For RT, changing "tegra210_cpu_pm_lock" from RWLOCK to RAW_SPINLOCK.
RWLOCK in PREEMPT_RT_FULL becomes sleeping lock which triggers BUG().
Changing it to RAW_SPINLOCK to use atomic lock.

 BUG: scheduling while atomic: swapper/3/0/0x00000002
 Modules linked in:
 CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.9.140-rt93-tegra #47
 Hardware name: Jetson-TX1 (DT)
 Call trace:
 [<ffffff800808b6d8>] dump_backtrace+0x0/0x198
 [<ffffff800808bc9c>] show_stack+0x24/0x30
 [<ffffff80084512d4>] dump_stack+0x98/0xbc
 [<ffffff80080e75d4>] __schedule_bug+0x64/0x80
 [<ffffff8008f63388>] __schedule+0x540/0x598
 [<ffffff8008f63428>] schedule+0x48/0xe0
 [<ffffff8008f64c34>] rt_spin_lock_slowlock+0x22c/0x308
 [<ffffff8008f66e38>] __rt_spin_lock+0x50/0x60
 [<ffffff8008f67024>] rt_read_lock+0x34/0x50
 [<ffffff80086bfc08>] tegra210_cpu_pm_exit+0x28/0x70
 [<ffffff8008b9c174>] arm_enter_idle_state+0xb4/0xc0
 [<ffffff8008b99acc>] cpuidle_enter_state+0x84/0x3a0
 [<ffffff8008b99e5c>] cpuidle_enter+0x34/0x48
 [<ffffff8008111120>] call_cpuidle+0x50/0x80
 [<ffffff8008111464>] cpu_startup_entry+0x174/0x208
 [<ffffff8008092510>] secondary_start_kernel+0x190/0x1f8
 [<0000000080f681a4>] 0x80f681a4

Bug 200511477

Change-Id: I0a10b6a23f0b7ad4323dc47ccb69e2384bda058f
Signed-off-by: Sumit Gupta <sumitg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2218264
GVS: Gerrit_Virtual_Submit
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

The enabling/disabling a idle state should apply to cluster instead of
on current running CPU only.

Bug 200523271

Change-Id: I1ba08c425c92ea6e7bbef1b753aa2c71d85c41ba
Signed-off-by: Vincent Zhang <xunz@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2128182
Reviewed-by: Joseph Lo <josephl@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Daniel Fu <danifu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

With trace_printk in the production code, following warning is printed
during boot:

    **********************************************************
    **   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **
    **                                                      **
    ** trace_printk() being used. Allocating extra memory.  **
    **                                                      **
    ** This means that this is a DEBUG kernel and it is     **
    ** unsafe for production use.                           **
    **                                                      **
    ** If you see this message and you are not debugging    **
    ** the kernel, report this immediately to your vendor!  **
    **                                                      **
    **   NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE NOTICE   **
    **********************************************************

commit 2184db46e425 tracing: Print nasty banner when trace_printk() is in use
has the rationale:

    Since trace_printk() should only be used for debugging purposes,
    there's no reason to waste memory on these buffers on a
    production system. That means, trace_printk() should never be used
    unless a developer is debugging their kernel. There's macro magic
    to allocate the buffers if trace_printk() is used anywhere in
    the kernel.

replace trace_printk with pr_debug. If these debug prints are desired,
a specialized trace event should be defined and used.

Bug 200432015
Bug 2462027

Change-Id: Ic15a36e3764368d45468a076bf03257f9e1d4f7d
Signed-off-by: Bo Yan <byan@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1922977
(cherry picked from commit e47b434e20a28c4755aad9ac8605fe22a18e4ec6)
Reviewed-on: https://git-master.nvidia.com/r/2028962
Reviewed-by: Vincent Zhang (SW) <xunz@nvidia.com>
Tested-by: Vincent Zhang (SW) <xunz@nvidia.com>
Reviewed-by: Daniel Fu <danifu@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

"nvidia-bug-report-tegra.sh" tries to read all files with
read permission from path "/sys/kernel/debug/*cpuidle*/".
Permission for "forced_idle_duration_us" node is wrongly
set to read also but only write to this node is supported.
So, reading it causes permission error. Changing permission
to only write for root user to fix the error.

Bug 200494482

Change-Id: Ibb41dad66e345c527addb796519cb73f792dd5ae
Signed-off-by: sumitg <sumitg@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/2022773
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Dipen Patel <dipenp@nvidia.com>
Reviewed-by: Bibek Basu <bbasu@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

Jira BPMP-2289

Change-Id: Ife20ea74421a33525ed6c2fed73001c890450e8f
Signed-off-by: Alex Frid <afrid@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1790758
Reviewed-by: Srikar Srimath Tirumala <srikars@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Sivaram Nair <sivaramn@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

Update the Tegra Processor Power Model for Linux v4.4. This is based
upon the following work by Wei Ni <wni@nvidia.com>:
Reviewed-on: https://chromium-review.googlesource.com/269914
Reviewed-on: https://chromium-review.googlesource.com/283720

Bug 1811732

Change-Id: I9a847cccee322853096093b1ad5110bc4650f3f2
Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-on: http://git-master/r/1255679
GVS: Gerrit_Virtual_Submit
Reviewed-by: Jon Mayo <jmayo@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1786276
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

This patch fixes below error as tv64 is removed since Kernel 4.10:
    In file included from drivers/soc/tegra/pm-tegra210.c:17:0:
    drivers/soc/tegra/pm-tegra210.c: In function ‘idle_write’:
    drivers/soc/tegra/pm-tegra210.c:404:56: error:
    	request for member ‘tv64’ in something not a structure or union

Bug 200419967

Change-Id: I20060d8126aee91926224677b282decfabdb38bc
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1749214
GVS: Gerrit_Virtual_Submit
Reviewed-by: Joseph Lo <josephl@nvidia.com>
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

There are some code paths remaining for the SC7-at-idle
support. However, that feature was never completed and there are no
plans to revive it. So the code can be removed.

Bug 200369181

Change-Id: I60cf718b0af5d8efedb4bb88158f8a7323d868ef
Signed-off-by: Timo Alho <talho@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1660816
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1749106
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Joseph Lo <josephl@nvidia.com>
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

The driver is supposed to call tegra_bpmp_send_receive() functions
to send mails to bpmp firmware without caring about return values.
However, the two function calls in the driver put the input mails
as output parameters while leaving the input parameters to NULL.

(Note: input means Kernel -> bpmp; output means bpmp -> kernel)

This creates a problem that bpmp doesn't receive the correct mail.
So this change just simply fixes the bug.

Bug 200344135

Change-Id: I40d3df0db0c629f5692c4b7002a817a0c9af8a36
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1586646
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1749105
GVS: Gerrit_Virtual_Submit
Reviewed-by: Joseph Lo <josephl@nvidia.com>
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

When CONFIG_ARM_CPUIDLE is not set, the cpuidle_devices becomes
NULL. So this patch just let the drvier handle this corner case.

Bug 200344135

Change-Id: I34d220f52fe12fac1db234f23f805049ae17e492
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1566743
Signed-off-by: Nicolin Chen <nicolinc@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1749104
GVS: Gerrit_Virtual_Submit
Reviewed-by: Joseph Lo <josephl@nvidia.com>
Reviewed-by: Sachin Nikam <snikam@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>