Fix bug #67705 (extensive backtracking in rule regular expression)

2026-03-24 00:02:20 +01:00 · 2014-08-04 00:01:57 -07:00
parent bff84d7c18
commit eeaec70758
4 changed files with 76 additions and 21 deletions
--- a/4
+++ b/4
@@ -5,6 +5,10 @@ PHP                                                                        NEWS
 - COM:
  . Fixed missing type checks in com_event_sink (Yussuf Khalil, Stas).

+- Fileinfo:
+  . Fixed bug #67705 (extensive backtracking in rule regular expression).
+    (CVE-2014-3538) (Remi)
+
 - Milter:
  . Fixed bug #67715 (php-milter does not build and crashes randomly). (Mike)

--- a/ext/fileinfo/data_file.c
+++ b/ext/fileinfo/data_file.c
@@ -115198,7 +115198,7 @@ const unsigned char php_magic_database[2606480] = {
 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
 0x00, 0x00, 0x40, 0x00, 0x3D, 0x1B, 0x11, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
-0x00, 0x00, 0x00, 0x00, 0x34, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
+0x00, 0x00, 0x00, 0x00, 0x34, 0x00, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
 0x5E, 0x5C, 0x73, 0x7B, 0x30, 0x2C, 0x31, 0x30, 0x30, 0x7D, 0x42, 0x45, 0x47, 0x49, 0x4E, 0x5C, 
 0x73, 0x7B, 0x30, 0x2C, 0x31, 0x30, 0x30, 0x7D, 0x5B, 0x7B, 0x5D, 0x00, 0x00, 0x00, 0x00, 0x00, 
 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 
--- a/ext/fileinfo/libmagic/softmagic.c
+++ b/ext/fileinfo/libmagic/softmagic.c
@@ -58,7 +58,7 @@ private int32_t mprint(struct magic_set *, struct magic *);
 private int32_t moffset(struct magic_set *, struct magic *);
 private void mdebug(uint32_t, const char *, size_t);
 private int mcopy(struct magic_set *, union VALUETYPE *, int, int,
-    const unsigned char *, uint32_t, size_t, size_t);
+    const unsigned char *, uint32_t, size_t, struct magic *);
 private int mconvert(struct magic_set *, struct magic *, int);
 private int print_sep(struct magic_set *, int);
 private int handle_annotation(struct magic_set *, struct magic *);
@@ -1003,7 +1003,7 @@ mdebug(uint32_t offset, const char *str, size_t len)

 private int
 mcopy(struct magic_set *ms, union VALUETYPE *p, int type, int indir,
-    const unsigned char *s, uint32_t offset, size_t nbytes, size_t linecnt)
+    const unsigned char *s, uint32_t offset, size_t nbytes, struct magic *m)
 {
 	/*
 	 * Note: FILE_SEARCH and FILE_REGEX do not actually copy
@@ -1023,15 +1023,24 @@ mcopy(struct magic_set *ms, union VALUETYPE *p, int type, int indir,
 			const char *last;	/* end of search region */
 			const char *buf;	/* start of search region */
 			const char *end;
-			size_t lines;
+			size_t lines, linecnt, bytecnt;

+			linecnt = m->str_range;
+			bytecnt = linecnt * 80;
+
+			if (bytecnt == 0) {
+				bytecnt = 8192;
+			}
+			if (bytecnt > nbytes) {
+				bytecnt = nbytes;
+			}
 			if (s == NULL) {
 				ms->search.s_len = 0;
 				ms->search.s = NULL;
 				return 0;
 			}
 			buf = RCAST(const char *, s) + offset;
-			end = last = RCAST(const char *, s) + nbytes;
+			end = last = RCAST(const char *, s) + bytecnt;
 			/* mget() guarantees buf <= last */
 			for (lines = linecnt, b = buf; lines && b < end &&
 			     ((b = CAST(const char *,
@@ -1044,7 +1053,7 @@ mcopy(struct magic_set *ms, union VALUETYPE *p, int type, int indir,
 					b++;
 			}
 			if (lines)
-				last = RCAST(const char *, s) + nbytes;
+				last = RCAST(const char *, s) + bytecnt;

 			ms->search.s = buf;
 			ms->search.s_len = last - buf;
@@ -1118,7 +1127,6 @@ mget(struct magic_set *ms, const unsigned char *s, struct magic *m,
    int *need_separator, int *returnval)
 {
 	uint32_t soffset, offset = ms->offset;
-	uint32_t count = m->str_range;
 	int rv, oneed_separator;
 	char *sbuf, *rbuf;
 	union VALUETYPE *p = &ms->ms_value;
@@ -1130,13 +1138,12 @@ mget(struct magic_set *ms, const unsigned char *s, struct magic *m,
 	}

 	if (mcopy(ms, p, m->type, m->flag & INDIR, s, (uint32_t)(offset + o),
-	    (uint32_t)nbytes, count) == -1)
+	    (uint32_t)nbytes, m) == -1)
 		return -1;

 	if ((ms->flags & MAGIC_DEBUG) != 0) {
 		fprintf(stderr, "mget(type=%d, flag=%x, offset=%u, o=%zu, "
-		    "nbytes=%zu, count=%u)\n", m->type, m->flag, offset, o,
-		    nbytes, count);
+		    "nbytes=%zu)\n", m->type, m->flag, offset, o, nbytes);
 		mdebug(offset, (char *)(void *)p, sizeof(union VALUETYPE));
 	}

@@ -1627,7 +1634,7 @@ mget(struct magic_set *ms, const unsigned char *s, struct magic *m,
 			if ((ms->flags & MAGIC_DEBUG) != 0)
 				fprintf(stderr, "indirect +offs=%u\n", offset);
 		}
-		if (mcopy(ms, p, m->type, 0, s, offset, nbytes, count) == -1)
+		if (mcopy(ms, p, m->type, 0, s, offset, nbytes, m) == -1)
 			return -1;
 		ms->offset = offset;

@@ -2057,7 +2064,7 @@ magiccheck(struct magic_set *ms, struct magic *m)
 			zval *retval;
 			zval *subpats;
 			char *haystack;
-			
+
 			MAKE_STD_ZVAL(retval);
 			ALLOC_INIT_ZVAL(subpats);
 			
--- a/ext/fileinfo/magicdata.patch
+++ b/ext/fileinfo/magicdata.patch
@@ -1,4 +1,4 @@
-Patches applied to file sources tree before generating magic.mgc
+Patches applied to file 5.14 sources tree before generating magic.mgc
 and before running create_data_file.php to create data_file.c.


@@ -17,14 +17,6 @@ diff --git a/magic/Magdir/commands b/magic/Magdir/commands
 index 67c3eee..4a7d8dd 100644
 --- a/magic/Magdir/commands
 +++ b/magic/Magdir/commands
-@@ -1,6 +1,6 @@
- 
- #------------------------------------------------------------------------------
-# $File: commands,v 1.44 2013/02/05 15:20:47 christos Exp $
-+# $File: commands,v 1.45 2013/02/06 14:18:52 christos Exp $
- # commands:  file(1) magic for various shells and interpreters
- #
- #0	string/w	:			shell archive or script for antique kernel text
@@ -49,7 +49,7 @@
 !:mime	text/x-awk
 0	string/wt	#!\ /usr/bin/awk	awk script text executable
@@ -36,4 +28,56 @@ index 67c3eee..4a7d8dd 100644
 0	string/wt	#!\ /bin/rc	Plan 9 rc shell script text executable
 -- 
 1.8.5.5
+From 0b478f445b6b7540b58af5d1fe583fa9e48fd745 Mon Sep 17 00:00:00 2001
+From: Christos Zoulas <christos@zoulas.com>
+Date: Wed, 28 May 2014 19:52:36 +0000
+Subject: [PATCH] further optimize awk by not looking for the BEGIN regex until
+ we found the BEGIN (Jan Kaluza)
+
+---
+ magic/Magdir/commands | 5 +++--
+ 1 file changed, 3 insertions(+), 2 deletions(-)
+
+diff --git a/magic/Magdir/commands b/magic/Magdir/commands
+index bfffdef..26b2869 100644
+--- a/magic/Magdir/commands
+++ b/magic/Magdir/commands
+@@ -49,7 +49,8 @@
+ !:mime	text/x-awk
+ 0	string/wt	#!\ /usr/bin/awk	awk script text executable
+ !:mime	text/x-awk
+-0	regex		=^\\s{0,100}BEGIN\\s{0,100}[{]	awk script text
+0       search/16384    BEGIN
+>0	regex		=^\\s{0,100}BEGIN\\s{0,100}[{]	awk script text
+ 
+ # AT&T Bell Labs' Plan 9 shell
+ 0	string/wt	#!\ /bin/rc	Plan 9 rc shell script text executable
+-- 
+2.0.3
+
+From 71a8b6c0d758acb0f73e2e51421a711b5e9d6668 Mon Sep 17 00:00:00 2001
+From: Christos Zoulas <christos@zoulas.com>
+Date: Fri, 30 May 2014 16:48:44 +0000
+Subject: [PATCH] Limit regex search for BEGIN to the first 4K of the file.
+
+---
+ magic/Magdir/commands | 5 ++---
+ 1 file changed, 2 insertions(+), 3 deletions(-)
+
+diff --git a/magic/Magdir/commands b/magic/Magdir/commands
+index 26b2869..bcd0f43 100644
+--- a/magic/Magdir/commands
+++ b/magic/Magdir/commands
+@@ -49,8 +49,7 @@
+ !:mime	text/x-awk
+ 0	string/wt	#!\ /usr/bin/awk	awk script text executable
+ !:mime	text/x-awk
+-0       search/16384    BEGIN
+->0	regex		=^\\s{0,100}BEGIN\\s{0,100}[{]	awk script text
+0	regex/4096	=^\\s{0,100}BEGIN\\s{0,100}[{]	awk script text
+ 
+ # AT&T Bell Labs' Plan 9 shell
+ 0	string/wt	#!\ /bin/rc	Plan 9 rc shell script text executable
+-- 
+2.0.3