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>
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>
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>
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>
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>
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>
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>
fiq_debugger is taking over uart, so it is necessary to disable
original uart in DT file. It can be done manually or by overlay.
Change-Id: I9f50ec15b0e22e602d73b9f745fc8666f8925d09
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
This reverts commit 5b42ae3eda.
This recereated arch/arm/include/asm/mach/mmc.h include file has
no active user in android-4.x kernels. Also all the necessary bits
are already moved to include/linux/amba/mmci.h.
6ef297f86b (ARM: 5720/1: Move MMCI header to amba include dir)
Change-Id: Ibf258b355d17f54f49b777a8f6e0089e9b59a3a5
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
This reverts commit 541632275e.
mmc_platform_data (or arch/arm/include/asm/mach/mmc.h in general)
has no active user in android-4.x kernels. Also all the necessary
bits are already moved to include/linux/amba/mmci.h.
6ef297f86b (ARM: 5720/1: Move MMCI header to amba include dir)
Change-Id: Iff384eb527327bf88543408e0257241c1fd99a43
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
If CONFIG_DUAL_ROLE_USB_INTF is disabled but the exported functions
are referenced, the build will result in warnings such as:
In file included from include/linux/usb/class-dual-role.h:112:13:
warning: ‘dual_role_instance_changed’ defined but not used
[-Wunused-function]
These stub functions should be static inline.
Change-Id: I5a9ef58dca32306fac5a4c7f28cdaa36fa8ae078
Signed-off-by: Jack Pham <jackp@codeaurora.org>
(cherry picked from commit 2d152dbb0743526b21d6bbefe097f874c027f860)
(cherry picked from commit 8ad66cafaa10e6ba94ff79a8dbc2cc437c6bfe93)
This reverts commit 91fa97e1e5.
This patch is no longer valid. There are no users for this status irq and
callback in android-4.x. The Qcom platform (mach-msm/qsd8x50, HTC Dream..)
and SDCC controller (msm_sdcc) using this status IRQ and callback are
dropped from mainline sometime back.
27842bb18b00 (mmc: Remove msm_sdcc driver)
c0c89fafa2 (ARM: Remove mach-msm and associated ARM architecture code)
Change-Id: Ia38e42a06dc184395f79c1ec1d306bf9775704d5
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
Here is just the quick fix for tracefs with selinux.
just add tracefs to the list of whitelisted filesystem
types in selinux_is_sblabel_mnt(), but the right fix would be to
generalize this logic as described in the last item on the todo list,
https://bitbucket.org/seandroid/wiki/wiki/ToDo
Change-Id: I2aa803ccffbcd2802a7287514da7648e6b364157
Signed-off-by: Yongqin Liu <yongqin.liu@linaro.org>
This reverts commit 0840b80cb9.
This patch is already upstreamed in v4.4, commit
658d4aed59 (HID: hid-multitouch: Filter collections by application usage.),
and further fixed/cleaned up afterwards in commits
c2ef8f21ea (HID: multitouch: add support for trackpads),
76f5902aeb (HID: hid-multitouch: Simplify setup and frame synchronization) et al.
By having this duplicate patch in AOSP we are doing redundant
checks for Touchscreen and Touchpad devices.
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
In upstream commit ca6fb06518
(tcp: attach SYNACK messages to request sockets instead of
listener)
http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=ca6fb0651883
The building of synack messages was changed, which made it so
the skb->sk points to a casted request_sock. This is problematic,
as there is no sk_socket in a request_sock. So when the qtaguid_mt
function tries to access the sk->sk_socket, it accesses uninitialized
memory.
After looking at how other netfilter implementations handle this,
I realized there was a skb_to_full_sk() helper added, which the
xt_qtaguid code isn't yet using.
This patch adds its use, and resovles panics seen when accessing
uninitialzed memory when processing synack packets.
Reported-by: YongQin Liu <yongquin.liu@linaro.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Kernel panic when type "cat /sys/kernel/debug/shrinker"
Unable to handle kernel paging request at virtual address 0af37d40
pgd = d4dec000
[0af37d40] *pgd=00000000
Internal error: Oops: 5 [#1] PREEMPT SMP ARM
[<c0bb8f24>] (_raw_spin_lock) from [<c020aa08>] (list_lru_count_one+0x14/0x28)
[<c020aa08>] (list_lru_count_one) from [<c02309a8>] (super_cache_count+0x40/0xa0)
[<c02309a8>] (super_cache_count) from [<c01f6ab0>] (debug_shrinker_show+0x50/0x90)
[<c01f6ab0>] (debug_shrinker_show) from [<c024fa5c>] (seq_read+0x1ec/0x48c)
[<c024fa5c>] (seq_read) from [<c022e8f8>] (__vfs_read+0x20/0xd0)
[<c022e8f8>] (__vfs_read) from [<c022f0d0>] (vfs_read+0x7c/0x104)
[<c022f0d0>] (vfs_read) from [<c022f974>] (SyS_read+0x44/0x9c)
[<c022f974>] (SyS_read) from [<c0107580>] (ret_fast_syscall+0x0/0x3c)
Code: e1a04000 e3a00001 ebd66b39 f594f000 (e1943f9f)
---[ end trace 60c74014a63a9688 ]---
Kernel panic - not syncing: Fatal exception
shrink_control.nid is used but not initialzed, same for
shrink_control.memcg.
This reverts commit b0e7a582b2.
Change-Id: I108de88fa4baaef99a53c4e4c6a1d8c4b4804157
Reported-by: Xiaowen Liu <xiaowen.liu@freescale.com>
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
This reverts commit 78d36d2111.
Drop this duplicate patch. This patch is already upstreamed in v4.4. Commits
5c73fceb8c (SELinux: Enable setting security contexts on rootfs inodes.),
12f348b9dc (SELinux: rename SE_SBLABELSUPP to SBLABEL_MNT), and
b43e725d8d (SELinux: use a helper function to determine seclabel),
for reference.
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
This reverts commit 43e1b4f528.
This patch is part of code which is already upstreamed in v4.4. Commits
5c73fceb8c (SELinux: Enable setting security contexts on rootfs inodes.),
12f348b9dc (SELinux: rename SE_SBLABELSUPP to SBLABEL_MNT), and
b43e725d8d (SELinux: use a helper function to determine seclabel).
for reference.
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
Allows FUSE to report to inotify that it is acting
as a layered filesystem. The userspace component
returns a string representing the location of the
underlying file. If the string cannot be resolved
into a path, the top level path is returned instead.
bug: 23904372
Change-Id: Iabdca0bbedfbff59e9c820c58636a68ef9683d9f
Signed-off-by: Daniel Rosenberg <drosen@google.com>
Update seq_printf() usage in xt_qtaguid to align
with changes from mainline commit 6798a8caaf
"fs/seq_file: convert int seq_vprint/seq_printf/etc...
returns to void".
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
This reverts commit 5c7566a29b.
This patch revert some changes in net/netfilter/xt_qtaguid.c as well.
I'll submit another patch to restore those changes.
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
This reverts commit 8d3a6c1538.
This series of patches revert AOSP UID_STAT and NET_ACTIVITY_STATS drivers.
I could not find any meaningful usage of these interfaces in AOSP master.
UID_STAT driver expose "/proc/uid_stat/*" interfaces but it is only
used in AOSP master as in what appears be an out of date bandwidth
test in frameworks/base and in somewhat recent battery utils test
in external/chromium-trace project.
NET_ACTIVITY_STATS driver expose "/proc/net/stat/activity" interface
but I can not track its usage anywhere in AOSP at all.
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
Drivers should use extcon moving forward.
Documentation/extcon/porting-android-switch-class describes
how to port existing switch class drivers to extcon.
This reverts commit e4b8e66e0a.
Change-Id: I5b622c7ab4c0cb9670f8903f259a99888f503c1a
IP_NF_TARGET_{MASQUERADE,NETMAP,REDIRECT} configs,
already enabled in android-base.cfg for tethering,
are of no use if CONFIG_IP_NF_NAT is not enabled.
Don't rely on platform config for that and enable
CONFIG_IP_NF_NAT in android-base.cfg as well.
Change-Id: Ic72bcebbd925b142b09539466bf963188c83108a
Signed-off-by: Amit Pundir <amit.pundir@linaro.org>
Backport notes:
Backport uses kernel_module_from_file not kernel_read_file hook.
kernel_read_file replaced kernel_module_from_file in the 4.6 kernel.
There are no inode_security_() helper functions (also introduced in
4.6) so the inode lookup is done using the file_inode() helper which
is standard for kernel version < 4.6.
(Cherry picked from commit 61d612ea731e57dc510472fb746b55cdc017f371)
Utilize existing kernel_read_file hook on kernel module load.
Add module_load permission to the system class.
Enforces restrictions on kernel module origin when calling the
finit_module syscall. The hook checks that source type has
permission module_load for the target type.
Example for finit_module:
allow foo bar_file:system module_load;
Similarly restrictions are enforced on kernel module loading when
calling the init_module syscall. The hook checks that source
type has permission module_load with itself as the target object
because the kernel module is sourced from the calling process.
Example for init_module:
allow foo foo:system module_load;
Bug: 27824855
Change-Id: I64bf3bd1ab2dc735321160642dc6bbfa996f8068
Signed-off-by: Jeff Vander Stoep <jeffv@google.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
When CONFIG_MMC_SIMULATE_MAX_SPEED is enabled, Expose max_read_speed,
max_write_speed and cache_size default module parameters and sysfs
controls to simulate a slow eMMC device. Default values are 0 (off),
0 (off) and 4 MB respectively.
Signed-off-by: Mark Salyzyn <salyzyn@google.com>
Bug: 26976972
Change-Id: I342bfbd8b85f9b790e3f0e1e4e51a900ae07e05d
Only apply the interactive boost when the interactive governor is
enabled. This seems like the right thing to do.
This was originally reviewed on
https://chromium-review.googlesource.com/273501
Change-Id: I5f4a7320683eada099f9a4253e3d6b0f03057fe8
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Dmitry Torokhov <dtor@google.com>
cpufreq_interactive_speedchange_task() is running as a separate kernel
thread and is calling __cpufreq_driver_target(), which requires callers
to hold policy->rwsem for writing to prevent racing with other parts of
the kernel trying to adjust the frequency, for example kernel thermal
throttling. Let's change the code to take policy->rwsem and while at it
refactor the code a bit.
This was originally 2 changes reviewed at:
https://chromium-review.googlesource.com/246273https://chromium-review.googlesource.com/256120
Change-Id: Icc2d97c6c1b929acd2ee32e8c81d81fd2af778ab
Signed-off-by: Dmitry Torokhov <dtor@chromium.org>
Reviewed-by: Dylan Reid <dgreid@chromium.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Signed-off-by: Dmitry Torokhov <dtor@google.com>
Add a sysfs entry that allows user space to determine whether dm-verity
has come across correctable errors on the underlying block device.
Bug: 22655252
Bug: 27928374
Change-Id: I80547a2aa944af2fb9ffde002650482877ade31b
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
(cherry picked from commit 7911fad5f0a2cf5afc2215657219a21e6630e001)
If ignore_zero_blocks is enabled dm-verity will return zeroes for blocks
matching a zero hash without validating the content.
Change-Id: I728fa4b2586b29f2793ea5cb014289892819d249
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
(cherry picked from commit 0cc37c2df4fa0aa702f9662edce4b7ce12c86b7a)
Add support for correcting corrupted blocks using Reed-Solomon.
This code uses RS(255, N) interleaved across data and hash
blocks. Each error-correcting block covers N bytes evenly
distributed across the combined total data, so that each byte is a
maximum distance away from the others. This makes it possible to
recover from several consecutive corrupted blocks with relatively
small space overhead.
In addition, using verity hashes to locate erasures nearly doubles
the effectiveness of error correction. Being able to detect
corrupted blocks also improves performance, because only corrupted
blocks need to corrected.
For a 2 GiB partition, RS(255, 253) (two parity bytes for each
253-byte block) can correct up to 16 MiB of consecutive corrupted
blocks if erasures can be located, and 8 MiB if they cannot, with
16 MiB space overhead.
Change-Id: Ife4f8889f7fbf0974bf3ed4be6d3322ae9b4cb0e
Signed-off-by: Sami Tolvanen <samitolvanen@google.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
(cherry picked from commit a739ff3f543afbb4a041c16cd0182c8e8d366e70)
