September 2023 - U-Boot

[PATCH v2] cmd: sf: Fix the flash_is_unlocked api size parameter
by Venkatesh Yadav Abbarapu 21 Sep '23

21 Sep '23

When flash erase is called with size parameter, code is checking if sectors are locked or not. But for checking, the whole device length minus offset is used instead of actual size which should be erased. That's why when only some sectors are locked it is not possible to erase unlocked sectors. The length is calculated as "length = max_chipsize - offset", flash_is_unlocked() api is getting updated with length which is incorrect. Fix this flash_is_unlocked() api by passing the size parameter. ZynqMP> sf erase 0 100000 len=0x8000000 which is flash size size=0x100000 We need to update the size in the flash_is_unlocked() api and not the length. Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu(a)amd.com> --- Changes in v2: - Updated the commit description. --- cmd/sf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/cmd/sf.c b/cmd/sf.c index 55bef2f769..977ffcb63c 100644 --- a/cmd/sf.c +++ b/cmd/sf.c @@ -366,7 +366,7 @@ static int do_spi_flash_erase(int argc, char *const argv[]) } if (flash->flash_is_unlocked && - !flash->flash_is_unlocked(flash, offset, len)) { + !flash->flash_is_unlocked(flash, offset, size)) { printf("ERROR: flash area is locked\n"); return CMD_RET_FAILURE; } -- 2.17.1

3 2

[PATCH v6 0/7] J721S2 DTS Sync from v6.6-rc1 to u-boot
by Manorit Chawdhry 21 Sep '23

21 Sep '23

The sync tries to ensure that U-boot remains functional with the updated Linux DTS and all the fixes from Linux move to U-boot during the sync. The series tries to sync from Linux v6.6-rc1 along with addition of the documentation for J721S2 that had been previously missing. DMA fixes [0] are currently being upstreamed to Linux. Test Logs are included in [1] [0]: https://lore.kernel.org/all/20230810174356.3322583-1-vigneshr@ti.com/ [1]: https://gist.github.com/manorit2001/dbad09fd00e8b7c3872e85874c8e648c Signed-off-by: Manorit Chawdhry <m-chawdhry(a)ti.com> --- Changes in v6: - Update the subject prefixes ( Nishanth ) - Link to v5: https://lore.kernel.org/r/20230919-b4-upstream-j721s2-r5-pinmux-v5-0-a54d6e… --- Manorit Chawdhry (6): Revert "arm: dts: k3-j7*: ddr: Update to 0.10 version of DDR config tool" arm: mach-k3: j721s2: Add mcu_timer0 id to the dev list arm: dts: k3-j721s2: Sync from Linux tag v6.6-rc1 arm: dts: k3-am68: Sync from Linux tag v6.6-rc1 docs: board: ti: Add j721s2_evm documentation board: ti: j721s2: MAINTAINERS: Update the MAINTAINERS File. Udit Kumar (1): configs: j721s2_evm_r5_defconfig: Increase malloc pool size in DRAM arch/arm/dts/k3-am68-sk-base-board-u-boot.dtsi | 75 +- arch/arm/dts/k3-am68-sk-base-board.dts | 524 +++++++++----- arch/arm/dts/k3-am68-sk-r5-base-board.dts | 151 +--- arch/arm/dts/k3-am68-sk-som.dtsi | 112 +-- arch/arm/dts/k3-j721e-ddr-evm-lp4-4266.dtsi | 98 +-- .../dts/k3-j721s2-common-proc-board-u-boot.dtsi | 66 +- arch/arm/dts/k3-j721s2-common-proc-board.dts | 376 ++++++---- arch/arm/dts/k3-j721s2-ddr-evm-lp4-4266.dtsi | 464 ++++++------ arch/arm/dts/k3-j721s2-main.dtsi | 777 ++++++++++++++++++++- arch/arm/dts/k3-j721s2-mcu-wakeup.dtsi | 374 +++++++++- arch/arm/dts/k3-j721s2-r5-common-proc-board.dts | 158 +---- arch/arm/dts/k3-j721s2-som-p0.dtsi | 172 ++--- arch/arm/dts/k3-j721s2-thermal.dtsi | 101 +++ arch/arm/dts/k3-j721s2.dtsi | 12 +- arch/arm/dts/k3-serdes.h | 204 ++++++ arch/arm/mach-k3/j721s2/dev-data.c | 1 + board/ti/j721s2/MAINTAINERS | 11 +- configs/j721s2_evm_r5_defconfig | 1 + doc/board/ti/j721s2_evm.rst | 341 +++++++++ doc/board/ti/k3.rst | 1 + 20 files changed, 2859 insertions(+), 1160 deletions(-) --- base-commit: 2fe4b54556ea6271237b35de68dc458bfceab94c change-id: 20230816-b4-upstream-j721s2-r5-pinmux-25c4cd61b258 Best regards, -- Manorit Chawdhry <m-chawdhry(a)ti.com>

1 7

[PATCH v7 0/2] Add support to fetch baudrate from dtb
by Venkatesh Yadav Abbarapu 21 Sep '23

21 Sep '23

In this patch series - Add support in Kconfig and convert for armada boards - Fetch baudrate from the dtb and update Changes in v7: - Updated the commit message. Changes in v6: - Changed Kconfig OF_SERIAL_DT_BAUD to OF_SERIAL_BAUD. - Added IS_ENABLED() in place of #ifdef CONFIG_OF_SERIAL_AUD Changes in v5: - Adding DEFAULT_ENV_IS_RW Kconfig in missing files - Updating DEFAULT_ENV_IS_RW to CONFIG_DEFAULT_ENV_IS_RW Changes in v4: - Moved SERIAL_DT_BAUD to another patch - Added doc file for fetching serial baudrate from DT. - Changed Kconfig SERIAL_DT_BAUD to OF_SERIAL_DT_BAUD - Added function docs wherever required. - Moved changes from fdtdec api to ofnode - Changed serial_get_valid_baudrate to check_valid_baudrate - Added function fetch_baud_from_dtb to fetch baud from DT - Used dectoul() for parsing baudrate Changes in v3: - Add SERIAL_DT_BAUD to Kconfig - Moved DEFAULT_ENV_IS_RW to Kconfig also updated armada files - Moved filler changes from zynqmp.h to generic file env_default.h - Removed ENV_RW_FILLER and added padding in the generic file env_default.h. - Print baudrate parameter properly when SERIAL_DT is enabled. Changes in v2: - Changed to #ifdef from #if CONFIG_IS_ENABLED to enable patching in spl. - Added SPL_ENV_SUPPORT dependency in SERIAL_DT_BAUD to allow SPL compilation. - Moved DEFAULT_ENV_IS_RW to Kconfig also updated armada files - Moved ENV_RW_FILLER to generic file env_default.h. - Increased the ENV_RW_FILLER padding to support 8000000 baud. Algapally Santosh Sagar (2): configs: Add support in Kconfig and convert for armada boards serial: zynqmp: Fetch baudrate from dtb and update configs/eDPU_defconfig | 1 + configs/mvebu_db-88f3720_defconfig | 1 + configs/mvebu_espressobin-88f3720_defconfig | 1 + configs/uDPU_defconfig | 1 + doc/README.serial_dt_baud | 41 ++++++++++++++++++++ drivers/core/ofnode.c | 18 +++++++++ drivers/serial/Kconfig | 15 +++++++ drivers/serial/serial-uclass.c | 43 +++++++++++++++++++++ include/configs/mvebu_armada-37xx.h | 1 - include/dm/ofnode.h | 14 ++++++- include/env_default.h | 8 +++- include/env_internal.h | 2 +- include/serial.h | 15 +++++++ 13 files changed, 155 insertions(+), 6 deletions(-) create mode 100644 doc/README.serial_dt_baud -- 2.17.1

2 3

[PATCH] env: ti: ti_common.env: Fix get_overlaystring for FIT Image
by Manorit Chawdhry 21 Sep '23

21 Sep '23

After the refactor with conf- nodes in fitImage, overlaystring wasn't didn't handle the new conf- nodes in FIT Booting. Fix get_overlaystring to handle conf- nodes. Fixes: 837833a724b7 ("environment: ti: Add get_fit_config command to get FIT config string") Reported-by: Aniket Limaye <a-limaye(a)ti.com> Signed-off-by: Manorit Chawdhry <m-chawdhry(a)ti.com> --- Test Logs: => setenv name_overlays ti/k3-dt.dtbo ti/k3-dt1.dtbo => run get_overlaystring => printenv overlaystring overlaystring=#conf-ti_k3-dt.dtbo#conf-ti_k3-dt1.dtbo --- include/env/ti/ti_common.env | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/include/env/ti/ti_common.env b/include/env/ti/ti_common.env index e87a41a6590b..4d0590912614 100644 --- a/include/env/ti/ti_common.env +++ b/include/env/ti/ti_common.env @@ -16,9 +16,14 @@ addr_fit=0x90000000 name_fit=fitImage update_to_fit=setenv loadaddr ${addr_fit}; setenv bootfile ${name_fit} get_overlaystring= - for overlay in $name_overlays; - do; - setenv overlaystring ${overlaystring}'#'${overlay}; + for overlay in $name_overlays; do; + if test ${boot_fit} -eq 1; + then + setexpr name_fit_overlay gsub / _ conf-${overlay}; + setenv overlaystring ${overlaystring}'#'${name_fit_overlay}; + else + setenv overlaystring ${overlaystring}'#'${overlay}; + fi; done; get_fit_config=setexpr name_fit_config gsub / _ conf-${fdtfile} run_fit=run get_fit_config; bootm ${addr_fit}#${name_fit_config}${overlaystring} --- base-commit: 2fe4b54556ea6271237b35de68dc458bfceab94c change-id: 20230915-b4-upstream-overlaystring-207e28b8c5fb Best regards, -- Manorit Chawdhry <m-chawdhry(a)ti.com>

1 1

[RFC PATCH 0/4] mtd: ubi: Enable accessing RO filesystems in UBI vols
by Sam Edwards 21 Sep '23

21 Sep '23

Hi UBI maintainers, My target's rootfs is a read-only filesystem stored in a static UBI volume, mounted via a "ubiblock" device after boot. I'd also like to keep the boot files in the same filesystem, so that it's all coupled together. To that end, I'm working on a patchset so that U-Boot can read static UBI volumes as read-only block devices (paralleling Linux's ubiblock mechanism). I'm very happy with how the first 3 patches in this series turned out (so I'm not asking about them per se, though feedback is certainly welcome). The fourth is where I got stuck: while the code definitely works, it requires bringing DM headers into disk/part.c, which certainly will not fly in mainline. I need to plumb this through drivers/block/blk-uclass.c to do it "properly." Part of the problem here is that these are *volumes,* but they are (optionally) *named.* In U-Boot's current view of block devices, it is only a partition, and not a volume, that may have a name. These aren't "partitions" in U-Boot's view, since a partition is a contiguous slice of a bigger block device, while these are (logically) separate block devices. So, I would need to invent a new function that can look up a named (sub)volume. I also probably need to invent new syntax, so that I'm not overloading the 0:1 syntax for partitions. I'm not especially beholden to the ':', but I do want to use the same syntax for volume numbers and volume names, to mirror Linux's `ubi.block=x,y` syntax. I'm also trying to reclaim the name "ubi" to refer to a UBI volume, while U-Boot currently thinks it should refer to the presently-mounted UBIFS. In my current series, the meaning of "ubi" depends on whether ubifs is mounted, for backwards-compatibility. If this isn't palpable, I could consider other options like "ubivol"/"ubiblock"/"ubiblk"/"ubistatic"/... So, the feedback I'm hoping for would be: 1) What is a good syntax for referring to a logical volume by name or ID? Keeping in mind this may affect more than just UBI, if e.g. U-Boot learns to peer inside LVM2s in the future. 2) What should I call the block functions for looking up a block device's subvolume by type+parentidx+{name,ID}? 3) Is my strategy of reclaiming "ubi" sound, or should I be conceding that to UBIFS and using a new type name for static UBI volumes? 4) Does my choose_blksz_for_volume() function make sense, or should I always be using a preferred block size (like 512) if possible? Cheers, Sam Sam Edwards (4): mtd: ubi: register UBI attachments as DM devices mtd: ubi: bind block device driver for static volumes disk: part: fall-through if "ubi" requested but ubifs not mounted HACK: enable access to `ubi 0:volname` block devices cmd/ubi.c | 11 +++ disk/part.c | 70 +++++++++++-- drivers/mtd/ubi/Makefile | 1 + drivers/mtd/ubi/ubi-uclass.c | 184 +++++++++++++++++++++++++++++++++++ include/dm/uclass-id.h | 1 + include/ubi_uboot.h | 5 + 6 files changed, 264 insertions(+), 8 deletions(-) create mode 100644 drivers/mtd/ubi/ubi-uclass.c -- 2.41.0

2 7

[PATCH v3] clk: ti: Notify AVS driver upon setting clock rate
by Udit Kumar 21 Sep '23

21 Sep '23

AVS is enabled at R5 SPL stage, on few platforms like J721E and J7200 clk-k3 is used instead if clk-sci driver. Add support in clk-k3 driver as well to notify AVS driver on setting clock rate so that voltage is changed accordingly. Also add support in clk-sci driver to do AVS notification before or after clock setting based upon current rate and new clock rate. Cc: Keerthy <j-keerthy(a)ti.com> Signed-off-by: Udit Kumar <u-kumar1(a)ti.com> --- Boot logs with v3: https://gist.github.com/uditkumarti/12a753a822c0aec9ca3028648c29dfa8 Change log: Change in v3: - Updated AVS notification before/after clock setting based upon new and current rate - Added modifed notification in clk-sci driver as well - v2 link https://lore.kernel.org/all/20230919140408.2608521-1-u-kumar1@ti.com/ Changes in v2: - Kept clk-sci.c as is because this is used by AM64 and AM65 platforms - v1 link https://lore.kernel.org/all/20230919060649.2518147-1-u-kumar1@ti.com drivers/clk/ti/clk-k3.c | 11 +++++++++++ drivers/clk/ti/clk-sci.c | 11 ++++++++--- 2 files changed, 19 insertions(+), 3 deletions(-) diff --git a/drivers/clk/ti/clk-k3.c b/drivers/clk/ti/clk-k3.c index ba925fa3c4..e8d8f8adf3 100644 --- a/drivers/clk/ti/clk-k3.c +++ b/drivers/clk/ti/clk-k3.c @@ -11,6 +11,7 @@ #include <errno.h> #include <soc.h> #include <clk-uclass.h> +#include <k3-avs.h> #include "k3-clk.h" #define PLL_MIN_FREQ 800000000 @@ -242,7 +243,13 @@ static ulong ti_clk_set_rate(struct clk *clk, ulong rate) const struct clk_ops *ops; ulong new_rate, rem; ulong diff, new_diff; + int set_avs_after_clock = 1; + if (rate >= ti_clk_get_rate(clk) && IS_ENABLED(CONFIG_K3_AVS0)) { + k3_avs_notify_freq(data->map[clk->id].dev_id, + data->map[clk->id].clk_id, rate); + set_avs_after_clock = 0; + } /* * We must propagate rate change to parent if current clock type * does not allow setting it. @@ -339,6 +346,10 @@ static ulong ti_clk_set_rate(struct clk *clk, ulong rate) } } + if (set_avs_after_clock && IS_ENABLED(CONFIG_K3_AVS0)) + k3_avs_notify_freq(data->map[clk->id].dev_id, + data->map[clk->id].clk_id, rate); + return new_rate; } diff --git a/drivers/clk/ti/clk-sci.c b/drivers/clk/ti/clk-sci.c index 74df5a397b..1926843dcc 100644 --- a/drivers/clk/ti/clk-sci.c +++ b/drivers/clk/ti/clk-sci.c @@ -91,12 +91,14 @@ static ulong ti_sci_clk_set_rate(struct clk *clk, ulong rate) const struct ti_sci_handle *sci = data->sci; const struct ti_sci_clk_ops *cops = &sci->ops.clk_ops; int ret; + u8 set_avs_after_clock = 1; debug("%s(clk=%p, rate=%lu)\n", __func__, clk, rate); -#ifdef CONFIG_K3_AVS0 - k3_avs_notify_freq(clk->id, clk->data, rate); -#endif + if (rate >= ti_sci_clk_get_rate(clk) && IS_ENABLED(CONFIG_K3_AVS0)) { + k3_avs_notify_freq(clk->id, clk->data, rate); + set_avs_after_clock = 0; + } ret = cops->set_freq(sci, clk->id, clk->data, 0, rate, ULONG_MAX); if (ret) { @@ -104,6 +106,9 @@ static ulong ti_sci_clk_set_rate(struct clk *clk, ulong rate) return ret; } + if (set_avs_after_clock && IS_ENABLED(CONFIG_K3_AVS0)) + k3_avs_notify_freq(clk->id, clk->data, rate); + return rate; } -- 2.34.1

3 2

[PATCH v4 1/2] x86: coreboot: Document cbmem console struct
by Simon Glass 21 Sep '23

21 Sep '23

Coreboot changed a few years ago to include an overflow flag. Update the structure to match this. This comes from coreboot commit: 6f5ead14b4 ("mb/google/nissa/var/joxer: Update eMMC DLL settings") Note: There are several implementations of this in coreboot. I have chosen to follow the one in src/lib/cbmem_console.c Signed-off-by: Simon Glass <sjg(a)chromium.org> --- Changes in v4: - Reword commit and change title Changes in v3: - Drop __packed as it does nothing useful arch/x86/include/asm/coreboot_tables.h | 17 +++++++++++++++-- 1 file changed, 15 insertions(+), 2 deletions(-) diff --git a/arch/x86/include/asm/coreboot_tables.h b/arch/x86/include/asm/coreboot_tables.h index 4de137fbab9d..0dfb64babb96 100644 --- a/arch/x86/include/asm/coreboot_tables.h +++ b/arch/x86/include/asm/coreboot_tables.h @@ -299,11 +299,24 @@ struct cb_vdat { #define CB_TAG_TIMESTAMPS 0x0016 #define CB_TAG_CBMEM_CONSOLE 0x0017 +#define CBMC_CURSOR_MASK ((1 << 28) - 1) +#define CBMC_OVERFLOW BIT(31) + +/* + * struct cbmem_console - In-memory console buffer for coreboot + * + * Structure describing console buffer. It is overlaid on a flat memory area, + * with body covering the extent of the memory. Once the buffer is full, + * output will wrap back around to the start of the buffer. The high bit of the + * cursor field gets set to indicate that this happened. If the underlying + * storage allows this, the buffer will persist across multiple boots and append + * to the previous log. + */ struct cbmem_console { u32 size; u32 cursor; - char body[0]; -} __packed; + u8 body[0]; +}; #define CB_TAG_MRC_CACHE 0x0018 -- 2.42.0.283.g2d96d420d3-goog

2 5

[PATCH v2 0/8] Add SM uclass and Meson SM driver
by Alexey Romanov 21 Sep '23

21 Sep '23

Hello! At the moment, there is no single general approach to using secure monitor in U-Boot, there is only the smc_call() function, over which everyone builds their own add-ons. This patchset is designed to solve this problem by adding a new uclass - SM_UCLASS. This UCLASS export following generic API: 1. sm_call() - generic SMC call to the secure-monitor 2. sm_call_read() - retrieve data from secure-monitor 3. sm_call_write() - send data to secure-monitor In the future, it is necessary to completely get rid of raw smc_call() calls, replacing them with the use of SM_UCLASS based drivers. V2: - Add SM UCLASS - Add SM sandbox driver - Add test for sandbox driver - Meson Secure Monitor driver now based on SM_UCLASS - Fix include order in arch/arm/mach-meson/sm.c Also, during the discussion in V1 of this patchset, it was discussed to create MESON_SM_UCLASS, but I considered such a uclass to be too arch-specific. That's why I suggest SM_UCLASS, which is not so arch-specific: secure monitor can used for whole ARM devices, not only for Amlogic SoC's. Alexey Romanov (8): drivers: introduce Secure Monitor uclass sandbox: add sandobx sm uclass driver sandbox: dts: add meson secure monitor node sandbox: add tests for UCLASS_SM sandbox: defconfig: enable CONFIG_SM option drivers: introduce Meson Secure Monitor driver arch: meson: sm: set correct order of the includes arch: meson: use secure monitor driver MAINTAINERS | 1 + arch/arm/mach-meson/Kconfig | 1 + arch/arm/mach-meson/sm.c | 116 +++++++++++---------- arch/sandbox/dts/test.dts | 4 + configs/sandbox_defconfig | 1 + drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/sm/Kconfig | 9 ++ drivers/sm/Makefile | 5 + drivers/sm/meson-sm.c | 198 ++++++++++++++++++++++++++++++++++++ drivers/sm/sandbox-sm.c | 76 ++++++++++++++ drivers/sm/sm-uclass.c | 55 ++++++++++ include/dm/uclass-id.h | 1 + include/meson/sm.h | 19 ++++ include/sandbox-sm.h | 18 ++++ include/sm-uclass.h | 72 +++++++++++++ include/sm.h | 67 ++++++++++++ test/dm/Makefile | 1 + test/dm/sm.c | 65 ++++++++++++ 19 files changed, 656 insertions(+), 56 deletions(-) create mode 100644 drivers/sm/Kconfig create mode 100644 drivers/sm/Makefile create mode 100644 drivers/sm/meson-sm.c create mode 100644 drivers/sm/sandbox-sm.c create mode 100644 drivers/sm/sm-uclass.c create mode 100644 include/meson/sm.h create mode 100644 include/sandbox-sm.h create mode 100644 include/sm-uclass.h create mode 100644 include/sm.h create mode 100644 test/dm/sm.c -- 2.25.1

2 16

[PATCH 4/4] spl: atf: Reduce SPL code size
by Massimo Pegorer 21 Sep '23

21 Sep '23

Previous fix commit increases code size of a few bytes. This minor rework finally reduces SPL size of about 64-72 bytes (tested with buildman on several aarch64 boards). Signed-off-by: Massimo Pegorer <massimo.pegorer+oss(a)gmail.com> --- common/spl/spl_atf.c | 42 +++++++++++++++++++----------------------- 1 file changed, 19 insertions(+), 23 deletions(-) diff --git a/common/spl/spl_atf.c b/common/spl/spl_atf.c index 0d8db2d14e..fa79f73be2 100644 --- a/common/spl/spl_atf.c +++ b/common/spl/spl_atf.c @@ -209,7 +209,19 @@ static void spl_enter_atf(uintptr_t bl31_entry, uintptr_t bl32_entry, atf_entry(bl31_params, (void *)fdt_addr); } -static int spl_fit_images_find(void *blob, int os) +static void spl_fit_images_get_entry(void *blob, int node, uintptr_t *entry_p) +{ + ulong val; + + if (fit_image_get_entry(blob, node, &val)) + if (fit_image_get_load(blob, node, &val)) + return; + + debug("%s: entry point 0x%lx\n", __func__, val); + *entry_p = val; +} + +static int spl_fit_images_get_os_entry(void *blob, int os, uintptr_t *entry_p) { int parent, node, ndepth = 0; const void *data; @@ -231,41 +243,28 @@ static int spl_fit_images_find(void *blob, int os) if (!data) continue; - if (genimg_get_os_id(data) == os) - return node; + if (genimg_get_os_id(data) == os) { + spl_fit_images_get_entry(blob, node, entry_p); + return 0; + } }; return -FDT_ERR_NOTFOUND; } -void spl_fit_images_get_entry(void *blob, int node, uintptr_t *entry_p) -{ - ulong val; - - if (fit_image_get_entry(blob, node, &val)) - if (fit_image_get_load(blob, node, &val)) - return; - - debug("%s: entry point 0x%lx\n", __func__, val); - *entry_p = val; -} - void spl_invoke_atf(struct spl_image_info *spl_image) { uintptr_t bl32_entry = 0; uintptr_t bl33_entry = CONFIG_TEXT_BASE; void *blob = spl_image->fdt_addr; uintptr_t platform_param = (uintptr_t)blob; - int node; /* * Find (in /fit-images) the TEE binary entry point address * (or load address if entry point is missing) and pass it as * the BL3-2 entry point. This is optional. */ - node = spl_fit_images_find(blob, IH_OS_TEE); - if (node >= 0) - spl_fit_images_get_entry(blob, node, &bl32_entry); + spl_fit_images_get_os_entry(blob, IH_OS_TEE, &bl32_entry); /* * Find (in /fit-images) the U-Boot binary entry point address @@ -273,10 +272,7 @@ void spl_invoke_atf(struct spl_image_info *spl_image) * the BL3-3 entry point. * This will need to be extended to support Falcon mode. */ - - node = spl_fit_images_find(blob, IH_OS_U_BOOT); - if (node >= 0) - spl_fit_images_get_entry(blob, node, &bl33_entry); + spl_fit_images_get_os_entry(blob, IH_OS_U_BOOT, &bl33_entry); /* * If ATF_NO_PLATFORM_PARAM is set, we override the platform -- 2.34.1

2 1

[PATCH 3/4] spl: atf: Fix potential but not occurring bug
by Massimo Pegorer 21 Sep '23

21 Sep '23

Bug: function spl_fit_images_get_entry returns uninitialized variable val if both fit_image_get_entry() and fit_image_get_load() fail (note that both of them do not set val on failure). Fix: use val only if entry or load address has been retrieved successfully. Note: in real world a failure of fit_image_get_load() will make current SPL flow to break on loading images, see spl_load_fit_image(), before entering spl_invoke_atf(). Signed-off-by: Massimo Pegorer <massimo.pegorer+oss(a)gmail.com> --- common/spl/spl_atf.c | 15 +++++++-------- 1 file changed, 7 insertions(+), 8 deletions(-) diff --git a/common/spl/spl_atf.c b/common/spl/spl_atf.c index 3d022ed4e7..0d8db2d14e 100644 --- a/common/spl/spl_atf.c +++ b/common/spl/spl_atf.c @@ -238,17 +238,16 @@ static int spl_fit_images_find(void *blob, int os) return -FDT_ERR_NOTFOUND; } -uintptr_t spl_fit_images_get_entry(void *blob, int node) +void spl_fit_images_get_entry(void *blob, int node, uintptr_t *entry_p) { ulong val; - int ret; - ret = fit_image_get_entry(blob, node, &val); - if (ret) - ret = fit_image_get_load(blob, node, &val); + if (fit_image_get_entry(blob, node, &val)) + if (fit_image_get_load(blob, node, &val)) + return; debug("%s: entry point 0x%lx\n", __func__, val); - return val; + *entry_p = val; } void spl_invoke_atf(struct spl_image_info *spl_image) @@ -266,7 +265,7 @@ void spl_invoke_atf(struct spl_image_info *spl_image) */ node = spl_fit_images_find(blob, IH_OS_TEE); if (node >= 0) - bl32_entry = spl_fit_images_get_entry(blob, node); + spl_fit_images_get_entry(blob, node, &bl32_entry); /* * Find (in /fit-images) the U-Boot binary entry point address @@ -277,7 +276,7 @@ void spl_invoke_atf(struct spl_image_info *spl_image) node = spl_fit_images_find(blob, IH_OS_U_BOOT); if (node >= 0) - bl33_entry = spl_fit_images_get_entry(blob, node); + spl_fit_images_get_entry(blob, node, &bl33_entry); /* * If ATF_NO_PLATFORM_PARAM is set, we override the platform -- 2.34.1

2 1