kconfig: Introduce cc-option-str

Kconfig has offered "def_bool" and "def_tristate" as a shorthand for a type definition plus a default since before the git era, but has never offered the equivalent for the other three types. Conor Dooley ran into

this gap[1] when fixing a symbol that had been given the wrong type:

Unfortunately, there is no such thing as "def_string", but in this case we can use "default" to propagate the value of ...

Nothing in the grammar requires the restriction. The rule that consumes a default is already type agnostic. Add the three missing types. No changes are needed to existing diagnostics. E.g. declaring a symbol "bool" and then assigning it with "def_string" still reports

warning: ignoring type redefinition of 'CONFLICT' from 'bool' to 'string'

Added tests for the types.

Build tested ARCH=x86_64 with GCC 16.2.0. Tests pass with "make testconfig".

Link https://lore.kernel.org/all/[email protected]/ [1] open here

Assisted-by LLM

Signed-off-by Kees Cook <[email protected]>

---
Cc: Nathan Chancellor <[email protected]>
Cc: Nicolas Schier <[email protected]>
Cc: Julian Braha <[email protected]>
Cc: Jonathan Corbet <[email protected]>
Cc: Shuah Khan <[email protected]>
Cc: Randy Dunlap <[email protected]>
Cc: Masahiro Yamada <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Cc: Nicolas Pitre <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Andy Shevchenko <[email protected]>
Cc: Andrew Jones <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>

---
 scripts/kconfig/tests/def_type/Kconfig        | 28 +++++++++++++++++++
 scripts/kconfig/tests/def_type/guard_n.config |  1 +
 scripts/kconfig/tests/def_type/guard_y.config |  1 +
 scripts/kconfig/tests/def_type/__init__.py    | 18 ++++++++++++
 .../kconfig/tests/def_type/expected_guard_n   |  9 ++++++
 .../kconfig/tests/def_type/expected_guard_y   | 11 ++++++++
 scripts/kconfig/kconfig-sym-check.pl          |  2 +-
 scripts/kconfig/lexer.l                       |  3 ++
 scripts/kconfig/parser.y                      |  6 ++++
 Documentation/kbuild/kconfig-language.rst     | 13 ++++++++-
 10 files changed, 90 insertions(+), 2 deletions(-)
 create mode 100644 scripts/kconfig/tests/def_type/Kconfig
 create mode 100644 scripts/kconfig/tests/def_type/guard_n.config
 create mode 100644 scripts/kconfig/tests/def_type/guard_y.config
 create mode 100644 scripts/kconfig/tests/def_type/__init__.py
 create mode 100644 scripts/kconfig/tests/def_type/expected_guard_n
 create mode 100644 scripts/kconfig/tests/def_type/expected_guard_y

scripts/kconfig/tests/def_type/Kconfig new file +28 -0

new file mode 100644
index 000000000000..fbee37a63179
@@ -0,0 +1,28 @@
+# SPDX-License-Identifier: GPL-2.0
+# The def_<type> shorthands: a type definition plus a default value.
+
+config MODULES
+	bool "Enable loadable module support"
+	modules
+	default y
+
+config GUARD
+	bool "Guard symbol"
+
+config DEF_BOOL
+	def_bool GUARD
+
+config DEF_TRISTATE
+	def_tristate m if GUARD
+
+config DEF_STRING
+	def_string "guarded" if GUARD
+	default "fallback"
+
+config DEF_INT
+	def_int 64 if GUARD
+	default 32
+
+config DEF_HEX
+	def_hex 0xdead if GUARD
+	default 0x0

scripts/kconfig/tests/def_type/guard_n.config new file +1 -0

new file mode 100644
index 000000000000..ed9ad6c1a2d4
@@ -0,0 +1 @@
+# CONFIG_GUARD is not set

scripts/kconfig/tests/def_type/guard_y.config new file +1 -0

new file mode 100644
index 000000000000..afe35b084542
@@ -0,0 +1 @@
+CONFIG_GUARD=y

scripts/kconfig/tests/def_type/__init__.py new file +18 -0

new file mode 100644
index 000000000000..1ebaf5da07c8
@@ -0,0 +1,18 @@
+# SPDX-License-Identifier: GPL-2.0
+"""
+Set a symbol's type and its default value in one line.
+
+"def_bool", "def_tristate", "def_string", "def_int" and "def_hex" are
+shorthand for a type definition plus a "default" property.  Check that
+each one sets the type, and that an "if" on the shorthand does not
+disturb the usual default cascade: the shorthand is only the first arm
+of the list, so a later "default" still applies when its condition is
+not met.
+"""
+
+def test(conf):
+    assert conf.olddefconfig(dot_config='guard_y.config') == 0
+    assert conf.config_matches('expected_guard_y')
+
+    assert conf.olddefconfig(dot_config='guard_n.config') == 0
+    assert conf.config_matches('expected_guard_n')

scripts/kconfig/tests/def_type/expected_guard_n new file +9 -0

new file mode 100644
index 000000000000..14719720a8b9
@@ -0,0 +1,9 @@
+#
+# Automatically generated file; DO NOT EDIT.
+# Main menu
+#
+CONFIG_MODULES=y
+# CONFIG_GUARD is not set
+CONFIG_DEF_STRING="fallback"
+CONFIG_DEF_INT=32
+CONFIG_DEF_HEX=0x0

scripts/kconfig/tests/def_type/expected_guard_y new file +1 -0

new file mode 100644
index 000000000000..b2844072d0b8
@@ -0,0 +1,11 @@
+#

scripts/kconfig/kconfig-sym-check.pl +1 -1

index daa5285fdefc..c8dd07f8b27c 100755
@@ -90,7 +90,7 @@ foreach my $file (@files) {
 			next;
 		}
 
-		if (/^\s*(default|def_bool|def_tristate|select|depends\s+on|imply|visible\s+if|range|if|bool|tristate|int|hex|string|prompt)\s+(.+)\s*$/) {
+		if (/^\s*(default|def_bool|def_tristate|def_string|def_int|def_hex|select|depends\s+on|imply|visible\s+if|range|if|bool|tristate|int|hex|string|prompt)\s+(.+)\s*$/) {
 			my $s = $2;
 			$s =~ s/"(?:[^"\\]|\\.)*"|'(?:[^'\\]|\\.)*'//g;
 			$s =~ s/#.*//;

scripts/kconfig/lexer.l +3 -0

index a6155422b4a6..1fa521199d4a 100644
@@ -105,6 +105,9 @@ n	[A-Za-z0-9_-]
 "comment"		return T_COMMENT;
 "config"		return T_CONFIG;
 "def_bool"		return T_DEF_BOOL;
+"def_hex"		return T_DEF_HEX;
+"def_int"		return T_DEF_INT;
+"def_string"		return T_DEF_STRING;
 "def_tristate"		return T_DEF_TRISTATE;
 "default"		return T_DEFAULT;
 "depends"		return T_DEPENDS;

scripts/kconfig/parser.y +6 -0

index 5fb6f07b6ad2..2174baf2b3fd 100644
@@ -53,6 +53,9 @@ struct menu *current_menu, *current_entry, *current_choice;
 %token T_CONFIG
 %token T_DEFAULT
 %token T_DEF_BOOL
+%token T_DEF_HEX
+%token T_DEF_INT
+%token T_DEF_STRING
 %token T_DEF_TRISTATE
 %token T_DEPENDS
 %token T_ENDCHOICE

@@ -309,6 +312,9 @@ type:
 default:
 	  T_DEFAULT		{ $$ = S_UNKNOWN; }
 	| T_DEF_BOOL		{ $$ = S_BOOLEAN; }
+	| T_DEF_HEX		{ $$ = S_HEX; }
+	| T_DEF_INT		{ $$ = S_INT; }
+	| T_DEF_STRING		{ $$ = S_STRING; }
 	| T_DEF_TRISTATE	{ $$ = S_TRISTATE; }
 
 /* if entry */

Documentation/kbuild/kconfig-language.rst +5 -1

index d9338407c1c6..00402d43e0dc 100644
@@ -113,11 +113,22 @@ applicable everywhere (see syntax).
 
 - type definition + default value::
 
-	"def_bool"/"def_tristate" <expr> ["if" <expr>]
+	"def_bool" <expr> ["if" <expr>]
+	"def_tristate" <expr> ["if" <expr>]
+	"def_string" <expr> ["if" <expr>]
+	"def_int" <expr> ["if" <expr>]
+	"def_hex" <expr> ["if" <expr>]
 
   This is a shorthand notation for a type definition plus a value.
   Optionally dependencies for this default value can be added with "if".
 

On 9/19/26 01:59, Kees Cook wrote:
[...]
the X condition here only applies to the value of y, but not to the type declaration of bool. Besides hurting readability, I can imagine a user making a mistake by attempting something like this:

def_bool 'y' if X
def_tristate 'y' if !X

thinking that they're making the type conditional. Of course, the interpreter warns if this is attempted, so you won't actually find any of these in the tree.

The second problem, is that since the order of defaults matters and conditions can shadow each other, def_<type> makes it harder for users to get defaults right. In the past, I've seen several config options with bugged defaults due to 'default' + 'def_<type>' [2][3][4].

Quoted 1 line

Signed-off-by Kees Cook <[email protected]>

Tested-by Julian Braha <[email protected]>

Reviewed-by Julian Braha <[email protected]>

Quoted 4 lines

The second problem, is that since the order of defaults matters and conditions can shadow each other, def_<type> makes it harder for users to get defaults right. In the past, I've seen several config options with bugged defaults due to 'default' + 'def_<type>' [2][3][4].

This is the one I, too, got worried about. From your example:

config GUEST_STATE_BUFFER_TEST

(quoted diff) +1 -2

-	def_tristate n
+	def_tristate KUNIT_ALL_TESTS
 	prompt "Enable Guest State Buffer unit tests"
 	depends on KUNIT
 	depends on KVM_BOOK3S_HV_POSSIBLE
-	default KUNIT_ALL_TESTS
 	help

Is there a way to detect this state? (I assume as a follow-up; it's a general problem.)

$(cc-option-bit) returns the flag it was given when the compiler accepts it, and an empty string otherwise. It was originally only used internally for machine bit flags, but it is generally useful[1].

Rename it to $(cc-option-str), which says what it returns rather than what it was first used for, and extract the probe logic from the existing $(cc-option) macro to reuse it. Additionally add support for an optional fallback, much like the Kbuild cc-option macro. This can be used as:

	config CC_OPT_FOO
		def_string "$(cc-option-str,-fnew-option,-fold-option)"

If the fallback is omitted, it is just an empty string.

Build tested ARCH=x86_64 defconfig with GCC 16.2.0.

Assisted-by LLM

Signed-off-by Kees Cook <[email protected]>

Add the linker counterpart to $(cc-option-str), sharing a $(ld-probe) with $(ld-option) the way the compiler macros share $(cc-probe). Like $(cc-option-str), it can now be used as:

	config LD_OPT_FOO
		def_string "$(ld-option-str,--new-option,--old-option)"

An omitted fallback expands to the empty string.

Build tested ARCH=x86_64 defconfig with GCC 16.2.0, and checked against GNU ld 2.47.

Assisted-by LLM

Signed-off-by Kees Cook <[email protected]>

Hi,

While reviewing the recent build speedup series[1], it became clear that we needed a general solution in kconfig to avoid repeating detected toolchain options. Implement cc-option-str and ld-option-str, and while at it, create "def_string" and friends to match "def_bool".

-Kees

[1] https://lore.kernel.org/all/20260918054058.GA384222@ax162/

Kees Cook (3):
kconfig: Add "def_string", "def_int" and "def_hex"
kconfig: Replace "cc-option-bit" with "cc-option-str"
kconfig: Add "ld-option-str"

Quoted 7 lines · Kees Cook

Hi,

While reviewing the recent build speedup series[1], it became clear that we needed a general solution in kconfig to avoid repeating detected toolchain options. Implement cc-option-str and ld-option-str, and while at it, create "def_string" and friends to match "def_bool".

Quoted 8 lines

-Kees

[1] https://lore.kernel.org/all/20260918054058.GA384222@ax162/

Kees Cook (3):
kconfig: Add "def_string", "def_int" and "def_hex"
kconfig: Replace "cc-option-bit" with "cc-option-str"
kconfig: Add "ld-option-str"

Thanks, exactly what I had in mind. This seems reasonable to me and passes my basic testing. I will let folks chime in for a few days then apply this to kbuild-next so that Lorenzo can use it as part of the speed up series.

--
Cheers,
Nathan

kconfig: Introduce cc-option-str

Hi,

While reviewing the recent build speedup series[1], it became clear that we needed a general solution in kconfig to avoid repeating detected toolchain options. Implement cc-option-str and ld-option-str, and while at it, create "def_string" and friends to match "def_bool".

[...]