March 2021 - U-Boot

[PATCH] video: Fix line padding calculation for 16 and 24 BPP bitmaps
by Sylwester Nawrocki 11 Apr '21

11 Apr '21

Each row in the pixel array in the bitmap file is padded if necessary so the row size is always a multiple of 4 bytes. In current code the complement of row size to a multiple of 4 bytes is further unnecessarily multiplied by the pixel size. This results in incorrect displaying of bitmaps having row size that is not a multiple of 4 bytes. Fix this by removing the unnecessary multiplication. Tested with 24BPP bitmap and XRGB32 display. Signed-off-by: Sylwester Nawrocki <s.nawrocki(a)samsung.com> --- drivers/video/video_bmp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/video/video_bmp.c b/drivers/video/video_bmp.c index 5a4d12c68d..5537378781 100644 --- a/drivers/video/video_bmp.c +++ b/drivers/video/video_bmp.c @@ -328,7 +328,7 @@ int video_bmp_display(struct udevice *dev, ulong bmp_image, int x, int y, for (j = 0; j < width; j++) fb_put_word(&fb, &bmap); - bmap += (padded_width - width) * 2; + bmap += (padded_width - width); fb -= width * 2 + priv->line_length; } break; @@ -352,7 +352,7 @@ int video_bmp_display(struct udevice *dev, ulong bmp_image, int x, int y, } } fb -= priv->line_length + width * (bpix / 8); - bmap += (padded_width - width) * 3; + bmap += (padded_width - width); } break; #endif /* CONFIG_BMP_24BPP */ -- 2.17.1

3 2

[PATCH v3 0/4] psci: add support for SYSTEM_RESET2 and PSCI_FEATURES
by Igor Opaniuk 11 Apr '21

11 Apr '21

From: Igor Opaniuk <igor.opaniuk(a)foundries.io> 1. Adds support for: * PSCI_FEATURES, introduced in PSCI 1.0. This provides API that allows discovering whether a specific PSCI function is implemented and its features. * SYSTEM_RESET2, introduced in PSCI 1.1, which extends existing SYSTEM_RESET. It provides support for vendor-specific resets, providing reset_type as an additional param. 2. PSCI sysreset driver is refactored to use new API. 3. do_reset cmd is extended, optional param added for providing type of reset CI: https://dev.azure.com/igoropaniuk/u-boot/_build/results?buildId=20&view=res… Changes in v3: - Drop RFC tag - Add usage doc for reset cmd - Reimplement param handling for reset cmd - Droped updates in reset usage string Changes in v2: - do_reset cmd updates Igor Opaniuk (4): psci: add features/reset2 support sysreset: psci: use psci driver exported functions sysreset: provide type of reset in do_reset cmd doc: usage: add usage details for reset cmd cmd/boot.c | 2 +- doc/usage/index.rst | 1 + doc/usage/reset.rst | 26 ++++++++++++ drivers/firmware/psci.c | 68 ++++++++++++++++++++++++++++++ drivers/sysreset/sysreset-uclass.c | 11 ++++- drivers/sysreset/sysreset_psci.c | 8 +--- include/linux/psci.h | 3 ++ 7 files changed, 111 insertions(+), 8 deletions(-) create mode 100644 doc/usage/reset.rst -- 2.25.1

4 11

[PATCH v3 00/12] Add support for pinmux status command on beaglebone
by Dario Binacchi 11 Apr '21

11 Apr '21

The series was born from the need to check the pinmux setting of a peripheral on a beaglebone board. I then ran the 'pinmux status -a' command but it failed because some operations (get_pin_muxing, get_pin_name and get_pins_count) were missing in the 'pinctrl-single' driver. The patch series can be cleanly applied to the HEAD of the master which at the time of release points to 08cca188c1 ("Merge https://gitlab.denx.de/u-boot/custodians/u-boot-usb") commit The series was archived with the status 'Changes Requested' because the Sandbox tests failed. I tried them locally and couldn't reproduce. Lokesh Vutla also verified that execution outside of gitlab-ci did not fail. It's been a while since then. I am available to investigate the problem on your indications with changes to the test patch. Changes in v3: - Added Pratyush Yadav review tag. - Updated commit message. - Added Pratyush Yadav review tag. - Added Pratyush Yadav review tag. - Added Simon Glass review tag. - Added Simon Glass review tag. - Added Simon Glass review tag. Changes in v2: - Added Simon Glass review tag. - Updated commit message. - Added Simon Glass review tag. - Added Pratyush Yadav review tag. - Updated commit message - Split in 2 commits - Check dev_read_addr_size return value. - Updated commit message. - Added Simon Glass review tag. - Updated commit message. - Remove pointer to access functions. - Added Simon Glass review tag. - Added Simon Glass review tag. - Added Simon Glass review tag. - Added error checking when the 'width' property is missing. - Fix coding style. Dario Binacchi (12): pinctrl: single: fix format of structure documentation pinctrl: single: fix the loop counter variable type pinctrl: single: fix offset management pinctrl: single: fix debug messages formatting pinctrl: single: get register area size by device API pinctrl: single: check "register-width" DT property pinctrl: single: change function mask default value pinctrl: single: add register access functions pinctrl: single: add get_pins_count operation pinctrl: single: add get_pin_name operation pinctrl: single: add get_pin_muxing operation test: pinmux: add test for 'pinctrl-single' driver arch/sandbox/dts/test.dts | 72 +++++ configs/sandbox_defconfig | 1 + drivers/pinctrl/pinctrl-single.c | 474 +++++++++++++++++++++++++++---- test/dm/pinmux.c | 91 +++++- 4 files changed, 573 insertions(+), 65 deletions(-) -- 2.17.1

2 14

[PATCH 01/13] arm: nanopi2: Remove unused code
by Tom Rini 10 Apr '21

10 Apr '21

This platform did not ever enable CONFIG_REVISION_TAG, so the code to set the board_rev environment variable was never enabled. This particular symbol is also only for use with the REVISION ATAG and this platform is new enough to have never supported an ATAG-based Linux Kernel. Cc: Stefan Bosch <stefan_b(a)posteo.net> Signed-off-by: Tom Rini <trini(a)konsulko.com> --- I'd be happy to see this patch replaced by one that enables what I think you meant to be doing and by default. Thanks! --- board/friendlyarm/nanopi2/board.c | 13 ------------- 1 file changed, 13 deletions(-) diff --git a/board/friendlyarm/nanopi2/board.c b/board/friendlyarm/nanopi2/board.c index 68980536abe9..6e546853b863 100644 --- a/board/friendlyarm/nanopi2/board.c +++ b/board/friendlyarm/nanopi2/board.c @@ -294,16 +294,6 @@ static void set_ether_addr(void) env_set("ethaddr", ethaddr); } -#ifdef CONFIG_REVISION_TAG -static void set_board_rev(void) -{ - char info[64] = {0, }; - - snprintf(info, ARRAY_SIZE(info), "%02x", get_board_rev()); - env_set("board_rev", info); -} -#endif - static void set_dtb_name(void) { char info[64] = {0, }; @@ -435,9 +425,6 @@ int board_late_init(void) { bd_update_env(); -#ifdef CONFIG_REVISION_TAG - set_board_rev(); -#endif set_dtb_name(); set_ether_addr(); -- 2.17.1

8 30

[PATCH] Add support for stack-protector
by Joel Peshkin 10 Apr '21

10 Apr '21

Cc: Simon Glass <sjg(a)chromium.org> Cc: Bin Meng <bmeng.cn(a)gmail.com> Cc: Jagan Teki <jagan(a)amarulasolutions.com> Cc: Kever Yang <kever.yang(a)rock-chips.com> Cc: Heinrich Schuchardt <xypron.glpk(a)gmx.de> Cc: AKASHI Takahiro <takahiro.akashi(a)linaro.org> Cc: Usama Arif <usama.arif(a)arm.com> Cc: Sam Protsenko <joe.skb7(a)gmail.com> Cc: Masahiro Yamada <masahiroy(a)kernel.org> Cc: Philippe Reynes <philippe.reynes(a)softathome.com> Cc: Eugeniu Rosca <roscaeugeniu(a)gmail.com> Cc: Jan Kiszka <jan.kiszka(a)siemens.com> Signed-off-by: Joel Peshkin <joel.peshkin(a)broadcom.com> --- Makefile | 4 ++++ common/Kconfig | 15 +++++++++++++++ common/Makefile | 2 ++ common/stackprot.c | 17 +++++++++++++++++ scripts/Makefile.spl | 6 ++++++ 5 files changed, 44 insertions(+) create mode 100644 common/stackprot.c diff --git a/Makefile b/Makefile index 3ee4cc00dd..6e7a81ec7d 100644 --- a/Makefile +++ b/Makefile @@ -677,7 +677,11 @@ else KBUILD_CFLAGS += -O2 endif +ifeq ($(CONFIG_STACKPROTECTOR),y) +KBUILD_CFLAGS += $(call cc-option,-fstack-protector-strong) +else KBUILD_CFLAGS += $(call cc-option,-fno-stack-protector) +endif KBUILD_CFLAGS += $(call cc-option,-fno-delete-null-pointer-checks) # disable stringop warnings in gcc 8+ diff --git a/common/Kconfig b/common/Kconfig index 2bce8c9ba1..e30c3c4ab8 100644 --- a/common/Kconfig +++ b/common/Kconfig @@ -595,6 +595,21 @@ config TPL_HASH and the algorithms it supports are defined in common/hash.c. See also CMD_HASH for command-line access. +config STACKPROTECTOR + bool "Stack Protector buffer overflow detection" + default n + help + Enable stack smash detection through gcc built-in stack-protector + canary logic + +config SPL_STACKPROTECTOR + bool "Stack Protector buffer overflow detection for SPL" + default n + +config TPL_STACKPROTECTOR + bool "Stack Protector buffer overflow detection for SPL" + default n + endmenu menu "Update support" diff --git a/common/Makefile b/common/Makefile index bcf352d016..fe71e18317 100644 --- a/common/Makefile +++ b/common/Makefile @@ -138,3 +138,5 @@ obj-$(CONFIG_CMD_LOADB) += xyzModem.o obj-$(CONFIG_$(SPL_TPL_)YMODEM_SUPPORT) += xyzModem.o obj-$(CONFIG_AVB_VERIFY) += avb_verify.o +obj-$(CONFIG_$(SPL_TPL_)STACKPROTECTOR) += stackprot.o + diff --git a/common/stackprot.c b/common/stackprot.c new file mode 100644 index 0000000000..7c95b8544f --- /dev/null +++ b/common/stackprot.c @@ -0,0 +1,17 @@ +// SPDX-License-Identifier: GPL-2.0+ +/* + * Copyright 2021 Broadcom + */ + +#include <common.h> + +DECLARE_GLOBAL_DATA_PTR; + +unsigned long __stack_chk_guard = 0xfeedf00ddeadbeef; + +void __stack_chk_fail(void) +{ + panic("Stack smashing detected in function: %p relocated from %p", + __builtin_return_address(0), + __builtin_return_address(0) - gd->reloc_off); +} diff --git a/scripts/Makefile.spl b/scripts/Makefile.spl index 9f1f7445d7..1505e4e851 100644 --- a/scripts/Makefile.spl +++ b/scripts/Makefile.spl @@ -63,6 +63,12 @@ include $(srctree)/scripts/Makefile.lib KBUILD_CFLAGS += -ffunction-sections -fdata-sections LDFLAGS_FINAL += --gc-sections +ifeq ($(CONFIG_$(SPL_TPL_)STACKPROTECTOR),y) +KBUILD_CFLAGS += -fstack-protector-strong +else +KBUILD_CFLAGS += -fno-stack-protector +endif + # FIX ME cpp_flags := $(KBUILD_CPPFLAGS) $(PLATFORM_CPPFLAGS) $(UBOOTINCLUDE) \ $(NOSTDINC_FLAGS) -- 2.27.0

4 27

[PATCH V2] arm: stm32mp1: Set soc_type, soc_pkg, soc_rev env variables
by Marek Vasut 09 Apr '21

09 Apr '21

Split up get_soc_name(), clean the decoding up a bit, and set up environment variables which contain the SoC type, package, revision. This is useful on SoMs, where multiple SoC options are populated. Signed-off-by: Marek Vasut <marex(a)denx.de> Cc: Patrick Delaunay <patrick.delaunay(a)st.com> Cc: Patrice Chotard <patrice.chotard(a)st.com> --- V2: Fix pkg = 0 , which should be *pkg = 0 --- arch/arm/mach-stm32mp/cpu.c | 105 ++++++++++++++++++------------------ 1 file changed, 53 insertions(+), 52 deletions(-) diff --git a/arch/arm/mach-stm32mp/cpu.c b/arch/arm/mach-stm32mp/cpu.c index bc2db535beb..ba8771940ef 100644 --- a/arch/arm/mach-stm32mp/cpu.c +++ b/arch/arm/mach-stm32mp/cpu.c @@ -352,89 +352,78 @@ u32 get_cpu_package(void) return get_otp(BSEC_OTP_PKG, PKG_SHIFT, PKG_MASK); } -void get_soc_name(char name[SOC_NAME_SIZE]) +static const char * const soc_type[] = { + "????", + "151C", "151A", "151F", "151D", + "153C", "153A", "153F", "153D", + "157C", "157A", "157F", "157D" +}; + +static const char * const soc_pkg[] = { "??", "AD", "AC", "AB", "AA" }; +static const char * const soc_rev[] = { "?", "A", "B", "Z" }; + +static void get_cpu_string_offsets(unsigned int *type, unsigned int *pkg, + unsigned int *rev) { - char *cpu_s, *cpu_r, *pkg; + u32 cpu_type = get_cpu_type(); + u32 ct = cpu_type & ~(BIT(7) | BIT(0)); + u32 cm = ((cpu_type & BIT(7)) >> 6) | (cpu_type & BIT(0)); + u32 cp = get_cpu_package(); - /* MPUs Part Numbers */ - switch (get_cpu_type()) { - case CPU_STM32MP157Fxx: - cpu_s = "157F"; - break; - case CPU_STM32MP157Dxx: - cpu_s = "157D"; - break; - case CPU_STM32MP157Cxx: - cpu_s = "157C"; - break; - case CPU_STM32MP157Axx: - cpu_s = "157A"; - break; - case CPU_STM32MP153Fxx: - cpu_s = "153F"; - break; - case CPU_STM32MP153Dxx: - cpu_s = "153D"; + /* Bits 0 and 7 are the ACDF, 00:C 01:A 10:F 11:D */ + switch (ct) { + case CPU_STM32MP151Cxx: + *type = cm + 1; break; case CPU_STM32MP153Cxx: - cpu_s = "153C"; - break; - case CPU_STM32MP153Axx: - cpu_s = "153A"; - break; - case CPU_STM32MP151Fxx: - cpu_s = "151F"; - break; - case CPU_STM32MP151Dxx: - cpu_s = "151D"; + *type = cm + 5; break; - case CPU_STM32MP151Cxx: - cpu_s = "151C"; - break; - case CPU_STM32MP151Axx: - cpu_s = "151A"; + case CPU_STM32MP157Cxx: + *type = cm + 9; break; default: - cpu_s = "????"; + *type = 0; break; } /* Package */ - switch (get_cpu_package()) { + switch (cp) { case PKG_AA_LBGA448: - pkg = "AA"; - break; case PKG_AB_LBGA354: - pkg = "AB"; - break; case PKG_AC_TFBGA361: - pkg = "AC"; - break; case PKG_AD_TFBGA257: - pkg = "AD"; + *pkg = cp; break; default: - pkg = "??"; + *pkg = 0; break; } - /* REVISION */ + /* Revision */ switch (get_cpu_rev()) { case CPU_REVA: - cpu_r = "A"; + *rev = 1; break; case CPU_REVB: - cpu_r = "B"; + *rev = 2; break; case CPU_REVZ: - cpu_r = "Z"; + *rev = 3; break; default: - cpu_r = "?"; + *rev = 0; break; } +} + +void get_soc_name(char name[SOC_NAME_SIZE]) +{ + unsigned int type, pkg, rev; - snprintf(name, SOC_NAME_SIZE, "STM32MP%s%s Rev.%s", cpu_s, pkg, cpu_r); + get_cpu_string_offsets(&type, &pkg, &rev); + + snprintf(name, SOC_NAME_SIZE, "STM32MP%s%s Rev.%s", + soc_type[type], soc_pkg[pkg], soc_rev[rev]); } #if defined(CONFIG_DISPLAY_CPUINFO) @@ -626,11 +615,23 @@ static int setup_serial_number(void) return 0; } +static void setup_soc_type_pkg_rev(void) +{ + unsigned int type, pkg, rev; + + get_cpu_string_offsets(&type, &pkg, &rev); + + env_set("soc_type", soc_type[type]); + env_set("soc_pkg", soc_pkg[pkg]); + env_set("soc_rev", soc_rev[rev]); +} + int arch_misc_init(void) { setup_boot_mode(); setup_mac_address(); setup_serial_number(); + setup_soc_type_pkg_rev(); return 0; } -- 2.30.2

3 2

[PATCH] imx: imx6ull: fix pinmux sel_input value for uart5 pins
by Yuichiro Goto 09 Apr '21

09 Apr '21

sel_input value for the following uart5 pins is different between i.MX6UL and i.MX6ULL: MX6_PAD_UART5_TX_DATA__UART5_DTE_RX MX6_PAD_UART5_RX_DATA__UART5_DCE_RX MX6_PAD_ENET1_RX_EN__UART5_DCE_RTS MX6_PAD_ENET1_TX_DATA0__UART5_DTE_RTS MX6_PAD_CSI_DATA02__UART5_DCE_RTS As sel_input value for the second one is fixed by the previous commit, fix the rest. Signed-off-by: Yuichiro Goto <goto(a)k-tech.co.jp> --- arch/arm/include/asm/arch-mx6/mx6ull_pins.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/arm/include/asm/arch-mx6/mx6ull_pins.h b/arch/arm/include/asm/arch-mx6/mx6ull_pins.h index de4a1abc..842d0caa 100644 --- a/arch/arm/include/asm/arch-mx6/mx6ull_pins.h +++ b/arch/arm/include/asm/arch-mx6/mx6ull_pins.h @@ -326,7 +326,7 @@ enum { MX6_PAD_UART5_TX_DATA__UART5_DCE_TX = IOMUX_PAD(0x0348, 0x00BC, 0, 0x0000, 0, 0), - MX6_PAD_UART5_TX_DATA__UART5_DTE_RX = IOMUX_PAD(0x0348, 0x00BC, 0, 0x0644, 4, 0), + MX6_PAD_UART5_TX_DATA__UART5_DTE_RX = IOMUX_PAD(0x0348, 0x00BC, 0, 0x0644, 6, 0), MX6_PAD_UART5_TX_DATA__ENET2_CRS = IOMUX_PAD(0x0348, 0x00BC, 1, 0x0000, 0, 0), MX6_PAD_UART5_TX_DATA__I2C2_SCL = IOMUX_PAD(0x0348, 0x00BC, IOMUX_CONFIG_SION | 2, 0x05AC, 2, 0), MX6_PAD_UART5_TX_DATA__CSI_DATA14 = IOMUX_PAD(0x0348, 0x00BC, 3, 0x04FC, 0, 0), @@ -363,7 +363,7 @@ enum { MX6_PAD_ENET1_RX_DATA1__USDHC2_LCTL = IOMUX_PAD(0x0354, 0x00C8, 8, 0x0000, 0, 0), MX6_PAD_ENET1_RX_EN__ENET1_RX_EN = IOMUX_PAD(0x0358, 0x00CC, 0, 0x0000, 0, 0), - MX6_PAD_ENET1_RX_EN__UART5_DCE_RTS = IOMUX_PAD(0x0358, 0x00CC, 1, 0x0640, 3, 0), + MX6_PAD_ENET1_RX_EN__UART5_DCE_RTS = IOMUX_PAD(0x0358, 0x00CC, 1, 0x0640, 5, 0), MX6_PAD_ENET1_RX_EN__UART5_DTE_CTS = IOMUX_PAD(0x0358, 0x00CC, 1, 0x0000, 0, 0), MX6_PAD_ENET1_RX_EN__CSI_DATA18 = IOMUX_PAD(0x0358, 0x00CC, 3, 0x050C, 0, 0), MX6_PAD_ENET1_RX_EN__FLEXCAN2_TX = IOMUX_PAD(0x0358, 0x00CC, 4, 0x0000, 0, 0), @@ -373,7 +373,7 @@ enum { MX6_PAD_ENET1_TX_DATA0__ENET1_TDATA00 = IOMUX_PAD(0x035C, 0x00D0, 0, 0x0000, 0, 0), MX6_PAD_ENET1_TX_DATA0__UART5_DCE_CTS = IOMUX_PAD(0x035C, 0x00D0, 1, 0x0000, 0, 0), - MX6_PAD_ENET1_TX_DATA0__UART5_DTE_RTS = IOMUX_PAD(0x035C, 0x00D0, 1, 0x0640, 4, 0), + MX6_PAD_ENET1_TX_DATA0__UART5_DTE_RTS = IOMUX_PAD(0x035C, 0x00D0, 1, 0x0640, 6, 0), MX6_PAD_ENET1_TX_DATA0__CSI_DATA19 = IOMUX_PAD(0x035C, 0x00D0, 3, 0x0510, 0, 0), MX6_PAD_ENET1_TX_DATA0__FLEXCAN2_RX = IOMUX_PAD(0x035C, 0x00D0, 4, 0x0588, 1, 0), MX6_PAD_ENET1_TX_DATA0__GPIO2_IO03 = IOMUX_PAD(0x035C, 0x00D0, 5, 0x0000, 0, 0), @@ -1012,7 +1012,7 @@ enum { MX6_PAD_CSI_DATA02__EIM_AD02 = IOMUX_PAD(0x0478, 0x01EC, 4, 0x0000, 0, 0), MX6_PAD_CSI_DATA02__GPIO4_IO23 = IOMUX_PAD(0x0478, 0x01EC, 5, 0x0000, 0, 0), MX6_PAD_CSI_DATA02__SAI1_RX_SYNC = IOMUX_PAD(0x0478, 0x01EC, 6, 0x0000, 0, 0), - MX6_PAD_CSI_DATA02__UART5_DCE_RTS = IOMUX_PAD(0x0478, 0x01EC, 8, 0x0640, 5, 0), + MX6_PAD_CSI_DATA02__UART5_DCE_RTS = IOMUX_PAD(0x0478, 0x01EC, 8, 0x0640, 7, 0), MX6_PAD_CSI_DATA02__UART5_DTE_CTS = IOMUX_PAD(0x0478, 0x01EC, 8, 0x0000, 0, 0), MX6_PAD_CSI_DATA03__CSI_DATA05 = IOMUX_PAD(0x047C, 0x01F0, 0, 0x04CC, 0, 0), -- 2.17.1

2 1

[PATCH 1/1] imx6: icorem6: chmod 644 enigcam.bmp
by Heinrich Schuchardt 09 Apr '21

09 Apr '21

Bitmap files should not be executable. Signed-off-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> --- tools/logos/engicam.bmp | Bin 1 file changed, 0 insertions(+), 0 deletions(-) mode change 100755 => 100644 tools/logos/engicam.bmp diff --git a/tools/logos/engicam.bmp b/tools/logos/engicam.bmp old mode 100755 new mode 100644 -- 2.30.2

3 2

[PATCH] doc: imx: psb: Document usage of SRC_GPR10 PERSIST_SECONDARY_BOOT for A/B switching
by Marek Vasut 09 Apr '21

09 Apr '21

Document SRC_GPR10 PERSIST_SECONDARY_BOOT functionality. This is useful for reliable bootloader A/B updates, as it permits switching between two copies of bootloader at different offsets of the same storage. The switch happens in case one copy is corrupted OR can be enforced by user. This functionality is present at least since i.MX53, however is poorly documented in all known SoC datasheets, hence this document aims to clarify the usage, currently on i.MX7D and i.MX8MM. Signed-off-by: Marek Vasut <marex(a)denx.de> # Original MX7D work, this document Signed-off-by: Igor Opaniuk <igor.opaniuk(a)foundries.io> # All the MX8M work Cc: Christoph Niedermaier <cniedermaier(a)dh-electronics.de> Cc: Fabio Estevam <festevam(a)gmail.com> Cc: Harald Seiler <hws(a)denx.de> Cc: Igor Opaniuk <igor.opaniuk(a)foundries.io> Cc: Jan Kiszka <jan.kiszka(a)siemens.com> Cc: Ludwig Zenz <lzenz(a)dh-electronics.com> Cc: Marcel Ziswiler <marcel.ziswiler(a)toradex.com> Cc: Peng Fan <peng.fan(a)nxp.com> Cc: Stefano Babic <sbabic(a)denx.de> Cc: Ye Li <ye.li(a)nxp.com> Cc: uboot-imx <uboot-imx(a)nxp.com> --- NOTE: This all was pieced together from various scraps of information. Links to further reading below. - MX53/Vybrid https://community.nxp.com/t5/Vybrid-Processors/Vybrid-secondary-image-table… - MX53 https://community.nxp.com/t5/i-MX-Processors/i-MX53-Redundant-Boot-blog-arc… - MX8M https://community.nxp.com/t5/i-MX-Processors/i-MX-8M-Mini-eMMC-Boot-Fusing/… - MX8M https://review.trustedfirmware.org/plugins/gitiles/TF-A/trusted-firmware-a/… --- doc/imx/index.rst | 9 +++ doc/imx/misc/index.rst | 9 +++ doc/imx/misc/psb.rst | 174 +++++++++++++++++++++++++++++++++++++++++ doc/index.rst | 11 +++ 4 files changed, 203 insertions(+) create mode 100644 doc/imx/index.rst create mode 100644 doc/imx/misc/index.rst create mode 100644 doc/imx/misc/psb.rst diff --git a/doc/imx/index.rst b/doc/imx/index.rst new file mode 100644 index 00000000000..b225b1d1837 --- /dev/null +++ b/doc/imx/index.rst @@ -0,0 +1,9 @@ +.. SPDX-License-Identifier: GPL-2.0+ + +NXP i.MX Machine-specific doc +============================= + +.. toctree:: + :maxdepth: 2 + + misc/index diff --git a/doc/imx/misc/index.rst b/doc/imx/misc/index.rst new file mode 100644 index 00000000000..85fbdb6588f --- /dev/null +++ b/doc/imx/misc/index.rst @@ -0,0 +1,9 @@ +.. SPDX-License-Identifier: GPL-2.0+ + +Miscellaneous +============= + +.. toctree:: + :maxdepth: 2 + + psb diff --git a/doc/imx/misc/psb.rst b/doc/imx/misc/psb.rst new file mode 100644 index 00000000000..bde6b1c0207 --- /dev/null +++ b/doc/imx/misc/psb.rst @@ -0,0 +1,174 @@ +i.MX7D/i.MX8MM SRC_GPR10 PERSIST_SECONDARY_BOOT for bootloader A/B switching +============================================================================ + +Introduction +------------ +Since at least iMX53 until iMX8MM, it is possible to have two copies of +bootloader in SD/eMMC and switch between them. The switch is triggered +either by the BootROM in case the bootloader image is faulty OR can be +enforced by the user. + +Operation +--------- + #. Upon Power-On Reset (POR) + + - SRC_GPR10 bit PERSIST_SECONDARY_BOOT is set to 0 + - BootROM attempts to start bootloader A-copy + + - if A-copy valid + + - BootROM starts A-copy + - END + + - if A-copy NOT valid + + - BootROM sets SRC_GPR10 bit PERSIST_SECONDARY_BOOT to 1 + - BootROM triggers WARM reset, GOTO 1) + - END + + #. Upon COLD Reset + + - GOTO 1) + - END + + #. Upon WARM Reset + + - SRC_GPR10 bit PERSIST_SECONDARY_BOOT is retained + + - if SRC_GPR10 bit PERSIST_SECONDARY_BOOT is 0 + + - BootROM attempts to start bootloader A-copy + + - if A-copy valid + + - BootROM starts A-copy + - END + + - if A-copy NOT valid + + - BootROM sets SRC_GPR10 bit PERSIST_SECONDARY_BOOT to 1 + - BootROM triggers WARM reset. GOTO 1.3) + - END + + - if SRC_GPR10 bit PERSIST_SECONDARY_BOOT is 1 + + - BootROM attempts to start bootloader B-copy + + - if B-copy valid + + - BootROM starts B-copy + - END + + - if B-copy NOT valid + - System hangs + - END + +Setup +----- +The bootloader A-copy must be placed at predetermined offset in SD/eMMC. The +bootloader B-copy area offset is determined by an offset stored in Secondary +Image Table (SIT). The SIT must be placed at predetermined offset in SD/eMMC. + +The following table contains offset of SIT, bootloader A-copy and recommended +bootloader B-copy offset. The offsets are in 512 Byte sector units (that is +offset 0x1 means 512 Bytes from the start of SD/eMMC card data partition). +For details on the addition of two numbers in recommended B-copy offset, see +SIT format below. + ++----------+--------------------+-----------------------+-----------------------------+ +| SoC | SIT offset (fixed) | A-copy offset (fixed) | B-copy offset (recommended) | ++----------+--------------------+-----------------------+-----------------------------+ +| iMX7D | 0x1 | 0x2 | 0x800+0x2 | ++----------+--------------------+-----------------------+-----------------------------+ +| iMX8MM | 0x41 | 0x42 | 0x1000+0x42 | ++----------+--------------------+-----------------------+-----------------------------+ + +SIT format +~~~~~~~~~~ +SIT is a 20 byte long structure containing of 5 32-bit words. Those encode +bootloader B-copy area offset (called "firstSectorNumber"), magic value +(called "tag") that is always 0x00112233, and three unused words set to 0. +SIT is documented in [1] and [2]. Example SIT are below:: + + $ hexdump -vC sit-mx7d.bin + 00000000 00 00 00 00 + 00000004 00 00 00 00 + 00000008 33 22 11 00 <--- This is the "tag" + 0000000c 00 08 00 00 <--- This is the "firstSectorNumber" + 00000010 00 00 00 00 + + $ hexdump -vC sit-mx8mm.bin + 00000000 00 00 00 00 + 00000004 00 00 00 00 + 00000008 33 22 11 00 <--- This is the "tag" + 0000000c 00 10 00 00 <--- This is the "firstSectorNumber" + 00000010 00 00 00 00 + +B-copy area offset ("firstSectorNumber") is offset, in units of 512 Byte +sectors, that is added to the start of boot media when switching between +A-copy and B-copy. For A-copy, this offset is 0x0. For B-copy, this offset +is determined by SIT (e.g. if firstSectorNumber is 0x1000 as it is above +in sit-mx8mm.bin, then the B-copy offset is 0x1000 sectors = 2 MiB). + +Bootloader A-copy (e.g. u-boot.imx or flash.bin) is placed at fixed offset +from A-copy area offset (e.g. 0x2 sectors from sector 0x0 for iMX7D, which +means u-boot.imx A-copy must be written to sector 0x2). + +The same applies to bootloader B-copy, which is placed at fixed offset from +B-copy area offset determined by SIT (e.g. 0x2 sectors from sector 0x800 [see +sit-mx7d.bin example above, this can be changed in SIT firstSectorNumber] for +iMX7D, which means u-boot.imx B-copy must be written to sector 0x802) + +**WARNING:** +B-copy area offset ("firstSectorNumber") is NOT equal to bootloader +(image, which is u-boot.imx or flash.bin) B-copy offset. + +To generate SIT, use for example the following bourne shell printf command:: + +$ printf '\x0\x0\x0\x0\x0\x0\x0\x0\x33\x22\x11\x00\x00\x08\x00\x00\x0\x0\x0\x0' > sit-mx7d.bin +$ printf '\x0\x0\x0\x0\x0\x0\x0\x0\x33\x22\x11\x00\x00\x10\x00\x00\x0\x0\x0\x0' > sit-mx8mm.bin + +Write bootloader A/B copy and SIT to SD/eMMC +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Examples of writing SIT and two copies of bootloader to SD or eMMC: + +- iMX8MM, SD card at /dev/sdX, Linux command line + :: + + $ dd if=sit-mx8mm.bin of=/dev/sdX bs=512 seek=65 + $ dd if=flash.bin of=/dev/sdX bs=512 seek=66 + $ dd if=flash.bin of=/dev/sdX bs=512 seek=4162 + +- iMX8MM, eMMC 1 data partition, U-Boot command line + :: + + => mmc partconf 1 0 0 0 + + => dhcp ${loadaddr} sit-mx8mm.bin + => mmc dev 1 + => mmc write ${loadaddr} 0x41 0x1 + + => dhcp ${loadaddr} flash.bin + => setexpr blkcnt ${filesize} + 0x1ff && setexpr blkcnt + => mmc dev 1 + => mmc write ${loadaddr} 0x42 ${blkcnt} + => mmc write ${loadaddr} 0x1042 ${blkcnt} + +WARM reset into B-copy using WDT +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +To perform a reboot into B-copy, the PERSIST_SECONDARY_BOOT must be set +in SRC_GPR0 register. Example on iMX8MM:: + + => mw 0x30390098 0x40000000 + +A WARM reset can be triggered using WDT as follows:: + + => mw.w 0x30280000 0x25 + +References +---------- + +.. [1] i.MX 7Dual Applications Processor Reference Manual, Rev. 1, 01/2018 ; section 6.6.5.3.5 Redundant boot support for expansion device +.. [2] i.MX 8M Mini Applications Processor Reference Manual, Rev. 3, 11/2020 ; section 6.1.5.4.5 Redundant boot support for expansion device diff --git a/doc/index.rst b/doc/index.rst index 4c44955d67f..5d5bb73846c 100644 --- a/doc/index.rst +++ b/doc/index.rst @@ -87,6 +87,17 @@ implementation. arch/index +Machine-specific doc +-------------------- + +These books provide programming details about machine-specific +implementation. + +.. toctree:: + :maxdepth: 2 + + imx/index + Board-specific doc ------------------ -- 2.30.2

3 2

[PATCH] pico-imx6ul: Pass the PMIC I2C address in pmic_get()
by Fabio Estevam 09 Apr '21

09 Apr '21

Pass "pfuze3000@8" in pmic_get() so that the PMIC node can be found in the devicetree. Signed-off-by: Fabio Estevam <festevam(a)gmail.com> --- board/technexion/pico-imx6ul/pico-imx6ul.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/board/technexion/pico-imx6ul/pico-imx6ul.c b/board/technexion/pico-imx6ul/pico-imx6ul.c index 62a54d0c8eb5..682c88dee78d 100644 --- a/board/technexion/pico-imx6ul/pico-imx6ul.c +++ b/board/technexion/pico-imx6ul/pico-imx6ul.c @@ -159,7 +159,7 @@ int power_init_board(void) struct udevice *dev; int ret, dev_id, rev_id; - ret = pmic_get("pfuze3000", &dev); + ret = pmic_get("pfuze3000@8", &dev); if (ret == -ENODEV) return 0; if (ret != 0) -- 2.25.1

3 2