September 2024 - U-Boot

[PATCH 0/3] Implement signing FIT images during image build
by al.kochet＠gmail.com 18 Oct '24

18 Oct '24

From: Alexander Kochetkov <al.kochet(a)gmail.com> Hello! I've done verified boot on Radxa Rock 3A. I've embedded public key in U-Boot SPL and signed FIT image configuration. All the work was done during U-Boot image build. For some use cases building and signing images in one go will be much simple, than building unsigned images and signing later. For example SPL-image for rk3568 called idbloader.img consist of TPL, U-boot SPL and U-boot SPL DTB with public key. So in order to assemble signed idbloader.img lately we have to keep all the intermediate files used during build. To embed public key, I've replaced u-boot-spl node with blob-ext and generated u-boot-spl-with-pubkey-dtb blob using u-boot-spl-pubkey-dtb entry. To sign FIT image I've used newly implemented fit property 'fit,sign'. I haven't sign FIT image nodes, because I had realized that signing configuration is safe and sufficient for verified boot. But I doubt. So I've left that signing scheme in the test. What do you think, is using signed configuration and signed images at the same time is much safer or doesn't provide any benefits? Now I thinking about implementing configuration option, something like FIT_SIGNATURE_KEYDIR. The value of the option will be passed to binman using -I. Alsi I want to embed another public key in the configuration DTB, so it will be used to verify kernel FIT. But I couldn't figure out how to do it using binman. &binman { u-boot-spl-with-pubkey-dtb { filename = "u-boot-spl-with-pubkey-dtb.bin"; u-boot-spl-nodtb { }; u-boot-spl-pubkey-dtb { algo = "sha256,rsa2048"; required = "conf"; key-name-hint = "uboot-spl"; }; }; simple-bin { ... mkimage { ... #ifdef CONFIG_ROCKCHIP_EXTERNAL_TPL rockchip-tpl { }; #elif defined(CONFIG_TPL) u-boot-tpl { }; #endif blob-ext { filename = "u-boot-spl-with-pubkey-dtb.bin"; }; }; fit: fit { ... fit,sign; ... configurations { default = "@config-DEFAULT-SEQ"; @config-SEQ { ... #ifdef CONFIG_SPL_FIT_SIGNATURE signature { algo = "sha256,rsa2048"; key-name-hint = "uboot-spl"; sign-images = "firmware", "loadables", "fdt"; }; #endif }; }; }; }; } Alexander Kochetkov (3): binman: fix passing loadables to mkimage on first run image-host: fix 'unknown error' error message binman: implement signing FIT images during image build tools/binman/btool/mkimage.py | 5 +- tools/binman/entries.rst | 7 ++ tools/binman/etype/fit.py | 57 +++++++++++++- tools/binman/ftest.py | 95 ++++++++++++++++++++++++ tools/binman/test/326_fit_signature.dts | 98 +++++++++++++++++++++++++ tools/binman/test/326_rsa2048.key | 28 +++++++ tools/binman/test/327_fit_signature.dts | 98 +++++++++++++++++++++++++ tools/binman/test/328_fit_signature.dts | 61 +++++++++++++++ tools/image-host.c | 2 +- 9 files changed, 446 insertions(+), 5 deletions(-) create mode 100644 tools/binman/test/326_fit_signature.dts create mode 100644 tools/binman/test/326_rsa2048.key create mode 100644 tools/binman/test/327_fit_signature.dts create mode 100644 tools/binman/test/328_fit_signature.dts -- 2.17.1

2 10

[PATCH] buildman: Correct some warnings about regex strings
by Simon Glass 18 Oct '24

18 Oct '24

With Python 3.12 some warnings have cropped up. Fix them. Signed-off-by: Simon Glass <sjg(a)chromium.org> --- tools/buildman/builder.py | 6 +++--- tools/buildman/toolchain.py | 8 ++++---- 2 files changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/buildman/builder.py b/tools/buildman/builder.py index 4090d328b30..cbf1345281b 100644 --- a/tools/buildman/builder.py +++ b/tools/buildman/builder.py @@ -34,7 +34,7 @@ from u_boot_pylib.terminal import tprint # Error in reading or end of file. # << # which indicates that BREAK_ME has an empty default -RE_NO_DEFAULT = re.compile(b'$(\w+)$ \[] $NEW$') +RE_NO_DEFAULT = re.compile(br'$(\w+)$ \[] $NEW$') # Symbol types which appear in the bloat feature (-B). Others are silently # dropped when reading in the 'nm' output @@ -374,9 +374,9 @@ class Builder: self._re_function = re.compile('(.*): In function.*') self._re_files = re.compile('In file included from.*') - self._re_warning = re.compile('(.*):(\d*):(\d*): warning: .*') + self._re_warning = re.compile(r'(.*):(\d*):(\d*): warning: .*') self._re_dtb_warning = re.compile('(.*): Warning .*') - self._re_note = re.compile('(.*):(\d*):(\d*): note: this is the location of the previous.*') + self._re_note = re.compile(r'(.*):(\d*):(\d*): note: this is the location of the previous.*') self._re_migration_warning = re.compile(r'^={21} WARNING ={22}\n.*\n=+\n', re.MULTILINE | re.DOTALL) diff --git a/tools/buildman/toolchain.py b/tools/buildman/toolchain.py index 6ca79c2c0f9..da8fd1b072c 100644 --- a/tools/buildman/toolchain.py +++ b/tools/buildman/toolchain.py @@ -438,12 +438,12 @@ class Toolchains: This converts ${blah} within the string to the value of blah. This function works recursively. + Resolved string + Args: var_dict: Dictionary containing variables and their values args: String containing make arguments Returns: - Resolved string - >>> bsettings.setup(None) >>> tcs = Toolchains() >>> tcs.Add('fred', False) @@ -454,7 +454,7 @@ class Toolchains: >>> tcs.ResolveReferences(var_dict, 'this=${oblique}_set${first}nd') 'this=OBLIQUE_setfi2ndrstnd' """ - re_var = re.compile('(\$\{[-_a-z0-9A-Z]{1,}\})') + re_var = re.compile(r'(\$\{[-_a-z0-9A-Z]{1,}\})') while True: m = re_var.search(args) @@ -493,7 +493,7 @@ class Toolchains: self._make_flags['target'] = brd.target arg_str = self.ResolveReferences(self._make_flags, self._make_flags.get(brd.target, '')) - args = re.findall("(?:\".*?\"|\S)+", arg_str) + args = re.findall(r"(?:\".*?\"|\S)+", arg_str) i = 0 while i < len(args): args[i] = args[i].replace('"', '') -- 2.43.0

1 1

[PATCH 1/3] u_boot_pylib: Support running coverage on selected functions
by Simon Glass 18 Oct '24

18 Oct '24

At present run_test_coverage() assumes you want code coverage for the entire code base. This is the normal situation, but sometimes it is useful to see the coverage provided by just a single test. Add support for this. Signed-off-by: Simon Glass <sjg(a)chromium.org> --- tools/u_boot_pylib/test_util.py | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/tools/u_boot_pylib/test_util.py b/tools/u_boot_pylib/test_util.py index 857ce58c98c..dd671965263 100644 --- a/tools/u_boot_pylib/test_util.py +++ b/tools/u_boot_pylib/test_util.py @@ -23,8 +23,9 @@ except: use_concurrent = False -def run_test_coverage(prog, filter_fname, exclude_list, build_dir, required=None, - extra_args=None, single_thread='-P1'): +def run_test_coverage(prog, filter_fname, exclude_list, build_dir, + required=None, extra_args=None, single_thread='-P1', + args=None): """Run tests and check that we get 100% coverage Args: @@ -42,6 +43,7 @@ def run_test_coverage(prog, filter_fname, exclude_list, build_dir, required=None single_thread (str): Argument string to make the tests run single-threaded. This is necessary to get proper coverage results. The default is '-P0' + args (list of str): List of tests to run, or None to run all Raises: ValueError if the code coverage is not 100% @@ -66,9 +68,10 @@ def run_test_coverage(prog, filter_fname, exclude_list, build_dir, required=None 'coverage') cmd = ('%s%s run ' - '--omit "%s" %s %s %s %s' % (prefix, covtool, ','.join(glob_list), - prog, extra_args or '', test_cmd, - single_thread or '-P1')) + '--omit "%s" %s %s %s %s %s' % (prefix, covtool, ','.join(glob_list), + prog, extra_args or '', test_cmd, + single_thread or '-P1', + ' '.join(args) if args else '')) os.system(cmd) stdout = command.output(covtool, 'report') lines = stdout.splitlines() -- 2.34.1

3 7

[PATCH] sandbox_spl: Restrict use of UPL when not enabled
by Simon Glass 18 Oct '24

18 Oct '24

With sandbox_spl we want to use the file-based boot in CI, so that this flow is tested. The recent UPL change enabled booting via that method, thus overriding the file-based boot. Correct this by using UPL only when the --upl flag is given. Signed-off-by: Simon Glass <sjg(a)chromium.org> Fixes: 91fde8e1764 ("sandbox: Add an SPL loader for UPL") --- arch/sandbox/cpu/spl.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/sandbox/cpu/spl.c b/arch/sandbox/cpu/spl.c index 1c33a520c64..76638362564 100644 --- a/arch/sandbox/cpu/spl.c +++ b/arch/sandbox/cpu/spl.c @@ -55,9 +55,10 @@ void board_init_f(ulong flag) void board_boot_order(u32 *spl_boot_list) { + struct sandbox_state *state = state_get_current(); + spl_boot_list[0] = BOOT_DEVICE_VBE; - spl_boot_list[1] = BOOT_DEVICE_UPL; - spl_boot_list[2] = BOOT_DEVICE_BOARD; + spl_boot_list[1] = state->upl ? BOOT_DEVICE_UPL : BOOT_DEVICE_BOARD; } static int spl_board_load_file(struct spl_image_info *spl_image, -- 2.43.0

2 3

[PATCH] efi_loader: Don't allocate from Special purpose memory
by Ilias Apalodimas 17 Oct '24

17 Oct '24

The EFI spec defines special-purpose memory in §7.2. That memory serves as a hint to the OS to avoid allocating this memory for core OS data or code that can not be relocated. So let's ignore it when allocating from conventional memory. Signed-off-by: Ilias Apalodimas <ilias.apalodimas(a)linaro.org> --- lib/efi_loader/efi_memory.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/lib/efi_loader/efi_memory.c b/lib/efi_loader/efi_memory.c index c6f1dd09456e..74732e37f8aa 100644 --- a/lib/efi_loader/efi_memory.c +++ b/lib/efi_loader/efi_memory.c @@ -422,7 +422,8 @@ static efi_status_t efi_check_allocated(u64 addr, bool must_be_allocated) if (addr >= start && addr < end) { if (must_be_allocated ^ - (item->desc.type == EFI_CONVENTIONAL_MEMORY)) + (item->desc.type == EFI_CONVENTIONAL_MEMORY) && + !(item->desc.attribute & EFI_MEMORY_SP)) return EFI_SUCCESS; else return EFI_NOT_FOUND; @@ -460,6 +461,9 @@ static uint64_t efi_find_free_memory(uint64_t len, uint64_t max_addr) if (desc->type != EFI_CONVENTIONAL_MEMORY) continue; + if (desc->attribute & EFI_MEMORY_SP) + continue; + /* Out of bounds for max_addr */ if ((ret + len) > max_addr) continue; -- 2.45.2

3 5

[PATCH] ARM: stm32: Add script to install U-Boot from SD/eMMC to SPI NOR on DH STM32MP15xx DHSOM
by Marek Vasut 17 Oct '24

17 Oct '24

Make the dh_update_sd_to_sf script generic, rename it to dh_update_sdmmc_to_sf and implement two specific dh_update_sd_to_sf and dh_update_emmc_to_sf scripts which load U-Boot from either SD or eMMC and install it into SPI NOR. Signed-off-by: Marek Vasut <marex(a)denx.de> --- Cc: Patrice Chotard <patrice.chotard(a)foss.st.com> Cc: Patrick Delaunay <patrick.delaunay(a)foss.st.com> Cc: Tom Rini <trini(a)konsulko.com> Cc: u-boot(a)dh-electronics.com Cc: u-boot(a)lists.denx.de Cc: uboot-stm32(a)st-md-mailman.stormreply.com --- include/configs/stm32mp15_dh_dhsom.h | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/include/configs/stm32mp15_dh_dhsom.h b/include/configs/stm32mp15_dh_dhsom.h index de39b19219d..fa759c938b3 100644 --- a/include/configs/stm32mp15_dh_dhsom.h +++ b/include/configs/stm32mp15_dh_dhsom.h @@ -38,16 +38,26 @@ "setenv loadaddr1 && " \ "setenv sblkcnt && " \ "setenv ublkcnt\0" \ - "dh_update_sd_to_sf=" /* Erase SPI NOR and install U-Boot from SD */ \ + "dh_update_sdmmc_to_sf=" /* Erase SPI NOR and install U-Boot from SD/eMMC */ \ "setexpr loadaddr1 ${loadaddr} + 0x1000000 && " \ - "load mmc 0:4 ${loadaddr1} /boot/u-boot-spl.stm32 && " \ + "load ${dh_update_iface} ${dh_update_dev} " \ + "${loadaddr1} /boot/u-boot-spl.stm32 && " \ "env set filesize1 ${filesize} && " \ - "load mmc 0:4 ${loadaddr} /boot/u-boot.itb && " \ + "load ${dh_update_iface} ${dh_update_dev} " \ + "${loadaddr} /boot/u-boot.itb && " \ "sf probe && sf erase 0 0x200000 && " \ "sf update ${loadaddr1} 0 ${filesize1} && " \ "sf update ${loadaddr1} 0x40000 ${filesize1} && " \ "sf update ${loadaddr} 0x80000 ${filesize} && " \ "env set filesize1 && env set loadaddr1\0" \ + "dh_update_sd_to_sf=" /* Erase SPI NOR and install U-Boot from SD */ \ + "setenv dh_update_iface mmc && " \ + "setenv dh_update_dev 0:4 && " \ + "run dh_update_sdmmc_to_sf\0" \ + "dh_update_emmc_to_sf=" /* Erase SPI NOR and install U-Boot from eMMC */ \ + "setenv dh_update_iface mmc && " \ + "setenv dh_update_dev 1:4 && " \ + "run dh_update_sdmmc_to_sf\0" \ "stdin=serial\0" \ "stdout=serial\0" \ "stderr=serial\0" \ -- 2.45.2

2 4

[PATCH v2] cmd: setexpr: fix no matching string in gsub return empty value
by Massimiliano Minella 16 Oct '24

16 Oct '24

From: Massimiliano Minella <massimiliano.minella(a)se.com> In gsub, when the destination string is empty, the string 't' is provided and the regular expression doesn't match, then the final result is an empty string. Example: => echo ${foo} => setenv foo => setexpr foo gsub e a bar => echo ${foo} => The variable ${foo} should contain "bar" and the lack of match shouldn't be considered an error. This patch fixes the erroneous behavior by removing the return statement and breaking out of the loop in case of lack of match. Also add a test for the no match case. Signed-off-by: Massimiliano Minella <massimiliano.minella(a)se.com> --- Changes in V2: - update documentation to describe the behavior cmd/setexpr.c | 9 ++++----- doc/usage/cmd/setexpr.rst | 1 + test/cmd/setexpr.c | 10 ++++++++++ 3 files changed, 15 insertions(+), 5 deletions(-) diff --git a/cmd/setexpr.c b/cmd/setexpr.c index 233471f6cb..ab76824a32 100644 --- a/cmd/setexpr.c +++ b/cmd/setexpr.c @@ -216,14 +216,12 @@ int setexpr_regex_sub(char *data, uint data_size, char *nbuf, uint nbuf_size, if (res == 0) { if (loop == 0) { debug("%s: No match\n", data); - return 1; } else { - break; + debug("## MATCH ## %s\n", data); } + break; } - debug("## MATCH ## %s\n", data); - if (!s) return 1; @@ -540,7 +538,8 @@ U_BOOT_CMD( " - For each substring matching the regular expression <r> in the\n" " string <t>, substitute the string <s>. The result is\n" " assigned to <name>. If <t> is not supplied, use the old\n" - " value of <name>\n" + " value of <name>. If no substring matching <r> is found in <t>,\n" + " assign <t> to <name>.\n" "setexpr name sub r s [t]\n" " - Just like gsub(), but replace only the first matching substring" #endif diff --git a/doc/usage/cmd/setexpr.rst b/doc/usage/cmd/setexpr.rst index d245a13ca8..593a0ea91e 100644 --- a/doc/usage/cmd/setexpr.rst +++ b/doc/usage/cmd/setexpr.rst @@ -39,6 +39,7 @@ setexpr name gsub <r> <s> [<t>] string <t>, substitute the string <s>. The result is assigned to <name>. If <t> is not supplied, use the old value of <name>. + If no substring matching <r> is found in <t>, assign <t> to <name>. setexpr name sub <r> <s> [<t>] Just like gsub(), but replace only the first matching substring diff --git a/test/cmd/setexpr.c b/test/cmd/setexpr.c index 312593e1e3..ee329e94b8 100644 --- a/test/cmd/setexpr.c +++ b/test/cmd/setexpr.c @@ -179,6 +179,16 @@ static int setexpr_test_regex(struct unit_test_state *uts) val = env_get("mary"); ut_asserteq_str("this is a test", val); + /* No match */ + ut_assertok(run_command("setenv fred 'this is a test'", 0)); + ut_assertok(run_command("setenv mary ''", 0)); + ut_assertok(run_command("setexpr fred gsub us is \"${fred}\"", 0)); + ut_assertok(run_command("setexpr mary gsub us is \"${fred}\"", 0)); + val = env_get("fred"); + ut_asserteq_str("this is a test", val); + val = env_get("mary"); + ut_asserteq_str("this is a test", val); + unmap_sysmem(buf); return 0; -- 2.43.0

4 6

[PATCH 0/2] board: libre-computer: Add support for Libre Computer aml-a311d-cc & aml-s905d3-cc boards
by Neil Armstrong 16 Oct '24

16 Oct '24

Add support for the Libre Computer aml-a311d-cc "Alta" & aml-s905d3-cc "Solitude" boards: https://libre.computer/products/aml-a311d-cc/ https://libre.computer/products/aml-s905d3-cc/ The Alta & Solitude boards have a Credit Card form factor, similar to the the previous "Le Potato" card, but with the Amlogic A311D or S903D3 SoCS, MIPI DSI and CSI connectors. PoE header and a single USB2 Type-C connector replacing the microUSB one for power and USB 2.0. The boards have an embedded SPI NOR flash, and EFI Capsule support is added. Signed-off-by: Neil Armstrong <neil.armstrong(a)linaro.org> --- Neil Armstrong (2): ARM: meson: add support for Libre Computer aml-a311d-cc ARM: meson: add support for Libre Computer aml-s905d3-cc .../dts/meson-g12b-a311d-libretech-cc-u-boot.dtsi | 15 +++ .../dts/meson-sm1-s905d3-libretech-cc-u-boot.dtsi | 15 +++ board/libre-computer/aml-a311d-cc/MAINTAINERS | 7 ++ board/libre-computer/aml-a311d-cc/Makefile | 6 ++ board/libre-computer/aml-a311d-cc/aml-a311d-cc.c | 44 +++++++++ board/libre-computer/aml-s905d3-cc/MAINTAINERS | 7 ++ board/libre-computer/aml-s905d3-cc/Makefile | 6 ++ board/libre-computer/aml-s905d3-cc/aml-s905d3-cc.c | 44 +++++++++ configs/aml-a311d-cc_defconfig | 108 +++++++++++++++++++++ configs/aml-s905d3-cc_defconfig | 108 +++++++++++++++++++++ doc/board/amlogic/aml-a311d-cc.rst | 46 +++++++++ doc/board/amlogic/aml-s905d3-cc.rst | 46 +++++++++ doc/board/amlogic/index.rst | 2 + 13 files changed, 454 insertions(+) --- base-commit: 13fd7a17858e5f2555a1d68e4afa75fceda0dac0 change-id: 20240920-u-boot-topic-libre-computer-solitude-alta-c3b22cbd16df Best regards, -- Neil Armstrong <neil.armstrong(a)linaro.org>

2 5

[PATCH v6 08/12] efi_loader: Disable ANSI output for tests
by Simon Glass 15 Oct '24

15 Oct '24

We don't want ANSI characters written in tests since it is a pain to check the output with ut_assert_nextline() et al. Provide a way to tests to request that ANSI characters not be sent. Add a proper function comment while we are here, to encourage others. Signed-off-by: Simon Glass <sjg(a)chromium.org> --- (no changes since v1) include/efi_loader.h | 21 ++++++++++++++++++++- lib/efi_loader/efi_console.c | 26 +++++++++++++++++--------- 2 files changed, 37 insertions(+), 10 deletions(-) diff --git a/include/efi_loader.h b/include/efi_loader.h index f84852e384f..82b90ee0f1d 100644 --- a/include/efi_loader.h +++ b/include/efi_loader.h @@ -531,8 +531,27 @@ efi_status_t efi_bootmgr_delete_boot_option(u16 boot_index); efi_status_t efi_bootmgr_run(void *fdt); /* search the boot option index in BootOrder */ bool efi_search_bootorder(u16 *bootorder, efi_uintn_t num, u32 target, u32 *index); -/* Set up console modes */ + +/** + * efi_setup_console_size() - update the mode table. + * + * By default the only mode available is 80x25. If the console has at least 50 + * lines, enable mode 80x50. If we can query the console size and it is neither + * 80x25 nor 80x50, set it as an additional mode. + */ void efi_setup_console_size(void); + +/** + * efi_console_set_ansi() - Set whether ANSI characters should be emitted + * + * These characters mess up tests which use ut_assert_nextline(). Call this + * function to tell efi_loader not to emit these characters when starting up the + * terminal + * + * @allow_ansi: Allow emitting ANSI characters + */ +void efi_console_set_ansi(bool allow_ansi); + /* Set up load options from environment variable */ efi_status_t efi_env_set_load_options(efi_handle_t handle, const char *env_var, u16 **load_options); diff --git a/lib/efi_loader/efi_console.c b/lib/efi_loader/efi_console.c index cea50c748aa..569fc9199bc 100644 --- a/lib/efi_loader/efi_console.c +++ b/lib/efi_loader/efi_console.c @@ -30,6 +30,17 @@ struct cout_mode { __maybe_unused static struct efi_object uart_obj; +/* + * suppress emission of ANSI codes for use by unit tests. Leave it as 0 for the + * default behaviour + */ +static bool no_ansi; + +void efi_console_set_ansi(bool allow_ansi) +{ + no_ansi = !allow_ansi; +} + static struct cout_mode efi_cout_modes[] = { /* EFI Mode 0 is 80x25 and always present */ { @@ -348,13 +359,6 @@ static int __maybe_unused query_vidconsole(int *rows, int *cols) return 0; } -/** - * efi_setup_console_size() - update the mode table. - * - * By default the only mode available is 80x25. If the console has at least 50 - * lines, enable mode 80x50. If we can query the console size and it is neither - * 80x25 nor 80x50, set it as an additional mode. - */ void efi_setup_console_size(void) { int rows = 25, cols = 80; @@ -362,8 +366,12 @@ void efi_setup_console_size(void) if (IS_ENABLED(CONFIG_VIDEO)) ret = query_vidconsole(&rows, &cols); - if (ret) - ret = query_console_serial(&rows, &cols); + if (ret) { + if (no_ansi) + ret = 0; + else + ret = query_console_serial(&rows, &cols); + } if (ret) return; -- 2.43.0

5 20

[2nd RESEND PATCH v2 0/6] mtd: Make sure UBIFS does not do multi-pass page programming on flashes that don't support it
by tkuw584924＠gmail.com 15 Oct '24

15 Oct '24

From: Takahiro Kuwano <Takahiro.Kuwano(a)infineon.com> Resent after modifying commit message in #5 (mention Linux commit first) and collecting A-b tags. This series is equivalent to the one for Linux MTD submitted by Pratyush Yadav. https://patchwork.ozlabs.org/project/linux-mtd/list/?series=217759&state=* Changes in v2: - Fix an issue in setting macronix_octal_fixups - Rework fixup hooks Takahiro Kuwano (6): mtd: ubi: Do not zero out EC and VID on ECC-ed NOR flashes mtd: spi-nor: Allow flashes to specify MTD writesize mtd: spi-nor: Check nor->info before setting macronix_octal_fixups mtd: spi-nor: Replace default_init() hook with late_init() mtd: spi-nor: Call spi_nor_post_sfdp_fixups() only after spi_nor_parse_sfdp() mtd: spi-nor: Set ECC unit size to MTD writesize in Infineon SEMPER flashes drivers/mtd/spi/spi-nor-core.c | 88 +++++++++++++++++++++------------- drivers/mtd/ubi/build.c | 4 +- drivers/mtd/ubi/io.c | 9 +++- include/linux/mtd/spi-nor.h | 1 + 4 files changed, 65 insertions(+), 37 deletions(-) -- 2.34.1

3 8