On Mon, Jun 08, 2026 at 07:19:28PM +0900, Masami Hiramatsu wrote:

On Fri, 05 Jun 2026 05:03:37 -0700
Breno Leitao [off-list ref] wrote:

quoted

Call xbc_prepend_embedded_cmdline() in setup_arch() right after the
CONFIG_CMDLINE merge and before strscpy(command_line, ...) so the
build-time-rendered embedded bootconfig "kernel" subtree is part of
boot_command_line by the time parse_early_param() runs. early_param()
handlers (mem=, earlycon=, loglevel=, ...) now see values supplied via
CONFIG_BOOT_CONFIG_EMBED_FILE without parsing bootconfig at runtime.

Gate the prepend on the bootconfig opt-in: only fold in the embedded
kernel.* keys when "bootconfig" is present on the command line, or
CONFIG_BOOT_CONFIG_FORCE is set. Applying the embedded cmdline
unconditionally would (a) diverge from how embedded init.* keys are
treated and (b) break fail-safe recovery: a malformed embedded
console=/mem= could panic the boot with no way for the admin to disable
it by dropping "bootconfig" from the bootloader cmdline.
cmdline_find_option_bool() runs before parse_early_param(), so the gate
is cheap and correctly ordered.

Select ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG so the user-visible
CONFIG_BOOT_CONFIG_EMBED_CMDLINE option becomes selectable on x86.

This seems like a dummy config. what code does depend on this flag?

No C code reads ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG directly — it's
a silent gating symbol, the same ARCH_SUPPORTS_* idiom as
ARCH_SUPPORTS_CFI, ARCH_SUPPORTS_LTO_CLANG, etc.

Its only role is the depends on line of BOOT_CONFIG_EMBED_CMDLINE: an
arch selects it once its setup_arch() calls
xbc_prepend_embedded_cmdline(), and that makes the user-visible
BOOT_CONFIG_EMBED_CMDLINE selectable.

Right now, only x86 supports embedded bootconfig, thus, only x86 does
the following (last patch):

	config X86
	+       select ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG

So, no other platform can see CONFIG_BOOT_CONFIG_EMBED_CMDLINE.

quoted

--- a/init/main.c
+++ b/init/main.c

@@ -378,12 +378,15 @@ static void __init setup_boot_config(void)
 	int pos, ret;
 	size_t size;
 	char *err;
+	bool from_embedded = false;
 
 	/* Cut out the bootconfig data even if we have no bootconfig option */
 	data = get_boot_config_from_initrd(&size);
 	/* If there is no bootconfig in initrd, try embedded one. */
-	if (!data)
+	if (!data) {
 		data = xbc_get_embedded_bootconfig(&size);
+		from_embedded = true;

Even from embedded bootconfig, if the arch set 
ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG=n, this must be applied to
the cmdline as we are doing.

Right — that path is preserved. When the arch doesn't select
ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG, BOOT_CONFIG_EMBED_CMDLINE is
unselectable, so xbc_embedded_cmdline_applied() is the no-op stub
returning false.

quoted

 	strscpy(tmp_cmdline, boot_command_line, COMMAND_LINE_SIZE);
 	err = parse_args("bootconfig", tmp_cmdline, NULL, 0, 0, 0, NULL,

@@ -421,8 +424,17 @@ static void __init setup_boot_config(void)
 	} else {
 		xbc_get_info(&ret, NULL);
 		pr_info("Load bootconfig: %ld bytes %d nodes\n", (long)size, ret);
-		/* keys starting with "kernel." are passed via cmdline */
-		extra_command_line = xbc_make_cmdline("kernel");
+		/*
+		 * keys starting with "kernel." are passed via cmdline. When
+		 * this bootconfig came from the embedded source and
+		 * setup_arch() already prepended the rendered "kernel" subtree
+		 * to boot_command_line, rendering again here would duplicate
+		 * the keys in saved_command_line and make accumulating handlers
+		 * (console=, earlycon=, ...) re-register the same value. Skip
+		 * only when the prepend really happened.

Also, this should mention ARCH_SUPPORTS_CMDLINE_FROM_BOOTCONFIG=n case.

Ack, I will update

Thanks for the review,
--breno