December 2024 - U-Boot

[PATCH v4 00/15] Make EFI memory allocations synchronous with LMB
by Sughosh Ganu 20 Jan '25

20 Jan '25

This is part two of the series to have the EFI and LMB modules have a coherent view of memory. Part one of this goal was to change the LMB module to have a global and persistent memory map. Those patches have now been applied to the next branch. These patches are changing the EFI memory allocation API's such that they rely on the LMB module to allocate RAM memory. This fixes the current scenario where the EFI memory module has no visibility of the allocations/reservations made by the LMB module. One thing to note here is that this is limited to the RAM memory region, i.e. the EFI_CONVENTIONAL_MEMORY type. Any other memory type that is to be added to the EFI memory map, still gets handled by the EFI memory module. Changes since V3: * Add comments for the LMB_NOOVERWRITE and LMB_NONOTIFY flags * Drop use of is_addr_in_ram() function * Drop use of CONFIG_MEM_MAP_UPDATE_NOTIFY symbol to check if the notification needs to be sent. * s/lmb_notify/lmb_should_notify * Put a check for EFI_LOADER in the lmb_should_notify() function Some test logs to highlight the issue that is being fixed by the series. Without patch series -------------------- lmb_dump_all: memory.count = 0x1 memory[0] [0x40000000-0x820fffff], 0x42100000 bytes flags: none reserved.count = 0x3 reserved[0] [0xe100000-0xeffffff], 0x00f00000 bytes flags: no-map reserved[1] [0x42000000-0x421fffff], 0x00200000 bytes flags: no-map reserved[2] [0x7f77da00-0x820fffff], 0x02982600 bytes flags: no-overwrite => efidebug memmap -- does not show regions allocated by lmb Missing TPMv2 device for EFI_TCG_PROTOCOL Type Start End Attributes ================ ================ ================ ========== CONVENTIONAL 0000000040000000-000000007f751000 WB BOOT DATA 000000007f751000-000000007f756000 WB RUNTIME DATA 000000007f756000-000000007f757000 WB|RT BOOT DATA 000000007f757000-000000007f758000 WB RUNTIME DATA 000000007f758000-000000007f77a000 WB|RT BOOT DATA 000000007f77a000-000000007f781000 WB BOOT CODE 000000007f781000-00000000807b5000 WB RUNTIME DATA 00000000807b5000-00000000807b6000 WB|RT BOOT CODE 00000000807b6000-00000000817c0000 WB RUNTIME CODE 00000000817c0000-00000000817d0000 WB|RT BOOT CODE 00000000817d0000-0000000082100000 WB => Trying to allocate EFI memory with already allocated region succeeds(should fail) --------------------------------------------------------------------------------- => efi_mem alloc 2000 42000000 Address returned 0x42000000 => efidebug memmap Type Start End Attributes ================ ================ ================ ========== CONVENTIONAL 0000000040000000-0000000042000000 WB BOOT DATA 0000000042000000-0000000042002000 WB CONVENTIONAL 0000000042002000-000000007f751000 WB BOOT DATA 000000007f751000-000000007f756000 WB RUNTIME DATA 000000007f756000-000000007f757000 WB|RT BOOT DATA 000000007f757000-000000007f758000 WB RUNTIME DATA 000000007f758000-000000007f77a000 WB|RT BOOT DATA 000000007f77a000-000000007f781000 WB BOOT CODE 000000007f781000-00000000807b5000 WB RUNTIME DATA 00000000807b5000-00000000807b6000 WB|RT BOOT CODE 00000000807b6000-00000000817c0000 WB RUNTIME CODE 00000000817c0000-00000000817d0000 WB|RT BOOT CODE 00000000817d0000-0000000082100000 WB => With patch series ----------------- lmb_dump_all: memory.count = 0x1 memory[0] [0x40000000-0x820fffff], 0x42100000 bytes flags: none reserved.count = 0x4 reserved[0] [0xe100000-0xeffffff], 0x00f00000 bytes flags: no-map reserved[1] [0x42000000-0x421fffff], 0x00200000 bytes flags: no-map reserved[2] [0x7f74f000-0x7f77dfff], 0x0002f000 bytes flags: no-notify, no-overwrite reserved[3] [0x7f77ea00-0x820fffff], 0x02981600 bytes flags: no-overwrite => efidebug memmap Type Start End Attributes ================ ================ ================ ========== BOOT DATA 000000000e100000-000000000f000000 WB CONVENTIONAL 0000000040000000-0000000042000000 WB BOOT DATA 0000000042000000-0000000042200000 WB CONVENTIONAL 0000000042200000-000000007f74e000 WB BOOT DATA 000000007f74e000-000000007f753000 WB RUNTIME DATA 000000007f753000-000000007f754000 WB|RT BOOT DATA 000000007f754000-000000007f755000 WB RUNTIME DATA 000000007f755000-000000007f777000 WB|RT BOOT DATA 000000007f777000-00000000807b6000 WB RUNTIME DATA 00000000807b6000-00000000807b7000 WB|RT BOOT DATA 00000000807b7000-00000000817c0000 WB RUNTIME CODE 00000000817c0000-00000000817d0000 WB|RT BOOT DATA 00000000817d0000-0000000082100000 WB Trying to allocate EFI memory with already allocated region fails ----------------------------------------------------------------- => efi_mem alloc 2000 42000000 efi_allocate_pages failed 800000000000000e => Trying to allocate EFI memory with non-allocated region succeeds ---------------------------------------------------------------- => efi_mem alloc 2000 42200000 Address returned 0x42200000 => efidebug memmap Type Start End Attributes ================ ================ ================ ========== BOOT DATA 000000000e100000-000000000f000000 WB CONVENTIONAL 0000000040000000-0000000042000000 WB BOOT DATA 0000000042000000-0000000042202000 WB CONVENTIONAL 0000000042202000-000000007f74d000 WB BOOT DATA 000000007f74d000-000000007f752000 WB RUNTIME DATA 000000007f752000-000000007f753000 WB|RT BOOT DATA 000000007f753000-000000007f754000 WB RUNTIME DATA 000000007f754000-000000007f776000 WB|RT BOOT DATA 000000007f776000-00000000807b5000 WB RUNTIME DATA 00000000807b5000-00000000807b6000 WB|RT BOOT DATA 00000000807b6000-00000000817c0000 WB RUNTIME CODE 00000000817c0000-00000000817d0000 WB|RT BOOT DATA 00000000817d0000-0000000082100000 WB => lmb_dump_all: memory.count = 0x1 memory[0] [0x40000000-0x820fffff], 0x42100000 bytes flags: none reserved.count = 0x5 reserved[0] [0xe100000-0xeffffff], 0x00f00000 bytes flags: no-map reserved[1] [0x42000000-0x421fffff], 0x00200000 bytes flags: no-map reserved[2] [0x42200000-0x42201fff], 0x00002000 bytes flags: no-notify, no-overwrite reserved[3] [0x7f74e000-0x7f77cfff], 0x0002f000 bytes flags: no-notify, no-overwrite reserved[4] [0x7f77da00-0x820fffff], 0x02982600 bytes flags: no-overwrite Sughosh Ganu (15): lmb: add versions of the lmb API with flags lmb: add a flag to allow suppressing memory map change notification lmb: add and reserve memory above ram_top efi: memory: use the lmb API's for allocating and freeing memory lmb: notify of any changes to the LMB memory map efi_memory: do not add U-Boot memory to the memory map ti: k3: remove efi_add_known_memory() function definition stm32mp: remove efi_add_known_memory() function definition lmb: allow for boards to specify memory map layerscape: use the lmb API's to add RAM memory x86: e820: use the lmb API for adding RAM memory efi_memory: do not add RAM memory to the memory map lmb: remove call to efi_lmb_reserve() efi_memory: rename variable to highlight overlap with free memory lmb: replace the double-underscore with single-underscore for all functions arch/arm/cpu/armv8/fsl-layerscape/cpu.c | 8 +- arch/arm/mach-k3/common.c | 11 -- arch/arm/mach-stm32mp/dram_init.c | 11 -- arch/x86/lib/e820.c | 47 +++-- include/efi_loader.h | 27 ++- include/lmb.h | 63 ++++++ lib/Kconfig | 19 ++ lib/efi_loader/Kconfig | 1 + lib/efi_loader/efi_memory.c | 217 ++++++--------------- lib/lmb.c | 242 ++++++++++++++++++------ 10 files changed, 389 insertions(+), 257 deletions(-) -- 2.34.1

5 22

[PATCH 1/1] test: str_ut.c depends on CONFIG_STRTO
by Heinrich Schuchardt 19 Jan '25

19 Jan '25

The string conversion functions are implemented in lib/strto.c which is only compiled if CONFIG_STRTO=y. Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt(a)canonical.com> --- test/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/test/Makefile b/test/Makefile index 145c952d2c3..efd5ba7acfd 100644 --- a/test/Makefile +++ b/test/Makefile @@ -21,7 +21,7 @@ ifneq ($(CONFIG_HUSH_PARSER),) obj-$(CONFIG_$(XPL_)CMDLINE) += hush/ endif obj-$(CONFIG_$(XPL_)CMDLINE) += print_ut.o -obj-$(CONFIG_$(XPL_)CMDLINE) += str_ut.o +obj-$(CONFIG_$(PHASE_)STRTO) += str_ut.o obj-$(CONFIG_UT_TIME) += time_ut.o obj-y += ut.o -- 2.45.2

2 2

[PATCH v7 0/9] add the support of sha256_hmac and sha256_hkdf
by Philippe Reynes 19 Jan '25

19 Jan '25

This serie adds the support of sha256_hmac and sha256_hkdf. A first version was sent several months ago just before the integration of mbedtls. This new version is based on mbedtls. The first patch of this serie add the support of hkdf using mbedtls. v7: - add missing reviewed-by v6: - use const for test - set buffer to 0 before using to to compute hmac v5: - use mbedtls to compute hmac - add tests for hkdf - enabled mbedtls in sandbox_defconfig - some cleanup v4: - fix computation of the hmac - add more test for hmac - move sha256 common functions to sha256_common.c - some cleanup v3: - also define sha256_hmac for legacy sha256 - add some #if to define function only when needed Philippe Reynes (9): tools: kwbimage.h: use linux/compiler_attributes.h tools: renesas_spkgimage.h: use linux/compiler_attributes.h mbedtls: enable support of hkdf lib: sha256: move common function to sha256_common.c lib: sha256: add feature sha256_hmac test: lib: add test for sha256_hmac lib: mbedtls: sha256: add support of key derivation test: lib: add test for key derivation configs: sandbox: enable mbedtls configs/sandbox_defconfig | 2 + include/u-boot/sha256.h | 24 +++ lib/Makefile | 1 + lib/mbedtls/Kconfig | 14 ++ lib/mbedtls/Makefile | 2 + lib/mbedtls/mbedtls_def_config.h | 4 + lib/mbedtls/sha256.c | 59 ++++--- lib/sha256.c | 67 ++++--- lib/sha256_common.c | 50 ++++++ test/lib/Makefile | 2 + test/lib/test_sha256_hkdf.c | 198 +++++++++++++++++++++ test/lib/test_sha256_hmac.c | 294 +++++++++++++++++++++++++++++++ tools/Makefile | 1 + tools/kwbimage.h | 6 +- tools/renesas_spkgimage.h | 6 +- 15 files changed, 670 insertions(+), 60 deletions(-) create mode 100644 lib/sha256_common.c create mode 100644 test/lib/test_sha256_hkdf.c create mode 100644 test/lib/test_sha256_hmac.c -- 2.25.1

3 14

[PATCH v2 1/4] net: phy: Port set/clear bits from Linux
by Paul Barker 19 Jan '25

19 Jan '25

To simply porting phy drivers from Linux to U-Boot, define phy_set_bits() and phy_clear_bits() functions with a similar API to those used in Linux. The U-Boot versions of these functions include the `devad` argument which is not present in the Linux versions, to keep them aligned with the other phy functions in U-Boot. Signed-off-by: Paul Barker <paul.barker.ct(a)bp.renesas.com> Reviewed-by: Marek Vasut <marek.vasut+renesas(a)mailbox.org> --- Changes v1->v2: - Split out of series adding RZ/G2L Ethernet support [1] - Added Marek's Reviewed-by tag [1]: https://lore.kernel.org/all/20241024152448.102-1-paul.barker.ct@bp.renesas.… include/phy.h | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/include/phy.h b/include/phy.h index 36785031eeb0..510b0a21831b 100644 --- a/include/phy.h +++ b/include/phy.h @@ -333,6 +333,28 @@ int gen10g_startup(struct phy_device *phydev); int gen10g_shutdown(struct phy_device *phydev); int gen10g_discover_mmds(struct phy_device *phydev); +/** + * phy_set_bits - Convenience function for setting bits in a PHY register + * @phydev: the phy_device struct + * @regnum: register number to write + * @val: bits to set + */ +static inline int phy_set_bits(struct phy_device *phydev, int devad, u32 regnum, u16 val) +{ + return phy_modify(phydev, devad, regnum, 0, val); +} + +/** + * phy_clear_bits - Convenience function for clearing bits in a PHY register + * @phydev: the phy_device struct + * @regnum: register number to write + * @val: bits to clear + */ +static inline int phy_clear_bits(struct phy_device *phydev, int devad, u32 regnum, u16 val) +{ + return phy_modify(phydev, devad, regnum, val, 0); +} + /** * U_BOOT_PHY_DRIVER() - Declare a new U-Boot driver * @__name: name of the driver -- 2.43.0

3 11

[PATCH 0/9] Enable arm64 host support in Gitlab
by Tom Rini 18 Jan '25

18 Jan '25

Hey all, First, thanks to Simon Glass and also Linaro, we now have access to a few fast arm64 host machines in our Gitlab instance, to use as CI runners. This series finishes the work that I pushed earlier and Simon had started that enables arm64 hosts to be used for most things now. The first notable change, especially if you use this on your own Gitlab instance is that "DEFAULT_TAG" is now unused and we instead have: - DEFAULT_ALL_TAG: - DEFAULT_ARM64_TAG: - DEFAULT_AMD64_TAG: - DEFAULT_FAST_AMD64_TAG: This lets us say that some jobs can be run on all runners, because they are small enough that anything we'd connect to CI is fast enough and it also does not depend on the underlying host architecture. Next we have tags for any arm64 host, or any amd64 host. Finally, we have a tag for fast amd64 hosts. What these last three are for is that we have a few jobs that need to run on amd64 hosts and so we have to restrict them there, but we also have now reworked the world build jobs to build (almost) everything in a single job and on the fast amd64 machines this is still as quick as the old way was, in practice. To reach this point, we say that the Xtensa jobs can only run on amd64 hosts. Our targets only work with the binary-only toolchain and so this is a reasonable limit and we exclude them from the world build jobs. We also need to deal with ensuring the right toolchain is used regardless what the host architecture is and that we don't use the host toolchain by accident. Finally, because some of these changes needed to be worked out in the linter, fix some of the general warnings that notes as well. -- Tom

2 24

[PATCH] squashfs: Fix sqfs_inode_size() for xattr related SQFS_LSYMLINK_TYPE
by nvbolhuis＠gmail.com 18 Jan '25

18 Jan '25

From: Norbert van Bolhuis <nvbolhuis(a)gmail.com> A squashfs filesystem with extended attributes (xattrs) may have inodes of type SQFS_LSYMLINK_TYPE. This might cause u-boot to fail to handle the filesystem since it assumes a SYMLINK_TYPE and LSYMLINK_TYPE inode are the same size. This is wrong, see: https://github.com/plougher/squashfs-tools/blob/master/squashfs-tools/read_… Using the mksquashfs '-no-xattrs' argument is probably best, but the mksquashfs '-xattrs' argument is the default. This patch fixes squashfs image handling by making sure parsing the uncompressed inode_table (with sqfs_find_inode) succeeeds. The only change needed is correctly determining the size of a SQFS_LSYMLINK_TYPE inode. Signed-off-by: Norbert van Bolhuis <nvbolhuis(a)gmail.com> --- fs/squashfs/sqfs_inode.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/squashfs/sqfs_inode.c b/fs/squashfs/sqfs_inode.c index bb3ccd37e3..ce9a8ff8e2 100644 --- a/fs/squashfs/sqfs_inode.c +++ b/fs/squashfs/sqfs_inode.c @@ -19,7 +19,9 @@ int sqfs_inode_size(struct squashfs_base_inode *inode, u32 blk_size) { - switch (get_unaligned_le16(&inode->inode_type)) { + u16 inode_type = get_unaligned_le16(&inode->inode_type); + + switch (inode_type) { case SQFS_DIR_TYPE: return sizeof(struct squashfs_dir_inode); @@ -87,7 +89,7 @@ int sqfs_inode_size(struct squashfs_base_inode *inode, u32 blk_size) get_unaligned_le32(&symlink->symlink_size), &size)) return -EINVAL; - return size; + return (inode_type == SQFS_SYMLINK_TYPE) ? size : size + sizeof(u32); } case SQFS_BLKDEV_TYPE: -- 2.47.1

2 1

[PATCH] qconfig.py: Update resync message
by Tom Rini 18 Jan '25

18 Jan '25

When using qconfig.py to resync defconfig files and commit the result, update it to say: Resync all defconfig files using qconfig.py Signed-off-by: Tom Rini <trini(a)konsulko.com> --- tools/qconfig.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/qconfig.py b/tools/qconfig.py index 058d72cf4bcc..259adbe1bc9e 100755 --- a/tools/qconfig.py +++ b/tools/qconfig.py @@ -1617,7 +1617,7 @@ def add_commit(configs): '\n '.join(configs)) else: msg = 'configs: Resync with savedefconfig' - msg += '\n\nRsync all defconfig files using moveconfig.py' + msg += '\n\nResync all defconfig files using qconfig.py' subprocess.call(['git', 'commit', '-s', '-m', msg]) -- 2.43.0

2 2

[RFC] Adding bootsource and reset-source* to /chosen
by Quentin Schulz 18 Jan '25

18 Jan '25

Hi all, In U-Boot, there are multiple stages which can be loaded from different media each. On Rockchip SoCs, it mostly goes this way, SoC BootROM loads from a medium (storage or over USB) and executes the TPL (or SPL if no TPL supported), the TPL goes back to BootROM and reads the SPL from the same medium and executes it. The SPL will favor the same medium for loading the U-Boot FIT image (containing TF-A BL31, OP-TEE OS BL32 and U-Boot proper image), but in case it fails, falls back to other possible media. In U-Boot we have the storage medium used for loading the U-Boot FIT image stored in the DTB under /chosen/u-boot,spl-boot-device property. But we do not have a property for the medium the BootROM loaded the TPL (or SPL) from. This could be useful information. For example, I have some CI where I want to check whether the BootROM loaded the TPL from the expected location (important when checking that they all work and don't depend on a bootloader flashed in a fallback medium for example). Another usecase could be to detect whenever one loads the TPL from USB (not a USB stick, some USB transfer; see rockusb for Rockchip for example) and decide to do something else in U-Boot or the OS depending on that (could be starting DFU for flashing the device, or fastboot for loading the rest of the system). Barebox currently does that by setting /chosen/bootsource to the full path of the device the BootROM says it booted from, but that isn't part of the DT spec so before I add support in U-Boot for that, I'm asking whether that's fine or if we should correct course in Barebox and use another name and/or location. I've also been made aware of other "sources" that are added to the /chosen node in Barebox, reset-source, reset-source-instance and reset-source-device. I am not entirely sure what those exactly are for BUT it seems that this is of interest to U-Boot as well since we do try to report a reset cause as read from the CPU for Rockchip SoCs, c.f. https://elixir.bootlin.com/u-boot/v2024.10/source/arch/arm/mach-rockchip/cp…. I believe /chosen/reset-source would be the reason, e.g. "POR", "RST", "WDG", "WKE", "THERM", etc. /chosen/reset-source-device contains the node path to the device that triggered that, that could be useful in the event there are more than one thermal trip point (e.g. CPU or GPU), or more than one hardware watchdog (I for example have a system with potentially 4: SoC, PMIC, 2 separate in an external MCU). I guess /chosen/reset-source-instance is for HW blocks that expose the same reset but from two different sources which aren't modeled in Device Tree. Again, this would be useful for the bootloader or OS to decide to do something else based on the reset cause (e.g. have data on why some devices reboot in the field, e.g. because of thermals, power loss, brownout, etc..). I'm reaching out to the devicetree-spec community before I add code in U-Boot to implement this so we don't add unnecessary backward compatibility code, were the properties or locations to be modified to match the spec. Does anyone disagree with adding this to the spec? Does anyone have another suggestion for the naming or location of the properties? Would you prefer a patch to the spec and discuss "code" directly there? Cheers, Quentin

3 3

[PATCH 00/14] Add support for Ethernet interfaces on RZ/G2L
by Paul Barker 18 Jan '25

18 Jan '25

This patch series enables the usage of both Ethernet interfaces on the Renesas RZ/G2L SMARC EVK board. This requires changes to the RZ/G2L clock driver, RZ/G2L pinctrl driver, ksz9131 phy driver and ravb network driver. Once all drivers have the required support, we enable the relevant options in the RZ/G2L defconfig and cherry-pick required devicetree changes from Linux v6.12-rc4. Paul Barker (14): clk: rzg2l: Ignore enable for core clocks pinctrl: rzg2l: Support 2.5V PVDD for Ethernet interfaces pinctrl: rzg2l: Support Ethernet TXC output enable pinctrl: rzg2l: Drop unnecessary scope net: phy: Port set/clear bits from Linux net: phy: ksz90x1: Handle ksz9131 LED errata net: phy: ksz90x1: Load skew values from device tree net: phy: ksz90x1: Simplify ksz9131_config_rgmii_delay net: ravb: Support up to two instances net: ravb: Simplify max-speed handling in ravb_of_to_plat net: ravb: Add RZ/G2L Support renesas_rzg2l_smarc_defconfig: Enable networking support arm64: dts: renesas: rzg2l: Enable Ethernet TXC output arm64: dts: renesas: rzg2l: Set Ethernet PVDD to 1.8V arch/arm/mach-renesas/Kconfig | 1 + configs/renesas_rzg2l_smarc_defconfig | 6 + drivers/clk/renesas/rzg2l-cpg.c | 8 + drivers/net/Kconfig | 2 + drivers/net/phy/micrel_ksz90x1.c | 161 +++++++++++-- drivers/net/ravb.c | 216 ++++++++++++++++-- drivers/pinctrl/renesas/rzg2l-pfc.c | 83 +++++-- .../src/arm64/renesas/rzg2l-smarc-som.dtsi | 86 ++++--- include/phy.h | 22 ++ include/renesas/rzg2l-pfc.h | 4 + 10 files changed, 505 insertions(+), 84 deletions(-) -- 2.43.0

3 41

[PATCH v3 0/4] Rockchip: Allow to silent TPL/SPL debug console
by Lukasz Czechowski 17 Jan '25

17 Jan '25

Currently it is not possible to completely silent the debug console in i.e. PX30, what might be required in some embedded devices. Disabling DEBUG_UART results in errors, due to missing symbols. In particular, rockchip sdram driver performs calls to debug_uart functions (i.e. printascii), defined directly in serial driver. This patch series fixes the dependencies in Kconfig and provides dummy implementation of debug_uart functions that are linked in case serial driver is not used. changes in v2: - Update commit titles - Create separate SPL and TPL symbols for RAM_ROCKCHIP_DEBUG changes in v3: - Dropped [PATCH v2 3/5] ram: rockchip: Add separate RAM_ROCKCHIP_DEBUG config for TPL/SPL As suggested by Kever, dependency on DEBUG_UART is sufficient and this patch was not needed. Lukasz Czechowski (4): debug_uart: Replace debug functions with dummies if CONFIG_DEBUG_UART is not set ram: rockchip: Fix dependency of RAM_ROCKCHIP_DEBUG rockchip: px30: Weaken dependency TPL/SPL serial rockchip: px30: Fix hard dependency to DEBUG_UART_BOARD_INIT arch/arm/mach-rockchip/Kconfig | 5 ++--- drivers/ram/rockchip/Kconfig | 1 + include/debug_uart.h | 16 ++++++++++++++++ 3 files changed, 19 insertions(+), 3 deletions(-) -- 2.43.0

5 16