February 2024 - U-Boot

[ANN] U-Boot v2024.04-rc2 released
by Tom Rini 14 Feb '24

14 Feb '24

Hey all, I'm a day late as I missed my calendar reminder, but here's -rc2 now. I am hopeful really just about everything that needs to come in, is now in, but I also know I have at least one more series to grab for TI platforms. And I'll grab the series I posted today that should fix the current CI failure that's related to not being able to download a file from Google drive. In terms of a changelog, git log --merges v2024.01-rc1..v2024.04-rc2 contains what I've pulled but as always, better PR messages and tags will provide better results here. I hope to get back on schedule and that means the rest of the rcs every other Monday, and with final release on Tuesday, April 2nd, 2024. Thanks all! -- Tom

1 0

[PATCH v1] Revert "board: verdin-am62: set cpu core voltage depending on speed grade"
by Francesco Dolcini 13 Feb '24

13 Feb '24

From: Francesco Dolcini <francesco.dolcini(a)toradex.com> This reverts commit d2099587d661c6ca2309256c0e04c06e26c8d34c. According to TI changing the VDD_CORE while the SoC is running is not allowed, the voltage must be set before the AM62 device reset is released, revert this change therefore. The correct solution would be to program the PMIC during manufactoring according to the speed grade of the SoC. Link: https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1… Fixes: d2099587d661 ("board: verdin-am62: set cpu core voltage depending on speed grade") Signed-off-by: Francesco Dolcini <francesco.dolcini(a)toradex.com> --- board/toradex/verdin-am62/verdin-am62.c | 47 ------------------------- 1 file changed, 47 deletions(-) diff --git a/board/toradex/verdin-am62/verdin-am62.c b/board/toradex/verdin-am62/verdin-am62.c index 395eb365a0bf..e948fc16ba99 100644 --- a/board/toradex/verdin-am62/verdin-am62.c +++ b/board/toradex/verdin-am62/verdin-am62.c @@ -14,13 +14,10 @@ #include <fdt_support.h> #include <init.h> #include <k3-ddrss.h> -#include <power/regulator.h> #include <spl.h> #include "../common/tdx-cfg-block.h" -#define VDD_CORE_REG "buck1" - DECLARE_GLOBAL_DATA_PTR; int board_init(void) @@ -53,37 +50,9 @@ int board_fit_config_name_match(const char *name) } #endif -static u32 get_vdd_core_nominal(void) -{ - int core_uvolt; - - switch (k3_get_speed_grade()) { - case 'G': - case 'K': - case 'S': - core_uvolt = 750000; - break; - case 'T': - default: - core_uvolt = 850000; - break; - } - return core_uvolt; -} - #if IS_ENABLED(CONFIG_OF_LIBFDT) && IS_ENABLED(CONFIG_OF_BOARD_SETUP) int ft_board_setup(void *blob, struct bd_info *bd) { - int core_uvolt; - - core_uvolt = get_vdd_core_nominal(); - if (core_uvolt != 850000) { - do_fixup_by_path_u32(blob, "/bus@f0000/i2c@20000000/pmic@30/regulators/buck1", - "regulator-max-microvolt", core_uvolt, 0); - do_fixup_by_path_u32(blob, "/bus@f0000/i2c@20000000/pmic@30/regulators/buck1", - "regulator-min-microvolt", core_uvolt, 0); - } - return ft_common_board_setup(blob, bd); } #endif @@ -118,22 +87,6 @@ static void select_dt_from_module_version(void) int board_late_init(void) { - int ret; - int core_uvolt; - struct udevice *dev = NULL; - - core_uvolt = get_vdd_core_nominal(); - if (core_uvolt != 850000) { - /* Set CPU core voltage to 0.75V for slower speed grades */ - ret = regulator_get_by_devname(VDD_CORE_REG, &dev); - if (ret) - pr_err("VDD CORE Regulator get error: %d\n", ret); - - ret = regulator_set_value_force(dev, core_uvolt); - if (ret) - pr_err("VDD CORE Regulator value setting error: %d\n", ret); - } - select_dt_from_module_version(); return 0; -- 2.39.2

2 1

[PATCH v1] cmd: hash: fix param count check
by Igor Opaniuk 13 Feb '24

13 Feb '24

From: Igor Opaniuk <igor.opaniuk(a)gmail.com> Add correct check for parameter count. This fixes this issue when `hash` cmd is invoked without params: => hash data abort pc : [<bf739204>] lr : [<ba6effa8>] reloc pc : [<60019204>] lr : [<5afcffa8>] sp : ba6dd9c8 ip : bf7391f0 fp : bf74ec14 r10: 00000001 r9 : ba6dfea0 r8 : bf7ea030 r7 : 00000000 r6 : ba6effa8 r5 : 00000000 r4 : ffffffff r3 : bf7c257c r2 : 00000001 r1 : 00000000 r0 : bf7e6e34 Flags: nZCv IRQs off FIQs on Mode SVC_32 Code: e5934004 e1a0e003 e59f3050 e2444001 (e5f4c001) Resetting CPU ... resetting ... Signed-off-by: Igor Opaniuk <igor.opaniuk(a)gmail.com> --- cmd/hash.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) diff --git a/cmd/hash.c b/cmd/hash.c index e163cd67742..5534a735fa7 100644 --- a/cmd/hash.c +++ b/cmd/hash.c @@ -14,15 +14,22 @@ #include <hash.h> #include <linux/ctype.h> +#if IS_ENABLED(CONFIG_HASH_VERIFY) +#define HARGS 6 +#else +#define HARGS 5 +#endif + static int do_hash(struct cmd_tbl *cmdtp, int flag, int argc, char *const argv[]) { char *s; int flags = HASH_FLAG_ENV; -#ifdef CONFIG_HASH_VERIFY - if (argc < 4) + if (argc < (HARGS - 1)) return CMD_RET_USAGE; + +#if IS_ENABLED(CONFIG_HASH_VERIFY) if (!strcmp(argv[1], "-v")) { flags |= HASH_FLAG_VERIFY; argc--; @@ -37,18 +44,12 @@ static int do_hash(struct cmd_tbl *cmdtp, int flag, int argc, return hash_command(*argv, flags, cmdtp, flag, argc - 1, argv + 1); } -#ifdef CONFIG_HASH_VERIFY -#define HARGS 6 -#else -#define HARGS 5 -#endif - U_BOOT_CMD( hash, HARGS, 1, do_hash, "compute hash message digest", "algorithm address count [[*]hash_dest]\n" " - compute message digest [save to env var / *address]" -#ifdef CONFIG_HASH_VERIFY +#if IS_ENABLED(CONFIG_HASH_VERIFY) "\nhash -v algorithm address count [*]hash\n" " - verify message digest of memory area to immediate value, \n" " env var or *address" -- 2.34.1

3 2

[PATCH] memory: ti-gpmc: Fix lock up at A53 SPL during NAND boot on AM64-EVM
by Roger Quadros 13 Feb '24

13 Feb '24

AM64 ES2.0 bootrom seems to enable WAIT0EDGEDETECTION interrupt. This causes a lockup at A53 SPL when accessing NAND controller or ELM registers. A good option would be to softrest GPMC block at probe but this cannot be done for AM64 as SOFTRESET bit is marked as reserved in SYSCONFIG register. Fix the issue by disabling all IRQs at probe. Signed-off-by: Roger Quadros <rogerq(a)kernel.org> --- drivers/memory/ti-gpmc.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/memory/ti-gpmc.c b/drivers/memory/ti-gpmc.c index 0b8674339e..8877b8f438 100644 --- a/drivers/memory/ti-gpmc.c +++ b/drivers/memory/ti-gpmc.c @@ -1196,6 +1196,12 @@ static int gpmc_probe(struct udevice *dev) gpmc_cfg = (struct gpmc *)priv->base; gpmc_base = priv->base; + /* + * Disable all IRQs as some bootroms might leave them enabled + * and that will cause a lock-up later + */ + gpmc_write_reg(GPMC_IRQENABLE, 0); + priv->l3_clk = devm_clk_get(dev, "fck"); if (IS_ERR(priv->l3_clk)) return PTR_ERR(priv->l3_clk); base-commit: 28760ce8640ff6266bd1c1c568a4a231576f3919 -- 2.34.1

2 1

[PATCH] poplar: add myself as co-maintainer
by Igor Opaniuk 13 Feb '24

13 Feb '24

Add myself as co-maintainer for Poplar board, as I'm currently working on it (re-testing releases, addressing issues etc). CC: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz(a)gmail.com> CC: Shawn Guo <shawn.guo(a)linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk(a)foundries.io> --- board/hisilicon/poplar/MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/board/hisilicon/poplar/MAINTAINERS b/board/hisilicon/poplar/MAINTAINERS index bfd4a9be66b..01efc96b0d3 100644 --- a/board/hisilicon/poplar/MAINTAINERS +++ b/board/hisilicon/poplar/MAINTAINERS @@ -1,6 +1,7 @@ Poplar BOARD M: Jorge Ramirez-Ortiz <jorge.ramirez.ortiz(a)gmail.com> M: Shawn Guo <shawn.guo(a)linaro.org> +M: Igor Opaniuk <igor.opaniuk(a)gmail.com> S: Maintained F: board/hisilicon/poplar F: doc/board/hisilicon/poplar.rst -- 2.34.1

5 8

[PATCH v2] boot: add support for button commands
by Caleb Connolly 13 Feb '24

13 Feb '24

With the relatively new button API in U-Boot, it's now much easier to model the common usecase of mapping arbitrary actions to different buttons during boot - for example entering fastboot mode, setting some additional kernel cmdline arguments, or booting with a custom recovery ramdisk, to name a few. Historically, this functionality has been implemented in board code, making it fixed for a given U-Boot binary and requiring the code be duplicated and modified for every board. Implement a generic abstraction to run an arbitrary command during boot when a specific button is pressed. The button -> command mapping is configured via environment variables with the following format: button_cmd_N_name=<button label> button_cmd_N=<command to run> Where N is the mapping number starting from 0. For example: button_cmd_0_name=vol_down button_cmd_0=fastboot usb 0 This will cause the device to enter fastboot mode if volume down is held during boot. After we enter the cli loop the button commands are no longer valid, this allows the buttons to additionally be used for navigating a boot menu. Tested-by: Svyatoslav Ryhel <clamor95(a)gmail.com> # Tegra30 Signed-off-by: Caleb Connolly <caleb.connolly(a)linaro.org> --- Changes since v1: * make get_button_cmd() static. * use #if IS_ENABLED(CONFIG_BUTTON_CMD) instead of #ifdef CONFIG_BUTTON_CMD. * Kconfig fixes --- boot/Kconfig | 15 +++++++ common/Makefile | 1 + common/button_cmd.c | 83 +++++++++++++++++++++++++++++++++++++++ common/main.c | 3 ++ doc/usage/environment.rst | 4 ++ include/button.h | 9 +++++ 6 files changed, 115 insertions(+) create mode 100644 common/button_cmd.c diff --git a/boot/Kconfig b/boot/Kconfig index fbc49c5bca47..882835731ea9 100644 --- a/boot/Kconfig +++ b/boot/Kconfig @@ -20,6 +20,21 @@ config TIMESTAMP loaded that does not, the message 'Wrong FIT format: no timestamp' is shown. +config BUTTON_CMD + bool "Support for running a command if a button is held during boot" + depends on CMDLINE + depends on BUTTON + help + For many embedded devices it's useful to enter a special flashing mode + such as fastboot mode when a button is held during boot. This option + allows arbitrary commands to be assigned to specific buttons. These will + be run after "preboot" if the button is held. Configuration is done via + the environment variables "button_cmd_N_name" and "button_cmd_N" where n is + the button number (starting from 0). e.g: + + "button_cmd_0_name=vol_down" + "button_cmd_0=fastboot usb 0" + menuconfig FIT bool "Flattened Image Tree (FIT)" select HASH diff --git a/common/Makefile b/common/Makefile index cdeadf72026c..53105a6ce12a 100644 --- a/common/Makefile +++ b/common/Makefile @@ -10,6 +10,7 @@ obj-y += main.o obj-y += exports.o obj-$(CONFIG_HUSH_PARSER) += cli_hush.o obj-$(CONFIG_AUTOBOOT) += autoboot.o +obj-$(CONFIG_BUTTON_CMD) += button_cmd.o # # boards obj-y += board_f.o diff --git a/common/button_cmd.c b/common/button_cmd.c new file mode 100644 index 000000000000..b6a8434d6f29 --- /dev/null +++ b/common/button_cmd.c @@ -0,0 +1,83 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Copyright (c) 2023 Linaro Ltd. + * Author: Caleb Connolly <caleb.connolly(a)linaro.org> + */ + +#include <button.h> +#include <command.h> +#include <env.h> +#include <log.h> +#include <vsprintf.h> + +/* Some sane limit "just in case" */ +#define MAX_BTN_CMDS 32 + +struct button_cmd { + bool pressed; + const char *btn_name; + const char *cmd; +}; + +/* + * Button commands are set via environment variables, e.g.: + * button_cmd_N_name=Volume Up + * button_cmd_N=fastboot usb 0 + * + * This function will retrieve the command for the given button N + * and populate the cmd struct with the command string and pressed + * state of the button. + * + * Returns 1 if a command was found, 0 otherwise. + */ +static int get_button_cmd(int n, struct button_cmd *cmd) +{ + const char *cmd_str; + struct udevice *btn; + char buf[24]; + + snprintf(buf, sizeof(buf), "button_cmd_%d_name", n); + cmd->btn_name = env_get(buf); + if (!cmd->btn_name) + return 0; + + button_get_by_label(cmd->btn_name, &btn); + if (!btn) { + log_err("No button labelled '%s'\n", cmd->btn_name); + return 0; + } + + cmd->pressed = button_get_state(btn) == BUTTON_ON; + /* If the button isn't pressed then cmd->cmd will be unused so don't waste + * cycles reading it + */ + if (!cmd->pressed) + return 1; + + snprintf(buf, sizeof(buf), "button_cmd_%d", n); + cmd_str = env_get(buf); + if (!cmd_str) { + log_err("No command set for button '%s'\n", cmd->btn_name); + return 0; + } + + cmd->cmd = cmd_str; + + return 1; +} + +void process_button_cmds(void) +{ + struct button_cmd cmd = {0}; + int i = 0; + + while (get_button_cmd(i++, &cmd) && i < MAX_BTN_CMDS) { + if (!cmd.pressed) + continue; + + log_info("BTN '%s'> %s\n", cmd.btn_name, cmd.cmd); + run_command(cmd.cmd, CMD_FLAG_ENV); + /* Don't run commands for multiple buttons */ + return; + } +} diff --git a/common/main.c b/common/main.c index 7c70de2e59a8..717e8b7e8bd2 100644 --- a/common/main.c +++ b/common/main.c @@ -8,6 +8,7 @@ #include <common.h> #include <autoboot.h> +#include <button.h> #include <bootstage.h> #include <cli.h> #include <command.h> @@ -61,6 +62,8 @@ void main_loop(void) efi_launch_capsules(); } + process_button_cmds(); + s = bootdelay_process(); if (cli_process_fdt(&s)) cli_secure_boot_cmd(s); diff --git a/doc/usage/environment.rst b/doc/usage/environment.rst index c57b717caaf3..ce5a9627f025 100644 --- a/doc/usage/environment.rst +++ b/doc/usage/environment.rst @@ -190,6 +190,10 @@ bootm_size bootstopkeysha256, bootdelaykey, bootstopkey See README.autoboot +button_cmd_0, button_cmd_0_name ... button_cmd_N, button_cmd_N_name + Used to map commands to run when a button is held during boot. + See CONFIG_BUTTON_CMD. + updatefile Location of the software update file on a TFTP server, used by the automatic software update feature. Please refer to diff --git a/include/button.h b/include/button.h index 207f4a0f4dbd..8d38e521324d 100644 --- a/include/button.h +++ b/include/button.h @@ -74,4 +74,13 @@ enum button_state_t button_get_state(struct udevice *dev); */ int button_get_code(struct udevice *dev); +#if IS_ENABLED(CONFIG_BUTTON_CMD) +/* Process button command mappings specified in the environment, + * running the commands for buttons which are pressed + */ +void process_button_cmds(void); +#else +static inline void process_button_cmds(void) {} +#endif /* CONFIG_BUTTON_CMD */ + #endif -- 2.42.1

4 9

Booting from NFSv4 file system?
by Cedric Blancher 13 Feb '24

13 Feb '24

Good morning! Does U-Boot support booting from a NFSv4 file system? Explicitly neither NFSv2 or NFSv3 will work in our case, as both protocol versions are depreciated and no longer allowed by our IT department. Ced -- Cedric Blancher <cedric.blancher(a)gmail.com> [https://plus.google.com/u/0/+CedricBlancher/] Institute Pasteur

2 2

[PATCH 1/3] mtd: spi-nor-ids: Add support for ESMT/EON EN25Q80B
by Frieder Schrempf 13 Feb '24

13 Feb '24

From: Frieder Schrempf <frieder.schrempf(a)kontron.de> The datasheet can be found here: https://www.esmt.com.tw/upload/pdf/ESMT/datasheets/EN25Q80B_Ver.E.pdf Signed-off-by: Frieder Schrempf <frieder.schrempf(a)kontron.de> --- drivers/mtd/spi/spi-nor-ids.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mtd/spi/spi-nor-ids.c b/drivers/mtd/spi/spi-nor-ids.c index 38a287487ed..4e83b8c94c9 100644 --- a/drivers/mtd/spi/spi-nor-ids.c +++ b/drivers/mtd/spi/spi-nor-ids.c @@ -80,6 +80,7 @@ const struct flash_info spi_nor_ids[] = { #endif #ifdef CONFIG_SPI_FLASH_EON /* EON */ /* EON -- en25xxx */ + { INFO("en25q80b", 0x1c3014, 0, 64 * 1024, 16, SECT_4K) }, { INFO("en25q32b", 0x1c3016, 0, 64 * 1024, 64, 0) }, { INFO("en25q64", 0x1c3017, 0, 64 * 1024, 128, SECT_4K) }, { INFO("en25q128b", 0x1c3018, 0, 64 * 1024, 256, 0) }, -- 2.43.0

3 7

[PATCH v2 0/7] AVB: cosmetic adjustments/improvements
by Igor Opaniuk 13 Feb '24

13 Feb '24

This is the first patch series for incoming major improvements for AVB implementation, that include: - Simplify and add more context for debug/error prints where it's needed. - Move SPDX license identifiers to the first line, so it conforms to license placement rule. - Use mmc_switch_part() only for eMMC. - Rework do_avb_verify_part, take into account device lock state for setting correct androidboot.verifiedbootstate. - Adjust AVB documentation, sync usage info with the one cmd/avb.c. Igor Opaniuk (7): common: avb_verify: don't call mmc_switch_part for SD avb: move SPDX license identifiers to the first line common: avb_verify: rework error/debug prints cmd: avb: rework prints common: avb_verify: add str_avb_io_error/str_avb_slot_error cmd: avb: rework do_avb_verify_part doc: android: avb: sync usage details cmd/avb.c | 173 +++++++++++++------------ common/avb_verify.c | 89 ++++++++++--- doc/android/avb2.rst | 16 ++- include/avb_verify.h | 7 +- test/py/tests/test_android/test_avb.py | 3 +- 5 files changed, 178 insertions(+), 110 deletions(-) -- 2.34.1

5 16

[RESEND PATCH v1] MAINTAINERS: add custodian tree info for AVB/AB
by Igor Opaniuk 13 Feb '24

13 Feb '24

From: Igor Opaniuk <igor.opaniuk(a)gmail.com> Add information about a custodian tree [1] for AVB/AB, which is maintained by Mattijs Korpershoek. [1] https://source.denx.de/u-boot/custodians/u-boot-dfu Signed-off-by: Igor Opaniuk <igor.opaniuk(a)gmail.com> --- MAINTAINERS | 2 ++ 1 file changed, 2 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 46ba17647f3..32f3ee30710 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -62,6 +62,7 @@ M: Igor Opaniuk <igor.opaniuk(a)gmail.com> M: Mattijs Korpershoek <mkorpershoek(a)baylibre.com> R: Sam Protsenko <semen.protsenko(a)linaro.org> S: Maintained +T: git https://source.denx.de/u-boot/custodians/u-boot-dfu.git F: boot/android_ab.c F: cmd/ab_select.c F: doc/android/ab.rst @@ -72,6 +73,7 @@ ANDROID AVB M: Igor Opaniuk <igor.opaniuk(a)gmail.com> M: Mattijs Korpershoek <mkorpershoek(a)baylibre.com> S: Maintained +T: git https://source.denx.de/u-boot/custodians/u-boot-dfu.git F: cmd/avb.c F: common/avb_verify.c F: doc/android/avb2.rst -- 2.34.1

2 2