GitLab

The Memory Controller provides a control register to check
if the video memory can be resized. The previous bootloader
might have locked this feature, which will be reflected by
this register.

This patch reads the control register before processing
a video memory resize request. An error code, -ENOTSUP,
is returned if the feature is locked.

Change-Id: Ia1d67f7a94aa15c6b18ff5c9b9b952e179596ae3
Signed-off-by: Anthony Zhou <anzhou@nvidia.com>

This patch enables on-chip timer1 interrupts for Tegra210 platforms.

Change-Id: Ic7417dc0e69264d7c28aa012fe2322cd30838f3e
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

Tegra210B01 SoCs support atomic context save for the two SE
hardware engines. Tegra210 SoCs have support for only one SE
engine and support a software based save/restore mechanism
instead.

This patch updates the SE driver to make this change.

Change-Id: Ia5e5ed75d0fe011f17809684bbc2ed2338925946
Signed-off-by: Harvey Hsieh <hhsieh@nvidia.com>

Tegra194 supports upto 64GB of DRAM, whereas the previous SoCs support
upto 32GB DRAM. This patch moves the common DRAM base/end macros to
individual Tegra SoC headers to fix this anomaly.

Change-Id: I1a9f386b67c2311baab289e726d95cef6954071b
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

Platforms that do not support bpmp firmware, do not need access
to the PMC block from outside of the CPU complex. The agents
running on the CPU can always access the PMC through the EL3
exception space.

This patch restricts non-secure world access to the PMC block on
such platforms.

Change-Id: I2c4318dc07ddf6407c1700595e0f4aac377ba258
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch skips past the signature header added to the sc7entry-fw
binary by the previous level bootloader. Currently, the size of
the header is 1KB, so adjust the start address and the binary size
at the time of copy.

Change-Id: Id0494548009749035846d54df417a960c640c8f9
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds SiP handler for Tegra210 platforms to service
read/write requests for PMC block. None of the secure registers
are accessible to the NS world though.

Change-Id: I7dc1f10c6a6ee6efc642ddcfb1170fb36d3accff
Signed-off-by: kalyani chidambaram <kalyanic@nvidia.com>

This patch puts all the DMA masters in reset before starting the System
Suspend sequence. This helps us make sure that there are no rogue agents
in the system trying to over-write the SC7 Entry Firmware with their own.

Change-Id: I7eb39999d229951e612fbfeb9f86c4efb8f98b5a
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds support to enter System Suspend on Tegra210 platforms
without the traditional BPMP firmware. The BPMP firmware will no longer
be supported on Tegra210 platforms and its functionality will be
divided across the CPU and sc7entry-fw.

The sc7entry-fw takes care of performing the hardware sequence required
to enter System Suspend (SC7 power state) from the COP. The CPU is required
to load this firmware to the internal RAM of the COP and start the sequence.
The CPU also make sure that the COP is off after cold boot and is only
powered on when we want to start the actual System Suspend sequence.

The previous bootloader loads the firmware to TZDRAM and passes its base and
size as part of the boot parameters. The EL3 layer is supposed to sanitize
the parameters before touching the firmware blob.

To assist the warmboot code with the PMIC discovery, EL3 is also supposed to
program PMC's scratch register #210, with appropriate values. Without these
settings the warmboot code wont be able to get the device out of System
Suspend.

Change-Id: I5a7b868512dbfd6cfefd55acf3978a1fd7ebf1e2
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch removes support for powering down a CPU cluster on
Tegra210 platforms as none of them actually use it.

Change-Id: I9665634cf2b5b7b8a1b5a2700cae152dc9165fe3
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds support to enter/exit to/from cluster idle power
state on Tegra210 platforms that do not load BPMP firmware.

The CPU initates the cluster idle sequence on the last standing
CPU, by following these steps:

Entry
-----
* stop other CPUs from waking up
* program the PWM pinmux to tristate for OVR PMIC
* program the flow controller to enter CC6 state
* skip L1 $ flush during cluster power down, as L2 $ is inclusive
  of L1 $ on Cortex-A57 CPUs

Exit
----
* program the PWM pinmux to un-tristate for OVR PMIC
* allow other CPUs to wake up

This patch also makes sure that cluster idle state entry is not
enabled until CL-DVFS is ready.

Change-Id: I54cf31bf72b4a09d9bf9d2baaed6ee5a963c7808
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch moves the PLAT_PHY_ADDR_SPACE_SIZE & PLAT_VIRT_ADDR_SPACE
macros to tegra_def.h, to define the virtual/physical address space
size on the platform.

Change-Id: I1c5d264c7ffc1af0e7b14cc16ae2c0416efc76f6
Signed-off-by: Steven Kao <skao@nvidia.com>

This patch enables the watchdog timer's interrupt as an FIQ
interrupt to the CPU. The interrupt generated by the watchdog
is connected to the flow controller for power management reasons,
and needs to be routed to the GICD for it to reach the CPU.

Change-Id: I9437b516da2c5d763eca72694ed7f3c7389b3d9e
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

Whenever the VPR memory is resized, the GPU is put into reset first
and then the new VPR parameters are programmed to the memory controller
block. There exists a scenario, where the GPU might be out before we
program the new VPR parameters. This means, the GPU would still be
using older settings and leak secrets.

This patch puts the GPU back into reset, if it is out of reset after
resizing VPR, to mitigate this hole.

Change-Id: I38a1000e3803f80909efcb02e27da4bd46909931
Signed-off-by: Jeetesh Burman <jburman@nvidia.com>

This patch cleans the Memory controller's interrupt status
register, before exiting to the non-secure world during
cold boot. This is required as we observed that the MC's
arbitration bit is set before exiting the secure world.

Change-Id: Iacd01994d03b3b9cbd7b8a57fe7ab5b04e607a9f
Signed-off-by: Harvey Hsieh <hhsieh@nvidia.com>

If ECID is valid, we can use force instantiation
otherwise, we should use reseed for random data
generation for RNG operations in SE context save
DNI because we are not keeping software save
sequence in main.

Change-Id: I73d650e6f45db17b780834b8de4c10501e05c8f3
Signed-off-by: Samuel Payne <spayne@nvidia.com>

This change ports the software based SE context save routines.
The software implements the context save sequence for SE/SE2 and
PKA1. The context save routine is intended to be invoked from
the ATF SC7 entry.

Change-Id: I9aa156d6e7e22a394bb10cb0c3b05fc303f08807
Signed-off-by: Marvin Hsu <marvinh@nvidia.com>

MISRA Rule 10.3, the value of an expression shall not be assigned to
an object with a narrower essential type or of a different essential
type category.

The essential type of a enum member is anonymous enum, the enum member
should be casted to the right type when using it.

Both UL and ULL suffix equal to uint64_t constant in compiler
aarch64-linux-gnu-gcc, to avoid confusing, only keep U and ULL suffix
in platform code. So in some case, cast a constant to uint32_t is
necessary.

Change-Id: I1aae8cba81ef47481736e7f95f53570de7013187
Signed-off-by: Anthony Zhou <anzhou@nvidia.com>

This patch enables clocks to the SE and Entropy block and gets them
out of reset, before starting the context save operation.

Change-Id: Ic196be8fb833dfd04c0e8d460c07058429999613
Signed-off-by: Samuel Payne <spayne@nvidia.com>

This patch disables SMMU hardware before suspending the SE
block, for the context save operation to complete. The NS
word will re-enable SMMU when we exit System Suspend.

Change-Id: I4d5cd982ea6780db5c38b124550d847e3928c60d
Signed-off-by: Samuel Payne <spayne@nvidia.com>

This patch memmaps all the IRAM memory banks during boot. The BPMP
firmware might place the channels in any of the IRAMs, so it is better
to map all the banks to avoid surprises.

Change-Id: Ia009a65d227ee50fbb23e511ce509daf41b877ee
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds the driver to communicate with the BPMP processor
for power management use cases. BPMP controls the entry into cluster
and system power states. The Tegra210 platform port queries the BPMP
to calculate the target state for the cluster. In case BPMP does not
allow CCx entry, the core enters a power down state.

Change-Id: I9c40aef561607a0b02c49b7f8118570eb9105cc9
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds the implementation of the SE atomic context save
sequence. The atomic context-save consistently saves to the TZRAM
carveout; thus there is no need to declare context save buffer or
map MMU region in TZRAM for context save. The atomic context-save
routine is responsible to validate the context-save progress
counter, where CTX_SAVE_CNT=133(SE1)/646(SE2), and the SE error
status to ensure the context save procedure complete successfully.

Change-Id: Ic80843902af70e76415530266cb158f668976c42
Signed-off-by: Marvin Hsu <marvinh@nvidia.com>
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

Enforce full include path for includes. Deprecate old paths.

The following folders inside include/lib have been left unchanged:

- include/lib/cpus/${ARCH}
- include/lib/el3_runtime/${ARCH}

The reason for this change is that having a global namespace for
includes isn't a good idea. It defeats one of the advantages of having
folders and it introduces problems that are sometimes subtle (because
you may not know the header you are actually including if there are two
of them).

For example, this patch had to be created because two headers were
called the same way: e0ea0928 ("Fix gpio includes of mt8173 platform
to avoid collision."). More recently, this patch has had similar
problems: 46f9b2c3 ("drivers: add tzc380 support").

This problem was introduced in commit 4ecca339

 ("Move include and
source files to logical locations"). At that time, there weren't too
many headers so it wasn't a real issue. However, time has shown that
this creates problems.

Platforms that want to preserve the way they include headers may add the
removed paths to PLAT_INCLUDES, but this is discouraged.

Change-Id: I39dc53ed98f9e297a5966e723d1936d6ccf2fc8f
Signed-off-by: Antonio Nino Diaz <antonio.ninodiaz@arm.com>

All identifiers, regardless of use, that start with two underscores are
reserved. This means they can't be used in header guards.

The style that this project is now to use the full name of the file in
capital letters followed by 'H'. For example, for a file called
"uart_example.h", the header guard is UART_EXAMPLE_H.

The exceptions are files that are imported from other projects:

- CryptoCell driver
- dt-bindings folders
- zlib headers

Change-Id: I50561bf6c88b491ec440d0c8385c74650f3c106e
Signed-off-by: Antonio Nino Diaz <antonio.ninodiaz@arm.com>

This patch moves the GPU reset state check, during VideoMem resize, to the
common SiP handler, to reduce code duplication.

Change-Id: I3818c5f104b809da83dc2a61d6a8149606f81c13
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds explicit casts (U(x)) to integers in the tegra_def.h
headers, to make them compatible with whatever operation they're used
in [MISRA-C Rule 10.1]

Change-Id: Ic5fc611aad986a2c6e6e6f625e0753ab9b69eb02
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

For SoCs T132 and T210, the header file 'platform_def.h' used to include
'tegra_def.h' and vice versa. This patch breaks this circular dependency
by making 'tegra_def.h' independent.

Change-Id: I45a00a84e6ab8b93d5e9242a9ff65f03e9102a96
Signed-off-by: Sandrine Bailleux <sandrine.bailleux@arm.com>

To make software license auditing simpler, use SPDX[0] license
identifiers instead of duplicating the license text in every file.

NOTE: Files that have been imported by FreeBSD have not been modified.

[0]: https://spdx.org/



Change-Id: I80a00e1f641b8cc075ca5a95b10607ed9ed8761a
Signed-off-by: dp-arm <dimitris.papastamos@arm.com>

This patch adds the Tegra micro-seconds controller to the
memory map. This allows us to use the delay_timer functionality.

Change-Id: Ia8b148a871949bfede539974cacbe0e93ec7e77c
Signed-off-by: Steven Kao <skao@nvidia.com>
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch moves the chip specific memory controller driver defines to
the appropriate tegra_def.h files, for future compatibility.

Change-Id: I3179fb771d8b32e913ca29bd94af95f4b2fc1961
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch adds support to identify the underlying platform
on which we are running. The currently supported platforms
are actual silicon and simulation platforms.

Change-Id: Iadf96e79ec663b3dbd1a18e9bb95ffcdb82fc8af
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch modifies the secure IRQ registration process to allow platforms
to specify the target CPUs as well as the owner of the IRQ. IRQs "owned"
by the EL3 would return INTR_TYPE_EL3 whereas those owned by the Trusted
OS would return INTR_TYPE_S_EL1 as a result.

Change-Id: I528f7c8220d0ae0c0f354e78d69e188abb666ef6
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

Tegra chips support multiple FIQ interrupt sources. These interrupts
are enabled in the GICD/GICC interfaces by the tegra_gic driver. A
new FIQ handler would be added in a subsequent change which can be
registered by the platform code.

This patch adds the GIC programming as part of the tegra_gic_setup()
which now takes an array of all the FIQ interrupts to be enabled for
the platform. The Tegra132 and Tegra210 platforms right now do not
register for any FIQ interrupts themselves, but will definitely use
this support in the future.

Change-Id: I0ea164be901cd6681167028fea0567399f18d4b8
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

The platform power states, PLAT_MAX_RET_STATE and PLAT_MAX_OFF_STATE,
can change on Tegra SoCs and so should be defined per-soc.

This patch moves these macro definitions to individual SoC's tegra_def.h
files.

Change-Id: Ib9b2752bc4d79cef6f79bee49882d340f71977a2
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch introduces a function to secure the on-chip TZRAM memory. The
Tegra132 and Tegra210 chips do not have a compelling use case to lock the
TZRAM. The trusted OS owns the TZRAM aperture on these chips and so it
can take care of locking the aperture. This might not be true for future
chips and this patch makes the TZRAM programming flexible.

Change-Id: I3ac9f1de1b792ccd23d4ded274784bbab2ea224a
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

The BL2 fills in the UART controller ID to be used as the normal as
well as the crash console on Tegra platforms. The controller ID to
UART controller base address mapping is handled by each Tegra SoC
the base addresses might change across Tegra chips.

This patch adds the handler to parse the platform params to get the
UART ID for the per-soc handlers.

Change-Id: I4d167b20a59aaf52a31e2a8edf94d8d6f89598fa
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch enables the processor retention and L2/CPUECTLR read/write
access from the NS world only for Cortex-A57 CPUs on the Tegra SoCs.

Change-Id: I9941a67686ea149cb95d80716fa1d03645325445
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

The Memory Select Switch Controller routes any CPU transactions to
the appropriate slave depending on the transaction address. During
system suspend, it loses all config settings and hence the CPU has
to restore them during resume.

This patch restores the controller's settings for enabling WRAP to
INCR burst type conversions on the master ports, for any incoming
requests from the AXI slave ports.

Tested by performing multiple system suspend cycles.
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

This patch implements the get_sys_suspend_power_state() handler required by
the PSCI SYSTEM_SUSPEND API. The intent of this handler is to return the
appropriate State-ID field which can be utilized in `affinst_suspend()` to
suspend to system affinity level.
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>