From: Kan Liang
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_DATA_PAGE_SIZE.
Signed-off-by: Kan Liang
---
No changes since V1
tools/perf/tests/sample-parsing.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/tools/perf/tests/sample
From: Stephane Eranian
Adds the infrastructure to sample the code address page size.
Introduce a new --code-page-size option for perf record.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
tools/include/uapi/linux/perf_event.h| 4 +++-
tools/perf
From: Kan Liang
Now, "--phys-data" is the only option which impacts the sort order.
A simple "if else" is enough to handle the option. But there will be
more options added, e.g. "--data-page-size", which also impact the sort
order. The code will become too com
From: Stephane Eranian
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_CODE_PAGE_SIZE.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
tools/perf/tests/sample-parsing.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff
the generic support is covered. The large PEBS will be disabled
with this sample type.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
include/linux/perf_event.h | 1 +
include/uapi/linux/perf_event.h | 4 +++-
kernel/events/core.c| 11
From: Kan Liang
Add a new sort option "data_page_size" for --mem-mode sort. With this
option applied, perf can sort and report by sample's data page size.
Here is an example.
perf report --stdio --mem-mode
--sort=comm,symbol,phys_daddr,data_page_size
# To display the perf.data h
C Coverage]
# ... ..
#
#
69.56% dtlb [.] GetTickCount 4K -
17.93% dtlb [.] Calibrate 4K -
-
11.40% dtlb [.] __gettimeofday4K -
-
Signed-off-by: Stephane Eranian
Signed-off-by:
:401095 4K
dtlb mem-loads:uP:4010cc 4K
dtlb mem-loads:uP:440b6f 4K
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V2
tools/perf/Documentation/perf-script.txt | 2 +-
tools/perf/builtin-script.c | 76
From: Kan Liang
Support new sample type PERF_SAMPLE_DATA_PAGE_SIZE for page size.
Add new option --data-page-size to record sample data page size.
Signed-off-by: Kan Liang
---
No changes since V1
tools/include/uapi/linux/perf_event.h| 16 +++-
tools/perf/Documentation/perf
From: Kan Liang
Add option --data-page-size in "perf mem" to record/report data page
size.
Here are some examples.
perf mem --phys-data --data-page-size report -D
# PID, TID, IP, ADDR, PHYS ADDR, DATA PAGE SIZE, LOCAL WEIGHT, DSRC,
# SYMBOL
20134 20134 0xff
From: Kan Liang
Display the data page size if it is available.
Can be configured by the user, for example:
perf script --fields comm,event,phys_addr,data_page_size
dtlb mem-loads:uP:3fec82ea8 4K
dtlb mem-loads:uP:3fec82e90 4K
dtlb mem
Commit-ID: 00ae831dfe4474ef6029558f5eb3ef0332d80043
Gitweb: https://git.kernel.org/tip/00ae831dfe4474ef6029558f5eb3ef0332d80043
Author: Kan Liang
AuthorDate: Fri, 25 Jan 2019 11:59:01 -0800
Committer: Borislav Petkov
CommitDate: Tue, 29 Jan 2019 16:37:35 +0100
x86/cpu: Add Atom
From: Kan Liang
Some PCI uncore PMUs cannot be registered on a 8-socket system (HPE
Superdome Flex).
To understand which Socket the PCI uncore PMUs belong to, perf retrieves
the local Node ID of the uncore device from CPUNODEID(0xC0) of the PCI
configuration space, and the mapping between
From: Kan Liang
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_DATA_PAGE_SIZE.
Signed-off-by: Kan Liang
---
No changes since V1
tools/perf/tests/sample-parsing.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/tools/perf/tests/sample
From: Kan Liang
Add a new sort option "data_page_size" for --mem-mode sort. With this
option applied, perf can sort and report by sample's data page size.
Here is an example.
perf report --stdio --mem-mode
--sort=comm,symbol,phys_daddr,data_page_size
# To display the perf.data h
From: Kan Liang
Support new sample type PERF_SAMPLE_DATA_PAGE_SIZE for page size.
Add new option --data-page-size to record sample data page size.
Signed-off-by: Kan Liang
---
No changes since V1
tools/include/uapi/linux/perf_event.h| 16 +++-
tools/perf/Documentation/perf
From: Stephane Eranian
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_CODE_PAGE_SIZE.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
tools/perf/tests/sample-parsing.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff
From: Kan Liang
Add option --data-page-size in "perf mem" to record/report data page
size.
Here are some examples.
perf mem --phys-data --data-page-size report -D
# PID, TID, IP, ADDR, PHYS ADDR, DATA PAGE SIZE, LOCAL WEIGHT, DSRC,
# SYMBOL
20134 20134 0xff
:401095 4K
dtlb mem-loads:uP:4010cc 4K
dtlb mem-loads:uP:440b6f 4K
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
Changes since V1:
- Use ULL to replace UL for 32bit.
tools/perf/Documentation/perf-script.txt | 2 +-
tools/perf
From: Stephane Eranian
Adds the infrastructure to sample the code address page size.
Introduce a new --code-page-size option for perf record.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
tools/include/uapi/linux/perf_event.h| 4 +++-
tools/perf
C Coverage]
# ... ..
#
#
69.56% dtlb [.] GetTickCount 4K -
17.93% dtlb [.] Calibrate 4K -
-
11.40% dtlb [.] __gettimeofday4K -
-
Signed-off-by: Stephane Eranian
Signed-off-by:
From: Kan Liang
Current perf can report both virtual address and physical address, but
it doesn't report page size. Users have no idea how large the utilized
page is. They cannot promote/demote large pages to optimize memory use.
Add a new sample type for data page size.
Current perf already
From: Kan Liang
Now, "--phys-data" is the only option which impacts the output format
and sort order. A simple "if else" is enough to handle the option.
But there will be more options added, e.g. "--data-page-size", which
also impact the output format and sort o
From: Kan Liang
Display the data page size if it is available.
Can be configured by the user, for example:
perf script --fields comm,event,phys_addr,data_page_size
dtlb mem-loads:uP:3fec82ea8 4K
dtlb mem-loads:uP:3fec82e90 4K
dtlb mem
the generic support is covered. The large PEBS will be disabled
with this sample type.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
No changes since V1
include/linux/perf_event.h | 1 +
include/uapi/linux/perf_event.h | 4 +++-
kernel/events/core.c| 11
From: Kan Liang
A microcode patch is also needed for Goldmont while counter freezing
feature is enabled. Otherwise, there will be some issues, e.g. PMI lost.
Signed-off-by: Kan Liang
---
Changes since V5:
- Apply the new name
arch/x86/events/intel/core.c | 9 +++--
1 file changed, 7
From: Kan Liang
Clean up counter freezing quirk to use the new facility to check for
min microcode revisions.
Rename the counter freezing quirk related functions. Because other
platforms, e.g. Goldmont, also needs to call the quirk.
Only check the boot CPU, assuming models and features
From: Kan Liang
Clean up SNB pebs quirk to use the new facility to check for min
microcode revisions.
Only check the boot CPU, assuming models and features are consistent
over all CPUs.
Signed-off-by: Kan Liang
---
Changes since V5:
- New patch to address Peter's comments.
arch/x86/events
Signed-off-by: Andi Kleen
Signed-off-by: Kan Liang
---
Changes since V5:
- The microcode patch has already been merged into future platforms.
So this patch is actually one-off things. The quirks is still used in
the patch.
Rename pebs_isolated to pebs_no_isolation. The default value is
From: Kan Liang
For bug workarounds or checks, it is useful to check for specific
microcode revisions.
Add a new generic function to match the CPU with stepping.
Add the other function to check the min microcode revisions for
the matched CPU.
A new table format is introduced to facilitate
From: Kan Liang
Add a new sort option "data_page_size" for --mem-mode sort. With this
option applied, perf can sort and report by sample's data page size.
Here is an example.
perf report --stdio --mem-mode
--sort=comm,symbol,phys_daddr,data_page_size
# To display the perf.data h
the generic support is covered. The large PEBS will be disabled
with this sample type.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
include/linux/perf_event.h | 1 +
include/uapi/linux/perf_event.h | 4 +++-
kernel/events/core.c| 11 ++-
3 files changed, 14
From: Kan Liang
Current perf can report both virtual address and physical address, but
it doesn't report page size. Users have no idea how large the utilized
page is. They cannot promote/demote large pages to optimize memory use.
Add a new sample type for data page size.
Current perf already
From: Stephane Eranian
Adds the infrastructure to sample the code address page size.
Introduce a new --code-page-size option for perf record.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
tools/include/uapi/linux/perf_event.h| 4 +++-
tools/perf/Documentation/perf
From: Kan Liang
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_DATA_PAGE_SIZE.
Signed-off-by: Kan Liang
---
tools/perf/tests/sample-parsing.c | 7 ++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/tools/perf/tests/sample-parsing.c
b/tools/perf
From: Kan Liang
Display the data page size if it is available.
Can be configured by the user, for example:
perf script --fields comm,event,phys_addr,data_page_size
dtlb mem-loads:uP:3fec82ea8 4K
dtlb mem-loads:uP:3fec82e90 4K
dtlb mem
C Coverage]
# ... ..
#
#
69.56% dtlb [.] GetTickCount 4K -
17.93% dtlb [.] Calibrate 4K -
-
11.40% dtlb [.] __gettimeofday4K -
-
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
:401095 4K
dtlb mem-loads:uP:4010cc 4K
dtlb mem-loads:uP:440b6f 4K
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
tools/perf/Documentation/perf-script.txt | 2 +-
tools/perf/builtin-script.c | 76
From: Stephane Eranian
Extend sample-parsing test cases to support new sample type
PERF_SAMPLE_CODE_PAGE_SIZE.
Signed-off-by: Stephane Eranian
Signed-off-by: Kan Liang
---
tools/perf/tests/sample-parsing.c | 6 +-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/tools/perf
From: Kan Liang
Add option --data-page-size in "perf mem" to record/report data page
size.
Here are some examples.
perf mem --phys-data --data-page-size report -D
# PID, TID, IP, ADDR, PHYS ADDR, DATA PAGE SIZE, LOCAL WEIGHT, DSRC,
# SYMBOL
20134 20134 0xff
From: Kan Liang
Support new sample type PERF_SAMPLE_DATA_PAGE_SIZE for page size.
Add new option --data-page-size to record sample data page size.
Signed-off-by: Kan Liang
---
tools/include/uapi/linux/perf_event.h| 16 +++-
tools/perf/Documentation/perf-record.txt | 3
From: Kan Liang
Now, "--phys-data" is the only option which impacts the output format
and sort order. A simple "if else" is enough to handle the option.
But there will be more options added, e.g. "--data-page-size", which
also impact the output format and sort o
From: Kan Liang
A ucode patch is also needed for Goldmont while counter freezing feature
is enabled. Otherwise, there will be some issues, e.g. PMI lost.
Signed-off-by: Kan Liang
---
- New patch, merged from https://lkml.org/lkml/2018/10/3/25
arch/x86/events/intel/core.c | 5 +
1 file
From: Kan Liang
For bug workarounds or checks it is useful to check for specific
microcode revisions.
Add a new generic function to match the CPU with stepping.
Add the other function to check the min microcode revisions for
the matched CPU.
A new table format is introduced to facilitate
From: Andi Kleen
KVM added a workaround for PEBS events leaking into guests with
commit 26a4f3c08de4 ("perf/x86: disable PEBS on a guest entry.")
This uses the VT entry/exit list to add an extra disable of the
PEBS_ENABLE MSR.
Intel also added a fix for this issue to microcode updates on
From: Kan Liang
Clean up counter freezing quirk to use the new facility to check for
min microcode revisions.
Rename the counter freezing quirk related functions. Because other
platforms, e.g. Goldmont, also needs to call the quirk.
Signed-off-by: Kan Liang
---
- New patch, merged from https
From: Kan Liang
When a task, which is sampled by a PEBS event with a fixed period, is
sched_in, the fixed period will always be used as new period for
counter. It's inaccurate, because the left period from last sched_out
isn't taken into account.
The auto-reload feature is implicitly enabled
From: Kan Liang
When a task, which is sampled by a PEBS event with a fixed period, is
sched_in, the fixed period will always be used as new period for
counter. It's inaccurate, because the left period from last sched_out
isn't taken into account.
The auto-reload feature is implicitly enabled
From: Kan Liang
In the process of handling a guest overflow, KVM unconditionally
reprograms perf counters before entering guest. The reprogramming brings
very high overhead. For common case, (e.g. vCPU still runs on the same
CPU), it's unnecessary.
Here is current process of handling
From: Kan Liang
In the process of handling a guest overflow, KVM unconditionally
reprograms perf counters before entering guest. The reprogramming brings
very high overhead. For common case, (e.g. vCPU still runs on the same
CPU), it's unnecessary.
Here is current process of handling
Commit-ID: f4a0742b3cc1d03b2ff448017b8c714a77e5a261
Gitweb: https://git.kernel.org/tip/f4a0742b3cc1d03b2ff448017b8c714a77e5a261
Author: Kan Liang
AuthorDate: Wed, 21 Nov 2018 08:49:39 -0800
Committer: Arnaldo Carvalho de Melo
CommitDate: Wed, 21 Nov 2018 22:39:59 -0300
perf pmu: Move
Commit-ID: f4a0742b3cc1d03b2ff448017b8c714a77e5a261
Gitweb: https://git.kernel.org/tip/f4a0742b3cc1d03b2ff448017b8c714a77e5a261
Author: Kan Liang
AuthorDate: Wed, 21 Nov 2018 08:49:39 -0800
Committer: Arnaldo Carvalho de Melo
CommitDate: Wed, 21 Nov 2018 22:39:59 -0300
perf pmu: Move
Commit-ID: 3b54411a4485c9f1d50f003810ecfceca82a09ae
Gitweb: https://git.kernel.org/tip/3b54411a4485c9f1d50f003810ecfceca82a09ae
Author: Kan Liang
AuthorDate: Wed, 14 Nov 2018 13:24:15 -0800
Committer: Arnaldo Carvalho de Melo
CommitDate: Wed, 21 Nov 2018 22:39:57 -0300
perf vendor
Commit-ID: 3b54411a4485c9f1d50f003810ecfceca82a09ae
Gitweb: https://git.kernel.org/tip/3b54411a4485c9f1d50f003810ecfceca82a09ae
Author: Kan Liang
AuthorDate: Wed, 14 Nov 2018 13:24:15 -0800
Committer: Arnaldo Carvalho de Melo
CommitDate: Wed, 21 Nov 2018 22:39:57 -0300
perf vendor
From: Kan Liang
The weak functions, strcmp_cpuid_str() and get_cpuid_str(), are defined
in pmu.c.
Most of the cpuid related functions, including *_cpuid_str()'s
declaration and platform specific definition, are in header.c/h.
To make the declaration and definition of all cpuid related functions
From: Kan Liang
The weak functions, strcmp_cpuid_str() and get_cpuid_str(), are defined
in pmu.c.
Most of the cpuid related functions, including *_cpuid_str()'s
declaration and platform specific definition, are in header.c/h.
To make the declaration and definition of all cpuid related functions
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
The client IMC bandwidth events return very huge result.
perf stat -e uncore_imc/data_reads/ -e uncore_imc/data_writes/ -I
1 -a
10.000117222 34,788.76 MiB uncore_imc/data_reads/
10.000117222 8.26 MiB uncore_imc/data_writes/
20.000374584 34,842.89 MiB uncore_imc
From: Kan Liang
Perf tools cannot find the proper event list for Cascadelake server.
Because Cascadelake server and Skylake server have the same CPU model
number, which are used by perf tools to find the event list.
The stepping for Skylake server is up to 4.
The stepping for Cascadelake server
From: Kan Liang
Perf tools cannot find the proper event list for Cascadelake server.
Because Cascadelake server and Skylake server have the same CPU model
number, which are used by perf tools to find the event list.
The stepping for Skylake server is up to 4.
The stepping for Cascadelake server
Commit-ID: 4d47d6407ac7b4b442a4e717488a3bb137398b6c
Gitweb: https://git.kernel.org/tip/4d47d6407ac7b4b442a4e717488a3bb137398b6c
Author: Kan Liang
AuthorDate: Fri, 19 Oct 2018 10:04:19 -0700
Committer: Ingo Molnar
CommitDate: Mon, 12 Nov 2018 05:03:26 +0100
perf/x86/intel/uncore
Commit-ID: 4d47d6407ac7b4b442a4e717488a3bb137398b6c
Gitweb: https://git.kernel.org/tip/4d47d6407ac7b4b442a4e717488a3bb137398b6c
Author: Kan Liang
AuthorDate: Fri, 19 Oct 2018 10:04:19 -0700
Committer: Ingo Molnar
CommitDate: Mon, 12 Nov 2018 05:03:26 +0100
perf/x86/intel/uncore
Commit-ID: c10a8de0d32e95b0b8c7c17b6dc09baea5a5a899
Gitweb: https://git.kernel.org/tip/c10a8de0d32e95b0b8c7c17b6dc09baea5a5a899
Author: Kan Liang
AuthorDate: Fri, 19 Oct 2018 10:04:18 -0700
Committer: Ingo Molnar
CommitDate: Mon, 12 Nov 2018 05:03:24 +0100
perf/x86/intel/uncore: Add
Commit-ID: c10a8de0d32e95b0b8c7c17b6dc09baea5a5a899
Gitweb: https://git.kernel.org/tip/c10a8de0d32e95b0b8c7c17b6dc09baea5a5a899
Author: Kan Liang
AuthorDate: Fri, 19 Oct 2018 10:04:18 -0700
Committer: Ingo Molnar
CommitDate: Mon, 12 Nov 2018 05:03:24 +0100
perf/x86/intel/uncore: Add
From: Kan Liang
The main event processing thread may hang if the ring buffer event
processing timeouts.
Analysis from David Miller:
"It hangs the event thread, because the ui call waits for a keypress
but the display thread will eat them up and the event thread thus
hangs in s
From: Kan Liang
The main event processing thread may hang if the ring buffer event
processing timeouts.
Analysis from David Miller:
"It hangs the event thread, because the ui call waits for a keypress
but the display thread will eat them up and the event thread thus
hangs in s
From: Kan Liang
The physical addresses for the last several samples are always lost in
large PEBS. For example,
#perf record -e mem-loads:uP --phys-data -c1 -- ./dtlb
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.103 MB perf.data (2661 samples
From: Kan Liang
To calculate the physical address, perf needs to walk the pages tables.
The related mapping may has already been removed from pages table in
some cases (e.g. large PEBS). The virtual address recorded in the first
PEBS records may already be unmapped before draining PEBS buffers
From: Kan Liang
The physical addresses for the last several samples are always lost in
large PEBS. For example,
#perf record -e mem-loads:uP --phys-data -c1 -- ./dtlb
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.103 MB perf.data (2661 samples
From: Kan Liang
To calculate the physical address, perf needs to walk the pages tables.
The related mapping may has already been removed from pages table in
some cases (e.g. large PEBS). The virtual address recorded in the first
PEBS records may already be unmapped before draining PEBS buffers
From: Kan Liang
Coffee Lake has 8 core product which has 8 Cboxes. The 8th CBOX is
mapped into different MSR space.
Increase the num_boxes to 8 to handle the new products. It will not
impact the previous platforms, SkyLake, KabyLake and earlier CoffeeLake.
Because the num_boxes
From: Kan Liang
Coffee Lake has 8 core product which has 8 Cboxes. The 8th CBOX is
mapped into different MSR space.
Increase the num_boxes to 8 to handle the new products. It will not
impact the previous platforms, SkyLake, KabyLake and earlier CoffeeLake.
Because the num_boxes
From: Kan Liang
KabyLake and CoffeeLake has the same client uncore events as SkyLake.
Add the PCI IDs for KabyLake Y, U, S processor line and CoffeeLake U,
H, S processor line.
Signed-off-by: Kan Liang
---
arch/x86/events/intel/uncore_snb.c | 115 -
1 file changed
From: Kan Liang
KabyLake and CoffeeLake has the same client uncore events as SkyLake.
Add the PCI IDs for KabyLake Y, U, S processor line and CoffeeLake U,
H, S processor line.
Signed-off-by: Kan Liang
---
arch/x86/events/intel/uncore_snb.c | 115 -
1 file changed
From: Kan Liang
A ucode patch is also needed for Goldmont while counter freezing feature
is enabled. Otherwise, there will be some issues, e.g. PMI lost.
Add a quirk to check microcode version. If the system starts with the
wrong ucode, leave the counter-freezing feature permanently disabled
From: Kan Liang
A ucode patch is also needed for Goldmont while counter freezing feature
is enabled. Otherwise, there will be some issues, e.g. PMI lost.
Add a quirk to check microcode version. If the system starts with the
wrong ucode, leave the counter-freezing feature permanently disabled
Commit-ID: 7c5314b88da6d5af98239786772a1c44cc5eb67d
Gitweb: https://git.kernel.org/tip/7c5314b88da6d5af98239786772a1c44cc5eb67d
Author: Kan Liang
AuthorDate: Wed, 8 Aug 2018 00:12:08 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 10:14:33 +0200
perf/x86/intel: Add quirk
Commit-ID: 7c5314b88da6d5af98239786772a1c44cc5eb67d
Gitweb: https://git.kernel.org/tip/7c5314b88da6d5af98239786772a1c44cc5eb67d
Author: Kan Liang
AuthorDate: Wed, 8 Aug 2018 00:12:08 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 10:14:33 +0200
perf/x86/intel: Add quirk
Commit-ID: ba12d20edc5caf9835006d8f3efd4ed18465c75b
Gitweb: https://git.kernel.org/tip/ba12d20edc5caf9835006d8f3efd4ed18465c75b
Author: Kan Liang
AuthorDate: Wed, 8 Aug 2018 00:12:06 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 10:14:30 +0200
perf/x86/intel: Factor out
Commit-ID: ba12d20edc5caf9835006d8f3efd4ed18465c75b
Gitweb: https://git.kernel.org/tip/ba12d20edc5caf9835006d8f3efd4ed18465c75b
Author: Kan Liang
AuthorDate: Wed, 8 Aug 2018 00:12:06 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 10:14:30 +0200
perf/x86/intel: Factor out
Commit-ID: 9d92cfeaf5215158d26d2991be7f7ff865cb98f3
Gitweb: https://git.kernel.org/tip/9d92cfeaf5215158d26d2991be7f7ff865cb98f3
Author: Kan Liang
AuthorDate: Fri, 21 Sep 2018 07:07:06 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 09:38:02 +0200
perf/x86/intel/uncore: Fix
Commit-ID: 9d92cfeaf5215158d26d2991be7f7ff865cb98f3
Gitweb: https://git.kernel.org/tip/9d92cfeaf5215158d26d2991be7f7ff865cb98f3
Author: Kan Liang
AuthorDate: Fri, 21 Sep 2018 07:07:06 -0700
Committer: Ingo Molnar
CommitDate: Tue, 2 Oct 2018 09:38:02 +0200
perf/x86/intel/uncore: Fix
From: Kan Liang
The counters on M3UPI Link 0 and Link 3 don't count. Writing 0 to
these counters may causes system crash on some machines.
The PCI BDF addresses of M3UPI in current code are incorrect.
The correct addresses should be
D18:F1 0x204D
D18:F2 0x204E
D18:F5 0x204D
Signed-off
From: Kan Liang
The counters on M3UPI Link 0 and Link 3 don't count. Writing 0 to
these counters may causes system crash on some machines.
The PCI BDF addresses of M3UPI in current code are incorrect.
The correct addresses should be
D18:F1 0x204D
D18:F2 0x204E
D18:F5 0x204D
Signed-off
From: Kan Liang
LBR can bring big overhead when the benchmark has high context switches.
For example, a sub benchmark of Dacapo, avrora.
Baseline: java -jar dacapo-9.12-MR1-bach.jar avrora -n 20
With LBR: perf record --branch-filter any,u -- java -jar
dacapo-9.12-MR1-bach.jar avrora -n 20
From: Kan Liang
LBR can bring big overhead when the benchmark has high context switches.
For example, a sub benchmark of Dacapo, avrora.
Baseline: java -jar dacapo-9.12-MR1-bach.jar avrora -n 20
With LBR: perf record --branch-filter any,u -- java -jar
dacapo-9.12-MR1-bach.jar avrora -n 20
From: Kan Liang
Display the page size if it is available.
Can be configured by the user, for example:
perf script --fields comm,event,phys_addr,page_size
dtlb mem-loads:uP:3fec82ea8 4K
dtlb mem-loads:uP:3fec82e90 4K
dtlb mem-loads:uP
From: Kan Liang
Display the page size if it is available.
Can be configured by the user, for example:
perf script --fields comm,event,phys_addr,page_size
dtlb mem-loads:uP:3fec82ea8 4K
dtlb mem-loads:uP:3fec82e90 4K
dtlb mem-loads:uP
From: Kan Liang
Add a new sort option "page_size" for --mem-mode sort. With this
option applied, perf can sort and report by sample's page size.
Here is an example.
perf report --stdio --mem-mode --sort=comm,symbol,phys_daddr,page_size
# To display the perf.data header info,
From: Kan Liang
Support new sample type PERF_SAMPLE_PAGE_SIZE for page size.
Add new option --page-size to record sample page size.
Signed-off-by: Kan Liang
---
tools/include/uapi/linux/perf_event.h| 13 -
tools/perf/Documentation/perf-record.txt | 3 +++
tools/perf/builtin
From: Kan Liang
Add a new sort option "page_size" for --mem-mode sort. With this
option applied, perf can sort and report by sample's page size.
Here is an example.
perf report --stdio --mem-mode --sort=comm,symbol,phys_daddr,page_size
# To display the perf.data header info,
From: Kan Liang
Support new sample type PERF_SAMPLE_PAGE_SIZE for page size.
Add new option --page-size to record sample page size.
Signed-off-by: Kan Liang
---
tools/include/uapi/linux/perf_event.h| 13 -
tools/perf/Documentation/perf-record.txt | 3 +++
tools/perf/builtin
From: Kan Liang
Now, "--phys-data" is the only option which impacts the output format and
sort order. A simple "if else" is enough to handle the option.
But there will be more options added, e.g. "--page-size", which also
impact the output format and sort order. Th
From: Kan Liang
Now, "--phys-data" is the only option which impacts the output format and
sort order. A simple "if else" is enough to handle the option.
But there will be more options added, e.g. "--page-size", which also
impact the output format and sort order. Th
From: Kan Liang
Add option --page-size in "perf mem" to record/report page size.
Here are some examples.
perf mem --phys-data --page-size report -D
# PID, TID, IP, ADDR, PHYS ADDR, PAGE SIZE, LOCAL WEIGHT, DSRC, SYMBOL
20134 20134 0xb5bd2fd0 0x0169a274e96a308 0x00044
From: Kan Liang
Add option --page-size in "perf mem" to record/report page size.
Here are some examples.
perf mem --phys-data --page-size report -D
# PID, TID, IP, ADDR, PHYS ADDR, PAGE SIZE, LOCAL WEIGHT, DSRC, SYMBOL
20134 20134 0xb5bd2fd0 0x0169a274e96a308 0x00044
1001 - 1100 of 3488 matches
Mail list logo