ci: travis: delete toolchain binaries after download

Since build.git has been upgraded to GCC 8.2 [1], we are running out of
disk space when running the Travis CI script. This is because the GCC 8.2

ci: travis: delete toolchain binaries after download

Since build.git has been upgraded to GCC 8.2 [1], we are running out of
disk space when running the Travis CI script. This is because the GCC 8.2
binaries are much larger than 6.2:

218M gcc-arm-8.2-2018.08-x86_64-aarch64-linux-gnu.tar.xz
213M gcc-arm-8.2-2018.08-x86_64-arm-linux-gnueabihf.tar.xz
94M gcc-linaro-6.2.1-2016.11-x86_64_aarch64-linux-gnu.tar.xz
88M gcc-linaro-6.2.1-2016.11-x86_64_arm-linux-gnueabihf.tar.xz

This commit makes some room by deleting the packages once they have been
extracted.

Link: [1] https://github.com/OP-TEE/build/commit/53119b77e9f2
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

symbolize: use CROSS_COMPILE directive if available

Get CROSS_COMPILE prefix directive if defined in the environment.
If CROSS_COMPILE is not defined, default to aarch64-linux-gnu- or
arm-linux-gnue

symbolize: use CROSS_COMPILE directive if available

Get CROSS_COMPILE prefix directive if defined in the environment.
If CROSS_COMPILE is not defined, default to aarch64-linux-gnu- or
arm-linux-gnueabihf- as done prior this change according to the
architecture read from the ELF file.

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

show more ...

abort.c: always save VFP state with thread context

Saving VFP state requires a thread context, if none is available print
abort info and panic().

Reviewed-by: Volodymyr Babchuk <volodymyr_babchuk@e

abort.c: always save VFP state with thread context

Saving VFP state requires a thread context, if none is available print
abort info and panic().

Reviewed-by: Volodymyr Babchuk <volodymyr_babchuk@epam.com>
Fixes: cfa34ec63699 ("abort.c: manipulate with VFP state only if thread is active")
Tested-by: Jens Wiklander <jens.wiklander@linaro.org> (Hikey AArch64 pager)
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

core: trace out size of the pager physical page pool

Add a nice info trace about the size of the pager physical
page pool.

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

core: trace out size of the pager physical page pool

Add a nice info trace about the size of the pager physical
page pool.

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

show more ...

core: correct unpaged constraint on GIC driver

Release of secondary boot cores on 32bit machine use SMC that
issue a SGI on secondary core. Since the interrupt is raised
from the monitor mode, the r

core: correct unpaged constraint on GIC driver

Release of secondary boot cores on 32bit machine use SMC that
issue a SGI on secondary core. Since the interrupt is raised
from the monitor mode, the related GIC driver resources must be
tagged as unpaged.

This change costs around 300 bytes of unpaged resident memory.

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

show more ...

core: correct dt_find_compatible_driver() if no driver found

By convention the compatible array for driver probing ends with
a NULL compatible string ID reference.

This change ensures such NULL ref

core: correct dt_find_compatible_driver() if no driver found

By convention the compatible array for driver probing ends with
a NULL compatible string ID reference.

This change ensures such NULL reference is properly handled as
the end of the compatible array references.

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

show more ...

Allow mixed declaration and code

Removes the -Wdeclaration-after-statement compiler flag to allow mixed
declaration and code

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

Allow mixed declaration and code

Removes the -Wdeclaration-after-statement compiler flag to allow mixed
declaration and code

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

show more ...

optee_msg.h: remove OPTEE_MSG_ATTR_FRAGMENT

OPTEE_MSG_ATTR_FRAGMENT isn't defined. Remove it from comments and replace
with OPTEE_MSG_ATTR_NONCONTIG where applicable.

Acked-by: Jerome Forissier <je

optee_msg.h: remove OPTEE_MSG_ATTR_FRAGMENT

OPTEE_MSG_ATTR_FRAGMENT isn't defined. Remove it from comments and replace
with OPTEE_MSG_ATTR_NONCONTIG where applicable.

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

show more ...

core: reference count registered shared memory

* Adds a reference counter to registered shared memory to make sure that
it's not unregistered while in use.
* Updates entry_std to use the reference

core: reference count registered shared memory

* Adds a reference counter to registered shared memory to make sure that
it's not unregistered while in use.
* Updates entry_std to use the reference counting functions

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

show more ...

entry_std: use READ_ONCE() to read from shared memory

Use the READ_ONCE() macro everywhere shared memory is read to make sure
that unexpected values can not be used.

Reviewed-by: Volodymyr Babchuk

entry_std: use READ_ONCE() to read from shared memory

Use the READ_ONCE() macro everywhere shared memory is read to make sure
that unexpected values can not be used.

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

show more ...

ta: link.mk: evaluate $(sm) values in targets

$(sm) var isn't properly evalutated as the link.mk is included (final
evaluation is done when the rule is triggered). Add dynamic target
generation base

ta: link.mk: evaluate $(sm) values in targets

$(sm) var isn't properly evalutated as the link.mk is included (final
evaluation is done when the rule is triggered). Add dynamic target
generation based on the current value of $(sm)

Fixes: https://github.com/OP-TEE/optee_os/issues/2537
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Igor Opaniuk <igor.opaniuk@linaro.org>

show more ...

mk: cleanvars.mk: clean spec-srcs var

Clean $(spec-srcs) variable, as the value can be accumulated in
ta/mk/build-user-ta.mk

Fixes linking issues when multiple TAs are compiled as Early TAs
(using

mk: cleanvars.mk: clean spec-srcs var

Clean $(spec-srcs) variable, as the value can be accumulated in
ta/mk/build-user-ta.mk

Fixes linking issues when multiple TAs are compiled as Early TAs
(using CFG_IN_TREE_EARLY_TAS)

$ make V=1
...
out/arm/ta/avb/user_ta_header.o:(.data+0x0): multiple definition of `trace_level'
out/arm/ta/avb/user_ta_header.o:(.data+0x0): first defined here
out/arm/ta/avb/user_ta_header.o: In function `tahead_get_trace_level':
optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: multiple definition of `trace_ext_prefix'
out/arm/ta/avb/user_ta_header.o:optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: first defined here
out/arm/ta/avb/user_ta_header.o: In function `tahead_get_trace_level':
optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: multiple definition of `ta_head'
out/arm/ta/avb/user_ta_header.o:optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: first defined here
out/arm/ta/avb/user_ta_header.o: In function `tahead_get_trace_level':
optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: multiple definition of `ta_heap'
out/arm/ta/avb/user_ta_header.o:optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: first defined here
out/arm/ta/avb/user_ta_header.o: In function `tahead_get_trace_level':
optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: multiple definition of `ta_heap_size'
out/arm/ta/avb/user_ta_header.o:optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: first defined here
out/arm/ta/avb/user_ta_header.o: In function `tahead_get_trace_level':
optee_os/out/arm/export-ta_arm32/src/user_ta_header.c:102: multiple definition of `ta_props'

Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Igor Opaniuk <igor.opaniuk@linaro.org>

show more ...

core: arm: imx: Support psci feature query

Support PSCI call to query features list. So that non-secure world
knows what features are supported by ATF. The feature list is based
on current implement

core: arm: imx: Support psci feature query

Support PSCI call to query features list. So that non-secure world
knows what features are supported by ATF. The feature list is based
on current implemented iMX psci functions.

Signed-off-by: Jun Nie <jun.nie@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Acked-by: Peng Fan <peng.fan@nxp.com>

show more ...

Fix copying dirty values in copy_in_params()

If the OP-TEE driver from the rich OS specifies a message with a number
of params < TEE_NUM_PARAMS, copy_in_params() will copy in undefined
values from p

Fix copying dirty values in copy_in_params()

If the OP-TEE driver from the rich OS specifies a message with a number
of params < TEE_NUM_PARAMS, copy_in_params() will copy in undefined
values from pt[i] (where i >= the number of params). This is because the
pt array is an uninitialized local value, and per the C99 standard
6.7.8:

If an object that has automatic storage duration is not initialized
explicitly, its value is indeterminate.

This change fixes this issue by clearing out the unused parts of pt.

Signed-off-by: Christopher Tam <godtamit@google.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

scripts/symbolize.py: improve error message when a command is not found

It is a common mistake to invoke symbolize.py while not having the proper
toolchain is one's $PATH. The script then exits with

scripts/symbolize.py: improve error message when a command is not found

It is a common mistake to invoke symbolize.py while not having the proper
toolchain is one's $PATH. The script then exits with a stack dump that is
unhelpful to users.
Improve the situation by showing an explicit "command not found" message
and the name of the command.

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

show more ...

scripts/symbolize.py: code cleanup

Fix pep8 errors.

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

core: fix tee_tadb_ta_create() panic

Fixes a panic triggered in tee_tadb_ta_create(). Before this patch
tee_tadb_ta_create() was calling tadb_put() if tee_tadb_open() failed.
This is incorrect as th

core: fix tee_tadb_ta_create() panic

Fixes a panic triggered in tee_tadb_ta_create(). Before this patch
tee_tadb_ta_create() was calling tadb_put() if tee_tadb_open() failed.
This is incorrect as the reference counter hasn't been increased then.
This patch fixes that by only calling tadb_put() once tee_tadb_open()
has succeeded.

Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Signed-off-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

entry_std: use READ_ONCE in strategic places

Code that deals with command buffers follows rule
"read once, validate, use". Problem is that compiler
does not know about this rule, so it can optimize

entry_std: use READ_ONCE in strategic places

Code that deals with command buffers follows rule
"read once, validate, use". Problem is that compiler
does not know about this rule, so it can optimize out
temporary variables and read data twice from the shared
buffer.

READ_ONCE() will ensure that compiler will not try to
optimize such reads.

Signed-off-by: Volodymyr Babchuk <vlad.babchuk@gmail.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

io.h: add READ_ONCE macro

Compiler can rearrange memory reads and writes if it does
not see any dependency on them. This can be troublesome
if we deal with memory which is shared with non-secure wor

io.h: add READ_ONCE macro

Compiler can rearrange memory reads and writes if it does
not see any dependency on them. This can be troublesome
if we deal with memory which is shared with non-secure world.

READ_ONCE macro ensures that compiler will read memory only once.
It is simple wrapper over __compiler_atomic_load(), but it's name
emphasizes it's function.

Signed-off-by: Volodymyr Babchuk <vlad.babchuk@gmail.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

aosp_optee.mk: Fix compilation for build outside of source

Trying to specify Android build location outside of souce directory's
default 'out' using OUT_DIR_COMMON_BASE env variable, results in erro

aosp_optee.mk: Fix compilation for build outside of source

Trying to specify Android build location outside of souce directory's
default 'out' using OUT_DIR_COMMON_BASE env variable, results in errors
like
"...
<built-in>:1:10: fatal error: 'conf.h' file not found
"

Signed-off-by: Jassi Brar <jaswinder.singh@linaro.org>
Tested-by: Victor Chong <victor.chong@linaro.org> (hikey620)
Acked-by: Victor Chong <victor.chong@linaro.org>
Acked-by: Jerome Forissier <jerome.forissier@linaro.org>
[jf: wrap commit text]
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

show more ...

documentation: clarify $(ARCH)

$(ARCH) has to be 'arm' for both 32-bit and 64-bit Arm builds.

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

documentation: clarify $(ARCH)

$(ARCH) has to be 'arm' for both 32-bit and 64-bit Arm builds.

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

show more ...

README.md: do not mention the checkpatch license

checkpatch.pl being covered by the GPL is not a convincing reason for
not having a local copy of the script. Nothing in OP-TEE is a derivative
work o

README.md: do not mention the checkpatch license

checkpatch.pl being covered by the GPL is not a convincing reason for
not having a local copy of the script. Nothing in OP-TEE is a derivative
work of checkpatch, and a "mere aggregation" is even explicitly covered in
the GPL text. It isn't necessary to be concerned with the output of
checkpatch, since that is not included in OP-TEE, either.

Therefore, remove the sentence that mentions the checkpatch license.

Reported-by: David Brown <david.brown@linaro.org>
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Acked-by: David Brown <david.brown@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

show more ...

porting guide: fix path

Markdown format is no longer used (superseded by rst format, see [1])

Link: [1] https://github.com/ARM-software/arm-trusted-firmware/commit/6f6257476754
Signed-off-by: Pierr

porting guide: fix path

Markdown format is no longer used (superseded by rst format, see [1])

Link: [1] https://github.com/ARM-software/arm-trusted-firmware/commit/6f6257476754
Signed-off-by: Pierre-Jean Texier <texier.pj2@gmail.com>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
[jf: shorten URL and move it to a Link: tag]
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>

show more ...

plat-rpi3: Modify addresses to work with upstream TF-A.

The upstream arm-trusted-firmware reserves 0x10100000-0x11000000 for
secure DRAM. Change the address according to the upstream TF-A.

To help

plat-rpi3: Modify addresses to work with upstream TF-A.

The upstream arm-trusted-firmware reserves 0x10100000-0x11000000 for
secure DRAM. Change the address according to the upstream TF-A.

To help troubleshoot discrepancies, this commit goes well with the TF-A at
commit aa49bde8a3e8 ("rpi3: Move NS-DRAM out of the protected region")

Acked-by: Jerome Forissier <jerome.forissier@linaro.org>
Tested-by: Igor Opaniuk <igor.opaniuk@linaro.org>
Reviewed-by: Igor Opaniuk <igor.opaniuk@linaro.org>
Signed-off-by: Ying-Chun Liu (PaulLiu) <paul.liu@linaro.org>

show more ...

ta: link.mk: filter out dependency and command files from $(MAKEFILE_LIST)

Prevents ta.lds from being needlessly re-generated, such as in the
following test case:

$ make -s out/arm-plat-vexpress/t

ta: link.mk: filter out dependency and command files from $(MAKEFILE_LIST)

Prevents ta.lds from being needlessly re-generated, such as in the
following test case:

$ make -s out/arm-plat-vexpress/ta/avb/ta.lds
$ make -s out/arm-plat-vexpress/core/kernel/console.o
$ make out/arm-plat-vexpress/ta/avb/ta.lds
CHK out/arm-plat-vexpress/conf.mk
CHK out/arm-plat-vexpress/include/generated/conf.h
CPP out/arm-plat-vexpress/ta/avb/ta.lds

The last line should not hapen because ta.lds exists already, and building
the unrelated console.o should not trigger a re-build.

The cause of the re-build is the dependency on $(MAKEFILE_LIST) which
contains *all* the files that record previous build command and
dependencies such as out/arm-plat-vexpress/core/kernel/.console.o.{d,cmd}.

Filter out those files, thus keeping only the static Makefiles (*.mk).
This is still more than needed since all the core *.mk are present, but
listing only the TA-related makefiles seems tricky (has to work both
inside and outside the OP-TEE build environment), and removing
$(MAKEFILE_LIST) altogether presents a risk that TA developers using the
TA dev kit will complain that their TA doesn't get re-built after a flag
is changed in a Makefile.

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

show more ...