Unbound workqueues are going to be NUMA-affine. Add wq_numa_tbl_len
and wq_numa_possible_cpumask[] in preparation. The former is the
highest NUMA node ID + 1 and the latter is masks of possibles CPUs for
each NUMA node.
This patch only introduces these. Future patches will make use of
them.
v2: NUMA initialization move into wq_numa_init(). Also, the possible
cpumask array is not created if there aren't multiple nodes on the
system. wq_numa_enabled bool added.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
The scheduled NUMA affinity support for unbound workqueues would need
to walk workqueues list and pool related operations on each workqueue.
Move wq_pool_mutex locking out of get/put_unbound_pool() to their
callers so that pool operations can be performed while walking the
workqueues list, which is also protected by wq_pool_mutex.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Lai Jiangshan <laijs@cn.fujitsu.com>
29c91e9912 ("workqueue: implement attribute-based unbound worker_pool
management") implemented attrs based worker_pool matching. It tried
to avoid false negative when comparing cpumasks with custom hash
function; unfortunately, the hash and comparison functions fail to
ignore CPUs which are not possible. It incorrectly assumed that
bitmap_copy() skips leftover bits in the last word of bitmap and
cpumask_equal() ignores impossible CPUs.
This patch updates attrs->cpumask handling such that impossible CPUs
are properly ignored.
* Hash and copy functions no longer do anything special. They expect
their callers to clear impossible CPUs.
* alloc_workqueue_attrs() initializes the cpumask to cpu_possible_mask
instead of setting all bits and explicit cpumask_setall() for
unbound_std_wq_attrs[] in init_workqueues() is dropped.
* apply_workqueue_attrs() is now responsible for ignoring impossible
CPUs. It makes a copy of @attrs and clears impossible CPUs before
doing anything else.
Signed-off-by: Tejun Heo <tj@kernel.org>
8864b4e59 ("workqueue: implement get/put_pwq()") implemented pwq
(pool_workqueue) refcnting which frees workqueue when the last pwq
goes away. It determined whether it was the last pwq by testing
wq->pwqs is empty. Unfortunately, the test was done outside wq->mutex
and multiple pwq release could race and try to free wq multiple times
leading to oops.
Test wq->pwqs emptiness while holding wq->mutex.
Signed-off-by: Tejun Heo <tj@kernel.org>
Switch to using SIMPLE_DEV_PM_OPS macro to declare the driver's
pm_ops. It reduces code size. Also, CONFIG_PM_SLEEP is added to
suspend/ resume functions to prevent build warnings when
CONFIG_PM_SLEEP is not selected.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Add CONFIG_PM_SLEEP to suspend/resume functions to fix the following
build warning when CONFIG_PM_SLEEP is not selected. This is because
sleep PM callbacks defined by SIMPLE_DEV_PM_OPS are only used when
the CONFIG_PM_SLEEP is enabled.
drivers/leds/leds-bd2802.c:766:12: warning: 'bd2802_suspend' defined but not used [-Wunused-function]
drivers/leds/leds-bd2802.c:776:12: warning: 'bd2802_resume' defined but not used [-Wunused-function]
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Now LP55xx provides automatic clock detection API, lp55xx_is_extclk_used().
The clock configuration can be done by the driver itself.
(a) Concept
The default value is set by each driver with clock selection.
The internal clock selection bit is updated in case that the external clock
is not detected or clock rate is not 32KHz.
(b) Change on LP55xx platform data
The clock configuration is done automatically, so no need to define
'update_config' in the platform side.
Correlated information are removed in the documentations and header.
(c) Definitions moved from header to driver files
CONFIG register values are moved each driver, LP5521 and LP5562.
Not necessary definitions are removed also.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Program execution is timed with 32768Hz clock in the LP55xx family devices.
To run LED functionalities, LP55xx devices provide two options.
One is using internal clock. The other is using external clock.
This patch enables external clock detection automatically.
If external clock is not detected, then the internal clock will be used in the
LP55xx driver.
Valid clock rate is 32768Hz in LP55xx devices.
This new API is used in each LP55xx driver like LP5521 and LP5562.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
This patch fixes a regression introduced by commit 72052fcc10
("leds: leds-ns2: add device tree binding").
When the driver is initialized with device tree data, platform_data
pointer is NULL. This causes a kernel oops at removal.
To fix this bug, num_leds is moved into driver_data and platform_data
is not longer used from ns2_led_remove().
Signed-off-by: Simon Guinot <simon.guinot@sequanux.org>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Call to led_pwm_set() can happen inside atomic context, like triggers.
If the PWM call can sleep, defer using a worker.
Signed-off-by: Florian Vaussard <florian.vaussard@epfl.ch>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
According to a sysfs documentation(Documentation/filesystem/sysfs.txt),
scnprintf() should be used in a read operation method.
It guarantees safe buffer size(PAGE_SIZE) which is allocated by the sysfs.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
LM355x and LM3642 support flash and torch functionality.
(Camera driver) (LED trigger for camera) (LED driver)
Turn on the flash ...> ledtrig_flash_ctrl(true) ...> LM355x or LM3642
brightness ctrl
<actual driving here>
Flash/torch LEDs are controlled by other driver using LED camera trigger
APIs, ledtrig_flash_ctrl()/ledtrig_torch_ctrl().
Then, actual device control is activated by each LED driver such like
LM355x or LM3642.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Some LED devices support flash/torch functionality through the LED subsystem.
This patch enables direct LED trigger controls by the driver.
Flash on/off and torch on/off can be done simply by other driver space.
Two trigger APIs are added, ledtrig_flash_ctrl() and ledtrig_torch_ctrl().
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Macros are used in case that an inline function doesn't work.
Otherwise, use an empty inline function.
(a) Case of !CONFIG_LEDS_TRIGGERS
Following macros are replaced with inline functions.
led_trigger_register_simple()
led_trigger_unregister_simple()
led_trigger_event()
To make inline types, the structure, 'led_trigger' should be defined.
This structure has no member at all.
(b) Case of !CONFIG_LEDS_TRIGGER_IDE_DISK
ledtrig_ide_activity() macro is replaced with an inline function as well.
(c) DEFINE_LED_TRIGGER() and DEFINE_LED_TRIGGER_GLOBAL()
Struct 'led_trigger' is defined both cases, with CONFIG_LEDS_TRIGGERS and
without CONFIG_LEDS_TRIGGERS.
Those macros are moved out of CONFIG_LED_TRIGGERS because of no-dependency
on CONFIG_LEDS_TRIGGERS.
(d) Fix build errors in mmc-core driver
After replacing macros with inline functions, following build errors occur.
(condition: CONFIG_LEDS_TRIGGERS is not set)
drivers/mmc/core/core.c: In function 'mmc_request_done':
drivers/mmc/core/core.c:164:25: error: 'struct mmc_host' has no member named 'led'
drivers/mmc/core/core.c: In function 'mmc_start_request':
drivers/mmc/core/core.c:254:24: error: 'struct mmc_host' has no member named 'led'
make[3]: *** [drivers/mmc/core/core.o] Error 1
The reason of these errors is non-existent member variable, 'led'.
It is only valid when CONFIG_LEDS_TRIGGERS is set.
But now, it can be used without this dependency.
To fix build errors, member 'led' is always used without its config option in
'include/linux/mmc/host.h'.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
This eliminates having an #ifdef returning NULL for the case
when OF is disabled.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Provide some trace, though the hardware is most likely non-functional if
this happens.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
The gpio_request_one() flags parameter was set to:
GPIOF_DIR_OUT | !!brightness
GPIOF_DIR_OUT and GPIOF_DIR_IN are defined as below:
GPIOF_DIR_OUT (0 << 0)
GPIOF_DIR_IN (1 << 0)
So, when '!!brightness' is 1, the gpio pin can be set as input,
instead of output.
To prevent this problem, GPIOF_OUT_INIT flags should be used when
using gpio_request_one().
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
The devm_gpio_request_one() flags parameter was set to:
GPIOF_DIR_OUT | gpio_get_value(template->cmd)
GPIOF_DIR_OUT and GPIOF_DIR_IN are defined as below:
GPIOF_DIR_OUT (0 << 0)
GPIOF_DIR_IN (1 << 0)
So, when 'gpio_get_value(template->cmd)' is 1, the gpio pin can
be set as input, instead of output.
To prevent this problem, GPIOF_OUT_INIT flags should be used when
using devm_gpio_request_one().
Same goes for 'gpio_get_value(template->slow)' case.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
The devm_gpio_request_one() flags parameter was set to:
GPIOF_DIR_OUT | state
GPIOF_DIR_OUT and GPIOF_DIR_IN are defined as below:
GPIOF_DIR_OUT (0 << 0)
GPIOF_DIR_IN (1 << 0)
So, when 'state' is 1, the gpio pin can be set as input, instead
of output.
To prevent this problem, GPIOF_OUT_INIT flags should be used when
using devm_gpio_request_one().
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
CONFIG_HOTPLUG was removed, so __devexit or __exit of remove()
should not be used.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
CONFIG_HOTPLUG was removed, so __devexit or __exit of remove()
should not be used.
Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
For better driver management, new subdirectory, 'trigger' is created.
All LED trigger drivers are moved into this directory.
Internal header, 'leds.h' is included in each LED trigger drivers.
Fix the location of header file, "leds.h" -> "../leds.h" in driver files.
One exception is here, 'ledtrig-timer.c'.
There is no need to include 'leds.h'. so '#include "leds.h"' line was removed.
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
LP5562 can drive up to 4 channels, RGB and White.
LEDs can be controlled directly via the led class control interface.
LP55xx common driver
LP5562 is one of LP55xx family device, so LP55xx common code are used.
On the other hand, chip specific configuration is defined in the structure
'lp55xx_device_config'
LED pattern data
LP5562 has also internal program memory which is used for running various LED
patterns. LP5562 driver supports the firmware interface and the predefined
pattern data as well.
LP5562 device attributes: 'led_pattern' and 'engine_mux'
A 'led_pattern' is an index code which runs the predefined pattern data.
And 'engine_mux' is updated with the firmware interface is activated.
Detailed description has been updated in the documentation files,
'leds-lp55xx.txt' and 'leds-lp5562.txt'.
Changes on the header file
LP5562 configurable definitions are added.
Pattern RGB data is fixed as constant value.
(No side effect on other devices, LP5521 or LP5523.)
(cooloney@gmail.com: remove redundant mutex_unlock(). Reported by Dan
Carpenter <dan.carpenter@oracle.com>)
Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Signed-off-by: Bryan Wu <cooloney@gmail.com>
Matthew reported kernels fail the pci_eisa probe and are later successful
with the virtual_eisa_root_init force probe without slot0.
The reason for that is: PNP probing is before pci_eisa_init gets called
as pci_eisa_init is called via pci_driver.
pnp 00:0f has 0xc80 - 0xc84 reserved.
[ 9.700409] pnp 00:0f: [io 0x0c80-0x0c84]
so eisa_probe will fail from pci_eisa_init
==>eisa_root_register
==>eisa_probe path.
as force_probe is not set in pci_eisa_root, it will bail early when
slot0 is not probed and initialized.
Try to use subsys_initcall_sync instead, and will keep following sequence:
pci_subsys_init
pci_eisa_init_early
pnpacpi_init/isapnp_init
After this patch EISA can be initialized properly, and PNP overlapping
resource will not be reserved.
[ 10.104434] system 00:0f: [io 0x0c80-0x0c84] could not be reserved
Reported-by: Matthew Whitehead <mwhitehe@redhat.com>
Tested-by: Matthew Whitehead <mwhitehe@redhat.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Matthew found that 3.8.3 is having problems with an old (ancient)
PCI-to-EISA bridge, the Intel 82375. It worked with the 3.2 kernel.
He identified the 82375, but doesn't assign the struct resource *res
pointer inside the struct eisa_root_device, and panics.
pci_eisa_init() was using bus->resource[] directly instead of
pci_bus_resource_n(). The bus->resource[] array is a PCI-internal
implementation detail, and after commit 45ca9e97 (PCI: add helpers for
building PCI bus resource lists) and commit 0efd5aab (PCI: add struct
pci_host_bridge_window with CPU/bus address offset), bus->resource[] is not
used for PCI root buses any more.
The 82375 is a subtractive-decode PCI device, so handle it the same
way we handle PCI-PCI bridges in subtractive-decode mode in
pci_read_bridge_bases().
[bhelgaas: changelog]
Reported-by: Matthew Whitehead <mwhitehe@redhat.com>
Tested-by: Matthew Whitehead <mwhitehe@redhat.com>
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org # v3.3+
Conflicts:
net/mac80211/sta_info.c
net/wireless/core.h
Two minor conflicts in wireless. Overlapping additions of extern
declarations in net/wireless/core.h and a bug fix overlapping with
the addition of a boolean parameter to __ieee80211_key_free().
Signed-off-by: David S. Miller <davem@davemloft.net>
code and data to use hwmod and omap_device. This is a prerequisite for
moving the hwmod code out of arch/arm.
Basic test logs are available at:
http://www.pwsan.com/omap/testlogs/sham_aes_integration_devel_3.10/20130330155313/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)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=1t1o
-----END PGP SIGNATURE-----
Merge tag 'omap-devel-a-for-3.10' of git://git.kernel.org/pub/scm/linux/kernel/git/pjw/omap-pending into omap-for-v3.10/crypto
For OMAP2+ SoCs, convert the SHA/MD5 and AES accelerator integration
code and data to use hwmod and omap_device. This is a prerequisite for
moving the hwmod code out of arch/arm.
Basic test logs are available at:
http://www.pwsan.com/omap/testlogs/sham_aes_integration_devel_3.10/20130330155313/
In the conversion to pinctrl, an error in the pins for the rebuild
LED was introduced. This patch assigns the correct pins and includes
the correct name for the LED in kirkwood-iomega_ix2_200.dts.
Signed-off-by: Nigel Roberts <nigel@nobiscuit.com>
Cc: <stable@vger.kernel.org> # v3.8.x
Signed-off-by: Jason Cooper <jason@lakedaemon.net>
Commit 90173882ed ("omap: add dsp platform
device") used CONFIG_BRIDGE_DVFS were it obviously meant
CONFIG_TIDSPBRIDGE_DVFS. Fix that.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Tony Lindgren <tony@atomide.com>
The Kconfig symbol MACH_OMAP_HTCWIZARD got added in v2.6.30. It has
never been used. Its entry can safely be removed.
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Initial motivation was to avoid the confusing exit when when '/' is
pressed in non verbose mode, as specified in the help line searches
are only available in verbose mode.
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-20xezxim2y4agmkx7f3sucll@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Now that the map browser shares the input routine with the hists
browser, there is no need for using any libnewt routine, so remove all
traces except for honouring NO_NEWT=1 on the makefile command line as an
indication that TUI support is not needed, in fact it just sets
NO_SLANG=1.
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-wae5o7xca9m52bj1re28jc5j@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Instead of an ad-hoc, libnewt based equivalent.
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-elrijp95pijt66y6mmij4xm1@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The python/perf.so binding lacks dependency for libtraceevent.a so that
it cause the following error building python/perf.so. This patch
introduce the dependency for it.
$ make python/perf.so
CHK -fstack-protector-all
CHK -Wstack-protector
CHK -Wvolatile-register-var
CHK -D_FORTIFY_SOURCE=2
CHK bionic
CHK libelf
CHK libdw
CHK libunwind
CHK -DLIBELF_MMAP
CHK libaudit
CHK libnewt
CHK gtk2
CHK -DHAVE_GTK_INFO_BAR
CHK perl
CHK python
CHK python version
CHK libbfd
CHK -DHAVE_STRLCPY
CHK -DHAVE_ON_EXIT
CHK -DBACKTRACE_SUPPORT
CHK libnuma
GEN python/perf.so
x86_64-pc-linux-gnu-gcc: error: ../lib/traceevent/libtraceevent.a: No such file or directory
error: command 'x86_64-pc-linux-gnu-gcc' failed with exit status 1
cp: cannot stat 'python_ext_build/lib/perf.so': No such file or directory
make: *** [python/perf.so] Error 1
Signed-off-by: Naohiro Aota <naota@elisp.net>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/87wqswzznx.fsf@locke.i-did-not-set--mail-host-address--so-tickle-me
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This reverts commit 0439539f72.
This caused this segfault:
[root@sandy linux]# perf sched rec
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 1.306 MB perf.data (~57062 samples) ]
perf
[root@sandy linux]# perf sched lat
perf: builtin-sched.c:781: thread_atoms_search: Assertion `!(thread != atoms->thread)' failed.
Aborted (core dumped)
[root@sandy linux]#
Further investigation is needed to check that even with machine__remove_thread()
not really deleting the thread referenced in the PERF_RECORD_EXIT (it goes to
machine->dead_threads, because references may still exist to them in things like
hist, etc) some event later comes for this dead thread and then
machine__findnew_thread() will create a new thead instance that will not be the
same as the one referenced by work_atoms->thread in thread_atoms_search().
For now just revert this patch to get the 'perf sched lat' back working.
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
echo Link: http://lkml.kernel.org/n/tip-`ranpwd -l 24`@git.kernel.org
Link: http://lkml.kernel.org/n/tip-hg4s6e5txiwqe00h8rdg1sin@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The symbol addresses in a dso have relative offsets from the start of a
mapping. So in order to ouput correct offset value from @ip, one of
them should be converted.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1359040242-8269-19-git-send-email-eranian@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Leverages the PERF_RECORD_MISC_MMAP_DATA bit in the RECORD_MMAP record
header. When the bit is set then the mapping type is set to
MAP__VARIABLE.
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-17-git-send-email-eranian@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This new command is a wrapper on top of perf record and perf report to
make it easier to configure for memory access profiling.
To record loads:
$ perf mem -t load rec .....
To record stores:
$ perf mem -t store rec .....
To get the report:
$ perf mem -t load rep
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-15-git-send-email-eranian@google.com
[ Fixed minor conflict with 66857b5 "Sort command-list.txt alphabetically" ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch adds the --mem-mode option to perf report.
This mode requires a perf.data file created with memory access samples.
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-13-git-send-email-eranian@google.com
[ Removed duplicates in the --sort help, man page needs updating,
Fixed minor conflict with 328ccda "perf report: Add --no-demangle option" ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We use the -W option to obtain the cost of the memory accesses.
Data address sampling is obtained via the -d option.
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-14-git-send-email-eranian@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch adds the sorting and histogram support
functions to enable profiling of memory accesses.
The following sorting orders are added:
- symbol_daddr: data address symbol (or raw address)
- dso_daddr: data address shared object
- locked: access uses locked transaction
- tlb : TLB access
- mem : memory level of the access (L1, L2, L3, RAM, ...)
- snoop: access snoop mode
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-12-git-send-email-eranian@google.com
[ committer note: changed to cope with fc5871ed, the move of methods to
machine.[ch], and the rename of dsrc to data_src, to match the change
made in the PERF_SAMPLE_DSRC in a previous patch. ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf record has a new option -W that enables weightened sampling.
Add sorting support in top/report for the average weight per sample and the
total weight sum. This allows to both compare relative cost per event
and the total cost over the measurement period.
Add the necessary glue to perf report, record and the library.
v2: Merge with new hist refactoring.
v3: Fix manpage. Remove value check.
Rename global_weight to weight and weight to local_weight.
v4: Readd sort keys to manpage
v5: Move weight to end
v6: Move weight to template
v7: Rename weight key.
Original patch from Andi modified by Stephane Eranian <eranian@google.com>
to include ONLY the weight supporting code and apply to pristine 3.8.0-rc4.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-6-git-send-email-eranian@google.com
[ committer note: changed to cope with fc5871ed and the hists_link perf test entry ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Type of mapping was lost and made it hard for a tool
to distinguish code vs. data mmaps. Perf has the ability
to distinguish the two.
Use a bit in the header->misc bitmask to keep track of
the mmap type. If PERF_RECORD_MISC_MMAP_DATA is set then
the mapping is not executable (!VM_EXEC). If not set, then
the mapping is executable.
Signed-off-by: Stephane Eranian <eranian@google.com>
Cc: peterz@infradead.org
Cc: ak@linux.intel.com
Cc: acme@redhat.com
Cc: jolsa@redhat.com
Cc: namhyung.kim@lge.com
Link: http://lkml.kernel.org/r/1359040242-8269-16-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull arch/tile fix from Chris Metcalf:
"This change allows newer Tilera boot tools to work correctly with
current (and stable) kernels by using the right filename to get the
initramfs from the Tilera hypervisor filesystem."
* 'stable' of git://git.kernel.org/pub/scm/linux/kernel/git/cmetcalf/linux-tile:
tile: expect new initramfs name from hypervisor file system