When executing the command `$ make -f versal.mk image` when following the
instructions in:

https://github.com/OP-TEE/optee_docs/blob/master/building/devices/versal.rst#configuring-and-building-for-vck190



an error is being encountered when trying to recreate this build using Ubuntu
20.04.5 LTS

`../u-boot/tools/mkimage -f versal/fitImage-versal-vck190.its versal/versal-vck190.ub`
`versal/fitImage-versal-vck190.its:17.20-19.15: Warning (unit_address_vs_reg): /images/kernel/hash@1: node has a unit name, but no reg property`
`versal/fitImage-versal-vck190.its:28.20-30.15: Warning (unit_address_vs_reg): /images/fdt-1/hash@1: node has a unit name, but no reg property`
`versal/fitImage-versal-vck190.its:39.20-41.15: Warning (unit_address_vs_reg): /images/fdt-2/hash@1: node has a unit name, but no reg property`
`versal/fitImage-versal-vck190.its:50.20-52.15: Warning (unit_address_vs_reg): /images/ramdisk/hash@1: node has a unit name, but no reg property`
`versal/fitImage-versal-vck190.its:57.16-65.11: Warning (unit_address_vs_reg): /configurations/conf@1: node has a unit name, but no reg property`
`versal/fitImage-versal-vck190.its:62.20-64.15: Warning (unit_address_vs_reg): /configurations/conf@1/hash@1: node has a unit name, but no reg property`
`../u-boot/tools/mkimage: verify_header failed for FIT Image support with exit code 1`
`make: *** [versal.mk:157: fitimage] Error 1`

This commit resolves the issue.

Signed-off-by: Nathan Menhorn <nathan.menhorn@amd.com>
Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

Support the build repo for Xilinx/AMD ZCU104, ZCU106 and Ultra96
boards.

Signed-off-by: Ibai Erkiaga <ibai.erkiaga-elorza@amd.com>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Fix ITS warning messages related to usage of @ in non memory related
nodes.

i.e.
zynqmp/fitImage-zynqmp-zcu102.its:17.20-19.15: Warning (unit_address_vs_reg): /images/kernel/hash@1: node has a unit name, but no reg or ranges property

Signed-off-by: Ibai Erkiaga <ibai.erkiaga-elorza@amd.com>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Use DTS overlay to add reserved memory areas for Trusted Firmware and
OP-TEE and mark as no-map in order to ensure kernel is not using these
memory areas.

Signed-off-by: Ibai Erkiaga <ibai.erkiaga-elorza@amd.com>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Adds CFG_CORE_PAUTH=y when configured with PAUTH=y. Also adds
pauth-impdef=on to QEMU_CPU to avoid performance degradation when
testing with CFG_CORE_PAUTH=y.

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

Add support for building the SPMC tests and the corresponding SPs in
xtest. Default behaviour is to use the already build binaries.

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

Move a comment line which accidentally added a whitespace to the end of
SP_PACKAGING_METHOD default value.

Fixes: db9b8f09 ("fvp: add support for Secure Partitions in the FIP")
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

System install:

The Versal ACAP user will need to download and install Petalinux and
the Versal ACAP BSP package [1]

  $ ls
    xilinx-vck190-v2022.1-04191534.bsp

  $ mkdir -p optee-project
  $ cd optee-project
  $ repo init -u https://github.com/OP-TEE/manifest.git -m versal.xml
  $ repo sync -j4 --no-clone-bundle
  $ cd build
  $ make -j8 toolchains
  $ make -j8

The BSP will need to be upacked somewhere in the filesystem; to do that, once
Petalinux has been installed do the following:

  $ cd ../../
  $  . /path/to/petalinux.2022.1/settings.sh
  $  petalinux-create --type project -s xilinx-vck190-v2022.1-04191534.bsp
  $ ls
    xilinx-vck190-2022.1

To build the Versal ACAP image, versal/bootImage-versal-vck190.bif needs to be
manually modified so some of the firmware files can be reached from the BSP.

Once done

  $ cd optee-project/build
  $ make -f versal.mk image

At that point the images are in optee-project/build/versal

To boot from JTAG:
  $ cd versal
  $ ./boot_jtag.sh

To boot from SD card create a big enough bootable partition on the SD and copy
the images
  $ cp BOOT.BIN  /mnt/usr/boot_partition/
  $ cp versal-vck190.ub  /mnt/usr/boot_partition/

Plug it in the device, set the boot switches and power up the board

  At the u-boot shell stop, load the itb and boot it.

  uboot> mmc dev 0
  uboot> fatload mmc 0:1 0x20000000 versal-vck190.ub
  uboot> bootm 0x20000000

  At the login prompt type "root" to enter and execute xtests

  [    2.391905] mmc0: new high speed SDHC card at address aaaa
  [    2.397907] mmcblk0: mmc0:aaaa SS08G 7.40 GiB
  OP-TEE embedded distrib for versal-vck190
  buildroot login: root
  # xtest
  Run test suite with level=0
  TEE test application started over default TEE instance
  [...]

[1] https://xilinx-wiki.atlassian.net/wiki/spaces/A/pages/2347204609/2022.1+Release



Signed-off-by: Jorge Ramirez-Ortiz <jorge@foundries.io>
Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

In some environments it has been noted that the Rust setup fails with a
netowrk error [1]. It seems to be an issue with the built-in Git
library, a workaround being to tell the Rust setup to use the Git
command instead. This commit does exactly that, regardless of the
environment.

Note: I chose not to restore the OP-TEE Rust build commands removed by
[1] in order to reduce the size of the image ($TOP/optee_rust is 1.8GB
after building) and avoid disk space issues in the "make check (QEMUv8)"
job which uses this image but doesn't build the Rust tests. The "make
check-rust (QEMUv8)" job will do whataver it takes.

Link: [1] https://github.com/jforissier/docker_optee_os_ci/commit/6d7f39aec9436168e5e4ce44c4b6f7791575e313


Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Acked-by: Yuan Zhuang <zhuangyuan04@baidu.com>

This commit introduces a new packaging option for Secure Partitions.
Instead of embedding the images into the OP-TEE binary, TF-A offers a
mechanism to encapsulate an SP image and its manifest into an SP package
and add that to the FIP [1].

TF-A needs two config options to enable this:
 - SP_LAYOUT_FILE: This json file contains the path of the SP images and
   their corresponding manifests. It's generated by Trusted Services.
 - ARM_BL2_SP_LIST_DTS: This dts snippet describes where each SP package
   should be loaded by BL2.

Link: [1] https://trustedfirmware-a.readthedocs.io/en/v2.6/components/secure-partition-manager.html#secure-partition-packages


Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Signed-off-by: Johann Neuhauser <jneuhauser@dh-electronics.com>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Signed-off-by: Johann Neuhauser <jneuhauser@dh-electronics.com>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Signed-off-by: Johann Neuhauser <jneuhauser@dh-electronics.com>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

In order to use login methods `TEEC_LOGIN_USER` and `TEEC_LOGIN_GROUP`,
a user space application must initialize the tokens with the SHA1
hashed UUID of the user or group.

Select host-pkgconf, util-linux and libuuid in order to use `libuuid`
and add supporting libraries to optee_client for this type of ACL based
login initialization.

Signed-off-by: Eero Aaltonen <eero.aaltonen@vaisala.com>
Acked-by: Jerome Forissier <jerome.forissier@linaro.org>

Buildroot parallel build is allowed when BR2_PER_PACKAGE_DIRECTORIES is
'y', see commit e9bff1c8 ("common.mk: buildroot: enable top-level
parallel build"). When disabled, Buildroot is not supposed to spawn
several make jobs simultaneously. However for some reason it seems it
can happen, see the IBART failure at [1]. The steps:
 >>> optee_client_ext 1.0 Fixing libtool files
and:
 >>> pcsc-lite 1.9.4 Fixing libtool files
...seem to be running simultaneously and touching the same path (since
per-package is disabled), which results in:
 mv: cannot stat '/home/optee/devel/qemu/out-br/host/arm-buildroot-linux-gnueabihf/sysroot/usr/lib/libpcsclite.la.fixed': No such file or directory
 package/pkg-generic.mk:331: recipe for target '/home/optee/devel/qemu/out-br/build/pcsc-lite-1.9.4/.stamp_staging_installed' failed

I confirmed the fact by adding a 'sleep 5' after "Fixing libtool files"
in package/pkg-generic.mk, as well as a 'ps -ef | grep sleep | grep -v
grep || :' just before the sleep. Then when building with
'make -j32 BR2_PER_PACKAGE_DIRECTORIES=n' (32 core machine) the 'sleep 5'
is displayed occasionally.

To avoid that, let's explicitly invoke make with -j1 when
BR2_PER_PACKAGE_DIRECTORIES is disabled.

Link: [1] https://optee.mooo.com:5000/logs/OP-TEE/optee_os/5487/1027234243/9147c347fb53612f62bf7de1c17e47a246d61ef2#02-qemu-build

 line 16398:
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Reviewed-by: Joakim Bech <joakim.bech@linaro.org>

On an Aarch64 host, "make toolchains" does not use parallel jobs when
building the Aarch64 GCC and the following warning is shown:

 $ make -j32 toolchains
 [...]
 make[1]: warning: jobserver unavailable: using -j1.  Add `+' to parent make rule.

Therefore, add the '+' sign as recommended. This speeds up the build
quite a bit on a 32-core machine for example.

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

The warning on option "-t" is remove in below commit.

https://gitlab.gnome.org/GNOME/gnome-terminal/-/commit/54da2fe919634e2d6a6aab392b36449947b812cd



And it seens work again, so add it back to make the
terminal title understandable.

Signed-off-by: Mark-PK Tsai <mark-pk.tsai@mediatek.com>
Reviewed-by: Jerome Forissier <jerome.forissier@linaro.org>

Increase the size of the guest (DomU) RAM from 128 to 160 MB to fix
an issue with xtest 6018:

  * regression_6018 Large object
  o regression_6018.1 Storage id: 00000001
  writing 0
  writing 1
  writing 2
  writing 3
  writing 4
  writing 5
  regression_6000.c:1895: fs_write(&sess, obj, block, block_size) has an unexpected value: 0xffff3024 = TEE_ERROR_TARGET_DEAD, expected 0x0 = TEEC_SUCCESS
    regression_6018.1 FAILED

For some reason, this test passes in CI [1] but not on my laptop
("make XEN_BOOT=y run", login: "root", run /bin/domu, login: "test"
then run "xtest 6018"). tee-supplicant receives error ENOSPC when
trying to write to /data/tee. Since the root FS in the guest is a
ramdisk, increasing the RAM seems to be the appropriate solution.

Link: [1] https://github.com/OP-TEE/optee_os/blob/578f89d28d2b66c35128f095989c355fbc2ef3ab/.azure-pipelines.yml#L256


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

Configures U-Boot with stm32mp15_defconfig that was introduced in
U-Boot tag v2021.10 for devices booted with OP-TEE and U-Boot stored
in TF-A FIP container. From U-Boot v2021.10 onward,
stm32mp15_trusted_defconfig is dedicated to boot scheme that do not
use TF-A FIP format.

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

The regular expression used to detect TEE core panics or assertion
failures must not allow multi-line text to match, otherwise the following
input may trigger it (test case: xtest 1007 with "make COMPILER=clang").

E/TC:? 0 TA panicked with code 0xbeef
E/LD:  Status of TA 5b9e0e40-2636-11e1-ad9e-0002a5d5c51b
E/LD:   arch: arm
E/LD:  region  0: va 0x00102000 pa 0x0e300000 size 0x002000 flags rw-s (ldelf)
E/LD:  region  1: va 0x00104000 pa 0x0e302000 size 0x00e000 flags r-xs (ldelf)
E/LD:  region  2: va 0x00112000 pa 0x0e310000 size 0x001000 flags rw-s (ldelf)
E/LD:  region  3: va 0x00113000 pa 0x0e311000 size 0x004000 flags rw-s (ldelf)
E/LD:  region  4: va 0x00117000 pa 0x0e315000 size 0x001000 flags r--s
E/LD:  region  5: va 0x00118000 pa 0x0e410000 size 0x003000 flags rw-s (stack)
E/LD:  region  6: va 0x00184000 pa 0x00001000 size 0x015000 flags r-xs [0]
E/LD:  region  7: va 0x00199000 pa 0x00016000 size 0x0e5000 flags rw-s [0]
E/LD:  assertion 'maps[map_idx].sz == sz' failed at ldelf/ta_elf.c:1451 in ta_elf_print_mappings()

In addition to using '[^n]*' instead of '.*', a couple of things need
fixing too:

- In Expect, `^` and `$` are the beginning and end of the current matching
buffer, not the beginning and end of a line. Therefore `^` is not
appropriate here.
- `\n` should not be matched because if it is consumed for example at the
end of an expression to mean "end of line", then it cannot be available to
another expression to mean "beginning of line".

Fixes: 09900ef9 ("qemu-check.exp: trap only OP-TEE Core assert or panic")
Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>

Build the Linux test applications used by Trusted Services to invoke the
services provided by Secure Partitions.

Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Build the Linux drivers used by Trusted Services to access FF-A from
user space.

Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Add support for building Secure Partitions implemented by the Trusted
Services project. The patch contains helper functions to invoke the TS
buildsystem in order to build SPs and include them into the OP-TEE bin.
It also adds a new top level Makefile which configures OP-TEE and TF-A
for the S-EL1 SPMC scenario on FVP platform.

Link: https://www.trustedfirmware.org/projects/trusted-services/


Acked-by: Jerome Forissier <jerome.forissier@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>
Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

The Measured Boot option not only enables the relevant config options
for TF-A and optee_os, but also building the fTPM TA and its Linux
driver. This is unnecessary when the consumer of the Event Log is not
fTPM but e.g. the PSA Attestation SP (as introduced in the next commit).

This commit modifies the MEASURED_BOOT option to only set the TF-A and
optee_os config, and adds the MEASURED_BOOT_FTPM option for the fTPM
related components. To keep backwards compatibility if MEASURED_BOOT is
selected, by default MEASURED_BOOT_FTPM is enabled too.

Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Add support for sharing a host directory with the FVP.

Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Add support for running the Base Platform FVP. This model offers more
configuration options and more peripherals than the Foundation model.
It's available free of charge on the Arm Developer website.

Link: https://developer.arm.com/downloads/-/arm-ecosystem-models


Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Refactor the FVP path and argument variables to prepare for adding a new
FVP platform in the next commit.

Reviewed-by: Gyorgy Szing <gyorgy.szing@arm.com>
Reviewed-by: Jens Wiklander <jens.wiklander@linaro.org>
Signed-off-by: Balint Dobszay <balint.dobszay@arm.com>

Adds Xen upstream patch "optee: enable OPTEE_SMC_SEC_CAP_MEMREF_NULL
capability" ([1] present in Xen 4.16.0) in order to avoid failures in
the GlobalPlatform test suite [2]:

 optee/build $ export GP_PACKAGE=~/work/TEE_Initial_Configuration-Test_Suite_v2_0_0_2-2017_06_09.7z
 optee/build $ make -j10 XEN_BOOT=y run
 ...
 $ xtest gp_
 ...
 gp_10081 FAILED first error at gp_10000.c:1653
 gp_10082 FAILED first error at gp_10000.c:1673
 gp_10083 FAILED first error at gp_10000.c:1693
 gp_10155 FAILED first error at gp_10000.c:3346
 gp_10156 FAILED first error at gp_10000.c:3366
 gp_10157 FAILED first error at gp_10000.c:3386
 gp_10229 FAILED first error at gp_10000.c:5039
 gp_10230 FAILED first error at gp_10000.c:5059
 gp_10231 FAILED first error at gp_10000.c:5079
 gp_10303 FAILED first error at gp_10000.c:6732
 gp_10304 FAILED first error at gp_10000.c:6752
 gp_10305 FAILED first error at gp_10000.c:6772
 gp_10632 FAILED first error at gp_10000.c:15608
 gp_10633 FAILED first error at gp_10000.c:15629
 gp_10634 FAILED first error at gp_10000.c:15650
 gp_10706 FAILED first error at gp_10000.c:17375
 gp_10707 FAILED first error at gp_10000.c:17396
 gp_10708 FAILED first error at gp_10000.c:17417
 gp_10780 FAILED first error at gp_10000.c:19142
 gp_10781 FAILED first error at gp_10000.c:19163
 gp_10782 FAILED first error at gp_10000.c:19184
 gp_10854 FAILED first error at gp_10000.c:20909
 gp_10855 FAILED first error at gp_10000.c:20930
 gp_10856 FAILED first error at gp_10000.c:20951

For each failed test there is a TA log:

 E/TA:  checkUpdateParameterMemref:195 Received buffer is not NULL but expected to be NULL.

Link: [1] https://xenbits.xen.org/gitweb/?p=xen.git;a=commit;h=d4fb5f166c2b
Link: [2] https://github.com/OP-TEE/optee_os/issues/4897


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

"make optee-rust OPTEE_RUST_ENABLE=y" always runs the setup.sh script
in $(ROOT)/optee_rust. The script clones about 600 MB of data. That
can't be done each time especially since the check-rust target depends
on optee-rust.

Signed-off-by: Jerome Forissier <jerome.forissier@linaro.org>
Tested-by: Yuan Zhuang <zhuangyuan04@baidu.com>
Acked-by: Yuan Zhuang <zhuangyuan04@baidu.com>

Adds a Xen bug fix from upstream [1]. With this we will be able to get
rid of a revert in the linux kernel [2].

Link: [1] https://github.com/xen-project/xen/commit/1c3ed9c908732d19660fbe83580674d585464d4c
Link: [2] https://github.com/linaro-swg/linux/commit/665750f493d3fcf307952f19c0442c092b76d852


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

Commit 8b808915 ("qemu_v8: xen: apply patch to fix issue with RPC
buffer release") really did only one part of the job. Only the Xen tools
are built with Buildroot; the hypervisor itself (which needs the patch
mentioned in the commit) is actually built from the top-level xen/
directory. As a result the issue is stil present.
This commit changes the QEMUv8 makefiles as well as the Buildroot
configuration for Xen so that both the hypervisor and the tools are
built. As for the Xen configuration (enabling OP-TEE support), it is
introduced via a patch to the defconfig file.

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

Adds a patch to revert a Xen commit [1] which causes a regression with
OP-TEE. There is no more logs to the secure console after these lines:

 I/TC: Added guest 1
 D/TC:2   tee_entry_exchange_capabilities:100 Asynchronous notifications are disabled
 D/TC:2   tee_entry_exchange_capabilities:109 Dynamic shared memory is enabled

...and TAs fail to open sessions with TEEC_ERROR_COMMUNICATION.

This is to prepare building Xen with Buildroot. The version of Xen
pulled by $(ROOT)/buildroot/package/xen is 4.14.3 and is affected by this
issue while the one we are currently using in $(ROOT)/xen is 4.14.1
and doesn't have the regression.

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

When C source code is built into the Rust TA, the Rust crate `cc`[1]
uses gcc compiler defined in the env var `CC`. Setting `CC` in
`optee_rust_examples_ext.mk` fixes the building error of those Rust
TAs in buildroot.

[1] https://crates.io/crates/cc



Signed-off-by: Yuan Zhuang <zhuangyuan04@baidu.com>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

In the new version of Teaclave TrustZone SDK, OPTEE_DIR is checked
in setup.sh. If OPTEE_DIR is not set, it will download OP-TEE source
code.
Set OPTEE_DIR as the directory of OP-TEE repo.

Signed-off-by: Yuan Zhuang <zhuangyuan04@baidu.com>
Acked-by: Joakim Bech <joakim.bech@linaro.org>
Acked-by: Jens Wiklander <jens.wiklander@linaro.org>

With LOG_LEVEL=50 and DEBUG=1, BL31 becomes too big, with LOG_LEVEL=40
we will miss a few debug prints, but it will at produce a valid binary
and more importantly, you'll get an elf-file with all debug symbols,
i.e., it's GDB makes up for the missing prints.

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

CFG_FLASH_SIZE can be used to generate only linux-4g or linux-8g instead
of both.

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

Adds a patch to be applied by Buildroot when building Xen. This fixes a
race condition in RPC buffer management, which causes random failures in
the OP-TEE CI jobs [1]. The patch is being upstreamed [2].

Link: [1] https://dev.azure.com/OPTEE/optee_os/_build/results?buildId=2827&view=logs&j=705748f3-7146-5e86-79af-1c0266d20a8c&t=d3458a7d-1a81-5639-8e63-8927d807e53a&l=7812
Link: [2] https://lists.xenproject.org/archives/html/xen-devel/2022-05/msg00245.html
Fixes: https://github.com/OP-TEE/optee_os/issues/5288


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

Introduces the option MEMTAG which enabled tagging of memory using the
ARM Memory Tagging Extension (MTE).

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

Updates the matching rules to only trap OP-TEE Core assert or panic.
Panic or assert in a TA may very well be part of a test case.

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

Adds a missing linux dependency for the flash-image target.

This should fix erros like:
  CC [M]  drivers/gpu/drm/nouveau/nvkm/subdev/i2c/bus.o
mformat -i .../build/../out/boot.img.fat -n 64 -h 255 -T 131072 -v "BOOT IMG" -C ::
mcopy -i .../build/../out/boot.img.fat .../build/../linux/arch/arm64/boot/Image ::
.../build/../linux/arch/arm64/boot/Image: No such file or directory

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