firefly-linux-kernel-4.4.55.git
8 years agosched: Energy-aware wake-up task placement
Morten Rasmussen [Sat, 9 May 2015 19:03:19 +0000 (20:03 +0100)]
sched: Energy-aware wake-up task placement

Let available compute capacity and estimated energy impact select
wake-up target cpu when energy-aware scheduling is enabled and the
system in not over-utilized (above the tipping point).

energy_aware_wake_cpu() attempts to find group of cpus with sufficient
compute capacity to accommodate the task and find a cpu with enough spare
capacity to handle the task within that group. Preference is given to
cpus with enough spare capacity at the current OPP. Finally, the energy
impact of the new target and the previous task cpu is compared to select
the wake-up target cpu.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Determine the current sched_group idle-state
Dietmar Eggemann [Tue, 27 Jan 2015 14:04:17 +0000 (14:04 +0000)]
sched: Determine the current sched_group idle-state

To estimate the energy consumption of a sched_group in
sched_group_energy() it is necessary to know which idle-state the group
is in when it is idle. For now, it is assumed that this is the current
idle-state (though it might be wrong). Based on the individual cpu
idle-states group_idle_state() finds the group idle-state.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched, cpuidle: Track cpuidle state index in the scheduler
Morten Rasmussen [Tue, 27 Jan 2015 13:48:07 +0000 (13:48 +0000)]
sched, cpuidle: Track cpuidle state index in the scheduler

The idle-state of each cpu is currently pointed to by rq->idle_state but
there isn't any information in the struct cpuidle_state that can used to
look up the idle-state energy model data stored in struct
sched_group_energy. For this purpose is necessary to store the idle
state index as well. Ideally, the idle-state data should be unified.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Add over-utilization/tipping point indicator
Morten Rasmussen [Sat, 9 May 2015 15:49:57 +0000 (16:49 +0100)]
sched: Add over-utilization/tipping point indicator

Energy-aware scheduling is only meant to be active while the system is
_not_ over-utilized. That is, there are spare cycles available to shift
tasks around based on their actual utilization to get a more
energy-efficient task distribution without depriving any tasks. When
above the tipping point task placement is done the traditional way based
on load_avg, spreading the tasks across as many cpus as possible based
on priority scaled load to preserve smp_nice. Below the tipping point we
want to use util_avg instead. We need to define a criteria for when we
make the switch.

The util_avg for each cpu converges towards 100% (1024) regardless of
how many task additional task we may put on it. If we define
over-utilized as:

sum_{cpus}(rq.cfs.avg.util_avg) + margin > sum_{cpus}(rq.capacity)

some individual cpus may be over-utilized running multiple tasks even
when the above condition is false. That should be okay as long as we try
to spread the tasks out to avoid per-cpu over-utilization as much as
possible and if all tasks have the _same_ priority. If the latter isn't
true, we have to consider priority to preserve smp_nice.

For example, we could have n_cpus nice=-10 util_avg=55% tasks and
n_cpus/2 nice=0 util_avg=60% tasks. Balancing based on util_avg we are
likely to end up with nice=-10 tasks sharing cpus and nice=0 tasks
getting their own as we 1.5*n_cpus tasks in total and 55%+55% is less
over-utilized than 55%+60% for those cpus that have to be shared. The
system utilization is only 85% of the system capacity, but we are
breaking smp_nice.

To be sure not to break smp_nice, we have defined over-utilization
conservatively as when any cpu in the system is fully utilized at it's
highest frequency instead:

cpu_rq(any).cfs.avg.util_avg + margin > cpu_rq(any).capacity

IOW, as soon as one cpu is (nearly) 100% utilized, we switch to load_avg
to factor in priority to preserve smp_nice.

With this definition, we can skip periodic load-balance as no cpu has an
always-running task when the system is not over-utilized. All tasks will
be periodic and we can balance them at wake-up. This conservative
condition does however mean that some scenarios that could benefit from
energy-aware decisions even if one cpu is fully utilized would not get
those benefits.

For system where some cpus might have reduced capacity on some cpus
(RT-pressure and/or big.LITTLE), we want periodic load-balance checks as
soon a just a single cpu is fully utilized as it might one of those with
reduced capacity and in that case we want to migrate it.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Estimate energy impact of scheduling decisions
Morten Rasmussen [Tue, 6 Jan 2015 17:34:05 +0000 (17:34 +0000)]
sched: Estimate energy impact of scheduling decisions

Adds a generic energy-aware helper function, energy_diff(), that
calculates energy impact of adding, removing, and migrating utilization
in the system.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Extend sched_group_energy to test load-balancing decisions
Morten Rasmussen [Fri, 2 Jan 2015 14:21:56 +0000 (14:21 +0000)]
sched: Extend sched_group_energy to test load-balancing decisions

Extended sched_group_energy() to support energy prediction with usage
(tasks) added/removed from a specific cpu or migrated between a pair of
cpus. Useful for load-balancing decision making.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Calculate energy consumption of sched_group
Morten Rasmussen [Thu, 18 Dec 2014 14:47:18 +0000 (14:47 +0000)]
sched: Calculate energy consumption of sched_group

For energy-aware load-balancing decisions it is necessary to know the
energy consumption estimates of groups of cpus. This patch introduces a
basic function, sched_group_energy(), which estimates the energy
consumption of the cpus in the group and any resources shared by the
members of the group.

NOTE: The function has five levels of identation and breaks the 80
character limit. Refactoring is necessary.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Highest energy aware balancing sched_domain level pointer
Morten Rasmussen [Fri, 2 Jan 2015 17:08:52 +0000 (17:08 +0000)]
sched: Highest energy aware balancing sched_domain level pointer

Add another member to the family of per-cpu sched_domain shortcut
pointers. This one, sd_ea, points to the highest level at which energy
model is provided. At this level and all levels below all sched_groups
have energy model data attached.

Partial energy model information is possible but restricted to providing
energy model data for lower level sched_domains (sd_ea and below) and
leaving load-balancing on levels above to non-energy-aware
load-balancing. For example, it is possible to apply energy-aware
scheduling within each socket on a multi-socket system and let normal
scheduling handle load-balancing between sockets.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Relocated cpu_util() and change return type
Morten Rasmussen [Thu, 11 Dec 2014 15:25:29 +0000 (15:25 +0000)]
sched: Relocated cpu_util() and change return type

Move cpu_util() to an earlier position in fair.c and change return
type to unsigned long as negative usage doesn't make much sense. All
other load and capacity related functions use unsigned long including
the caller of cpu_util().

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Compute cpu capacity available at current frequency
Morten Rasmussen [Tue, 13 Jan 2015 14:11:28 +0000 (14:11 +0000)]
sched: Compute cpu capacity available at current frequency

capacity_orig_of() returns the max available compute capacity of a cpu.
For scale-invariant utilization tracking and energy-aware scheduling
decisions it is useful to know the compute capacity available at the
current OPP of a cpu.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agoarm64: Cpu invariant scheduler load-tracking and capacity support
Juri Lelli [Thu, 30 Apr 2015 10:53:48 +0000 (11:53 +0100)]
arm64: Cpu invariant scheduler load-tracking and capacity support

Provides the scheduler with a cpu scaling correction factor for more
accurate load-tracking and cpu capacity handling.

The Energy Model (EM) (in fact the capacity value of the last element
of the capacity states vector of the core (MC) level sched_group_energy
structure) is used as the source for this cpu scaling factor.

The cpu capacity value depends on the micro-architecture and the
maximum frequency of the cpu.

The maximum frequency part should not be confused with the frequency
invariant scheduler load-tracking support which deals with frequency
related scaling due to DFVS functionality.

Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agoarm: Cpu invariant scheduler load-tracking and capacity support
Dietmar Eggemann [Fri, 10 Jul 2015 12:57:19 +0000 (13:57 +0100)]
arm: Cpu invariant scheduler load-tracking and capacity support

Provides the scheduler with a cpu scaling correction factor for more
accurate load-tracking and cpu capacity handling.

The Energy Model (EM) (in fact the capacity value of the last element
of the capacity states vector of the core (MC) level sched_group_energy
structure) is used instead of the arm arch specific cpu_efficiency and
dtb property 'clock-frequency' values as the source for this cpu
scaling factor.

The cpu capacity value depends on the micro-architecture and the
maximum frequency of the cpu.

The maximum frequency part should not be confused with the frequency
invariant scheduler load-tracking support which deals with frequency
related scaling due to DFVS functionality.

Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched: Introduce SD_SHARE_CAP_STATES sched_domain flag
Morten Rasmussen [Tue, 13 Jan 2015 13:50:46 +0000 (13:50 +0000)]
sched: Introduce SD_SHARE_CAP_STATES sched_domain flag

cpufreq is currently keeping it a secret which cpus are sharing
clock source. The scheduler needs to know about clock domains as well
to become more energy aware. The SD_SHARE_CAP_STATES domain flag
indicates whether cpus belonging to the sched_domain share capacity
states (P-states).

There is no connection with cpufreq (yet). The flag must be set by
the arch specific topology code.

cc: Russell King <linux@arm.linux.org.uk>
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Initialize energy data structures
Dietmar Eggemann [Fri, 14 Nov 2014 16:20:20 +0000 (16:20 +0000)]
sched: Initialize energy data structures

The sched_group_energy (sge) pointer of the first sched_group (sg) in
the sched_domain (sd) is initialized to point to the appropriate (in
terms of sd level and cpu) sge data defined in the arch and so to the
correct part of the Energy Model (EM).

Energy-aware scheduling allows that a system has only EM data up to a
certain sd level (so called highest energy aware balancing sd level).
A check in init_sched_energy() enforces that all sd's below this sd
level contain EM data.

The 'int cpu' parameter of sched_domain_energy_f requires that
check_sched_energy_data() makes sure that all cpus spanned by a sg
are provisioned with the same EM data.

This patch has also been tested with feature FORCE_SD_OVERLAP enabled.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched: Introduce energy data structures
Dietmar Eggemann [Fri, 14 Nov 2014 16:08:45 +0000 (16:08 +0000)]
sched: Introduce energy data structures

The struct sched_group_energy represents the per sched_group related
data which is needed for energy aware scheduling. It contains:

  (1) number of elements of the idle state array
  (2) pointer to the idle state array which comprises 'power consumption'
      for each idle state
  (3) number of elements of the capacity state array
  (4) pointer to the capacity state array which comprises 'compute
      capacity and power consumption' tuples for each capacity state

The struct sched_group obtains a pointer to a struct sched_group_energy.

The function pointer sched_domain_energy_f is introduced into struct
sched_domain_topology_level which will allow the arch to pass a particular
struct sched_group_energy from the topology shim layer into the scheduler
core.

The function pointer sched_domain_energy_f has an 'int cpu' parameter
since the folding of two adjacent sd levels via sd degenerate doesn't work
for all sd levels. I.e. it is not possible for example to use this feature
to provide per-cpu energy in sd level DIE on ARM's TC2 platform.

It was discussed that the folding of sd levels approach is preferable
over the cpu parameter approach, simply because the user (the arch
specifying the sd topology table) can introduce less errors. But since
it is not working, the 'int cpu' parameter is the only way out. It's
possible to use the folding of sd levels approach for
sched_domain_flags_f and the cpu parameter approach for the
sched_domain_energy_f at the same time though. With the use of the
'int cpu' parameter, an extra check function has to be provided to make
sure that all cpus spanned by a sched group are provisioned with the same
energy data.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched: Make energy awareness a sched feature
Morten Rasmussen [Tue, 13 Jan 2015 13:45:51 +0000 (13:45 +0000)]
sched: Make energy awareness a sched feature

This patch introduces the ENERGY_AWARE sched feature, which is
implemented using jump labels when SCHED_DEBUG is defined. It is
statically set false when SCHED_DEBUG is not defined. Hence this doesn't
allow energy awareness to be enabled without SCHED_DEBUG. This
sched_feature knob will be replaced later with a more appropriate
control knob when things have matured a bit.

ENERGY_AWARE is based on per-entity load-tracking hence FAIR_GROUP_SCHED
must be enable. This dependency isn't checked at compile time yet.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Documentation for scheduler energy cost model
Morten Rasmussen [Tue, 13 Jan 2015 13:43:28 +0000 (13:43 +0000)]
sched: Documentation for scheduler energy cost model

This documentation patch provides an overview of the experimental
scheduler energy costing model, associated data structures, and a
reference recipe on how platforms can be characterized to derive energy
models.

Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Prevent unnecessary active balance of single task in sched group
Morten Rasmussen [Thu, 2 Jul 2015 16:16:34 +0000 (17:16 +0100)]
sched: Prevent unnecessary active balance of single task in sched group

Scenarios with the busiest group having just one task and the local
being idle on topologies with sched groups with different numbers of
cpus manage to dodge all load-balance bailout conditions resulting the
nr_balance_failed counter to be incremented. This eventually causes a
pointless active migration of the task. This patch prevents this by not
incrementing the counter when the busiest group only has one task.
ASYM_PACKING migrations and migrations due to reduced capacity should
still take place as these are explicitly captured by
need_active_balance().

A better solution would be to not attempt the load-balance in the first
place, but that requires significant changes to the order of bailout
conditions and statistics gathering.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Enable idle balance to pull single task towards cpu with higher capacity
Dietmar Eggemann [Mon, 26 Jan 2015 19:47:28 +0000 (19:47 +0000)]
sched: Enable idle balance to pull single task towards cpu with higher capacity

We do not want to miss out on the ability to pull a single remaining
task from a potential source cpu towards an idle destination cpu. Add an
extra criteria to need_active_balance() to kick off active load balance
if the source cpu is over-utilized and has lower capacity than the
destination cpu.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched: Consider spare cpu capacity at task wake-up
Morten Rasmussen [Mon, 6 Jul 2015 14:01:10 +0000 (15:01 +0100)]
sched: Consider spare cpu capacity at task wake-up

find_idlest_group() selects the wake-up target group purely
based on group load which leads to suboptimal choices in low load
scenarios. An idle group with reduced capacity (due to RT tasks or
different cpu type) isn't necessarily a better target than a lightly
loaded group with higher capacity.

The patch adds spare capacity as an additional group selection
parameter. The target group is now selected based on the following
criteria:

1. Return the group with the cpu with most spare capacity and this
capacity is significant if such group exists. Significant spare capacity
is currently at least 20% to spare.

2. Return the group with the lowest load, unless it is the local group
in which case NULL is returned and the search is continued at the next
(lower) level.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Add cpu capacity awareness to wakeup balancing
Morten Rasmussen [Sat, 9 May 2015 18:53:49 +0000 (19:53 +0100)]
sched: Add cpu capacity awareness to wakeup balancing

Wakeup balancing is completely unaware of cpu capacity, cpu utilization
and task utilization. The task is preferably placed on a cpu which is
idle in the instant the wakeup happens. New tasks
(SD_BALANCE_{FORK,EXEC} are placed on an idle cpu in the idlest group if
such can be found, otherwise it goes on the least loaded one. Existing
tasks (SD_BALANCE_WAKE) are placed on the previous cpu or an idle cpu
sharing the same last level cache unless the wakee_flips heuristic in
wake_wide() decides to fallback to considering cpus outside SD_LLC.
Hence existing tasks are not guaranteed to get a chance to migrate to a
different group at wakeup in case the current one has reduced cpu
capacity (due RT/IRQ pressure or different uarch e.g. ARM big.LITTLE).
They may eventually get pulled by other cpus doing
periodic/idle/nohz_idle balance, but it may take quite a while before it
happens.

This patch adds capacity awareness to find_idlest_{group,queue} (used by
SD_BALANCE_{FORK,EXEC} and SD_BALANCE_WAKE under certain circumstances)
such that groups/cpus that can accommodate the waking task based on task
utilization are preferred. In addition, wakeup of existing tasks
(SD_BALANCE_WAKE) is sent through find_idlest_{group,queue} also if the
task doesn't fit the capacity of the previous cpu to allow it to escape
(override wake_affine) when necessary instead of relying on
periodic/idle/nohz_idle balance to eventually sort it out.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agosched: Store system-wide maximum cpu capacity in root domain
Dietmar Eggemann [Thu, 7 May 2015 17:46:15 +0000 (18:46 +0100)]
sched: Store system-wide maximum cpu capacity in root domain

To be able to compare the capacity of the target cpu with the highest
cpu capacity of the system in the wakeup path, store the system-wide
maximum cpu capacity in the root domain.

cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agoarm: Update arch_scale_cpu_capacity() to reflect change to define
Morten Rasmussen [Tue, 14 Apr 2015 15:25:31 +0000 (16:25 +0100)]
arm: Update arch_scale_cpu_capacity() to reflect change to define

arch_scale_cpu_capacity() is no longer a weak function but a #define
instead. Include the #define in topology.h.

cc: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
8 years agoarm64: Enable frequency invariant scheduler load-tracking support
Dietmar Eggemann [Fri, 25 Sep 2015 16:15:11 +0000 (17:15 +0100)]
arm64: Enable frequency invariant scheduler load-tracking support

Defines arch_scale_freq_capacity() to use cpufreq implementation.

Including <linux/cpufreq.h> in topology.h like for the arm arch doesn't
work because of CONFIG_COMPAT=y (Kernel support for 32-bit EL0).
That's why cpufreq_scale_freq_capacity() has to be declared extern in
topology.h.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agoarm: Enable frequency invariant scheduler load-tracking support
Dietmar Eggemann [Wed, 23 Sep 2015 11:47:48 +0000 (12:47 +0100)]
arm: Enable frequency invariant scheduler load-tracking support

Defines arch_scale_freq_capacity() to use cpufreq implementation.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agocpufreq: Frequency invariant scheduler load-tracking support
Dietmar Eggemann [Thu, 17 Sep 2015 15:10:56 +0000 (16:10 +0100)]
cpufreq: Frequency invariant scheduler load-tracking support

Implements cpufreq_scale_freq_capacity() to provide the scheduler with a
frequency scaling correction factor for more accurate load-tracking.

The factor is:

current_freq(cpu) << SCHED_CAPACITY_SHIFT / max_freq(cpu)

In fact, freq_scale should be a struct cpufreq_policy data member. But
this would require that the scheduler hot path (__update_load_avg()) would
have to grab the cpufreq lock. This can be avoided by using per-cpu data
initialized to SCHED_CAPACITY_SCALE for freq_scale.

Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
8 years agosched/fair: Fix new task's load avg removed from source CPU in wake_up_new_task()
Yuyang Du [Wed, 16 Dec 2015 23:34:27 +0000 (07:34 +0800)]
sched/fair: Fix new task's load avg removed from source CPU in wake_up_new_task()

If a newly created task is selected to go to a different CPU in fork
balance when it wakes up the first time, its load averages should
not be removed from the source CPU since they are never added to
it before. The same is also applicable to a never used group entity.

Fix it in remove_entity_load_avg(): when entity's last_update_time
is 0, simply return. This should precisely identify the case in
question, because in other migrations, the last_update_time is set
to 0 after remove_entity_load_avg().

Reported-by: Steve Muckle <steve.muckle@linaro.org>
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
[peterz: cfs_rq_last_update_time]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <Juri.Lelli@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Patrick Bellasi <patrick.bellasi@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: http://lkml.kernel.org/r/20151216233427.GJ28098@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
8 years agoFROMLIST: pstore: drop pmsg bounce buffer
Mark Salyzyn [Wed, 31 Aug 2016 15:09:04 +0000 (08:09 -0700)]
FROMLIST: pstore: drop pmsg bounce buffer

(from https://lkml.org/lkml/2016/9/1/428)
(cherry pick from android-3.10 commit b58133100b38f2bf83cad2d7097417a3a196ed0b)

Removing a bounce buffer copy operation in the pmsg driver path is
always better. We also gain in overall performance by not requesting
a vmalloc on every write as this can cause precious RT tasks, such
as user facing media operation, to stall while memory is being
reclaimed. Added a write_buf_user to the pstore functions, a backup
platform write_buf_user that uses the small buffer that is part of
the instance, and implemented a ramoops write_buf_user that only
supports PSTORE_TYPE_PMSG.

Signed-off-by: Mark Salyzyn <salyzyn@google.com>
Bug: 31057326
Change-Id: I4cdee1cd31467aa3e6c605bce2fbd4de5b0f8caa

8 years agoUPSTREAM: usercopy: remove page-spanning test for now
Kees Cook [Wed, 7 Sep 2016 16:54:34 +0000 (09:54 -0700)]
UPSTREAM: usercopy: remove page-spanning test for now

A custom allocator without __GFP_COMP that copies to userspace has been
found in vmw_execbuf_process[1], so this disables the page-span checker
by placing it behind a CONFIG for future work where such things can be
tracked down later.

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1373326

Reported-by: Vinson Lee <vlee@freedesktop.org>
Fixes: f5509cc18daa ("mm: Hardened usercopy")
Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: I4177c0fb943f14a5faf5c70f5e54bf782c316f43
(cherry picked from commit 8e1f74ea02cf4562404c48c6882214821552c13f)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: usercopy: force check_object_size() inline
Kees Cook [Wed, 7 Sep 2016 16:39:32 +0000 (09:39 -0700)]
UPSTREAM: usercopy: force check_object_size() inline

Just for good measure, make sure that check_object_size() is always
inlined too, as already done for copy_*_user() and __copy_*_user().

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: Ibfdf4790d03fe426e68d9a864c55a0d1bbfb7d61
(cherry picked from commit a85d6b8242dc78ef3f4542a0f979aebcbe77fc4e)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoBACKPORT: usercopy: fold builtin_const check into inline function
Kees Cook [Wed, 31 Aug 2016 23:04:21 +0000 (16:04 -0700)]
BACKPORT: usercopy: fold builtin_const check into inline function

Instead of having each caller of check_object_size() need to remember to
check for a const size parameter, move the check into check_object_size()
itself. This actually matches the original implementation in PaX, though
this commit cleans up the now-redundant builtin_const() calls in the
various architectures.

Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: I348809399c10ffa051251866063be674d064b9ff
(cherry picked from 81409e9e28058811c9ea865345e1753f8f677e44)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: x86/uaccess: force copy_*_user() to be inlined
Kees Cook [Tue, 6 Sep 2016 18:56:01 +0000 (11:56 -0700)]
UPSTREAM: x86/uaccess: force copy_*_user() to be inlined

As already done with __copy_*_user(), mark copy_*_user() as __always_inline.
Without this, the checks for things like __builtin_const_p() won't work
consistently in either hardened usercopy nor the recent adjustments for
detecting usercopy overflows at compile time.

The change in kernel text size is detectable, but very small:

 text      data     bss     dec      hex     filename
12118735  5768608 14229504 32116847 1ea106f vmlinux.before
12120207  5768608 14229504 32118319 1ea162f vmlinux.after

Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: I284c85c2a782145f46655a91d4f83874c90eba61
(cherry picked from commit e6971009a95a74f28c58bbae415c40effad1226c)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: HID: core: prevent out-of-bound readings
Benjamin Tissoires [Tue, 19 Jan 2016 11:34:58 +0000 (12:34 +0100)]
UPSTREAM: HID: core: prevent out-of-bound readings

(cherry picked from commit 50220dead1650609206efe91f0cc116132d59b3f)

Plugging a Logitech DJ receiver with KASAN activated raises a bunch of
out-of-bound readings.

The fields are allocated up to MAX_USAGE, meaning that potentially, we do
not have enough fields to fit the incoming values.
Add checks and silence KASAN.

Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Change-Id: Iaf25e882a6696884439d7091b5fbb0b350d893d3
Bug: 30951261

8 years agoAndroid: Fix build breakages.
Mohan Srinivasan [Thu, 8 Sep 2016 00:39:42 +0000 (17:39 -0700)]
Android: Fix build breakages.

The IO latency histogram change broke allmodconfig and
allnoconfig builds. This fixes those breakages.

Change-Id: I9cdae655b40ed155468f3cef25cdb74bb56c4d3e
Signed-off-by: Mohan Srinivasan <srmohan@google.com>
8 years agoUPSTREAM: tty: Prevent ldisc drivers from re-using stale tty fields
Peter Hurley [Fri, 27 Nov 2015 19:30:21 +0000 (14:30 -0500)]
UPSTREAM: tty: Prevent ldisc drivers from re-using stale tty fields

(cherry picked from commit dd42bf1197144ede075a9d4793123f7689e164bc)

Line discipline drivers may mistakenly misuse ldisc-related fields
when initializing. For example, a failure to initialize tty->receive_room
in the N_GIGASET_M101 line discipline was recently found and fixed [1].
Now, the N_X25 line discipline has been discovered accessing the previous
line discipline's already-freed private data [2].

Harden the ldisc interface against misuse by initializing revelant
tty fields before instancing the new line discipline.

[1]
    commit fd98e9419d8d622a4de91f76b306af6aa627aa9c
    Author: Tilman Schmidt <tilman@imap.cc>
    Date:   Tue Jul 14 00:37:13 2015 +0200

    isdn/gigaset: reset tty->receive_room when attaching ser_gigaset

[2] Report from Sasha Levin <sasha.levin@oracle.com>
    [  634.336761] ==================================================================
    [  634.338226] BUG: KASAN: use-after-free in x25_asy_open_tty+0x13d/0x490 at addr ffff8800a743efd0
    [  634.339558] Read of size 4 by task syzkaller_execu/8981
    [  634.340359] =============================================================================
    [  634.341598] BUG kmalloc-512 (Not tainted): kasan: bad access detected
    ...
    [  634.405018] Call Trace:
    [  634.405277] dump_stack (lib/dump_stack.c:52)
    [  634.405775] print_trailer (mm/slub.c:655)
    [  634.406361] object_err (mm/slub.c:662)
    [  634.406824] kasan_report_error (mm/kasan/report.c:138 mm/kasan/report.c:236)
    [  634.409581] __asan_report_load4_noabort (mm/kasan/report.c:279)
    [  634.411355] x25_asy_open_tty (drivers/net/wan/x25_asy.c:559 (discriminator 1))
    [  634.413997] tty_ldisc_open.isra.2 (drivers/tty/tty_ldisc.c:447)
    [  634.414549] tty_set_ldisc (drivers/tty/tty_ldisc.c:567)
    [  634.415057] tty_ioctl (drivers/tty/tty_io.c:2646 drivers/tty/tty_io.c:2879)
    [  634.423524] do_vfs_ioctl (fs/ioctl.c:43 fs/ioctl.c:607)
    [  634.427491] SyS_ioctl (fs/ioctl.c:622 fs/ioctl.c:613)
    [  634.427945] entry_SYSCALL_64_fastpath (arch/x86/entry/entry_64.S:188)

Cc: Tilman Schmidt <tilman@imap.cc>
Cc: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Peter Hurley <peter@hurleysoftware.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Change-Id: Ibed6feadfb9706d478f93feec3b240aecfc64af3
Bug: 30951112

8 years agoUPSTREAM: netfilter: nfnetlink: correctly validate length of batch messages
Phil Turnbull [Tue, 2 Feb 2016 18:36:45 +0000 (13:36 -0500)]
UPSTREAM: netfilter: nfnetlink: correctly validate length of batch messages

(cherry picked from commit c58d6c93680f28ac58984af61d0a7ebf4319c241)

If nlh->nlmsg_len is zero then an infinite loop is triggered because
'skb_pull(skb, msglen);' pulls zero bytes.

The calculation in nlmsg_len() underflows if 'nlh->nlmsg_len <
NLMSG_HDRLEN' which bypasses the length validation and will later
trigger an out-of-bound read.

If the length validation does fail then the malformed batch message is
copied back to userspace. However, we cannot do this because the
nlh->nlmsg_len can be invalid. This leads to an out-of-bounds read in
netlink_ack:

    [   41.455421] ==================================================================
    [   41.456431] BUG: KASAN: slab-out-of-bounds in memcpy+0x1d/0x40 at addr ffff880119e79340
    [   41.456431] Read of size 4294967280 by task a.out/987
    [   41.456431] =============================================================================
    [   41.456431] BUG kmalloc-512 (Not tainted): kasan: bad access detected
    [   41.456431] -----------------------------------------------------------------------------
    ...
    [   41.456431] Bytes b4 ffff880119e79310: 00 00 00 00 d5 03 00 00 b0 fb fe ff 00 00 00 00  ................
    [   41.456431] Object ffff880119e79320: 20 00 00 00 10 00 05 00 00 00 00 00 00 00 00 00   ...............
    [   41.456431] Object ffff880119e79330: 14 00 0a 00 01 03 fc 40 45 56 11 22 33 10 00 05  .......@EV."3...
    [   41.456431] Object ffff880119e79340: f0 ff ff ff 88 99 aa bb 00 14 00 0a 00 06 fe fb  ................
                                            ^^ start of batch nlmsg with
                                               nlmsg_len=4294967280
    ...
    [   41.456431] Memory state around the buggy address:
    [   41.456431]  ffff880119e79400: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
    [   41.456431]  ffff880119e79480: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
    [   41.456431] >ffff880119e79500: 00 00 00 00 fc fc fc fc fc fc fc fc fc fc fc fc
    [   41.456431]                                ^
    [   41.456431]  ffff880119e79580: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
    [   41.456431]  ffff880119e79600: fc fc fc fc fc fc fc fc fc fc fb fb fb fb fb fb
    [   41.456431] ==================================================================

Fix this with better validation of nlh->nlmsg_len and by setting
NFNL_BATCH_FAILURE if any batch message fails length validation.

CAP_NET_ADMIN is required to trigger the bugs.

Fixes: 9ea2aa8b7dba ("netfilter: nfnetlink: validate nfnetlink header from batch")
Signed-off-by: Phil Turnbull <phil.turnbull@oracle.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Change-Id: Id3e15c40cb464bf2791af907c235d8a316b2449c
Bug: 30947055

8 years agocpuset: Make cpusets restore on hotplug
Riley Andrews [Tue, 6 Sep 2016 22:16:25 +0000 (15:16 -0700)]
cpuset: Make cpusets restore on hotplug

This deliberately changes the behavior of the per-cpuset
cpus file to not be effected by hotplug. When a cpu is offlined,
it will be removed from the cpuset/cpus file. When a cpu is onlined,
if the cpuset originally requested that that cpu was part of the cpuset,
that cpu will be restored to the cpuset. The cpus files still
have to be hierachical, but the ranges no longer have to be out of
the currently online cpus, just the physically present cpus.

Change-Id: I22cdf33e7d312117bcefba1aeb0125e1ada289a9
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
8 years agoUPSTREAM: mm/slub: support left redzone
Joonsoo Kim [Tue, 15 Mar 2016 21:55:12 +0000 (14:55 -0700)]
UPSTREAM: mm/slub: support left redzone

SLUB already has a redzone debugging feature.  But it is only positioned
at the end of object (aka right redzone) so it cannot catch left oob.
Although current object's right redzone acts as left redzone of next
object, first object in a slab cannot take advantage of this effect.
This patch explicitly adds a left red zone to each object to detect left
oob more precisely.

Background:

Someone complained to me that left OOB doesn't catch even if KASAN is
enabled which does page allocation debugging.  That page is out of our
control so it would be allocated when left OOB happens and, in this
case, we can't find OOB.  Moreover, SLUB debugging feature can be
enabled without page allocator debugging and, in this case, we will miss
that OOB.

Before trying to implement, I expected that changes would be too
complex, but, it doesn't look that complex to me now.  Almost changes
are applied to debug specific functions so I feel okay.

Signed-off-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Change-Id: Ib893a17ecabd692e6c402e864196bf89cd6781a5
(cherry picked from commit d86bd1bece6fc41d59253002db5441fe960a37f6)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: Make the hardened user-copy code depend on having a hardened allocator
Linus Torvalds [Fri, 19 Aug 2016 19:47:01 +0000 (12:47 -0700)]
UPSTREAM: Make the hardened user-copy code depend on having a hardened allocator

The kernel test robot reported a usercopy failure in the new hardened
sanity checks, due to a page-crossing copy of the FPU state into the
task structure.

This happened because the kernel test robot was testing with SLOB, which
doesn't actually do the required book-keeping for slab allocations, and
as a result the hardening code didn't realize that the task struct
allocation was one single allocation - and the sanity checks fail.

Since SLOB doesn't even claim to support hardening (and you really
shouldn't use it), the straightforward solution is to just make the
usercopy hardening code depend on the allocator supporting it.

Reported-by: kernel test robot <xiaolong.ye@intel.com>
Cc: Kees Cook <keescook@chromium.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Change-Id: I37d51f866f873341bf7d5297249899b852e1c6ce
(cherry picked from commit 6040e57658eee6eb1315a26119101ca832d1f854)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoAndroid: MMC/UFS IO Latency Histograms.
Mohan Srinivasan [Fri, 26 Aug 2016 01:31:01 +0000 (18:31 -0700)]
Android: MMC/UFS IO Latency Histograms.

This patch adds a new sysfs node (latency_hist) and reports
IO (svc time) latency histograms. Disabled by default, can be
enabled by echoing 0 into latency_hist, stats can be cleared
by writing 2 into latency_hist. This commit fixes the 32 bit
build breakage in the previous commit. Tested on both 32 bit
and 64 bit arm devices.

Bug: 30677035
Change-Id: I9a615a16616d80f87e75676ac4d078a5c429dcf9
Signed-off-by: Mohan Srinivasan <srmohan@google.com>
8 years agoUPSTREAM: usercopy: fix overlap check for kernel text
Josh Poimboeuf [Mon, 22 Aug 2016 16:53:59 +0000 (11:53 -0500)]
UPSTREAM: usercopy: fix overlap check for kernel text

When running with a local patch which moves the '_stext' symbol to the
very beginning of the kernel text area, I got the following panic with
CONFIG_HARDENED_USERCOPY:

  usercopy: kernel memory exposure attempt detected from ffff88103dfff000 (<linear kernel text>) (4096 bytes)
  ------------[ cut here ]------------
  kernel BUG at mm/usercopy.c:79!
  invalid opcode: 0000 [#1] SMP
  ...
  CPU: 0 PID: 4800 Comm: cp Not tainted 4.8.0-rc3.after+ #1
  Hardware name: Dell Inc. PowerEdge R720/0X3D66, BIOS 2.5.4 01/22/2016
  task: ffff880817444140 task.stack: ffff880816274000
  RIP: 0010:[<ffffffff8121c796>] __check_object_size+0x76/0x413
  RSP: 0018:ffff880816277c40 EFLAGS: 00010246
  RAX: 000000000000006b RBX: ffff88103dfff000 RCX: 0000000000000000
  RDX: 0000000000000000 RSI: ffff88081f80dfa8 RDI: ffff88081f80dfa8
  RBP: ffff880816277c90 R08: 000000000000054c R09: 0000000000000000
  R10: 0000000000000005 R11: 0000000000000006 R12: 0000000000001000
  R13: ffff88103e000000 R14: ffff88103dffffff R15: 0000000000000001
  FS:  00007fb9d1750800(0000) GS:ffff88081f800000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00000000021d2000 CR3: 000000081a08f000 CR4: 00000000001406f0
  Stack:
   ffff880816277cc8 0000000000010000 000000043de07000 0000000000000000
   0000000000001000 ffff880816277e60 0000000000001000 ffff880816277e28
   000000000000c000 0000000000001000 ffff880816277ce8 ffffffff8136c3a6
  Call Trace:
   [<ffffffff8136c3a6>] copy_page_to_iter_iovec+0xa6/0x1c0
   [<ffffffff8136e766>] copy_page_to_iter+0x16/0x90
   [<ffffffff811970e3>] generic_file_read_iter+0x3e3/0x7c0
   [<ffffffffa06a738d>] ? xfs_file_buffered_aio_write+0xad/0x260 [xfs]
   [<ffffffff816e6262>] ? down_read+0x12/0x40
   [<ffffffffa06a61b1>] xfs_file_buffered_aio_read+0x51/0xc0 [xfs]
   [<ffffffffa06a6692>] xfs_file_read_iter+0x62/0xb0 [xfs]
   [<ffffffff812224cf>] __vfs_read+0xdf/0x130
   [<ffffffff81222c9e>] vfs_read+0x8e/0x140
   [<ffffffff81224195>] SyS_read+0x55/0xc0
   [<ffffffff81003a47>] do_syscall_64+0x67/0x160
   [<ffffffff816e8421>] entry_SYSCALL64_slow_path+0x25/0x25
  RIP: 0033:[<00007fb9d0c33c00>] 0x7fb9d0c33c00
  RSP: 002b:00007ffc9c262f28 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
  RAX: ffffffffffffffda RBX: fffffffffff8ffff RCX: 00007fb9d0c33c00
  RDX: 0000000000010000 RSI: 00000000021c3000 RDI: 0000000000000004
  RBP: 00000000021c3000 R08: 0000000000000000 R09: 00007ffc9c264d6c
  R10: 00007ffc9c262c50 R11: 0000000000000246 R12: 0000000000010000
  R13: 00007ffc9c2630b0 R14: 0000000000000004 R15: 0000000000010000
  Code: 81 48 0f 44 d0 48 c7 c6 90 4d a3 81 48 c7 c0 bb b3 a2 81 48 0f 44 f0 4d 89 e1 48 89 d9 48 c7 c7 68 16 a3 81 31 c0 e8 f4 57 f7 ff <0f> 0b 48 8d 90 00 40 00 00 48 39 d3 0f 83 22 01 00 00 48 39 c3
  RIP  [<ffffffff8121c796>] __check_object_size+0x76/0x413
   RSP <ffff880816277c40>

The checked object's range [ffff88103dfff000ffff88103e000000) is
valid, so there shouldn't have been a BUG.  The hardened usercopy code
got confused because the range's ending address is the same as the
kernel's text starting address at 0xffff88103e000000.  The overlap check
is slightly off.

Fixes: f5509cc18daa ("mm: Hardened usercopy")
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: I839dbf4ddbb4d9874026a42abed557eb9b3f8bef
(cherry picked from commit 94cd97af690dd9537818dc9841d0ec68bb1dd877)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: usercopy: avoid potentially undefined behavior in pointer math
Eric Biggers [Fri, 19 Aug 2016 19:15:22 +0000 (12:15 -0700)]
UPSTREAM: usercopy: avoid potentially undefined behavior in pointer math

check_bogus_address() checked for pointer overflow using this expression,
where 'ptr' has type 'const void *':

ptr + n < ptr

Since pointer wraparound is undefined behavior, gcc at -O2 by default
treats it like the following, which would not behave as intended:

(long)n < 0

Fortunately, this doesn't currently happen for kernel code because kernel
code is compiled with -fno-strict-overflow.  But the expression should be
fixed anyway to use well-defined integer arithmetic, since it could be
treated differently by different compilers in the future or could be
reported by tools checking for undefined behavior.

Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Change-Id: I73b13be651cf35c03482f2014bf2c3dd291518ab
(cherry picked from commit 7329a655875a2f4bd6984fe8a7e00a6981e802f3)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoUPSTREAM: unsafe_[get|put]_user: change interface to use a error target label
Linus Torvalds [Mon, 8 Aug 2016 20:02:01 +0000 (13:02 -0700)]
UPSTREAM: unsafe_[get|put]_user: change interface to use a error target label

When I initially added the unsafe_[get|put]_user() helpers in commit
5b24a7a2aa20 ("Add 'unsafe' user access functions for batched
accesses"), I made the mistake of modeling the interface on our
traditional __[get|put]_user() functions, which return zero on success,
or -EFAULT on failure.

That interface is fairly easy to use, but it's actually fairly nasty for
good code generation, since it essentially forces the caller to check
the error value for each access.

In particular, since the error handling is already internally
implemented with an exception handler, and we already use "asm goto" for
various other things, we could fairly easily make the error cases just
jump directly to an error label instead, and avoid the need for explicit
checking after each operation.

So switch the interface to pass in an error label, rather than checking
the error value in the caller.  Best do it now before we start growing
more users (the signal handling code in particular would be a good place
to use the new interface).

So rather than

if (unsafe_get_user(x, ptr))
... handle error ..

the interface is now

unsafe_get_user(x, ptr, label);

where an error during the user mode fetch will now just cause a jump to
'label' in the caller.

Right now the actual _implementation_ of this all still ends up being a
"if (err) goto label", and does not take advantage of any exception
label tricks, but for "unsafe_put_user()" in particular it should be
fairly straightforward to convert to using the exception table model.

Note that "unsafe_get_user()" is much harder to convert to a clever
exception table model, because current versions of gcc do not allow the
use of "asm goto" (for the exception) with output values (for the actual
value to be fetched).  But that is hopefully not a limitation in the
long term.

[ Also note that it might be a good idea to switch unsafe_get_user() to
  actually _return_ the value it fetches from user space, but this
  commit only changes the error handling semantics ]

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Change-Id: Ib905a84a04d46984320f6fd1056da4d72f3d6b53
(cherry picked from commit 1bd4403d86a1c06cb6cc9ac87664a0c9d3413d51)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoBACKPORT: arm64: mm: fix location of _etext
Ard Biesheuvel [Thu, 23 Jun 2016 13:53:17 +0000 (15:53 +0200)]
BACKPORT: arm64: mm: fix location of _etext

As Kees Cook notes in the ARM counterpart of this patch [0]:

  The _etext position is defined to be the end of the kernel text code,
  and should not include any part of the data segments. This interferes
  with things that might check memory ranges and expect executable code
  up to _etext.

In particular, Kees is referring to the HARDENED_USERCOPY patch set [1],
which rejects attempts to call copy_to_user() on kernel ranges containing
executable code, but does allow access to the .rodata segment. Regardless
of whether one may or may not agree with the distinction, it makes sense
for _etext to have the same meaning across architectures.

So let's put _etext where it belongs, between .text and .rodata, and fix
up existing references to use __init_begin instead, which unlike _end_rodata
includes the exception and notes sections as well.

The _etext references in kaslr.c are left untouched, since its references
to [_stext, _etext) are meant to capture potential jump instruction targets,
and so disregarding .rodata is actually an improvement here.

[0] http://article.gmane.org/gmane.linux.kernel/2245084
[1] http://thread.gmane.org/gmane.linux.kernel.hardened.devel/2502

Reported-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
(cherry picked from commit 9fdc14c55cd6579d619ccd9d40982e0805e62b6d)
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoBACKPORT: ARM: 8583/1: mm: fix location of _etext
Kees Cook [Thu, 23 Jun 2016 20:28:47 +0000 (21:28 +0100)]
BACKPORT: ARM: 8583/1: mm: fix location of _etext

The _etext position is defined to be the end of the kernel text code,
and should not include any part of the data segments. This interferes
with things that might check memory ranges and expect executable code
up to _etext. Just to be conservative, leave the kernel resource as
it was, using __init_begin instead of _etext as the end mark.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Change-Id: Ida514d1359dbe6f782f562ce29b4ba09ae72bfc0
(cherry picked from commit 14c4a533e0996f95a0a64dfd0b6252d788cebc74)
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
8 years agoBACKPORT: Don't show empty tag stats for unprivileged uids
Mohamad Ayyash [Wed, 11 May 2016 20:18:35 +0000 (13:18 -0700)]
BACKPORT: Don't show empty tag stats for unprivileged uids

BUG: 27577101
BUG: 27532522
Signed-off-by: Mohamad Ayyash <mkayyash@google.com>
8 years agoUPSTREAM: tcp: fix use after free in tcp_xmit_retransmit_queue()
Eric Dumazet [Wed, 17 Aug 2016 12:56:26 +0000 (05:56 -0700)]
UPSTREAM: tcp: fix use after free in tcp_xmit_retransmit_queue()

(cherry picked from commit bb1fceca22492109be12640d49f5ea5a544c6bb4)

When tcp_sendmsg() allocates a fresh and empty skb, it puts it at the
tail of the write queue using tcp_add_write_queue_tail()

Then it attempts to copy user data into this fresh skb.

If the copy fails, we undo the work and remove the fresh skb.

Unfortunately, this undo lacks the change done to tp->highest_sack and
we can leave a dangling pointer (to a freed skb)

Later, tcp_xmit_retransmit_queue() can dereference this pointer and
access freed memory. For regular kernels where memory is not unmapped,
this might cause SACK bugs because tcp_highest_sack_seq() is buggy,
returning garbage instead of tp->snd_nxt, but with various debug
features like CONFIG_DEBUG_PAGEALLOC, this can crash the kernel.

This bug was found by Marco Grassi thanks to syzkaller.

Fixes: 6859d49475d4 ("[TCP]: Abstract tp->highest_sack accessing & point to next skb")
Reported-by: Marco Grassi <marco.gra@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>
Cc: Yuchung Cheng <ycheng@google.com>
Cc: Neal Cardwell <ncardwell@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Reviewed-by: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change-Id: I58bb02d6e4e399612e8580b9e02d11e661df82f5
Bug: 31183296

8 years agoANDROID: base-cfg: drop SECCOMP_FILTER config
Amit Pundir [Fri, 2 Sep 2016 04:43:21 +0000 (10:13 +0530)]
ANDROID: base-cfg: drop SECCOMP_FILTER config

Don't need to set SECCOMP_FILTER explicitly since CONFIG_SECCOMP=y will
select that config anyway.

Fixes: a49dcf2e745c ("ANDROID: base-cfg: enable SECCOMP config")
Change-Id: Iff18ed4d2db5a55b9f9480d5ecbeef7b818b3837
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoUPSTREAM: [media] xc2028: unlock on error in xc2028_set_config()
Dan Carpenter [Wed, 3 Feb 2016 15:34:00 +0000 (13:34 -0200)]
UPSTREAM: [media] xc2028: unlock on error in xc2028_set_config()

(cherry picked from commit 210bd104c6acd31c3c6b8b075b3f12d4a9f6b60d)

We have to unlock before returning -ENOMEM.

Fixes: 8dfbcc4351a0 ('[media] xc2028: avoid use after free')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Change-Id: I7b6ba9fde5c6e29467e6de23d398af2fe56e2547
Bug: 30946097

8 years agoUPSTREAM: [media] xc2028: avoid use after free
Mauro Carvalho Chehab [Thu, 28 Jan 2016 11:22:44 +0000 (09:22 -0200)]
UPSTREAM: [media] xc2028: avoid use after free

(cherry picked from commit 8dfbcc4351a0b6d2f2d77f367552f48ffefafe18)

If struct xc2028_config is passed without a firmware name,
the following trouble may happen:

[11009.907205] xc2028 5-0061: type set to XCeive xc2028/xc3028 tuner
[11009.907491] ==================================================================
[11009.907750] BUG: KASAN: use-after-free in strcmp+0x96/0xb0 at addr ffff8803bd78ab40
[11009.907992] Read of size 1 by task modprobe/28992
[11009.907994] =============================================================================
[11009.907997] BUG kmalloc-16 (Tainted: G        W      ): kasan: bad access detected
[11009.907999] -----------------------------------------------------------------------------

[11009.908008] INFO: Allocated in xhci_urb_enqueue+0x214/0x14c0 [xhci_hcd] age=0 cpu=3 pid=28992
[11009.908012]  ___slab_alloc+0x581/0x5b0
[11009.908014]  __slab_alloc+0x51/0x90
[11009.908017]  __kmalloc+0x27b/0x350
[11009.908022]  xhci_urb_enqueue+0x214/0x14c0 [xhci_hcd]
[11009.908026]  usb_hcd_submit_urb+0x1e8/0x1c60
[11009.908029]  usb_submit_urb+0xb0e/0x1200
[11009.908032]  usb_serial_generic_write_start+0xb6/0x4c0
[11009.908035]  usb_serial_generic_write+0x92/0xc0
[11009.908039]  usb_console_write+0x38a/0x560
[11009.908045]  call_console_drivers.constprop.14+0x1ee/0x2c0
[11009.908051]  console_unlock+0x40d/0x900
[11009.908056]  vprintk_emit+0x4b4/0x830
[11009.908061]  vprintk_default+0x1f/0x30
[11009.908064]  printk+0x99/0xb5
[11009.908067]  kasan_report_error+0x10a/0x550
[11009.908070]  __asan_report_load1_noabort+0x43/0x50
[11009.908074] INFO: Freed in xc2028_set_config+0x90/0x630 [tuner_xc2028] age=1 cpu=3 pid=28992
[11009.908077]  __slab_free+0x2ec/0x460
[11009.908080]  kfree+0x266/0x280
[11009.908083]  xc2028_set_config+0x90/0x630 [tuner_xc2028]
[11009.908086]  xc2028_attach+0x310/0x8a0 [tuner_xc2028]
[11009.908090]  em28xx_attach_xc3028.constprop.7+0x1f9/0x30d [em28xx_dvb]
[11009.908094]  em28xx_dvb_init.part.3+0x8e4/0x5cf4 [em28xx_dvb]
[11009.908098]  em28xx_dvb_init+0x81/0x8a [em28xx_dvb]
[11009.908101]  em28xx_register_extension+0xd9/0x190 [em28xx]
[11009.908105]  em28xx_dvb_register+0x10/0x1000 [em28xx_dvb]
[11009.908108]  do_one_initcall+0x141/0x300
[11009.908111]  do_init_module+0x1d0/0x5ad
[11009.908114]  load_module+0x6666/0x9ba0
[11009.908117]  SyS_finit_module+0x108/0x130
[11009.908120]  entry_SYSCALL_64_fastpath+0x16/0x76
[11009.908123] INFO: Slab 0xffffea000ef5e280 objects=25 used=25 fp=0x          (null) flags=0x2ffff8000004080
[11009.908126] INFO: Object 0xffff8803bd78ab40 @offset=2880 fp=0x0000000000000001

[11009.908130] Bytes b4 ffff8803bd78ab30: 01 00 00 00 2a 07 00 00 9d 28 00 00 01 00 00 00  ....*....(......
[11009.908133] Object ffff8803bd78ab40: 01 00 00 00 00 00 00 00 b0 1d c3 6a 00 88 ff ff  ...........j....
[11009.908137] CPU: 3 PID: 28992 Comm: modprobe Tainted: G    B   W       4.5.0-rc1+ #43
[11009.908140] Hardware name:                  /NUC5i7RYB, BIOS RYBDWi35.86A.0350.2015.0812.1722 08/12/2015
[11009.908142]  ffff8803bd78a000 ffff8802c273f1b8 ffffffff81932007 ffff8803c6407a80
[11009.908148]  ffff8802c273f1e8 ffffffff81556759 ffff8803c6407a80 ffffea000ef5e280
[11009.908153]  ffff8803bd78ab40 dffffc0000000000 ffff8802c273f210 ffffffff8155ccb4
[11009.908158] Call Trace:
[11009.908162]  [<ffffffff81932007>] dump_stack+0x4b/0x64
[11009.908165]  [<ffffffff81556759>] print_trailer+0xf9/0x150
[11009.908168]  [<ffffffff8155ccb4>] object_err+0x34/0x40
[11009.908171]  [<ffffffff8155f260>] kasan_report_error+0x230/0x550
[11009.908175]  [<ffffffff81237d71>] ? trace_hardirqs_off_caller+0x21/0x290
[11009.908179]  [<ffffffff8155e926>] ? kasan_unpoison_shadow+0x36/0x50
[11009.908182]  [<ffffffff8155f5c3>] __asan_report_load1_noabort+0x43/0x50
[11009.908185]  [<ffffffff8155ea00>] ? __asan_register_globals+0x50/0xa0
[11009.908189]  [<ffffffff8194cea6>] ? strcmp+0x96/0xb0
[11009.908192]  [<ffffffff8194cea6>] strcmp+0x96/0xb0
[11009.908196]  [<ffffffffa13ba4ac>] xc2028_set_config+0x15c/0x630 [tuner_xc2028]
[11009.908200]  [<ffffffffa13bac90>] xc2028_attach+0x310/0x8a0 [tuner_xc2028]
[11009.908203]  [<ffffffff8155ea78>] ? memset+0x28/0x30
[11009.908206]  [<ffffffffa13ba980>] ? xc2028_set_config+0x630/0x630 [tuner_xc2028]
[11009.908211]  [<ffffffffa157a59a>] em28xx_attach_xc3028.constprop.7+0x1f9/0x30d [em28xx_dvb]
[11009.908215]  [<ffffffffa157aa2a>] ? em28xx_dvb_init.part.3+0x37c/0x5cf4 [em28xx_dvb]
[11009.908219]  [<ffffffffa157a3a1>] ? hauppauge_hvr930c_init+0x487/0x487 [em28xx_dvb]
[11009.908222]  [<ffffffffa01795ac>] ? lgdt330x_attach+0x1cc/0x370 [lgdt330x]
[11009.908226]  [<ffffffffa01793e0>] ? i2c_read_demod_bytes.isra.2+0x210/0x210 [lgdt330x]
[11009.908230]  [<ffffffff812e87d0>] ? ref_module.part.15+0x10/0x10
[11009.908233]  [<ffffffff812e56e0>] ? module_assert_mutex_or_preempt+0x80/0x80
[11009.908238]  [<ffffffffa157af92>] em28xx_dvb_init.part.3+0x8e4/0x5cf4 [em28xx_dvb]
[11009.908242]  [<ffffffffa157a6ae>] ? em28xx_attach_xc3028.constprop.7+0x30d/0x30d [em28xx_dvb]
[11009.908245]  [<ffffffff8195222d>] ? string+0x14d/0x1f0
[11009.908249]  [<ffffffff8195381f>] ? symbol_string+0xff/0x1a0
[11009.908253]  [<ffffffff81953720>] ? uuid_string+0x6f0/0x6f0
[11009.908257]  [<ffffffff811a775e>] ? __kernel_text_address+0x7e/0xa0
[11009.908260]  [<ffffffff8104b02f>] ? print_context_stack+0x7f/0xf0
[11009.908264]  [<ffffffff812e9846>] ? __module_address+0xb6/0x360
[11009.908268]  [<ffffffff8137fdc9>] ? is_ftrace_trampoline+0x99/0xe0
[11009.908271]  [<ffffffff811a775e>] ? __kernel_text_address+0x7e/0xa0
[11009.908275]  [<ffffffff81240a70>] ? debug_check_no_locks_freed+0x290/0x290
[11009.908278]  [<ffffffff8104a24b>] ? dump_trace+0x11b/0x300
[11009.908282]  [<ffffffffa13e8143>] ? em28xx_register_extension+0x23/0x190 [em28xx]
[11009.908285]  [<ffffffff81237d71>] ? trace_hardirqs_off_caller+0x21/0x290
[11009.908289]  [<ffffffff8123ff56>] ? trace_hardirqs_on_caller+0x16/0x590
[11009.908292]  [<ffffffff812404dd>] ? trace_hardirqs_on+0xd/0x10
[11009.908296]  [<ffffffffa13e8143>] ? em28xx_register_extension+0x23/0x190 [em28xx]
[11009.908299]  [<ffffffff822dcbb0>] ? mutex_trylock+0x400/0x400
[11009.908302]  [<ffffffff810021a1>] ? do_one_initcall+0x131/0x300
[11009.908306]  [<ffffffff81296dc7>] ? call_rcu_sched+0x17/0x20
[11009.908309]  [<ffffffff8159e708>] ? put_object+0x48/0x70
[11009.908314]  [<ffffffffa1579f11>] em28xx_dvb_init+0x81/0x8a [em28xx_dvb]
[11009.908317]  [<ffffffffa13e81f9>] em28xx_register_extension+0xd9/0x190 [em28xx]
[11009.908320]  [<ffffffffa0150000>] ? 0xffffffffa0150000
[11009.908324]  [<ffffffffa0150010>] em28xx_dvb_register+0x10/0x1000 [em28xx_dvb]
[11009.908327]  [<ffffffff810021b1>] do_one_initcall+0x141/0x300
[11009.908330]  [<ffffffff81002070>] ? try_to_run_init_process+0x40/0x40
[11009.908333]  [<ffffffff8123ff56>] ? trace_hardirqs_on_caller+0x16/0x590
[11009.908337]  [<ffffffff8155e926>] ? kasan_unpoison_shadow+0x36/0x50
[11009.908340]  [<ffffffff8155e926>] ? kasan_unpoison_shadow+0x36/0x50
[11009.908343]  [<ffffffff8155e926>] ? kasan_unpoison_shadow+0x36/0x50
[11009.908346]  [<ffffffff8155ea37>] ? __asan_register_globals+0x87/0xa0
[11009.908350]  [<ffffffff8144da7b>] do_init_module+0x1d0/0x5ad
[11009.908353]  [<ffffffff812f2626>] load_module+0x6666/0x9ba0
[11009.908356]  [<ffffffff812e9c90>] ? symbol_put_addr+0x50/0x50
[11009.908361]  [<ffffffffa1580037>] ? em28xx_dvb_init.part.3+0x5989/0x5cf4 [em28xx_dvb]
[11009.908366]  [<ffffffff812ebfc0>] ? module_frob_arch_sections+0x20/0x20
[11009.908369]  [<ffffffff815bc940>] ? open_exec+0x50/0x50
[11009.908374]  [<ffffffff811671bb>] ? ns_capable+0x5b/0xd0
[11009.908377]  [<ffffffff812f5e58>] SyS_finit_module+0x108/0x130
[11009.908379]  [<ffffffff812f5d50>] ? SyS_init_module+0x1f0/0x1f0
[11009.908383]  [<ffffffff81004044>] ? lockdep_sys_exit_thunk+0x12/0x14
[11009.908394]  [<ffffffff822e6936>] entry_SYSCALL_64_fastpath+0x16/0x76
[11009.908396] Memory state around the buggy address:
[11009.908398]  ffff8803bd78aa00: 00 00 fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[11009.908401]  ffff8803bd78aa80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[11009.908403] >ffff8803bd78ab00: fc fc fc fc fc fc fc fc 00 00 fc fc fc fc fc fc
[11009.908405]                                            ^
[11009.908407]  ffff8803bd78ab80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[11009.908409]  ffff8803bd78ac00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[11009.908411] ==================================================================

In order to avoid it, let's set the cached value of the firmware
name to NULL after freeing it. While here, return an error if
the memory allocation fails.

Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Change-Id: I945c841dcfb45de2056267e4aa50bbe176b527cf
Bug: 30946097

8 years agoANDROID: base-cfg: enable SECCOMP config
Yongqin Liu [Thu, 1 Sep 2016 16:36:04 +0000 (22:06 +0530)]
ANDROID: base-cfg: enable SECCOMP config

Enable following seccomp configs

CONFIG_SECCOMP=y
CONFIG_SECCOMP_FILTER=y

Otherwise we will get mediacode error like this on Android N:

E /system/bin/mediaextractor: libminijail: prctl(PR_SET_SECCOMP, SECCOMP_MODE_FILTER): Invalid argument

Change-Id: I2477b6a2cfdded5c0ebf6ffbb6150b0e5fe2ba12
Signed-off-by: Yongqin Liu <yongqin.liu@linaro.org>
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoANDROID: rcu_sync: Export rcu_sync_lockdep_assert
Guenter Roeck [Wed, 31 Aug 2016 16:52:16 +0000 (09:52 -0700)]
ANDROID: rcu_sync: Export rcu_sync_lockdep_assert

x86_64:allmodconfig fails to build with the following error.

ERROR: "rcu_sync_lockdep_assert" [kernel/locking/locktorture.ko] undefined!

Introduced by commit 3228c5eb7af2 ("RFC: FROMLIST: locking/percpu-rwsem:
Optimize readers and reduce global impact"). The applied upstream version
exports the missing symbol, so let's do the same.

Change-Id: If4e516715c3415fe8c82090f287174857561550d
Fixes: 3228c5eb7af2 ("RFC: FROMLIST: locking/percpu-rwsem: Optimize ...")
Signed-off-by: Guenter Roeck <groeck@chromium.org>
8 years agoRFC: FROMLIST: cgroup: reduce read locked section of cgroup_threadgroup_rwsem during...
Balbir Singh [Wed, 10 Aug 2016 19:43:06 +0000 (15:43 -0400)]
RFC: FROMLIST: cgroup: reduce read locked section of cgroup_threadgroup_rwsem during fork

cgroup_threadgroup_rwsem is acquired in read mode during process exit
and fork.  It is also grabbed in write mode during
__cgroups_proc_write().  I've recently run into a scenario with lots
of memory pressure and OOM and I am beginning to see

systemd

 __switch_to+0x1f8/0x350
 __schedule+0x30c/0x990
 schedule+0x48/0xc0
 percpu_down_write+0x114/0x170
 __cgroup_procs_write.isra.12+0xb8/0x3c0
 cgroup_file_write+0x74/0x1a0
 kernfs_fop_write+0x188/0x200
 __vfs_write+0x6c/0xe0
 vfs_write+0xc0/0x230
 SyS_write+0x6c/0x110
 system_call+0x38/0xb4

This thread is waiting on the reader of cgroup_threadgroup_rwsem to
exit.  The reader itself is under memory pressure and has gone into
reclaim after fork. There are times the reader also ends up waiting on
oom_lock as well.

 __switch_to+0x1f8/0x350
 __schedule+0x30c/0x990
 schedule+0x48/0xc0
 jbd2_log_wait_commit+0xd4/0x180
 ext4_evict_inode+0x88/0x5c0
 evict+0xf8/0x2a0
 dispose_list+0x50/0x80
 prune_icache_sb+0x6c/0x90
 super_cache_scan+0x190/0x210
 shrink_slab.part.15+0x22c/0x4c0
 shrink_zone+0x288/0x3c0
 do_try_to_free_pages+0x1dc/0x590
 try_to_free_pages+0xdc/0x260
 __alloc_pages_nodemask+0x72c/0xc90
 alloc_pages_current+0xb4/0x1a0
 page_table_alloc+0xc0/0x170
 __pte_alloc+0x58/0x1f0
 copy_page_range+0x4ec/0x950
 copy_process.isra.5+0x15a0/0x1870
 _do_fork+0xa8/0x4b0
 ppc_clone+0x8/0xc

In the meanwhile, all processes exiting/forking are blocked almost
stalling the system.

This patch moves the threadgroup_change_begin from before
cgroup_fork() to just before cgroup_canfork().  There is no nee to
worry about threadgroup changes till the task is actually added to the
threadgroup.  This avoids having to call reclaim with
cgroup_threadgroup_rwsem held.

tj: Subject and description edits.

Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Acked-by: Zefan Li <lizefan@huawei.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: stable@vger.kernel.org # v4.2+
Signed-off-by: Tejun Heo <tj@kernel.org>
[jstultz: Cherry-picked from:
 git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup.git 568ac888215c7f]
Change-Id: Ie8ece84fb613cf6a7b08cea1468473a8df2b9661
Signed-off-by: John Stultz <john.stultz@linaro.org>
8 years agoRFC: FROMLIST: cgroup: avoid synchronize_sched() in __cgroup_procs_write()
Peter Zijlstra [Thu, 11 Aug 2016 16:54:13 +0000 (18:54 +0200)]
RFC: FROMLIST: cgroup: avoid synchronize_sched() in __cgroup_procs_write()

The current percpu-rwsem read side is entirely free of serializing insns
at the cost of having a synchronize_sched() in the write path.

The latency of the synchronize_sched() is too high for cgroups. The
commit 1ed1328792ff talks about the write path being a fairly cold path
but this is not the case for Android which moves task to the foreground
cgroup and back around binder IPC calls from foreground processes to
background processes, so it is significantly hotter than human initiated
operations.

Switch cgroup_threadgroup_rwsem into the slow mode for now to avoid the
problem, hopefully it should not be that slow after another commit
80127a39681b ("locking/percpu-rwsem: Optimize readers and reduce global
impact").

We could just add rcu_sync_enter() into cgroup_init() but we do not want
another synchronize_sched() at boot time, so this patch adds the new helper
which doesn't block but currently can only be called before the first use.

Cc: Tejun Heo <tj@kernel.org>
Cc: Paul McKenney <paulmck@linux.vnet.ibm.com>
Reported-by: John Stultz <john.stultz@linaro.org>
Reported-by: Dmitry Shmidt <dimitrysh@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
[jstultz: backported to 4.4]
Change-Id: I34aa9c394d3052779b56976693e96d861bd255f2
Mailing-list-URL: https://lkml.org/lkml/2016/8/11/557
Signed-off-by: John Stultz <john.stultz@linaro.org>
8 years agoRFC: FROMLIST: locking/percpu-rwsem: Optimize readers and reduce global impact
Peter Zijlstra [Tue, 9 Aug 2016 15:44:12 +0000 (08:44 -0700)]
RFC: FROMLIST: locking/percpu-rwsem: Optimize readers and reduce global impact

Currently the percpu-rwsem switches to (global) atomic ops while a
writer is waiting; which could be quite a while and slows down
releasing the readers.

This patch cures this problem by ordering the reader-state vs
reader-count (see the comments in __percpu_down_read() and
percpu_down_write()). This changes a global atomic op into a full
memory barrier, which doesn't have the global cacheline contention.

This also enables using the percpu-rwsem with rcu_sync disabled in order
to bias the implementation differently, reducing the writer latency by
adding some cost to readers.

Mailing-list-URL: https://lkml.org/lkml/2016/8/9/181
Cc: Paul McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[jstultz: Backported to 4.4]
Change-Id: I8ea04b4dca2ec36f1c2469eccafde1423490572f
Signed-off-by: John Stultz <john.stultz@linaro.org>
8 years agonet: ipv6: Fix ping to link-local addresses.
Lorenzo Colitti [Fri, 12 Aug 2016 16:13:38 +0000 (01:13 +0900)]
net: ipv6: Fix ping to link-local addresses.

ping_v6_sendmsg does not set flowi6_oif in response to
sin6_scope_id or sk_bound_dev_if, so it is not possible to use
these APIs to ping an IPv6 address on a different interface.
Instead, it sets flowi6_iif, which is incorrect but harmless.

Stop setting flowi6_iif, and support various ways of setting oif
in the same priority order used by udpv6_sendmsg.

[Backport of net 5e457896986e16c440c97bb94b9ccd95dd157292]

Bug: 29370996
Change-Id: Ibe1b9434c00ed96f1e30acb110734c6570b087b8
Tested: https://android-review.googlesource.com/#/c/254470/
Signed-off-by: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
8 years agoipv6: fix endianness error in icmpv6_err
Hannes Frederic Sowa [Sat, 11 Jun 2016 18:32:06 +0000 (20:32 +0200)]
ipv6: fix endianness error in icmpv6_err

IPv6 ping socket error handler doesn't correctly convert the new 32 bit
mtu to host endianness before using.

[Cherry-pick of net dcb94b88c09ce82a80e188d49bcffdc83ba215a6]

Bug: 29370996
Change-Id: Iea0ca79f16c2a1366d82b3b0a3097093d18da8b7
Cc: Lorenzo Colitti <lorenzo@google.com>
Fixes: 6d0bfe22611602f ("net: ipv6: Add IPv6 support to the ping socket.")
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
8 years agoANDROID: dm: android-verity: Allow android-verity to be compiled as an independent...
Badhri Jagan Sridharan [Tue, 23 Aug 2016 18:32:37 +0000 (11:32 -0700)]
ANDROID: dm: android-verity: Allow android-verity to be compiled as an independent module

Exports the device mapper callbacks of linear and dm-verity-target
methods.

Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I0358be0615c431dce3cc78575aaac4ccfe3aacd7

8 years agoMerge branch 'linux-linaro-lsk-v4.4' into linux-linaro-lsk-v4.4-android
Alex Shi [Tue, 30 Aug 2016 02:27:13 +0000 (10:27 +0800)]
Merge branch 'linux-linaro-lsk-v4.4' into linux-linaro-lsk-v4.4-android

Conflicts:
arch/arm/Kconfig

8 years agoMerge remote-tracking branch 'v4.4/topic/mm-kaslr-pax_usercopy' into linux-linaro...
Alex Shi [Sat, 27 Aug 2016 03:27:14 +0000 (11:27 +0800)]
Merge remote-tracking branch 'v4.4/topic/mm-kaslr-pax_usercopy' into linux-linaro-lsk-v4.4

8 years agomm: SLUB hardened usercopy support
Kees Cook [Thu, 23 Jun 2016 22:24:05 +0000 (15:24 -0700)]
mm: SLUB hardened usercopy support

Under CONFIG_HARDENED_USERCOPY, this adds object size checking to the
SLUB allocator to catch any copies that may span objects. Includes a
redzone handling fix discovered by Michael Ellerman.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
Reviwed-by: Laura Abbott <labbott@redhat.com>
(cherry picked from commit ed18adc1cdd00a5c55a20fbdaed4804660772281)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm: SLAB hardened usercopy support
Kees Cook [Thu, 23 Jun 2016 22:20:59 +0000 (15:20 -0700)]
mm: SLAB hardened usercopy support

Under CONFIG_HARDENED_USERCOPY, this adds object size checking to the
SLAB allocator to catch any copies that may span objects.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
(cherry picked from commit 04385fc5e8fffed84425d909a783c0f0c587d847)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agos390/uaccess: Enable hardened usercopy
Kees Cook [Thu, 7 Jul 2016 18:38:39 +0000 (11:38 -0700)]
s390/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on s390.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 97433ea4fda62349bfa42089455593cbcb57e06c)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agosparc/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:10:13 +0000 (15:10 -0700)]
sparc/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on sparc.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 9d9208a15800f9f06f102f9aac1e8b323c3b8575)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agopowerpc/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:10:01 +0000 (15:10 -0700)]
powerpc/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on powerpc.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
(cherry picked from commit 1d3c1324746fed0e34a5b94d3ed303e7521ed603)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoia64/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:09:50 +0000 (15:09 -0700)]
ia64/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on ia64.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 73d35887e24da77e8d1321b2e92bd9b9128e2fc2)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoarm64/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:59:42 +0000 (15:59 -0700)]
arm64/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on arm64. As done by KASAN in -next,
renames the low-level functions to __arch_copy_*_user() so a static inline
can do additional work before the copy.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit faf5b63e294151d6ac24ca6906d6f221bd3496cd)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoARM: uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:06:53 +0000 (15:06 -0700)]
ARM: uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on arm.

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit dfd45b6103c973bfcea2341d89e36faf947dbc33)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86/uaccess: Enable hardened usercopy
Kees Cook [Thu, 23 Jun 2016 22:04:01 +0000 (15:04 -0700)]
x86/uaccess: Enable hardened usercopy

Enables CONFIG_HARDENED_USERCOPY checks on x86. This is done both in
copy_*_user() and __copy_*_user() because copy_*_user() actually calls
down to _copy_*_user() and not __copy_*_user().

Based on code from PaX and grsecurity.

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
(cherry picked from commit 5b710f34e194c6b7710f69fdb5d798fdf35b98c1)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: remove more uaccess_32.h complexity
Linus Torvalds [Mon, 23 May 2016 00:21:27 +0000 (17:21 -0700)]
x86: remove more uaccess_32.h complexity

I'm looking at trying to possibly merge the 32-bit and 64-bit versions
of the x86 uaccess.h implementation, but first this needs to be cleaned
up.

For example, the 32-bit version of "__copy_from_user_inatomic()" is
mostly the special cases for the constant size, and it's actually almost
never relevant.  Most users aren't actually using a constant size
anyway, and the few cases that do small constant copies are better off
just using __get_user() instead.

So get rid of the unnecessary complexity.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit bd28b14591b98f696bc9f94c5ba2e598ca487dfd)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: remove pointless uaccess_32.h complexity
Linus Torvalds [Sun, 22 May 2016 21:19:37 +0000 (14:19 -0700)]
x86: remove pointless uaccess_32.h complexity

I'm looking at trying to possibly merge the 32-bit and 64-bit versions
of the x86 uaccess.h implementation, but first this needs to be cleaned
up.

For example, the 32-bit version of "__copy_to_user_inatomic()" is mostly
the special cases for the constant size, and it's actually never
relevant.  Every user except for one aren't actually using a constant
size anyway, and the one user that uses it is better off just using
__put_user() instead.

So get rid of the unnecessary complexity.

[ The same cleanup should likely happen to __copy_from_user_inatomic()
  as well, but that one has a lot more users that I need to take a look
  at first ]

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 5b09c3edecd37ec1a52fbd5ae97a19734edc7a77)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: fix SMAP in 32-bit environments
Linus Torvalds [Tue, 23 Feb 2016 22:58:52 +0000 (14:58 -0800)]
x86: fix SMAP in 32-bit environments

In commit 11f1a4b9755f ("x86: reorganize SMAP handling in user space
accesses") I changed how the stac/clac instructions were generated
around the user space accesses, which then made it possible to do
batched accesses efficiently for user string copies etc.

However, in doing so, I completely spaced out, and didn't even think
about the 32-bit case.  And nobody really even seemed to notice, because
SMAP doesn't even exist until modern Skylake processors, and you'd have
to be crazy to run 32-bit kernels on a modern CPU.

Which brings us to Andy Lutomirski.

He actually tested the 32-bit kernel on new hardware, and noticed that
it doesn't work.  My bad.  The trivial fix is to add the required
uaccess begin/end markers around the raw accesses in <asm/uaccess_32.h>.

I feel a bit bad about this patch, just because that header file really
should be cleaned up to avoid all the duplicated code in it, and this
commit just expands on the problem.  But this just fixes the bug without
any bigger cleanup surgery.

Reported-and-tested-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit de9e478b9d49f3a0214310d921450cf5bb4a21e6)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoUse the new batched user accesses in generic user string handling
Linus Torvalds [Thu, 17 Dec 2015 18:05:19 +0000 (10:05 -0800)]
Use the new batched user accesses in generic user string handling

This converts the generic user string functions to use the batched user
access functions.

It makes a big difference on Skylake, which is the first x86
microarchitecture to implement SMAP.  The STAC/CLAC instructions are not
very fast, and doing them for each access inside the loop that copies
strings from user space (which is what the pathname handling does for
every pathname the kernel uses, for example) is very inefficient.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 9fd4470ff4974c41b1db43c3b355b9085af9c12a)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoAdd 'unsafe' user access functions for batched accesses
Linus Torvalds [Thu, 17 Dec 2015 17:57:27 +0000 (09:57 -0800)]
Add 'unsafe' user access functions for batched accesses

The naming is meant to discourage random use: the helper functions are
not really any more "unsafe" than the traditional double-underscore
functions (which need the address range checking), but they do need even
more infrastructure around them, and should not be used willy-nilly.

In addition to checking the access range, these user access functions
require that you wrap the user access with a "user_acess_{begin,end}()"
around it.

That allows architectures that implement kernel user access control
(x86: SMAP, arm64: PAN) to do the user access control in the wrapping
user_access_begin/end part, and then batch up the actual user space
accesses using the new interfaces.

The main (and hopefully only) use for these are for core generic access
helpers, initially just the generic user string functions
(strnlen_user() and strncpy_from_user()).

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 5b24a7a2aa2040c8c50c3b71122901d01661ff78)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agox86: reorganize SMAP handling in user space accesses
Linus Torvalds [Thu, 17 Dec 2015 17:45:09 +0000 (09:45 -0800)]
x86: reorganize SMAP handling in user space accesses

This reorganizes how we do the stac/clac instructions in the user access
code.  Instead of adding the instructions directly to the same inline
asm that does the actual user level access and exception handling, add
them at a higher level.

This is mainly preparation for the next step, where we will expose an
interface to allow users to mark several accesses together as being user
space accesses, but it does already clean up some code:

 - the inlined trivial cases of copy_in_user() now do stac/clac just
   once over the accesses: they used to do one pair around the user
   space read, and another pair around the write-back.

 - the {get,put}_user_ex() macros that are used with the catch/try
   handling don't do any stac/clac at all, because that happens in the
   try/catch surrounding them.

Other than those two cleanups that happened naturally from the
re-organization, this should not make any difference. Yet.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 11f1a4b9755f5dbc3e822a96502ebe9b044b14d8)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agomm: Hardened usercopy
Kees Cook [Tue, 7 Jun 2016 18:05:33 +0000 (11:05 -0700)]
mm: Hardened usercopy

This is the start of porting PAX_USERCOPY into the mainline kernel. This
is the first set of features, controlled by CONFIG_HARDENED_USERCOPY. The
work is based on code by PaX Team and Brad Spengler, and an earlier port
from Casey Schaufler. Additional non-slab page tests are from Rik van Riel.

This patch contains the logic for validating several conditions when
performing copy_to_user() and copy_from_user() on the kernel object
being copied to/from:
- address range doesn't wrap around
- address range isn't NULL or zero-allocated (with a non-zero copy size)
- if on the slab allocator:
  - object size must be less than or equal to copy size (when check is
    implemented in the allocator, which appear in subsequent patches)
- otherwise, object must not span page allocations (excepting Reserved
  and CMA ranges)
- if on the stack
  - object must not extend before/after the current process stack
  - object must be contained by a valid stack frame (when there is
    arch/build support for identifying stack frames)
- object must not overlap with kernel text

Signed-off-by: Kees Cook <keescook@chromium.org>
Tested-by: Valdis Kletnieks <valdis.kletnieks@vt.edu>
Tested-by: Michael Ellerman <mpe@ellerman.id.au>
(cherry picked from commit f5509cc18daa7f82bcc553be70df2117c8eedc16)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
Conflicts:
skip debug_page_ref and KCOV_INSTRUMENT in mm/Makefile

8 years agomm: Implement stack frame object validation
Kees Cook [Tue, 12 Jul 2016 23:19:48 +0000 (16:19 -0700)]
mm: Implement stack frame object validation

This creates per-architecture function arch_within_stack_frames() that
should validate if a given object is contained by a kernel stack frame.
Initial implementation is on x86.

This is based on code from PaX.

Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 0f60a8efe4005ab5e65ce000724b04d4ca04a199)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
Conflicts:
skip EBPF_JIT in arch/x86/Kconfig

8 years agomm: Add is_migrate_cma_page
Laura Abbott [Tue, 19 Jul 2016 22:00:04 +0000 (15:00 -0700)]
mm: Add is_migrate_cma_page

Code such as hardened user copy[1] needs a way to tell if a
page is CMA or not. Add is_migrate_cma_page in a similar way
to is_migrate_isolate_page.

[1]http://article.gmane.org/gmane.linux.kernel.mm/155238

Signed-off-by: Laura Abbott <labbott@redhat.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
(cherry picked from commit 7c15d9bb8231f998ae7dc0b72415f5215459f7fb)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
8 years agoUPSTREAM: usb: gadget: configfs: add mutex lock before unregister gadget
Winter Wang [Wed, 27 Jul 2016 02:03:19 +0000 (10:03 +0800)]
UPSTREAM: usb: gadget: configfs: add mutex lock before unregister gadget

There may be a race condition if f_fs calls unregister_gadget_item in
ffs_closed() when unregister_gadget is called by UDC store at the same time.
this leads to a kernel NULL pointer dereference:

[  310.644928] Unable to handle kernel NULL pointer dereference at virtual address 00000004
[  310.645053] init: Service 'adbd' is being killed...
[  310.658938] pgd = c9528000
[  310.662515] [00000004] *pgd=19451831, *pte=00000000, *ppte=00000000
[  310.669702] Internal error: Oops: 817 [#1] PREEMPT SMP ARM
[  310.675211] Modules linked in:
[  310.678294] CPU: 0 PID: 1537 Comm: ->transport Not tainted 4.1.15-03725-g793404c #2
[  310.685958] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
[  310.692493] task: c8e24200 ti: c945e000 task.ti: c945e000
[  310.697911] PC is at usb_gadget_unregister_driver+0xb4/0xd0
[  310.703502] LR is at __mutex_lock_slowpath+0x10c/0x16c
[  310.708648] pc : [<c075efc0>]    lr : [<c0bfb0bc>]    psr: 600f0113
<snip..>
[  311.565585] [<c075efc0>] (usb_gadget_unregister_driver) from [<c075e2b8>] (unregister_gadget_item+0x1c/0x34)
[  311.575426] [<c075e2b8>] (unregister_gadget_item) from [<c076fcc8>] (ffs_closed+0x8c/0x9c)
[  311.583702] [<c076fcc8>] (ffs_closed) from [<c07736b8>] (ffs_data_reset+0xc/0xa0)
[  311.591194] [<c07736b8>] (ffs_data_reset) from [<c07738ac>] (ffs_data_closed+0x90/0xd0)
[  311.599208] [<c07738ac>] (ffs_data_closed) from [<c07738f8>] (ffs_ep0_release+0xc/0x14)
[  311.607224] [<c07738f8>] (ffs_ep0_release) from [<c023e030>] (__fput+0x80/0x1d0)
[  311.614635] [<c023e030>] (__fput) from [<c014e688>] (task_work_run+0xb0/0xe8)
[  311.621788] [<c014e688>] (task_work_run) from [<c010afdc>] (do_work_pending+0x7c/0xa4)
[  311.629718] [<c010afdc>] (do_work_pending) from [<c010770c>] (work_pending+0xc/0x20)

for functions using functionFS, i.e. android adbd will close /dev/usb-ffs/adb/ep0
when usb IO thread fails, but switch adb from on to off also triggers write
"none" > UDC. These 2 operations both call unregister_gadget, which will lead
to the panic above.

add a mutex before calling unregister_gadget for api used in f_fs.

Signed-off-by: Winter Wang <wente.wang@nxp.com>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
8 years agoANDROID: dm-verity: adopt changes made to dm callbacks
Badhri Jagan Sridharan [Tue, 9 Aug 2016 19:47:37 +0000 (12:47 -0700)]
ANDROID: dm-verity: adopt changes made to dm callbacks

v4.4 introduced changes to the callbacks used for
dm-linear and dm-verity-target targets. Move to those headers
in dm-android-verity.

Verified on hikey while having
BOARD_USES_RECOVERY_AS_BOOT := true
BOARD_BUILD_SYSTEM_ROOT_IMAGE := true

BUG: 27339727
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: Ic64950c3b55f0a6eaa570bcedc2ace83bbf3005e

8 years agoUPSTREAM: ecryptfs: fix handling of directory opening
Al Viro [Wed, 4 May 2016 18:04:13 +0000 (14:04 -0400)]
UPSTREAM: ecryptfs: fix handling of directory opening

(cherry picked from commit 6a480a7842545ec520a91730209ec0bae41694c1)

First of all, trying to open them r/w is idiocy; it's guaranteed to fail.
Moreover, assigning ->f_pos and assuming that everything will work is
blatantly broken - try that with e.g. tmpfs as underlying layer and watch
the fireworks.  There may be a non-trivial amount of state associated with
current IO position, well beyond the numeric offset.  Using the single
struct file associated with underlying inode is really not a good idea;
we ought to open one for each ecryptfs directory struct file.

Additionally, file_operations both for directories and non-directories are
full of pointless methods; non-directories should *not* have ->iterate(),
directories should not have ->flush(), ->fasync() and ->splice_read().

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Change-Id: I4813ce803f270fdd364758ce1dc108b76eab226e
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoANDROID: net: core: fix UID-based routing
Amit Pundir [Thu, 11 Aug 2016 13:43:22 +0000 (19:13 +0530)]
ANDROID: net: core: fix UID-based routing

Fix RTA_UID enum to match it with the Android userspace code which
assumes RTA_UID=18.

With this patch all Android kernel networking unit tests mentioned here
https://source.android.com/devices/tech/config/kernel_network_tests.html
are success.

Without this patch multinetwork_test.py unit test fails.

Change-Id: I3ff36670f7d4e5bf5f01dce584ae9d53deabb3ed
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoANDROID: net: fib: remove duplicate assignment
Amit Pundir [Fri, 12 Aug 2016 05:54:50 +0000 (11:24 +0530)]
ANDROID: net: fib: remove duplicate assignment

Remove duplicate FRA_GOTO assignment.

Fixes: fd2cf795f3ab ("net: core: Support UID-based routing.")
Change-Id: I462c24b16fdef42ae2332571a0b95de3ef9d2e25
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
8 years agoFROMLIST: proc: Fix timerslack_ns CAP_SYS_NICE check when adjusting self
John Stultz [Mon, 1 Aug 2016 23:49:07 +0000 (16:49 -0700)]
FROMLIST: proc: Fix timerslack_ns CAP_SYS_NICE check when adjusting self

In changing from checking ptrace_may_access(p, PTRACE_MODE_ATTACH_FSCREDS)
to capable(CAP_SYS_NICE), I missed that ptrace_my_access succeeds
when p == current, but the CAP_SYS_NICE doesn't.

Thus while the previous commit was intended to loosen the needed
privledges to modify a processes timerslack, it needlessly restricted
a task modifying its own timerslack via the proc/<tid>/timerslack_ns
(which is permitted also via the PR_SET_TIMERSLACK method).

This patch corrects this by checking if p == current before checking
the CAP_SYS_NICE value.

Cc: Kees Cook <keescook@chromium.org>
Cc: "Serge E. Hallyn" <serge@hallyn.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
CC: Arjan van de Ven <arjan@linux.intel.com>
Cc: Oren Laadan <orenl@cellrox.com>
Cc: Ruchi Kandoi <kandoiruchi@google.com>
Cc: Rom Lemarchand <romlem@android.com>
Cc: Todd Kjos <tkjos@google.com>
Cc: Colin Cross <ccross@android.com>
Cc: Nick Kralevich <nnk@google.com>
Cc: Dmitry Shmidt <dimitrysh@google.com>
Cc: Elliott Hughes <enh@google.com>
Cc: Android Kernel Team <kernel-team@android.com>
Mailing-list-url: http://www.spinics.net/lists/kernel/msg2317488.html
Change-Id: Ia3e8aff07c2d41f55b6617502d33c39b7d781aac
Signed-off-by: John Stultz <john.stultz@linaro.org>
8 years agoANDROID: dm verity fec: pack the fec_header structure
Jeremy Compostella [Tue, 10 May 2016 11:10:20 +0000 (13:10 +0200)]
ANDROID: dm verity fec: pack the fec_header structure

The fec_header structure is generated build time and stored on disk.
The fec_header might be build on a 64 bits machine while it is read
per a 32 bits device or the other way around.  In such situations, the
fec_header fields are not aligned as expected by the device and it
fails to read the fec_header structure.

This patch makes the fec_header packed.

Change-Id: Idb84453e70cc11abd5ef3a0adfbb16f8b5feaf06
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
8 years agoANDROID: dm: android-verity: Verify header before fetching table
Badhri Jagan Sridharan [Thu, 7 Jul 2016 00:16:19 +0000 (17:16 -0700)]
ANDROID: dm: android-verity: Verify header before fetching table

Move header validation logic before reading the verity_table as
an invalid header implies the table is invalid as well.

(Cherry-picked from:
https://partner-android-review.git.corp.google.com/#/c/625203)

BUG: 29940612
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: Ib34d25c0854202f3e70df0a6d0ef1d96f0250c8e

8 years agoANDROID: dm: allow adb disable-verity only in userdebug
Badhri Jagan Sridharan [Mon, 27 Jun 2016 23:25:55 +0000 (16:25 -0700)]
ANDROID: dm: allow adb disable-verity only in userdebug

adb disable-verity was allowed when the phone is in the
unlocked state. Since the driver is now aware of the build
variant, honor "adb disable-verity" only in userdebug
builds.

(Cherry-picked from
https://partner-android-review.git.corp.google.com/#/c/622117)

BUG: 29276559
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I7ce9f38d8c7a62361392c5a8ccebb288f8a3a2ea

8 years agoANDROID: dm: mount as linear target if eng build
Badhri Jagan Sridharan [Sat, 18 Jun 2016 01:54:35 +0000 (18:54 -0700)]
ANDROID: dm: mount as linear target if eng build

eng builds dont have verity enabled i.e it does even
have verity metadata appended to the parition. Therefore
add rootdev as linear device and map the entire partition
if build variant is "eng".

(Cherry-picked based on
https://partner-android-review.git.corp.google.com/#/c/618690/)

BUG: 29276559
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I8f5c2289b842b820ca04f5773525e5449bb3f355

8 years agoANDROID: dm: use default verity public key
Badhri Jagan Sridharan [Fri, 20 May 2016 23:45:45 +0000 (16:45 -0700)]
ANDROID: dm: use default verity public key

If the dm-android-verity target does not provide a default
key try using the default public key from the system keyring.
The defualt verity keyid is passed as a kernel command line
argument veritykeyid=.

The order of the dm-android-verity params have been reversed
to facilitate the change.

Old format example:
dm="system none ro,0 1 android-verity Android:#7e4333f9bba00adfe0ede979e28ed1920492b40f /dev/mmcblk0p43"

New formats supported:
dm="system none ro,0 1 android-verity /dev/mmcblk0p43 Android:#7e4333f9bba00adfe0ede979e28ed1920492b40f"

(or)

dm="system none ro,0 1 android-verity /dev/mmcblk0p43"
when veritykeyid= is set in the kernel command line.

BUG: 28384658
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I506c89b053d835ab579e703eef2bc1f8487250de
(cherry picked from commit c5c74d0327729f35b576564976885596c6d0e7fb)

8 years agoANDROID: dm: fix signature verification flag
Badhri Jagan Sridharan [Fri, 20 May 2016 23:44:19 +0000 (16:44 -0700)]
ANDROID: dm: fix signature verification flag

The bug was that the signature verification was only
happening when verity was disabled. It should always
happen when verity is enabled.

Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I2d9354e240d36ea06fc68c2d18d8e87b823a4c2f
(cherry picked from commit 5364b5ca0b1a12a58283b51408e43fc36d4e4fe7)

8 years agoANDROID: dm: use name_to_dev_t
Jeremy Compostella [Fri, 15 Apr 2016 11:32:54 +0000 (13:32 +0200)]
ANDROID: dm: use name_to_dev_t

This patch makes android_verity_ctr() parse its block device string
parameter with name_to_dev_t().  It allows the use of less hardware
related block device reference like PARTUUID for instance.

Change-Id: Idb84453e70cc11abd5ef3a0adfbb16f8b5feaf07
Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
8 years agoANDROID: dm: rename dm-linear methods for dm-android-verity
Badhri Jagan Sridharan [Tue, 5 Apr 2016 18:18:16 +0000 (11:18 -0700)]
ANDROID: dm: rename dm-linear methods for dm-android-verity

This keeps linear_target as static variable and just exposes
the linear target methods for android-verity

Cherry-picked: https://android-review.googlesource.com/#/c/212858

Change-Id: I4a377e417b00afd9ecccdb3e605fea31a7df112e
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
(cherry picked from commit a6d1b091f40b25d97849487e29ec097bc5f568dd)

8 years agoANDROID: dm: Minor cleanup
Badhri Jagan Sridharan [Mon, 28 Mar 2016 21:41:21 +0000 (14:41 -0700)]
ANDROID: dm: Minor cleanup

Compacts the linear device arguments removing the
unnecessary variables.

Bug: 27175947
Change-Id: I157170eebe3c0f89a68ae05870a1060f188d0da0
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
8 years agoANDROID: dm: Mounting root as linear device when verity disabled
Badhri Jagan Sridharan [Mon, 21 Mar 2016 17:55:23 +0000 (10:55 -0700)]
ANDROID: dm: Mounting root as linear device when verity disabled

This CL makes android-verity target to be added as linear
dm device if when bootloader is unlocked and verity is disabled.

Bug: 27175947
Change-Id: Ic41ca4b8908fb2777263799cf3a3e25934d70f18
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
8 years agoANDROID: dm-android-verity: Rebase on top of 4.1
Badhri Jagan Sridharan [Tue, 9 Feb 2016 00:28:43 +0000 (16:28 -0800)]
ANDROID: dm-android-verity: Rebase on top of 4.1

Following CLs in upstream causes minor changes to dm-android-verity target.
1. keys: change asymmetric keys to use common hash definitions
2. block: Abstract out bvec iterator
Rebase dm-android-verity on top of these changes.

Bug: 27175947

Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: Icfdc3e7b3ead5de335a059cade1aca70414db415

8 years agoANDROID: dm: Add android verity target
Badhri Jagan Sridharan [Tue, 15 Dec 2015 04:09:39 +0000 (20:09 -0800)]
ANDROID: dm: Add android verity target

This device-mapper target is virtually a VERITY target. This
target is setup by reading the metadata contents piggybacked
to the actual data blocks in the block device. The signature
of the metadata contents are verified against the key included
in the system keyring. Upon success, the underlying verity
target is setup.

BUG: 27175947

Change-Id: I7e99644a0960ac8279f02c0158ed20999510ea97
Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
8 years agoANDROID: dm: fix dm_substitute_devices()
Jeremy Compostella [Mon, 2 May 2016 15:29:28 +0000 (17:29 +0200)]
ANDROID: dm: fix dm_substitute_devices()

When candidate is the last parameter, candidate_end points to the '\0'
character and not the DM_FIELD_SEP character.  In such a situation, we
should not move the candidate_end pointer one character backward.

Signed-off-by: Jeremy Compostella <jeremy.compostella@intel.com>
8 years agoANDROID: dm: Rebase on top of 4.1
Badhri Jagan Sridharan [Tue, 9 Feb 2016 00:47:41 +0000 (16:47 -0800)]
ANDROID: dm: Rebase on top of 4.1

1. "dm: optimize use SRCU and RCU" removes the use of dm_table_put.
2. "dm: remove request-based logic from make_request_fn wrapper" necessitates
    calling dm_setup_md_queue or else the request_queue's make_request_fn
    pointer ends being unset.

[    7.711600] Internal error: Oops - bad mode: 0 [#1] PREEMPT SMP
[    7.717519] CPU: 1 PID: 1 Comm: swapper/0 Tainted: G        W       4.1.15-02273-gb057d16-dirty #33
[    7.726559] Hardware name: HiKey Development Board (DT)
[    7.731779] task: ffffffc005f8acc0 ti: ffffffc005f8c000 task.ti: ffffffc005f8c000
[    7.739257] PC is at 0x0
[    7.741787] LR is at generic_make_request+0x8c/0x108
....
[    9.082931] Call trace:
[    9.085372] [<          (null)>]           (null)
[    9.090074] [<ffffffc0003f4ac0>] submit_bio+0x98/0x1e0
[    9.095212] [<ffffffc0001e2618>] _submit_bh+0x120/0x1f0
[    9.096165] cfg80211: Calling CRDA to update world regulatory domain
[    9.106781] [<ffffffc0001e5450>] __bread_gfp+0x94/0x114
[    9.112004] [<ffffffc00024a748>] ext4_fill_super+0x18c/0x2d64
[    9.117750] [<ffffffc0001b275c>] mount_bdev+0x194/0x1c0
[    9.122973] [<ffffffc0002450dc>] ext4_mount+0x14/0x1c
[    9.128021] [<ffffffc0001b29a0>] mount_fs+0x3c/0x194
[    9.132985] [<ffffffc0001d059c>] vfs_kern_mount+0x4c/0x134
[    9.138467] [<ffffffc0001d2168>] do_mount+0x204/0xbbc
[    9.143514] [<ffffffc0001d2ec4>] SyS_mount+0x94/0xe8
[    9.148479] [<ffffffc000c54074>] mount_block_root+0x120/0x24c
[    9.154222] [<ffffffc000c543e8>] mount_root+0x110/0x12c
[    9.159443] [<ffffffc000c54574>] prepare_namespace+0x170/0x1b8
[    9.165273] [<ffffffc000c53d98>] kernel_init_freeable+0x23c/0x260
[    9.171365] [<ffffffc0009b1748>] kernel_init+0x10/0x118
[    9.176589] Code: bad PC value
[    9.179807] ---[ end trace 75e1bc52ba364d13 ]---

Bug: 27175947

Signed-off-by: Badhri Jagan Sridharan <Badhri@google.com>
Change-Id: I952d86fd1475f0825f9be1386e3497b36127abd0

8 years agoCHROMIUM: dm: boot time specification of dm=
Will Drewry [Wed, 9 Jun 2010 22:47:38 +0000 (17:47 -0500)]
CHROMIUM: dm: boot time specification of dm=

This is a wrap-up of three patches pending upstream approval.
I'm bundling them because they are interdependent, and it'll be
easier to drop it on rebase later.

1. dm: allow a dm-fs-style device to be shared via dm-ioctl

Integrates feedback from Alisdair, Mike, and Kiyoshi.

Two main changes occur here:

- One function is added which allows for a programmatically created
mapped device to be inserted into the dm-ioctl hash table.  This binds
the device to a name and, optional, uuid which is needed by udev and
allows for userspace management of the mapped device.

- dm_table_complete() was extended to handle all of the final
functional changes required for the table to be operational once
called.

2. init: boot to device-mapper targets without an initr*

Add a dm= kernel parameter modeled after the md= parameter from
do_mounts_md.  It allows for device-mapper targets to be configured at
boot time for use early in the boot process (as the root device or
otherwise).  It also replaces /dev/XXX calls with major:minor opportunistically.

The format is dm="name uuid ro,table line 1,table line 2,...".  The
parser expects the comma to be safe to use as a newline substitute but,
otherwise, uses the normal separator of space.  Some attempt has been
made to make it forgiving of additional spaces (using skip_spaces()).

A mapped device created during boot will be assigned a minor of 0 and
may be access via /dev/dm-0.

An example dm-linear root with no uuid may look like:

root=/dev/dm-0  dm="lroot none ro, 0 4096 linear /dev/ubdb 0, 4096 4096 linear /dv/ubdc 0"

Once udev is started, /dev/dm-0 will become /dev/mapper/lroot.

Older upstream threads:
http://marc.info/?l=dm-devel&m=127429492521964&w=2
http://marc.info/?l=dm-devel&m=127429499422096&w=2
http://marc.info/?l=dm-devel&m=127429493922000&w=2

Latest upstream threads:
https://patchwork.kernel.org/patch/104859/
https://patchwork.kernel.org/patch/104860/
https://patchwork.kernel.org/patch/104861/

Bug: 27175947

Signed-off-by: Will Drewry <wad@chromium.org>
Review URL: http://codereview.chromium.org/2020011

Change-Id: I92bd53432a11241228d2e5ac89a3b20d19b05a31

8 years agoImplement memory_state_time, used by qcom,cpubw
James Carr [Sat, 30 Jul 2016 02:02:16 +0000 (19:02 -0700)]
Implement memory_state_time, used by qcom,cpubw

New driver memory_state_time tracks time spent in different DDR
frequency and bandwidth states.

Memory drivers such as qcom,cpubw can post updated state to the driver
after registering a callback. Processed by a workqueue

Bandwidth buckets are read in from device tree in the relevant qualcomm
section, can be defined in any quantity and spacing.

The data is exposed at /sys/kernel/memory_state_time, able to be read by
the Android framework.

Functionality is behind a config option CONFIG_MEMORY_STATE_TIME

Change-Id: I4fee165571cb975fb9eacbc9aada5e6d7dd748f0
Signed-off-by: James Carr <carrja@google.com>