CVE-2017-7186 fix (upstream rev 670)

Fix 32-bit non-UTF property test crash.

diff --git a/src/pcre2_internal.h b/src/pcre2_internal.h
index 56908708aa13b20a9518946dbb33600b6ef06fe6..7fb374d63e3a60e3f0b9a753babae7817b912821 100644 (file)
--- a/src/pcre2_internal.h
+++ b/src/pcre2_internal.h
@@ -1794,10 +1794,17 @@ typedef struct {
 /* UCD access macros */
 
 #define UCD_BLOCK_SIZE 128
-#define GET_UCD(ch) (PRIV(ucd_records) + \
+#define REAL_GET_UCD(ch) (PRIV(ucd_records) + \
         PRIV(ucd_stage2)[PRIV(ucd_stage1)[(int)(ch) / UCD_BLOCK_SIZE] * \
         UCD_BLOCK_SIZE + (int)(ch) % UCD_BLOCK_SIZE])
 
+#if PCRE2_CODE_UNIT_WIDTH == 32
+#define GET_UCD(ch) ((ch > MAX_UTF_CODE_POINT)? \
+  PRIV(dummy_ucd_record) : REAL_GET_UCD(ch))
+#else
+#define GET_UCD(ch) REAL_GET_UCD(ch)
+#endif
+
 #define UCD_CHARTYPE(ch)    GET_UCD(ch)->chartype
 #define UCD_SCRIPT(ch)      GET_UCD(ch)->script
 #define UCD_CATEGORY(ch)    PRIV(ucp_gentype)[UCD_CHARTYPE(ch)]
@@ -1854,6 +1861,9 @@ extern const uint8_t          PRIV(utf8_table4)[];
 #define _pcre2_default_compile_context PCRE2_SUFFIX(_pcre2_default_compile_context_)
 #define _pcre2_default_match_context   PCRE2_SUFFIX(_pcre2_default_match_context_)
 #define _pcre2_default_tables          PCRE2_SUFFIX(_pcre2_default_tables_)
+#if PCRE2_CODE_UNIT_WIDTH == 32
+#define _pcre2_dummy_ucd_record        PCRE2_SUFFIX(_pcre2_dummy_ucd_record_)
+#endif
 #define _pcre2_hspace_list             PCRE2_SUFFIX(_pcre2_hspace_list_)
 #define _pcre2_vspace_list             PCRE2_SUFFIX(_pcre2_vspace_list_)
 #define _pcre2_ucd_caseless_sets       PCRE2_SUFFIX(_pcre2_ucd_caseless_sets_)
@@ -1878,6 +1888,9 @@ extern const uint32_t                  PRIV(hspace_list)[];
 extern const uint32_t                  PRIV(vspace_list)[];
 extern const uint32_t                  PRIV(ucd_caseless_sets)[];
 extern const ucd_record                PRIV(ucd_records)[];
+#if PCRE2_CODE_UNIT_WIDTH == 32
+extern const ucd_record                PRIV(dummy_ucd_record)[];
+#endif
 extern const uint8_t                   PRIV(ucd_stage1)[];
 extern const uint16_t                  PRIV(ucd_stage2)[];
 extern const uint32_t                  PRIV(ucp_gbtable)[];

diff --git a/src/pcre2_ucd.c b/src/pcre2_ucd.c
index 116f537b387a10a000e9b48138379407410f687c..56aa29dbe153adb1331805b1971373615a015dc8 100644 (file)
--- a/src/pcre2_ucd.c
+++ b/src/pcre2_ucd.c
@@ -41,6 +41,20 @@ const uint32_t PRIV(ucd_caseless_sets)[] = {0};
 
 const char *PRIV(unicode_version) = "8.0.0";
 
+/* If the 32-bit library is run in non-32-bit mode, character values
+greater than 0x10ffff may be encountered. For these we set up a
+special record. */
+
+#if PCRE2_CODE_UNIT_WIDTH == 32
+const ucd_record PRIV(dummy_ucd_record)[] = {{
+  ucp_Common,    /* script */
+  ucp_Cn,        /* type unassigned */
+  ucp_gbOther,   /* grapheme break property */
+  0,             /* case set */
+  0,             /* other case */
+  }};
+#endif
+
 /* When recompiling tables with a new Unicode version, please check the
 types in this structure definition from pcre2_internal.h (the actual
 field names will be different):

diff --git a/testdata/testinput12 b/testdata/testinput12
index 14a7715151ae3ae15ca54f896f6ab65ae5cb143a..86f2c7447af516688c75bc3291e27112cc23ec01 100644 (file)
--- a/testdata/testinput12
+++ b/testdata/testinput12
@@ -343,4 +343,7 @@
 /./utf
     \x{110000}
 
+/\pP/ucp
+    \x{7fffffff}\=no_jit
+
 # End of testinput12

diff --git a/testdata/testoutput12-16 b/testdata/testoutput12-16
index 383a032fa4a57ce61995bfda4d73ff557d84d31c..b19f6c48318a336fd13a345a1fdf8df5a1a73038 100644 (file)
--- a/testdata/testoutput12-16
+++ b/testdata/testoutput12-16
@@ -1367,4 +1367,10 @@ Subject length lower bound = 2
     \x{110000}
 ** Failed: character \x{110000} is greater than 0x10ffff and so cannot be converted to UTF-16
 
+/\pP/ucp
+    \x{7fffffff}\=no_jit
+** Character \x{7fffffff} is greater than 0xffff and UTF-16 mode is not enabled.
+** Truncation will probably give the wrong result.
+No match
+
 # End of testinput12

diff --git a/testdata/testoutput12-32 b/testdata/testoutput12-32
index 95f1834049bd1807b97be97e34a5b333a0aed8e2..220f4f1485c54e7ebac333ed604c1e53fac8f58a 100644 (file)
--- a/testdata/testoutput12-32
+++ b/testdata/testoutput12-32
@@ -1361,4 +1361,8 @@ Subject length lower bound = 2
     \x{110000}
 Failed: error -28: UTF-32 error: code points greater than 0x10ffff are not defined at offset 0
 
+/\pP/ucp
+    \x{7fffffff}\=no_jit
+No match
+
 # End of testinput12