0d8b67
commit ee7a3144c9922808181009b7b3e50e852fb4999b
0d8b67
Author: Andreas Schwab <schwab@suse.de>
0d8b67
Date:   Mon Dec 21 08:56:43 2020 +0530
0d8b67
0d8b67
    Fix buffer overrun in EUC-KR conversion module (bz #24973)
0d8b67
    
0d8b67
    The byte 0xfe as input to the EUC-KR conversion denotes a user-defined
0d8b67
    area and is not allowed.  The from_euc_kr function used to skip two bytes
0d8b67
    when told to skip over the unknown designation, potentially running over
0d8b67
    the buffer end.
0d8b67
0d8b67
Conflicts:
0d8b67
	iconvdata/Makefile
0d8b67
0d8b67
Textual conflict in iconvdata/Makefile due to tests differences.
0d8b67
0d8b67
diff --git a/iconvdata/Makefile b/iconvdata/Makefile
0d8b67
index 4ec2741cdc..85009f3390 100644
0d8b67
--- a/iconvdata/Makefile
0d8b67
+++ b/iconvdata/Makefile
0d8b67
@@ -73,7 +73,7 @@ modules.so := $(addsuffix .so, $(modules))
0d8b67
 ifeq (yes,$(build-shared))
0d8b67
 tests = bug-iconv1 bug-iconv2 tst-loading tst-e2big tst-iconv4 bug-iconv4 \
0d8b67
 	tst-iconv6 bug-iconv5 bug-iconv6 tst-iconv7 bug-iconv8 bug-iconv9 \
0d8b67
-	bug-iconv10 bug-iconv11 bug-iconv12
0d8b67
+	bug-iconv10 bug-iconv11 bug-iconv12 bug-iconv13
0d8b67
 ifeq ($(have-thread-library),yes)
0d8b67
 tests += bug-iconv3
0d8b67
 endif
0d8b67
diff --git a/iconvdata/bug-iconv13.c b/iconvdata/bug-iconv13.c
0d8b67
new file mode 100644
0d8b67
index 0000000000..87aaff398e
0d8b67
--- /dev/null
0d8b67
+++ b/iconvdata/bug-iconv13.c
0d8b67
@@ -0,0 +1,53 @@
0d8b67
+/* bug 24973: Test EUC-KR module
0d8b67
+   Copyright (C) 2020 Free Software Foundation, Inc.
0d8b67
+   This file is part of the GNU C Library.
0d8b67
+
0d8b67
+   The GNU C Library is free software; you can redistribute it and/or
0d8b67
+   modify it under the terms of the GNU Lesser General Public
0d8b67
+   License as published by the Free Software Foundation; either
0d8b67
+   version 2.1 of the License, or (at your option) any later version.
0d8b67
+
0d8b67
+   The GNU C Library is distributed in the hope that it will be useful,
0d8b67
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
0d8b67
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
0d8b67
+   Lesser General Public License for more details.
0d8b67
+
0d8b67
+   You should have received a copy of the GNU Lesser General Public
0d8b67
+   License along with the GNU C Library; if not, see
0d8b67
+   <https://www.gnu.org/licenses/>.  */
0d8b67
+
0d8b67
+#include <errno.h>
0d8b67
+#include <iconv.h>
0d8b67
+#include <stdio.h>
0d8b67
+#include <support/check.h>
0d8b67
+
0d8b67
+static int
0d8b67
+do_test (void)
0d8b67
+{
0d8b67
+  iconv_t cd = iconv_open ("UTF-8//IGNORE", "EUC-KR");
0d8b67
+  TEST_VERIFY_EXIT (cd != (iconv_t) -1);
0d8b67
+
0d8b67
+  /* 0xfe (->0x7e : row 94) and 0xc9 (->0x49 : row 41) are user-defined
0d8b67
+     areas, which are not allowed and should be skipped over due to
0d8b67
+     //IGNORE.  The trailing 0xfe also is an incomplete sequence, which
0d8b67
+     should be checked first.  */
0d8b67
+  char input[4] = { '\xc9', '\xa1', '\0', '\xfe' };
0d8b67
+  char *inptr = input;
0d8b67
+  size_t insize = sizeof (input);
0d8b67
+  char output[4];
0d8b67
+  char *outptr = output;
0d8b67
+  size_t outsize = sizeof (output);
0d8b67
+
0d8b67
+  /* This used to crash due to buffer overrun.  */
0d8b67
+  TEST_VERIFY (iconv (cd, &inptr, &insize, &outptr, &outsize) == (size_t) -1);
0d8b67
+  TEST_VERIFY (errno == EINVAL);
0d8b67
+  /* The conversion should produce one character, the converted null
0d8b67
+     character.  */
0d8b67
+  TEST_VERIFY (sizeof (output) - outsize == 1);
0d8b67
+
0d8b67
+  TEST_VERIFY_EXIT (iconv_close (cd) != -1);
0d8b67
+
0d8b67
+  return 0;
0d8b67
+}
0d8b67
+
0d8b67
+#include <support/test-driver.c>
0d8b67
diff --git a/iconvdata/euc-kr.c b/iconvdata/euc-kr.c
0d8b67
index b0d56cf3ee..1045bae926 100644
0d8b67
--- a/iconvdata/euc-kr.c
0d8b67
+++ b/iconvdata/euc-kr.c
0d8b67
@@ -80,11 +80,7 @@ euckr_from_ucs4 (uint32_t ch, unsigned char *cp)
0d8b67
 									      \
0d8b67
     if (ch <= 0x9f)							      \
0d8b67
       ++inptr;								      \
0d8b67
-    /* 0xfe(->0x7e : row 94) and 0xc9(->0x59 : row 41) are		      \
0d8b67
-       user-defined areas.  */						      \
0d8b67
-    else if (__builtin_expect (ch == 0xa0, 0)				      \
0d8b67
-	     || __builtin_expect (ch > 0xfe, 0)				      \
0d8b67
-	     || __builtin_expect (ch == 0xc9, 0))			      \
0d8b67
+    else if (__glibc_unlikely (ch == 0xa0))				      \
0d8b67
       {									      \
0d8b67
 	/* This is illegal.  */						      \
0d8b67
 	STANDARD_FROM_LOOP_ERR_HANDLER (1);				      \
0d8b67
diff --git a/iconvdata/ksc5601.h b/iconvdata/ksc5601.h
0d8b67
index d3eb3a4ff8..f5cdc72797 100644
0d8b67
--- a/iconvdata/ksc5601.h
0d8b67
+++ b/iconvdata/ksc5601.h
0d8b67
@@ -50,15 +50,15 @@ ksc5601_to_ucs4 (const unsigned char **s, size_t avail, unsigned char offset)
0d8b67
   unsigned char ch2;
0d8b67
   int idx;
0d8b67
 
0d8b67
+  if (avail < 2)
0d8b67
+    return 0;
0d8b67
+
0d8b67
   /* row 94(0x7e) and row 41(0x49) are user-defined area in KS C 5601 */
0d8b67
 
0d8b67
   if (ch < offset || (ch - offset) <= 0x20 || (ch - offset) >= 0x7e
0d8b67
       || (ch - offset) == 0x49)
0d8b67
     return __UNKNOWN_10646_CHAR;
0d8b67
 
0d8b67
-  if (avail < 2)
0d8b67
-    return 0;
0d8b67
-
0d8b67
   ch2 = (*s)[1];
0d8b67
   if (ch2 < offset || (ch2 - offset) <= 0x20 || (ch2 - offset) >= 0x7f)
0d8b67
     return __UNKNOWN_10646_CHAR;
0d8b67
-- 
0d8b67
2.29.2
0d8b67