rpms/pcre2
6
0
Fork 0
Code Issues Pull Requests Releases Activity

10.30-RC1 bump

Browse Source
This commit is contained in:
Petr Písař 2017-07-20 10:46:05 +02:00
parent 0099a92309
commit 054cc8376c
22 changed files with 604 additions and 1993 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
.gitignore vendored
View File

@ -7,3 +7,4 @@
/pcre2-10.22.tar.bz2
/pcre2-10.23-RC1.tar.bz2
/pcre2-10.23.tar.bz2
/pcre2-10.30-RC1.tar.bz2

105
pcre2-10.23-Check-malloc-returns-in-pcre2test.patch
View File

@ -1,105 +0,0 @@
From 86c1fb487cd84b1971d4446e58ddb1602c95fdc3 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Thu, 23 Feb 2017 17:05:43 +0000
Subject: [PATCH] Check malloc returns in pcre2test.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@669 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
---
ChangeLog | 9 +++++++++
configure.ac | 6 +++---
src/pcre2.h | 6 +++---
src/pcre2test.c | 22 +++++++++++++++++++---
testdata/testinput2 | 3 +++
testdata/testoutput2 | 4 ++++
6 files changed, 41 insertions(+), 9 deletions(-)
diff --git a/src/pcre2test.c b/src/pcre2test.c
index 241c22c..01457e8 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -1365,8 +1365,7 @@ are supported. */
(test_mode == PCRE8_MODE && G(x,8)->f r (y)) || \
(test_mode == PCRE16_MODE && G(x,16)->f r (y)) || \
(test_mode == PCRE32_MODE && G(x,32)->f r (y)))
-
-
+
/* ----- Two out of three modes are supported ----- */
@@ -1775,7 +1774,7 @@ the three different cases. */
#define TESTFLD(x,f,r,y) ( \
(test_mode == G(G(PCRE,BITONE),_MODE) && G(x,BITONE)->f r (y)) || \
(test_mode == G(G(PCRE,BITTWO),_MODE) && G(x,BITTWO)->f r (y)))
-
+
#endif /* Two out of three modes */
@@ -6169,7 +6168,16 @@ if ((pat_patctl.control & CTL_POSIX) != 0)
if (msg[0] == 0) fprintf(outfile, "\n");
if (dat_datctl.oveccount > 0)
+ {
pmatch = (regmatch_t *)malloc(sizeof(regmatch_t) * dat_datctl.oveccount);
+ if (pmatch == NULL)
+ {
+ fprintf(outfile, "** Failed to get memory for recording matching "
+ "information (size set = %du)\n", dat_datctl.oveccount);
+ return PR_OK;
+ }
+ }
+
if ((dat_datctl.options & PCRE2_NOTBOL) != 0) eflags |= REG_NOTBOL;
if ((dat_datctl.options & PCRE2_NOTEOL) != 0) eflags |= REG_NOTEOL;
if ((dat_datctl.options & PCRE2_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
@@ -6305,6 +6313,14 @@ else
PCRE2_MATCH_DATA_CREATE(match_data, max_oveccount, NULL);
}
+if (CASTVAR(void *, match_data) == NULL)
+ {
+ fprintf(outfile, "** Failed to get memory for recording matching "
+ "information (size requested: %d)\n", dat_datctl.oveccount);
+ max_oveccount = 0;
+ return PR_OK;
+ }
+
/* Replacement processing is ignored for DFA matching. */
if (dat_datctl.replacement[0] != 0 && (dat_datctl.control & CTL_DFA) != 0)
diff --git a/testdata/testinput2 b/testdata/testinput2
index a700d5a..921d83c 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -4969,4 +4969,7 @@ a)"xI
/(?<a>a)(?<b>b)\g{b}\g{a}\g{a}\g{a}\g{a}(?<a>xx)(?<b>zz)/I,dupnames
+//
+ \=ovector=7777777777
+
# End of testinput2
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index ce8c667..9862ea3 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -15479,6 +15479,10 @@ First code unit = 'a'
Last code unit = 'z'
Subject length lower bound = 11
+//
+ \=ovector=7777777777
+** Invalid value in 'ovector=7777777777'
+
# End of testinput2
Error -63: PCRE2_ERROR_BADDATA (unknown error number)
Error -62: bad serialized data
--
2.7.4

51
pcre2-10.23-Close-serialization-file-in-pcre2test-after-any-erro.patch
View File

@ -1,51 +0,0 @@
From 23f3fd65eb57bdaa03701177bbd2043a08a3299e Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Tue, 21 Mar 2017 16:18:54 +0000
Subject: [PATCH] Close serialization file in pcre2test after any error.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@692 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.32.
diff --git a/src/pcre2test.c b/src/pcre2test.c
index 39eedbe..b6d3ec9 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -4456,6 +4456,7 @@ switch(cmd)
if (rc < 0)
{
serial_error(rc, "Serialization");
+ fclose(f);
break;
}
@@ -4469,6 +4470,7 @@ switch(cmd)
if (fwrite(serial, 1, serial_size, f) != serial_size)
{
fprintf(outfile, "** Wrong return from fwrite()\n");
+ fclose(f);
return PR_ABEND;
}
@@ -4496,6 +4498,7 @@ switch(cmd)
{
fprintf(outfile, "** Failed to get memory (size %lu) for #load\n",
(unsigned long int)serial_size);
+ fclose(f);
return PR_ABEND;
}
@@ -4503,6 +4506,7 @@ switch(cmd)
{
fprintf(outfile, "** Wrong return from fread()\n");
free(serial);
+ fclose(f);
return PR_ABEND;
}
fclose(f);
--
2.7.4

55
pcre2-10.23-Correct-an-incorrect-cast.patch
View File

@ -1,55 +0,0 @@
From 8ac12aa2c80b1a5eec8fdafa66611c8c7cde5af7 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Fri, 14 Apr 2017 12:14:41 +0000
Subject: [PATCH] Correct an incorrect cast.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@748 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
---
src/pcre2_valid_utf.c | 8 ++++----
2 files changed, 6 insertions(+), 4 deletions(-)
diff --git a/src/pcre2_valid_utf.c b/src/pcre2_valid_utf.c
index 3e18f12..96e8bff 100644
--- a/src/pcre2_valid_utf.c
+++ b/src/pcre2_valid_utf.c
@@ -7,7 +7,7 @@ and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Original API code Copyright (c) 1997-2012 University of Cambridge
- New API code Copyright (c) 2016 University of Cambridge
+ New API code Copyright (c) 2016-2017 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@@ -142,20 +142,20 @@ for (p = string; length > 0; p++)
if (c < 0xc0) /* Isolated 10xx xxxx byte */
{
- *erroroffset = (int)(p - string);
+ *erroroffset = (PCRE2_SIZE)(p - string);
return PCRE2_ERROR_UTF8_ERR20;
}
if (c >= 0xfe) /* Invalid 0xfe or 0xff bytes */
{
- *erroroffset = (int)(p - string);
+ *erroroffset = (PCRE2_SIZE)(p - string);
return PCRE2_ERROR_UTF8_ERR21;
}
ab = PRIV(utf8_table4)[c & 0x3f]; /* Number of additional bytes (1-5) */
if (length < ab) /* Missing bytes */
{
- *erroroffset = (int)(p - string);
+ *erroroffset = (PCRE2_SIZE)(p - string);
switch(ab - length)
{
case 1: return PCRE2_ERROR_UTF8_ERR1;
--
2.7.4

122
pcre2-10.23-Fix-32-bit-error-buffer-size-bug-in-pcre2test-Bugzil.patch
View File

@ -1,122 +0,0 @@
From 85f8f579ac7aa268290ed7bd3aca90d927a9b7e9 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Tue, 21 Mar 2017 17:46:21 +0000
Subject: [PATCH 1/2] Fix 32-bit error buffer size bug in pcre2test (Bugzilla
2079).
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@696 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
---
doc/pcre2api.3 | 10 +++++-----
src/pcre2_error.c | 2 +-
src/pcre2test.c | 7 ++++---
testdata/testinput2 | 2 ++
testdata/testoutput2 | 3 +++
diff --git a/doc/pcre2api.3 b/doc/pcre2api.3
index e0a434a..0a3d2ee 100644
--- a/doc/pcre2api.3
+++ b/doc/pcre2api.3
@@ -1,4 +1,4 @@
-.TH PCRE2API 3 "24 December 2016" "PCRE2 10.23"
+.TH PCRE2API 3 "21 March 2017" "PCRE2 10.30"
.SH NAME
PCRE2 - Perl-compatible regular expressions (revised API)
.sp
@@ -2633,8 +2633,8 @@ The internal recursion limit was reached.
A text message for an error code from any PCRE2 function (compile, match, or
auxiliary) can be obtained by calling \fBpcre2_get_error_message()\fP. The code
is passed as the first argument, with the remaining two arguments specifying a
-code unit buffer and its length, into which the text message is placed. Note
-that the message is returned in code units of the appropriate width for the
+code unit buffer and its length in code units, into which the text message is
+placed. The message is returned in code units of the appropriate width for the
library that is being used.
.P
The returned message is terminated with a trailing zero, and the function
@@ -3321,6 +3321,6 @@ Cambridge, England.
.rs
.sp
.nf
-Last updated: 23 December 2016
-Copyright (c) 1997-2016 University of Cambridge.
+Last updated: 21 March 2017
+Copyright (c) 1997-2017 University of Cambridge.
.fi
diff --git a/src/pcre2_error.c b/src/pcre2_error.c
index 9eab4fc..e25c7e6 100644
--- a/src/pcre2_error.c
+++ b/src/pcre2_error.c
@@ -271,7 +271,7 @@ distinct.
Arguments:
enumber error number
buffer where to put the message (zero terminated)
- size size of the buffer
+ size size of the buffer in code units
Returns: length of message if all is well
negative on error
diff --git a/src/pcre2test.c b/src/pcre2test.c
index b6d3ec9..9289656 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -2889,7 +2889,7 @@ if (pbuffer32_size < 4*len + 4)
{
if (pbuffer32 != NULL) free(pbuffer32);
pbuffer32_size = 4*len + 4;
- if (pbuffer32_size < 256) pbuffer32_size = 256;
+ if (pbuffer32_size < 512) pbuffer32_size = 512;
pbuffer32 = (uint32_t *)malloc(pbuffer32_size);
if (pbuffer32 == NULL)
{
@@ -7600,7 +7600,8 @@ if (arg_error != NULL)
int errcode;
char *endptr;
-/* Ensure the relevant non-8-bit buffer is available. */
+/* Ensure the relevant non-8-bit buffer is available. Ensure that it is at
+least 128 code units, because it is used for retrieving error messages. */
#ifdef SUPPORT_PCRE2_16
if (test_mode == PCRE16_MODE)
@@ -7620,7 +7621,7 @@ if (arg_error != NULL)
#ifdef SUPPORT_PCRE2_32
if (test_mode == PCRE32_MODE)
{
- pbuffer32_size = 256;
+ pbuffer32_size = 512;
pbuffer32 = (uint32_t *)malloc(pbuffer32_size);
if (pbuffer32 == NULL)
{
diff --git a/testdata/testinput2 b/testdata/testinput2
index 5a77e88..d62e975 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -5017,4 +5017,6 @@ a)"xI
/(?<!\1((?U)1((?U))))(*F)/never_backslash_c,alt_bsux,anchored,extended
+/\g{3/
+
# End of testinput2
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index 3288424..ebf5fbd 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -15570,6 +15570,9 @@ No match
/(?<!\1((?U)1((?U))))(*F)/never_backslash_c,alt_bsux,anchored,extended
+/\g{3/
+Failed: error 157 at offset 2: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number
+
# End of testinput2
Error -63: PCRE2_ERROR_BADDATA (unknown error number)
Error -62: bad serialized data
--
2.7.4

123
pcre2-10.23-Fix-32-bit-non-UTF-property-test-crash.patch
View File

@ -1,123 +0,0 @@
From be14c2f3f01dc2b6b574b3a86536ee8194945f7f Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Fri, 24 Feb 2017 18:25:32 +0000
Subject: [PATCH] Fix 32-bit non-UTF property test crash.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@670 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
diff --git a/src/pcre2_internal.h b/src/pcre2_internal.h
index 6a8774c..720bbc9 100644
--- a/src/pcre2_internal.h
+++ b/src/pcre2_internal.h
@@ -1774,10 +1774,17 @@ 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)]
@@ -1834,6 +1841,9 @@ 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_)
@@ -1858,6 +1868,9 @@ 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 116f537..56aa29d 100644
--- 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 c3b2bfc..decfe82 100644
--- a/testdata/testinput12
+++ b/testdata/testinput12
@@ -360,4 +360,7 @@
/[\s[:^ascii:]]/B,ucp
+/\pP/ucp
+ \x{7fffffff}\=no_jit
+
# End of testinput12
diff --git a/testdata/testoutput12-16 b/testdata/testoutput12-16
index 3b5a0cd..41e0a48 100644
--- a/testdata/testoutput12-16
+++ b/testdata/testoutput12-16
@@ -1415,4 +1415,10 @@ No match
End
------------------------------------------------------------------
+/\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 1496159..e9130b9 100644
--- a/testdata/testoutput12-32
+++ b/testdata/testoutput12-32
@@ -1409,4 +1409,8 @@ No match
End
------------------------------------------------------------------
+/\pP/ucp
+ \x{7fffffff}\=no_jit
+No match
+
# End of testinput12
--
2.7.4

37
pcre2-10.23-Fix-NULL-deference-if-pcre2_callout_enumerate-is-cal.patch
View File

@ -1,37 +0,0 @@
From d5790ad94d486405ecaac3d8f5796f642e57bf06 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Tue, 21 Mar 2017 16:31:54 +0000
Subject: [PATCH] Fix NULL deference if pcre2_callout_enumerate() is called
with a NULL pattern when Unicode support is available (Bugzilla 2076).
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@694 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
diff --git a/src/pcre2_pattern_info.c b/src/pcre2_pattern_info.c
index 0392a0b..5dd8ac5 100644
--- a/src/pcre2_pattern_info.c
+++ b/src/pcre2_pattern_info.c
@@ -261,11 +261,15 @@ pcre2_real_code *re = (pcre2_real_code *)code;
pcre2_callout_enumerate_block cb;
PCRE2_SPTR cc;
#ifdef SUPPORT_UNICODE
-BOOL utf = (re->overall_options & PCRE2_UTF) != 0;
+BOOL utf;
#endif
if (re == NULL) return PCRE2_ERROR_NULL;
+#ifdef SUPPORT_UNICODE
+utf = (re->overall_options & PCRE2_UTF) != 0;
+#endif
+
/* Check that the first field in the block is the magic number. If it is not,
return with PCRE2_ERROR_BADMAGIC. */
--
2.7.4

79
pcre2-10.23-Fix-bug-introduced-at-10.21-use-memory-allocator-fro.patch
View File

@ -1,79 +0,0 @@
From 51b522c0120d061f54317f8a1cede05328c97aca Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat, 25 Mar 2017 15:19:49 +0000
Subject: [PATCH] Fix bug introduced at 10.21: use memory allocator from the
pattern if no context is supplied to pcre2_match().
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Ported to 10.23:
commit 1b7e16d584f0560dd84d2a3260cf08e1a86cccc1
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat Mar 25 15:19:49 2017 +0000
Fix bug introduced at 10.21: use memory allocator from the pattern if no
context is supplied to pcre2_match().
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@707 6239d852-aaf2-0410-a92c-
79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2_match.c | 18 ++++++++++--------
1 file changed, 10 insertions(+), 8 deletions(-)
diff --git a/src/pcre2_match.c b/src/pcre2_match.c
index 78a9bac..52bb843 100644
--- a/src/pcre2_match.c
+++ b/src/pcre2_match.c
@@ -6518,11 +6518,6 @@ options |= (re->flags & FF) / ((FF & (~FF+1)) / (OO & (~OO+1)));
#undef FF
#undef OO
-/* A NULL match context means "use a default context" */
-
-if (mcontext == NULL)
- mcontext = (pcre2_match_context *)(&PRIV(default_match_context));
-
/* These two settings are used in the code for checking a UTF string that
follows immediately afterwards. Other values in the mb block are used only
during interpretive pcre_match() processing, not when the JIT support is in
@@ -6590,7 +6585,7 @@ if (utf && (options & PCRE2_NO_UTF_CHECK) == 0)
/* It is an error to set an offset limit without setting the flag at compile
time. */
-if (mcontext->offset_limit != PCRE2_UNSET &&
+if (mcontext != NULL && mcontext->offset_limit != PCRE2_UNSET &&
(re->overall_options & PCRE2_USE_OFFSET_LIMIT) == 0)
return PCRE2_ERROR_BADOFFSETLIMIT;
@@ -6609,7 +6604,15 @@ if (re->executable_jit != NULL && (options & ~PUBLIC_JIT_MATCH_OPTIONS) == 0)
}
#endif
-/* Carry on with non-JIT matching. */
+/* Carry on with non-JIT matching. A NULL match context means "use a default
+context", but we take the memory control functions from the pattern. */
+
+if (mcontext == NULL)
+ {
+ mcontext = (pcre2_match_context *)(&PRIV(default_match_context));
+ mb->memctl = re->memctl;
+ }
+else mb->memctl = mcontext->memctl;
anchored = ((re->overall_options | options) & PCRE2_ANCHORED) != 0;
firstline = (re->overall_options & PCRE2_FIRSTLINE) != 0;
@@ -6621,7 +6624,6 @@ bumpalong_limit = (mcontext->offset_limit == PCRE2_UNSET)?
mb->callout = mcontext->callout;
mb->callout_data = mcontext->callout_data;
-mb->memctl = mcontext->memctl;
#ifdef HEAP_MATCH_RECURSE
mb->stack_memctl = mcontext->stack_memctl;
#endif
--
2.7.4

119
pcre2-10.23-Fix-character-type-detection-when-32-bit-and-UCP-are.patch
View File

@ -1,119 +0,0 @@
From 749c2eb6994ee26c1fcf1eb7b57faf5c84e9878b Mon Sep 17 00:00:00 2001
From: zherczeg <zherczeg@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Fri, 31 Mar 2017 05:40:37 +0000
Subject: [PATCH] Fix character type detection when 32-bit and UCP are enabled
but UTF is not in JIT.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@720 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2_jit_compile.c | 31 +++++++++++++++++++++++++++++++
testdata/testinput12 | 2 +-
testdata/testoutput12-16 | 2 +-
testdata/testoutput12-32 | 2 +-
4 files changed, 34 insertions(+), 3 deletions(-)
diff --git a/src/pcre2_jit_compile.c b/src/pcre2_jit_compile.c
index 846510a..e93143d 100644
--- a/src/pcre2_jit_compile.c
+++ b/src/pcre2_jit_compile.c
@@ -588,6 +588,8 @@ the start pointers when the end of the capturing group has not yet reached. */
#define READ_CHAR_MAX 0x7fffffff
+#define INVALID_UTF_CHAR 888
+
static PCRE2_SPTR bracketend(PCRE2_SPTR cc)
{
SLJIT_ASSERT((*cc >= OP_ASSERT && *cc <= OP_ASSERTBACK_NOT) || (*cc >= OP_ONCE && *cc <= OP_SCOND));
@@ -3558,10 +3560,30 @@ static void do_getucd(compiler_common *common)
/* Search the UCD record for the character comes in TMP1.
Returns chartype in TMP1 and UCD offset in TMP2. */
DEFINE_COMPILER;
+#if PCRE2_CODE_UNIT_WIDTH == 32
+struct sljit_jump *jump;
+#endif
+
+#if defined SLJIT_DEBUG && SLJIT_DEBUG
+/* dummy_ucd_record */
+const ucd_record *record = GET_UCD(INVALID_UTF_CHAR);
+SLJIT_ASSERT(record->script == ucp_Common && record->chartype == ucp_Cn && record->gbprop == ucp_gbOther);
+SLJIT_ASSERT(record->caseset == 0 && record->other_case == 0);
+#endif
SLJIT_ASSERT(UCD_BLOCK_SIZE == 128 && sizeof(ucd_record) == 8);
sljit_emit_fast_enter(compiler, RETURN_ADDR, 0);
+
+#if PCRE2_CODE_UNIT_WIDTH == 32
+if (!common->utf)
+ {
+ jump = CMP(SLJIT_LESS, TMP1, 0, SLJIT_IMM, MAX_UTF_CODE_POINT + 1);
+ OP1(SLJIT_MOV, TMP1, 0, SLJIT_IMM, INVALID_UTF_CHAR);
+ JUMPHERE(jump);
+ }
+#endif
+
OP2(SLJIT_LSHR, TMP2, 0, TMP1, 0, SLJIT_IMM, UCD_BLOCK_SHIFT);
OP1(SLJIT_MOV_U8, TMP2, 0, SLJIT_MEM1(TMP2), (sljit_sw)PRIV(ucd_stage1));
OP2(SLJIT_AND, TMP1, 0, TMP1, 0, SLJIT_IMM, UCD_BLOCK_MASK);
@@ -5969,6 +5991,15 @@ if (needstype || needsscript)
if (needschar && !charsaved)
OP1(SLJIT_MOV, RETURN_ADDR, 0, TMP1, 0);
+#if PCRE2_CODE_UNIT_WIDTH == 32
+ if (!common->utf)
+ {
+ jump = CMP(SLJIT_LESS, TMP1, 0, SLJIT_IMM, MAX_UTF_CODE_POINT + 1);
+ OP1(SLJIT_MOV, TMP1, 0, SLJIT_IMM, INVALID_UTF_CHAR);
+ JUMPHERE(jump);
+ }
+#endif
+
OP2(SLJIT_LSHR, TMP2, 0, TMP1, 0, SLJIT_IMM, UCD_BLOCK_SHIFT);
OP1(SLJIT_MOV_U8, TMP2, 0, SLJIT_MEM1(TMP2), (sljit_sw)PRIV(ucd_stage1));
OP2(SLJIT_AND, TMP1, 0, TMP1, 0, SLJIT_IMM, UCD_BLOCK_MASK);
diff --git a/testdata/testinput12 b/testdata/testinput12
index decfe82..cca5dfa 100644
--- a/testdata/testinput12
+++ b/testdata/testinput12
@@ -361,6 +361,6 @@
/[\s[:^ascii:]]/B,ucp
/\pP/ucp
- \x{7fffffff}\=no_jit
+ \x{7fffffff}
# End of testinput12
diff --git a/testdata/testoutput12-16 b/testdata/testoutput12-16
index 41e0a48..33b8a33 100644
--- a/testdata/testoutput12-16
+++ b/testdata/testoutput12-16
@@ -1416,7 +1416,7 @@ No match
------------------------------------------------------------------
/\pP/ucp
- \x{7fffffff}\=no_jit
+ \x{7fffffff}
** Character \x{7fffffff} is greater than 0xffff and UTF-16 mode is not enabled.
** Truncation will probably give the wrong result.
No match
diff --git a/testdata/testoutput12-32 b/testdata/testoutput12-32
index e9130b9..1abeb59 100644
--- a/testdata/testoutput12-32
+++ b/testdata/testoutput12-32
@@ -1410,7 +1410,7 @@ No match
------------------------------------------------------------------
/\pP/ucp
- \x{7fffffff}\=no_jit
+ \x{7fffffff}
No match
# End of testinput12
--
2.7.4

187
pcre2-10.23-Fix-crash-for-forward-reference-in-lookbehind-with-P.patch
View File

@ -1,187 +0,0 @@
From 9db78c678d82cc166f77b007e4834354ecf690bd Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Thu, 16 Mar 2017 17:17:47 +0000
Subject: [PATCH] Fix crash for forward reference in lookbehind with
PCRE2_ANCHORED. Fixes oss-fuzz issue 865.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Ported to 10.23:
commit a74ea02e5006bae4484b508b04a010e90c4dfdc2
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Thu Mar 16 17:17:47 2017 +0000
Fix crash for forward reference in lookbehind with PCRE2_ANCHORED. Fixes
oss-fuzz issue 865.
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@681 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2_compile.c | 52 +++++++++++++++++++++++++++++++++++-----------------
testdata/testinput2 | 2 ++
testdata/testoutput2 | 2 ++
3 files changed, 39 insertions(+), 17 deletions(-)
diff --git a/src/pcre2_compile.c b/src/pcre2_compile.c
index 6d98a68..26831d7 100644
--- a/src/pcre2_compile.c
+++ b/src/pcre2_compile.c
@@ -8049,6 +8049,10 @@ the end of the branch, it is called to skip over an internal lookaround, and it
is also called to skip to the end of a class, during which it will never
encounter nested groups (but there's no need to have special code for that).
+When called to find the end of a branch or group, pptr must point to the first
+meta code inside the branch, not the branch-starting code. In other cases it
+can point to the item that causes the function to be called.
+
Arguments:
pptr current pointer to skip from
skiptype PSKIP_CLASS when skipping to end of class
@@ -8065,10 +8069,10 @@ parsed_skip(uint32_t *pptr, uint32_t skiptype)
{
uint32_t nestlevel = 0;
-for (pptr += 1;; pptr++)
+for (;; pptr++)
{
uint32_t meta = META_CODE(*pptr);
-
+
switch(meta)
{
default: /* Just skip over most items */
@@ -8160,11 +8164,12 @@ return pptr;
/* This is called for nested groups within a branch of a lookbehind whose
length is being computed. If all the branches in the nested group have the same
length, that is OK. On entry, the pointer must be at the first element after
-the group initializing code. Caching is used to improve processing speed when
-the same capturing group occurs many times.
+the group initializing code. On exit it points to OP_KET. Caching is used to
+improve processing speed when the same capturing group occurs many times.
Arguments:
pptrptr pointer to pointer in the parsed pattern
+ isinline FALSE if a reference or recursion; TRUE for inline group
errcodeptr pointer to the errorcode
lcptr pointer to the loop counter
group number of captured group or -1 for a non-capturing group
@@ -8175,27 +8180,29 @@ Returns: the group length or a negative number
*/
static int
-get_grouplength(uint32_t **pptrptr, int *errcodeptr, int *lcptr,
+get_grouplength(uint32_t **pptrptr, BOOL isinline, int *errcodeptr, int *lcptr,
int group, parsed_recurse_check *recurses, compile_block *cb)
{
int branchlength;
int grouplength = -1;
/* The cache can be used only if there is no possibility of there being two
-groups with the same number. */
+groups with the same number. We do not need to set the end pointer for a group
+that is being processed as a back reference or recursion, but we must do so for
+an inline group. */
-if (group > 0)
+if (group > 0 && (cb->external_flags & PCRE2_DUPCAPUSED) == 0)
{
uint32_t groupinfo = cb->groupinfo[group];
- if ((cb->external_flags & PCRE2_DUPCAPUSED) == 0)
+ if ((groupinfo & GI_NOT_FIXED_LENGTH) != 0) return -1;
+ if ((groupinfo & GI_SET_FIXED_LENGTH) != 0)
{
- if ((groupinfo & GI_NOT_FIXED_LENGTH) != 0) return -1;
- if ((groupinfo & GI_SET_FIXED_LENGTH) != 0)
- return groupinfo & GI_FIXED_LENGTH_MASK;
+ if (isinline) *pptrptr = parsed_skip(*pptrptr, PSKIP_KET);
+ return groupinfo & GI_FIXED_LENGTH_MASK;
}
}
-/* Scan the group */
+/* Scan the group. In this case we find the end pointer of necessity. */
for(;;)
{
@@ -8353,11 +8360,12 @@ for (;; pptr++)
}
break;
- /* Lookaheads can be ignored. */
+ /* Lookaheads can be ignored, but we must start the skip inside the group
+ so that it isn't treated as a group within the branch. */
case META_LOOKAHEAD:
case META_LOOKAHEADNOT:
- pptr = parsed_skip(pptr, PSKIP_KET);
+ pptr = parsed_skip(pptr + 1, PSKIP_KET);
if (pptr == NULL) goto PARSED_SKIP_FAILED;
break;
@@ -8455,15 +8463,24 @@ for (;; pptr++)
else if (*gptr == (META_CAPTURE | group)) break;
}
- gptrend = parsed_skip(gptr, PSKIP_KET);
+ /* We must start the search for the end of the group at the first meta code
+ inside the group. Otherwise it will be treated as an enclosed group. */
+
+ gptrend = parsed_skip(gptr + 1, PSKIP_KET);
if (gptrend == NULL) goto PARSED_SKIP_FAILED;
if (pptr > gptr && pptr < gptrend) goto ISNOTFIXED; /* Local recursion */
for (r = recurses; r != NULL; r = r->prev) if (r->groupptr == gptr) break;
if (r != NULL) goto ISNOTFIXED; /* Mutual recursion */
this_recurse.prev = recurses;
this_recurse.groupptr = gptr;
+
+ /* We do not need to know the position of the end of the group, that is,
+ gptr is not used after the call to get_grouplength(). Setting the second
+ argument FALSE stops it scanning for the end when the length can be found
+ in the cache. */
+
gptr++;
- grouplength = get_grouplength(&gptr, errcodeptr, lcptr, group,
+ grouplength = get_grouplength(&gptr, FALSE, errcodeptr, lcptr, group,
&this_recurse, cb);
if (grouplength < 0)
{
@@ -8500,7 +8517,8 @@ for (;; pptr++)
case META_NOCAPTURE:
pptr++;
CHECK_GROUP:
- grouplength = get_grouplength(&pptr, errcodeptr, lcptr, group, recurses, cb);
+ grouplength = get_grouplength(&pptr, TRUE, errcodeptr, lcptr, group,
+ recurses, cb);
if (grouplength < 0) return -1;
itemlength = grouplength;
break;
diff --git a/testdata/testinput2 b/testdata/testinput2
index 921d83c..43692ea 100644
--- a/testdata/testinput2
+++ b/testdata/testinput2
@@ -4972,4 +4972,6 @@ a)"xI
//
\=ovector=7777777777
+/(?<!\1((?U)1((?U))))(*F)/never_backslash_c,alt_bsux,anchored,extended
+
# End of testinput2
diff --git a/testdata/testoutput2 b/testdata/testoutput2
index 9862ea3..cf9c8a3 100644
--- a/testdata/testoutput2
+++ b/testdata/testoutput2
@@ -15483,6 +15483,8 @@ Subject length lower bound = 11
\=ovector=7777777777
** Invalid value in 'ovector=7777777777'
+/(?<!\1((?U)1((?U))))(*F)/never_backslash_c,alt_bsux,anchored,extended
+
# End of testinput2
Error -63: PCRE2_ERROR_BADDATA (unknown error number)
Error -62: bad serialized data
--
2.7.4

255
pcre2-10.23-Fix-crash-when-more-than-one-kind-of-push-was-set-in.patch
View File

@ -1,255 +0,0 @@
From fcc05a01a8c1f87b3286a27dffc508dd5b0ced31 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat, 6 May 2017 16:56:07 +0000
Subject: [PATCH] Fix crash when more than one kind of "push" was set in
pcre2test.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@767 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
---
src/pcre2test.c | 15 +++++++++------
testdata/testinput20 | 44 ++++++++++++++++++++++++++------------------
testdata/testoutput20 | 47 +++++++++++++++++++++++++++++------------------
diff --git a/src/pcre2test.c b/src/pcre2test.c
index a7fe7c0..cab604f 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -721,12 +721,15 @@ the first control word. Note that CTL_POSIX_NOSUB is always accompanied by
CTL_POSIX, so it doesn't need its own entries. */
static uint32_t exclusive_pat_controls[] = {
- CTL_POSIX | CTL_HEXPAT,
- CTL_POSIX | CTL_PUSH,
- CTL_POSIX | CTL_PUSHCOPY,
- CTL_POSIX | CTL_PUSHTABLESCOPY,
- CTL_POSIX | CTL_USE_LENGTH,
- CTL_EXPAND | CTL_HEXPAT };
+ CTL_POSIX | CTL_HEXPAT,
+ CTL_POSIX | CTL_PUSH,
+ CTL_POSIX | CTL_PUSHCOPY,
+ CTL_POSIX | CTL_PUSHTABLESCOPY,
+ CTL_POSIX | CTL_USE_LENGTH,
+ CTL_PUSH | CTL_PUSHCOPY,
+ CTL_PUSH | CTL_PUSHTABLESCOPY,
+ CTL_PUSHCOPY | CTL_PUSHTABLESCOPY,
+ CTL_EXPAND | CTL_HEXPAT };
/* Data controls that are mutually exclusive. At present these are all in the
first control word. */
diff --git a/testdata/testinput20 b/testdata/testinput20
index c87a07e..71f39ae 100644
--- a/testdata/testinput20
+++ b/testdata/testinput20
@@ -31,20 +31,20 @@
#load testsaved2
#pop info
- foofoo
+ foofoo
barbar
-
+
#pop mark
C
-\= Expect no match
- D
-
+\= Expect no match
+ D
+
#pop
- AmanaplanacanalPanama
+ AmanaplanacanalPanama
#pop info
metcalfe 33
-
+
# Check for an error when different tables are used.
/abc/push,tables=1
@@ -59,30 +59,30 @@
#pop should give an error
pqr
-
+
/abcd/pushcopy
abcd
-
+
#pop
- abcd
+ abcd
#pop should give an error
/abcd/push
#popcopy
abcd
-
+
#pop
- abcd
-
+ abcd
+
/abcd/push
#save testsaved1
#pop should give an error
#load testsaved1
-#popcopy
+#popcopy
abcd
-
+
#pop
abcd
@@ -91,10 +91,18 @@
/abcd/pushtablescopy
abcd
-#popcopy
+#popcopy
abcd
-
+
#pop
- abcd
+ abcd
+
+# Must only specify one of these
+
+//push,pushcopy
+
+//push,pushtablescopy
+
+//pushcopy,pushtablescopy
# End of testinput20
diff --git a/testdata/testoutput20 b/testdata/testoutput20
index db99866..d6265fd 100644
--- a/testdata/testoutput20
+++ b/testdata/testoutput20
@@ -40,25 +40,25 @@ Named capturing subpatterns:
Options: dupnames
Starting code units: b f
Subject length lower bound = 6
- foofoo
+ foofoo
0: foofoo
1: foo
barbar
0: barbar
1: <unset>
2: bar
-
+
#pop mark
C
0: C
1: C
MK: A
-\= Expect no match
- D
+\= Expect no match
+ D
No match, mark = A
-
+
#pop
- AmanaplanacanalPanama
+ AmanaplanacanalPanama
0: AmanaplanacanalPanama
1: <unset>
2: <unset>
@@ -78,7 +78,7 @@ Subject length lower bound = 3
0: metcalfe 33
1: metcalfe
2: 33
-
+
# Check for an error when different tables are used.
/abc/push,tables=1
@@ -97,13 +97,13 @@ Serialization failed: error -30: patterns do not all use the same character tabl
#pop should give an error
** Can't pop off an empty stack
pqr
-
+
/abcd/pushcopy
abcd
0: abcd
-
+
#pop
- abcd
+ abcd
0: abcd
#pop should give an error
@@ -113,21 +113,21 @@ Serialization failed: error -30: patterns do not all use the same character tabl
#popcopy
abcd
0: abcd
-
+
#pop
- abcd
+ abcd
0: abcd
-
+
/abcd/push
#save testsaved1
#pop should give an error
** Can't pop off an empty stack
#load testsaved1
-#popcopy
+#popcopy
abcd
0: abcd
-
+
#pop
abcd
0: abcd
@@ -139,12 +139,23 @@ Serialization failed: error -30: patterns do not all use the same character tabl
abcd
0: abcd
-#popcopy
+#popcopy
abcd
0: abcd
-
+
#pop
- abcd
+ abcd
0: abcd
+# Must only specify one of these
+
+//push,pushcopy
+** Not allowed together: push pushcopy
+
+//push,pushtablescopy
+** Not allowed together: push pushtablescopy
+
+//pushcopy,pushtablescopy
+** Not allowed together: pushcopy pushtablescopy
+
# End of testinput20
--
2.9.3

155
pcre2-10.23-Fix-lookbehind-with-zero-length-branch-in-DFA-matchi.patch
View File

@ -1,155 +0,0 @@
From 0dd869d72a472eca78ab323fd7318b3fae00593b Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat, 27 May 2017 15:49:29 +0000
Subject: [PATCH] Fix lookbehind with zero-length branch in DFA matching. Fixes
oss-fuzz 1959.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Petr Písař: Ported to 10.23:
commit 25926aac413c2b67359bc03c2a11a86a7a0718da
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat May 27 15:49:29 2017 +0000
Fix lookbehind with zero-length branch in DFA matching. Fixes oss-fuzz 1959.
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@807 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2_dfa_match.c | 30 +++++++++++-------------------
testdata/testinput6 | 6 ++++++
testdata/testoutput6 | 10 ++++++++++
3 files changed, 27 insertions(+), 19 deletions(-)
diff --git a/src/pcre2_dfa_match.c b/src/pcre2_dfa_match.c
index 519a147..7f0f9bf 100644
--- a/src/pcre2_dfa_match.c
+++ b/src/pcre2_dfa_match.c
@@ -375,14 +375,10 @@ internal_dfa_match(
{
stateblock *active_states, *new_states, *temp_states;
stateblock *next_active_state, *next_new_state;
-
const uint8_t *ctypes, *lcc, *fcc;
PCRE2_SPTR ptr;
PCRE2_SPTR end_code;
-PCRE2_SPTR first_op;
-
dfa_recursion_info new_recursive;
-
int active_count, new_count, match_count;
/* Some fields in the mb block are frequently referenced, so we load them into
@@ -417,21 +413,15 @@ active_states = (stateblock *)(workspace + 2);
next_new_state = new_states = active_states + wscount;
new_count = 0;
-first_op = this_start_code + 1 + LINK_SIZE +
- ((*this_start_code == OP_CBRA || *this_start_code == OP_SCBRA ||
- *this_start_code == OP_CBRAPOS || *this_start_code == OP_SCBRAPOS)
- ? IMM2_SIZE:0);
-
/* The first thing in any (sub) pattern is a bracket of some sort. Push all
the alternative states onto the list, and find out where the end is. This
makes is possible to use this function recursively, when we want to stop at a
matching internal ket rather than at the end.
-If the first opcode in the first alternative is OP_REVERSE, we are dealing with
-a backward assertion. In that case, we have to find out the maximum amount to
-move back, and set up each alternative appropriately. */
+If we are dealing with a backward assertion we have to find out the maximum
+amount to move back, and set up each alternative appropriately. */
-if (*first_op == OP_REVERSE)
+if (*this_start_code == OP_ASSERTBACK || *this_start_code == OP_ASSERTBACK_NOT)
{
size_t max_back = 0;
size_t gone_back;
@@ -476,15 +466,17 @@ if (*first_op == OP_REVERSE)
if (current_subject < mb->start_used_ptr)
mb->start_used_ptr = current_subject;
- /* Now we can process the individual branches. */
+ /* Now we can process the individual branches. There will be an OP_REVERSE at
+ the start of each branch, except when the length of the branch is zero. */
end_code = this_start_code;
do
{
- size_t back = (size_t)GET(end_code, 2+LINK_SIZE);
+ uint32_t revlen = (end_code[1+LINK_SIZE] == OP_REVERSE)? 1 + LINK_SIZE : 0;
+ size_t back = (revlen == 0)? 0 : (size_t)GET(end_code, 2+LINK_SIZE);
if (back <= gone_back)
{
- int bstate = (int)(end_code - start_code + 2 + 2*LINK_SIZE);
+ int bstate = (int)(end_code - start_code + 1 + LINK_SIZE + revlen);
ADD_NEW_DATA(-bstate, 0, (int)(gone_back - back));
}
end_code += GET(end_code, 1);
@@ -544,7 +536,7 @@ for (;;)
BOOL partial_newline = FALSE;
BOOL could_continue = reset_could_continue;
reset_could_continue = FALSE;
-
+
if (ptr > mb->last_used_ptr) mb->last_used_ptr = ptr;
/* Make the new state list into the active state list and empty the
@@ -597,7 +589,7 @@ for (;;)
int state_offset = current_state->offset;
int rrc;
int count;
-
+
/* A negative offset is a special case meaning "hold off going to this
(negated) state until the number of characters in the data field have
been skipped". If the could_continue flag was passed over from a previous
@@ -633,7 +625,7 @@ for (;;)
code = start_code + state_offset;
codevalue = *code;
-
+
/* If this opcode inspects a character, but we are at the end of the
subject, remember the fact for use when testing for a partial match. */
diff --git a/testdata/testinput6 b/testdata/testinput6
index be9b767..9fb96d2 100644
--- a/testdata/testinput6
+++ b/testdata/testinput6
@@ -4889,4 +4889,10 @@
/(02-)?[0-9]{3}-[0-9]{3}/
02-123-123
+/(?<=abc|)/
+ abcde\=aftertext
+
+/(?<=|abc)/
+ abcde\=aftertext
+
# End of testinput6
diff --git a/testdata/testoutput6 b/testdata/testoutput6
index 2930acc..f75e731 100644
--- a/testdata/testoutput6
+++ b/testdata/testoutput6
@@ -7691,4 +7691,14 @@ Failed: error -53: recursion limit exceeded
02-123-123
0: 02-123-123
+/(?<=abc|)/
+ abcde\=aftertext
+ 0:
+ 0+ abcde
+
+/(?<=|abc)/
+ abcde\=aftertext
+ 0:
+ 0+ abcde
+
# End of testinput6
--
2.9.4

231
pcre2-10.23-Fix-matching-offsets-from-regexec-in-the-POSIX-wrapp.patch
View File

@ -1,231 +0,0 @@
From 398087c31c9eeb80eecdc74440ada9ebe8e2a6d7 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat, 3 Jun 2017 16:42:58 +0000
Subject: [PATCH] Fix matching offsets from regexec() in the POSIX wrapper when
called with REG_STARTEND and a starting offset greater than zero.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Petr Písař: Ported to 10.23:
commit 4ed24ba49fc4a584c58509177e5a3ad6d1a000e4
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sat Jun 3 16:42:58 2017 +0000
Fix matching offsets from regexec() in the POSIX wrapper when called with
REG_STARTEND and a starting offset greater than zero.
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@818 6239d852-aaf2-0410-a92c-79f79f9480
69
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
doc/pcre2posix.3 | 30 ++++++++++++++++++------------
doc/pcre2test.1 | 14 ++++++++++++++
src/pcre2posix.c | 4 ++--
src/pcre2test.c | 16 +++++++++++++++-
testdata/testinput18 | 10 ++++++++++
testdata/testoutput18 | 17 +++++++++++++++++
6 files changed, 76 insertions(+), 15 deletions(-)
diff --git a/doc/pcre2posix.3 b/doc/pcre2posix.3
index 70a86d8..b37046b 100644
--- a/doc/pcre2posix.3
+++ b/doc/pcre2posix.3
@@ -1,4 +1,4 @@
-.TH PCRE2POSIX 3 "31 January 2016" "PCRE2 10.22"
+.TH PCRE2POSIX 3 "03 June 2017" "PCRE2 10.30"
.SH NAME
PCRE2 - Perl-compatible regular expressions (revised API)
.SH "SYNOPSIS"
@@ -204,15 +204,21 @@ function.
.sp
REG_STARTEND
.sp
-The string is considered to start at \fIstring\fP + \fIpmatch[0].rm_so\fP and
-to have a terminating NUL located at \fIstring\fP + \fIpmatch[0].rm_eo\fP
-(there need not actually be a NUL at that location), regardless of the value of
-\fInmatch\fP. This is a BSD extension, compatible with but not specified by
-IEEE Standard 1003.2 (POSIX.2), and should be used with caution in software
-intended to be portable to other systems. Note that a non-zero \fIrm_so\fP does
-not imply REG_NOTBOL; REG_STARTEND affects only the location of the string, not
-how it is matched. Setting REG_STARTEND and passing \fIpmatch\fP as NULL are
-mutually exclusive; the error REG_INVARG is returned.
+When this option is set, the string is considered to start at \fIstring\fP +
+\fIpmatch[0].rm_so\fP and to have a terminating NUL located at \fIstring\fP +
+\fIpmatch[0].rm_eo\fP (there need not actually be a NUL at that location),
+regardless of the value of \fInmatch\fP. However, the offsets of the matched
+string and any captured substrings are still given relative to the start of
+\fIstring\fP. (Before PCRE2 release 10.30 these were given relative to
+\fIstring\fP + \fIpmatch[0].rm_so\fP, but this differs from other
+implementations.)
+.P
+This is a BSD extension, compatible with but not specified by IEEE Standard
+1003.2 (POSIX.2), and should be used with caution in software intended to be
+portable to other systems. Note that a non-zero \fIrm_so\fP does not imply
+REG_NOTBOL; REG_STARTEND affects only the location of the string, not how it is
+matched. Setting REG_STARTEND and passing \fIpmatch\fP as NULL are mutually
+exclusive; the error REG_INVARG is returned.
.P
If the pattern was compiled with the REG_NOSUB flag, no data about any matched
strings is returned. The \fInmatch\fP and \fIpmatch\fP arguments of
@@ -271,6 +277,6 @@ Cambridge, England.
.rs
.sp
.nf
-Last updated: 31 January 2016
-Copyright (c) 1997-2016 University of Cambridge.
+Last updated: 03 June 2017
+Copyright (c) 1997-2017 University of Cambridge.
.fi
diff --git a/doc/pcre2test.1 b/doc/pcre2test.1
index bd7383e..f8ce5bf 100644
--- a/doc/pcre2test.1
+++ b/doc/pcre2test.1
@@ -1023,6 +1023,20 @@ wrapper API to be used, the only option-setting modifiers that have any effect
are \fBnotbol\fP, \fBnotempty\fP, and \fBnoteol\fP, causing REG_NOTBOL,
REG_NOTEMPTY, and REG_NOTEOL, respectively, to be passed to \fBregexec()\fP.
The other modifiers are ignored, with a warning message.
+.P
+There is one additional modifier that can be used with the POSIX wrapper. It is
+ignored (with a warning) if used for non-POSIX matching.
+.sp
+ posix_startend=<n>[:<m>]
+.sp
+This causes the subject string to be passed to \fBregexec()\fP using the
+REG_STARTEND option, which uses offsets to restrict which part of the string is
+searched. If only one number is given, the end offset is passed as the end of
+the subject string. For more detail of REG_STARTEND, see the
+.\" HREF
+\fBpcre2posix\fP
+.\"
+documentation.
.
.
.SS "Setting match controls"
diff --git a/src/pcre2posix.c b/src/pcre2posix.c
index 4ecc701..8be969a 100644
--- a/src/pcre2posix.c
+++ b/src/pcre2posix.c
@@ -338,8 +338,8 @@ if (rc >= 0)
if ((size_t)rc > nmatch) rc = (int)nmatch;
for (i = 0; i < (size_t)rc; i++)
{
- pmatch[i].rm_so = ovector[i*2];
- pmatch[i].rm_eo = ovector[i*2+1];
+ pmatch[i].rm_so = ovector[i*2] + so;
+ pmatch[i].rm_eo = ovector[i*2+1] + so;
}
for (; i < nmatch; i++) pmatch[i].rm_so = pmatch[i].rm_eo = -1;
return 0;
diff --git a/src/pcre2test.c b/src/pcre2test.c
index af05b68..ecb6811 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -182,7 +182,7 @@ void vms_setsymbol( char *, char *, int );
#endif
#endif
-#define CFORE_UNSET UINT32_MAX /* Unset value for cfail/cerror fields */
+#define CFORE_UNSET UINT32_MAX /* Unset value for startend/cfail/cerror fields */
#define DFA_WS_DIMENSION 1000 /* Size of DFA workspace */
#define DEFAULT_OVECCOUNT 15 /* Default ovector count */
#define JUNK_OFFSET 0xdeadbeef /* For initializing ovector */
@@ -502,6 +502,7 @@ typedef struct datctl { /* Structure for data line modifiers. */
uint32_t control; /* Must be in same position as patctl */
uint32_t control2; /* Must be in same position as patctl */
uint8_t replacement[REPLACE_MODSIZE]; /* So must this */
+ uint32_t startend[2];
uint32_t cerror[2];
uint32_t cfail[2];
int32_t callout_data;
@@ -615,6 +616,7 @@ static modstruct modlist[] = {
{ "ph", MOD_DAT, MOD_OPT, PCRE2_PARTIAL_HARD, DO(options) },
{ "posix", MOD_PAT, MOD_CTL, CTL_POSIX, PO(control) },
{ "posix_nosub", MOD_PAT, MOD_CTL, CTL_POSIX|CTL_POSIX_NOSUB, PO(control) },
+ { "posix_startend", MOD_DAT, MOD_IN2, 0, DO(startend) },
{ "ps", MOD_DAT, MOD_OPT, PCRE2_PARTIAL_SOFT, DO(options) },
{ "push", MOD_PAT, MOD_CTL, CTL_PUSH, PO(control) },
{ "pushcopy", MOD_PAT, MOD_CTL, CTL_PUSHCOPY, PO(control) },
@@ -6193,6 +6195,14 @@ if ((pat_patctl.control & CTL_POSIX) != 0)
}
}
+ if (dat_datctl.startend[0] != CFORE_UNSET)
+ {
+ pmatch[0].rm_so = dat_datctl.startend[0];
+ pmatch[0].rm_eo = (dat_datctl.startend[1] != 0)?
+ dat_datctl.startend[1] : len;
+ eflags |= REG_STARTEND;
+ }
+
if ((dat_datctl.options & PCRE2_NOTBOL) != 0) eflags |= REG_NOTBOL;
if ((dat_datctl.options & PCRE2_NOTEOL) != 0) eflags |= REG_NOTEOL;
if ((dat_datctl.options & PCRE2_NOTEMPTY) != 0) eflags |= REG_NOTEMPTY;
@@ -6252,6 +6262,9 @@ if ((dat_datctl.control & (CTL_DFA|CTL_FINDLIMITS)) == (CTL_DFA|CTL_FINDLIMITS))
dat_datctl.control &= ~CTL_FINDLIMITS;
}
+if (dat_datctl.startend[0] != CFORE_UNSET)
+ fprintf(outfile, "** \\=posix_startend ignored for non-POSIX matching\n");
+
/* ALLUSEDTEXT is not supported with JIT, but JIT is not used with DFA
matching, even if the JIT compiler was used. */
@@ -7407,6 +7420,7 @@ memset(&def_datctl, 0, sizeof(datctl));
def_datctl.oveccount = DEFAULT_OVECCOUNT;
def_datctl.copy_numbers[0] = -1;
def_datctl.get_numbers[0] = -1;
+def_datctl.startend[0] = def_datctl.startend[1] = CFORE_UNSET;
def_datctl.cerror[0] = def_datctl.cerror[1] = CFORE_UNSET;
def_datctl.cfail[0] = def_datctl.cfail[1] = CFORE_UNSET;
diff --git a/testdata/testinput18 b/testdata/testinput18
index bd1c6ad..d8d8084 100644
--- a/testdata/testinput18
+++ b/testdata/testinput18
@@ -109,4 +109,14 @@
/(?=(a\K))/
a
+/^d(e)$/posix
+ acdef\=posix_startend=2:4
+ acde\=posix_startend=2
+\= Expect no match
+ acdef
+ acdef\=posix_startend=2
+
+/^a\x{00}b$/posix
+ a\x{00}b\=posix_startend=0:3
+
# End of testdata/testinput18
diff --git a/testdata/testoutput18 b/testdata/testoutput18
index fd6fac3..0c13abc 100644
--- a/testdata/testoutput18
+++ b/testdata/testoutput18
@@ -168,4 +168,21 @@ Start of matched string is beyond its end - displaying from end to start.
0: a
1: a
+/^d(e)$/posix
+ acdef\=posix_startend=2:4
+ 0: de
+ 1: e
+ acde\=posix_startend=2
+ 0: de
+ 1: e
+\= Expect no match
+ acdef
+No match: POSIX code 17: match failed
+ acdef\=posix_startend=2
+No match: POSIX code 17: match failed
+
+/^a\x{00}b$/posix
+ a\x{00}b\=posix_startend=0:3
+ 0: a\x00b
+
# End of testdata/testinput18
--
2.9.4

40
pcre2-10.23-Fix-memory-leak-when-deserializing-invalid-data-Bugz.patch
View File

@ -1,40 +0,0 @@
From 0fece4355e2a5d494936d285eb200314112c9a8b Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Tue, 21 Mar 2017 16:25:01 +0000
Subject: [PATCH] Fix memory leak when deserializing invalid data (Bugzilla
2075).
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@693 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař: Ported to 10.23.
diff --git a/src/pcre2_serialize.c b/src/pcre2_serialize.c
index 0af26d8..d2cc603 100644
--- a/src/pcre2_serialize.c
+++ b/src/pcre2_serialize.c
@@ -7,7 +7,7 @@ and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Original API code Copyright (c) 1997-2012 University of Cambridge
- New API code Copyright (c) 2016 University of Cambridge
+ New API code Copyright (c) 2016-2017 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@@ -214,7 +214,10 @@ for (i = 0; i < number_of_codes; i++)
if (dst_re->magic_number != MAGIC_NUMBER ||
dst_re->name_entry_size > MAX_NAME_SIZE + IMM2_SIZE + 1 ||
dst_re->name_count > MAX_NAME_COUNT)
+ {
+ memctl->free(dst_re, memctl->memory_data);
return PCRE2_ERROR_BADSERIALIZEDDATA;
+ }
/* At the moment only one table is supported. */
--
2.7.4

101
pcre2-10.23-Fix-misbehaving-DFA-match-for-possessively-repeated-.patch
View File

@ -1,101 +0,0 @@
From a96183f8281c6e8b0a42c5a05146a96f9540b673 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Wed, 22 Mar 2017 15:12:06 +0000
Subject: [PATCH] Fix misbehaving DFA match for possessively repeated character
class (Bugzilla 2086).
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Ported to 10.23:
commit 46399965f23c3efc1197823d1cd54084b27d9174
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Wed Mar 22 15:12:06 2017 +0000
Fix misbehaving DFA match for possessively repeated character class (Bugzilla
2086).
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@698 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2_dfa_match.c | 10 ++++++----
testdata/testinput6 | 3 +++
testdata/testoutput6 | 4 ++++
3 files changed, 13 insertions(+), 4 deletions(-)
diff --git a/src/pcre2_dfa_match.c b/src/pcre2_dfa_match.c
index c909d61..519a147 100644
--- a/src/pcre2_dfa_match.c
+++ b/src/pcre2_dfa_match.c
@@ -544,7 +544,7 @@ for (;;)
BOOL partial_newline = FALSE;
BOOL could_continue = reset_could_continue;
reset_could_continue = FALSE;
-
+
if (ptr > mb->last_used_ptr) mb->last_used_ptr = ptr;
/* Make the new state list into the active state list and empty the
@@ -597,7 +597,7 @@ for (;;)
int state_offset = current_state->offset;
int rrc;
int count;
-
+
/* A negative offset is a special case meaning "hold off going to this
(negated) state until the number of characters in the data field have
been skipped". If the could_continue flag was passed over from a previous
@@ -633,7 +633,7 @@ for (;;)
code = start_code + state_offset;
codevalue = *code;
-
+
/* If this opcode inspects a character, but we are at the end of the
subject, remember the fact for use when testing for a partial match. */
@@ -2539,11 +2539,13 @@ for (;;)
if (isinclass)
{
int max = (int)GET2(ecode, 1 + IMM2_SIZE);
- if (*ecode == OP_CRPOSRANGE)
+
+ if (*ecode == OP_CRPOSRANGE && count >= (int)GET2(ecode, 1))
{
active_count--; /* Remove non-match possibility */
next_active_state--;
}
+
if (++count >= max && max != 0) /* Max 0 => no limit */
{ ADD_NEW(next_state_offset + 1 + 2 * IMM2_SIZE, 0); }
else
diff --git a/testdata/testinput6 b/testdata/testinput6
index ee6cd1b..be9b767 100644
--- a/testdata/testinput6
+++ b/testdata/testinput6
@@ -4886,4 +4886,7 @@
\= Expect recursion limit exceeded
a[00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]
+/(02-)?[0-9]{3}-[0-9]{3}/
+ 02-123-123
+
# End of testinput6
diff --git a/testdata/testoutput6 b/testdata/testoutput6
index 62f2674..2930acc 100644
--- a/testdata/testoutput6
+++ b/testdata/testoutput6
@@ -7687,4 +7687,8 @@ No match
a[00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]([00]
Failed: error -53: recursion limit exceeded
+/(02-)?[0-9]{3}-[0-9]{3}/
+ 02-123-123
+ 0: 02-123-123
+
# End of testinput6
--
2.7.4

183
pcre2-10.23-Fix-pcre2test-bug-for-global-match-with-zero-termina.patch
View File

@ -1,183 +0,0 @@
From 37fdecdb938eae9aece6c3b552b26d9054212a43 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sun, 19 Mar 2017 18:34:27 +0000
Subject: [PATCH] Fix pcre2test bug for global match with zero terminated
subject.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Ported to 10.23:
commit 0f66bd9a67d59124c7f81b44ee28b6c1f26fd789
Author: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Sun Mar 19 18:34:27 2017 +0000
Fix pcre2test bug for global match with zero terminated subject.
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@687 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2test.c | 38 ++++++++++++++++++++++----------------
testdata/testinput5 | 3 +++
testdata/testoutput5 | 4 ++++
3 files changed, 29 insertions(+), 16 deletions(-)
diff --git a/src/pcre2test.c b/src/pcre2test.c
index 01457e8..4220b46 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -5724,7 +5724,7 @@ Returns: PR_OK continue processing next line
static int
process_data(void)
{
-PCRE2_SIZE len, ulen;
+PCRE2_SIZE len, ulen, arg_ulen;
uint32_t gmatched;
uint32_t c, k;
uint32_t g_notempty = 0;
@@ -6087,6 +6087,7 @@ ENDSTRING:
SET(*q, 0);
len = CASTVAR(uint8_t *, q) - dbuffer; /* Length in bytes */
ulen = len/code_unit_size; /* Length in code units */
+arg_ulen = ulen; /* Value to use in match arg */
/* If the string was terminated by \= we must now interpret modifiers. */
@@ -6115,11 +6116,15 @@ if (pat_patctl.replacement[0] != 0 &&
}
/* We now have the subject in dbuffer, with len containing the byte length, and
-ulen containing the code unit length. Move the data to the end of the buffer so
-that a read over the end can be caught by valgrind or other means. If we have
-explicit valgrind support, mark the unused start of the buffer unaddressable.
-If we are using the POSIX interface, or testing zero-termination, we must
-include the terminating zero in the usable data. */
+ulen containing the code unit length, with a copy in arg_ulen for use in match
+function arguments (this gets changed to PCRE2_ZERO_TERMINATED when the
+zero_terminate modifier is present).
+
+Move the data to the end of the buffer so that a read over the end can be
+caught by valgrind or other means. If we have explicit valgrind support, mark
+the unused start of the buffer unaddressable. If we are using the POSIX
+interface, or testing zero-termination, we must include the terminating zero in
+the usable data. */
c = code_unit_size * (((pat_patctl.control & CTL_POSIX) +
(dat_datctl.control & CTL_ZERO_TERMINATE) != 0)? 1:0);
@@ -6250,7 +6255,7 @@ if ((dat_datctl.control & (CTL_ALLUSEDTEXT|CTL_DFA)) == CTL_ALLUSEDTEXT &&
/* Handle passing the subject as zero-terminated. */
if ((dat_datctl.control & CTL_ZERO_TERMINATE) != 0)
- ulen = PCRE2_ZERO_TERMINATED;
+ arg_ulen = PCRE2_ZERO_TERMINATED;
/* The nullcontext modifier is used to test calling pcre2_[jit_]match() with a
NULL context. */
@@ -6452,7 +6457,7 @@ if (dat_datctl.replacement[0] != 0)
rlen = PCRE2_ZERO_TERMINATED;
else
rlen = (CASTVAR(uint8_t *, r) - rbuffer)/code_unit_size;
- PCRE2_SUBSTITUTE(rc, compiled_code, pp, ulen, dat_datctl.offset,
+ PCRE2_SUBSTITUTE(rc, compiled_code, pp, arg_ulen, dat_datctl.offset,
dat_datctl.options|xoptions, match_data, dat_context,
rbuffer, rlen, nbuffer, &nsize);
@@ -6534,7 +6539,7 @@ else for (gmatched = 0;; gmatched++)
start_time = clock();
for (i = 0; i < timeitm; i++)
{
- PCRE2_DFA_MATCH(capcount, compiled_code, pp, ulen,
+ PCRE2_DFA_MATCH(capcount, compiled_code, pp, arg_ulen,
dat_datctl.offset, dat_datctl.options | g_notempty, match_data,
use_dat_context, dfa_workspace, DFA_WS_DIMENSION);
}
@@ -6545,7 +6550,7 @@ else for (gmatched = 0;; gmatched++)
start_time = clock();
for (i = 0; i < timeitm; i++)
{
- PCRE2_JIT_MATCH(capcount, compiled_code, pp, ulen,
+ PCRE2_JIT_MATCH(capcount, compiled_code, pp, arg_ulen,
dat_datctl.offset, dat_datctl.options | g_notempty, match_data,
use_dat_context);
}
@@ -6556,7 +6561,7 @@ else for (gmatched = 0;; gmatched++)
start_time = clock();
for (i = 0; i < timeitm; i++)
{
- PCRE2_MATCH(capcount, compiled_code, pp, ulen,
+ PCRE2_MATCH(capcount, compiled_code, pp, arg_ulen,
dat_datctl.offset, dat_datctl.options | g_notempty, match_data,
use_dat_context);
}
@@ -6572,9 +6577,9 @@ else for (gmatched = 0;; gmatched++)
if ((dat_datctl.control & CTL_FINDLIMITS) != 0)
{
- capcount = check_match_limit(pp, ulen, PCRE2_ERROR_MATCHLIMIT, "match");
+ capcount = check_match_limit(pp, arg_ulen, PCRE2_ERROR_MATCHLIMIT, "match");
if (FLD(compiled_code, executable_jit) == NULL)
- (void)check_match_limit(pp, ulen, PCRE2_ERROR_RECURSIONLIMIT,
+ (void)check_match_limit(pp, arg_ulen, PCRE2_ERROR_RECURSIONLIMIT,
"recursion");
}
@@ -6604,7 +6609,7 @@ else for (gmatched = 0;; gmatched++)
dfa_workspace = (int *)malloc(DFA_WS_DIMENSION*sizeof(int));
if (dfa_matched++ == 0)
dfa_workspace[0] = -1; /* To catch bad restart */
- PCRE2_DFA_MATCH(capcount, compiled_code, pp, ulen,
+ PCRE2_DFA_MATCH(capcount, compiled_code, pp, arg_ulen,
dat_datctl.offset, dat_datctl.options | g_notempty, match_data,
use_dat_context, dfa_workspace, DFA_WS_DIMENSION);
if (capcount == 0)
@@ -6616,10 +6621,10 @@ else for (gmatched = 0;; gmatched++)
else
{
if ((pat_patctl.control & CTL_JITFAST) != 0)
- PCRE2_JIT_MATCH(capcount, compiled_code, pp, ulen, dat_datctl.offset,
+ PCRE2_JIT_MATCH(capcount, compiled_code, pp, arg_ulen, dat_datctl.offset,
dat_datctl.options | g_notempty, match_data, use_dat_context);
else
- PCRE2_MATCH(capcount, compiled_code, pp, ulen, dat_datctl.offset,
+ PCRE2_MATCH(capcount, compiled_code, pp, arg_ulen, dat_datctl.offset,
dat_datctl.options | g_notempty, match_data, use_dat_context);
if (capcount == 0)
{
@@ -7032,6 +7037,7 @@ else for (gmatched = 0;; gmatched++)
pp += end_offset * code_unit_size;
len -= end_offset * code_unit_size;
ulen -= end_offset;
+ if (arg_ulen != PCRE2_ZERO_TERMINATED) arg_ulen -= end_offset;
}
}
} /* End of global loop */
diff --git a/testdata/testinput5 b/testdata/testinput5
index e5a43e5..a574872 100644
--- a/testdata/testinput5
+++ b/testdata/testinput5
@@ -1763,4 +1763,7 @@
/[^\HH]/Bi,utf
+//g,utf
+ \=zero_terminate
+
# End of testinput5
diff --git a/testdata/testoutput5 b/testdata/testoutput5
index 9651fd1..26f9569 100644
--- a/testdata/testoutput5
+++ b/testdata/testoutput5
@@ -4232,4 +4232,8 @@ Failed: error 125 at offset 2: lookbehind assertion is not fixed length
End
------------------------------------------------------------------
+//g,utf
+ \=zero_terminate
+ 0:
+
# End of testinput5
--
2.7.4

73
pcre2-10.23-Previous-patch-was-not-quite-complete.patch
View File

@ -1,73 +0,0 @@
From 3091213a8f163aaad43390229380e22adb096787 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Tue, 21 Mar 2017 18:36:13 +0000
Subject: [PATCH 2/2] Previous patch was not quite complete.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@697 6239d852-aaf2-0410-a92c-79f79f948069
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2test.c | 15 +++++++++------
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git a/src/pcre2test.c b/src/pcre2test.c
index 9289656..b979dd1 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -1017,9 +1017,9 @@ are supported. */
if (test_mode == PCRE8_MODE) \
r = pcre2_get_error_message_8(a,G(b,8),G(G(b,8),_size)); \
else if (test_mode == PCRE16_MODE) \
- r = pcre2_get_error_message_16(a,G(b,16),G(G(b,16),_size)); \
+ r = pcre2_get_error_message_16(a,G(b,16),G(G(b,16),_size/2)); \
else \
- r = pcre2_get_error_message_32(a,G(b,32),G(G(b,32),_size))
+ r = pcre2_get_error_message_32(a,G(b,32),G(G(b,32),_size/4))
#define PCRE2_GET_OVECTOR_COUNT(a,b) \
if (test_mode == PCRE8_MODE) \
@@ -1399,6 +1399,9 @@ the three different cases. */
/* ----- Common macros for two-mode cases ----- */
+#define BYTEONE (BITONE/8)
+#define BYTETWO (BITTWO/8)
+
#define CASTFLD(t,a,b) \
((test_mode == G(G(PCRE,BITONE),_MODE))? (t)(G(a,BITONE)->b) : \
(t)(G(a,BITTWO)->b))
@@ -1481,9 +1484,9 @@ the three different cases. */
#define PCRE2_GET_ERROR_MESSAGE(r,a,b) \
if (test_mode == G(G(PCRE,BITONE),_MODE)) \
- r = G(pcre2_get_error_message_,BITONE)(a,G(b,BITONE),G(G(b,BITONE),_size)); \
+ r = G(pcre2_get_error_message_,BITONE)(a,G(b,BITONE),G(G(b,BITONE),_size/BYTEONE)); \
else \
- r = G(pcre2_get_error_message_,BITTWO)(a,G(b,BITTWO),G(G(b,BITTWO),_size))
+ r = G(pcre2_get_error_message_,BITTWO)(a,G(b,BITTWO),G(G(b,BITTWO),_size/BYTETWO))
#define PCRE2_GET_OVECTOR_COUNT(a,b) \
if (test_mode == G(G(PCRE,BITONE),_MODE)) \
@@ -1904,7 +1907,7 @@ the three different cases. */
#define PCRE2_DFA_MATCH(a,b,c,d,e,f,g,h,i,j) \
a = pcre2_dfa_match_16(G(b,16),(PCRE2_SPTR16)c,d,e,f,G(g,16),h,i,j)
#define PCRE2_GET_ERROR_MESSAGE(r,a,b) \
- r = pcre2_get_error_message_16(a,G(b,16),G(G(b,16),_size))
+ r = pcre2_get_error_message_16(a,G(b,16),G(G(b,16),_size/2))
#define PCRE2_GET_OVECTOR_COUNT(a,b) a = pcre2_get_ovector_count_16(G(b,16))
#define PCRE2_GET_STARTCHAR(a,b) a = pcre2_get_startchar_16(G(b,16))
#define PCRE2_JIT_COMPILE(r,a,b) r = pcre2_jit_compile_16(G(a,16),b)
@@ -2000,7 +2003,7 @@ the three different cases. */
#define PCRE2_DFA_MATCH(a,b,c,d,e,f,g,h,i,j) \
a = pcre2_dfa_match_32(G(b,32),(PCRE2_SPTR32)c,d,e,f,G(g,32),h,i,j)
#define PCRE2_GET_ERROR_MESSAGE(r,a,b) \
- r = pcre2_get_error_message_32(a,G(b,32),G(G(b,32),_size))
+ r = pcre2_get_error_message_32(a,G(b,32),G(G(b,32),_size/4))
#define PCRE2_GET_OVECTOR_COUNT(a,b) a = pcre2_get_ovector_count_32(G(b,32))
#define PCRE2_GET_STARTCHAR(a,b) a = pcre2_get_startchar_32(G(b,32))
#define PCRE2_JIT_COMPILE(r,a,b) r = pcre2_jit_compile_32(G(a,32),b)
--
2.7.4

457
pcre2-10.30-RC1-Distribute-forgotten-sljitProtExecAllocator.c.patch Normal file
View File

@ -0,0 +1,457 @@
From 953920c6fac83750ac8a4528672f1bd2c41d1fc6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Petr=20P=C3=ADsa=C5=99?= <ppisar@redhat.com>
Date: Thu, 20 Jul 2017 10:38:24 +0200
Subject: [PATCH] Distribute forgotten sljitProtExecAllocator.c
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
Makefile.am | 1 +
src/sljit/sljitProtExecAllocator.c | 421 +++++++++++++++++++++++++++++++++++++
2 files changed, 422 insertions(+)
create mode 100644 src/sljit/sljitProtExecAllocator.c
diff --git a/Makefile.am b/Makefile.am
index bbf23b8..7fa98c5 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -442,6 +442,7 @@ EXTRA_DIST += \
src/sljit/sljitNativeX86_32.c \
src/sljit/sljitNativeX86_64.c \
src/sljit/sljitNativeX86_common.c \
+ src/sljit/sljitProtExecAllocator.c \
src/sljit/sljitUtils.c
# Some of the JIT sources are also in separate files that are #included.
diff --git a/src/sljit/sljitProtExecAllocator.c b/src/sljit/sljitProtExecAllocator.c
new file mode 100644
index 0000000..8a5b2b3
--- /dev/null
+++ b/src/sljit/sljitProtExecAllocator.c
@@ -0,0 +1,421 @@
+/*
+ * Stack-less Just-In-Time compiler
+ *
+ * Copyright Zoltan Herczeg (hzmester@freemail.hu). All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without modification, are
+ * permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice, this list of
+ * conditions and the following disclaimer.
+ *
+ * 2. Redistributions in binary form must reproduce the above copyright notice, this list
+ * of conditions and the following disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER(S) AND CONTRIBUTORS ``AS IS'' AND ANY
+ * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+ * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT
+ * SHALL THE COPYRIGHT HOLDER(S) OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+ * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED
+ * TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR
+ * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
+ * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+/*
+ This file contains a simple executable memory allocator
+
+ It is assumed, that executable code blocks are usually medium (or sometimes
+ large) memory blocks, and the allocator is not too frequently called (less
+ optimized than other allocators). Thus, using it as a generic allocator is
+ not suggested.
+
+ How does it work:
+ Memory is allocated in continuous memory areas called chunks by alloc_chunk()
+ Chunk format:
+ [ block ][ block ] ... [ block ][ block terminator ]
+
+ All blocks and the block terminator is started with block_header. The block
+ header contains the size of the previous and the next block. These sizes
+ can also contain special values.
+ Block size:
+ 0 - The block is a free_block, with a different size member.
+ 1 - The block is a block terminator.
+ n - The block is used at the moment, and the value contains its size.
+ Previous block size:
+ 0 - This is the first block of the memory chunk.
+ n - The size of the previous block.
+
+ Using these size values we can go forward or backward on the block chain.
+ The unused blocks are stored in a chain list pointed by free_blocks. This
+ list is useful if we need to find a suitable memory area when the allocator
+ is called.
+
+ When a block is freed, the new free block is connected to its adjacent free
+ blocks if possible.
+
+ [ free block ][ used block ][ free block ]
+ and "used block" is freed, the three blocks are connected together:
+ [ one big free block ]
+*/
+
+/* --------------------------------------------------------------------- */
+/* System (OS) functions */
+/* --------------------------------------------------------------------- */
+
+/* 64 KByte. */
+#define CHUNK_SIZE 0x10000
+
+struct chunk_header {
+ void *executable;
+ int fd;
+};
+
+/*
+ alloc_chunk / free_chunk :
+ * allocate executable system memory chunks
+ * the size is always divisible by CHUNK_SIZE
+ allocator_grab_lock / allocator_release_lock :
+ * make the allocator thread safe
+ * can be empty if the OS (or the application) does not support threading
+ * only the allocator requires this lock, sljit is fully thread safe
+ as it only uses local variables
+*/
+
+#include <fcntl.h>
+
+#ifndef O_NOATIME
+#define O_NOATIME 0
+#endif
+
+#ifdef __O_TMPFILE
+#ifndef O_TMPFILE
+#define O_TMPFILE (__O_TMPFILE | O_DIRECTORY)
+#endif
+#endif
+
+int mkostemp(char *template, int flags);
+char *secure_getenv(const char *name);
+
+static SLJIT_INLINE int create_tempfile(void)
+{
+ int fd;
+
+ char tmp_name[256];
+ size_t tmp_name_len;
+ char *dir;
+ size_t len;
+
+#ifdef P_tmpdir
+ len = (P_tmpdir != NULL) ? strlen(P_tmpdir) : 0;
+
+ if (len > 0 && len < sizeof(tmp_name)) {
+ strcpy(tmp_name, P_tmpdir);
+ tmp_name_len = len;
+ }
+ else {
+ strcpy(tmp_name, "/tmp");
+ tmp_name_len = 4;
+ }
+#else
+ strcpy(tmp_name, "/tmp");
+ tmp_name_len = 4;
+#endif
+
+ dir = secure_getenv("TMPDIR");
+ if (dir) {
+ len = strlen(dir);
+ if (len > 0 && len < sizeof(tmp_name)) {
+ strcpy(tmp_name, dir);
+ tmp_name_len = len;
+ }
+ }
+
+ SLJIT_ASSERT(tmp_name_len > 0 && tmp_name_len < sizeof(tmp_name));
+
+ while (tmp_name_len > 0 && tmp_name[tmp_name_len - 1] == '/') {
+ tmp_name_len--;
+ tmp_name[tmp_name_len] = '\0';
+ }
+
+#ifdef O_TMPFILE
+ fd = open(tmp_name, O_TMPFILE | O_EXCL | O_RDWR | O_NOATIME | O_CLOEXEC, S_IRUSR | S_IWUSR);
+ if (fd != -1)
+ return fd;
+#endif
+
+ if (tmp_name_len + 7 >= sizeof(tmp_name))
+ {
+ return -1;
+ }
+
+ strcpy(tmp_name + tmp_name_len, "/XXXXXX");
+ fd = mkostemp(tmp_name, O_CLOEXEC | O_NOATIME);
+
+ if (fd == -1)
+ return fd;
+
+ if (unlink(tmp_name)) {
+ close(fd);
+ return -1;
+ }
+
+ return fd;
+}
+
+static SLJIT_INLINE struct chunk_header* alloc_chunk(sljit_uw size)
+{
+ struct chunk_header *retval;
+ int fd;
+
+ fd = create_tempfile();
+ if (fd == -1)
+ return NULL;
+
+ if (ftruncate(fd, size)) {
+ close(fd);
+ return NULL;
+ }
+
+ retval = (struct chunk_header *)mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0);
+
+ if (retval == MAP_FAILED) {
+ close(fd);
+ return NULL;
+ }
+
+ retval->executable = mmap(NULL, size, PROT_READ | PROT_EXEC, MAP_SHARED, fd, 0);
+
+ if (retval->executable == MAP_FAILED) {
+ munmap(retval, size);
+ close(fd);
+ return NULL;
+ }
+
+ retval->fd = fd;
+ return retval;
+}
+
+static SLJIT_INLINE void free_chunk(void *chunk, sljit_uw size)
+{
+ struct chunk_header *header = ((struct chunk_header *)chunk) - 1;
+
+ int fd = header->fd;
+ munmap(header->executable, size);
+ munmap(header, size);
+ close(fd);
+}
+
+/* --------------------------------------------------------------------- */
+/* Common functions */
+/* --------------------------------------------------------------------- */
+
+#define CHUNK_MASK (~(CHUNK_SIZE - 1))
+
+struct block_header {
+ sljit_uw size;
+ sljit_uw prev_size;
+ sljit_sw executable_offset;
+};
+
+struct free_block {
+ struct block_header header;
+ struct free_block *next;
+ struct free_block *prev;
+ sljit_uw size;
+};
+
+#define AS_BLOCK_HEADER(base, offset) \
+ ((struct block_header*)(((sljit_u8*)base) + offset))
+#define AS_FREE_BLOCK(base, offset) \
+ ((struct free_block*)(((sljit_u8*)base) + offset))
+#define MEM_START(base) ((void*)((base) + 1))
+#define ALIGN_SIZE(size) (((size) + sizeof(struct block_header) + 7) & ~7)
+
+static struct free_block* free_blocks;
+static sljit_uw allocated_size;
+static sljit_uw total_size;
+
+static SLJIT_INLINE void sljit_insert_free_block(struct free_block *free_block, sljit_uw size)
+{
+ free_block->header.size = 0;
+ free_block->size = size;
+
+ free_block->next = free_blocks;
+ free_block->prev = NULL;
+ if (free_blocks)
+ free_blocks->prev = free_block;
+ free_blocks = free_block;
+}
+
+static SLJIT_INLINE void sljit_remove_free_block(struct free_block *free_block)
+{
+ if (free_block->next)
+ free_block->next->prev = free_block->prev;
+
+ if (free_block->prev)
+ free_block->prev->next = free_block->next;
+ else {
+ SLJIT_ASSERT(free_blocks == free_block);
+ free_blocks = free_block->next;
+ }
+}
+
+SLJIT_API_FUNC_ATTRIBUTE void* sljit_malloc_exec(sljit_uw size)
+{
+ struct chunk_header *chunk_header;
+ struct block_header *header;
+ struct block_header *next_header;
+ struct free_block *free_block;
+ sljit_uw chunk_size;
+ sljit_sw executable_offset;
+
+ allocator_grab_lock();
+ if (size < (64 - sizeof(struct block_header)))
+ size = (64 - sizeof(struct block_header));
+ size = ALIGN_SIZE(size);
+
+ free_block = free_blocks;
+ while (free_block) {
+ if (free_block->size >= size) {
+ chunk_size = free_block->size;
+ if (chunk_size > size + 64) {
+ /* We just cut a block from the end of the free block. */
+ chunk_size -= size;
+ free_block->size = chunk_size;
+ header = AS_BLOCK_HEADER(free_block, chunk_size);
+ header->prev_size = chunk_size;
+ header->executable_offset = free_block->header.executable_offset;
+ AS_BLOCK_HEADER(header, size)->prev_size = size;
+ }
+ else {
+ sljit_remove_free_block(free_block);
+ header = (struct block_header*)free_block;
+ size = chunk_size;
+ }
+ allocated_size += size;
+ header->size = size;
+ allocator_release_lock();
+ return MEM_START(header);
+ }
+ free_block = free_block->next;
+ }
+
+ chunk_size = sizeof(struct chunk_header) + sizeof(struct block_header);
+ chunk_size = (chunk_size + size + CHUNK_SIZE - 1) & CHUNK_MASK;
+
+ chunk_header = alloc_chunk(chunk_size);
+ if (!chunk_header) {
+ allocator_release_lock();
+ return NULL;
+ }
+
+ executable_offset = (sljit_sw)((sljit_u8*)chunk_header->executable - (sljit_u8*)chunk_header);
+
+ chunk_size -= sizeof(struct chunk_header) + sizeof(struct block_header);
+ total_size += chunk_size;
+
+ header = (struct block_header *)(chunk_header + 1);
+
+ header->prev_size = 0;
+ header->executable_offset = executable_offset;
+ if (chunk_size > size + 64) {
+ /* Cut the allocated space into a free and a used block. */
+ allocated_size += size;
+ header->size = size;
+ chunk_size -= size;
+
+ free_block = AS_FREE_BLOCK(header, size);
+ free_block->header.prev_size = size;
+ free_block->header.executable_offset = executable_offset;
+ sljit_insert_free_block(free_block, chunk_size);
+ next_header = AS_BLOCK_HEADER(free_block, chunk_size);
+ }
+ else {
+ /* All space belongs to this allocation. */
+ allocated_size += chunk_size;
+ header->size = chunk_size;
+ next_header = AS_BLOCK_HEADER(header, chunk_size);
+ }
+ next_header->size = 1;
+ next_header->prev_size = chunk_size;
+ next_header->executable_offset = executable_offset;
+ allocator_release_lock();
+ return MEM_START(header);
+}
+
+SLJIT_API_FUNC_ATTRIBUTE void sljit_free_exec(void* ptr)
+{
+ struct block_header *header;
+ struct free_block* free_block;
+
+ allocator_grab_lock();
+ header = AS_BLOCK_HEADER(ptr, -(sljit_sw)sizeof(struct block_header));
+ header = AS_BLOCK_HEADER(header, -header->executable_offset);
+ allocated_size -= header->size;
+
+ /* Connecting free blocks together if possible. */
+
+ /* If header->prev_size == 0, free_block will equal to header.
+ In this case, free_block->header.size will be > 0. */
+ free_block = AS_FREE_BLOCK(header, -(sljit_sw)header->prev_size);
+ if (SLJIT_UNLIKELY(!free_block->header.size)) {
+ free_block->size += header->size;
+ header = AS_BLOCK_HEADER(free_block, free_block->size);
+ header->prev_size = free_block->size;
+ }
+ else {
+ free_block = (struct free_block*)header;
+ sljit_insert_free_block(free_block, header->size);
+ }
+
+ header = AS_BLOCK_HEADER(free_block, free_block->size);
+ if (SLJIT_UNLIKELY(!header->size)) {
+ free_block->size += ((struct free_block*)header)->size;
+ sljit_remove_free_block((struct free_block*)header);
+ header = AS_BLOCK_HEADER(free_block, free_block->size);
+ header->prev_size = free_block->size;
+ }
+
+ /* The whole chunk is free. */
+ if (SLJIT_UNLIKELY(!free_block->header.prev_size && header->size == 1)) {
+ /* If this block is freed, we still have (allocated_size / 2) free space. */
+ if (total_size - free_block->size > (allocated_size * 3 / 2)) {
+ total_size -= free_block->size;
+ sljit_remove_free_block(free_block);
+ free_chunk(free_block, free_block->size + sizeof(struct block_header));
+ }
+ }
+
+ allocator_release_lock();
+}
+
+SLJIT_API_FUNC_ATTRIBUTE void sljit_free_unused_memory_exec(void)
+{
+ struct free_block* free_block;
+ struct free_block* next_free_block;
+
+ allocator_grab_lock();
+
+ free_block = free_blocks;
+ while (free_block) {
+ next_free_block = free_block->next;
+ if (!free_block->header.prev_size &&
+ AS_BLOCK_HEADER(free_block, free_block->size)->size == 1) {
+ total_size -= free_block->size;
+ sljit_remove_free_block(free_block);
+ free_chunk(free_block, free_block->size + sizeof(struct block_header));
+ }
+ free_block = next_free_block;
+ }
+
+ SLJIT_ASSERT((total_size && free_blocks) || (!total_size && !free_blocks));
+ allocator_release_lock();
+}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_sw sljit_exec_offset(void* ptr)
+{
+ return ((struct block_header *)(ptr))[-1].executable_offset;
+}
--
2.9.4

31
pcre2-10.30-RC1-Fix-formatting-converted_length.patch Normal file
View File

@ -0,0 +1,31 @@
From 0c0a69a8312a3fd9e20465940634a4caa6d84c07 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Petr=20P=C3=ADsa=C5=99?= <ppisar@redhat.com>
Date: Thu, 20 Jul 2017 13:14:19 +0200
Subject: [PATCH] Fix formatting converted_length
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
size_t width varies among platforms.
Signed-off-by: Petr Písař <ppisar@redhat.com>
---
src/pcre2test.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/pcre2test.c b/src/pcre2test.c
index 5d1cbdc..3775772 100644
--- a/src/pcre2test.c
+++ b/src/pcre2test.c
@@ -5466,7 +5466,7 @@ if (pat_patctl.convert_type != CONVERT_UNSET)
if (rc != 0)
{
- fprintf(outfile, "** Pattern conversion error at offset %lu: ",
+ fprintf(outfile, "** Pattern conversion error at offset %zu: ",
converted_length);
convert_return = print_error_message(rc, "", "\n")? PR_SKIP:PR_ABEND;
}
--
2.9.4

86
pcre2-10.30-RC1-Put-back-pcre2_set_recursion_limit-as-a-real-functio.patch Normal file
View File

@ -0,0 +1,86 @@
From e5a7fbea3d4478a57346135b8965c943a11da9c1 Mon Sep 17 00:00:00 2001
From: ph10 <ph10@6239d852-aaf2-0410-a92c-79f79f948069>
Date: Fri, 21 Jul 2017 08:53:02 +0000
Subject: [PATCH] Put back pcre2_set_recursion_limit() as a real function,
synonym for pcre2_set_depth_limit() in order to preserve the ABI.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
git-svn-id: svn://vcs.exim.org/pcre2/code/trunk@846 6239d852-aaf2-0410-a92c-79f79f948069
Petr Písař <ppisar@redhat.com>: Ported to 10.30-RC1
---
src/pcre2.h.generic | 4 +++-
src/pcre2.h.in | 4 +++-
src/pcre2_context.c | 8 +++++++-
diff --git a/src/pcre2.h.generic b/src/pcre2.h.generic
index 7cafb4c..287b12a 100644
--- a/src/pcre2.h.generic
+++ b/src/pcre2.h.generic
@@ -489,6 +489,8 @@ PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
pcre2_set_offset_limit(pcre2_match_context *, PCRE2_SIZE); \
PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
+ pcre2_set_recursion_limit(pcre2_match_context *, uint32_t); \
+PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
pcre2_set_recursion_memory_management(pcre2_match_context *, \
void *(*)(PCRE2_SIZE, void *), void (*)(void *, void *), void *);
@@ -760,7 +762,7 @@ pcre2_compile are called by application code. */
#define pcre2_substring_number_from_name PCRE2_SUFFIX(pcre2_substring_number_from_name_)
/* Keep this old function name for backwards compatibility */
-#define pcre2_set_recursion_limit pcre2_set_depth_limit
+#define pcre2_set_recursion_limit PCRE2_SUFFIX(pcre2_set_recursion_limit_)
/* Keep this obsolete function for backwards compatibility: it is now a noop. */
#define pcre2_set_recursion_memory_management PCRE2_SUFFIX(pcre2_set_recursion_memory_management_)
diff --git a/src/pcre2.h.in b/src/pcre2.h.in
index 399ddcd..4a8e126 100644
--- a/src/pcre2.h.in
+++ b/src/pcre2.h.in
@@ -489,6 +489,8 @@ PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
pcre2_set_offset_limit(pcre2_match_context *, PCRE2_SIZE); \
PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
+ pcre2_set_recursion_limit(pcre2_match_context *, uint32_t); \
+PCRE2_EXP_DECL int PCRE2_CALL_CONVENTION \
pcre2_set_recursion_memory_management(pcre2_match_context *, \
void *(*)(PCRE2_SIZE, void *), void (*)(void *, void *), void *);
@@ -760,7 +762,7 @@ pcre2_compile are called by application code. */
#define pcre2_substring_number_from_name PCRE2_SUFFIX(pcre2_substring_number_from_name_)
/* Keep this old function name for backwards compatibility */
-#define pcre2_set_recursion_limit pcre2_set_depth_limit
+#define pcre2_set_recursion_limit PCRE2_SUFFIX(pcre2_set_recursion_limit_)
/* Keep this obsolete function for backwards compatibility: it is now a noop. */
#define pcre2_set_recursion_memory_management PCRE2_SUFFIX(pcre2_set_recursion_memory_management_)
diff --git a/src/pcre2_context.c b/src/pcre2_context.c
index 9e557da..2c14df0 100644
--- a/src/pcre2_context.c
+++ b/src/pcre2_context.c
@@ -431,10 +431,16 @@ mcontext->offset_limit = limit;
return 0;
}
-/* This function became obsolete at release 10.30. It is kept as a no-op for
+/* This function became obsolete at release 10.30. It is kept as a synonym for
backwards compatibility. */
PCRE2_EXP_DEFN int PCRE2_CALL_CONVENTION
+pcre2_set_recursion_limit(pcre2_match_context *mcontext, uint32_t limit)
+{
+return pcre2_set_depth_limit(mcontext, limit);
+}
+
+PCRE2_EXP_DEFN int PCRE2_CALL_CONVENTION
pcre2_set_recursion_memory_management(pcre2_match_context *mcontext,
void *(*mymalloc)(size_t, void *), void (*myfree)(void *, void *),
void *mydata)
--
2.9.4

104
pcre2.spec
View File

@ -1,16 +1,18 @@
# This is stable release:
#%%global rcversion RC1
%global rcversion RC1
Name: pcre2
Version: 10.23
Release: %{?rcversion:0.}8%{?rcversion:.%rcversion}%{?dist}
Version: 10.30
Release: %{?rcversion:0.}1%{?rcversion:.%rcversion}%{?dist}
%global myversion %{version}%{?rcversion:-%rcversion}
Summary: Perl-compatible regular expression library
Group: System Environment/Libraries
# the library: BSD
# the library: BSD with exceptions
# pcre2test (linked to GNU readline): BSD (linked to GPLv3+)
# COPYING: see LICENCE file
# LICENSE: BSD text and declares Public Domain
# LICENSE: BSD text with exceptions and
# Public Domain declaration
# for testdata
#Bundled
# src/sljit: BSD
#Not distributed in binary package
# aclocal.m4: FSFULLR and GPLv2+ with exception
# ar-lib: GPLv2+ with exception
@ -36,57 +38,15 @@ URL: http://www.pcre.org/
Source: ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre/%{?rcversion:Testing/}%{name}-%{myversion}.tar.bz2
# Do no set RPATH if libdir is not /usr/lib
Patch0: pcre2-10.10-Fix-multilib.patch
# Handle memmory allocation failures in pcre2test tool, in upstream after 10.23
Patch1: pcre2-10.23-Check-malloc-returns-in-pcre2test.patch
# Fix CVE-2017-7186 (a crash when finding a Unicode property for a character
# with a code point greater than 0x10ffff in UTF-32 library while UTF mode is
# disabled), upstream bug #2052, in upstream after 10.23
Patch2: pcre2-10.23-Fix-32-bit-non-UTF-property-test-crash.patch
# Fix an internal error for a forward reference in a lookbehind with
# PCRE2_ANCHORED, # oss-fuzz bug #865, in upstream after 10.23
Patch3: pcre2-10.23-Fix-crash-for-forward-reference-in-lookbehind-with-P.patch
# Fix a pcre2test bug for global match with zero terminated subject,
# upstream bug #2063, in upstream after 10.23
Patch4: pcre2-10.23-Fix-pcre2test-bug-for-global-match-with-zero-termina.patch
# Close serialization file in pcre2test after any error, upstream bug #2074,
# in upstream after 10.23
Patch5: pcre2-10.23-Close-serialization-file-in-pcre2test-after-any-erro.patch
# Fix a memory leak in pcre2_serialize_decode() when the input is invalid,
# upstream bug #2075, in upsream after 10.23.
Patch6: pcre2-10.23-Fix-memory-leak-when-deserializing-invalid-data-Bugz.patch
# Fix a potential NULL dereference in pcre2_callout_enumerate() if called with
# a NULL pattern pointer when Unicode support is available, upstream bug #2076,
# in upstream after 10.23
Patch7: pcre2-10.23-Fix-NULL-deference-if-pcre2_callout_enumerate-is-cal.patch
# 1/2 Fix 32-bit error buffer size bug in pcre2test, upstream bug #2079,
# in upstream after 10.23
Patch8: pcre2-10.23-Fix-32-bit-error-buffer-size-bug-in-pcre2test-Bugzil.patch
# 2/2 Fix 32-bit error buffer size bug in pcre2test, upstream bug #2079,
# in upstream after 10.23
Patch9: pcre2-10.23-Previous-patch-was-not-quite-complete.patch
# Fix DFA match for a possessively repeated character class, upstream bug #2086,
# in upstream after 10.23
Patch10: pcre2-10.23-Fix-misbehaving-DFA-match-for-possessively-repeated-.patch
# Use a memory allocator from the pattern if no context is supplied to
# pcre2_match(), in upsream after 10.23
Patch11: pcre2-10.23-Fix-bug-introduced-at-10.21-use-memory-allocator-fro.patch
# Fix CVE-2017-7186 in JIT mode (a crash when finding a Unicode property for
# a character with a code point greater than 0x10ffff in UTF-32 library while
# UTF mode is disabled), bug #1434504, upstream bug #2052,
# in upstream after 10.23
Patch12: pcre2-10.23-Fix-character-type-detection-when-32-bit-and-UCP-are.patch
# Fix an incorrect cast in UTF validation, upstream bug #2090,
# in upstream after 10.23
Patch13: pcre2-10.23-Correct-an-incorrect-cast.patch
# Fix a pcre2test crash on multiple push statements, upstream bug #2109,
# in upstream after 10.23
Patch14: pcre2-10.23-Fix-crash-when-more-than-one-kind-of-push-was-set-in.patch
# Fix DFA matching a lookbehind assertion that has a zero-length branch,
# PCRE2 oss-fuzz issue 1859, in upstream after 10.23
Patch15: pcre2-10.23-Fix-lookbehind-with-zero-length-branch-in-DFA-matchi.patch
# Fix returned offsets from regexec() when REG_STARTEND is used with starting offset
# greater than zero, upstream bug #2128, in upstream after 10.23
Patch16: pcre2-10.23-Fix-matching-offsets-from-regexec-in-the-POSIX-wrapp.patch
# Add forgotten sources for SELinux JIT allocator, in upstream after 10.23-RC1,
# <https://lists.exim.org/lurker/message/20170720.103409.cb67b89d.en.html>
Patch1: pcre2-10.30-RC1-Distribute-forgotten-sljitProtExecAllocator.c.patch
# Preserve pcre2-10.23 ABI, in upstream after 10.23-RC1
# <https://lists.exim.org/lurker/message/20170720.111013.7f8f78d3.en.html>
Patch2: pcre2-10.30-RC1-Put-back-pcre2_set_recursion_limit-as-a-real-functio.patch
# Correct formatting a size_t variable, in upstream after 10.23-RC1,
# <https://lists.exim.org/lurker/message/20170720.112359.fa1910da.en.html>
Patch3: pcre2-10.30-RC1-Fix-formatting-converted_length.patch
BuildRequires: autoconf
BuildRequires: automake
BuildRequires: coreutils
@ -94,6 +54,7 @@ BuildRequires: gcc
BuildRequires: libtool
BuildRequires: make
BuildRequires: readline-devel
Provides: bundled(sljit)
%description
PCRE2 is a re-working of the original PCRE (Perl-compatible regular
@ -115,7 +76,7 @@ restricted, and does not give full access to all of PCRE2's facilities.
%package utf16
Summary: UTF-16 variant of PCRE2
Group: Development/Libraries
Provides: bundled(sljit)
Conflicts: %{name}%{?_isa} < 10.21-4
%description utf16
@ -123,7 +84,7 @@ This is PCRE2 library working on UTF-16 strings.
%package utf32
Summary: UTF-32 variant of PCRE2
Group: Development/Libraries
Provides: bundled(sljit)
Conflicts: %{name}%{?_isa} < 10.21-4
%description utf32
@ -131,7 +92,6 @@ This is PCRE2 library working on UTF-32 strings.
%package devel
Summary: Development files for %{name}
Group: Development/Libraries
Requires: %{name}%{?_isa} = %{version}-%{release}
Requires: %{name}-utf16%{?_isa} = %{version}-%{release}
Requires: %{name}-utf32%{?_isa} = %{version}-%{release}
@ -144,8 +104,8 @@ pcre2posix.h.
%package static
Summary: Static library for %{name}
Group: Development/Libraries
Requires: %{name}-devel%{_isa} = %{version}-%{release}
Provides: bundled(sljit)
%description static
Library for static linking for %{name}.
@ -154,7 +114,6 @@ Library for static linking for %{name}.
Summary: Auxiliary utilities for %{name}
# pcre2test (linked to GNU readline): BSD (linked to GPLv3+)
License: BSD and GPLv3+
Group: Development/Tools
Requires: %{name}%{_isa} = %{version}-%{release}
%description tools
@ -166,19 +125,6 @@ Utilities demonstrating PCRE2 capabilities like pcre2grep or pcre2test.
%patch1 -p1
%patch2 -p1
%patch3 -p1
%patch4 -p1
%patch5 -p1
%patch6 -p1
%patch7 -p1
%patch8 -p1
%patch9 -p1
%patch10 -p1
%patch11 -p1
%patch12 -p1
%patch13 -p1
%patch14 -p1
%patch15 -p1
%patch16 -p1
# Because of multilib patch
libtoolize --copy --force
autoreconf -vif
@ -200,6 +146,7 @@ autoreconf -vif
--disable-coverage \
--disable-ebcdic \
--disable-fuzz-support \
--enable-jit-sealloc \
--disable-never-backslash-C \
--enable-newline-is-lf \
--enable-pcre2-8 \
@ -213,7 +160,7 @@ autoreconf -vif
--enable-pcre2test-libreadline \
--disable-rebuild-chartables \
--enable-shared \
--enable-stack-for-recursion \
--disable-silent-rules \
--enable-static \
--enable-unicode \
--disable-valgrind
@ -277,6 +224,11 @@ make %{?_smp_mflags} check VERBOSE=yes
%{_mandir}/man1/pcre2test.*
%changelog
* Thu Jul 20 2017 Petr Pisar <ppisar@redhat.com> - 10.30-0.1.RC1
- 10.30-RC1 bump
- Heap-based matching implementation replaced stack-based one
- SELinux-friendly JIT enabled
* Fri Jun 16 2017 Petr Pisar <ppisar@redhat.com> - 10.23-8
- Fix DFA matching a lookbehind assertion that has a zero-length branch
(PCRE2 oss-fuzz issue 1859)

2
sources
View File

@ -1 +1 @@
SHA512 (pcre2-10.23.tar.bz2) = 3e5910bd2405cc35934d91e4be760abe4f2e900202a20b6ba74adb7a3acb2b74b3bf9b0e97e8de10f8e8534133e0722e0bf0f5fb40d6c2c4520d1ed61749d456
SHA512 (pcre2-10.30-RC1.tar.bz2) = f72722e12822cb9f2831980cbc786536b1394350cac542cf280e8c14cf9a10386bb0ee2cd73ceba7d9c521ff8f1e0e79a8d89aaf93638ceb0cd7f9eabbfa2d07