Change-Id: I7940f174c635d9db490441557f81d21640985a11

This driver reserve the specified CPUs for the hypervisor to
use them for the other guests. The reservation is honored by
isolating the CPUs.

Change-Id: I46bbdf304dcace318bed5d5b470b8812b22130ed
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>

This driver reserve the specified CPUs for the hypervisor to use
it for the other guests. The reservation is done by isolating a CPU
so that this OS try not to use the CPU as much as possible. A CPU
can't be left in the isolation state while it is going offline. Since
the CPU is going offline and it can't be used by this OS, it is
considered as reserved. When such CPUs come online again, isolate
them to honor the reservation.

The following sysfs files are made available to configure this driver.

/sys/module/hyp_core_ctl/parameters/reserve_cpus:

Receive a cpumask string for the reservation. The default value
can be specified via a config option.

/sys/module/hyp_core_ctl/parameters/enable:

A boolean input to enable/disable the reservation. When reservation
is enabled, all the online CPUs in the reserve_cpus cpumask are
isolated. They get un-isolated after writing 0 to this file.

Change-Id: I0a266936c37546325f7bd63526b77cf888def92a
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>

* commit '5da5db28':
  "Revert "Revert "msm: camera: isp: improve flush logic to prevent race condition""
  msm: camera: cci: Fix burst read sequence
  msm: camera: crm: Force unreserve link
  msm: camera: isp: manage wait list properly
  msm: camera: icp: Change log type
  msm: camera: Support Linux_32 bit platform
  msm: camera: sync: Validate if fence is already signaled
  msm: camera: cci: Modify condition check to read CCI data
  msm: camera: isp: Report bubble on request in wait list
  msm: camera: sync: Add callback mechanism for signaled objects
  msm: camera: icp: Relaxing the aggression for Clock and BW voting

Change-Id: I2d5c5043f1b032a1a0c9406dc215c7054b8f48eb
Signed-off-by: Sridhar Gujje <sgujje@codeaurora.org>

Ignore events with no particular cpu associated, while
moving the events to zombie state, during event
release.

Change-Id: I11624f3734633ee6c55454be0cd32f638ed61024
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>

HW breakpoint was disabled in
'commit f2a3f3ef ("arch: arm64: disable HW breakpoint")'.
However, it is required for setting breakpoints using ptrace
syscall, which is required for certain userspace applications
to work. So, reenable it.

Change-Id: Ib2a49488b92b0f77d13871d151778b3f0feab273
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>

sched_task_util is tracking latency with WALT specific
variable, so track latency only when WALT is enabled, otherwise
do not track.

Change-Id: I54f347dfedae91303eecde5be0b4e9c9ad46af7b
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>

task_fits_max is called outside of SCHED_WALT and results in
compilation failure if SCHED_WALT is disabled.

Change-Id: I12a0b93a714d041716fc651b461e6388a676f4a2
Signed-off-by: Satya Durga Srinivasu Prabhala <satyap@codeaurora.org>

task_in_cum_window_demand() is used in WALT only, add stub function
for it when WALT disabled.

Change-Id: I6bf8467e86c57ddebb11c1c687b03e5702647c30
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>

Refactor rtg_target selection to avoid compilation issue
with WALT disabled.

Change-Id: Ieeba6c22844da76af0563de746e02cc7e7b95587
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>

A CPU is marked as reserved when a task is under active migration to
that CPU. The reserved CPUs are excluded from CPU selection in
task placement. This reservation scheme is currently tied to WALT.
So add a stub function for is_reserved() to avoid compilation issue
when WALT is disabled.

Change-Id: I77c6a8fd3f24f1763fd97fce18574613e9e947ff
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>

Currently the sched_boost is compiled only for WALT. But sched_boost
defines are accessed outside WALT in RT code. Move these definitions
outside WALT to fix compilation issues.

Change-Id: I38d006c9edb8c859aa10d0638aa271eaaa4db40d
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>

The schedutil is calling twice into the scheduler while querying
the load of each CPU. The cpu_util_freq() is called first to
get the CPU un-boosted load and WALT stats. Then boosted_cpu_util()
is called to get the boosted CPU load. This results in doing the
the same calculations twice in the scheduler and printing
sched_load_to_gov trace point twice. Fix this inefficiency by
changing boosted_cpu_util() to update WALT stats along with
boosted CPU load.

Change-Id: Ia825cafca6a25c56b0edb1ae8c55e7c7277f2968
Signed-off-by: Pavankumar Kondeti <pkondeti@codeaurora.org>
[clingutla@codeaurora.org: Resolved trivial merge conflicts.]
Signed-off-by: Lingutla Chandrasekhar <clingutla@codeaurora.org>

Add support for FBE to work with F2FS filesystem on eMMC
based devices. For F2FS+FBE on eMMC, the cryto data unit
size (CDU size) should be 4KB as F2FS encrypts/decrypts
the data at min. 4KB blocks with (inode|pgidx) as it's
corresponding data unit number (DUN).

Change-Id: I797d896e4a7fbd26fd5f02a1682b9c1d77f68fb6
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>

This fixes wrong bio merge into request which breaks encrypted ICE blocks.

Bug: 71554412
Change-Id: I65b5936add01aa62449d46791a736b734022a48c
Signed-off-by: Jaegeuk Kim <jaegeuk@google.com>
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>

F2FS uses dun for data encryption/decryption via ICE. Hence,
allow the code to use it if a req has a dun. Without this there
will be data corruption issues when FBE is enabled on F2FS.

Change-Id: I02817c2a8a8f75f70e8a9b0988010a6a13ec1860
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>

There seems to be an integration mistake due to which
the dun is not getting set properly in the read path.
This results into data corruption issues when FBE/ICE
is enabled on F2FS.

Change-Id: I938e460ce1f8c297b6af918ada91ac2cf02fa342
Signed-off-by: Sahitya Tummala <stummala@codeaurora.org>

When WLAN driver encounters FW timeout scenarios, it may check if the
FW is down by calling icnss_is_fw_down() API. Recovery should also be
considered as FW down, as FW has not completely recovered and ready
for WLAN communication.

Change-Id: I4c9e15aacc5605dca823e2cfcbf6f87152aaa78e
CRs-Fixed: 2218795
Signed-off-by: Sameer Thalappil <sameert@codeaurora.org>

Enable CPUIDLE and ARM cpuidle configs on SDM670
VM defconfig.

Change-Id: I2af047a60ca4ce9ce6dd728acb9feb6563f219d2
Signed-off-by: Srinivas Rao L <lsrao@codeaurora.org>

Enable vservices configuration on sdm670.

Change-Id: Iebbde77b5162239a7808f9c88471bbdb7d7283af
Signed-off-by: Neeraj Upadhyay <neeraju@codeaurora.org>

This change clears the threshold irq after draining the FIFO
buffer as oppose to immediately clearing it. Also added separate
complete events for threshold and read done.

Change-Id: I29f21545bb52a3c08ec1afffdd7d3f610a3055dc
Signed-off-by: Karthik Anantha Ram <kartanan@codeaurora.org>

This change forces unreserve link to be executed
even in the case of failures in unlinking devices
within the session. Improves logging in unlink
sequence.

Change-Id: Ie0736aaaec21d46a2cf2f3b78d09e7db01e93dd9
Signed-off-by: Venkat Chinta <vchinta@codeaurora.org>

Manage the request transition in lists properly in multiple
location.

Change-Id: Ifcf115b77bd663a22b95a4a5468a545df875e3d3
Signed-off-by: Junzhe Zou <jnzhezou@codeaurora.org>