Allocate the mcount record pages as a group of pages as big
as can be allocated and waste no more than a single page.
Grouping the mcount pages as much as possible helps with cache
locality, as we do not need to redirect with descriptors as we
cross from page to page. It also allows us to do more with the
records later on (sort them with bigger benefits).
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Records that are added to the function trace table are
permanently there, except for modules. By separating out the
modules to their own pages that can be freed in one shot
we can remove the "freed" flag and simplify some of the record
management.
Another benefit of doing this is that we can also move the
records around; sort them.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The stop machine method to modify all functions in the kernel
(some 20,000 of them) is the safest way to do so across all archs.
But some archs may not need this big hammer approach to modify code
on SMP machines, and can simply just update the code it needs.
Adding a weak function arch_ftrace_update_code() that now does the
stop machine, will also let any arch override this method.
If the arch needs to check the system and then decide if it can
avoid stop machine, it can still call ftrace_run_stop_machine() to
use the old method.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When gcc inlines a function, it does not mark it with the mcount
prologue, which in turn means that inlined functions are not traced
by the function tracer. But if CONFIG_OPTIMIZE_INLINING is set, then
gcc is allowed not to inline a function that is marked inline.
Depending on the options and the compiler, a function may or may
not be traced by the function tracer, depending on whether gcc
decides to inline a function or not. This has caused several
problems in the pass becaues gcc is not always consistent with
what it decides to inline between different gcc versions.
Some places should not be traced (like paravirt native_* functions)
and these are mostly marked as inline. When gcc decides not to
inline the function, and if that function should not be traced, then
the ftrace function tracer will suddenly break when it use to work
fine. This becomes even harder to debug when different versions of
gcc will not inline that function, making the same kernel and config
work for some gcc versions and not work for others.
By making all functions marked inline to not be traced will remove
the ambiguity that gcc adds when it comes to tracing functions marked
inline. All gcc versions will be consistent with what functions are
traced and having volatile working code will be removed.
Note, only the inline macro when CONFIG_OPTIMIZE_INLINING is set needs
to have notrace added, as the attribute __always_inline will force
the function to be inlined and then not traced.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Multiple users of the function tracer can register their functions
with the ftrace_ops structure. The accounting within ftrace will
update the counter on each function record that is being traced.
When the ftrace_ops filtering adds or removes functions, the
function records will be updated accordingly if the ftrace_ops is
still registered.
When a ftrace_ops is removed, the counter of the function records,
that the ftrace_ops traces, are decremented. When they reach zero
the functions that they represent are modified to stop calling the
mcount code.
When changes are made, the code is updated via stop_machine() with
a command passed to the function to tell it what to do. There is an
ENABLE and DISABLE command that tells the called function to enable
or disable the functions. But the ENABLE is really a misnomer as it
should just update the records, as records that have been enabled
and now have a count of zero should be disabled.
The DISABLE command is used to disable all functions regardless of
their counter values. This is the big off switch and is not the
complement of the ENABLE command.
To make matters worse, when a ftrace_ops is unregistered and there
is another ftrace_ops registered, neither the DISABLE nor the
ENABLE command are set when calling into the stop_machine() function
and the records will not be updated to match their counter. A command
is passed to that function that will update the mcount code to call
the registered callback directly if it is the only one left. This
means that the ftrace_ops that is still registered will have its callback
called by all functions that have been set for it as well as the ftrace_ops
that was just unregistered.
Here's a way to trigger this bug. Compile the kernel with
CONFIG_FUNCTION_PROFILER set and with CONFIG_FUNCTION_GRAPH not set:
CONFIG_FUNCTION_PROFILER=y
# CONFIG_FUNCTION_GRAPH is not set
This will force the function profiler to use the function tracer instead
of the function graph tracer.
# cd /sys/kernel/debug/tracing
# echo schedule > set_ftrace_filter
# echo function > current_tracer
# cat set_ftrace_filter
schedule
# cat trace
# tracer: nop
#
# entries-in-buffer/entries-written: 692/68108025 #P:4
#
# _-----=> irqs-off
# / _----=> need-resched
# | / _---=> hardirq/softirq
# || / _--=> preempt-depth
# ||| / delay
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
kworker/0:2-909 [000] .... 531.235574: schedule <-worker_thread
<idle>-0 [001] .N.. 531.235575: schedule <-cpu_idle
kworker/0:2-909 [000] .... 531.235597: schedule <-worker_thread
sshd-2563 [001] .... 531.235647: schedule <-schedule_hrtimeout_range_clock
# echo 1 > function_profile_enabled
# echo 0 > function_porfile_enabled
# cat set_ftrace_filter
schedule
# cat trace
# tracer: function
#
# entries-in-buffer/entries-written: 159701/118821262 #P:4
#
# _-----=> irqs-off
# / _----=> need-resched
# | / _---=> hardirq/softirq
# || / _--=> preempt-depth
# ||| / delay
# TASK-PID CPU# |||| TIMESTAMP FUNCTION
# | | | |||| | |
<idle>-0 [002] ...1 604.870655: local_touch_nmi <-cpu_idle
<idle>-0 [002] d..1 604.870655: enter_idle <-cpu_idle
<idle>-0 [002] d..1 604.870656: atomic_notifier_call_chain <-enter_idle
<idle>-0 [002] d..1 604.870656: __atomic_notifier_call_chain <-atomic_notifier_call_chain
The same problem could have happened with the trace_probe_ops,
but they are modified with the set_frace_filter file which does the
update at closure of the file.
The simple solution is to change ENABLE to UPDATE and call it every
time an ftrace_ops is unregistered.
Link: http://lkml.kernel.org/r/1323105776-26961-3-git-send-email-jolsa@redhat.com
Cc: stable@vger.kernel.org # 3.0+
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
add support for SG lists on dwc3 driver. With
this we can e.g. use VFS layer's SG lists on
storage gadgets so that we can start bigger
transfers and improve throughput.
Signed-off-by: Felipe Balbi <balbi@ti.com>
the LST bit is to be set on the last of a series
of consecutive TRBs. We had a workaround for a
problem where data would get corrupted but that
doesn't happen anymore. It's likely that it was
caused by some FPGA instability during development
phase.
Signed-off-by: Felipe Balbi <balbi@ti.com>
Some controllers support scatter/gather transfers
and that might be very useful for some gadget drivers.
This means that we can make use of larger buffer
allocations which means we will have less completion
IRQs overtime, thus improving the perceived performance.
Signed-off-by: Felipe Balbi <balbi@ti.com>
This is just a minor optimization for the long
if we have on the driver.
When we want to check that one input is true
and the other must be false, the bitwise XOR
operator will achieve that for us.
Tested-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Instead use container_of to retrieve the s3c_hsudc from the
struct usb_gadget pointer.
[ balbi@ti.com : changed verbose container_of() into
an already provided helper 'to_hsudc()' ]
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
As the memory region is requested through request_mem_region
use the correct paired method to release it in the error path
and don't go "beneath the API".
Reported-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
The udc has three supplies: vdda (3.3V), vddi (1.2V) and vddosc (1.8-3.3V).
Turn these on and off on start and stop calls.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
udc_start and udc_stop reduce code duplication in comparison to
start and stop generalising calls done by all drivers
(i.e. bind and unbind) and moving these calls to common code.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Instead of adding and deleting the gadget device in the start and stop
invocations. Use device_register in the probe method to initialize
and add the gadget device.
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Gadget drivers should be compilable on all architectures.
This patch removes one dependency on architecture-specific code.
Acked-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Heiko Stuebner <heiko@sntech.de>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Implement a simple test for the self-monitoring data from the
perf mmap data area control page:
6: x86 rdpmc test:
0: 6053
1: 60053
2: 600059
3: 6000059
4: 60000075
5: 600000247
Ok
The counts are expected to increase monotonically - these
are recovered via RDPMC, without calling into the kernel.
It might be nice to add logic to automagically turn these numbers into OK/FAIL.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Arun Sharma <asharma@fb.com>
Link: http://lkml.kernel.org/n/tip-evf5yii88ljdgmaihccbxxw1@git.kernel.org
[ various small improvements ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This change removes confusing extern qualifier, which doesn't have any practical
sense there.
Signed-off-by: Vladimir Zapolskiy <vladimir.zapolskiy@nokia.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>
Defer LED setting action to a workqueue.
This is more likely to send all LED change events in a single URB.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
If any userspace program has opened a keyboard device, the input core
de-activates the keyboard's LEDs upon suspend(). It does this by sending
individual EV_LED[LED_X]=0 events to the underlying device driver by
directly calling the driver's registered event() handler.
The usb-hid driver event() handler processes each request by immediately
attempting to submit a CTRL URB to turn off the LED. USB URB submission
is asynchronous. First the URB is added to the head of the ctrl queue.
Then, if the CTRL_RUNNING flag is false, the URB is submitted immediately
(and CTRL_RUNNING is set). If the CTRL_RUNNING flag was already true,
then the newly queued URB is submitted in the ctrl completion handler when
all previously submitted URBs have completed. When all queued URBs have
been submitted, the completion handler clears the CTRL_RUNNING flag.
In the 2-LED suspend case, at input suspend(), 2 LED event CTRL URBs get
queued, with only the first actually submitted. Soon after input
suspend() handler finishes, the usb-hid suspend() handler gets called.
Since this is NOT a PM_EVENT_AUTO suspend, the handler sets
REPORTED_IDLE, then waits for io to complete.
Unfortunately, this usually happens while the first LED request is
actually still being processed. Thus when the completion handler tries
to submit the second LED request it fails, since REPORTED_IDLE is
already set! This REPORTED_IDLE check failure causes the completion
handler to complete, however without clearing the CTRL_RUNNING flag.
This, in turn, means that the suspend() handler's wait_io() condition
is never satisfied, and instead it times out after 10 seconds, aborting
the original system suspend.
This patch changes the behavior to the following:
(1) allow completion handler to finish submitting all queued URBs, even if
REPORTED_IDLE is set. This guarantees that all URBs queued before the
hid-core suspend() call will be submitted before the system is
suspended.
(2) if REPORTED_IDLE is set and the URB queue is empty, queue, but
don't submit, new URB submission requests. These queued requests get
submitted when resume() flushes the URB queue. This is similar to the
existing behavior, however, any requests that arrive while the queue is
not yet empty will still get submitted before suspend.
(3) set the RUNNING flag when flushing the URB queue in resume().
This keeps URBs that were queued in (2) from colliding with any new
URBs that are being submitted during the resume process. The new URB
submission requests upon resume get properly queued behind the ones
being flushed instead of the current situation where they collide,
causing memory corruption and oopses.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
LED_ON was defined in the original version of the hid-core autosuspend patch.
However, during review, the setting and clearing of it was redone
using ledcount. The test was left in accidentally.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
Acked-by: Oliver Neukum <oneukum@suse.de>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
There is a small race between try_to_wake_up() and sched_move_task(),
which is trying to move the process being woken up.
try_to_wake_up() on CPU0 sched_move_task() on CPU1
--------------------------------+---------------------------------
raw_spin_lock_irqsave(p->pi_lock)
task_waking_fair()
->p.se.vruntime -= cfs_rq->min_vruntime
ttwu_queue()
->send reschedule IPI to CPU1
raw_spin_unlock_irqsave(p->pi_lock)
task_rq_lock()
-> tring to aquire both p->pi_lock and
rq->lock with IRQ disabled
task_move_group_fair()
-> p.se.vruntime
-= (old)cfs_rq->min_vruntime
+= (new)cfs_rq->min_vruntime
task_rq_unlock()
(via IPI)
sched_ttwu_pending()
raw_spin_lock(rq->lock)
ttwu_do_activate()
...
enqueue_entity()
child.se->vruntime += cfs_rq->min_vruntime
raw_spin_unlock(rq->lock)
As a result, vruntime of the process becomes far bigger than min_vruntime,
if (new)cfs_rq->min_vruntime >> (old)cfs_rq->min_vruntime.
This patch fixes this problem by just ignoring such process in
task_move_group_fair(), because the vruntime has already been normalized in
task_waking_fair().
Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/20111215143741.df82dd50.nishimura@mxp.nes.nec.co.jp
Signed-off-by: Ingo Molnar <mingo@elte.hu>
There is a small race between do_fork() and sched_move_task(), which is
trying to move the child.
do_fork() sched_move_task()
--------------------------------+---------------------------------
copy_process()
sched_fork()
task_fork_fair()
-> vruntime of the child is initialized
based on that of the parent.
-> we can see the child in "tasks" file now.
task_rq_lock()
task_move_group_fair()
-> child.se.vruntime
-= (old)cfs_rq->min_vruntime
+= (new)cfs_rq->min_vruntime
task_rq_unlock()
wake_up_new_task()
...
enqueue_entity()
child.se.vruntime += cfs_rq->min_vruntime
As a result, vruntime of the child becomes far bigger than min_vruntime,
if (new)cfs_rq->min_vruntime >> (old)cfs_rq->min_vruntime.
This patch fixes this problem by just ignoring such process in
task_move_group_fair(), because the vruntime has already been normalized in
task_fork_fair().
Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/20111215143607.2ee12c5d.nishimura@mxp.nes.nec.co.jp
Signed-off-by: Ingo Molnar <mingo@elte.hu>
There is a small race between task_fork_fair() and sched_move_task(),
which is trying to move the parent.
task_fork_fair() sched_move_task()
--------------------------------+---------------------------------
cfs_rq = task_cfs_rq(current)
-> cfs_rq is the "old" one.
curr = cfs_rq->curr
-> curr is set to the parent.
task_rq_lock()
dequeue_task()
->parent.se.vruntime -= (old)cfs_rq->min_vruntime
enqueue_task()
->parent.se.vruntime += (new)cfs_rq->min_vruntime
task_rq_unlock()
raw_spin_lock_irqsave(rq->lock)
se->vruntime = curr->vruntime
-> vruntime of the child is set to that of the parent
which has already been updated by sched_move_task().
se->vruntime -= (old)cfs_rq->min_vruntime.
raw_spin_unlock_irqrestore(rq->lock)
As a result, vruntime of the child becomes far bigger than expected,
if (new)cfs_rq->min_vruntime >> (old)cfs_rq->min_vruntime.
This patch fixes this problem by setting "cfs_rq" and "curr" after
holding the rq->lock.
Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Acked-by: Paul Turner <pjt@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Tejun Heo <tj@kernel.org>
Link: http://lkml.kernel.org/r/20111215143655.662676b0.nishimura@mxp.nes.nec.co.jp
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Remove cfs bandwidth period check from tg_set_cfs_period.
Invalid bandwidth period's lower/upper limits are denoted
by min_cfs_quota_period/max_cfs_quota_period repsectively,
and are checked against valid period in tg_set_cfs_bandwidth().
As pjt pointed out, negative input will result in very large unsigned
numbers and will be caught by the max allowed period test.
Signed-off-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Acked-by: Paul Turner <pjt@google.com>
[ammended changelog to mention negative values]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20111210135925.GA14593@linux.vnet.ibm.com
--
kernel/sched/core.c | 3 ---
1 file changed, 3 deletions(-)
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The current lock break relies on contention on the rq locks, something
which might never come because we've got IRQs disabled. Or will be
very likely because on anything with more than 2 cpus a synchronized
load-balance pass will very likely cause contention on the rq locks.
Also the sched_nr_migrate thing fails when it gets trapped the loops
of either the cgroup muck in load_balance_fair() or the move_tasks()
load condition.
Instead, use the new lb_flags field to propagate break/abort
conditions for all these loops and create a new loop outside the irq
disabled on the break being required.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/n/tip-tsceb6w61q0gakmsccix6xxi@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This event counts the number of reference core cpu cycles.
Reference means that the event increments at a constant rate which
is not subject to core CPU frequency adjustments. The event may
not count when the processor is in halted (low power) state.
As such, it may not be equivalent to wall clock time. However,
when the processor is not halted state, the event keeps
a constant correlation with wall clock time.
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1323559734-3488-3-git-send-email-eranian@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>