UEFI expects strings to be UTF-16. So add an option so that when EFI_LOADER is enabled, we can use the expected unicode string size.

Signed-off-by: Rob Clark robdclark@gmail.com --- Kconfig | 8 ++++++++ Makefile | 4 ++++ lib/efi_loader/Kconfig | 1 + 3 files changed, 13 insertions(+)

diff --git a/Kconfig b/Kconfig index c1451bceda..7319f1fa94 100644 --- a/Kconfig +++ b/Kconfig @@ -53,6 +53,14 @@ config CC_OPTIMIZE_FOR_SIZE

This option is enabled by default for U-Boot.

+config CC_SHORT_WCHAR + bool "Use 16b wchar" + default n + help + Enabling this option will pass "-fshort-wchar" to gcc, for + 16bit unicode strings. This is used by EFI_LOADER, as the + UEFI spec defines strings to be UTF-16. + config DISTRO_DEFAULTS bool "Select defaults suitable for booting general purpose Linux distributions" default y if ARCH_SUNXI || TEGRA diff --git a/Makefile b/Makefile index 50a002e72f..91b11f5a7b 100644 --- a/Makefile +++ b/Makefile @@ -590,6 +590,10 @@ else KBUILD_CFLAGS += -O2 endif

+ifdef CONFIG_CC_SHORT_WCHAR +KBUILD_CFLAGS += -fshort-wchar +endif + KBUILD_CFLAGS += $(call cc-option,-fno-stack-protector) KBUILD_CFLAGS += $(call cc-option,-fno-delete-null-pointer-checks)

diff --git a/lib/efi_loader/Kconfig b/lib/efi_loader/Kconfig index d2b6327119..e28ef51ad4 100644 --- a/lib/efi_loader/Kconfig +++ b/lib/efi_loader/Kconfig @@ -1,6 +1,7 @@ config EFI_LOADER bool "Support running EFI Applications in U-Boot" depends on (ARM || X86) && OF_LIBFDT + select CC_SHORT_WCHAR default y help Select this option if you want to run EFI applications (like grub2)

We'll eventually want these in a few places in efi_loader, and also vsprintf.

Signed-off-by: Rob Clark robdclark@gmail.com --- include/charset.h | 55 ++++++++++++++++++++++++++++++ lib/Makefile | 1 + lib/charset.c | 81 ++++++++++++++++++++++++++++++++++++++++++++ lib/efi_loader/efi_console.c | 17 ++-------- 4 files changed, 140 insertions(+), 14 deletions(-) create mode 100644 include/charset.h create mode 100644 lib/charset.c

diff --git a/include/charset.h b/include/charset.h new file mode 100644 index 0000000000..47ff6c7af1 --- /dev/null +++ b/include/charset.h @@ -0,0 +1,55 @@ +/* + * charset conversion utils + * + * Copyright (c) 2017 Rob Clark + * + * SPDX-License-Identifier: GPL-2.0+ + */ + +#ifndef __CHARSET_H_ +#define __CHARSET_H_ + +#define MAX_UTF8_PER_UTF16 4 + +/** + * utf16_strlen() - Get the length of an utf16 string + * + * Returns the number of 16 bit characters in an utf16 string, not + * including the terminating NULL character. + * + * @in the string to measure + * @return the string length + */ +size_t utf16_strlen(uint16_t *in); + +/** + * utf16_strnlen() - Get the length of a fixed-size utf16 string. + * + * Returns the number of 16 bit characters in an utf16 string, + * not including the terminating NULL character, but at most + * 'count' number of characters. In doing this, utf16_strnlen() + * looks at only the first 'count' characters. + * + * @in the string to measure + * @count the maximum number of characters to count + * @return the string length, up to a maximum of 'count' + */ +size_t utf16_strnlen(const uint16_t *in, size_t count); + +/** + * utf16_to_utf8() - Convert an utf16 string to utf8 + * + * Converts 'size' characters of the utf16 string 'src' to utf8 + * written to the 'dest' buffer. + * + * NOTE that a single utf16 character can generate up to 4 utf8 + * characters. See MAX_UTF8_PER_UTF16. + * + * @dest the destination buffer to write the utf8 characters + * @src the source utf16 string + * @size the number of utf16 characters to convert + * @return the pointer to the first unwritten byte in 'dest' + */ +uint8_t *utf16_to_utf8(uint8_t *dest, const uint16_t *src, size_t size); + +#endif /* __CHARSET_H_ */ diff --git a/lib/Makefile b/lib/Makefile index eacc7d6485..b88b6ebd53 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -19,6 +19,7 @@ obj-$(CONFIG_OF_LIVE) += of_live.o obj-$(CONFIG_CMD_DHRYSTONE) += dhry/

obj-$(CONFIG_AES) += aes.o +obj-y += charset.o obj-$(CONFIG_USB_TTY) += circbuf.o obj-y += crc7.o obj-y += crc8.o diff --git a/lib/charset.c b/lib/charset.c new file mode 100644 index 0000000000..eaff2e542e --- /dev/null +++ b/lib/charset.c @@ -0,0 +1,81 @@ +/* + * charset conversion utils + * + * Copyright (c) 2017 Rob Clark + * + * SPDX-License-Identifier: GPL-2.0+ + */ + +#include <common.h> +#include <charset.h> + +/* + * utf8/utf16 conversion mostly lifted from grub + */ + +size_t utf16_strlen(uint16_t *in) +{ + size_t i; + for (i = 0; in[i]; i++); + return i; +} + +size_t utf16_strnlen(const uint16_t *in, size_t count) +{ + size_t i; + for (i = 0; count-- && in[i]; i++); + return i; +} + +/* Convert UTF-16 to UTF-8. */ +uint8_t *utf16_to_utf8(uint8_t *dest, const uint16_t *src, size_t size) +{ + uint32_t code_high = 0; + + while (size--) { + uint32_t code = *src++; + + if (code_high) { + if (code >= 0xDC00 && code <= 0xDFFF) { + /* Surrogate pair. */ + code = ((code_high - 0xD800) << 10) + (code - 0xDC00) + 0x10000; + + *dest++ = (code >> 18) | 0xF0; + *dest++ = ((code >> 12) & 0x3F) | 0x80; + *dest++ = ((code >> 6) & 0x3F) | 0x80; + *dest++ = (code & 0x3F) | 0x80; + } else { + /* Error... */ + *dest++ = '?'; + /* *src may be valid. Don't eat it. */ + src--; + } + + code_high = 0; + } else { + if (code <= 0x007F) { + *dest++ = code; + } else if (code <= 0x07FF) { + *dest++ = (code >> 6) | 0xC0; + *dest++ = (code & 0x3F) | 0x80; + } else if (code >= 0xD800 && code <= 0xDBFF) { + code_high = code; + continue; + } else if (code >= 0xDC00 && code <= 0xDFFF) { + /* Error... */ + *dest++ = '?'; + } else if (code < 0x10000) { + *dest++ = (code >> 12) | 0xE0; + *dest++ = ((code >> 6) & 0x3F) | 0x80; + *dest++ = (code & 0x3F) | 0x80; + } else { + *dest++ = (code >> 18) | 0xF0; + *dest++ = ((code >> 12) & 0x3F) | 0x80; + *dest++ = ((code >> 6) & 0x3F) | 0x80; + *dest++ = (code & 0x3F) | 0x80; + } + } + } + + return dest; +} diff --git a/lib/efi_loader/efi_console.c b/lib/efi_loader/efi_console.c index 5ebce4b544..3fc82b8726 100644 --- a/lib/efi_loader/efi_console.c +++ b/lib/efi_loader/efi_console.c @@ -7,6 +7,7 @@ */

#include <common.h> +#include <charset.h> #include <efi_loader.h>

static bool console_size_queried; @@ -138,20 +139,8 @@ static efi_status_t EFIAPI efi_cout_reset(

static void print_unicode_in_utf8(u16 c) { - char utf8[4] = { 0 }; - char *b = utf8; - - if (c < 0x80) { - *(b++) = c; - } else if (c < 0x800) { - *(b++) = 192 + c / 64; - *(b++) = 128 + c % 64; - } else { - *(b++) = 224 + c / 4096; - *(b++) = 128 + c / 64 % 64; - *(b++) = 128 + c % 64; - } - + char utf8[MAX_UTF8_PER_UTF16] = { 0 }; + utf16_to_utf8((u8 *)utf8, &c, 1); puts(utf8); }

This is convenient for efi_loader which deals a lot with UTF-16. Only enabled with CC_SHORT_WCHAR, leaving room to add a UTF-32 version when CC_SHORT_WCHAR is not enabled.

Signed-off-by: Rob Clark robdclark@gmail.com --- examples/api/Makefile | 1 + lib/vsprintf.c | 31 +++++++++++++++++++++++++++++-- 2 files changed, 30 insertions(+), 2 deletions(-)

diff --git a/examples/api/Makefile b/examples/api/Makefile index dab6398bab..87c15d0f68 100644 --- a/examples/api/Makefile +++ b/examples/api/Makefile @@ -34,6 +34,7 @@ EXT_COBJ-y += lib/div64.o EXT_COBJ-y += lib/string.o EXT_COBJ-y += lib/time.o EXT_COBJ-y += lib/vsprintf.o +EXT_COBJ-y += lib/charset.o EXT_SOBJ-$(CONFIG_PPC) += arch/powerpc/lib/ppcstring.o ifeq ($(ARCH),arm) EXT_SOBJ-$(CONFIG_USE_ARCH_MEMSET) += arch/arm/lib/memset.o diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 874a2951f7..0678b49b01 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -17,6 +17,7 @@ #include <linux/ctype.h>

#include <common.h> +#include <charset.h>

#include <div64.h> #define noinline __attribute__((noinline)) @@ -270,6 +271,26 @@ static char *string(char *buf, char *end, char *s, int field_width, return buf; }

+static char *string16(char *buf, char *end, u16 *s, int field_width, + int precision, int flags) +{ + u16 *str = s ? s : (u16[]){'<','N','U','L','L','>','\0'}; + int utf16_len = utf16_strnlen(str, precision); + u8 utf8[utf16_len * MAX_UTF8_PER_UTF16]; + int utf8_len, i; + + utf8_len = utf16_to_utf8(utf8, str, utf16_len) - utf8; + + if (!(flags & LEFT)) + while (utf8_len < field_width--) + ADDCH(buf, ' '); + for (i = 0; i < utf8_len; ++i) + ADDCH(buf, utf8[i]); + while (utf8_len < field_width--) + ADDCH(buf, ' '); + return buf; +} + #ifdef CONFIG_CMD_NET static const char hex_asc[] = "0123456789abcdef"; #define hex_asc_lo(x) hex_asc[((x) & 0x0f)] @@ -528,8 +549,14 @@ repeat: continue;

case 's': - str = string(str, end, va_arg(args, char *), - field_width, precision, flags); + if (CONFIG_IS_ENABLED(CC_SHORT_WCHAR) && + qualifier == 'l') { + str = string16(str, end, va_arg(args, u16 *), + field_width, precision, flags); + } else { + str = string(str, end, va_arg(args, char *), + field_width, precision, flags); + } continue;

case 'p':

This works (roughly) the same way as linux's, but we currently always print lower-case (ie. we just keep %pUB and %pUL for compat with linux), mostly just because that is what uuid_bin_to_str() supports.

%pUb: 01020304-0506-0708-090a-0b0c0d0e0f10 %pUl: 04030201-0605-0807-090a-0b0c0d0e0f10

It will be used by a later efi_loader paths for efi variables and for device-path-to-text protocol, and also quite useful for debug prints of protocol GUIDs.

Signed-off-by: Rob Clark robdclark@gmail.com --- examples/api/Makefile | 1 + include/config_fallbacks.h | 1 + lib/vsprintf.c | 46 ++++++++++++++++++++++++++++++++++++++++++++-- 3 files changed, 46 insertions(+), 2 deletions(-)

diff --git a/examples/api/Makefile b/examples/api/Makefile index 87c15d0f68..899527267d 100644 --- a/examples/api/Makefile +++ b/examples/api/Makefile @@ -35,6 +35,7 @@ EXT_COBJ-y += lib/string.o EXT_COBJ-y += lib/time.o EXT_COBJ-y += lib/vsprintf.o EXT_COBJ-y += lib/charset.o +EXT_COBJ-$(CONFIG_LIB_UUID) += lib/uuid.o EXT_SOBJ-$(CONFIG_PPC) += arch/powerpc/lib/ppcstring.o ifeq ($(ARCH),arm) EXT_SOBJ-$(CONFIG_USE_ARCH_MEMSET) += arch/arm/lib/memset.o diff --git a/include/config_fallbacks.h b/include/config_fallbacks.h index 961a83d758..56b9de09f2 100644 --- a/include/config_fallbacks.h +++ b/include/config_fallbacks.h @@ -57,6 +57,7 @@

#if (CONFIG_IS_ENABLED(PARTITION_UUIDS) || \ CONFIG_IS_ENABLED(EFI_PARTITION) || \ + CONFIG_IS_ENABLED(EFI_LOADER) || \ defined(CONFIG_RANDOM_UUID) || \ defined(CONFIG_CMD_UUID) || \ defined(CONFIG_BOOTP_PXE)) && \ diff --git a/lib/vsprintf.c b/lib/vsprintf.c index 0678b49b01..71a995dee0 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -18,6 +18,7 @@

#include <common.h> #include <charset.h> +#include <uuid.h>

#include <div64.h> #define noinline __attribute__((noinline)) @@ -366,6 +367,40 @@ static char *ip4_addr_string(char *buf, char *end, u8 *addr, int field_width, } #endif

+#ifdef CONFIG_LIB_UUID +/* + * This works (roughly) the same way as linux's, but we currently always + * print lower-case (ie. we just keep %pUB and %pUL for compat with linux), + * mostly just because that is what uuid_bin_to_str() supports. + * + * %pUb: 01020304-0506-0708-090a-0b0c0d0e0f10 + * %pUl: 04030201-0605-0807-090a-0b0c0d0e0f10 + */ +static char *uuid_string(char *buf, char *end, u8 *addr, int field_width, + int precision, int flags, const char *fmt) +{ + char uuid[UUID_STR_LEN + 1]; + int str_format = UUID_STR_FORMAT_STD; + + switch (*(++fmt)) { + case 'L': + case 'l': + str_format = UUID_STR_FORMAT_GUID; + break; + case 'B': + case 'b': + /* this is the default */ + break; + default: + break; + } + + uuid_bin_to_str(addr, uuid, str_format); + + return string(buf, end, uuid, field_width, precision, flags); +} +#endif + /* * Show a '%p' thing. A kernel extension is that the '%p' is followed * by an extra set of alphanumeric characters that are extended format @@ -399,8 +434,8 @@ static char *pointer(const char *fmt, char *buf, char *end, void *ptr, flags); #endif

-#ifdef CONFIG_CMD_NET switch (*fmt) { +#ifdef CONFIG_CMD_NET case 'a': flags |= SPECIAL | ZEROPAD;

@@ -430,8 +465,15 @@ static char *pointer(const char *fmt, char *buf, char *end, void *ptr, precision, flags); flags &= ~SPECIAL; break; - } #endif +#ifdef CONFIG_LIB_UUID + case 'U': + return uuid_string(buf, end, ptr, field_width, precision, + flags, fmt); +#endif + default: + break; + } flags |= SMALL; if (field_width == -1) { field_width = 2*sizeof(void *);

Solves build issue:

Building current source for 134 boards (12 threads, 1 job per thread) arm: + lsxhl +examples/api/vsprintf.o: In function `string16': +lib/vsprintf.c:278: undefined reference to `memcpy' +examples/api/uuid.o: In function `uuid_bin_to_str': +lib/uuid.c:197: undefined reference to `memcpy' +lib/uuid.c:199: undefined reference to `memcpy' +make[3]: *** [examples/api/demo] Error 1 +make[2]: *** [examples/api] Error 2 +make[1]: *** [examples] Error 2 +make: *** [sub-make] Error 2 133 0 1 /134 sheevaplug

Signed-off-by: Rob Clark robdclark@gmail.com --- examples/api/glue.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)

diff --git a/examples/api/glue.c b/examples/api/glue.c index 8aabf32c89..575c1e55f3 100644 --- a/examples/api/glue.c +++ b/examples/api/glue.c @@ -416,3 +416,15 @@ void ub_display_clear(void) { syscall(API_DISPLAY_CLEAR, NULL); } + +__weak void *memcpy(void *dest, const void *src, size_t size) +{ + unsigned char *dptr = dest; + const unsigned char *ptr = src; + const unsigned char *end = src + size; + + while (ptr < end) + *dptr++ = *ptr++; + + return dest; +}