IOT2050-related enhancements
v2 cover IOT2050-related enhancements
(Almost) flushing our upstream queue for the IOT2050 device, this mostly brings board-specific changes such as:
- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devices
There are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags to CONFIG_NET (repost)
- tools: Add script for converting public key into device tree include
Changes in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"
Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Jan
Jan Kiszka (10):
env: Complete generic support for writable list
env: Couple networking-related variable flags to CONFIG_NET
tools: Add script for converting public key into device tree include
iot2050: Update firmware layout
iot2050: Add watchdog start to bootcmd
iot2050: Add CONFIG_ENV_FLAGS_LIST_STATIC
arm: dts: iot2050: Allow verifying U-Boot proper by SPL
iot2050: Add script for signing artifacts
arm: dts: iot2050: Optionally embed OTP programming data into image
doc: iot2050: Add a note about the watchdog firmware
Su Baocheng (2):
board: siemens: iot2050: Split the build for PG1 and PG2
arm: dts: iot2050: Use the auto generator nodes for fdt
chao zeng (1):
board: siemens: iot2050: use the named gpio to control the user-button
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 134 ++++++------------
board/siemens/iot2050/Kconfig | 35 ++++-
board/siemens/iot2050/board.c | 17 +--
...ot2050_defconfig => iot2050_pg1_defconfig} | 7 +-
...ot2050_defconfig => iot2050_pg2_defconfig} | 9 +-
doc/board/siemens/iot2050.rst | 79 ++++++++++-
env/Kconfig | 2 +
env/env.c | 22 +++
env/flags.c | 10 +-
include/configs/iot2050.h | 17 +++
include/env_flags.h | 4 +-
tools/binman/missing-blob-help | 14 +-
tools/iot2050-sign-fw.sh | 51 +++++++
tools/key2dtsi.py | 64 +++++++++
14 files changed, 334 insertions(+), 131 deletions(-)
copy configs/{iot2050_defconfig => iot2050_pg1_defconfig} (94%)
rename configs/{iot2050_defconfig => iot2050_pg2_defconfig} (92%)
create mode 100755 tools/iot2050-sign-fw.sh
create mode 100755 tools/key2dtsi.pySignature · 1 line
2.35.3
PATCH v2 1/13 env: Complete generic support for writable list
From: Jan Kiszka <[email protected]>
This completes what 890feecaab72 started by selecting ENV_APPEND and ENV_IS_NOWHERE and by moving this driver to top if the list. This ensures that load operations pick up both the default env and the permitted parts of the next-prio location. When writing though, we must use the regular ordering.
With this change, boards only need to define the list of writable variables but no longer have to provide a custom env_get_location implementation.
--- env/Kconfig | 2 ++ env/env.c | 22 ++++++++++++++++++++++ 2 files changed, 24 insertions(+)
env/Kconfig +2 -0
@@ -715,6 +715,8 @@ config ENV_APPEND config ENV_WRITEABLE_LIST bool "Permit write access only to listed variables" + select ENV_IS_NOWHERE + select ENV_APPEND help If defined, only environment variables which explicitly set the 'w' writeable flag can be written and modified at runtime. No variables
env/env.c +22 -0
@@ -133,6 +133,28 @@ __weak enum env_location arch_env_get_location(enum env_operation op, int prio)if (prio >=ARRAY_SIZE (env_locations))return ENVL_UNKNOWN ; +if (IS_ENABLED (CONFIG_ENV_WRITEABLE_LIST )) { +/* +* In writeable-list mode, ENVL_NOWHERE gains highest prio by +* virtually injecting it at prio 0. +*/ +if (prio ==0 ) { +/* +* Avoid the injection for write operations as that +* would block it. +*/ +if (op !=ENVOP_SAVE && op !=ENVOP_ERASE ) +return ENVL_NOWHERE ; + }else { +/* +* always subtract 1, also for writing because +* env_load_prio, which is used for writing, was +* initialized with that offset. +*/ + prio--; + } + } +return env_locations[prio]; }
Signature · 1 line
2.35.3
Re: [PATCH V2 01/13] env: Complete generic support for writable list
Hi,
Am 05.10.2022 um 10:33 schrieb Jan Kiszka:
Quoted 43 lines
From: Jan Kiszka <[email protected]>
This completes what
890feecaab72started by selectingENV_APPENDandENV_IS_NOWHEREand by moving this driver to top if the list. This ensures that load operations pick up both the default env and the permitted parts of the next-prio location. When writing though, we must use the regular ordering.With this change, boards only need to define the list of writable variables but no longer have to provide a custom
env_get_locationimplementation.CC Joe Hershberger <[email protected]>CC Marek Vasut <[email protected]>Signed-off-by Jan Kiszka <[email protected]>--- env/Kconfig | 2 ++ env/env.c | 22 ++++++++++++++++++++++ 2 files changed, 24 insertions(+)env/Kconfig +2
-0@@ -715,6 +715,8 @@ config ENV_APPEND config ENV_WRITEABLE_LIST bool "Permit write access only to listed variables" + select ENV_IS_NOWHERE + select ENV_APPEND help If defined, only environment variables which explicitly set the 'w' writeable flag can be written and modified at runtime. No variablesenv/env.c +2
-0@@ -133,6 +133,28 @@ __weak enum env_location arch_env_get_location(enum env_operation op, int prio)if (prio >=ARRAY_SIZE (env_locations))return ENVL_UNKNOWN ; +if (IS_ENABLED (CONFIG_ENV_WRITEABLE_LIST )) { +/*
env/env.c __weak enum env_location arch_env_get_location(enum env_operation op, int prio) · line 147
Quoted 10 lines
env/env.c +10
-0+* In writeable-list mode, ENVL_NOWHERE gains highest prio by +* virtually injecting it at prio 0. +*/ +if (prio ==0 ) { +/* +* Avoid the injection for write operations as that +* would block it. +*/ +if (op !=ENVOP_SAVE && op !=ENVOP_ERASE ) +return ENVL_NOWHERE ;
Is it consensus now to use ENVL_NOWHERE as synonym for default environment? If I remember correct this was rejected in the past and ENVL_NOWHERE should only be used if no enviroment is available.
Why don't you call env_set_default(NULL, 0) in env_load() before env_driver_lookup()?
env/env.c __weak enum env_location arch_env_get_location(enum env_operation op, int prio) · line 160
Quoted 13 lines
env/env.c +10
-0+ }else { +/* +* always subtract 1, also for writing because +* env_load_prio, which is used for writing, was +* initialized with that offset. +*/ + prio--; + } + } +return env_locations[prio]; }
Regards
Stefan
Re: [PATCH V2 01/13] env: Complete generic support for writable list
Quoted 62 lines · Stefan Herbrechtsmeier
Hi,
Am 05.10.2022 um 10:33 schrieb Jan Kiszka:
Quoted 44 lines
From: Jan Kiszka <[email protected]>
This completes what
890feecaab72started by selectingENV_APPENDandENV_IS_NOWHEREand by moving this driver to top if the list. This ensures that load operations pick up both the default env and the permitted parts of the next-prio location. When writing though, we must use the regular ordering.With this change, boards only need to define the list of writable variables but no longer have to provide a custom
env_get_locationimplementation.CC Joe Hershberger <[email protected]>CC Marek Vasut <[email protected]>Signed-off-by Jan Kiszka <[email protected]>---
env/Kconfig | 2 ++
env/env.c | 22 ++++++++++++++++++++++2 files changed, 24 insertions(+)env/Kconfig +0
-0@@ -715,6 +715,8 @@ config ENV_APPENDconfig
ENV_WRITEABLE_LIST
bool "Permit write access only to listed variables" + selectENV_IS_NOWHERE
+ selectENV_APPEND
help
If defined, only environment variables which explicitly set the 'w'
writeable flag can be written and modified at runtime. No variablesenv/env.c +1
-0@@ -133,6 +133,28 @@ __weak enum env_location arch_env_get_location(enum env_operation op, int prio) if (prio >= ARRAY_SIZE(env_locations)) return ENVL_UNKNOWN; + if (IS_ENABLED(CONFIG_ENV_WRITEABLE_LIST)) { + /*env/env.c __weak enum env_location · line 147
Quoted 10 lines
env/env.c +10
-0+* In writeable-list mode, ENVL_NOWHERE gains highest prio by +* virtually injecting it at prio 0. +*/ + if (prio == 0) { + /* + * Avoid the injection for write operations as that + * would block it. + */ + if (op != ENVOP_SAVE && op != ENVOP_ERASE) + return ENVL_NOWHERE;Is it consensus now to use
ENVL_NOWHEREas synonym for default environment? If I remember correct this was rejected in the past and ENVL_NOWHERE should only be used if no enviroment is available.
Quoted 2 lines
Why don't you call
env_set_default(NULL, 0)inenv_load()beforeenv_driver_lookup()?
Worth to explore... if that should avoid this logic here... Let me try.
Jan
Quoted 19 lines
Quoted 13 lines
env/env.c +10
-0+ } else { + /* + * always subtract 1, also for writing because + * env_load_prio, which is used for writing, was + * initialized with that offset. + */ + prio--; + } + } + return env_locations[prio]; }Quoted 1 line
Regards
Stefan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux
Re: [PATCH V2 01/13] env: Complete generic support for writable list
Hi Jan,
Am 27.10.2022 um 14:38 schrieb Jan Kiszka:
Quoted 62 lines · Stefan Herbrechtsmeier
Quoted 61 lines
Hi,
Am 05.10.2022 um 10:33 schrieb Jan Kiszka:
Quoted 44 lines
From: Jan Kiszka <[email protected]>
This completes what
890feecaab72started by selectingENV_APPENDandENV_IS_NOWHEREand by moving this driver to top if the list. This ensures that load operations pick up both the default env and the permitted parts of the next-prio location. When writing though, we must use the regular ordering.With this change, boards only need to define the list of writable variables but no longer have to provide a custom
env_get_locationimplementation.CC Joe Hershberger <[email protected]>CC Marek Vasut <[email protected]>Signed-off-by Jan Kiszka <[email protected]>--- env/Kconfig | 2 ++ env/env.c | 22 ++++++++++++++++++++++ 2 files changed, 24 insertions(+)env/Kconfig +2
-0@@ -715,6 +715,8 @@ config ENV_APPEND config ENV_WRITEABLE_LIST bool "Permit write access only to listed variables" + select ENV_IS_NOWHERE + select ENV_APPEND help If defined, only environment variables which explicitly setthe 'w'
writeable flag can be written and modified at runtime. No
variablesenv/env.c +1
-0@@ -133,6 +133,28 @@ __weak enum env_location arch_env_get_location(enum env_operation op, int prio) if (prio >= ARRAY_SIZE(env_locations)) return ENVL_UNKNOWN; + if (IS_ENABLED(CONFIG_ENV_WRITEABLE_LIST)) { + /*env/env.c __weak enum env_location · line 147
Quoted 10 lines
env/env.c +10
-0+* In writeable-list mode, ENVL_NOWHERE gains highest prio by +* virtually injecting it at prio 0. +*/ + if (prio == 0) { + /* + * Avoid the injection for write operations as that + * would block it. + */ + if (op != ENVOP_SAVE && op != ENVOP_ERASE) + return ENVL_NOWHERE;Is it consensus now to use
ENVL_NOWHEREas synonym for default environment? If I remember correct this was rejected in the past and ENVL_NOWHERE should only be used if no enviroment is available.
Quoted 3 lines
Quoted 2 lines
Why don't you call
env_set_default(NULL, 0)inenv_load()beforeenv_driver_lookup()?Worth to explore... if that should avoid this logic here... Let me try.
Additionally we have remove the #if !CONFIG_IS_ENABLED(ENV_APPEND) arround the return 0` in env_load() and the following change in env_export
env/common.c +0 -0
@@ -234,7 +234,7 @@ int env_export(env_t *env_out)
ssize_t len;
res = (char *)env_out->data;
- len = hexport_r(&env_htab, '\0', 0, &res, ENV_SIZE, 0, NULL); + len = hexport_r(&env_htab, '\0', H_EXTERNAL, &res, ENV_SIZE, 0, NULL); if (len < 0) {
pr_err("Cannot export environment: errno = %d\n", errno); return 1;
I can't remeber why the H_EXTERNAL in env_export was needed.
Regards
Stefan
PATCH v2 2/13 env: Couple networking-related variable flags to CONFIG_NET
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Boards may set networking variables programmatically, thus may have CONFIG_NET on but CONFIG_CMD_NET off. The IOT2050 is an example.
--- env/flags.c | 10 +++++----- include/env_flags.h | 4 ++-- 2 files changed, 7 insertions(+), 7 deletions(-)
env/flags.c +5 -5
@@ -22,7 +22,7 @@#include <env_internal.h> #endif -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET #define ENV_FLAGS_NET_VARTYPE_REPS "im"#else #define ENV_FLAGS_NET_VARTYPE_REPS ""
@@ -57,7 +57,7 @@ static const char * const env_flags_vartype_names[] = {"decimal" ,"hexadecimal" ,"boolean" , -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET"IP address" , "MAC address", #endif
@@ -211,7 +211,7 @@ static void skip_num(int hex, const char *value, const char **end, *end = value; } -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET int eth_validate_ethaddr_str (const char *addr) {const char *end;
@@ -244,7 +244,7 @@ static int _env_flags_validate_type(const char *value, enum env_flags_vartype type) {const char *end; -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET const char *cur;int i;#endif
@@ -273,7 +273,7 @@ static int _env_flags_validate_type(const char *value,if (value[1 ] !='\0' )return -1 ;break ; -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET case env_flags_vartype_ipaddr: cur = value;for (i =0 ; i <4 ; i++) {
include/env_flags.h +2 -2
@@ -12,7 +12,7 @@ enum env_flags_vartype { env_flags_vartype_decimal, env_flags_vartype_hex, env_flags_vartype_bool, -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET env_flags_vartype_ipaddr, env_flags_vartype_macaddr, #endif
@@ -111,7 +111,7 @@ enum env_flags_varaccess env_flags_parse_varaccess(const char *flags);*/ enum env_flags_varaccess env_flags_parse_varaccess_from_binflags (int binflags); -#ifdef CONFIG_CMD_NET +#ifdef CONFIG_NET /* * Check if a string has the format of an Ethernet MAC address */
Signature · 1 line
2.35.3
PATCH v2 3/13 tools: Add script for converting public key into device tree include
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Allows to create a public key device tree dtsi for inclusion into U-Boot SPL and proper during first build already. This can be achieved via CONFIG_DEVICE_TREE_INCLUDES.
--- tools/key2dtsi.py | 64 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 64 insertions(+) create mode 100755 tools/key2dtsi.py
tools/key2dtsi.py new file +64 -0
@@ -0,0 +1,64 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0-only +# +# Public key to dtsi converter. +# +# Copyright (c) Siemens AG, 2022 +# + +from argparse import ArgumentParser, FileType +from os.path import basename, splitext +from Cryptodome.PublicKey import RSA +from Cryptodome.Util.number import inverse + +def int_to_bytestr(n, length=None): + if not length: + length = (n.bit_length() + 7) // 8 + byte_array = n.to_bytes(length, 'big') + return ' '.join(['{:02x}'.format(byte) for byte in byte_array]) + +ap = ArgumentParser(description='Public key to dtsi converter') + +ap.add_argument('--hash', '-H', default='sha256', + help='hash to be used with key (default: sha256)') +ap.add_argument('--required-conf', '-c', action='store_true', + help='mark key required for configuration') +ap.add_argument('--required-image', '-i', action='store_true', + help='mark key required for image') +ap.add_argument('--spl', '-s', action='store_true', + help='mark key for usage in SPL') +ap.add_argument('key_file', metavar='KEY_FILE', type=FileType('r'), + help='key file (formats: X.509, PKCS#1, OpenSSH)') +ap.add_argument('dtsi_file', metavar='DTSI_FILE', type=FileType('w'), + help='dtsi output file') + +args = ap.parse_args() + +key_name, _ = splitext(basename(args.key_file.name)) + +key_data = args.key_file.read() +key = RSA.importKey(key_data) + +r_squared = (2**key.size_in_bits())**2 % key.n +n0_inverse = 2**32 - inverse(key.n, 2**32) + +out = args.dtsi_file +out.write('/ {\n') +out.write('\tsignature {\n') +out.write('\t\tkey-{} {{\n'.format(key_name)) +out.write('\t\t\tkey-name-hint = "{}";\n'.format(key_name)) +out.write('\t\t\talgo = "{},rsa{}";\n'.format(args.hash, key.size_in_bits())) +out.write('\t\t\trsa,num-bits = <{}>;\n'.format(key.size_in_bits())) +out.write('\t\t\trsa,modulus = [{}];\n'.format(int_to_bytestr(key.n))) +out.write('\t\t\trsa,exponent = [{}];\n'.format(int_to_bytestr(key.e, 8))) +out.write('\t\t\trsa,r-squared = [{}];\n'.format(int_to_bytestr(r_squared))) +out.write('\t\t\trsa,n0-inverse = <0x{:x}>;\n'.format(n0_inverse)) +if args.required_conf: + out.write('\t\t\trequired = "conf";\n') +elif args.required_image: + out.write('\t\t\trequired = "image";\n') +if args.spl: + out.write('\t\t\tu-boot,dm-spl;\n') +out.write('\t\t};\n') +out.write('\t};\n') +out.write('};\n')
Signature · 1 line
2.35.3
PATCH v2 4/13 board: siemens: iot2050: Split the build for PG1 and PG2
In reply to Jan KiszkaFrom: Su Baocheng <[email protected]>
Due to different signature keys, the PG1 and the PG2 boards can no longer use the same FSBL (tiboot3). This makes it impossible anyway to maintaine a single flash.bin for both variants, so we can also split the build.
A new target is added to indicates the build is for PG1 vs. PG2 boards. Hence now the variants have separated defconfig files.
The runtime board_is_sr1() check does make no sense anymore, so remove it and replace with build time check.
Documentation is updated accordingly. New binary artifacts are already available via meta-iot2050.
---
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 80 ++++++-------------
board/siemens/iot2050/Kconfig | 28 ++++++-
board/siemens/iot2050/board.c | 12 +--
...ot2050_defconfig => iot2050_pg1_defconfig} | 2 +-
...ot2050_defconfig => iot2050_pg2_defconfig} | 4 +-
doc/board/siemens/iot2050.rst | 15 +++-
6 files changed, 65 insertions(+), 76 deletions(-)
copy configs/{iot2050_defconfig => iot2050_pg1_defconfig} (99%)
rename configs/{iot2050_defconfig => iot2050_pg2_defconfig} (97%)arch/arm/dts/k3-am65-iot2050-boot-image.dtsi +25 -55
@@ -1,6 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 /* - * Copyright (c) Siemens AG, 2020-2021 + * Copyright (c) Siemens AG, 2020-2022 * * Authors: * Jan Kiszka <[email protected]>
@@ -17,7 +17,11 @@ blob-ext@0x000000 { offset = <0x000000>; - filename = "tiboot3.bin"; +#ifdef CONFIG_TARGET_IOT2050_A53_PG1 + filename = "seboot_pg1.bin"; +#else + filename = "seboot_pg2.bin"; +#endif missing-msg = "iot2050-seboot"; };
@@ -43,42 +47,30 @@ }; fdt-iot2050-basic { - description = "k3-am6528-iot2050-basic.dtb"; + description = "k3-am6528-iot2050-basic*.dtb"; type = "flat_dt"; arch = "arm64"; compression = "none"; blob { +#ifdef CONFIG_TARGET_IOT2050_A53_PG1 filename = "arch/arm/dts/k3-am6528-iot2050-basic.dtb"; - }; - }; - - fdt-iot2050-basic-pg2 { - description = "k3-am6528-iot2050-basic-pg2.dtb"; - type = "flat_dt"; - arch = "arm64"; - compression = "none"; - blob { +#else filename = "arch/arm/dts/k3-am6528-iot2050-basic-pg2.dtb"; +#endif }; }; fdt-iot2050-advanced { - description = "k3-am6548-iot2050-advanced.dtb"; + description = "k3-am6548-iot2050-advanced*.dtb"; type = "flat_dt"; arch = "arm64"; compression = "none"; blob { +#ifdef CONFIG_TARGET_IOT2050_A53_PG1 filename = "arch/arm/dts/k3-am6548-iot2050-advanced.dtb"; - }; - }; - - fdt-iot2050-advanced-pg2 { - description = "k3-am6548-iot2050-advanced-pg2.dtb"; - type = "flat_dt"; - arch = "arm64"; - compression = "none"; - blob { +#else filename = "arch/arm/dts/k3-am6548-iot2050-advanced-pg2.dtb"; +#endif }; };
@@ -108,30 +100,12 @@ #endif }; - conf-iot2050-basic-pg2 { - description = "iot2050-basic-pg2"; - firmware = "u-boot"; - fdt = "fdt-iot2050-basic-pg2"; -#ifdef CONFIG_WDT_K3_RTI_FW_FILE - loadables = "k3-rti-wdt-firmware"; -#endif - }; - conf-iot2050-advanced { description = "iot2050-advanced"; firmware = "u-boot"; fdt = "fdt-iot2050-advanced"; #ifdef CONFIG_WDT_K3_RTI_FW_FILE loadables = "k3-rti-wdt-firmware"; -#endif - }; - - conf-iot2050-advanced-pg2 { - description = "iot2050-advanced-pg2"; - firmware = "u-boot"; - fdt = "fdt-iot2050-advanced-pg2"; -#ifdef CONFIG_WDT_K3_RTI_FW_FILE - loadables = "k3-rti-wdt-firmware"; #endif }; };
@@ -150,28 +124,24 @@ fill-byte = [00]; }; - /* PG1 sysfw, basic variant */ + /* sysfw, basic variant */ blob-ext@0x6c0000 { offset = <0x6c0000>; - filename = "sysfw.itb"; +#ifdef CONFIG_TARGET_IOT2050_A53_PG1 + filename = "sysfw_sr1.itb"; +#else + filename = "sysfw_sr2.itb"; +#endif missing-msg = "iot2050-sysfw"; }; - /* PG1 sysfw, advanced variant */ + /* sysfw, advanced variant */ blob-ext@0x740000 { offset = <0x740000>; - filename = "sysfw.itb_HS"; - missing-msg = "iot2050-sysfw"; - }; - /* PG2 sysfw, basic variant */ - blob-ext@0x7c0000 { - offset = <0x7c0000>; - filename = "sysfw_sr2.itb"; - missing-msg = "iot2050-sysfw"; - }; - /* PG2 sysfw, advanced variant */ - blob-ext@0x840000 { - offset = <0x840000>; +#ifdef CONFIG_TARGET_IOT2050_A53_PG1 + filename = "sysfw_sr1.itb_HS"; +#else filename = "sysfw_sr2.itb_HS"; +#endif missing-msg = "iot2050-sysfw"; }; };
board/siemens/iot2050/Kconfig +24 -4
@@ -1,20 +1,40 @@ # SPDX-License-Identifier: GPL-2.0+ # -# Copyright (c) Siemens AG, 2018-2021 +# Copyright (c) Siemens AG, 2018-2022 # # Authors: # Le Jin <[email protected]> # Jan Kiszka <[email protected]> -config TARGET_IOT2050_A53 - bool "IOT2050 running on A53" +choice + prompt "Siemens SIMATIC IOT2050 boards" + optional + +config TARGET_IOT2050_A53_PG1 + bool "IOT2050 PG1 running on A53" + select IOT2050_A53_COMMON + help + This builds U-Boot for the Product Generation 1 (PG1) of the IOT2050 + devices. + +config TARGET_IOT2050_A53_PG2 + bool "IOT2050 PG2 running on A53" + select IOT2050_A53_COMMON + help + This builds U-Boot for the Product Generation 2 (PG2) of the IOT2050 + devices. + +endchoice + +config IOT2050_A53_COMMON + bool select ARM64 select SOC_K3_AM654 select BOARD_LATE_INIT select SYS_DISABLE_DCACHE_OPS select BINMAN -if TARGET_IOT2050_A53 +if IOT2050_A53_COMMON config SYS_BOARD default "iot2050"
board/siemens/iot2050/board.c +2 -10
@@ -55,14 +55,6 @@ static bool board_is_advanced(void)strstr ((char *)info->name,"IOT2050-ADVANCED" ) !=NULL ; } -static bool board_is_sr1 (void ) -{ -struct iot2050_info *info =IOT2050_INFO_DATA ; - -return info->magic ==IOT2050_INFO_MAGIC && - !strstr ((char *)info->name,"-PG2" ); -} -static void remove_mmc1_target (void ) {char *boot_targets =strdup (env_get ("boot_targets" ));
@@ -109,12 +101,12 @@ void set_board_info_env(void) }if (board_is_advanced ()) { -if (board_is_sr1 ()) +if (IS_ENABLED (CONFIG_TARGET_IOT2050_A53_PG1 )) fdtfile ="ti/k3-am6548-iot2050-advanced.dtb" ;else fdtfile ="ti/k3-am6548-iot2050-advanced-pg2.dtb" ; }else { -if (board_is_sr1 ()) +if (IS_ENABLED (CONFIG_TARGET_IOT2050_A53_PG1 )) fdtfile ="ti/k3-am6528-iot2050-basic.dtb" ;else fdtfile ="ti/k3-am6528-iot2050-basic-pg2.dtb" ;
configs/iot2050_pg1_defconfig renamed from configs/iot2050_defconfig +1 -1
@@ -8,7 +8,7 @@ CONFIG_SPL_LIBCOMMON_SUPPORT=y CONFIG_SPL_LIBGENERIC_SUPPORT=y CONFIG_NR_DRAM_BANKS=2 CONFIG_SOC_K3_AM654=y -CONFIG_TARGET_IOT2050_A53=y +CONFIG_TARGET_IOT2050_A53_PG1=y CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0x680000 CONFIG_ENV_SECT_SIZE=0x20000
configs/iot2050_pg2_defconfig renamed from configs/iot2050_defconfig +2 -2
@@ -8,13 +8,13 @@ CONFIG_SPL_LIBCOMMON_SUPPORT=y CONFIG_SPL_LIBGENERIC_SUPPORT=y CONFIG_NR_DRAM_BANKS=2 CONFIG_SOC_K3_AM654=y -CONFIG_TARGET_IOT2050_A53=y +CONFIG_TARGET_IOT2050_A53_PG2=y CONFIG_ENV_SIZE=0x20000 CONFIG_ENV_OFFSET=0x680000 CONFIG_ENV_SECT_SIZE=0x20000 CONFIG_DM_GPIO=y CONFIG_SPL_DM_SPI=y -CONFIG_DEFAULT_DEVICE_TREE="k3-am6528-iot2050-basic" +CONFIG_DEFAULT_DEVICE_TREE="k3-am6528-iot2050-basic-pg2" CONFIG_SPL_TEXT_BASE=0x80080000 CONFIG_SYS_PROMPT="IOT2050> " CONFIG_SPL_SERIAL=y
doc/board/siemens/iot2050.rst +11 -4
@@ -24,9 +24,10 @@ Binary dependencies can be found in https://github.com/siemens/meta-iot2050/tree/master/recipes-bsp/u-boot/files/prebuild. The following binaries from that source need to be present in the build folder: - - tiboot3.bin - - sysfw.itb - - sysfw.itb_HS + - seboot_pg1.bin + - sysfw_sr1.itb + - sysfw_sr1.itb_HS + - seboot_pg2.bin - sysfw_sr2.itb - sysfw_sr2.itb_HS
@@ -57,7 +58,13 @@ U-Boot: $ export ATF=/path/to/bl31.bin $ export TEE=/path/to/tee-pager_v2.bin - $ make iot2050_defconfig + + # configure for PG1 + $ make iot2050_pg1_defconfig + + # or configure for PG2 + $ make iot2050_pg2_defconfig + $ make Flashing
Signature · 1 line
2.35.3
PATCH v2 5/13 arm: dts: iot2050: Use the auto generator nodes for fdt
In reply to Jan KiszkaFrom: Su Baocheng <[email protected]>
Refactor according to the entry fit: Entry containing a FIT of document tools/binman/README.entries.
As the generator uses the device tree name for the config description, board_fit_config_name_match requires a small adjustment as well.
--- arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 44 ++++---------------- board/siemens/iot2050/board.c | 3 ++ configs/iot2050_pg1_defconfig | 1 + configs/iot2050_pg2_defconfig | 1 + 4 files changed, 12 insertions(+), 37 deletions(-)
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi +7 -37
@@ -32,6 +32,7 @@ fit@0x280000 { description = "U-Boot for IOT2050"; + fit,fdt-list = "of-list"; offset = <0x280000>; images { u-boot {
@@ -46,32 +47,11 @@ }; }; - fdt-iot2050-basic { - description = "k3-am6528-iot2050-basic*.dtb"; + @fdt-SEQ { + description = "fdt-NAME"; type = "flat_dt"; arch = "arm64"; compression = "none"; - blob { -#ifdef CONFIG_TARGET_IOT2050_A53_PG1 - filename = "arch/arm/dts/k3-am6528-iot2050-basic.dtb"; -#else - filename = "arch/arm/dts/k3-am6528-iot2050-basic-pg2.dtb"; -#endif - }; - }; - - fdt-iot2050-advanced { - description = "k3-am6548-iot2050-advanced*.dtb"; - type = "flat_dt"; - arch = "arm64"; - compression = "none"; - blob { -#ifdef CONFIG_TARGET_IOT2050_A53_PG1 - filename = "arch/arm/dts/k3-am6548-iot2050-advanced.dtb"; -#else - filename = "arch/arm/dts/k3-am6548-iot2050-advanced-pg2.dtb"; -#endif - }; }; #ifdef CONFIG_WDT_K3_RTI_FW_FILE
@@ -89,21 +69,11 @@ }; configurations { - default = "conf-iot2050-basic"; - - conf-iot2050-basic { - description = "iot2050-basic"; - firmware = "u-boot"; - fdt = "fdt-iot2050-basic"; -#ifdef CONFIG_WDT_K3_RTI_FW_FILE - loadables = "k3-rti-wdt-firmware"; -#endif - }; - - conf-iot2050-advanced { - description = "iot2050-advanced"; + default = "@config-DEFAULT-SEQ"; + @config-SEQ { + description = "NAME"; firmware = "u-boot"; - fdt = "fdt-iot2050-advanced"; + fdt = "fdt-SEQ"; #ifdef CONFIG_WDT_K3_RTI_FW_FILE loadables = "k3-rti-wdt-firmware"; #endif
board/siemens/iot2050/board.c +3 -0
@@ -154,6 +154,9 @@ int board_fit_config_name_match(const char *name)struct iot2050_info *info =IOT2050_INFO_DATA ;char upper_name[32 ]; +/* skip the prefix "k3-am65x8-" */ + name +=10 ; +if (info->magic !=IOT2050_INFO_MAGIC ||strlen (name) >=sizeof (upper_name))return -1 ;
configs/iot2050_pg1_defconfig +1 -0
@@ -69,6 +69,7 @@ CONFIG_CMD_TIME=y # CONFIG_ISO_PARTITION is not set CONFIG_OF_CONTROL=y CONFIG_SPL_OF_CONTROL=y +CONFIG_OF_LIST="k3-am6528-iot2050-basic k3-am6548-iot2050-advanced" CONFIG_SPL_MULTI_DTB_FIT=y CONFIG_SPL_OF_LIST="k3-am65-iot2050-spl" CONFIG_SPL_MULTI_DTB_FIT_NO_COMPRESSION=y
configs/iot2050_pg2_defconfig +1 -0
@@ -69,6 +69,7 @@ CONFIG_CMD_TIME=y # CONFIG_ISO_PARTITION is not set CONFIG_OF_CONTROL=y CONFIG_SPL_OF_CONTROL=y +CONFIG_OF_LIST="k3-am6528-iot2050-basic-pg2 k3-am6548-iot2050-advanced-pg2" CONFIG_SPL_MULTI_DTB_FIT=y CONFIG_SPL_OF_LIST="k3-am65-iot2050-spl" CONFIG_SPL_MULTI_DTB_FIT_NO_COMPRESSION=y
Signature · 1 line
2.35.3
PATCH v2 6/13 iot2050: Update firmware layout
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
The latest version of the binary-only firmware parts come in a combined form of FSBL and sysfw containers. This implies some layout changes to the generated firmware image but also makes handling of artifacts much simpler (4 files less). The env locations will not change, just the space reserved for U-Boot will shrink from 4 to 3 MB - still plenty of space left in practice.
Adjust configuration and documentation accordingly.
Along this change, add a new reservation for update commands of the user-controlled OTP part. A specific userspace tool will fill it, and the FSBL will evaluate it during boot. This reservation will use 64K of the former sysfw section.
--- arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 30 ++++++-------------- configs/iot2050_pg1_defconfig | 2 +- configs/iot2050_pg2_defconfig | 2 +- doc/board/siemens/iot2050.rst | 4 --- tools/binman/missing-blob-help | 8 +----- 5 files changed, 11 insertions(+), 35 deletions(-)
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi +8 -22
@@ -25,15 +25,15 @@ missing-msg = "iot2050-seboot"; }; - blob@0x080000 { - offset = <0x080000>; + blob@0x180000 { + offset = <0x180000>; filename = "tispl.bin"; }; - fit@0x280000 { + fit@0x380000 { description = "U-Boot for IOT2050"; fit,fdt-list = "of-list"; - offset = <0x280000>; + offset = <0x380000>; images { u-boot { description = "U-Boot";
@@ -94,25 +94,11 @@ fill-byte = [00]; }; - /* sysfw, basic variant */ - blob-ext@0x6c0000 { + /* OTP update command block */ + fill@0x6c0000 { offset = <0x6c0000>; -#ifdef CONFIG_TARGET_IOT2050_A53_PG1 - filename = "sysfw_sr1.itb"; -#else - filename = "sysfw_sr2.itb"; -#endif - missing-msg = "iot2050-sysfw"; - }; - /* sysfw, advanced variant */ - blob-ext@0x740000 { - offset = <0x740000>; -#ifdef CONFIG_TARGET_IOT2050_A53_PG1 - filename = "sysfw_sr1.itb_HS"; -#else - filename = "sysfw_sr2.itb_HS"; -#endif - missing-msg = "iot2050-sysfw"; + size = <0x010000>; + fill-byte = [ff]; }; }; };
configs/iot2050_pg1_defconfig +1 -1
@@ -52,7 +52,7 @@ CONFIG_SPL_POWER_DOMAIN=y # CONFIG_SPL_SPI_FLASH_TINY is not set CONFIG_SPL_SPI_FLASH_SFDP_SUPPORT=y CONFIG_SPL_SPI_LOAD=y -CONFIG_SYS_SPI_U_BOOT_OFFS=0x280000 +CONFIG_SYS_SPI_U_BOOT_OFFS=0x380000 CONFIG_SYS_MAXARGS=64 CONFIG_SYS_PBSIZE=1050 CONFIG_CMD_ASKENV=y
configs/iot2050_pg2_defconfig +1 -1
@@ -52,7 +52,7 @@ CONFIG_SPL_POWER_DOMAIN=y # CONFIG_SPL_SPI_FLASH_TINY is not set CONFIG_SPL_SPI_FLASH_SFDP_SUPPORT=y CONFIG_SPL_SPI_LOAD=y -CONFIG_SYS_SPI_U_BOOT_OFFS=0x280000 +CONFIG_SYS_SPI_U_BOOT_OFFS=0x380000 CONFIG_SYS_MAXARGS=64 CONFIG_SYS_PBSIZE=1050 CONFIG_CMD_ASKENV=y
doc/board/siemens/iot2050.rst +0 -4
@@ -25,11 +25,7 @@ https://github.com/siemens/meta-iot2050/tree/master/recipes-bsp/u-boot/files/pre The following binaries from that source need to be present in the build folder: - seboot_pg1.bin - - sysfw_sr1.itb - - sysfw_sr1.itb_HS - seboot_pg2.bin - - sysfw_sr2.itb - - sysfw_sr2.itb_HS Building --------
tools/binman/missing-blob-help +1 -7
@@ -21,13 +21,7 @@ Please read the section on SCP firmware in board/sunxi/README.sunxi64 iot2050-seboot: See the documentation for IOT2050 board. Your image is missing SEBoot which is mandatory for board startup. Prebuilt SEBoot located at -meta-iot2050/tree/master/recipes-bsp/u-boot/files/prebuild/tiboot3.bin. - -iot2050-sysfw: -See the documentation for IOT2050 board. Your image is missing system -firmware which is mandatory for board startup. Prebuilt system firmware -located at meta-iot2050/tree/master/recipes-bsp/u-boot/files/prebuild/ -with sysfw prefix. +meta-iot2050/tree/master/recipes-bsp/u-boot/files/prebuild/seboot_pg*.bin. k3-rti-wdt-firmware: If CONFIG_WDT_K3_RTI_LOAD_FW is enabled, a firmware image is needed for
Signature · 1 line
2.35.3
PATCH v2 7/13 iot2050: Add watchdog start to bootcmd
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Allows run-time control over watchdog auto-start and the timeout via setting the environment variable watchdog_timeout_ms. A value of zero means "do not start". Use CONFIG_WATCHDOG_TIMEOUT_MSECS as initial value and this to zero by default. Users can then enable the watchdog once the use and OS which picks it up during boot.
--- configs/iot2050_pg1_defconfig | 2 ++ configs/iot2050_pg2_defconfig | 2 ++ include/configs/iot2050.h | 9 +++++++++ 3 files changed, 13 insertions(+)
configs/iot2050_pg1_defconfig +2 -0
@@ -32,6 +32,7 @@ CONFIG_OF_BOARD_SETUP=y CONFIG_BOOTSTAGE=y CONFIG_SHOW_BOOT_PROGRESS=y CONFIG_SPL_SHOW_BOOT_PROGRESS=y +CONFIG_BOOTCOMMAND="run start_watchdog; run distro_bootcmd" CONFIG_CONSOLE_MUX=y # CONFIG_DISPLAY_CPUINFO is not set CONFIG_SPL_MAX_SIZE=0x58000
@@ -142,6 +143,7 @@ CONFIG_USB_DWC3_GENERIC=y CONFIG_USB_KEYBOARD=y # CONFIG_WATCHDOG is not set # CONFIG_WATCHDOG_AUTOSTART is not set +CONFIG_WATCHDOG_TIMEOUT_MSECS=0 CONFIG_WDT=y CONFIG_WDT_K3_RTI=y CONFIG_WDT_K3_RTI_LOAD_FW=y
configs/iot2050_pg2_defconfig +2 -0
@@ -32,6 +32,7 @@ CONFIG_OF_BOARD_SETUP=y CONFIG_BOOTSTAGE=y CONFIG_SHOW_BOOT_PROGRESS=y CONFIG_SPL_SHOW_BOOT_PROGRESS=y +CONFIG_BOOTCOMMAND="run start_watchdog; run distro_bootcmd" CONFIG_CONSOLE_MUX=y # CONFIG_DISPLAY_CPUINFO is not set CONFIG_SPL_MAX_SIZE=0x58000
@@ -142,6 +143,7 @@ CONFIG_USB_DWC3_GENERIC=y CONFIG_USB_KEYBOARD=y # CONFIG_WATCHDOG is not set # CONFIG_WATCHDOG_AUTOSTART is not set +CONFIG_WATCHDOG_TIMEOUT_MSECS=0 CONFIG_WDT=y CONFIG_WDT_K3_RTI=y CONFIG_WDT_K3_RTI_LOAD_FW=y
include/configs/iot2050.h +9 -0
@@ -15,6 +15,14 @@/* SPL Loader Configuration */ +#define WATCHDOG_ENV \ + "watchdog_timeout_ms=" __stringify(CONFIG_WATCHDOG_TIMEOUT_MSECS) "\0" \ + "start_watchdog=if test ${watchdog_timeout_ms} -gt 0; then " \ + "wdt dev watchdog@40610000; " \ + "wdt start ${watchdog_timeout_ms}; " \ + "echo Watchdog started, timeout ${watchdog_timeout_ms} ms; " \ + "fi\0" +/* U-Boot general configuration */ #define EXTRA_ENV_IOT2050_BOARD_SETTINGS \ "usb_pgood_delay=900\0"
@@ -43,6 +51,7 @@#define CONFIG_EXTRA_ENV_SETTINGS \ DEFAULT_LINUX_BOOT_ENV \ BOOTENV \ + WATCHDOG_ENV \ EXTRA_ENV_IOT2050_BOARD_SETTINGS#include <configs/ti_armv7_common.h>
Signature · 1 line
2.35.3
PATCH v2 8/13 iot2050: Add CONFIG_ENV_FLAGS_LIST_STATIC
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Will be needed when CONFIG_ENV_WRITEABLE_LIST is enabled. The listed variables shall remain writable, for informational purposes - they have to be considered untrusted because the persistent U-Boot env is not protected.
--- include/configs/iot2050.h | 8 ++++++++ 1 file changed, 8 insertions(+)
include/configs/iot2050.h +8 -0
@@ -56,4 +56,12 @@#include <configs/ti_armv7_common.h> +#ifdef CONFIG_ENV_WRITEABLE_LIST +/* relevant for secure boot with CONFIG_ENV_WRITEABLE_LIST=y */ +#define CONFIG_ENV_FLAGS_LIST_STATIC \ + "board_uuid:sw,board_name:sw,board_serial:sw,board_a5e:sw," \ + "mlfb:sw,fw_version:sw,seboot_version:sw," \ + "eth1addr:mw,eth2addr:mw,watchdog_timeout_ms:dw,boot_targets:sw" +#endif +#endif /* __CONFIG_IOT2050_H */
Signature · 1 line
2.35.3
PATCH v2 9/13 arm: dts: iot2050: Allow verifying U-Boot proper by SPL
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Add hashes and configuration signature stubs to prepare verified boot of main U-Boot by SPL.
--- arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 16 ++++++++++++++++ 1 file changed, 16 insertions(+)
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi +16 -0
@@ -14,6 +14,7 @@ filename = "flash.bin"; pad-byte = <0xff>; size = <0x8c0000>; + allow-repack; blob-ext@0x000000 { offset = <0x000000>;
@@ -45,6 +46,9 @@ entry = <0x80800000>; u-boot-nodtb { }; + hash { + algo = "sha256"; + }; }; @fdt-SEQ {
@@ -52,6 +56,9 @@ type = "flat_dt"; arch = "arm64"; compression = "none"; + hash { + algo = "sha256"; + }; }; #ifdef CONFIG_WDT_K3_RTI_FW_FILE
@@ -64,6 +71,9 @@ filename = CONFIG_WDT_K3_RTI_FW_FILE; missing-msg = "k3-rti-wdt-firmware"; }; + hash { + algo = "sha256"; + }; }; #endif };
@@ -77,10 +87,16 @@ #ifdef CONFIG_WDT_K3_RTI_FW_FILE loadables = "k3-rti-wdt-firmware"; #endif + signature { + sign-images = "firmware", "fdt", "loadables"; + }; }; }; }; + fdtmap { + }; + /* primary env */ fill@0x680000 { offset = <0x680000>;
Signature · 1 line
2.35.3
PATCH v2 10/13 iot2050: Add script for signing artifacts
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
There are many ways to get a signed firmware for the IOT2050 devices, namely for the parts under user-control. This script documents one way of doing it, given a signing key. Augment the board documentation with the required procedure around it.
--- doc/board/siemens/iot2050.rst | 52 +++++++++++++++++++++++++++++++++++ tools/iot2050-sign-fw.sh | 51 ++++++++++++++++++++++++++++++++++ 2 files changed, 103 insertions(+) create mode 100755 tools/iot2050-sign-fw.sh
doc/board/siemens/iot2050.rst +52 -0
@@ -79,3 +79,55 @@ Via external programmer Dediprog SF100 or SF600: .. code-block:: text $ dpcmd --vcc 2 -v -u flash.bin + +Signing (optional) +------------------ + +To enable verified boot for the firmware artifacts after the Siemens-managed +first-stage loader (seboot_pg*.bin), the following steps need to be taken +before and after the build: + +Generate dtsi holding the public key +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +.. code-block:: text + + tools/key2dtsi.py -c -s key.pem public-key.dtsi + +This will be used to embed the public key into U-Boot SPL and main so that each +step can validate signatures of the succeeding one. + +Adjust U-Boot configuration +^^^^^^^^^^^^^^^^^^^^^^^^^^^ + +Enabled at least the following options in U-Boot: + +.. code-block:: text + + CONFIG_SPL_FIT_SIGNATURE=y + CONFIG_DEVICE_TREE_INCLUDES="/path/to/public-key.dtsi" + CONFIG_RSA=y + +Note that there are more configuration changes needed in order to lock-down +the command line and the boot process of U-Boot for secure scenarios. These are +not in scope here. + +Build U-Boot +^^^^^^^^^^^^ + +See related section above. + +Sign flash.bin +^^^^^^^^^^^^^^ + +In the build folder still containing artifacts from step 3, invoke: + +.. code-block:: text + + tools/iot2050-sign-fw.sh /path/to/key.pem + +Flash signed flash.bin +^^^^^^^^^^^^^^^^^^^^^^ + +The signing has happen in-place in flash.bin, thus the flashing procedure +described above.
tools/iot2050-sign-fw.sh new file +51 -0
51 lines not on this page · raw message
Signature · 1 line
2.35.3
Re: [PATCH V2 10/13] iot2050: Add script for signing artifacts
Hi Jan,
Quoted 8 lines · Jan Kiszka
From: Jan Kiszka <[email protected]>
There are many ways to get a signed firmware for the IOT2050 devices, namely for the parts under user-control. This script documents one way of doing it, given a signing key. Augment the board documentation with the required procedure around it.
Quoted 6 lines
Signed-off-by Jan Kiszka <[email protected]>--- doc/board/siemens/iot2050.rst | 52 +++++++++++++++++++++++++++++++++++ tools/iot2050-sign-fw.sh | 51 ++++++++++++++++++++++++++++++++++ 2 files changed, 103 insertions(+) create mode 100755 tools/iot2050-sign-fw.sh
Please use binman for this. You can create new entry type for your needs. We want to avoid adding arch-specific scripts with no tests.
Regards,
Simon
Re: [PATCH V2 10/13] iot2050: Add script for signing artifacts
Quoted 18 lines · Simon Glass
Hi Jan,
Quoted 8 lines · Jan Kiszka
From: Jan Kiszka <[email protected]>
There are many ways to get a signed firmware for the IOT2050 devices, namely for the parts under user-control. This script documents one way of doing it, given a signing key. Augment the board documentation with the required procedure around it.
Quoted 6 lines
Signed-off-by Jan Kiszka <[email protected]>--- doc/board/siemens/iot2050.rst | 52 +++++++++++++++++++++++++++++++++++ tools/iot2050-sign-fw.sh | 51 ++++++++++++++++++++++++++++++++++ 2 files changed, 103 insertions(+) create mode 100755 tools/iot2050-sign-fw.sh
Quoted 2 lines
Please use binman for this. You can create new entry type for your needs. We want to avoid adding arch-specific scripts with no tests.
We will need a script in the foreseeable future, even when binman should be fixed /wrt replace - see how the certs need to be set up.
Jan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux
Re: [PATCH V2 10/13] iot2050: Add script for signing artifacts
Quoted 19 lines · Jan Kiszka
Quoted 18 lines · Simon Glass
Hi Jan,
Quoted 14 lines · Jan Kiszka
From: Jan Kiszka <[email protected]>
There are many ways to get a signed firmware for the IOT2050 devices, namely for the parts under user-control. This script documents one way of doing it, given a signing key. Augment the board documentation with the required procedure around it.
Signed-off-by Jan Kiszka <[email protected]>--- doc/board/siemens/iot2050.rst | 52 +++++++++++++++++++++++++++++++++++ tools/iot2050-sign-fw.sh | 51 ++++++++++++++++++++++++++++++++++ 2 files changed, 103 insertions(+) create mode 100755 tools/iot2050-sign-fw.sh
Quoted 5 lines
Quoted 2 lines
Please use binman for this. You can create new entry type for your needs. We want to avoid adding arch-specific scripts with no tests.
We will need a script in the foreseeable future, even when binman should be fixed /wrt replace - see how the certs need to be set up.
...and 'binman replace' is still broken:
# source/tools/binman/binman replace -i flash.bin -f tispl.bin_signed blob@0x180000 binman: Error 1 running 'mkimage -t -F /tmp/binman.ip2gc0oy/[email protected]': Usage: mkimage -l image -l ==> list image header information mkimage [-x] -A arch -O os -T type -C comp -a addr -e ep -n name -d data_file[:data_file...] image -A ==> set architecture to 'arch' -O ==> set operating system to 'os' -T ==> set image type to 'type' -C ==> set compression type 'comp' -a ==> set load address to 'addr' (hex) -e ==> set entry point to 'ep' (hex) -n ==> set image name to 'name' -d ==> use image data from 'datafile' -x ==> set XIP (execute in place) mkimage [-D dtc_options] [-f fit-image.its|-F] fit-image -D => set options for device tree compiler -f => input filename for FIT source Signing / verified boot options: [-k keydir] [-K dtb] [ -c <comment>] [-r] -k => set directory containing private keys -K => write public keys to this .dtb file -c => add comment in signature node -F => re-sign existing FIT image -r => mark keys used as 'required' in dtb mkimage -V ==> print version information and exit
Jan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux
PATCH v2 11/13 arm: dts: iot2050: Optionally embed OTP programming data into image
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
Use external blob otpcmd.bin to replace the 0xff filled OTP programming command block to create a firmware image that provisions the OTP on first boot. This otpcmd.bin is generated from the customer keys using steps described in the meta-iot2050 integration layer for the device.
Based on original patch by Baocheng Su.
--- arch/arm/dts/k3-am65-iot2050-boot-image.dtsi | 8 ++++++++ board/siemens/iot2050/Kconfig | 7 +++++++ doc/board/siemens/iot2050.rst | 8 ++++++++ tools/binman/missing-blob-help | 8 ++++++++ 4 files changed, 31 insertions(+)
arch/arm/dts/k3-am65-iot2050-boot-image.dtsi +8 -0
@@ -111,10 +111,18 @@ }; /* OTP update command block */ +#if CONFIG_IOT2050_EMBED_OTPCMD + blob-ext@0x6c0000 { + offset = <0x6c0000>; + size = <0x010000>; + filename = "otpcmd.bin"; + missing-msg = "iot2050-otpcmd"; +#else fill@0x6c0000 { offset = <0x6c0000>; size = <0x010000>; fill-byte = [ff]; +#endif }; }; };
board/siemens/iot2050/Kconfig +7 -0
@@ -49,4 +49,11 @@ config IOT2050_BOOT_SWITCH bool "Disable eMMC boot via USER button (Advanced version only)" default y +config IOT2050_EMBED_OTPCMD + bool "Embed OTP programming data" + help + Embed signed OTP programming data 'otpcmd.bin' into the firmware + image. This data will be evaluated and executed on first boot of the + device. + endif
doc/board/siemens/iot2050.rst +8 -0
@@ -27,6 +27,14 @@ The following binaries from that source need to be present in the build folder: - seboot_pg1.bin - seboot_pg2.bin +For building an image containing the OTP key provisioning data, below binary +needs to be present in the build folder: + + - otpcmd.bin + +Regarding how to generating this otpcmd.bin, please refer to: +https://github.com/siemens/meta-iot2050/tree/master/recipes-bsp/secure-boot-otp-provisioning/files/make-otpcmd.sh + Building --------
tools/binman/missing-blob-help +8 -0
@@ -23,6 +23,14 @@ See the documentation for IOT2050 board. Your image is missing SEBoot which is mandatory for board startup. Prebuilt SEBoot located at meta-iot2050/tree/master/recipes-bsp/u-boot/files/prebuild/seboot_pg*.bin. +iot2050-otpcmd: +See the documentation for IOT2050 board. Your image is missing OTP command data +block which is used for provisioning the customer keys to the board. +Please refer to +meta-iot2050/tree/master/recipes-bsp/secure-boot-otp-provisioning/files/make-otpcmd.sh +for how to generate this binary. If you are not using secure boot or do not +intend to provision the keys, disable CONFIG_IOT2050_EMBED_OTPCMD. + k3-rti-wdt-firmware: If CONFIG_WDT_K3_RTI_LOAD_FW is enabled, a firmware image is needed for the R5F core(s) to trigger the system reset. One possible source is
Signature · 1 line
2.35.3
PATCH v2 12/13 doc: iot2050: Add a note about the watchdog firmware
In reply to Jan KiszkaFrom: Jan Kiszka <[email protected]>
This is enabled by default, thus should be described as well.
--- doc/board/siemens/iot2050.rst | 4 ++++ 1 file changed, 4 insertions(+)
doc/board/siemens/iot2050.rst +4 -0
@@ -27,6 +27,10 @@ The following binaries from that source need to be present in the build folder: - seboot_pg1.bin - seboot_pg2.bin +When using the watchdog, a related firmware for the R5 core(s) is needed, e.g. +https://github.com/siemens/k3-rti-wdt. The name and location of the image is +configured via CONFIG_WDT_K3_RTI_FW_FILE. + For building an image containing the OTP key provisioning data, below binary needs to be present in the build folder:
Signature · 1 line
2.35.3
PATCH v2 13/13 board: siemens: iot2050: use the named gpio to control the user-button
In reply to Jan KiszkaFrom: chao zeng <[email protected]>
User-button is controlled by the mcu domain gpio number 25. But main0 main1 mcu domain all have gpio number 25.
To identify where the gpio is from, Using gpio controll base as the prefix to indicate the gpio resource.
--- board/siemens/iot2050/board.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
board/siemens/iot2050/board.c +1 -1
@@ -183,7 +183,7 @@ static bool user_button_pressed(void)memset (&gpio,0 ,sizeof (gpio)); -if (dm_gpio_lookup_name ("25" , &gpio) <0 || +if (dm_gpio_lookup_name ("gpio@42110000_25" , &gpio) <0 ||dm_gpio_request (&gpio,"USER button" ) <0 ||dm_gpio_set_dir_flags (&gpio,GPIOD_IS_IN ) <0 )return false;
Signature · 1 line
2.35.3
Quoted 14 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly brings board-specific changes such as:
- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET(repost)
- tools: Add script for converting public key into device tree include
Quoted 8 lines
Changes in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Any remaining questions/remarks for this series? Asking also as we would like to move forward with upstreaming that M.2 support which depends on this one here.
Jan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux
Quoted 20 lines · Jan Kiszka
Quoted 19 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly brings board-specific changes such as:
- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET(repost)
- tools: Add script for converting public key into device tree includeChanges in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"
Quoted 6 lines
Quoted 2 lines
Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Any remaining questions/remarks for this series? Asking also as we would like to move forward with upstreaming that M.2 support which depends on this one here.
Marek, I believe you had concerns about the environment side of v1 here, is v2 more to your liking? Thanks!
Signature · 1 line
Tom
Quoted 24 lines · Tom Rini
Quoted 23 lines · Jan Kiszka
Quoted 21 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly brings board-specific changes such as:
- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET(repost)
- tools: Add script for converting public key into device tree includeChanges in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Quoted 6 lines
Quoted 3 lines
Any remaining questions/remarks for this series? Asking also as we would like to move forward with upstreaming that M.2 support which depends on this one here.
Marek, I believe you had concerns about the environment side of v1 here, is v2 more to your liking? Thanks!
Nothing changed in V2 in that aspect as far as I can tell.
Quoted 28 lines · Marek Vasut
Quoted 27 lines · Tom Rini
Quoted 26 lines · Jan Kiszka
Quoted 24 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly
brings board-specific changes such as:- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET(repost)
- tools: Add script for converting public key into device tree includeChanges in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Quoted 8 lines
Quoted 6 lines
Quoted 3 lines
Any remaining questions/remarks for this series? Asking also as we would like to move forward with upstreaming that M.2 support which depends on this one here.
Marek, I believe you had concerns about the environment side of v1 here, is v2 more to your liking? Thanks!
Nothing changed in V2 in that aspect as far as I can tell.
Well... patch 1 changes significantly and is much less invasive - as we discussed around v1. Can you please update your comments then? I'm lost now what else you would like to see changed.
Jan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux
Quoted 33 lines · Jan Kiszka
Quoted 32 lines · Marek Vasut
Quoted 31 lines · Tom Rini
Quoted 29 lines · Jan Kiszka
Quoted 17 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly
brings board-specific changes such as:- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET
(repost)
- tools: Add script for converting public key into device tree
includeQuoted 7 lines
Changes in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Any remaining questions/remarks for this series? Asking also as we would like to move forward with upstreaming that M.2 support which depends on this one here.
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
Marek, I believe you had concerns about the environment side of v1 here, is v2 more to your liking? Thanks!
Nothing changed in V2 in that aspect as far as I can tell.
Well... patch 1 changes significantly and is much less invasive - as we discussed around v1. Can you please update your comments then? I'm lost now what else you would like to see changed.
IIRC my comment was that env prio ordering is board specific and so it should be in board/ , wasn't it ?
Quoted 41 lines · Marek Vasut
Quoted 40 lines · Jan Kiszka
Quoted 39 lines · Marek Vasut
Quoted 37 lines · Tom Rini
Quoted 24 lines · Jan Kiszka
Quoted 23 lines · Jan Kiszka
(Almost) flushing our upstream queue for the IOT2050 device, this mostly
brings board-specific changes such as:- updated build process and firmware layout for PG1 vs. PG2 devices
- more watchdog preparations
- preparations for verified boot on IOT2050 Advanced devicesThere are also some generic extensions in this series which are dependencies for the above:
- env: Complete generic support for writable list
- env: Couple networking-related variable flags toCONFIG_NET(repost)
- tools: Add script for converting public key into device tree includeChanges in v2:
- rebased over latest master
- reworked patch 1 to be less invasive to the code
- added "iot2050: use the named gpio to control the user-button"Quoted 8 lines
Quoted 2 lines
Still in our backlog is support for a new variant that comes with M.2 slots. But that requires some more rework first.
Any remaining questions/remarks for this series? Asking also as we would
like to move forward with upstreaming that M.2 support which depends on
this one here.Marek, I believe you had concerns about the environment side of v1 here,
is v2 more to your liking? Thanks!
Quoted 8 lines
Quoted 5 lines
Quoted 1 line
Nothing changed in V2 in that aspect as far as I can tell.
Well... patch 1 changes significantly and is much less invasive - as we discussed around v1. Can you please update your comments then? I'm lost now what else you would like to see changed.
IIRC my comment was that env prio ordering is board specific and so it should be in board/ , wasn't it ?
Please re-read the old discussion to its end. It ended with how to make the approach less invasive.
The whole patch is about making it no longer board specific - because it isn't for the variable protection scenario under secure boot. As I pointed out in that thread, we have multiple downstream boards using exactly this very same pattern already.
Thanks,
Jan
Signature · 2 lines
Siemens AG, Technology Competence Center Embedded Linux