core: sm_a32.S: add missing .cantunwind

Adds missing .cantunwind directives to the two alternative exception
vectors sm_vect_table_a15 and sm_vect_table_bpiall.

Reviewed-by: Jerome Forissier <jerom

core: sm_a32.S: add missing .cantunwind

Adds missing .cantunwind directives to the two alternative exception
vectors sm_vect_table_a15 and sm_vect_table_bpiall.

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

show more ...

core: correct alignment for exception vectors

The FUNC and LOCAL_FUNC macros changes section so specifying alignment
before the function will be lost. This may break for instance usage of
sm_vect_ta

core: correct alignment for exception vectors

The FUNC and LOCAL_FUNC macros changes section so specifying alignment
before the function will be lost. This may break for instance usage of
sm_vect_table, depending on configuration. Fix this by passing required
alignment as a parameter to the macro.

Fixes: a31e8303cf2e ("Remove '.section .text.<name>' and use function macros instead")
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

Add alignment parameter to FUNC() and LOCAL_FUNC() macros

Adds a default alignment parameter to FUNC() and LOCAL_FUNC() assembly
macros. This is needed since the function macro also sets the sectio

Add alignment parameter to FUNC() and LOCAL_FUNC() macros

Adds a default alignment parameter to FUNC() and LOCAL_FUNC() assembly
macros. This is needed since the function macro also sets the section
of the function making any alignment directives before the function
macro useless.

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

show more ...

ci: travis: download checkpatch.pl from linux-next

The Travis CI script obtains checkpatch.pl from a mirror of the master
branch of the Linux repository. Unfortunately, kernel v5.9 has a buggy
check

ci: travis: download checkpatch.pl from linux-next

The Travis CI script obtains checkpatch.pl from a mirror of the master
branch of the Linux repository. Unfortunately, kernel v5.9 has a buggy
checkpatch.pl [1]:

Global symbol "$gitroot" requires explicit package name [...]

The fix [2] is available in linux-next already but due to the Linux
release schedule it won't make it to the master branch until the end of
the v5.10-rc1 merge window, which is expected two weeks after the v5.9
release.

Downloading checkpatch.pl from linux-next seems to be a better idea in
general. Of course it might break more often due to the unstable nature
of linux-next, but on the other hand fixes appear much more quickly
than in master. A faster update time is even more beneficial in case
new features are added (such as the --kconfig-prefix option I submitted
some time ago [3]) because new patches have to wait up to several weeks
if they miss the merge window.

Therefore, change the checkpatch URL from linux master to linux-next to
fix the above error immediately and hopefully be more convenient in the
future.

Link: [1] https://travis-ci.org/github/OP-TEE/optee_os/builds/737636665#L264
Link: [2] https://lkml.org/lkml/2020/10/18/167
Link: [3] https://lkml.org/lkml/2020/8/18/212
Signed-off-by: Jerome Forissier <jerome@forissier.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: deprecate register_dynamic_shm()

register_dynamic_shm() and register_ddr() does almost the same thing but
they cannot be used both in the same platform build. To straighten out
this confusion

core: deprecate register_dynamic_shm()

register_dynamic_shm() and register_ddr() does almost the same thing but
they cannot be used both in the same platform build. To straighten out
this confusion we're switching to use only register_ddr() which is a bit
more flexible than register_dynamic_shm() since the former automatically
creates holes in the memory map where there otherwise would be a fatal
conflict.

register_dynamic_shm() is deprecated and for backwards compatibility
reimplemented as a special variant of register_ddr().

MEM_AREA_RAM_NSEC is kept only as a way of establish a static map of
non-secure memory. For example CFG_BCM_ELOG_BASE for plat-bcm.

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

show more ...

core: rpmb: check return value of encrypt_block()

Added error handling if call of function encrypt_block() fails
in tee_rpmb_req_pack().

Signed-off-by: Stefan Schmidt <snst@meek.de>
Reviewed-by: Je

core: rpmb: check return value of encrypt_block()

Added error handling if call of function encrypt_block() fails
in tee_rpmb_req_pack().

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

show more ...

ta_dev_kit.mk: make sure that libutils is linked second time

libgcc depends on __getauxval symbol from libuils. As, generally libutils
is linked before libgcc, we will get "unresolved symbol" error.

ta_dev_kit.mk: make sure that libutils is linked second time

libgcc depends on __getauxval symbol from libuils. As, generally libutils
is linked before libgcc, we will get "unresolved symbol" error. To resolve
this dependency we need to link libutils second time - after libgcc.

Signed-off-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

link.mk: implement support for libnames-after-libgcc variable

Newer versions of libgcc depend on external __getauxval() symbol, which is
now provided by libutils. But libgcc is linked after libutils

link.mk: implement support for libnames-after-libgcc variable

Newer versions of libgcc depend on external __getauxval() symbol, which is
now provided by libutils. But libgcc is linked after libutils, so linker
can't resolve that symbol. We can't include libgcc into linking group with
libtutils, because libgcc provides symbols that conflict with libutil's
ones, like __aeabi_idiv with friends for instance.

So, to resolve libgcc dependency on libutils we need to link with libutils
second time. To make things more generic, we will introduce
$(libnames-after-libgcc) variable for libraries that should be linked after
libgcc.

Signed-off-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

libutils: provide empty __getauxval() implementation

Never version of libgcc are built with LSE implementation in mind. To
determine if LSE is available on platform it calls __getauxval(), so in
som

libutils: provide empty __getauxval() implementation

Never version of libgcc are built with LSE implementation in mind. To
determine if LSE is available on platform it calls __getauxval(), so in
some cases we can get undefined reference to __getauxval() error.

Prominent case is libgcc_eh.a library, which is used by C++ TAs. Exception
handler depends on atomic operations, so it tries to call
init_have_lse_atomics() first. This function in turn calls __getauxval(),
which causes linking error.

In the future we can make __getauxval() to return actual platform
capabilities.

Signed-off-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

early_ta: use scattered array helpers

Simplifies the core linker script by replacing the hard coded
.rodata.early_ta section with use of SCATTERED_ARRAY_DEFINE_PG_ITEM()
instead.

Reviewed-by: Jerom

early_ta: use scattered array helpers

Simplifies the core linker script by replacing the hard coded
.rodata.early_ta section with use of SCATTERED_ARRAY_DEFINE_PG_ITEM()
instead.

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

show more ...

Update CHANGELOG for 3.11.0

Update CHANGELOG for 3.11.0 and collect Tested-by tags.

Signed-off-by: Jerome Forissier <jerome@forissier.org>
Tested-by: Jens Wiklander <jens.wiklander@linaro.org> (Jun

Update CHANGELOG for 3.11.0

Update CHANGELOG for 3.11.0 and collect Tested-by tags.

Signed-off-by: Jerome Forissier <jerome@forissier.org>
Tested-by: Jens Wiklander <jens.wiklander@linaro.org> (Juno)
Tested-by: Jerome Forissier <jerome@forissier.org> (QEMU)
Tested-by: Jerome Forissier <jerome@forissier.org> (QEMUv8)
Tested-by: Jerome Forissier <jerome@forissier.org> (HiKey960)
Tested-by: Jens Wiklander <jens.wiklander@linaro.org> (Hikey GP v1.1)
Tested-by: Jens Wiklander <jens.wiklander@linaro.org> (FVP)
Tested-by: Igor Opaniuk <igor.opaniuk@gmail.com> (Poplar)
Tested-by: Ricardo Salveti <ricardo@foundries.io> (ZynqMP)
Tested-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com> (Rcar H3)
Tested-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com> (Rcar H3/virtualization)
Tested-by: Rouven Czerwinski <r.czerwinski@pengutronix.de> (ccimx6ulsbcpro)
Tested-by: Rouven Czerwinski <r.czerwinski@pengutronix.de> (STM32MP157DK2)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6dlsabreauto)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6dlsabresd)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6qpsabreauto)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6ulevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6ullevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6ulzevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx6sllevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx7dsabresd)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx7ulpevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx8mmevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx8mnevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx8mqevk)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx8qmmek)
Tested-by: Clement Faure <clement.faure@nxp.com> (mx8qxpmek)
Tested-by: Etienne Carriere <etienne.carriere@linaro.org> (stm32mp1)

show more ...

plat: zynqmp: use generic_ram_layout for defining dram layout

Switch to the generic generic_ram_layout header file for defining the
default dram layout. This allow allows the user to easily customiz

plat: zynqmp: use generic_ram_layout for defining dram layout

Switch to the generic generic_ram_layout header file for defining the
default dram layout. This allow allows the user to easily customize the
default dram base and size via CFG_TZDRAM_START/CFG_TZDRAM_SIZE.

Default values are still the same as previously set by platform_config.

Signed-off-by: Ricardo Salveti <ricardo@foundries.io>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

scripts: support both pycryptodome and pycryptodomex

Current actual Python library for cryptography is being distributed in
two flavors: pycroptodome and pycroptodomex. They are basically the same
l

scripts: support both pycryptodome and pycryptodomex

Current actual Python library for cryptography is being distributed in
two flavors: pycroptodome and pycroptodomex. They are basically the same
library, but with different import names:

- pycryptodome provides 'Crypto' module and indented to directly replace
old pycrypto library

- pycryptodomex provides 'Cryptodome' module and is intended for old
distributions, where pycrypto is still present

Most of the modern Linux distributions provide both of the libraries, so
there is no difference which one is to use. But some of them (like
Yocto/Poky) provide only one.

This patches makes scripts agnostic to a crypto library flavor being used
by trying to import Cryptodome first and then Crypto if first import fails.

Signed-off-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

MAINTAINERS: maintain plat-zynqmp

Update MAINTAINERS with myself as reviewer for Xilinx Zynq
UltraScale+ MPSOC (plat-zynqmp).

Signed-off-by: Ricardo Salveti <ricardo@foundries.io>
Acked-by: Jerome

MAINTAINERS: maintain plat-zynqmp

Update MAINTAINERS with myself as reviewer for Xilinx Zynq
UltraScale+ MPSOC (plat-zynqmp).

Signed-off-by: Ricardo Salveti <ricardo@foundries.io>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

plat: zynqmp: force disable core ALSR

Disable core ASLR for two reasons:
1. There is no source for ALSR seed, as ATF does not provide a
DTB to OP-TEE. Hardware RNG is also not currently supported

plat: zynqmp: force disable core ALSR

Disable core ASLR for two reasons:
1. There is no source for ALSR seed, as ATF does not provide a
DTB to OP-TEE. Hardware RNG is also not currently supported.
2. OP-TEE does not boot with enabled CFG_CORE_ASLR.

Further investigation is needed to see why enabled ASLR causes
OP-TEE to not boot properly.

Signed-off-by: Ricardo Salveti <ricardo@foundries.io>
Acked-by: Jerome Forissier <jerome@forissier.org>

show more ...

core: rpmb: fix off-by-one in block index check

The max block index check in tee_rpmb_req_pack() is incorrect and would
fail when trying to access the last block of the partition.

Signed-off-by: Je

core: rpmb: fix off-by-one in block index check

The max block index check in tee_rpmb_req_pack() is incorrect and would
fail when trying to access the last block of the partition.

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

show more ...

drivers: scmi-msg: refine local header files

Change scmi-msg drivers to define/declare protocol resources in
protocol specific local file instead of from local common.h file.

Signed-off-by: Etienne

drivers: scmi-msg: refine local header files

Change scmi-msg drivers to define/declare protocol resources in
protocol specific local file instead of from local common.h file.

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

show more ...

drivers: scmi-msg: fix indentation

Fix minor indentation issue in voltage_domain.h.

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

drivers: scmi-msg: Set SCMI version to 3

Update SCMI version to 3 as supported voltage domain protocol is
introduced in the version 3 of the specification.

Fixes: 006d89b8f49f ("drivers: scmi-msg:

drivers: scmi-msg: Set SCMI version to 3

Update SCMI version to 3 as supported voltage domain protocol is
introduced in the version 3 of the specification.

Fixes: 006d89b8f49f ("drivers: scmi-msg: add SCMI Voltage Domain protocol")
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Reviewed-by: Jerome Forissier <jerome@forissier.org>

show more ...

plat-stm32mp1: scmi: expose PWR regulators

Expose SoC PWR regulators as SCMI voltage domains to the non-secure
world. They are exposed to SCMI agent scmi0 that exposes SoC resources
used by non-secu

plat-stm32mp1: scmi: expose PWR regulators

Expose SoC PWR regulators as SCMI voltage domains to the non-secure
world. They are exposed to SCMI agent scmi0 that exposes SoC resources
used by non-secure world but controlled by secure world since RCC TZ
hardening configuration.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

plat-stm32mp1: scmi: prepare platform SCMI voltage regulators

Define generic resources for stm32mp1 platform to support voltage
regulators exposed to SCMI agents through the voltage domain protocol.

plat-stm32mp1: scmi: prepare platform SCMI voltage regulators

Define generic resources for stm32mp1 platform to support voltage
regulators exposed to SCMI agents through the voltage domain protocol.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

plat-stm32mp1: introduce PWR regulators

Introduce a voltage regulator driver for the voltage controllers
driven through PWR interface of stm32mp1 SoCs.

Signed-off-by: Etienne Carriere <etienne.carr

plat-stm32mp1: introduce PWR regulators

Introduce a voltage regulator driver for the voltage controllers
driven through PWR interface of stm32mp1 SoCs.

Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

MAINTAINERS: plat-ti, plat-k3: remove Andrew's address

Andrew's email address is bouncing, remove it.

Signed-off-by: Jerome Forissier <jerome@forissier.org>
Acked-by: Joakim Bech <joakim.bech@linar

MAINTAINERS: plat-ti, plat-k3: remove Andrew's address

Andrew's email address is bouncing, remove it.

Signed-off-by: Jerome Forissier <jerome@forissier.org>
Acked-by: Joakim Bech <joakim.bech@linaro.org>

show more ...

core: ffa: Fix the FF-A version returned to Secure Partition

The current Firmware Framework(FF-A) specification defines the major
version of 0x1 and minor version of 0x0. Return these values when
re

core: ffa: Fix the FF-A version returned to Secure Partition

The current Firmware Framework(FF-A) specification defines the major
version of 0x1 and minor version of 0x0. Return these values when
requested through the FFA_VERSION call instead of returning the
FFA_VERSION function id

Signed-off-by: Sughosh Ganu <sughosh.ganu@linaro.org>
Acked-by: Jerome Forissier <jerome@forissier.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

Update revision for release tag 3.11.0-rc1

Signed-off-by: Jerome Forissier <jerome@forissier.org>