patches: Add option to disable character maps

Musl's `iconv` includes lots of character sets by default. Provide an
option to disable most of them to save on binary size.

Signed-off-by: Tu Dinh Ngoc <dinhngoc.tu@irit.fr>
Reviewed-by: Sergiu Moga <sergiu.moga@protonmail.com>
Reviewed-by: Eduard Vintilă <eduard.vintila47@gmail.com>
Approved-by: Razvan Deaconescu <razvand@unikraft.io>
Tested-by: Unikraft CI <monkey@unikraft.io>
GitHub-Closes: #40

diff --git a/Config.uk b/Config.uk
index 29a847e59634df57f5957912b5c48f54f28ba744..dc4c26ce23bff7a04b6b2cdf95bffacfc5a6e5b2 100644 (file)
--- a/Config.uk
+++ b/Config.uk
@@ -98,6 +98,11 @@ config LIBMUSL_LOCALE
   bool "liblocale"
   default y
 
+config LIBMUSL_LOCALE_LEGACY
+  bool "Enable legacy locales"
+  depends on LIBMUSL_LOCALE
+  default y
+
 config LIBMUSL_MALLOC
   bool "libmalloc"
   default y

diff --git a/patches/0021-locale-iconv-Add-character-map-option.patch b/patches/0021-locale-iconv-Add-character-map-option.patch
new file mode 100644 (file)
index 0000000..757d695
--- /dev/null
+++ b/patches/0021-locale-iconv-Add-character-map-option.patch
@@ -0,0 +1,119 @@
+From 9bde85765c75459a1a43060a69d8dc39e1eaeba0 Mon Sep 17 00:00:00 2001
+From: Tu Dinh Ngoc <dinhngoc.tu@irit.fr>
+Date: Tue, 4 Apr 2023 20:07:08 +0000
+Subject: [PATCH] locale/iconv: Add character map option
+
+Musl's `iconv` includes lots of character sets by default. Provide an option
+to disable most of them.
+
+Signed-off-by: Tu Dinh Ngoc <dinhngoc.tu@irit.fr>
+---
+ src/locale/codepages.h |  3 ++-
+ src/locale/iconv.c     | 10 ++++++++++
+ 2 files changed, 12 insertions(+), 1 deletion(-)
+
+diff --git a/src/locale/codepages.h b/src/locale/codepages.h
+index 4e236ef..bec76af 100644
+--- a/src/locale/codepages.h
++++ b/src/locale/codepages.h
+@@ -2,6 +2,7 @@
+ "latin1\0"
+ "\0\100"
+
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+ "iso88592\0"
+ "\0\50"
+ "\240\20\364\127\116\244\334\364\324\51\250\124\65\125\126\156\265\42\27\134"
+@@ -317,4 +318,4 @@
+ "\115\70\361\4\24\121\110\221\313\76\374\344\243\317\77\134\334\63\5\25"
+ "\125\130\161\5\26\131\150\41\13\65\326\110\63\115\65\60\304\40\303\14"
+ "\64\324\140\303\15\70\344\60\313\66\334\144\243\315\47"
+-
++#endif
+diff --git a/src/locale/iconv.c b/src/locale/iconv.c
+index 3047c27..0d17ace 100644
+--- a/src/locale/iconv.c
++++ b/src/locale/iconv.c
+@@ -48,6 +48,7 @@ static const unsigned char charmaps[] =
+ "utf16\0\0\312"
+ "ucs4\0utf32\0\0\313"
+ "ucs2\0\0\314"
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+ "eucjp\0\0\320"
+ "shiftjis\0sjis\0\0\321"
+ "iso2022jp\0\0\322"
+@@ -56,6 +57,7 @@ static const unsigned char charmaps[] =
+ "gb2312\0\0\332"
+ "big5\0bigfive\0cp950\0big5hkscs\0\0\340"
+ "euckr\0ksc5601\0ksx1001\0cp949\0\0\350"
++#endif
+ #include "codepages.h"
+ ;
+
+@@ -66,6 +68,7 @@ static const unsigned short legacy_chars[] = {
+ #include "legacychars.h"
+ };
+
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+ static const unsigned short jis0208[84][94] = {
+ #include "jis0208.h"
+ };
+@@ -89,6 +92,7 @@ static const unsigned short ksc[93][94] = {
+ static const unsigned short rev_jis[] = {
+ #include "revjis.h"
+ };
++#endif
+
+ static int fuzzycmp(const unsigned char *a, const unsigned char *b)
+ {
+@@ -205,6 +209,7 @@ static unsigned legacy_map(const unsigned char *map, unsigned c)
+       return x < 256 ? x : legacy_chars[x-256];
+ }
+
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+ static unsigned uni_to_jis(unsigned c)
+ {
+       unsigned nel = sizeof rev_jis / sizeof *rev_jis;
+@@ -223,6 +228,7 @@ static unsigned uni_to_jis(unsigned c)
+               }
+       }
+ }
++#endif
+
+ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restrict out, size_t *restrict outb)
+ {
+@@ -319,6 +325,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
+                       }
+                       type = scd->state;
+                       continue;
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+               case SHIFT_JIS:
+                       if (c < 128) break;
+                       if (c-0xa1 <= 0xdf-0xa1) {
+@@ -518,6 +525,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
+                       c = ksc[c][d];
+                       if (!c) goto ilseq;
+                       break;
++#endif
+               default:
+                       if (!c) break;
+                       c = legacy_map(map, c);
+@@ -559,6 +567,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
+                               }
+                       }
+                       goto subst;
++#if CONFIG_LIBMUSL_LOCALE_LEGACY
+               case SHIFT_JIS:
+                       if (c < 128) goto revout;
+                       if (c == 0xa5) {
+@@ -632,6 +641,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
+                       *(*out)++ = 'B';
+                       *outb -= 8;
+                       break;
++#endif
+               case UCS2:
+                       totype = UCS2BE;
+               case UCS2BE:
+--
+2.34.1
+