10 Mar, 2010
29 commits
-
I overlooked the perf_disable()/perf_enable() calls in
intel_pmu_handle_irq(), (pointed out by Markus) so we should not
explicitly disable_all/enable_all pebs counters in the drain functions,
these are already disabled and enabling them early is confusing.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Calling ioctl(PERF_EVENT_IOC_DISABLE) on a thottled counter would result
in a double disable, cure this by using x86_pmu_{start,stop} for
throttle/unthrottle and teach x86_pmu_stop() to check ->active_mask.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
It turns out the LBR is massively unreliable on certain CPUs, so code the
fixup a little more defensive to avoid crashing the kernel.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Some CPUs have errata where the LBR is not cleared on Power-On. So always
clear the LBRs before use.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
This CPU has just too many handycaps to be really useful.
Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Fix up the ds allocation error path, where we could free @buffer before
we used it.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Since there's now two users for this, place it in a common header.
Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: Masami Hiramatsu
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Expose the full PEBS record using PERF_SAMPLE_RAW
Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Saner PERF_CAPABILITIES support, which also exposes pebs_trap. Use that
latter to make PEBS's use of LBR conditional since a fault-like pebs
should already report the correct IP.( As of this writing there is no known hardware that implements
!pebs_trap )Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Use the PERF_RECORD_MISC_EXACT information to measure the success rate of
the PEBS fix-up.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Use the LBR to fix up the PEBS IP+1 issue.
As said, PEBS reports the next instruction, here we use the LBR to find
the last branch and from that construct the actual IP. If the IP matches
the LBR-TO, we use LBR-FROM, otherwise we use the LBR-TO address as the
beginning of the last basic block and decode forward.Once we find a match to the current IP, we use the previous location.
This patch introduces a new ABI element: PERF_RECORD_MISC_EXACT, which
conveys that the reported IP (PERF_SAMPLE_IP) is the exact instruction
that caused the event (barring CPU errata).The fixup can fail due to various reasons:
1) LBR contains invalid data (quite possible)
2) part of the basic block got paged out
3) the reported IP isn't part of the basic block (see 1)Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: Masami Hiramatsu
Cc: "Zhang, Yanmin"
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Implement simple suport Intel Last-Branch-Record, it supports all
hardware that implements FREEZE_LBRS_ON_PMI, but does not (yet) implement
the LBR config register.The Intel LBR is a FIFO of From,To addresses describing the last few
branches the hardware took.This patch does not add perf interface to the LBR, but merely provides an
interface for internal use.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Minimal userspace interface to the new 'precise' events flag.
Can be used like "perf top -e r00c0p" which will use PEBS to sample
retired instructions.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
This patch implements support for Intel Precise Event Based Sampling,
which is an alternative counter mode in which the counter triggers a
hardware assist to collect information on events. The hardware assist
takes a trap like snapshot of a subset of the machine registers.This data is written to the Intel Debug-Store, which can be programmed
with a data threshold at which to raise a PMI.With the PEBS hardware assist being trap like, the reported IP is always
one instruction after the actual instruction that triggered the event.This implements a simple PEBS model that always takes a single PEBS event
at a time. This is done so that the interaction with the rest of the
system is as expected (freq adjust, period randomization, lbr,
callchains, etc.).It adds an ABI element: perf_event_attr::precise, which indicates that we
wish to use this (constrained, but precise) mode.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Try to avoid useless rotation and PMU disables.
[ Could be improved by keeping a nr_runnable count to better account
for the < PERF_STAT_INACTIVE counters ]Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
hw_perf_enable() would enable already enabled events.
This causes problems with code that assumes that ->enable/->disable calls
are balanced (like the LBR code does).What happens is that events that were already running and left in place
would get enabled again.Avoid this by only enabling new events that match their previous
assignment.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
hw_perf_enable() would disable events that were not yet enabled.
This causes problems with code that assumes that ->enable/->disable calls
are balanced (like the LBR code does).What happens is that we disable newly added counters that match their
previous assignment, even though they are not yet programmed on the
hardware.Avoid this by only doing the first pass over the existing events.
Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Make sure n_added is properly accounted so that we can rely on the value
to reflect the number of added counters. This is needed if its going to
be used for more than a boolean check.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Calling ioctl(PERF_EVENT_IOC_DISABLE) on a thottled counter would result
in a double disable, cure this by using x86_pmu_{start,stop} for
throttle/unthrottle and teach x86_pmu_stop() to check ->active_mask.Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
pmu::start should undo pmu::stop, make it so.
Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
There is no concurrency on these variables, so don't use LOCK'ed ops.
As to the intel_pmu_handle_irq() status bit clean, nobody uses that so
remove it all together.Signed-off-by: Peter Zijlstra
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
Pass the full perf_event into the x86_pmu functions so that those may
make use of more than the hw_perf_event, and while doing this, remove the
superfluous second argument.Signed-off-by: Peter Zijlstra
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
The second and third argument to x86_perf_event_update() are superfluous
since they are simple expressions of the first argument. Hence remove
them.Signed-off-by: Peter Zijlstra
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
The second and third argument to x86_perf_event_set_period() are
superfluous since they are simple expressions of the first argument.
Hence remove them.Signed-off-by: Peter Zijlstra
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
Explicitly use intel_pmu_{disable,enable}_all() in intel_pmu_handle_irq()
to avoid the NMI race conditions in perf_{disable,enable}Signed-off-by: Peter Zijlstra
Cc: Arnaldo Carvalho de Melo
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
LKML-Reference:
Signed-off-by: Ingo Molnar -
Currently we always call hw_perf_disable(), even if its already disabled,
this seems superflous, esp. since it cannot be made NMI safe (see further
patches).Signed-off-by: Peter Zijlstra
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
Remove the hw_perf_event_*() hotplug hooks in favour of per PMU hotplug
notifiers. This has the advantage of reducing the static weak interface
as well as exposing all hotplug actions to the PMU.Use this to fix x86 hotplug usage where we did things in ONLINE which
should have been done in UP_PREPARE or STARTING.Signed-off-by: Peter Zijlstra
Cc: Paul Mundt
Cc: paulus@samba.org
Cc: eranian@google.com
Cc: robert.richter@amd.com
Cc: fweisbec@gmail.com
Cc: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
This makes it easier to extend perf_sample_data and fixes a bug on arm
and sparc, which failed to set ->raw to NULL, which can cause crashes
when combined with PERF_SAMPLE_RAW.It also optimizes PowerPC and tracepoint, because the struct
initialization is forced to zero out the whole structure.Signed-off-by: Peter Zijlstra
Acked-by: Jean Pihet
Reviewed-by: Frederic Weisbecker
Acked-by: David S. Miller
Cc: Jamie Iles
Cc: Paul Mackerras
Cc: Stephane Eranian
Cc: stable@kernel.org
LKML-Reference:
Signed-off-by: Ingo Molnar -
Conflicts:
tools/perf/util/probe-event.cMerge reason: Pick up -rc1 and resolve the conflict as well.
Signed-off-by: Ingo Molnar
09 Mar, 2010
3 commits
-
Acked-by: Peter Zijlstra
Acked-by: Paul Mackerras
Acked-by: Arnaldo Carvalho de Melo
LKML-Reference:
Signed-off-by: Ingo Molnar -
* git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core-2.6: (62 commits)
msi-laptop: depends on RFKILL
msi-laptop: Detect 3G device exists by standard ec command
msi-laptop: Add resume method for set the SCM load again
msi-laptop: Support some MSI 3G netbook that is need load SCM
msi-laptop: Add threeg sysfs file for support query 3G state by standard 66/62 ec command
msi-laptop: Support standard ec 66/62 command on MSI notebook and nebook
Driver core: create lock/unlock functions for struct device
sysfs: fix for thinko with sysfs_bin_attr_init()
sysfs: Kill unused sysfs_sb variable.
sysfs: Pass super_block to sysfs_get_inode
driver core: Use sysfs_rename_link in device_rename
sysfs: Implement sysfs_rename_link
sysfs: Pack sysfs_dirent more tightly.
sysfs: Serialize updates to the vfs inode
sysfs: windfarm: init sysfs attributes
sysfs: Use sysfs_attr_init and sysfs_bin_attr_init on module dynamic attributes
sysfs: Document sysfs_attr_init and sysfs_bin_attr_init
sysfs: Use sysfs_attr_init and sysfs_bin_attr_init on dynamic attributes
sysfs: Use one lockdep class per sysfs attribute.
sysfs: Only take active references on attributes.
...
08 Mar, 2010
8 commits
-
* git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
[WATCHDOG] i6300esb.c: change platform_driver to pci_driver
[WATCHDOG] i6300esb: fix unlock register with
[WATCHDOG] drivers/watchdog/wdt.c:wdt_ioctl(): make `ident' non-static
[WATCHDOG] change reboot_notifier to platform-shutdown method.
[WATCHDOG] watchdog_info constify
[WATCHDOG] gef_wdt: Author corrections following split of GE Fanuc joint venture
[WATCHDOG] iTCO_wdt: clean up probe(), modify err msg
[WATCHDOG] ep93xx: watchdog timer driver for TS-72xx SBCs cleanup
[WATCHDOG] support for max63xx watchdog timer chips
[WATCHDOG] ep93xx: added platform side support for TS-72xx WDT driver
[WATCHDOG] ep93xx: implemented watchdog timer driver for TS-72xx SBCs -
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6: (26 commits)
ALSA: hdmi - show debug message on changing audio infoframe
ALSA: hdmi - merge common code for intelhdmi and nvhdmi
ALSA: hda - Add ASRock mobo to MSI blacklist
ALSA: hda: uninitialized variable fix
ALSA: hda: Use LPIB for a Biostar Microtech board
ALSA: usb/audio.h: Fix field order
ALSA: fix jazz16 compile (udelay)
ALSA: hda: Use LPIB for Dell Latitude 131L
ALSA: hda - Build hda_eld into snd-hda-codec module
ALSA: hda - Support NVIDIA MCP89 and GT21x hdmi audio
ALSA: hda - Support max codecs to 8 for nvidia hda controller
ALSA: riptide: clean up while loop
ALSA: usbaudio - remove debug "SAMPLE BYTES" printk line
ALSA: timer - pass real event in snd_timer_notify1() to instance callback
ALSA: oxygen: change || to &&
ALSA: opti92x: use PnP data to select Master Control port
ASoC: fix ak4104 register array access
ASoC: soc_pcm_open: Add missing bailout tag
ALSA: usbaudio: Fix wrong bitrate for Creative Creative VF0470 Live Cam
ALSA: ua101: removing debugging code
... -
* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6:
[S390] smsgiucv_app: deliver z/VM CP special messages (SMSG) as uevents
[S390] smsgiucv: declare char pointers as "const"
[S390] dasd: automatic recognition of read-only devices
[S390] remove unused qdio flags in zfcp and qeth
[S390] Cleanup xtime usage
[S390] qdio: add missing bracket
[S390] cio: fix init_count in case of recognition after steal lock
[S390] dasd: security and PSF update patch for EMC CKD ioctl
[S390] hvc_iucv: allocate memory buffers for IUCV in zone DMA
[S390] uaccess: make sure copy_from_user_overflow is builtin -
Remove the EXPERIMENTAL flag from FS-Cache so that Ubuntu can make use of the
facility.Signed-off-by: Christian Kujau
Signed-off-by: David Howells
Signed-off-by: Linus Torvalds -
Since the intel 6300esb platform uses a dedicated PCI ID for
it's watchdog timer (00:1d.4), it's better and easier to have
this as a pci_driver.Signed-off-by: Wim Van Sebroeck
-
The smsgiucv_app driver registers a callback with the smsgiucv driver
to receive z/VM CP special messages (SMSG) starting with "APP".When the callback is called for special messages, the driver creates
an uevent for the received message. The uevent consists of additional
environment data containing the message prefix ("APP"), message sender,
and message content.udev rules can be used to trigger application specific actions through
matching the content or sender of the special message.Signed-off-by: Hendrik Brueckner
Signed-off-by: Martin Schwidefsky -
Declare the smsgiucv prefix char pointer as "const" and use
use const char pointers in callback functions.Signed-off-by: Hendrik Brueckner
Signed-off-by: Martin Schwidefsky -
In z/VM it is possible to attach a device as read-only. To prevent
unintentional write requests and subsequent I/O errors, we can detect
this configuration using the z/VM DIAG 210 interface and set the
respective linux block device to read-only as well.Signed-off-by: Stefan Weinhuber
Signed-off-by: Martin Schwidefsky
