patchset updates CascadelakeX events to v1.08 and
updates SkylakeX events to v1.21.
The events have been tested on CascadelakeX and SkylakeX
servers with latest perf/core branch.
Jin Yao (2
This patchset updates CascadelakeX events to v1.08 and
updates SkylakeX events to v1.21.
The events have been tested on CascadelakeX and SkylakeX
servers with latest perf/core branch.
Jin Yao (2):
perf vendor events: Update CascadelakeX events to v1.08
perf vendor events: Update SkylakeX
Hi Jiri,
On 5/28/2020 12:28 AM, Jiri Olsa wrote:
On Wed, May 27, 2020 at 09:49:11PM +0800, Jin, Yao wrote:
Hi Jiri,
On 5/27/2020 6:28 PM, Jiri Olsa wrote:
On Wed, May 27, 2020 at 02:31:03PM +0800, Jin, Yao wrote:
SNIP
Thanks
Jin Yao
Issue is found!
It looks we can't set "p
Hi Jiri,
On 5/27/2020 6:28 PM, Jiri Olsa wrote:
On Wed, May 27, 2020 at 02:31:03PM +0800, Jin, Yao wrote:
SNIP
Thanks
Jin Yao
Issue is found!
It looks we can't set "pos->leader = pos" in either for_each_group_member()
or in for_each_group_evsel() because it may
Hi Jiri,
On 5/27/2020 2:31 PM, Jin, Yao wrote:
Hi Jiri,
On 5/27/2020 11:20 AM, Jin, Yao wrote:
Hi Jiri,
On 5/26/2020 7:51 PM, Jiri Olsa wrote:
On Mon, May 25, 2020 at 02:55:58PM +0800, Jin Yao wrote:
SNIP
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index 2a9de6491700
Hi Jiri,
On 5/27/2020 11:20 AM, Jin, Yao wrote:
Hi Jiri,
On 5/26/2020 7:51 PM, Jiri Olsa wrote:
On Mon, May 25, 2020 at 02:55:58PM +0800, Jin Yao wrote:
SNIP
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index 2a9de6491700..1161cffc0688 100644
--- a/tools/perf/util
Hi Jiri,
On 5/26/2020 7:51 PM, Jiri Olsa wrote:
On Mon, May 25, 2020 at 02:55:58PM +0800, Jin Yao wrote:
SNIP
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index 2a9de6491700..1161cffc0688 100644
--- a/tools/perf/util/evlist.c
+++ b/tools/perf/util/evlist.c
@@ -1704,3
main div.c:44
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
tools/perf/Documentation/perf-diff.txt | 4 +
tools/perf/builtin-diff.c | 133 ++---
2 files chang
In previous patch, we have created evsel_streams array
This patch returns the specified evsel_streams according to the
evsel_idx.
Signed-off-by: Jin Yao
---
tools/perf/util/stream.c | 11 +++
tools/perf/util/stream.h | 3 +++
2 files changed, 14 insertions(+)
diff --git a/tools/perf
e,
cycles: 2, hits: 4.08%
--
main div.c:42
compute_flag div.c:28
Signed-off-by: Jin Yao
---
tools/perf/util/callchain.c | 13
tools/perf/util/callchain.h | 2 +
tools/perf/util/stream.c| 125
tools/perf/util
way.
v3:
---
v2 has 14 patches, it's hard to review.
v3 is only 7 patches for basic stream comparison.
Jin Yao (7):
perf util: Create streams
perf util: Get the evsel_streams by evsel_idx
perf util: Compare two streams
perf util: Link stream pair
perf util: Calculate the sum of
x27;t consider the case that
source code is changed).
The matching logic is, compare the chain string first. If it's not
matched, fallback to dso address comparison.
Signed-off-by: Jin Yao
---
tools/perf/util/callchain.c | 54 +
tools/perf/util/callch
We have used callchain_node->hit to measure the hot level of one
stream. This patch calculates the sum of hits of total streams.
Then in next patch, we can use following formula to report hot
percent for one stream.
hot percent = callchain_node->hit / sum of total hits
Signed-off-by: J
which streams are matched.
Signed-off-by: Jin Yao
---
tools/perf/util/stream.c | 53
tools/perf/util/stream.h | 4 +++
2 files changed, 57 insertions(+)
diff --git a/tools/perf/util/stream.c b/tools/perf/util/stream.c
index 7f538d1085ef..76896a790798
p N hottest streams in a stream array.
So now we can get the per-event top N hottest streams.
Signed-off-by: Jin Yao
---
tools/perf/util/Build| 1 +
tools/perf/util/stream.c | 152 +++
tools/perf/util/stream.h | 30
3 files changed, 183 inserti
;{A,B}','{C,D,E}',F.
Fixes: 6a4bb04caacc8 ("perf tools: Enable grouping logic for parsed events")
Signed-off-by: Jin Yao
---
tools/perf/builtin-stat.c | 3 +++
tools/perf/util/evlist.c | 49 +++
tools/perf/util/evlist.h | 5
The evlist may consist of some event + group combinations.
For example, perf stat -e "A,{B,C},D".
This patch testes the event in evlist to see if it has the
correct leader and correct nr_members.
Signed-off-by: Jin Yao
---
tools/perf/tests/parse-eve
Hi Jiri,
On 5/22/2020 5:53 PM, Jiri Olsa wrote:
On Thu, May 21, 2020 at 02:22:40PM +0800, Jin Yao wrote:
SNIP
---
tools/perf/builtin-stat.c | 3 +++
tools/perf/util/evlist.c | 32
tools/perf/util/evlist.h | 5 +
3 files changed, 40 insertions
alid.
So when get_group_fd tries to access FD(leader, 1, 0), access
violation will happen.
This patch ensures that the grouped events with same cpu maps
before we go to get_group_fd.
If the cpu maps are not matched, we force to disable the group.
Fixes: 6a4bb04caacc8 ("perf tools: E
Hi Jiri,
On 5/20/2020 3:50 PM, Jiri Olsa wrote:
On Wed, May 20, 2020 at 01:36:40PM +0800, Jin, Yao wrote:
Hi Jiri,
On 5/18/2020 11:28 AM, Jin, Yao wrote:
Hi Jiri,
On 5/15/2020 4:33 PM, Jiri Olsa wrote:
On Fri, May 15, 2020 at 02:04:57PM +0800, Jin, Yao wrote:
SNIP
I think I get the root
Hi Jiri,
On 5/18/2020 11:28 AM, Jin, Yao wrote:
Hi Jiri,
On 5/15/2020 4:33 PM, Jiri Olsa wrote:
On Fri, May 15, 2020 at 02:04:57PM +0800, Jin, Yao wrote:
SNIP
I think I get the root cause. That should be a serious bug in get_group_fd,
access violation!
For a group mixed with system-wide
83 ("perf stat: Update or print per-thread stats")
Signed-off-by: Jin Yao
---
tools/perf/builtin-stat.c | 70 +++
1 file changed, 41 insertions(+), 29 deletions(-)
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index e0c1ad23c768..f
201 - 300 of 1001 matches
Mail list logo