drivers: scmi-msg: remove unused SCMI channel agent name

Remove field agent_name from struct scmi_msg_channel since unused.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by

drivers: scmi-msg: remove unused SCMI channel agent name

Remove field agent_name from struct scmi_msg_channel since unused.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

drivers: scmi-msg: fix header file inline comments

Fix inline comments for scmi-msg API functions description.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Fori

drivers: scmi-msg: fix header file inline comments

Fix inline comments for scmi-msg API functions description.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

drivers: scmi-msg: rename agent_id to channel_id

Rename agent_id reference to channel_id to avoid confusion with the
agent identifiers used in SCMI protocol to identify agent, whereas
the drivers on

drivers: scmi-msg: rename agent_id to channel_id

Rename agent_id reference to channel_id to avoid confusion with the
agent identifiers used in SCMI protocol to identify agent, whereas
the drivers only reference an SCMI channel, whatever the agent ID
associated with the channel and knowing that an SCMI agent can have
several channels to communicate with the SCMI platform/server.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

core: arm: enforce LTC multi-threading protection

Remove CFG_LTC_OPTEE_THREAD switch and enable or disable
_CFG_CORE_LTC_OPTEE_THREAD based on multi-thread support
since multi-threading mandates thr

core: arm: enforce LTC multi-threading protection

Remove CFG_LTC_OPTEE_THREAD switch and enable or disable
_CFG_CORE_LTC_OPTEE_THREAD based on multi-thread support
since multi-threading mandates thread protection means.

Suggested-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

plat-stm32mp1: simplify device memory mapping

Register device memory by cluster range rather than by device interface
as the later is likely to grow as new devices are added whereas the
overall stat

plat-stm32mp1: simplify device memory mapping

Register device memory by cluster range rather than by device interface
as the later is likely to grow as new devices are added whereas the
overall static mapped may not change.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

plat-stm32mp1: update from deprecated register_dynamic_shm()

Use macro register_ddr() rather than register_dynamic_shm() that is
deprecated.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro

plat-stm32mp1: update from deprecated register_dynamic_shm()

Use macro register_ddr() rather than register_dynamic_shm() that is
deprecated.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

drivers: imx_rngb: random number generator

Add support for the RNG(B) as described in the i.MX 6ULL Applications
Processor Reference Manual, Rev 1, 11/2017.

Tested on an imx6ull based board.

Signe

drivers: imx_rngb: random number generator

Add support for the RNG(B) as described in the i.MX 6ULL Applications
Processor Reference Manual, Rev 1, 11/2017.

Tested on an imx6ull based board.

Signed-off-by: Jorge Ramirez-Ortiz <jorge@foundries.io>
Reviewed-by: Clement Faure <clement.faure@nxp.com>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: drivers: stm32_rng.c: include thread.h

The implementation makes use of thread_mask_exceptions() and
thread_unmask_exceptions() functions, therefore, include thread.h to avoid
compilation error

core: drivers: stm32_rng.c: include thread.h

The implementation makes use of thread_mask_exceptions() and
thread_unmask_exceptions() functions, therefore, include thread.h to avoid
compilation errors.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

core: driver: stpmic1: do not use TEE_Result as return type

stpmic1_regulator_levels_mv() uses TEE_Result as return type.
The caller on core/arch/arm/plat-stm32mp1/scmi_server.c does
not check the r

core: driver: stpmic1: do not use TEE_Result as return type

stpmic1_regulator_levels_mv() uses TEE_Result as return type.
The caller on core/arch/arm/plat-stm32mp1/scmi_server.c does
not check the return value, therefore, change it to void.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

core: mm: move core_memprot.h to core/include/mm

This commit moves core_memprot.h to core/include/mm since it is
architecture-independent.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.co

core: mm: move core_memprot.h to core/include/mm

This commit moves core_memprot.h to core/include/mm since it is
architecture-independent.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

core: kernel: pm.c: add missing header stdlib.h

This commit explicitly includes stdlib.h to use realloc() function.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Jerome

core: kernel: pm.c: add missing header stdlib.h

This commit explicitly includes stdlib.h to use realloc() function.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

drivers: gic: fix the off-by-one error

The gd->max_it should refer to the largest support interrupt id.
Fix the off-by-one errors so that the interrupt with the largest
id can be correctly handled.

drivers: gic: fix the off-by-one error

The gd->max_it should refer to the largest support interrupt id.
Fix the off-by-one errors so that the interrupt with the largest
id can be correctly handled.

Acked-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Fangsuo Wu <fangsuowu@asrmicro.com>

show more ...

plat-ls: Add DSPI driver for NXP LS Platforms

This patch adds DSPI driver for Layerscape Platforms.
DSPI compilation is enabled by default for LX2160A-QDS and LX2160A-RDB.

Signed-off-by: Carl Lamb

plat-ls: Add DSPI driver for NXP LS Platforms

This patch adds DSPI driver for Layerscape Platforms.
DSPI compilation is enabled by default for LX2160A-QDS and LX2160A-RDB.

Signed-off-by: Carl Lamb <calamb@microsoft.com>
Signed-off-by: Manish Tomar <manish.tomar@nxp.com>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

core: kernel: move ldelf_loader.h to core/include/kernel

To make reuse of ldelf_loader.h when porting OP-TEE OS to other
architectures, this commit moves it to core/include/kernel.

Signed-off-by: M

core: kernel: move ldelf_loader.h to core/include/kernel

To make reuse of ldelf_loader.h when porting OP-TEE OS to other
architectures, this commit moves it to core/include/kernel.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move linker.h to core/include/kernel

To make reuse of linker.h when porting OP-TEE OS to other architectures,
this commit moves it to core/include/kernel.

Signed-off-by: Marouene Boub

core: kernel: move linker.h to core/include/kernel

To make reuse of linker.h when porting OP-TEE OS to other architectures,
this commit moves it to core/include/kernel.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move user_ta.h to core/include/kernel

This commit moves user_ta.h to core/include/kernel since it is included by
sources which are not located in arch tree and it does not contain any

core: kernel: move user_ta.h to core/include/kernel

This commit moves user_ta.h to core/include/kernel since it is included by
sources which are not located in arch tree and it does not contain any
architecture-dependent definition.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move unwind.h to core/include/kernel

unwind.h is used by sources which are not located in arch tree.
In addition, it does not contain any architecture-dependent definition.
This commit

core: kernel: move unwind.h to core/include/kernel

unwind.h is used by sources which are not located in arch tree.
In addition, it does not contain any architecture-dependent definition.
This commit move unwind.h to core/include/kernel.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move timer.h to core/include/kernel

timer.h does not contain any architecture-dependent definition, therefore,
move it to core/include/kernel.

Signed-off-by: Marouene Boubakri <maroue

core: kernel: move timer.h to core/include/kernel

timer.h does not contain any architecture-dependent definition, therefore,
move it to core/include/kernel.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move time_source.h to core/include/kernel

The time_source.h does not contain any architecture-dependent definition,
hence, move it to core/include/kernel instead.

Signed-off-by: Marou

core: kernel: move time_source.h to core/include/kernel

The time_source.h does not contain any architecture-dependent definition,
hence, move it to core/include/kernel instead.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: time_source.h: remove un-implemented function

arm_prng_add_jitter_entropy() in time_source.h is defined but never
implemented or referenced. This commit removes the prototype.

Signed-

core: kernel: time_source.h: remove un-implemented function

arm_prng_add_jitter_entropy() in time_source.h is defined but never
implemented or referenced. This commit removes the prototype.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: move pseudo_ta.h to core/include/kernel

The pseudo_ta.h is used by core/kernel/tee_ta_manager.c and it does not
contain any architecture-dependent definition, hence, move it to
core/in

core: kernel: move pseudo_ta.h to core/include/kernel

The pseudo_ta.h is used by core/kernel/tee_ta_manager.c and it does not
contain any architecture-dependent definition, hence, move it to
core/include/kernel instead.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: kernel: make linker.h architecture agnostic

Other architectures may not have vector table. To make reuse of the
linker.h, this commit adds the start/end addresses of the exception index
table

core: kernel: make linker.h architecture agnostic

Other architectures may not have vector table. To make reuse of the
linker.h, this commit adds the start/end addresses of the exception index
table only if we are building for ARM.

Signed-off-by: Marouene Boubakri <marouene.boubakri@nxp.com>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: Add ffa_msg_send_req and ffa_msg_send_resp

Add FF-A message handling of FFA_MSG_SEND_DIRECT_REQ_32 and
FFA_MSG_SEND_DIRECT_RESP_32 messages.
The FFA_MSG_SEND_DIRECT_RESP_32 message is used to

core: Add ffa_msg_send_req and ffa_msg_send_resp

Add FF-A message handling of FFA_MSG_SEND_DIRECT_REQ_32 and
FFA_MSG_SEND_DIRECT_RESP_32 messages.
The FFA_MSG_SEND_DIRECT_RESP_32 message is used to send a message from
one FF-A endpoint to a second FF-A endpoint. FFA_MSG_SEND_DIRECT_RESP_32
is used to send a response message from the callee to the caller.

A separate abstraction is made (spmc_sp_handler) for handling FF-A
messages received from secure endpoints (SP).
thread_spmc_msg_recv() is still used for handling all FF-A messages
coming from the Normal World. When a FFA_MSG_SEND_DIRECT_REQ_32 message
is received for the Normal World a new kernel thread is started and
spmc_sp_msg_handler() is run. spmc_sp_msg_handler() is used as the main
message handler for handling the FF-A messages coming from SPs.

Whenever a message is sent to the normal world the thread is finished
and the FF-A message is forwarded to the Normal World.

Whenever a SVC exception is received from an SP, the whole context is
saved and executions is resumed in S-EL1. The first 8 register (FF-A
registers x0-x7) contain the FF-A message and are used to handle the
message by the SPMC.

Signed-off-by: Jelle Sels <jelle.sels@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

core: Initiate and load Secure Partitions

Secure Partitions (SP) are S-El0 execution service defined in the Arm
FF-A specification.
The Secure Partitions are loaded as the last part of the boot proc

core: Initiate and load Secure Partitions

Secure Partitions (SP) are S-El0 execution service defined in the Arm
FF-A specification.
The Secure Partitions are loaded as the last part of the boot process.
A Secure Partitions can be added to image using the SP_PATHS build
option. The SPs are loaded using ldelf.

Signed-off-by: Jelle Sels <jelle.sels@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...

core: Rename abort_print_current_ta

Rename abort_print_current_ta to abort_print_current_ts as it is used by
both SPs and TAs.

Signed-off-by: Jelle Sels <jelle.sels@arm.com>
Reviewed-by: Jens Wikla

core: Rename abort_print_current_ta

Rename abort_print_current_ta to abort_print_current_ts as it is used by
both SPs and TAs.

Signed-off-by: Jelle Sels <jelle.sels@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Etienne Carriere <etienne.carriere@linaro.org>

show more ...