August 2025 - Concept

[PATCH 0/8] efi: A few minor improvements
by Simon Glass 28 Aug '25

28 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series mostly tidies up the efidebug command, but includes a few other pieces as well. Simon Glass (8): efi_loader: Drop unnecessary efi_dp_str() efidebug: doc: Indent the media output correctly doc: efidebug: Support sorting with efidebug memmap doc: Add documentation for about efidebug boot boot: Move obtaining the label into a common file bootstd: Use the new label finder in bootflow_menu_add() efi: Use the console mux by default with the EFI app efi: Set the log category throughout lib/efi_client/ boot/bootflow.c | 27 +++++ boot/bootflow_menu.c | 7 +- cmd/bootflow.c | 16 +-- cmd/efidebug.c | 57 ++++++++++- doc/usage/cmd/efidebug.rst | 195 +++++++++++++++++++++++++++++++++--- include/bootflow.h | 8 ++ include/efi_loader.h | 2 - lib/efi_client/Kconfig | 1 + lib/efi_client/app_run.c | 2 + lib/efi_client/efi.c | 2 + lib/efi_client/efi_info.c | 2 + lib/efi_client/efi_vars.c | 2 + lib/efi_client/sdram.c | 2 + lib/efi_client/stub.c | 2 + lib/efi_client/stub_arm64.c | 1 + lib/efi_client/stub_x86.c | 2 + lib/efi_client/sync_dt.c | 2 + 17 files changed, 286 insertions(+), 44 deletions(-) -- 2.43.0 base-commit: 1ec67f521cb32e0e423873256fd537aed2a5314a branch: apc

1 8

[PATCH 00/20] efi: App and devicetree improvements
by Simon Glass 28 Aug '25

28 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series improves U-Boot's EFI app support and adds new functionality for managing devicetree reserved-memory regions. The main focus is on enhancing memory management when running as an EFI application. A new mechanism is introduced to sync EFI reserved-memory regions (such as runtime services and ACPI tables) with the devicetree's /reserved-memory nodes. This ensures that Linux doesn't inadvertently use memory regions that EFI firmware has reserved. The main improvements are: - New 'fdt reserved' command to display devicetree reserved-memory regions - New 'efi memsync' command to synchronize EFI memory map with devicetree - Enhance EFI application support with better memory region handling - Improve EFI debugging commands and device path handling - Better integrate EFI boot services and devicetree setup There are a few other minor thigns: - Enhance efidebug commands with media device and driver information - Expose EFI device-paths in a few places in bootstd - Improve bootflow device reporting - ARM-specific fixes for exception level handling This series addresses many of the issues found in booting Linux on Qualcomm X-Elite laptops using the EFI app. Simon Glass (20): efi: Set the efi_media device name when binding efi: app: Avoid switching exception levels on ARM efi: app: Avoid freeing memory on exit efi: Add a way to show media devices efi: Add all messaging types for device paths efi: Convert a device-path to a uclass and name efi: Add a uclass column to efidebug media command efi: Implement efidebug show drivers in the app boot: Correct filename in bootmeth_common_read_file() boot: Show the device path for EFI bootflows boot: Show the underlying bootflow device in the app efi: Use abuf when reading EFI variables efi: Add a command to help with shim efi: Add an efi subcommand to show the loaded image fdt: Add a function to show the reserved-memory nodes fdt: Add a command to show reserved-memory regions efi: Provide a way to sync EFI reserved-memory to fdt fdt: Provide a command to sync the reserve-memory node efi: app: Add a simple-framebuffer node if enabled efi: app: Sync EFI reserved-memory to the devicetree boot/bootmeth-uclass.c | 2 +- boot/fdt_support.c | 54 ++++++++ cmd/Kconfig | 9 ++ cmd/Makefile | 1 + cmd/bootflow.c | 38 +++++- cmd/efi.c | 37 ++++- cmd/efidebug.c | 73 +++++++--- cmd/fdt.c | 7 +- cmd/nvedit_efi.c | 62 ++++++--- cmd/shim.c | 65 +++++++++ doc/usage/cmd/efi.rst | 50 +++++++ doc/usage/cmd/efidebug.rst | 24 +++- doc/usage/cmd/fdt.rst | 26 ++++ doc/usage/cmd/shim.rst | 101 ++++++++++++++ doc/usage/index.rst | 1 + include/efi.h | 51 +++++++ include/efi_api.h | 20 ++- include/efi_device_path.h | 25 ++++ include/fdt_support.h | 7 + include/shim.h | 13 ++ lib/efi/device_path.c | 125 +++++++++++++++++ lib/efi_client/app_run.c | 21 ++- lib/efi_client/efi.c | 249 +++++++++++++++++++++++++++++++++- lib/efi_client/efi_app.c | 12 +- lib/efi_client/efi_app_init.c | 21 ++- test/cmd/fdt.c | 40 ++++++ 26 files changed, 1072 insertions(+), 62 deletions(-) create mode 100644 cmd/shim.c create mode 100644 doc/usage/cmd/shim.rst create mode 100644 include/shim.h -- 2.43.0 base-commit: 88123cd086d8999443f84e8311913b7e94b6aa9b branch: apb

1 20

[PATCH 0/6] Provide basic support for a mouse
by Simon Glass 25 Aug '25

25 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series introduces a mouse uclass and provides drivers for two types of mouse: sandbox (using SDL) and USB (potentially usable on any hardware). Note that this was written four years ago, so may need some tweaking. I have tidied it up so that it builds and that (after running sandbox with the -Dl flags), 'mouse dump' shows mouse events. Simon Glass (6): Add a few floating-point routines input: Add mouse support sandbox: sdl: Add support for mouse input usb: Add a USB mouse driver input: Add a command to show mouse input sandbox: Enable the mouse arch/sandbox/cpu/sdl.c | 48 ++++- arch/sandbox/dts/sandbox.dtsi | 4 + arch/sandbox/include/asm/sdl.h | 4 + cmd/Kconfig | 9 + cmd/Makefile | 1 + cmd/mouse.c | 70 +++++++ drivers/input/Kconfig | 17 ++ drivers/input/Makefile | 4 + drivers/input/mouse-uclass.c | 28 +++ drivers/input/sandbox_mouse.c | 35 ++++ drivers/input/usb_mouse.c | 325 +++++++++++++++++++++++++++++++++ include/dm/uclass-id.h | 1 + include/mouse.h | 78 ++++++++ include/usb.h | 1 + include/vsprintf.h | 25 +++ lib/Kconfig | 13 ++ lib/Makefile | 1 + lib/strtof.c | 200 ++++++++++++++++++++ 18 files changed, 859 insertions(+), 5 deletions(-) create mode 100644 cmd/mouse.c create mode 100644 drivers/input/mouse-uclass.c create mode 100644 drivers/input/sandbox_mouse.c create mode 100644 drivers/input/usb_mouse.c create mode 100644 include/mouse.h create mode 100644 lib/strtof.c -- 2.43.0 base-commit: 03cd5e055b153666b836bbc0c3d3b3c7ade52bf0 branch: gu

1 6

[PATCH v2 00/16] console: Refactor in preparation for the pager
by Simon Glass 25 Aug '25

25 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series tidies up a few small things the serial and console areas: - Move detection of serial-console size to the serial module - Provide a Kconfig to disable serial detection - Fix some missing driver-model flags in stdio devices - Move console docs into the main documentation This series is marked v2 since some of the patches were sent in any earlier series. With various tweaks that series grew too large to be sent as a single series. Changes in v2: - Add new patch to provide a way to tell if the video is visible - Add new patch to move environment handling lower in the file - Add new patch to refactor handling of the result in on_console() Simon Glass (16): bdinfo: Show the serial device video: Mark the vidconsole as a driver model stdio device video: Mark keyboards as driver model stdio devices serial: Generalise the code to check the terminal size serial: Cache the terminal size sandbox: Add a function to detect terminal connection sandbox: Provide a way to tell if the video is visible post: Include config.h in the header cli: Avoid sending ANSI characters to a non-terminal console: Update coninfo to use sdev instead of dev console: Update conio command to show uclass console: Move environment handling lower in the file console: Refactor handling of the result in on_console() doc: Remove obsolete text in README.console doc: Move console docs to rST doc: Tidy up the console docs a little arch/sandbox/cpu/os.c | 5 + arch/sandbox/cpu/start.c | 4 + arch/sandbox/cpu/state.c | 14 +++ arch/sandbox/include/asm/state.h | 15 +++ cmd/Kconfig | 8 ++ cmd/bdinfo.c | 3 +- cmd/console.c | 45 ++++++--- common/cli.c | 3 +- common/console.c | 130 ++++++++++++------------- doc/README.console | 100 ------------------- doc/usage/console.rst | 62 ++++++++++++ doc/usage/index.rst | 1 + drivers/input/keyboard-uclass.c | 2 +- drivers/serial/Kconfig | 15 +++ drivers/serial/serial-uclass.c | 156 ++++++++++++++++++++++++++++++ drivers/video/vidconsole-uclass.c | 2 +- include/os.h | 8 ++ include/post.h | 2 + include/serial.h | 63 ++++++++++++ include/video.h | 21 ++++ lib/efi_loader/efi_console.c | 124 +----------------------- test/cmd/bdinfo.c | 4 +- 22 files changed, 484 insertions(+), 303 deletions(-) delete mode 100644 doc/README.console create mode 100644 doc/usage/console.rst -- 2.43.0 base-commit: 03cd5e055b153666b836bbc0c3d3b3c7ade52bf0 branch: clia2

1 16

[PATCH 00/16] Introduce a pager for the console
by Simon Glass 22 Aug '25

22 Aug '25

From: Simon Glass <sjg(a)chromium.org> Some U-Boot commands can produce a lot of output and this can run off top of the display. This series introduces a simple pager feature, to allow the output to be read, before pressing SPACE to continue. A fixed buffer size (4K) is used, which sets a limit on the permmited output before paging fails and the output is simply written all in one lot. In most cases this is plenty, but 'env print' does print the whole environment as one string, so if the legacy distro-boot scripts are used it could happen. The default number of visible lines is set with a Kconfig option. Where a video terminal is being used, the page size is set to match that. In general U-Boot cannot guess the number of visible lines, since a serial terminal may be in use. There is also a 'pager' environment variable which can override any detected value. This initial series only counts newlines. It has no support for counting characters so the result will be sub-optimal if very long lines are written. This could be addressed in a future patch. Future work may detect the terminal size by sending an ANSI sequence. Simon Glass (16): console: Add the basic pager implementation console: Add tests for the basic functionality console: Provide a way to output without the pager console: Rename console-put functions to prepare for pager console: Plumb in the pager console: test: Allow tests to bypass the pager console: Support paging with single characters console: Add a few more paging tests console: Reset the pager when entering a new command console: Implement an environment var to control the pager console: Get the pager lines from the vidconsole efi: Enable the console pager for the app doc: Remove obsolete text in README.console doc: Move console docs to rST doc: Tidy up the console docs a little pager: doc: Add mention of the pager feature common/Kconfig | 20 ++ common/Makefile | 2 + common/cli_readline.c | 13 +- common/console.c | 99 +++-- common/pager.c | 208 +++++++++++ doc/README.console | 100 ------ doc/usage/console.rst | 78 ++++ doc/usage/environment.rst | 8 + doc/usage/index.rst | 1 + include/asm-generic/global_data.h | 11 + include/env_callback.h | 7 + include/pager.h | 178 +++++++++ lib/efi_client/Kconfig | 1 + test/common/Makefile | 1 + test/common/pager.c | 576 ++++++++++++++++++++++++++++++ test/test-main.c | 3 + 16 files changed, 1183 insertions(+), 123 deletions(-) create mode 100644 common/pager.c delete mode 100644 doc/README.console create mode 100644 doc/usage/console.rst create mode 100644 include/pager.h create mode 100644 test/common/pager.c -- 2.43.0 base-commit: 96d39714e407ff11177434b835ab1a7d46c8c695 branch: cli

1 16

[PATCH 0/7] efi: Improvements to env print -e
by Simon Glass 21 Aug '25

21 Aug '25

From: Simon Glass <sjg(a)chromium.org> Printing EFI variables can be quite verbose, with many hundreds of lines of text. Part of this is because GUIDs and hex dumps are included. It can also be hard to find a few variables visually in an unsorted list. This series makes the feature more user-friendly: - Puts verbose output behind a -v flag, so that 'env print -e' behaves more like 'env print' - Adds a -s option to sort the list Simon Glass (7): efi: Move guid used for variables to common files efi: Enable access to efi variables from the app efi: Change env print -e to show only names by default efi: Rename var_name64 to name in efi_dump_var_all() efi: Refactor error handling in efi_dump_var_all() efi: nvedit: Place variables in an alist before printing them cmd: Add -s flag to printenv -e for sorted output cmd/Kconfig | 1 + cmd/nvedit.c | 6 +- cmd/nvedit_efi.c | 164 ++++++++++++------ doc/usage/cmd/env.rst | 9 +- doc/usage/cmd/printenv.rst | 84 ++++++++- include/efi.h | 2 + include/efi_loader.h | 1 - lib/efi/device_path.c | 3 + lib/efi_loader/efi_var_common.c | 3 - .../py/tests/test_efi_secboot/test_authvar.py | 46 ++--- 10 files changed, 231 insertions(+), 88 deletions(-) -- 2.43.0 base-commit: 96d39714e407ff11177434b835ab1a7d46c8c695 branch: loadt

1 7

[PATCH 00/22] efi: Improvements for the EFI app on ARM
by Simon Glass 20 Aug '25

20 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series provides a number of minor improvements for the EFI app when running on ARM machines (on top of Tianocore, for example): - Tidy up various comments - Show the ARM exception level in bdinfo - Tidy up the output of 'meminfo' - Get the addr_find command running - Reduce verbosity when not debugging - Show the model at the top of the diplay when using vidconsole - Allow faking the boot right into the actual OS jump (for debugging) Changes in v2: - Make the messages longer and more explanatory - Use the existing current_el() function - Add a colon so it is clear that 'load_os type' is not a compound term Simon Glass (22): arm: bootm: Add some debugging arm: Fix swtiching typo arm: Show the exception level with bdinfo arm: Drop kernel_entry for arm64 meminfo: Allow for up to 10 hex digits event: Add a dummy function for event_notify() boot: Improve debugging in bootm_load_os() boot: Pass flags to the bootm_final event fdt: Add debugging for fdt_simplefb efi: app: Show only a summary of disks and partitions efi: Fix up the addr_find command efi: app: Tidy up some stale comments in setup_memory() efi: app: Allocate pages in any region efi: app: Pick up the SMBIOS table efi: app: Use EFI_PAGE_SHIFT instead of 12 in free_memory() efi: app: Print the final message before freeing memory efi: app: Use the relocated global_data efi: app: Only show the memory map when debugging efi: app: Add a simplefb node to the devicetree efi: arm: Increase the cyclic timeout again efi: app: Show the model when the vidconsole starts efi: app: Enable the cat command arch/arm/lib/bdinfo.c | 4 ++++ arch/arm/lib/bootm.c | 15 +++++++------ board/efi/efi-arm_app/board.c | 4 ++++ boot/bootm.c | 3 ++- boot/bootm_final.c | 13 ++++++----- boot/fdt_simplefb.c | 4 ++++ cmd/Kconfig | 3 ++- cmd/addr_find.c | 40 +++++++++++---------------------- cmd/meminfo.c | 9 ++++---- configs/efi-arm_app64_defconfig | 5 ++++- include/bootm.h | 13 +---------- include/event.h | 17 +++++++++++++- include/event_decl.h | 27 ++++++++++++++++++++++ lib/Kconfig | 3 ++- lib/efi_client/efi_app.c | 37 +++++++++++++++++++++--------- lib/efi_client/efi_app_init.c | 10 ++++++--- test/cmd/meminfo.c | 2 +- 17 files changed, 135 insertions(+), 74 deletions(-) create mode 100644 include/event_decl.h -- 2.43.0 base-commit: 12f3fc10fd44ee320b79236d7ff0ea3c2d78c12e branch: loadp2

3 31

[PATCH] efi: app: Don't bother showing the memory details in banner
by Simon Glass 20 Aug '25

20 Aug '25

From: Simon Glass <sjg(a)chromium.org> This information is not useful to users, only for debugging. So use log_debug() for it. Signed-off-by: Simon Glass <sjg(a)chromium.org> Suggested-by: Heinrich Schuchardt <xypron.glpk(a)gmx.de> --- lib/efi_client/efi_app.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/lib/efi_client/efi_app.c b/lib/efi_client/efi_app.c index df51c5a4de9..f583fa0b7a5 100644 --- a/lib/efi_client/efi_app.c +++ b/lib/efi_client/efi_app.c @@ -116,19 +116,19 @@ static efi_status_t setup_memory(struct efi_priv *priv) ret = boot->allocate_pages(EFI_ALLOCATE_MAX_ADDRESS, priv->image_data_type, pages, &addr); if (ret) { - log_info("(any address) "); + log_debug("(any address) "); ret = boot->allocate_pages(EFI_ALLOCATE_ANY_PAGES, priv->image_data_type, pages, &addr); } if (ret) { - log_info("(using pool %lx) ", ret); + log_debug("(using pool %lx) ", ret); priv->ram_base = (ulong)efi_malloc(priv, CONFIG_EFI_RAM_SIZE, &ret); if (!priv->ram_base) return ret; priv->use_pool_for_malloc = true; } else { - log_info("(using allocated RAM address %lx) ", (ulong)addr); + log_debug("(using allocated RAM address %lx) ", (ulong)addr); priv->ram_base = addr; } gd->ram_base = addr; -- 2.43.0 base-commit: d33a7d75c25c9c19019919897d248687c15f0974 branch: loadu

1 0

[PATCH v2 0/5] ACPI fixes
by Simon Glass 20 Aug '25

20 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series collects a few fix-ups for the recent BGRT series and tweaks the ACPI tables to use U-Boot as the 'creator' of the tables. It also adds a new logo for U-Boot (with text below). Changes in v2: - Add new patch to explain why boot-services data is used for BGRT image Simon Glass (5): tools: Add the logo with text under it acpi: Explain why boot-services data is used for BGRT image acpi: Use the logo with text underneath acpi: Check ACPI revision before accessing XSDT acpi: Use the U-Boot ACPI ID and version arch/x86/lib/acpi_nhlt.c | 2 +- arch/x86/lib/acpi_table.c | 2 +- include/acpi/acpi_table.h | 5 +-- lib/acpi/acpi.c | 8 +++-- lib/acpi/acpi_extra.c | 11 +++++- lib/acpi/acpi_table.c | 2 +- lib/acpi/bgrt_image.bmp | Bin 43926 -> 56574 bytes test/dm/acpi.c | 58 +++++++++++++++++++------------ tools/logos/u_boot_logo_text.svg | Bin 0 -> 10733 bytes 9 files changed, 55 insertions(+), 33 deletions(-) create mode 100644 tools/logos/u_boot_logo_text.svg -- 2.43.0 base-commit: d33a7d75c25c9c19019919897d248687c15f0974 branch: bgr2

1 5

[PATCH 00/14] efi: app: Support booting an OS
by Simon Glass 20 Aug '25

20 Aug '25

From: Simon Glass <sjg(a)chromium.org> This series completes the work to get the EFI app booting an EFI application (e.g. Ubuntu): - efidebug dh now works, and is slightly enhanced - device paths which refer to the underlying EFI layer can now be requested by the app - a test is provided, for EFI on ARM booting into Ubuntu using the sjg lab Simon Glass (14): test/py: Support a timing report when max_count is 0 efi: Drop EFI_CALL() from efidebug command efi: Move driver-binding guid to a common file efi: Drop the BS macro with efi debug efi: Use a function to obtain boot services in efi debug efi: Use a local var for a handle with efidebug dh efi: Show the device name only with the EFI loader efi: Update efidebug dh to work with the app efi: Add the component-name2 protocol to efidebug dh efi: Show component names and other info with efidebug dh efi: Add device-path support for EFI media devices efi: Add a little debugging to calculate_paths() efi: app: Allow booting an EFI app test: efi: Add a test for the app booting Ubuntu via EFI cmd/efidebug.c | 62 ++++++++++++------ include/efi.h | 3 + include/efi_api.h | 15 +++++ include/efi_loader.h | 2 - lib/efi/Kconfig | 2 +- lib/efi/device_path.c | 120 +++++++++++++++++++++++++++++++++- lib/efi/run.c | 8 ++- lib/efi_loader/efi_boottime.c | 8 ++- lib/uuid.c | 1 + test/py/conftest.py | 5 +- test/py/tests/test_distro.py | 57 ++++++++++++++-- 11 files changed, 251 insertions(+), 32 deletions(-) -- 2.43.0 base-commit: 9e77a93eda06d17d0f9fa20d4936caf4b055576c branch: loadr

1 14