boot/optee-os: new package
PATCH v4 1/7 boot/optee-os: new package
OP-TEE OS is maintained by the OP-TEE project. It provides an open source solution for development and integration of secure services for Armv7-A and Armv8-A CPU based platforms supporting the TrustZone technology. This technology enables CPUs to concurrently host a secure world as the OP-TEE OS and a non-secure world as a Linux based OS.
The OP-TEE project maintains other packages to leverage OP-TEE on Linux kernel based OSes. An OP-TEE interface driver is available in the Linux kernel since 4.12 upon CONFIG_OPTEE.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
https://www.op-tee.org/
https://github.com/OP-TEE/optee_os
---
Changes v3 -> v4
- Upgrade from OP-TEE 3.3.0 to 3.4.0. No local patch required.
Changes v2 -> v3
- Add an entry in file DEVELOPERS.
- Fix BR2_ARM_CPU_ARMV7 into BR2_ARM_CPU_ARMV7A and replace dependency on BR2_aarch64 with dep on BR2_ARM_CPU_ARMV8A which is more accurate.
- Many cleaning in description sections of Config.in and layout reordering.
- Reordering in optee-os.mk layout.
- Correct dependency openssl into host-openssl.
- Use OPTEE_OS_INSTALL_STAGING_CMDS for installs in the staging dir.
- Clean in-tree TAs install command (s/@(foreach ...)/$(INSTALL) ...)
- BR2_TARGET_OPTEE_OS_SERVICES selects BR2_TARGET_OPTEE_OS_CORE instead of depending on BR2_TARGET_OPTEE_OS_SDK. This because core build also builds the in-tree TAs.
- Replace common optee-os.hash with per-version .hash files. Support the released 3.x tags from OP-TEE project.
- Remove useless indirection in OPTEE_OS_INSTALL_STAGING_CMDS definition.
- Fix issue of Aarch64 OS to attempt to build Aarch32 user mode support while the selected cross compilation toolchain cannot compile for Aarch32 targets. OP-TEE OS 3.3.0 is patched to backport support for CFG_USER_TA_TARGETS directive, allowing to restrict build to Aarch64.
Changes v1 -> v2:
- Replace dependency on BR2_arm with BR2_ARM_CPU_ARMV7 as BR2_arm
is enabled for non Armv7 targets.
- Correct build dependencies on OpenSSL and pycrypto.
Remove patch on package python scripts since pycrypto dependency
is now handled.
- Correct location of in-tree services TAs (s/ta_services/ta/).
Remvoe OPTEE_OS_BUILD_SERVICES as service TAs are already built built when OP-TEE OS core is built. Correct BR2_TARGET_OPTEE_OS_SERVICES options: it only installs the
- Fix bad reference in Config.in package description.
- Fix wrong hash for the optee-os v3.3.0 tarball.
- Fix bad use of OPTEE_OS_VERSION where it is the value content that
is expected: $(OPTEE_OS_VERSION).
- Clarify output build directory name: use out/.
- Minor replace use if/endif with use of depends on in Config.mk.
- Add missing dependency of BR2_TARGET_OPTEE_OS_SERVICES on BR2_TARGET_OPTEE_OS_SDK.
- Change commit header comment to "boot/optee-os: new package".
--- DEVELOPERS | 3 ++ boot/Config.in | 1 + boot/optee-os/3.4.0/optee-os.hash | 4 ++ boot/optee-os/Config.in | 101 ++++++++++++++++++++++++++++++++++++++ boot/optee-os/optee-os.mk | 97 ++++++++++++++++++++++++++++++++++++ 5 files changed, 206 insertions(+) create mode 100644 boot/optee-os/3.4.0/optee-os.hash create mode 100644 boot/optee-os/Config.in create mode 100644 boot/optee-os/optee-os.mk
DEVELOPERS +3 -0
@@ -681,6 +681,9 @@ F: package/mxsldr/ N: Ernesto L. Williams Jr <[email protected]> F: package/szip/ +N: Etienne Carriere <[email protected]> +F: boot/optee-os/ + N: Eugene Tarassov <[email protected]> F: package/tcf-agent/
boot/Config.in +1 -0
@@ -13,6 +13,7 @@ source "boot/gummiboot/Config.in" source "boot/lpc32xxcdl/Config.in" source "boot/mv-ddr-marvell/Config.in" source "boot/mxs-bootlets/Config.in" +source "boot/optee-os/Config.in" source "boot/riscv-pk/Config.in" source "boot/s500-bootloader/Config.in" source "boot/shim/Config.in"
boot/optee-os/3.4.0/optee-os.hash new file +4 -0
@@ -0,0 +1,4 @@ +# From https://github.com/OP-TEE/optee_os/archive/3.4.0.tar.gz +sha256 51d42ac7aa780ec8d8ee471eff689a29a7621aacace046722b1490b62ec2d481 optee-os-3.4.0.tar.gz +# Locally computed +sha256 fda8385993f112d7ca61b88b54ba5b4cbeec7e43a0f9b317d5186703c1985e8f LICENSE
boot/optee-os/Config.in new file +101 -0
101 lines not on this page · raw message
boot/optee-os/optee-os.mk new file +97 -0
97 lines not on this page · raw message
PATCH v4 2/7 optee-client: new package
OP-TEE client API library and supplicant daemon from the OP-TEE project are packaged in package/optee-client. An init script launches the tee-supplicant deamon. Package is added to the Security menu of BR configuration.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0.
Changes v2 -> v3:
- Add an entry in file DEVELOPERS.
- Clean Config.in layout and description sections.
- Drop BR2_PACKAGE_OPTEE_CLIENT_SYNCED_VERSION.
- Clean optee-client.mk layout.
- Remove OPTEE_CLIENT_INSTALL_STAGING indirection.
- Replace optee-client.hash with per-version optee-client.hash files.
- Support the released 3.x tags from OP-TEE project.
- Correct license tag (client is BSD-2-Clause instead of BSD-3-Clause).
- Rewrite the init script to use start-stop-daemon
Changes v1 -> v2:
- Add option BR2_PACKAGE_OPTEE_CLIENT_SYNCED_VERSION to ensure OP-TEE client version is synced with OP-TEE OS version when the later if enabled.
- Remove useless OPTEE_CLIENT_INSTALL_IMAGE=YES.
--- DEVELOPERS | 1 + package/Config.in | 1 + package/optee-client/3.4.0/optee-client.hash | 4 ++ package/optee-client/Config.in | 55 ++++++++++++++++++++++++++++ package/optee-client/S30optee | 49 +++++++++++++++++++++++++ package/optee-client/optee-client.mk | 25 +++++++++++++ 6 files changed, 135 insertions(+) create mode 100644 package/optee-client/3.4.0/optee-client.hash create mode 100644 package/optee-client/Config.in create mode 100644 package/optee-client/S30optee create mode 100644 package/optee-client/optee-client.mk
DEVELOPERS +1 -0
@@ -683,6 +683,7 @@ F: package/szip/ N: Etienne Carriere <[email protected]> F: boot/optee-os/ +F: package/optee-client/ N: Eugene Tarassov <[email protected]> F: package/tcf-agent/
package/Config.in +1 -0
@@ -2077,6 +2077,7 @@ endmenu menu "Security" source "package/checkpolicy/Config.in" + source "package/optee-client/Config.in" source "package/paxtest/Config.in" source "package/policycoreutils/Config.in" source "package/refpolicy/Config.in"
package/optee-client/3.4.0/optee-client.hash new file +4 -0
@@ -0,0 +1,4 @@ +# From https://github.com/OP-TEE/optee_client/archive/3.4.0.tar.gz +sha256 7c131a44f4d73acb8816fba88bcd7e6a18537f0a522ae426a20e2ca7dc46a6ec optee-client-3.4.0.tar.gz +# Locally computed +sha256 fda8385993f112d7ca61b88b54ba5b4cbeec7e43a0f9b317d5186703c1985e8f LICENSE
package/optee-client/Config.in new file +55 -0
55 lines not on this page · raw message
package/optee-client/S30optee new file +49 -0
49 lines not on this page · raw message
package/optee-client/optee-client.mk new file +25 -0
@@ -0,0 +1,25 @@ +################################################################################ +# +# optee-client +# +################################################################################ + +OPTEE_CLIENT_VERSION = $(call qstrip,$(BR2_PACKAGE_OPTEE_CLIENT_VERSION)) +OPTEE_CLIENT_LICENSE = BSD-2-Clause +OPTEE_CLIENT_LICENSE_FILES = LICENSE +OPTEE_CLIENT_INSTALL_STAGING = YES + +ifeq ($(BR2_PACKAGE_OPTEE_CLIENT_CUSTOM_GIT),y) +OPTEE_CLIENT_SITE = $(call qstrip,$(BR2_PACKAGE_OPTEE_CLIENT_CUSTOM_REPO_URL)) +OPTEE_CLIENT_SITE_METHOD = git +BR_NO_CHECK_HASH_FOR += $(OPTEE_CLIENT_SOURCE) +else +OPTEE_CLIENT_SITE = $(call github,OP-TEE,optee_client,$(OPTEE_CLIENT_VERSION)) +endif + +define OPTEE_CLIENT_INSTALL_INIT_SYSV + $(INSTALL) -m 0755 -D $(OPTEE_CLIENT_PKGDIR)/S30optee \ + $(TARGET_DIR)/etc/init.d/S30optee +endef + +$(eval $(cmake-package))
PATCH v4 2/7 optee-client: new package
Hello ?tienne,
I have applied to the next branch after a few changes (see below), but there are some build failures with some configurations, which will have to be fixed. I'm giving more details below as well.
Quoted 16 lines · On Wed, 30 Jan 2019 11:47:24 +0100 Etienne Carriere…
package/optee-client/3.4.0/optee-client.hash new file +4
-0@@ -0,0 +1,4 @@ +# From https://github.com/OP-TEE/optee_client/archive/3.4.0.tar.gz +sha256 7c131a44f4d73acb8816fba88bcd7e6a18537f0a522ae426a20e2ca7dc46a6ec optee-client-3.4.0.tar.gz +# Locally computed +sha256 fda8385993f112d7ca61b88b54ba5b4cbeec7e43a0f9b317d5186703c1985e8f LICENSEpackage/optee-client/Config.in new file +0
-0
package/optee-client/Config.in line 2
Quoted 2 lines
package/optee-client/Config.in new file +2
-0@@ -0,0 +1,55 @@ +config BR2_PACKAGE_OPTEE_CLIENT + bool "optee-client"
The following dependency:
depends on !BR2_STATIC_LIBS
is needed, because optee-client tries to unconditionally build a shared library, instead of obeying to the standard BUILD_SHARED_LIBS/BUILD_STATIC_LIBS CMake variables.
Quoted 10 lines
package/optee-client/Config.in new file +10
-0+if BR2_PACKAGE_OPTEE_CLIENT + +choice + prompt "version" + default BR2_PACKAGE_OPTEE_CLIENT_LATEST + help + Select the version of OP-TEE client you want to use + +config BR2_PACKAGE_OPTEE_CLIENT_LATEST + bool "3.4.0"
package/optee-client/Config.in line 34
Quoted 10 lines
package/optee-client/Config.in new file +10
-0+ help + This fetches the registered release tag from the + OP-TEE official Git repository. + +config BR2_PACKAGE_OPTEE_CLIENT_CUSTOM_GIT + bool "Custom Git repository" + help + Sync with a specific OP-TEE Git repository. + +endchoice
I don't really see why a version selection was needed, so I've dropped it entirely. Is there anything really HW/SoC specific in optee-client ?
There are two build failures that I noticed.
With the following defconfig:
BR2_arm=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_TOOLCHAIN_EXTERNAL_CUSTOM=y
BR2_TOOLCHAIN_EXTERNAL_DOWNLOAD=y BR2_TOOLCHAIN_EXTERNAL_URL="http://autobuild.buildroot.org/toolchains/tarballs/armv5-ctng-linux-gnueabi.tar.xz" BR2_TOOLCHAIN_EXTERNAL_CUSTOM_PREFIX="armv5-ctng-linux-gnueabi" BR2_TOOLCHAIN_EXTERNAL_GCC_4_8=y
BR2_TOOLCHAIN_EXTERNAL_HEADERS_3_12=y
BR2_TOOLCHAIN_EXTERNAL_CUSTOM_GLIBC=y
BR2_TOOLCHAIN_EXTERNAL_CXX=y
BR2_INIT_NONE=y
BR2_SYSTEM_BIN_SH_NONE=y
# BR2_PACKAGE_BUSYBOX is not set
BR2_PACKAGE_OPTEE_CLIENT=y
# BR2_TARGET_ROOTFS_TAR is not set
You will get:
/home/thomas/projets/outputs/armv5-ctng-linux-gnueabi/build/optee-client-3.4.0/libteec/src/tee_client_api.c: In function 'TEEC_InitializeContext':/home/thomas/projets/outputs/armv5-ctng-linux-gnueabi/build/optee-client-3.4.0/libteec/src/tee_client_api.c:149:28: error: 'gen_caps' may be used uninitialized in this function [-Werror=maybe-uninitialized]
ctx->reg_mem = gen_caps & TEE_GEN_CAP_REG_MEM;
^cc1: all warnings being treated as errors
With this other defconfig:
BR2_or1k=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_TOOLCHAIN_EXTERNAL_DOWNLOAD=y BR2_TOOLCHAIN_EXTERNAL_URL="http://autobuild.buildroot.net/toolchains/tarballs/br-openrisc-full-2018.02-925-gd4158df.tar.bz2" BR2_TOOLCHAIN_EXTERNAL_GCC_5=y
BR2_TOOLCHAIN_EXTERNAL_HEADERS_4_16=y
BR2_TOOLCHAIN_EXTERNAL_LOCALE=y
# BR2_TOOLCHAIN_EXTERNAL_HAS_THREADS_DEBUG is not set
BR2_TOOLCHAIN_EXTERNAL_CXX=y
BR2_INIT_NONE=y
BR2_SYSTEM_BIN_SH_NONE=y
# BR2_PACKAGE_BUSYBOX is not set
BR2_PACKAGE_OPTEE_CLIENT=y
# BR2_TARGET_ROOTFS_TAR is not set
You will get this failure:
/home/thomas/projets/buildroot/output/build/optee-client-3.4.0/libteec/src/tee_client_api.c: In function ?TEEC_OpenSession?:/home/thomas/projets/buildroot/output/build/optee-client-3.4.0/libteec/src/tee_client_api.c:507:8: error: cast increases required alignment of target type [-Werror=cast-align]
arg = (struct tee_ioctl_open_session_arg *)buf;
^
/home/thomas/projets/buildroot/output/build/optee-client-3.4.0/libteec/src/tee_client_api.c: In function ?TEEC_InvokeCommand?:
/home/thomas/projets/buildroot/output/build/optee-client-3.4.0/libteec/src/tee_client_api.c:581:8: error: cast increases required alignment of target type [-Werror=cast-align]
arg = (struct tee_ioctl_invoke_arg *)buf;
^cc1: all warnings being treated as errors
Since I wanted to move forward with merging the OP-TEE support, I applied this patch despite these build failures (a number of other configurations are building fine), but we will need those build issues to be resolved.
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 3/7 optee-examples: new package
In reply to Etienne CarriereThis package generates embedded Linux based OS userland client applications and OP-TEE OS trusted applications all embedded in the file system. These applications shows how to use the APIs OP-TEE OS is based on, both in the non secure and secure worlds.
Package is added next to the OP-TEE client package in the BR package configuration.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0 with an added patch to fix an issue reported by recent GCC toolchains.
---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0.
Changes v2 -> v3:
- Add an entry in file DEVELOPERS.
- Clean Config.in layout and description sections.
- Drop BR2_PACKAGE_OPTEE_EXAMPLES_SYNCED_VERSION.
- Clean optee-examples.mk layout.
- Replace common optee-examples.hash with per-version .hash files.
- Simplify install command in iotee-test.mk
- Patch optee_test 3.3.0 against an issue in the CMake files.
- Remove dependency of BR2_PACKAGE_OPTEE_EXAMPLES. As the package depends on BR2_TARGET_OPTEE_OS, leave it to optee-os to define the supported architectures.
Changes v1 -> v2:
- Replace BR2_arm with BR2_ARM_CPU_ARMV7 as OP-TEE supports only BR2_ARM_CPU_ARMV7 architectures among the 32bit Arm machines.
- Select OP-TEE client and add dependency on OP-TEE OS.
- Add option BR2_PACKAGE_OPTEE_EXAMPLES_SYNCED_VERSION to ensure OP-TEE examples version is synced with OP-TEE OS version.
- Do not force output build directory, rely on native path: out/.
- Replace if/endif with depends on in Config.in.
- Remove useless OPTEE_EXAMPLES_INSTALL_STAGING=YES.
- Add package official URL in Config.in package description.
--- DEVELOPERS | 1 + package/Config.in | 1 + .../0001-fix-deprecated-size_t-type-for-size.patch | 37 ++++++++++++++ package/optee-examples/3.4.0/optee-examples.hash | 4 ++ package/optee-examples/Config.in | 59 ++++++++++++++++++++++ package/optee-examples/optee-examples.mk | 44 ++++++++++++++++ 6 files changed, 146 insertions(+) create mode 100644 package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch create mode 100644 package/optee-examples/3.4.0/optee-examples.hash create mode 100644 package/optee-examples/Config.in create mode 100644 package/optee-examples/optee-examples.mk
DEVELOPERS +1 -0
@@ -684,6 +684,7 @@ F: package/szip/ N: Etienne Carriere <[email protected]> F: boot/optee-os/ F: package/optee-client/ +F: package/optee-examples/ N: Eugene Tarassov <[email protected]> F: package/tcf-agent/
package/Config.in +1 -0
@@ -2078,6 +2078,7 @@ endmenu menu "Security" source "package/checkpolicy/Config.in" source "package/optee-client/Config.in" + source "package/optee-examples/Config.in" source "package/paxtest/Config.in" source "package/policycoreutils/Config.in" source "package/refpolicy/Config.in"
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch new file +37 -0
@@ -0,0 +1,37 @@ +commit fca0caba591f31f66325826c74acd26aefd52dee +Author: Etienne Carriere <[email protected]> +Date: Tue Dec 18 22:35:16 2018 +0100 + + secure_storage: fix deprecated size_t type for size + + size_t types is an deprecated type used in GPD API v1.0. + Update + + Error reported by GCC 7.3-2018.05: + secure_storage_ta.c:203:6: warning: passing argument 4 of 'TEE_ReadObjectData' from incompatible pointer type [-Wincompatible-pointer-types] + &read_bytes); + + Signed-off-by: Etienne Carriere <[email protected]> + +diff --git a/secure_storage/ta/secure_storage_ta.c b/secure_storage/ta/secure_storage_ta.c +index d120e47..3ccc12d 100644 +--- a/secure_storage/ta/secure_storage_ta.c ++++ b/secure_storage/ta/secure_storage_ta.c +@@ -146,7 +146,7 @@ static TEE_Result read_raw_object(uint32_t param_types, TEE_Param params[4]) + TEE_ObjectHandle object; + TEE_ObjectInfo object_info; + TEE_Result res; +- size_t read_bytes; ++ uint32_t read_bytes; + char *obj_id; + size_t obj_id_sz; + char *data; +@@ -202,7 +202,7 @@ static TEE_Result read_raw_object(uint32_t param_types, TEE_Param params[4]) + res = TEE_ReadObjectData(object, data, object_info.dataSize, + &read_bytes); + if (res != TEE_SUCCESS || read_bytes != object_info.dataSize) { +- EMSG("TEE_ReadObjectData failed 0x%08x, read %u over %u", ++ EMSG("TEE_ReadObjectData failed 0x%08x, read %" PRIu32 " over %u", + res, read_bytes, object_info.dataSize); + goto exit; + }
package/optee-examples/3.4.0/optee-examples.hash new file +4 -0
@@ -0,0 +1,4 @@ +# From https://github.com/linaro-swg/optee_examples/archive/3.4.0.tar.gz +sha256 d833753980ac438c1675787857bb8352997352212334274de9419770097ce039 optee-examples-3.4.0.tar.gz +# Locally computed +sha256 6f1ef8449cb82ae79d2155605f7985bdf0f08e7ab5007de9b4362e8bf28733b9 LICENSE
package/optee-examples/Config.in new file +59 -0
59 lines not on this page · raw message
package/optee-examples/optee-examples.mk new file +44 -0
44 lines not on this page · raw message
PATCH v4 3/7 optee-examples: new package
Hello ?tienne,
Quoted 5 lines · On Wed, 30 Jan 2019 11:47:25 +0100 Etienne Carriere…
This package generates embedded Linux based OS userland client applications and OP-TEE OS trusted applications all embedded in the file system. These applications shows how to use the APIs OP-TEE OS is based on, both in the non secure and secure worlds.
Quoted 8 lines
Package is added next to the OP-TEE client package in the BR package configuration.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0 with an added patch to fix an issue reported by recent GCC toolchains.
Signed-off-by Etienne Carriere <[email protected]>
I have applied to next, but with a number of changes, see below.
Quoted 10 lines
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch new file +4
-0@@ -0,0 +1,37 @@ +commit fca0caba591f31f66325826c74acd26aefd52dee +Author: Etienne Carriere <[email protected]> +Date: Tue Dec 18 22:35:16 2018 +0100 +
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch line 14
Quoted 10 lines
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch new file +10
-0+ secure_storage: fix deprecated size_t type for size + + size_t types is an deprecated type used in GPD API v1.0. + Update + + Error reported by GCC 7.3-2018.05: + secure_storage_ta.c:203:6: warning: passing argument 4 of 'TEE_ReadObjectData' from incompatible pointer type [-Wincompatible-pointer-types] + &read_bytes); + + Signed-off-by: Etienne Carriere <[email protected]>
Patches should be generated with "git format-patch" and not "git show". Indeed, patches generated by "git format-patch" can conveniently be applied by "git am".
I've re-generated the patch to use the proper format.
Quoted 16 lines
package/optee-examples/3.4.0/optee-examples.hash new file +4
-0@@ -0,0 +1,4 @@ +# From https://github.com/linaro-swg/optee_examples/archive/3.4.0.tar.gz +sha256 d833753980ac438c1675787857bb8352997352212334274de9419770097ce039 optee-examples-3.4.0.tar.gz +# Locally computed +sha256 6f1ef8449cb82ae79d2155605f7985bdf0f08e7ab5007de9b4362e8bf28733b9 LICENSEpackage/optee-examples/Config.in new file +0
-0
package/optee-examples/Config.in line 3
Quoted 3 lines
package/optee-examples/Config.in new file +3
-0@@ -0,0 +1,59 @@ +config BR2_PACKAGE_OPTEE_EXAMPLES + bool "optee-examples" + depends on BR2_TARGET_OPTEE_OS
In fact, building OP-TEE itself is not sufficient, because this does not guarantee the SDK has been built, so I have added a:
select BR2_TARGET_OPTEE_OS_SDK
here.
package/optee-examples/Config.in line 4
Quoted 1 line
package/optee-examples/Config.in new file +1
-0+ select BR2_PACKAGE_OPTEE_CLIENT
This now depends on !BR2_STATIC_LIBS, so I've propagated this dependency.
Quoted 24 lines
package/optee-examples/Config.in new file +24
-0+ help + Enable the OP-TEE examples package that brings examples of + implementation of OP-TEE non-secure client applications and + secure trusted applications. OP-TEE examples is a + component delivered by the OP-TEE project. + + Trusted application binary files are installed in the target + directory /lib/optee_armtz as other trusted applications. + At runtime OP-TEE OS can load trusted applications from this + non-secure filesystem/directory into the secure world for + execution. + + https://github.com/linaro-swg/optee_examples + +if BR2_PACKAGE_OPTEE_EXAMPLES + +choice + prompt "version" + default BR2_PACKAGE_OPTEE_EXAMPLES_LATEST + help + Select the version of OP-TEE exmaples you want to use + +config BR2_PACKAGE_OPTEE_EXAMPLES_LATEST + bool "3.4.0"
package/optee-examples/Config.in line 38
Quoted 10 lines
package/optee-examples/Config.in new file +10
-0+ help + This fetches the registered release tag from the + OP-TEE official Git repository. + +config BR2_PACKAGE_OPTEE_EXAMPLES_CUSTOM_GIT + bool "Custom Git repository" + help + Sync with a specific OP-TEE Git repository. + +endchoice
Here as well, I don't think the version selection makes a lot of sense, so I dropped it.
package/optee-examples/Config.in line 42
Quoted 4 lines
package/optee-examples/Config.in new file +4
-0+ifeq ($(BR2_aarch64),y) +OPTEE_EXAMPLES_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm64 +endif +ifeq ($(BR2_arm),y)
Changed:
endif
ifeq ($(BR2_arm),y)
to just:
else ifeq ($(BR2_arm),y)
package/optee-examples/Config.in line 53
Quoted 11 lines
package/optee-examples/Config.in new file +11
-0+OPTEE_EXAMPLES_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm32 +endif + +# Trusted Application are not built from CMake due to ta_dev_kit dependencies. +# We must build and install them on target. +define OPTEE_EXAMPLES_BUILD_TAS + @$(foreach f,$(wildcard $(@D)/*/ta/Makefile), \ + $(TARGET_CONFIGURE_OPTS) \ + $(MAKE) CROSS_COMPILE=$(TARGET_CROSS) \ + TA_DEV_KIT_DIR=$(OPTEE_EXAMPLES_SDK) \ + O=out -C $(dir $f) all &&) true
Convoluted syntax here. Changed to:
$(foreach f,$(wildcard $(@D)/*/ta/Makefile), \ $(TARGET_CONFIGURE_OPTS) \ $(MAKE) CROSS_COMPILE=$(TARGET_CROSS) \ TA_DEV_KIT_DIR=$(OPTEE_EXAMPLES_SDK) \ O=out -C $(dir $f) all )
Also: could this be fixed upstream, so that the CMake build system builds/installs everything ?
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 3/7 optee-examples: new package
Hello Thomas,
Quoted 48 lines · Thomas Petazzoni
Hello ?tienne,
Quoted 5 lines · On Wed, 30 Jan 2019 11:47:25 +0100 Etienne Carriere…
This package generates embedded Linux based OS userland client applications and OP-TEE OS trusted applications all embedded in the file system. These applications shows how to use the APIs OP-TEE OS is based on, both in the non secure and secure worlds.
Quoted 8 lines
Package is added next to the OP-TEE client package in the BR package configuration.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0 with an added patch to fix an issue reported by recent GCC toolchains.
Signed-off-by Etienne Carriere <[email protected]>I have applied to next, but with a number of changes, see below.
Quoted 10 lines
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch new file +4
-0@@ -0,0 +1,37 @@ +commit fca0caba591f31f66325826c74acd26aefd52dee +Author: Etienne Carriere <[email protected]> +Date: Tue Dec 18 22:35:16 2018 +0100 +package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch line 14
Quoted 10 lines
package/optee-examples/3.4.0/0001-fix-deprecated-size_t-type-for-size.patch new file +10
-0+ secure_storage: fix deprecated size_t type for size + + size_t types is an deprecated type used in GPD API v1.0. + Update + + Error reported by GCC 7.3-2018.05: + secure_storage_ta.c:203:6: warning: passing argument 4 of 'TEE_ReadObjectData' from incompatible pointer type [-Wincompatible-pointer-types] + &read_bytes); + + Signed-off-by: Etienne Carriere <[email protected]>Patches should be generated with "git format-patch" and not "git show". Indeed, patches generated by "git format-patch" can conveniently be applied by "git am".
Quoted 1 line
I've re-generated the patch to use the proper format.
Ok, sorry, I'll be careful for the next ones.
Quoted 62 lines
Quoted 17 lines
package/optee-examples/3.4.0/optee-examples.hash new file +4
-0@@ -0,0 +1,4 @@ +# From https://github.com/linaro-swg/optee_examples/archive/3.4.0.tar.gz +sha256 d833753980ac438c1675787857bb8352997352212334274de9419770097ce039 optee-examples-3.4.0.tar.gz +# Locally computed +sha256 6f1ef8449cb82ae79d2155605f7985bdf0f08e7ab5007de9b4362e8bf28733b9 LICENSEpackage/optee-examples/Config.in new file +0
-0package/optee-examples/Config.in line 3
Quoted 3 lines
package/optee-examples/Config.in new file +3
-0@@ -0,0 +1,59 @@ +config BR2_PACKAGE_OPTEE_EXAMPLES + bool "optee-examples" + depends on BR2_TARGET_OPTEE_OSIn fact, building OP-TEE itself is not sufficient, because this does not guarantee the SDK has been built, so I have added a:
select
BR2_TARGET_OPTEE_OS_SDKhere.
package/optee-examples/Config.in line 4
Quoted 1 line
package/optee-examples/Config.in new file +1
-0+ select BR2_PACKAGE_OPTEE_CLIENTThis now depends on !BR2_STATIC_LIBS, so I've propagated this dependency.
Quoted 29 lines
package/optee-examples/Config.in new file +29
-0+ help + Enable the OP-TEE examples package that brings examples of + implementation of OP-TEE non-secure client applications and + secure trusted applications. OP-TEE examples is a + component delivered by the OP-TEE project. + + Trusted application binary files are installed in the target + directory /lib/optee_armtz as other trusted applications. + At runtime OP-TEE OS can load trusted applications from this + non-secure filesystem/directory into the secure world for + execution. + + https://github.com/linaro-swg/optee_examples + +if BR2_PACKAGE_OPTEE_EXAMPLES + +choice + prompt "version" + default BR2_PACKAGE_OPTEE_EXAMPLES_LATEST + help + Select the version of OP-TEE exmaples you want to use + +config BR2_PACKAGE_OPTEE_EXAMPLES_LATEST + bool "3.4.0" + help + This fetches the registered release tag from the + OP-TEE official Git repository. + +config BR2_PACKAGE_OPTEE_EXAMPLES_CUSTOM_GIT
Quoted 8 lines
package/optee-examples/Config.in line 38
Quoted 5 lines
package/optee-examples/Config.in new file +5
-0+ bool "Custom Git repository" + help + Sync with a specific OP-TEE Git repository. + +endchoiceHere as well, I don't think the version selection makes a lot of sense, so I dropped it.
Fair.
Quoted 37 lines
package/optee-examples/Config.in line 42
Quoted 6 lines
package/optee-examples/Config.in new file +4
-0+ifeq ($(BR2_aarch64),y) +OPTEE_EXAMPLES_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm64 +endif +ifeq ($(BR2_arm),y)Changed:
endif
ifeq ($(BR2_arm),y)to just:
else ifeq ($(BR2_arm),y)
package/optee-examples/Config.in line 53
Quoted 11 lines
package/optee-examples/Config.in new file +11
-0+OPTEE_EXAMPLES_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm32 +endif + +# Trusted Application are not built from CMake due to ta_dev_kit dependencies. +# We must build and install them on target. +define OPTEE_EXAMPLES_BUILD_TAS + @$(foreach f,$(wildcard $(@D)/*/ta/Makefile), \ + $(TARGET_CONFIGURE_OPTS) \ + $(MAKE) CROSS_COMPILE=$(TARGET_CROSS) \ + TA_DEV_KIT_DIR=$(OPTEE_EXAMPLES_SDK) \ + O=out -C $(dir $f) all &&) trueConvoluted syntax here. Changed to:
$(foreach f,$(wildcard $(@D)/*/ta/Makefile), \ $(TARGET_CONFIGURE_OPTS) \ $(MAKE) CROSS_COMPILE=$(TARGET_CROSS) \ TA_DEV_KIT_DIR=$(OPTEE_EXAMPLES_SDK) \ O=out -C $(dir $f) all )
Quoted 2 lines
Also: could this be fixed upstream, so that the CMake build system builds/installs everything ?
That a tricky part. The binaries that executes in the secure world, that is Trusted Application, i.e those built from the optee-test package as well as the OP-TEE OS core (optee-os) uses build sequence written with GNU Makefile that we (I mean people contributing to op-tee) did not manage yet to port to CMake. This is an open task and it would greatly help integration of OP-TEE in many build env, but it is not available yet. Sorry
Best regards,
etienne
Quoted 8 lines
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
PATCH v4 4/7 optee-test: new package
In reply to Etienne CarriereOP-TEE test package provide test materials as part of the OP-TEE project helping platforms to verify their OP-TEE components against a set of regression and performance tests.
Package is added in the BR package configuration next to the OP-TEE client package.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0 with an added patch to fix an issue reported by recent GCC toolchains.
---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. Local patches for 3.3.0 are not applicable. Add a local patch to fix a loop optimization issue reported by recent GCC.
Changes v2 -> v3:
- Add an entry in file DEVELOPERS.
- Clean Config.in layout and description sections.
- Drop BR2_PACKAGE_OPTEE_TEST_SYNCED_VERSION.
- Clean optee-test.mk layout.
- Replace common optee-test.hash with per-version .hash files.
- Patch optee_benchmark 3.3.0 against an issue reported by GCC warns.
- Remove dependency of BR2_PACKAGE_OPTEE_EXAMPLES and Arm architecture. As the package depends on BR2_TARGET_OPTEE_OS, leave it to optee-os to define the supported architectures.
Changes v1 -> v2:
- Replace BR2_arm with BR2_ARM_CPU_ARMV7 as OP-TEE supports only BR2_ARM_CPU_ARMV7 architectures among the 32bit Arm machines.
- Add missing dependency on BR2_TARGET_OPTEE_OS and select BR2_PACKAGE_OPTEE_CLIENT when enabled.
- Add option BR2_PACKAGE_OPTEE_TEST_SYNCED_VERSION to ensure OP-TEE test version is synced with OP-TEE OS version.
- Fix official repo URL in Config.in package description.
- Remove useless OPTEE_TEST_INSTALL_STAGING=YES.
- Do not force output build directory and rely on native one: out/.
--- DEVELOPERS | 1 + package/Config.in | 1 + ...-regression-4100-update-string-conversion.patch | 67 ++++++++++++++++++++++ package/optee-test/3.4.0/optee-test.hash | 4 ++ package/optee-test/Config.in | 63 ++++++++++++++++++++ package/optee-test/optee-test.mk | 45 +++++++++++++++ 6 files changed, 181 insertions(+) create mode 100644 package/optee-test/3.4.0/0001-regression-4100-update-string-conversion.patch create mode 100644 package/optee-test/3.4.0/optee-test.hash create mode 100644 package/optee-test/Config.in create mode 100644 package/optee-test/optee-test.mk
DEVELOPERS +1 -0
@@ -685,6 +685,7 @@ N: Etienne Carriere <[email protected]> F: boot/optee-os/ F: package/optee-client/ F: package/optee-examples/ +F: package/optee-test/ N: Eugene Tarassov <[email protected]> F: package/tcf-agent/
package/Config.in +1 -0
@@ -2079,6 +2079,7 @@ menu "Security" source "package/checkpolicy/Config.in" source "package/optee-client/Config.in" source "package/optee-examples/Config.in" + source "package/optee-test/Config.in" source "package/paxtest/Config.in" source "package/policycoreutils/Config.in" source "package/refpolicy/Config.in"
package/optee-test/3.4.0/0001-regression-4100-update-string-conversion.patch new file +67 -0
67 lines not on this page · raw message
package/optee-test/3.4.0/optee-test.hash new file +4 -0
@@ -0,0 +1,4 @@ +# From https://github.com/OP-TEE/optee_test/archive/3.4.0.tar.gz +sha256 755904c5b845763a2460c32c21100a57c713009b6b88cc3fc21f0e5be8645e2b optee-test-3.4.0.tar.gz +# Locally computed +sha256 6e6810981f0ddab9e0d44399d0700a15d9f760a3c2843cc866659c2074139ae7 LICENSE.md
package/optee-test/Config.in new file +63 -0
63 lines not on this page · raw message
package/optee-test/optee-test.mk new file +45 -0
45 lines not on this page · raw message
Thomas Petazzoni
AppliedPATCH v4 4/7 optee-test: new package
Quoted 11 lines · On Wed, 30 Jan 2019 11:47:26 +0100 Etienne Carriere…
OP-TEE test package provide test materials as part of the OP-TEE project helping platforms to verify their OP-TEE components against a set of regression and performance tests.
Package is added in the BR package configuration next to the OP-TEE client package.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0 with an added patch to fix an issue reported by recent GCC toolchains.
Quoted 7 lines
Signed-off-by Etienne Carriere <[email protected]>---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. Local patches for 3.3.0 are not applicable. Add a local patch to fix a loop optimization issue reported by recent GCC.
Applied to next with the same changes as for optee-examples:
[Thomas:
- drop version selection
- propagate !BR2_STATIC_LIBS dependency of optee-client
- make sure BR2_TARGET_OPTEE_OS_SDK is selected
- use a patch generated by git format-patch
- simplify the construct to build the examples]
Best regards,
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 4/7 optee-test: new package
In reply to Etienne CarriereHello,
package/optee-test/optee-test.mk line 26
Quoted 6 lines · On Wed, 30 Jan 2019 11:47:26 +0100 Etienne Carriere…
package/optee-test/optee-test.mk new file +6
-0+ifeq ($(BR2_aarch64),y) +OPTEE_TEST_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm64 +endif +ifeq ($(BR2_arm),y) +OPTEE_TEST_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm32 +endif
Second thought about this: this the SDK is installed by OPTEE-OS, it should perhaps be boot/optee-os/optee-os.mk that provides a variable pointing to the SDK. It can then be used by all packages that use the SDK instead of duplicating this OPTEE-OS specific knowledge in all users of the SDK.
Could you fix this by a set of follow-up patches ?
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 4/7 optee-test: new package
Quoted 19 lines · Thomas Petazzoni
Hello,
package/optee-test/optee-test.mk line 26
Quoted 6 lines · On Wed, 30 Jan 2019 11:47:26 +0100 Etienne Carriere…
package/optee-test/optee-test.mk new file +6
-0+ifeq ($(BR2_aarch64),y) +OPTEE_TEST_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm64 +endif +ifeq ($(BR2_arm),y) +OPTEE_TEST_SDK = $(STAGING_DIR)/lib/optee/export-ta_arm32 +endifSecond thought about this: this the SDK is installed by OPTEE-OS, it should perhaps be
boot/optee-os/optee-os.mkthat provides a variable pointing to the SDK. It can then be used by all packages that use the SDK instead of duplicating this OPTEE-OS specific knowledge in all users of the SDK.
Quoted 1 line
Could you fix this by a set of follow-up patches ?
Ok, will do. Indeed would look better.
Regards,
Etienne
Quoted 8 lines
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
PATCH v4 5/7 optee-benchmark: new package
In reply to Etienne CarriereOP-TEE performance benchmark tools for the OP-TEE project.
This packages generates embedded Linux based OS materials used to retrieve execution timing information on invocation of the OP-TEE secure services.
It is added next to the OP-TEE client package in BR configuration.
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. No local patch required.
Changes v2 -> v3:
- Add an entry in file DEVELOPERS.
- Clean Config.in layout and description sections.
- Drop BR2_PACKAGE_OPTEE_BENCHMARK_SYNCED_VERSION.
- Clean optee-benchmark.mk layout.
- Replace common optee-benchmark.hash with per-version .hash files.
- Add a patch on 3.3.0 to fix an issue reported through GCC warns.
Changes v1 -> v2:
- Add dependency on OP-TEE client.
- Add option BR2_PACKAGE_OPTEE_BENCHMARK_SYNCED_VERSION to ensure OP-TEE benchmark version is synced with OP-TEE client version.
- Remove useless OPTEE_BENCHMARK_INSTALL_STAGING and OPTEE_BENCHMARK_INSTALL_IMAGES.
- Remove unused BR2_PACKAGE_OPTEE_BENCHMARK_GIT_REFERENCE.
- Remove useless _INSTALL_STAGING/_INSTALL_IMAGES=YES.
--- DEVELOPERS | 1 + package/Config.in | 1 + package/optee-benchmark/3.4.0/optee-benchmark.hash | 2 + package/optee-benchmark/Config.in | 55 ++++++++++++++++++++++ package/optee-benchmark/optee-benchmark.mk | 22 +++++++++ 5 files changed, 81 insertions(+) create mode 100644 package/optee-benchmark/3.4.0/optee-benchmark.hash create mode 100644 package/optee-benchmark/Config.in create mode 100644 package/optee-benchmark/optee-benchmark.mk
DEVELOPERS +1 -0
@@ -683,6 +683,7 @@ F: package/szip/ N: Etienne Carriere <[email protected]> F: boot/optee-os/ +F: package/optee-benchmark/ F: package/optee-client/ F: package/optee-examples/ F: package/optee-test/
package/Config.in +1 -0
@@ -2077,6 +2077,7 @@ endmenu menu "Security" source "package/checkpolicy/Config.in" + source "package/optee-benchmark/Config.in" source "package/optee-client/Config.in" source "package/optee-examples/Config.in" source "package/optee-test/Config.in"
package/optee-benchmark/3.4.0/optee-benchmark.hash new file +2 -0
@@ -0,0 +1,2 @@ +# From https://github.com/linaro-swg/optee_benchmark/archive/3.4.0.tar.gz +sha256 e5e868a06a9dcc8cc444b3e72c65f57670b0811091be62edbe0d03d13c75e716 optee-benchmark-3.4.0.tar.gz
package/optee-benchmark/Config.in new file +55 -0
55 lines not on this page · raw message
package/optee-benchmark/optee-benchmark.mk new file +22 -0
@@ -0,0 +1,22 @@ +################################################################################ +# +# optee-benchmarch +# +################################################################################ + +OPTEE_BENCHMARK_VERSION = $(call qstrip,$(BR2_PACKAGE_OPTEE_BENCHMARK_VERSION)) +OPTEE_BENCHMARK_LICENSE = BSD-2-Clause + +OPTEE_BENCHMARK_DEPENDENCIES = optee-client libyaml + +ifeq ($(BR2_PACKAGE_OPTEE_BENCHMARK_LATEST),y) +OPTEE_BENCHMARK_SITE = $(call github,linaro-swg,optee_benchmark,$(OPTEE_BENCHMARK_VERSION)) +endif + +ifeq ($(BR2_PACKAGE_OPTEE_BENCHMARK_CUSTOM_GIT),y) +OPTEE_BENCHMARK_SITE = $(call qstrip,$(BR2_PACKAGE_OPTEE_BENCHMARK_CUSTOM_REPO_URL)) +OPTEE_BENCHMARK_SITE_METHOD = git +BR_NO_CHECK_HASH_FOR += $(OPTEE_BENCHMARK_SOURCE) +endif + +$(eval $(cmake-package))
PATCH v4 5/7 optee-benchmark: new package
Hello,
Quoted 8 lines · On Wed, 30 Jan 2019 11:47:27 +0100 Etienne Carriere…
OP-TEE performance benchmark tools for the OP-TEE project.
This packages generates embedded Linux based OS materials used to retrieve execution timing information on invocation of the OP-TEE secure services.
It is added next to the OP-TEE client package in BR configuration.
Quoted 8 lines
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
Signed-off-by Etienne Carriere <[email protected]>---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. No local patch required.
I have applied to next with the following changes:
[Thomas:
- drop version selection
- propagate the dependency of optee-client]
Another thing I noticed is that your .hash files do not have the hashes for the license files. Could you send a patch adding those ?
Thanks,
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 5/7 optee-benchmark: new package
Quoted 23 lines · Thomas Petazzoni
Hello,
Quoted 8 lines · On Wed, 30 Jan 2019 11:47:27 +0100 Etienne Carriere…
OP-TEE performance benchmark tools for the OP-TEE project.
This packages generates embedded Linux based OS materials used to retrieve execution timing information on invocation of the OP-TEE secure services.
It is added next to the OP-TEE client package in BR configuration.
Quoted 8 lines
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
Signed-off-by Etienne Carriere <[email protected]>---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. No local patch required.
Quoted 8 lines
I have applied to next with the following changes:
[Thomas:
- drop version selection
- propagate the dependency of optee-client]Another thing I noticed is that your .hash files do not have the hashes for the license files. Could you send a patch adding those ?
Sure, will do.
Thanks,
etienne
Quoted 8 lines
Thanks,
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
PATCH v4 5/7 optee-benchmark: new package
Quoted 28 lines · Etienne Carriere
Quoted 25 lines · Thomas Petazzoni
Hello,
Quoted 8 lines · On Wed, 30 Jan 2019 11:47:27 +0100 Etienne Carriere…
OP-TEE performance benchmark tools for the OP-TEE project.
This packages generates embedded Linux based OS materials used to retrieve execution timing information on invocation of the OP-TEE secure services.
It is added next to the OP-TEE client package in BR configuration.
Quoted 8 lines
This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
Signed-off-by Etienne Carriere <[email protected]>---
Changes v3 -> v4:
- Upgrade from OP-TEE release 3.3.0 to 3.4.0. No local patch required.I have applied to next with the following changes:
Quoted 8 lines
Quoted 6 lines
[Thomas:
- drop version selection
- propagate the dependency of optee-client]Another thing I noticed is that your .hash files do not have the hashes for the license files. Could you send a patch adding those ?
Sure, will do.
For info, the package currently lacks of LICENSE file. This should be fixed in next OP-TEE release 3.5.0. Hash file will be updated when bumping to that version.
Regards,
etienne
Quoted 12 lines
Thanks,
etienneQuoted 8 lines
Thanks,
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
In reply to Etienne CarriereThis change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
The target Linux based normal world embeds the standard minimal filesystem with OP-TEE non-secure components embedded files from OP-TEE test, examples and benchmark packages.
The Linux custom configuration is dumped from the vexpress_defconfig with few added fragments: OP-TEE driver and 9p for virtual filesystem to ease file manipulation and exchanges through Qemu virtfs support.
The standard way for booting OP-TEE with a non-secure world companion use the Arm Trusted Firmware-A as bootloader. OP-TEE OS provides the BL32 image and U-boot the BL33 image. The proposed board enables OP-TEE and U-boot build for this. However package boot/arm-trusted-firmware needs few change support building Armv7-A targets.
Therefore the proposed board allows one to build the images but not yet to run the target with the built Qemu host tool.
---
Changes v3 -> v4
- No change.
Changes v2 -> v3
- New change to introduce a board that at least builds Armv7-A OP-TEE.
--- board/qemu/armv7a-tz-virt/linux.config | 167 ++++++++++++++++++++++++++++++++ board/qemu/armv7a-tz-virt/readme.txt | 11 +++ board/qemu/armv7a-tz-virt/u-boot.config | 3 + configs/qemu_armv7a_tz_virt_defconfig | 41 ++++++++ 4 files changed, 222 insertions(+) create mode 100644 board/qemu/armv7a-tz-virt/linux.config create mode 100644 board/qemu/armv7a-tz-virt/readme.txt create mode 100644 board/qemu/armv7a-tz-virt/u-boot.config create mode 100644 configs/qemu_armv7a_tz_virt_defconfig
board/qemu/armv7a-tz-virt/linux.config new file +167 -0
@@ -0,0 +1,167 @@ +# CONFIG_LOCALVERSION_AUTO is not set +CONFIG_SYSVIPC=y +CONFIG_IKCONFIG=y +CONFIG_IKCONFIG_PROC=y +CONFIG_LOG_BUF_SHIFT=14 +CONFIG_CGROUPS=y +CONFIG_CPUSETS=y +# CONFIG_UTS_NS is not set +# CONFIG_IPC_NS is not set +# CONFIG_PID_NS is not set +# CONFIG_NET_NS is not set +CONFIG_BLK_DEV_INITRD=y +CONFIG_PROFILING=y +CONFIG_OPROFILE=y +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +# CONFIG_LBDAF is not set +# CONFIG_BLK_DEV_BSG is not set +# CONFIG_IOSCHED_DEADLINE is not set +# CONFIG_IOSCHED_CFQ is not set +CONFIG_ARCH_VEXPRESS=y +CONFIG_ARCH_VEXPRESS_DCSCB=y +CONFIG_ARCH_VEXPRESS_TC2_PM=y +# CONFIG_SWP_EMULATE is not set +CONFIG_SMP=y +CONFIG_HAVE_ARM_ARCH_TIMER=y +CONFIG_MCPM=y +CONFIG_VMSPLIT_2G=y +CONFIG_NR_CPUS=8 +CONFIG_ARM_PSCI=y +CONFIG_AEABI=y +CONFIG_CMA=y +CONFIG_ZBOOT_ROM_TEXT=0x0 +CONFIG_ZBOOT_ROM_BSS=0x0 +CONFIG_CMDLINE="console=ttyAMA0" +CONFIG_CPU_IDLE=y +CONFIG_CPU_IDLE_MULTIPLE_DRIVERS=y +CONFIG_VFP=y +CONFIG_NEON=y +# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_UNIX=y +CONFIG_INET=y +CONFIG_IP_PNP=y +CONFIG_IP_PNP_DHCP=y +CONFIG_IP_PNP_BOOTP=y +# CONFIG_IPV6 is not set +# CONFIG_WIRELESS is not set +CONFIG_NET_9P=y +CONFIG_NET_9P_VIRTIO=y +CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" +CONFIG_DEVTMPFS=y +CONFIG_MTD=y +CONFIG_MTD_CMDLINE_PARTS=y +CONFIG_MTD_BLOCK=y +CONFIG_MTD_CFI=y +CONFIG_MTD_CFI_INTELEXT=y +CONFIG_MTD_CFI_AMDSTD=y +CONFIG_MTD_PHYSMAP=y +CONFIG_MTD_PHYSMAP_OF=y +CONFIG_MTD_PLATRAM=y +CONFIG_MTD_UBI=y +CONFIG_PROC_DEVICETREE=y +CONFIG_VIRTIO_BLK=y +# CONFIG_SCSI_PROC_FS is not set +CONFIG_BLK_DEV_SD=y +CONFIG_SCSI_VIRTIO=y +CONFIG_ATA=y +# CONFIG_SATA_PMP is not set +CONFIG_NETDEVICES=y +CONFIG_VIRTIO_NET=y +CONFIG_SMC91X=y +CONFIG_SMSC911X=y +# CONFIG_WLAN is not set +CONFIG_INPUT_EVDEV=y +# CONFIG_SERIO_SERPORT is not set +CONFIG_SERIO_AMBAKMI=y +CONFIG_LEGACY_PTY_COUNT=16 +CONFIG_SERIAL_AMBA_PL011=y +CONFIG_SERIAL_AMBA_PL011_CONSOLE=y +CONFIG_VIRTIO_CONSOLE=y +CONFIG_HW_RANDOM=y +CONFIG_HW_RANDOM_VIRTIO=y +CONFIG_I2C=y +CONFIG_I2C_VERSATILE=y +CONFIG_SENSORS_VEXPRESS=y +CONFIG_REGULATOR=y +CONFIG_REGULATOR_VEXPRESS=y +CONFIG_FB=y +CONFIG_FB_ARMCLCD=y +CONFIG_FRAMEBUFFER_CONSOLE=y +CONFIG_LOGO=y +# CONFIG_LOGO_LINUX_MONO is not set +# CONFIG_LOGO_LINUX_VGA16 is not set +CONFIG_SOUND=y +CONFIG_SND=y +CONFIG_SND_MIXER_OSS=y +CONFIG_SND_PCM_OSS=y +# CONFIG_SND_DRIVERS is not set +CONFIG_SND_ARMAACI=y +CONFIG_HID_DRAGONRISE=y +CONFIG_HID_GYRATION=y +CONFIG_HID_TWINHAN=y +CONFIG_HID_NTRIG=y +CONFIG_HID_PANTHERLORD=y +CONFIG_HID_PETALYNX=y +CONFIG_HID_SAMSUNG=y +CONFIG_HID_SONY=y +CONFIG_HID_SUNPLUS=y +CONFIG_HID_GREENASIA=y +CONFIG_HID_SMARTJOYPLUS=y +CONFIG_HID_TOPSEED=y +CONFIG_HID_THRUSTMASTER=y +CONFIG_HID_ZEROPLUS=y +CONFIG_USB=y +CONFIG_USB_ANNOUNCE_NEW_DEVICES=y +CONFIG_USB_MON=y +CONFIG_USB_STORAGE=y +CONFIG_USB_ISP1760=y +CONFIG_MMC=y +CONFIG_MMC_ARMMMCI=y +CONFIG_NEW_LEDS=y +CONFIG_LEDS_CLASS=y +CONFIG_LEDS_GPIO=y +CONFIG_LEDS_TRIGGERS=y +CONFIG_LEDS_TRIGGER_HEARTBEAT=y +CONFIG_LEDS_TRIGGER_CPU=y +CONFIG_RTC_CLASS=y +CONFIG_RTC_DRV_PL031=y +CONFIG_VIRTIO_BALLOON=y +CONFIG_VIRTIO_MMIO=y +CONFIG_VIRTIO_MMIO_CMDLINE_DEVICES=y +CONFIG_EXT2_FS=y +CONFIG_EXT3_FS=y +# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set +# CONFIG_EXT3_FS_XATTR is not set +CONFIG_EXT4_FS=y +CONFIG_VFAT_FS=y +CONFIG_TMPFS=y +#CONFIG_JFFS2_FS is not set +CONFIG_UBIFS_FS=y +CONFIG_CRAMFS=y +CONFIG_SQUASHFS=y +CONFIG_SQUASHFS_LZO=y +CONFIG_NFS_FS=y +CONFIG_ROOT_NFS=y +CONFIG_9P_FS=y +CONFIG_NLS_CODEPAGE_437=y +CONFIG_NLS_ISO8859_1=y +CONFIG_DEBUG_INFO=y +CONFIG_DEBUG_FS=y +CONFIG_MAGIC_SYSRQ=y +CONFIG_DEBUG_KERNEL=y +CONFIG_DETECT_HUNG_TASK=y +# CONFIG_SCHED_DEBUG is not set +CONFIG_DEBUG_USER=y +# CONFIG_CRYPTO_ANSI_CPRNG is not set +# CONFIG_CRYPTO_HW is not set +### Enable OP-TEE +CONFIG_TEE=y +CONFIG_OPTEE=y +### Enable 9P VFS +CONFIG_NET_9P=y +CONFIG_NET_9P_VIRTIO=y +CONFIG_9P_FS=y +CONFIG_9P_FS_POSIX_ACL=y
board/qemu/armv7a-tz-virt/readme.txt new file +11 -0
@@ -0,0 +1,11 @@ +Board qemu_armv7a_tz_virt builds a Qemu Armv7-A target with +OP-TEE running in the TrustZone secure world setup and a Linux based +OS running in the non-secure world. + +This setup is usually booted with the Arm Trsuted Firmware-A (TF-A from +package boot/arm-trusted-firmware). However the current Buildroot package +needs few changes to build TF-A for OP-TEE support. + +Until BR arm-trusted-firmware is updated this board allows one to only +build the secure and non-secure boot images if not the BIOS for the Qemu +host.
board/qemu/armv7a-tz-virt/u-boot.config new file +3 -0
@@ -0,0 +1,3 @@ +CONFIG_SYS_TEXT_BASE=0x60000000 +CONFIG_BOOTCOMMAND="fdt addr ${fdt_addr} && fdt resize 1000 && smhload zImage ${kernel_addr_r} && smhload rootfs.cpio.gz ${ramdisk_addr_r} ramdisk_addr_end && setenv bootargs console=ttyAMA0,115200 earlyprintk=serial,ttyAMA0,115200 && fdt chosen ${ramdisk_addr_r} ${ramdisk_addr_end} && bootz ${kernel_addr_r} - ${fdt_addr}" +CONFIG_SEMIHOSTING=y
configs/qemu_armv7a_tz_virt_defconfig new file +41 -0
41 lines not on this page · raw message
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Hello,
Quoted 17 lines · On Wed, 30 Jan 2019 11:47:28 +0100 Etienne Carriere…
This change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
The target Linux based normal world embeds the standard minimal filesystem with OP-TEE non-secure components embedded files from OP-TEE test, examples and benchmark packages.
The Linux custom configuration is dumped from the
vexpress_defconfigwith few added fragments: OP-TEE driver and 9p for virtual filesystem to ease file manipulation and exchanges through Qemu virtfs support.The standard way for booting OP-TEE with a non-secure world companion use the Arm Trusted Firmware-A as bootloader. OP-TEE OS provides the BL32 image and U-boot the BL33 image. The proposed board enables OP-TEE and U-boot build for this. However package boot/arm-trusted-firmware needs few change support building Armv7-A targets.
Quoted 4 lines
Therefore the proposed board allows one to build the images but not yet to run the target with the built Qemu host tool.
Signed-off-by Etienne Carriere <[email protected]>
I don't know what is the opinion of Peter, Arnout and Yann, but I think this PATCH 6/7 and PATCH 7/7, instead of adding more defconfigs should instead add test cases to our runtime test infrastructure in support/testing/. Indeed:
- We probably don't want to have Qemu defconfigs for every possible feature in Buildroot
- A runtime test case, even if it's indeed a bit less visible than a defconfig, still documents a configuration that "works" for a given feature.
- A runtime test case allows to really runtime test the feature by booting Qemu.
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Thomas, Etienne, All,
On 2019-02-17 23:12 +0100, Thomas Petazzoni spake thusly:
Quoted 11 lines
Quoted 2 lines · On Wed, 30 Jan 2019 11:47:28 +0100 Etienne Carriere…
This change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
I don't know what is the opinion of Peter, Arnout and Yann, but I think this PATCH 6/7 and PATCH 7/7, instead of adding more defconfigs should instead add test cases to our runtime test infrastructure in support/testing/. Indeed:
- We probably don't want to have Qemu defconfigs for every possible feature in Buildroot
However, I would not be opposed to having _one_ defconfig that can be used as a reference / starting-point.
Quoted 5 lines
- A runtime test case, even if it's indeed a bit less visible than a defconfig, still documents a configuration that "works" for a given feature.
- A runtime test case allows to really runtime test the feature by booting Qemu.
Agreed: adding a runtiem test should indeed be provided, whether we have a defconfig or not.
Regards,
Yann E. MORIN.
Quoted 10 lines
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 561 099 427 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Hello all,
Quoted 13 lines · Yann E. MORIN
Thomas, Etienne, All,
On 2019-02-17 23:12 +0100, Thomas Petazzoni spake thusly:
Quoted 9 lines
Quoted 2 lines · On Wed, 30 Jan 2019 11:47:28 +0100 Etienne Carriere…
This change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
I don't know what is the opinion of Peter, Arnout and Yann, but I think this PATCH 6/7 and PATCH 7/7, instead of adding more defconfigs should instead add test cases to our runtime test infrastructure in support/testing/. Indeed:
Quoted 5 lines
Quoted 2 lines
- We probably don't want to have Qemu defconfigs for every possible feature in Buildroot
However, I would not be opposed to having _one_ defconfig that can be used as a reference / starting-point.
Is the Qemu emulator the best candidate for such starting point. I think it is as one can use it to experience Arm specific OP-TEE package without needing specific HW but a standard Linux host.
I would have preferred proposing a change in the already available Qemu Armv7 as qemu_arm_vexpress_defconfig is but I fear enabling TrustZone support in Qemu will break other nice Qemu features ones are used to (graphics?).
Maybe I can find a real HW for which BR can store a defconfig that enables OP-TEE.
Quoted 6 lines
Quoted 5 lines
- A runtime test case, even if it's indeed a bit less visible than a defconfig, still documents a configuration that "works" for a given feature.
- A runtime test case allows to really runtime test the feature by booting Qemu.
Quoted 9 lines
Agreed: adding a runtiem test should indeed be provided, whether we have a defconfig or not.
Regards,
Yann E. MORIN.Quoted 2 lines
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
I think I can prepare that. Or I will ask few help on the ML if I can't find my way.
The initial intention when adding these defconfig to my patch series was to answer a request from patch v3 (i think) review where Thomas asked for something that could b used to check OP-TEE at least builds, if possible boots, from a BR build. I understand that maybe you though more of such runtime test, rather than a defconfig.
Regards,
etienne
Quoted 17 lines
Quoted 9 lines
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com--
.-----------------.--------------------.------------------.--------------------.| Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 561 099 427 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. |'------------------------------^-------^------------------^--------------------'
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Hello,
Quoted 7 lines · On Mon, 18 Feb 2019 22:28:10 +0100 Etienne Carriere…
Quoted 2 lines
However, I would not be opposed to having _one_ defconfig that can be used as a reference / starting-point.
Is the Qemu emulator the best candidate for such starting point. I think it is as one can use it to experience Arm specific OP-TEE package without needing specific HW but a standard Linux host.
Quoted 7 lines
I would have preferred proposing a change in the already available Qemu Armv7 as
qemu_arm_vexpress_defconfigis but I fear enabling TrustZone support in Qemu will break other nice Qemu features ones are used to (graphics?).Maybe I can find a real HW for which BR can store a defconfig that enables OP-TEE.
I think Yann didn't say that a Qemu defconfig was not good, he said quite the opposite: that having one Qemu defconfig that uses OP-TEE would be useful.
Note: I am not sure I agree because if we go down this road, we would have lots of Qemu defconfigs demonstrating lots of different features.
Quoted 5 lines
Quoted 2 lines
Quoted 2 lines
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
I think I can prepare that. Or I will ask few help on the ML if I can't find my way.
Sure, feel free to ask questions. The runtime test infrastructure is not documented, but there are numerous existing test cases that you should help you getting started.
You can list existing tests by doing:
./support/testing/run-tests -l
and run one specific test by doing:
./support/testing/run-tests -d /path/to/some/build/dir tests.fs.test_ext.TestExt2
Option -k is really useful during development, as it keeps the build artifacts instead of removing them once the test is completed.
Quoted 5 lines
The initial intention when adding these defconfig to my patch series was to answer a request from patch v3 (i think) review where Thomas asked for something that could b used to check OP-TEE at least builds, if possible boots, from a BR build. I understand that maybe you though more of such runtime test, rather than a defconfig.
Yeah, maybe I wasn't clear back then, sorry about that.
Best regards,
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
In reply to Thomas PetazzoniQuoted 27 lines · Thomas Petazzoni
Hello,
Quoted 17 lines · On Wed, 30 Jan 2019 11:47:28 +0100 Etienne Carriere…
This change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
The target Linux based normal world embeds the standard minimal filesystem with OP-TEE non-secure components embedded files from OP-TEE test, examples and benchmark packages.
The Linux custom configuration is dumped from the
vexpress_defconfigwith few added fragments: OP-TEE driver and 9p for virtual filesystem to ease file manipulation and exchanges through Qemu virtfs support.The standard way for booting OP-TEE with a non-secure world companion use the Arm Trusted Firmware-A as bootloader. OP-TEE OS provides the BL32 image and U-boot the BL33 image. The proposed board enables OP-TEE and U-boot build for this. However package boot/arm-trusted-firmware needs few change support building Armv7-A targets.
Quoted 4 lines
Therefore the proposed board allows one to build the images but not yet to run the target with the built Qemu host tool.
Signed-off-by Etienne Carriere <[email protected]>
Quoted 7 lines
I don't know what is the opinion of Peter, Arnout and Yann, but I think this PATCH 6/7 and PATCH 7/7, instead of adding more defconfigs should instead add test cases to our runtime test infrastructure in support/testing/. Indeed:
- We probably don't want to have Qemu defconfigs for every possible feature in Buildroot
I don't quite agree. I think we *do* want to have defconfigs that demonstrate
major features. For example, I like that we have qt5 defconfigs for several platforms.
And for those defconfigs, obviously it is great if it can be qemu-based, as
Etienne pointed out. (For the Qt5 ones obviously they can't be qemu-based, and indeed currently we don't have any qemu-based feature defconfigs.)
So IMO this TrustZone defconfig is a good thing.
Quoted 4 lines
- A runtime test case, even if it's indeed a bit less visible than a defconfig, still documents a configuration that "works" for a given feature.
So I would propose a runtime test that uses that defconfig.
Regards,
Arnout
Quoted 11 lines
- A runtime test case allows to really runtime test the feature by booting Qemu.
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
Thanks!
Thomas
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Quoted 47 lines · Arnout Vandecappelle
Quoted 27 lines · Thomas Petazzoni
Hello,
Quoted 17 lines · On Wed, 30 Jan 2019 11:47:28 +0100 Etienne Carriere…
This change introduces a Qemu board for an Armv7-A target executing with OP-TEE secure world services.
The target Linux based normal world embeds the standard minimal filesystem with OP-TEE non-secure components embedded files from OP-TEE test, examples and benchmark packages.
The Linux custom configuration is dumped from the
vexpress_defconfigwith few added fragments: OP-TEE driver and 9p for virtual filesystem to ease file manipulation and exchanges through Qemu virtfs support.The standard way for booting OP-TEE with a non-secure world companion use the Arm Trusted Firmware-A as bootloader. OP-TEE OS provides the BL32 image and U-boot the BL33 image. The proposed board enables OP-TEE and U-boot build for this. However package boot/arm-trusted-firmware needs few change support building Armv7-A targets.
Quoted 4 lines
Therefore the proposed board allows one to build the images but not yet to run the target with the built Qemu host tool.
Signed-off-by Etienne Carriere <[email protected]>Quoted 7 lines
I don't know what is the opinion of Peter, Arnout and Yann, but I think this PATCH 6/7 and PATCH 7/7, instead of adding more defconfigs should instead add test cases to our runtime test infrastructure in support/testing/. Indeed:
- We probably don't want to have Qemu defconfigs for every possible feature in Buildroot
I don't quite agree. I think we *do* want to have defconfigs that demonstrate
major features. For example, I like that we have qt5 defconfigs for several platforms.And for those defconfigs, obviously it is great if it can be qemu-based, as
Etienne pointed out. (For the Qt5 ones obviously they can't be qemu-based, and indeed currently we don't have any qemu-based feature defconfigs.)
Quoted 9 lines
So IMO this TrustZone defconfig is a good thing.
Quoted 4 lines
- A runtime test case, even if it's indeed a bit less visible than a defconfig, still documents a configuration that "works" for a given feature.
So I would propose a runtime test that uses that defconfig.
Dear all,
I've prepared something to test the optee on qemu/arm through the runtime tests but few questions puzzle me.
1/ To use the board defconfig for the test, I created the file support/test/conf/qemu_xxx_defconfig as a symlink to configs/qemu_xxx_defconfig.
I'm not sure it is that nice. But it allowed to set the test config in the test_opee.py using:
class TestOpteeXtest(infra.basetest.BRTest):
with open(infra.filepath('conf/qemu_armv7a_tz_virt_defconfig'),
'r') as config_file:
config = "".join(line for line in config_file if line[:1]!='#')
If you think there is another more convenient way, feel free to suggest.
2/ To share this test I need few preliminary changes in arm-trusted-firmware. This will brings a series of dependent patches: tf-a updates + qemu board defconfig + optee runtime test script.
The series depends on branch next, as dependent on OP-TEE packages. That will make several versatile changes in a single patches series over the next branch.
I plan to submit the series to the ML, for the next.
Tell me if you rather have this work be reviewed in 2 steps: first tf-a changes, then once merged, qemu board and the optee runtime test.
Best regards,
etienne
Quoted 16 lines
Regards,
ArnoutQuoted 11 lines
- A runtime test case allows to really runtime test the feature by booting Qemu.
Etienne, would you be willing to convert those two configurations to the runtime test infrastructure ?
Thanks!
Thomas
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Hello ?tienne,
Quoted 9 lines · On Tue, 5 Mar 2019 10:14:07 +0100 Etienne Carriere…
I've prepared something to test the optee on qemu/arm through the runtime tests but few questions puzzle me.
1/ To use the board defconfig for the test, I created the file support/test/conf/qemu_xxx_defconfig as a symlink to configs/qemu_xxx_defconfig.
I'm not sure it is that nice. But it allowed to set the test config in thetest_opee.pyusing:
Quoted 4 lines
class TestOpteeXtest(infra.basetest.BRTest):
with open(infra.filepath('conf/qemu_armv7a_tz_virt_defconfig'),
'r') asconfig_file:config = "".join(line for line in config_file if line[:1]!='#')
infra.filepath() is as simple as:
def filepath(relpath):
return os.path.join(os.getcwd(), "support/testing", relpath)
so you can just do:
with open(os.path.join(os.getcwd(), "configs/qemu_armv7a_tz_virt_defconfig"))
however, I see one down-side with this: defconfigs are usually using an internal toolchain, so they take a long time to build. We typically try to use external toolchains for most runtime tests, to make them faster and therefore more usable. But perhaps you can take the defconfig + inject a few lines of configuration to use an external toolchain.
Quoted 10 lines
2/ To share this test I need few preliminary changes in arm-trusted-firmware. This will brings a series of dependent patches: tf-a updates + qemu board defconfig + optee runtime test script.
The series depends on branch next, as dependent on OP-TEE packages. That will make several versatile changes in a single patches series over the next branch.I plan to submit the series to the ML, for the next.
Tell me if you rather have this work be reviewed in 2 steps: first tf-a changes, then once merged, qemu board and the optee runtime test.
Note: next is going to be merged back in master very soon now, since 2019.02 has been released.
If you already have all the changes, then please send them in one single series, it makes it clearer why the preparation patches are needed.
Also, I have seen your series improving various things in optee related packages, I intend to have a look very soon.
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com
PATCH v4 6/7 configs/qemu_armv7a_tz_virt: Armv7-A emulation with TrustZone services
Quoted 20 lines · Thomas Petazzoni
Hello ?tienne,
Quoted 9 lines · On Tue, 5 Mar 2019 10:14:07 +0100 Etienne Carriere…
I've prepared something to test the optee on qemu/arm through the runtime tests but few questions puzzle me.
1/ To use the board defconfig for the test, I created the file support/test/conf/qemu_xxx_defconfig as a symlink to configs/qemu_xxx_defconfig.
I'm not sure it is that nice. But it allowed to set the test config in thetest_opee.pyusing:Quoted 4 lines
class TestOpteeXtest(infra.basetest.BRTest):
with open(infra.filepath('conf/qemu_armv7a_tz_virt_defconfig'),
'r') asconfig_file:config = "".join(line for line in config_file if line[:1]!='#')
Quoted 9 lines
infra.filepath()is as simple as:def filepath(relpath):
returnos.path.join(os.getcwd(), "support/testing", relpath)so you can just do:
with
open(os.path.join(os.getcwd(), "configs/qemu_armv7a_tz_virt_defconfig"))
Ok, i should have looked into :)
Thanks. I prefer to get rid of such symlink.
Quoted 6 lines
however, I see one down-side with this: defconfigs are usually using an internal toolchain, so they take a long time to build. We typically try to use external toolchains for most runtime tests, to make them faster and therefore more usable. But perhaps you can take the defconfig + inject a few lines of configuration to use an external toolchain.
Great, I'll go this way: defconfig + extra toolchain directives.
Quoted 11 lines
Quoted 10 lines
2/ To share this test I need few preliminary changes in arm-trusted-firmware. This will brings a series of dependent patches: tf-a updates + qemu board defconfig + optee runtime test script.
The series depends on branch next, as dependent on OP-TEE packages. That will make several versatile changes in a single patches series over the next branch.I plan to submit the series to the ML, for the next.
Tell me if you rather have this work be reviewed in 2 steps: first tf-a changes, then once merged, qemu board and the optee runtime test.
Quoted 7 lines
Note: next is going to be merged back in master very soon now, since 2019.02 has been released.
If you already have all the changes, then please send them in one single series, it makes it clearer why the preparation patches are needed.
Fine, thank.
I'll send soon.
Regards,
etienne
Quoted 11 lines
Also, I have seen your series improving various things in optee related packages, I intend to have a look very soon.
Thanks!
Thomas
--
Thomas Petazzoni, CTO, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
PATCH v4 7/7 configs/qemu_aarch64_tz_virt: AArch64 emulation with TrustZone services
In reply to Etienne CarriereThis change introduces a Qemu board for an AArch64 target executing with OP-TEE secure world services.
The target Linux based normal world embeds the standard minimal filesystem with OP-TEE non-secure components embedded files from OP-TEE client, test, examples and benchmark packages.
The Linux custom configuration is dumped from the qemu/aarch64-virt board with few added fragments: OP-TEE driver and 9p for virtual filesystem to ease file manipulation and exchanges through Qemu virtfs support.
The standard way for booting OP-TEE with a non-secure world companion use the Arm Trusted Firmware-A as bootloader. OP-TEE OS provides the BL32 image and U-boot the BL33 image. The proposed board enables OP-TEE and U-boot build for this. However package boot/arm-trusted-firmware needs few changes to build OP-TEE compliant TF-A images.
Therefore the proposed board allows one to build the images but not yet to run the target with the built Qemu host tool.
---
Changes v3 -> v4
- No change.
Changes v2 -> v3
- New change to introduce a board that at least builds Aarch64 OP-TEE.
--- board/qemu/aarch64-tz-virt/linux.config | 63 ++++++++++++++++++++++++++++++++ board/qemu/aarch64-tz-virt/readme.txt | 11 ++++++ board/qemu/aarch64-tz-virt/u-boot.config | 3 ++ configs/qemu_aarch64_tz_virt_defconfig | 47 ++++++++++++++++++++++++ 4 files changed, 124 insertions(+) create mode 100644 board/qemu/aarch64-tz-virt/linux.config create mode 100644 board/qemu/aarch64-tz-virt/readme.txt create mode 100644 board/qemu/aarch64-tz-virt/u-boot.config create mode 100644 configs/qemu_aarch64_tz_virt_defconfig
board/qemu/aarch64-tz-virt/linux.config new file +63 -0
@@ -0,0 +1,63 @@ +CONFIG_SYSVIPC=y +CONFIG_POSIX_MQUEUE=y +CONFIG_NO_HZ_IDLE=y +CONFIG_HIGH_RES_TIMERS=y +CONFIG_TASKSTATS=y +CONFIG_SCHED_AUTOGROUP=y +CONFIG_PROFILING=y +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +CONFIG_BLK_DEV_BSGLIB=y +CONFIG_ARCH_VEXPRESS=y +CONFIG_TRANSPARENT_HUGEPAGE=y +CONFIG_BINFMT_MISC=y +CONFIG_COMPAT=y +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_PACKET_DIAG=y +CONFIG_UNIX=y +CONFIG_NET_KEY=y +CONFIG_INET=y +CONFIG_IP_MULTICAST=y +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_BRIDGE=m +CONFIG_NET_SCHED=y +CONFIG_VSOCKETS=y +CONFIG_DEVTMPFS=y +CONFIG_DEVTMPFS_MOUNT=y +CONFIG_VIRTIO_BLK=y +CONFIG_BLK_DEV_SD=y +CONFIG_CHR_DEV_SG=y +CONFIG_SCSI_CONSTANTS=y +CONFIG_SCSI_LOGGING=y +CONFIG_SCSI_SCAN_ASYNC=y +CONFIG_SCSI_VIRTIO=y +CONFIG_ATA=y +CONFIG_NETDEVICES=y +CONFIG_DUMMY=y +CONFIG_MACVLAN=y +CONFIG_VIRTIO_NET=y +CONFIG_NLMON=y +CONFIG_VT_HW_CONSOLE_BINDING=y +CONFIG_SERIAL_AMBA_PL011=y +CONFIG_SERIAL_AMBA_PL011_CONSOLE=y +CONFIG_VIRTIO_CONSOLE=y +CONFIG_HW_RANDOM=y +CONFIG_HW_RANDOM_VIRTIO=y +CONFIG_RTC_CLASS=y +CONFIG_VIRTIO_MMIO=y +CONFIG_VIRTIO_MMIO_CMDLINE_DEVICES=y +CONFIG_MAILBOX=y +CONFIG_PL320_MBOX=y +CONFIG_EXT4_FS=y +CONFIG_TMPFS=y +CONFIG_TMPFS_POSIX_ACL=y +### Enable OP-TEE +CONFIG_TEE=y +CONFIG_OPTEE=y +### Enable 9P VFS +CONFIG_NET_9P=y +CONFIG_NET_9P_VIRTIO=y +CONFIG_9P_FS=y +CONFIG_9P_FS_POSIX_ACL=y +
board/qemu/aarch64-tz-virt/readme.txt new file +11 -0
@@ -0,0 +1,11 @@ +Board qemu_aarch64_tz_virt builds a Qemu Armv8-A target with +OP-TEE running in the TrustZone secure wolrd setup and a Linux based +OS running in the non-secure wolrd. + +This setup is usually booted with the Arm Trsuted Firmware-A (TF-A from +package boot/arm-trusted-firmware). However the current Buildroot package +needs few changes to build TF-A for OP-TEE support. + +Until BR arm-trusted-firmware is updated this board allows one to only +build the secure and non-secure boot images if not the BIOS for the Qemu +host.
board/qemu/aarch64-tz-virt/u-boot.config new file +3 -0
@@ -0,0 +1,3 @@ +CONFIG_SYS_TEXT_BASE=0x60000000 +CONFIG_BOOTCOMMAND="fdt addr ${fdt_addr} && fdt resize 1000 && smhload zImage ${kernel_addr_r} && smhload rootfs.cpio.gz ${ramdisk_addr_r} ramdisk_addr_end && setenv bootargs console=ttyAMA0,115200 earlyprintk=serial,ttyAMA0,115200 && fdt chosen ${ramdisk_addr_r} ${ramdisk_addr_end} && bootz ${kernel_addr_r} - ${fdt_addr}" +CONFIG_SEMIHOSTING=y
configs/qemu_aarch64_tz_virt_defconfig new file +47 -0
47 lines not on this page · raw message
Thomas Petazzoni
AppliedPATCH v4 1/7 boot/optee-os: new package
In reply to Etienne CarriereQuoted 14 lines · On Wed, 30 Jan 2019 11:47:23 +0100 Etienne Carriere…
OP-TEE OS is maintained by the OP-TEE project. It provides an open source solution for development and integration of secure services for Armv7-A and Armv8-A CPU based platforms supporting the TrustZone technology. This technology enables CPUs to concurrently host a secure world as the OP-TEE OS and a non-secure world as a Linux based OS.
The OP-TEE project maintains other packages to leverage OP-TEE on Linux kernel based OSes. An OP-TEE interface driver is available in the Linux kernel since 4.12 upon
CONFIG_OPTEE.This change references in Buildroot the today's latest OP-TEE revision release tagged 3.4.0.
Quoted 8 lines
https://www.op-tee.org/
https://github.com/OP-TEE/optee_osSigned-off-by Etienne Carriere <[email protected]>---
Changes v3 -> v4
- Upgrade from OP-TEE 3.3.0 to 3.4.0. No local patch required.
Applied to next with the following fixes:
[Thomas:
- pass CFG_ARM32_core=y when building for ARMv7, otherwise the build fails
- add a check that verifies that BR2_TARGET_OPTEE_OS_PLATFORM is not empty
- minor formatting fixes/adjustements.]
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com