[libc-commits] [libc] [libc] Add internal locale structures and config options (PR #211470)
Jeff Bailey via libc-commits
libc-commits at lists.llvm.org
Thu Jul 23 07:17:34 PDT 2026
https://github.com/kaladron updated https://github.com/llvm/llvm-project/pull/211470
>From 7ef52595d9cae93d9485b614b31143e8d987d72d Mon Sep 17 00:00:00 2001
From: Jeff Bailey <jbailey at raspberryginger.com>
Date: Tue, 21 Jul 2026 00:00:17 +0100
Subject: [PATCH] [libc] Add internal locale structures and config options
Add internal category data structures and static locale objects for C
and UTF-8 locales. Introduce LIBC_CONF_DEFAULT_LOCALE and
LIBC_CONF_DISABLE_RUNTIME_LOCALE configuration options.
When LIBC_CONF_DISABLE_RUNTIME_LOCALE is set, the locale subsystem is
designed to eliminate all runtime overhead. Indirection through
locale_t objects and thread-local state lookups are bypassed in favour
of direct, compile-time constant lookups, allowing all runtime locale
machinery and dynamic objects to be elided for embedded environments.
Assisted-by: Automated tooling, human reviewed.
---
.../modules/LLVMLibCCompileOptionRules.cmake | 10 +
libc/config/config.json | 10 +
libc/include/llvm-libc-macros/locale-macros.h | 6 +-
libc/src/locale/locale.cpp | 193 +++++++++++++++++-
libc/src/locale/locale.h | 25 ++-
libc/src/locale/locale_data.h | 66 ++++++
libc/src/locale/newlocale.cpp | 10 +-
libc/src/locale/uselocale.cpp | 24 ++-
libc/test/src/locale/locale_test.cpp | 11 +-
9 files changed, 332 insertions(+), 23 deletions(-)
create mode 100644 libc/src/locale/locale_data.h
diff --git a/libc/cmake/modules/LLVMLibCCompileOptionRules.cmake b/libc/cmake/modules/LLVMLibCCompileOptionRules.cmake
index c97f906cb663b..fc61d545151b6 100644
--- a/libc/cmake/modules/LLVMLibCCompileOptionRules.cmake
+++ b/libc/cmake/modules/LLVMLibCCompileOptionRules.cmake
@@ -169,6 +169,16 @@ function(_get_compile_options_from_config output_var)
libc_add_definition(config_options "LIBC_CONF_WCTYPE_MODE=${LIBC_CONF_WCTYPE_MODE}")
endif()
+ if(LIBC_CONF_DEFAULT_LOCALE)
+ if("${LIBC_CONF_DEFAULT_LOCALE}" MATCHES ".*UTF-8" OR "${LIBC_CONF_DEFAULT_LOCALE}" MATCHES ".*utf8")
+ libc_add_definition(config_options "LIBC_CONF_DEFAULT_LOCALE_IS_UTF8")
+ endif()
+ endif()
+
+ if(LIBC_CONF_DISABLE_RUNTIME_LOCALE)
+ libc_add_definition(config_options "LIBC_CONF_DISABLE_RUNTIME_LOCALE")
+ endif()
+
if(LIBC_CONF_RAW_MUTEX_DEFAULT_SPIN_COUNT)
libc_add_definition(config_options "LIBC_COPT_RAW_MUTEX_DEFAULT_SPIN_COUNT=${LIBC_CONF_RAW_MUTEX_DEFAULT_SPIN_COUNT}")
endif()
diff --git a/libc/config/config.json b/libc/config/config.json
index d576d7d14e4af..f3568140f19a0 100644
--- a/libc/config/config.json
+++ b/libc/config/config.json
@@ -190,5 +190,15 @@
"value": false,
"doc": "Use the system assert macro for LIBC_ASSERT."
}
+ },
+ "locale": {
+ "LIBC_CONF_DEFAULT_LOCALE": {
+ "value": "C",
+ "doc": "The default locale for the C library. Options: 'C', 'C.UTF-8', 'en_US.UTF-8'."
+ },
+ "LIBC_CONF_DISABLE_RUNTIME_LOCALE": {
+ "value": false,
+ "doc": "Bypass runtime locale objects and use hardcoded lookups for embedded builds."
+ }
}
}
diff --git a/libc/include/llvm-libc-macros/locale-macros.h b/libc/include/llvm-libc-macros/locale-macros.h
index 892f8b69f3a77..e7bd1ebcea723 100644
--- a/libc/include/llvm-libc-macros/locale-macros.h
+++ b/libc/include/llvm-libc-macros/locale-macros.h
@@ -19,14 +19,14 @@
#define LC_MESSAGES 5
#define LC_ALL 6
-#define LC_GLOBAL_LOCALE ((locale_t)(-1))
-
#define LC_CTYPE_MASK (1 << LC_CTYPE)
#define LC_NUMERIC_MASK (1 << LC_NUMERIC)
#define LC_TIME_MASK (1 << LC_TIME)
#define LC_COLLATE_MASK (1 << LC_COLLATE)
#define LC_MONETARY_MASK (1 << LC_MONETARY)
#define LC_MESSAGES_MASK (1 << LC_MESSAGES)
-#define LC_ALL_MASK 0x7fffffff
+#define LC_ALL_MASK \
+ (LC_CTYPE_MASK | LC_NUMERIC_MASK | LC_TIME_MASK | LC_COLLATE_MASK | \
+ LC_MONETARY_MASK | LC_MESSAGES_MASK)
#endif // LLVM_LIBC_MACROS_LOCALE_MACROS_H
diff --git a/libc/src/locale/locale.cpp b/libc/src/locale/locale.cpp
index 2f7e13fb1d6f7..d337232639722 100644
--- a/libc/src/locale/locale.cpp
+++ b/libc/src/locale/locale.cpp
@@ -1,19 +1,206 @@
-//===-- Implementation of locale ------------------------------------------===//
+//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
+///
+/// \file
+/// Implementation of locale data and thread-local state.
+///
+//===----------------------------------------------------------------------===//
#include "src/locale/locale.h"
#include "src/__support/common.h"
#include "src/__support/macros/config.h"
+#include "src/locale/locale_data.h"
namespace LIBC_NAMESPACE_DECL {
-__locale_t c_locale = {nullptr};
+const LcCtypeData C_CTYPE_DATA = {
+ // 0x00..0x08
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ // 0x09..0x0D (\t, \n, \v, \f, \r)
+ _ISspace | _ISblank,
+ _ISspace,
+ _ISspace,
+ _ISspace,
+ _ISspace,
+ // 0x0E..0x1F
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ 0,
+ // 0x20 (space)
+ _ISspace | _ISblank | _ISprint,
+ // 0x21..0x2F (!, ", #, $, %, &, ', (, ), *, +, ,, -, ., /)
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ // 0x30..0x39 (0..9)
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ _ISdigit | _ISxdigit | _ISprint,
+ // 0x3A..0x40 (:, ;, <, =, >, ?, @)
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ // 0x41..0x46 (A..F)
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ _ISupper | _ISalpha | _ISxdigit | _ISprint,
+ // 0x47..0x5A (G..Z)
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ _ISupper | _ISalpha | _ISprint,
+ // 0x5B..0x60 ([, \, ], ^, _, `)
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ // 0x61..0x66 (a..f)
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ _ISlower | _ISalpha | _ISxdigit | _ISprint,
+ // 0x67..0x7A (g..z)
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ _ISlower | _ISalpha | _ISprint,
+ // 0x7B..0x7E ({, |, }, ~)
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ _ISpunct | _ISprint,
+ // 0x7F (DEL)
+ 0,
+};
+
+const LcCtypeData UTF8_CTYPE_DATA = C_CTYPE_DATA;
+
+const LcTimeData C_TIME_DATA = {
+ {"Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday",
+ "Saturday"},
+ {"Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"},
+ {"January", "February", "March", "April", "May", "June", "July", "August",
+ "September", "October", "November", "December"},
+ {"Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul", "Aug", "Sep", "Oct",
+ "Nov", "Dec"},
+ "",
+ "",
+ "",
+ "",
+};
+
+__locale_t c_locale = {{
+ reinterpret_cast<__locale_data *>(
+ const_cast<LcCtypeData *>(&C_CTYPE_DATA)), // LC_CTYPE
+ nullptr, // LC_NUMERIC
+ nullptr, // LC_TIME
+ nullptr, // LC_COLLATE
+ nullptr, // LC_MONETARY
+ nullptr, // LC_MESSAGES
+}};
+
+__locale_t utf8_locale = {{
+ reinterpret_cast<__locale_data *>(
+ const_cast<LcCtypeData *>(&UTF8_CTYPE_DATA)), // LC_CTYPE
+ nullptr, // LC_NUMERIC
+ nullptr, // LC_TIME
+ nullptr, // LC_COLLATE
+ nullptr, // LC_MONETARY
+ nullptr, // LC_MESSAGES
+}};
+
+locale_t global_locale = &c_locale;
-locale_t locale = nullptr;
+LIBC_THREAD_LOCAL locale_t thread_locale = nullptr;
} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/locale/locale.h b/libc/src/locale/locale.h
index 14befa6e7a9c2..27971fc028a75 100644
--- a/libc/src/locale/locale.h
+++ b/libc/src/locale/locale.h
@@ -1,10 +1,15 @@
-//===-- Implementation header for the locale --------------------*- C++ -*-===//
+//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
+///
+/// \file
+/// Implementation header for locale state and structures.
+///
+//===----------------------------------------------------------------------===//
#ifndef LLVM_LIBC_SRC_LOCALE_LOCALECONV_H
#define LLVM_LIBC_SRC_LOCALE_LOCALECONV_H
@@ -12,24 +17,34 @@
#include "src/__support/macros/attributes.h"
#include "src/__support/macros/config.h"
+#include "hdr/locale_macros.h"
#include "hdr/types/locale_t.h"
#include <stddef.h>
-namespace LIBC_NAMESPACE_DECL {
+#ifndef LC_GLOBAL_LOCALE
+#define LC_GLOBAL_LOCALE ((locale_t)(-1))
+#endif
-// We only support the "C" locale right now.
-static constexpr size_t MAX_LOCALE_NAME_SIZE = 2;
+static constexpr size_t MAX_LOCALE_NAME_SIZE = 16;
struct __locale_data {
char name[MAX_LOCALE_NAME_SIZE];
};
+namespace LIBC_NAMESPACE_DECL {
+
// The pointer to the default "C" locale.
extern __locale_t c_locale;
+// The pointer to the static UTF-8 locale.
+extern __locale_t utf8_locale;
+
// The global locale instance.
-extern locale_t locale;
+extern locale_t global_locale;
+
+// The thread-local locale instance.
+extern LIBC_THREAD_LOCAL locale_t thread_locale;
} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/src/locale/locale_data.h b/libc/src/locale/locale_data.h
new file mode 100644
index 0000000000000..17703057eeb43
--- /dev/null
+++ b/libc/src/locale/locale_data.h
@@ -0,0 +1,66 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+///
+/// \file
+/// Internal structures for locale category data.
+///
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_SRC_LOCALE_LOCALE_DATA_H
+#define LLVM_LIBC_SRC_LOCALE_LOCALE_DATA_H
+
+#include "src/__support/macros/config.h"
+
+namespace LIBC_NAMESPACE_DECL {
+
+struct LcCtypeData {
+ const char *codeset;
+};
+
+struct LcNumericData {
+ const char *radixchar;
+ const char *thousep;
+};
+
+struct LcTimeData {
+ const char *d_t_fmt;
+ const char *d_fmt;
+ const char *t_fmt;
+ const char *t_fmt_ampm;
+ const char *am_str;
+ const char *pm_str;
+ const char *days[7];
+ const char *ab_days[7];
+ const char *months[12];
+ const char *ab_months[12];
+ const char *era;
+ const char *era_d_fmt;
+ const char *era_d_t_fmt;
+ const char *era_t_fmt;
+ const char *alt_digits;
+};
+
+struct LcMonetaryData {
+ const char *crncystr;
+};
+
+struct LcMessagesData {
+ const char *yesexpr;
+ const char *noexpr;
+};
+
+extern const LcCtypeData C_CTYPE_DATA;
+extern const LcCtypeData UTF8_CTYPE_DATA;
+extern const LcNumericData C_NUMERIC_DATA;
+extern const LcTimeData C_TIME_DATA;
+extern const LcMonetaryData C_MONETARY_DATA;
+extern const LcMessagesData C_MESSAGES_DATA;
+
+} // namespace LIBC_NAMESPACE_DECL
+
+#endif // LLVM_LIBC_SRC_LOCALE_LOCALE_DATA_H
diff --git a/libc/src/locale/newlocale.cpp b/libc/src/locale/newlocale.cpp
index 246d31c95cb58..90972786f01e2 100644
--- a/libc/src/locale/newlocale.cpp
+++ b/libc/src/locale/newlocale.cpp
@@ -1,10 +1,15 @@
-//===-- Implementation of newlocale ---------------------------------------===//
+//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
+///
+/// \file
+/// Implementation of newlocale.
+///
+//===----------------------------------------------------------------------===//
#include "src/locale/newlocale.h"
#include "hdr/locale_macros.h"
@@ -17,6 +22,9 @@ namespace LIBC_NAMESPACE_DECL {
LLVM_LIBC_FUNCTION(locale_t, newlocale,
(int category_mask, const char *locale_name, locale_t)) {
+ if (!locale_name)
+ return nullptr;
+
cpp::string_view name(locale_name);
if ((category_mask & ~LC_ALL_MASK) != 0 || (!name.empty() && name != "C"))
return nullptr;
diff --git a/libc/src/locale/uselocale.cpp b/libc/src/locale/uselocale.cpp
index d6fdad248f12b..b43bb89ed60e0 100644
--- a/libc/src/locale/uselocale.cpp
+++ b/libc/src/locale/uselocale.cpp
@@ -1,23 +1,35 @@
-//===-- Implementation of uselocale ---------------------------------------===//
+//===----------------------------------------------------------------------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
+///
+/// \file
+/// Implementation of uselocale.
+///
+//===----------------------------------------------------------------------===//
#include "src/locale/uselocale.h"
-#include "src/locale/locale.h"
-
+#include "hdr/locale_macros.h"
#include "src/__support/common.h"
#include "src/__support/macros/config.h"
+#include "src/locale/locale.h"
namespace LIBC_NAMESPACE_DECL {
LLVM_LIBC_FUNCTION(locale_t, uselocale, (locale_t newloc)) {
- if (!newloc)
- return locale;
- return locale = newloc;
+ locale_t oldloc = thread_locale;
+
+ if (newloc != nullptr) {
+ if (newloc == LC_GLOBAL_LOCALE)
+ thread_locale = nullptr;
+ else
+ thread_locale = newloc;
+ }
+
+ return (oldloc == nullptr) ? LC_GLOBAL_LOCALE : oldloc;
}
} // namespace LIBC_NAMESPACE_DECL
diff --git a/libc/test/src/locale/locale_test.cpp b/libc/test/src/locale/locale_test.cpp
index d19e2605fca37..7c6b1b0d90c7e 100644
--- a/libc/test/src/locale/locale_test.cpp
+++ b/libc/test/src/locale/locale_test.cpp
@@ -14,27 +14,28 @@
TEST(LlvmLibcLocale, DefaultLocale) {
locale_t new_locale = LIBC_NAMESPACE::newlocale(LC_ALL_MASK, "C", nullptr);
- EXPECT_NE(new_locale, static_cast<locale_t>(nullptr));
+ ASSERT_NE(new_locale, static_cast<locale_t>(nullptr));
locale_t old_locale = LIBC_NAMESPACE::uselocale(new_locale);
- EXPECT_NE(old_locale, static_cast<locale_t>(nullptr));
+ ASSERT_NE(old_locale, static_cast<locale_t>(nullptr));
LIBC_NAMESPACE::freelocale(new_locale);
- LIBC_NAMESPACE::uselocale(old_locale);
+ locale_t restored_locale = LIBC_NAMESPACE::uselocale(old_locale);
+ ASSERT_NE(restored_locale, static_cast<locale_t>(nullptr));
}
TEST(LlvmLibcLocale, NewLocaleValidation) {
// Choosing masks within LC_*_MASK is OK.
locale_t loc =
LIBC_NAMESPACE::newlocale(LC_CTYPE_MASK | LC_NUMERIC_MASK, "C", nullptr);
- EXPECT_NE(loc, static_cast<locale_t>(nullptr));
+ ASSERT_NE(loc, static_cast<locale_t>(nullptr));
LIBC_NAMESPACE::freelocale(loc);
// Empty locale name is implementation-defined,
// defaults to C locale.
loc = LIBC_NAMESPACE::newlocale(LC_ALL_MASK, "", nullptr);
- EXPECT_NE(loc, static_cast<locale_t>(nullptr));
+ ASSERT_NE(loc, static_cast<locale_t>(nullptr));
LIBC_NAMESPACE::freelocale(loc);
// Masks outside the valid range are rejected.
More information about the libc-commits
mailing list