linux-stable.git/kernel/smp.c, branch linux-2.6.37.y

smp_call_function_many: handle concurrent clearing of mask

2011-03-23T19:50:40+00:00

commit 723aae25d5cdb09962901d36d526b44d4be1051c upstream.

Mike Galbraith reported finding a lockup ("perma-spin bug") where the
cpumask passed to smp_call_function_many was cleared by other cpu(s)
while a cpu was preparing its call_data block, resulting in no cpu to
clear the last ref and unlock the block.

Having cpus clear their bit asynchronously could be useful on a mask of
cpus that might have a translation context, or cpus that need a push to
complete an rcu window.

Instead of adding a BUG_ON and requiring yet another cpumask copy, just
detect the race and handle it.

Note: arch_send_call_function_ipi_mask must still handle an empty
cpumask because the data block is globally visible before the that arch
callback is made.  And (obviously) there are no guarantees to which cpus
are notified if the mask is changed during the call; only cpus that were
online and had their mask bit set during the whole call are guaranteed
to be called.

Reported-by: Mike Galbraith 
Reported-by: Jan Beulich 
Acked-by: Jan Beulich 
Signed-off-by: Milton Miller 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

call_function_many: add missing ordering

2011-03-23T19:50:39+00:00

commit 45a5791920ae643eafc02e2eedef1a58e341b736 upstream.

Paul McKenney's review pointed out two problems with the barriers in the
2.6.38 update to the smp call function many code.

First, a barrier that would force the func and info members of data to
be visible before their consumption in the interrupt handler was
missing.  This can be solved by adding a smp_wmb between setting the
func and info members and setting setting the cpumask; this will pair
with the existing and required smp_rmb ordering the cpumask read before
the read of refs.  This placement avoids the need a second smp_rmb in
the interrupt handler which would be executed on each of the N cpus
executing the call request.  (I was thinking this barrier was present
but was not).

Second, the previous write to refs (establishing the zero that we the
interrupt handler was testing from all cpus) was performed by a third
party cpu.  This would invoke transitivity which, as a recient or
concurrent addition to memory-barriers.txt now explicitly states, would
require a full smp_mb().

However, we know the cpumask will only be set by one cpu (the data
owner) and any preivous iteration of the mask would have cleared by the
reading cpu.  By redundantly writing refs to 0 on the owning cpu before
the smp_wmb, the write to refs will follow the same path as the writes
that set the cpumask, which in turn allows us to keep the barrier in the
interrupt handler a smp_rmb instead of promoting it to a smp_mb (which
will be be executed by N cpus for each of the possible M elements on the
list).

I moved and expanded the comment about our (ab)use of the rcu list
primitives for the concurrent walk earlier into this function.  I
considered moving the first two paragraphs to the queue list head and
lock, but felt it would have been too disconected from the code.

Cc: Paul McKinney 
Signed-off-by: Milton Miller 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

call_function_many: fix list delete vs add race

2011-03-23T19:50:39+00:00

commit e6cd1e07a185d5f9b0aa75e020df02d3c1c44940 upstream.

Peter pointed out there was nothing preventing the list_del_rcu in
smp_call_function_interrupt from running before the list_add_rcu in
smp_call_function_many.

Fix this by not setting refs until we have gotten the lock for the list.
Take advantage of the wmb in list_add_rcu to save an explicit additional
one.

I tried to force this race with a udelay before the lock & list_add and
by mixing all 64 online cpus with just 3 random cpus in the mask, but
was unsuccessful.  Still, inspection shows a valid race, and the fix is
a extension of the existing protection window in the current code.

Reported-by: Peter Zijlstra 
Signed-off-by: Milton Miller 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

kernel/smp.c: fix smp_call_function_many() SMP race

2011-02-17T23:14:51+00:00

commit 6dc19899958e420a931274b94019e267e2396d3e upstream.

I noticed a failure where we hit the following WARN_ON in
generic_smp_call_function_interrupt:

                if (!cpumask_test_and_clear_cpu(cpu, data->cpumask))
                        continue;

                data->csd.func(data->csd.info);

                refs = atomic_dec_return(&data->refs);
                WARN_ON(refs < 0);      <-------------------------

We atomically tested and cleared our bit in the cpumask, and yet the
number of cpus left (ie refs) was 0.  How can this be?

It turns out commit 54fdade1c3332391948ec43530c02c4794a38172
("generic-ipi: make struct call_function_data lockless") is at fault.  It
removes locking from smp_call_function_many and in doing so creates a
rather complicated race.

The problem comes about because:

 - The smp_call_function_many interrupt handler walks call_function.queue
   without any locking.
 - We reuse a percpu data structure in smp_call_function_many.
 - We do not wait for any RCU grace period before starting the next
   smp_call_function_many.

Imagine a scenario where CPU A does two smp_call_functions back to back,
and CPU B does an smp_call_function in between.  We concentrate on how CPU
C handles the calls:

CPU A            CPU B                  CPU C              CPU D

smp_call_function
                                        smp_call_function_interrupt
                                            walks
					call_function.queue sees
					data from CPU A on list

                 smp_call_function

                                        smp_call_function_interrupt
                                            walks

                                        call_function.queue sees
                                          (stale) CPU A on list
							   smp_call_function int
							   clears last ref on A
							   list_del_rcu, unlock
smp_call_function reuses
percpu *data A
                                         data->cpumask sees and
                                         clears bit in cpumask
                                         might be using old or new fn!
                                         decrements refs below 0

set data->refs (too late!)

The important thing to note is since the interrupt handler walks a
potentially stale call_function.queue without any locking, then another
cpu can view the percpu *data structure at any time, even when the owner
is in the process of initialising it.

The following test case hits the WARN_ON 100% of the time on my PowerPC
box (having 128 threads does help :)

#include 
#include 

#define ITERATIONS 100

static void do_nothing_ipi(void *dummy)
{
}

static void do_ipis(struct work_struct *dummy)
{
	int i;

	for (i = 0; i < ITERATIONS; i++)
		smp_call_function(do_nothing_ipi, NULL, 1);

	printk(KERN_DEBUG "cpu %d finished\n", smp_processor_id());
}

static struct work_struct work[NR_CPUS];

static int __init testcase_init(void)
{
	int cpu;

	for_each_online_cpu(cpu) {
		INIT_WORK(&work[cpu], do_ipis);
		schedule_work_on(cpu, &work[cpu]);
	}

	return 0;
}

static void __exit testcase_exit(void)
{
}

module_init(testcase_init)
module_exit(testcase_exit)
MODULE_LICENSE("GPL");
MODULE_AUTHOR("Anton Blanchard");

I tried to fix it by ordering the read and the write of ->cpumask and
->refs.  In doing so I missed a critical case but Paul McKenney was able
to spot my bug thankfully :) To ensure we arent viewing previous
iterations the interrupt handler needs to read ->refs then ->cpumask then
->refs _again_.

Thanks to Milton Miller and Paul McKenney for helping to debug this issue.

[miltonm@bga.com: add WARN_ON and BUG_ON, remove extra read of refs before initial read of mask that doesn't help (also noted by Peter Zijlstra), adjust comments, hopefully clarify scenario ]
[miltonm@bga.com: remove excess tests]
Signed-off-by: Anton Blanchard 
Signed-off-by: Milton Miller 
Cc: Ingo Molnar 
Cc: "Paul E. McKenney" 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds 
Signed-off-by: Greg Kroah-Hartman

Typedef SMP call function pointer

2010-10-27T16:28:36+00:00

Typedef the pointer to the function to be called by smp_call_function() and
friends:

	typedef void (*smp_call_func_t)(void *info);

as it is used in a fair number of places.

Signed-off-by: David Howells 
cc: linux-arch@vger.kernel.org

generic-ipi: Fix deadlock in __smp_call_function_single

2010-09-10T14:48:40+00:00

Just got my 6 way machine to a state where cpu 0 is in an
endless loop within __smp_call_function_single.
All other cpus are idle.

The call trace on cpu 0 looks like this:

 __smp_call_function_single
 scheduler_tick
 update_process_times
 tick_sched_timer
 __run_hrtimer
 hrtimer_interrupt
 clock_comparator_work
 do_extint
 ext_int_handler
 ----> timer irq
 cpu_idle

__smp_call_function_single() got called from nohz_balancer_kick()
(inlined) with the remote cpu being 1, wait being 0 and the per
cpu variable remote_sched_softirq_cb (call_single_data) of the
current cpu (0).

Then it loops forever when it tries to grab the lock of the
call_single_data, since it is already locked and enqueued on cpu 0.

My theory how this could have happened: for some reason the
scheduler decided to call __smp_call_function_single() on it's own
cpu, and sends an IPI to itself. The interrupt stays pending
since IRQs are disabled. If then the hypervisor schedules the
cpu away it might happen that upon rescheduling both the IPI and
the timer IRQ are pending. If then interrupts are enabled again
it depends which one gets scheduled first.
If the timer interrupt gets delivered first we end up with the
local deadlock as seen in the calltrace above.

Let's make __smp_call_function_single() check if the target cpu is
the current cpu and execute the function immediately just like
smp_call_function_single does. That should prevent at least the
scenario described here.

It might also be that the scheduler is not supposed to call
__smp_call_function_single with the remote cpu being the current
cpu, but that is a different issue.

Signed-off-by: Heiko Carstens 
Acked-by: Peter Zijlstra 
Acked-by: Jens Axboe 
Cc: Venkatesh Pallipadi 
Cc: Suresh Siddha 
LKML-Reference: <20100910114729.GB2827@osiris.boeblingen.de.ibm.com>
Signed-off-by: Ingo Molnar

kernel/: convert cpu notifier to return encapsulate errno value

2010-05-27T16:12:48+00:00

By the previous modification, the cpu notifier can return encapsulate
errno value.  This converts the cpu notifiers for kernel/*.c

Signed-off-by: Akinobu Mita 
Cc: Ingo Molnar 
Cc: Peter Zijlstra 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h

2010-03-30T13:02:32+00:00

percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files.  percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.

percpu.h -> slab.h dependency is about to be removed.  Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability.  As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.

  http://userweb.kernel.org/~tj/misc/slabh-sweep.py

The script does the followings.

* Scan files for gfp and slab usages and update includes such that
  only the necessary includes are there.  ie. if only gfp is used,
  gfp.h, if slab is used, slab.h.

* When the script inserts a new include, it looks at the include
  blocks and try to put the new include such that its order conforms
  to its surrounding.  It's put in the include block which contains
  core kernel includes, in the same order that the rest are ordered -
  alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
  doesn't seem to be any matching order.

* If the script can't find a place to put a new include (mostly
  because the file doesn't have fitting include block), it prints out
  an error message indicating which .h file needs to be added to the
  file.

The conversion was done in the following steps.

1. The initial automatic conversion of all .c files updated slightly
   over 4000 files, deleting around 700 includes and adding ~480 gfp.h
   and ~3000 slab.h inclusions.  The script emitted errors for ~400
   files.

2. Each error was manually checked.  Some didn't need the inclusion,
   some needed manual addition while adding it to implementation .h or
   embedding .c file was more appropriate for others.  This step added
   inclusions to around 150 files.

3. The script was run again and the output was compared to the edits
   from #2 to make sure no file was left behind.

4. Several build tests were done and a couple of problems were fixed.
   e.g. lib/decompress_*.c used malloc/free() wrappers around slab
   APIs requiring slab.h to be added manually.

5. The script was run on all .h files but without automatically
   editing them as sprinkling gfp.h and slab.h inclusions around .h
   files could easily lead to inclusion dependency hell.  Most gfp.h
   inclusion directives were ignored as stuff from gfp.h was usually
   wildly available and often used in preprocessor macros.  Each
   slab.h inclusion directive was examined and added manually as
   necessary.

6. percpu.h was updated not to include slab.h.

7. Build test were done on the following configurations and failures
   were fixed.  CONFIG_GCOV_KERNEL was turned off for all tests (as my
   distributed build env didn't work with gcov compiles) and a few
   more options had to be turned off depending on archs to make things
   build (like ipr on powerpc/64 which failed due to missing writeq).

   * x86 and x86_64 UP and SMP allmodconfig and a custom test config.
   * powerpc and powerpc64 SMP allmodconfig
   * sparc and sparc64 SMP allmodconfig
   * ia64 SMP allmodconfig
   * s390 SMP allmodconfig
   * alpha SMP allmodconfig
   * um on x86_64 SMP allmodconfig

8. percpu.h modifications were reverted so that it could be applied as
   a separate patch and serve as bisection point.

Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.

Signed-off-by: Tejun Heo 
Guess-its-ok-by: Christoph Lameter 
Cc: Ingo Molnar 
Cc: Lee Schermerhorn

generic-ipi: Optimize accesses by using DEFINE_PER_CPU_SHARED_ALIGNED for IPI data

2010-01-18T08:02:59+00:00

The smp ipi data is passed around and given write access by
other cpus and should be separated from per-cpu data consumed by
this cpu.

Looking for hot lines, I saw call_function_data shared with
tick_cpu_sched.

Signed-off-by: Milton Miller 
Acked-by: Anton Blanchard 
Acked-by: Jens Axboe 
Cc: Andrew Morton 
Cc: Linus Torvalds 
Cc: Peter Zijlstra 
Cc: : Nick Piggin 
LKML-Reference: <20100118020051.GR12666@kryten>
Signed-off-by: Ingo Molnar

smp_call_function_any(): pass the node value to cpumask_of_node()

2010-01-16T20:15:39+00:00

The change in acpi_cpufreq to use smp_call_function_any causes a warning
when it is called since the function erroneously passes the cpu id to
cpumask_of_node rather than the node that the cpu is on.  Fix this.

cpumask_of_node(3): node > nr_node_ids(1)
Pid: 1, comm: swapper Not tainted 2.6.33-rc3-00097-g2c1f189 #223
Call Trace:
 [] cpumask_of_node+0x23/0x58
 [] smp_call_function_any+0x65/0xfa
 [] ? do_drv_read+0x0/0x2f
 [] get_cur_val+0xb0/0x102
 [] get_cur_freq_on_cpu+0x74/0xc5
 [] acpi_cpufreq_cpu_init+0x417/0x515
 [] ? __down_write+0xb/0xd
 [] cpufreq_add_dev+0x278/0x922

Signed-off-by: David John 
Cc: Suresh Siddha 
Cc: Rusty Russell 
Cc: Thomas Gleixner 
Cc: "H. Peter Anvin" 
Cc: Ingo Molnar 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds