archive/php-src
1
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2025-08-16 05:58:45 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

Bundle pcre-8.35 with sljit support

Browse source
This commit is contained in:
Dmitry Stogov 2014-06-06 18:46:19 +04:00
parent e499c3d259
commit dd0e96cca3
93 changed files with 47922 additions and 1852 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ext/pcre/config.w32
View file

@ -3,7 +3,7 @@
EXTENSION("pcre", "php_pcre.c", false /* never shared */,
"-Iext/pcre/pcrelib");
ADD_SOURCES("ext/pcre/pcrelib", "pcre_chartables.c pcre_ucd.c pcre_compile.c pcre_config.c pcre_exec.c pcre_fullinfo.c pcre_get.c pcre_globals.c pcre_maketables.c pcre_newline.c pcre_ord2utf8.c pcre_refcount.c pcre_study.c pcre_tables.c pcre_valid_utf8.c pcre_version.c pcre_xclass.c", "pcre");
ADD_SOURCES("ext/pcre/pcrelib", "pcre_chartables.c pcre_ucd.c pcre_compile.c pcre_config.c pcre_exec.c pcre_fullinfo.c pcre_get.c pcre_globals.c pcre_maketables.c pcre_newline.c pcre_ord2utf8.c pcre_refcount.c pcre_study.c pcre_tables.c pcre_valid_utf8.c pcre_version.c pcre_xclass.c pcre_jit_compile.c", "pcre");
ADD_DEF_FILE("ext\\pcre\\php_pcre.def");
AC_DEFINE('HAVE_BUNDLED_PCRE', 1, 'Using bundled PCRE library');

3
ext/pcre/config0.m4
View file

@ -58,7 +58,8 @@ PHP_ARG_WITH(pcre-regex,,
pcrelib/pcre_maketables.c pcrelib/pcre_newline.c \
pcrelib/pcre_ord2utf8.c pcrelib/pcre_refcount.c pcrelib/pcre_study.c \
pcrelib/pcre_tables.c pcrelib/pcre_valid_utf8.c \
pcrelib/pcre_version.c pcrelib/pcre_xclass.c"
pcrelib/pcre_version.c pcrelib/pcre_xclass.c \
pcrelib/pcre_jit_compile.c"
PHP_PCRE_CFLAGS="-DHAVE_CONFIG_H -I@ext_srcdir@/pcrelib"
PHP_NEW_EXTENSION(pcre, $pcrelib_sources php_pcre.c, no,,$PHP_PCRE_CFLAGS)
PHP_ADD_BUILD_DIR($ext_builddir/pcrelib)

6
ext/pcre/pcrelib/AUTHORS
View file

@ -8,7 +8,7 @@ Email domain: cam.ac.uk
University of Cambridge Computing Service,
Cambridge, England.
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
All rights reserved
@ -19,7 +19,7 @@ Written by: Zoltan Herczeg
Email local part: hzmester
Emain domain: freemail.hu
Copyright(c) 2010-2013 Zoltan Herczeg
Copyright(c) 2010-2014 Zoltan Herczeg
All rights reserved.
@ -30,7 +30,7 @@ Written by: Zoltan Herczeg
Email local part: hzmester
Emain domain: freemail.hu
Copyright(c) 2009-2013 Zoltan Herczeg
Copyright(c) 2009-2014 Zoltan Herczeg
All rights reserved.

120
ext/pcre/pcrelib/ChangeLog
View file

@ -1,6 +1,126 @@
ChangeLog for PCRE
------------------
Version 8.35 04-April-2014
--------------------------
1. A new flag is set, when property checks are present in an XCLASS.
When this flag is not set, PCRE can perform certain optimizations
such as studying these XCLASS-es.
2. The auto-possessification of character sets were improved: a normal
and an extended character set can be compared now. Furthermore
the JIT compiler optimizes more character set checks.
3. Got rid of some compiler warnings for potentially uninitialized variables
that show up only when compiled with -O2.
4. A pattern such as (?=ab\K) that uses \K in an assertion can set the start
of a match later then the end of the match. The pcretest program was not
handling the case sensibly - it was outputting from the start to the next
binary zero. It now reports this situation in a message, and outputs the
text from the end to the start.
5. Fast forward search is improved in JIT. Instead of the first three
characters, any three characters with fixed position can be searched.
Search order: first, last, middle.
6. Improve character range checks in JIT. Characters are read by an inprecise
function now, which returns with an unknown value if the character code is
above a certain treshold (e.g: 256). The only limitation is that the value
must be bigger than the treshold as well. This function is useful, when
the characters above the treshold are handled in the same way.
7. The macros whose names start with RAWUCHAR are placeholders for a future
mode in which only the bottom 21 bits of 32-bit data items are used. To
make this more memorable for those maintaining the code, the names have
been changed to start with UCHAR21, and an extensive comment has been added
to their definition.
8. Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
to the export list in Makefile.am (they were accidentally omitted from the
8.34 tarball).
9. The informational output from pcretest used the phrase "starting byte set"
which is inappropriate for the 16-bit and 32-bit libraries. As the output
for "first char" and "need char" really means "non-UTF-char", I've changed
"byte" to "char", and slightly reworded the output. The documentation about
these values has also been (I hope) clarified.
10. Another JIT related optimization: use table jumps for selecting the correct
backtracking path, when more than four alternatives are present inside a
bracket.
11. Empty match is not possible, when the minimum length is greater than zero,
and there is no \K in the pattern. JIT should avoid empty match checks in
such cases.
12. In a caseless character class with UCP support, when a character with more
than one alternative case was not the first character of a range, not all
the alternative cases were added to the class. For example, s and \x{17f}
are both alternative cases for S: the class [RST] was handled correctly,
but [R-T] was not.
13. The configure.ac file always checked for pthread support when JIT was
enabled. This is not used in Windows, so I have put this test inside a
check for the presence of windows.h (which was already tested for).
14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
The algorithm provides a way to skip certain starting offsets, and usually
faster than linear prefix searches.
15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
as for 'fr_FR' and 'french'. For some reason, however, it then used the
Windows-specific input and output files, which have 'french' screwed in.
So this could never have worked. One of the problems with locales is that
they aren't always the same. I have now updated RunTest so that it checks
the output of the locale test (test 3) against three different output
files, and it allows the test to pass if any one of them matches. With luck
this should make the test pass on some versions of Solaris where it was
failing. Because of the uncertainty, the script did not used to stop if
test 3 failed; it now does. If further versions of a French locale ever
come to light, they can now easily be added.
16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
there was a message during ./configure, but it did not stop. This now
provokes an error. The invalid example in README has been corrected.
If a value less than the minimum is given, the minimum value has always
been used, but now a warning is given.
17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
was a bug in the test system, which is now fixed. Also, the list of various
configurations that are tested for each release did not have one with both
16/32 bits and --enable-bar-anycrlf. It now does.
18. pcretest was missing "-C bsr" for displaying the \R default setting.
19. Little endian PowerPC systems are supported now by the JIT compiler.
20. The fast forward newline mechanism could enter to an infinite loop on
certain invalid UTF-8 input. Although we don't support these cases
this issue can be fixed by a performance optimization.
21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
not take account if existing stack usage. There is now a new global
variable called pcre_stack_guard that can be set to point to an external
function to check stack availability. It is called at the start of
processing every parenthesized group.
22. A typo in the code meant that in ungreedy mode the max/min qualifier
behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
not match "ab".
23. When UTF was disabled, the JIT program reported some incorrect compile
errors. These messages are silenced now.
24. Experimental support for ARM-64 and MIPS-64 has been added to the JIT
compiler.
25. Change all the temporary files used in RunGrepTest to be different to those
used by RunTest so that the tests can be run simultaneously, for example by
"make -j check".
Version 8.34 15-December-2013
-----------------------------

6
ext/pcre/pcrelib/LICENCE
View file

@ -24,7 +24,7 @@ Email domain: cam.ac.uk
University of Cambridge Computing Service,
Cambridge, England.
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
All rights reserved.
@ -35,7 +35,7 @@ Written by: Zoltan Herczeg
Email local part: hzmester
Emain domain: freemail.hu
Copyright(c) 2010-2013 Zoltan Herczeg
Copyright(c) 2010-2014 Zoltan Herczeg
All rights reserved.
@ -46,7 +46,7 @@ Written by: Zoltan Herczeg
Email local part: hzmester
Emain domain: freemail.hu
Copyright(c) 2009-2013 Zoltan Herczeg
Copyright(c) 2009-2014 Zoltan Herczeg
All rights reserved.

11
ext/pcre/pcrelib/NEWS
View file

@ -1,6 +1,17 @@
News about PCRE releases
------------------------
Release 8.35 04-April-2014
--------------------------
There have been performance improvements for classes containing non-ASCII
characters and the "auto-possessification" feature has been extended. Other
minor improvements have been implemented and bugs fixed. There is a new callout
feature to enable applications to do detailed stack checks at compile time, to
avoid running out of stack for deeply nested parentheses. The JIT compiler has
been extended with experimental support for ARM-64, MIPS-64, and PPC-LE.
Release 8.34 15-December-2013
-----------------------------

21
ext/pcre/pcrelib/README
View file

@ -85,11 +85,12 @@ documentation is supplied in two other forms:
1. There are files called doc/pcre.txt, doc/pcregrep.txt, and
doc/pcretest.txt in the source distribution. The first of these is a
concatenation of the text forms of all the section 3 man pages except
those that summarize individual functions. The other two are the text
forms of the section 1 man pages for the pcregrep and pcretest commands.
These text forms are provided for ease of scanning with text editors or
similar tools. They are installed in <prefix>/share/doc/pcre, where
<prefix> is the installation prefix (defaulting to /usr/local).
the listing of pcredemo.c and those that summarize individual functions.
The other two are the text forms of the section 1 man pages for the
pcregrep and pcretest commands. These text forms are provided for ease of
scanning with text editors or similar tools. They are installed in
<prefix>/share/doc/pcre, where <prefix> is the installation prefix
(defaulting to /usr/local).
2. A set of files containing all the documentation in HTML form, hyperlinked
in various ways, and rooted in a file called index.html, is distributed in
@ -372,12 +373,12 @@ library. They are also documented in the pcrebuild man page.
Of course, the relevant libraries must be installed on your system.
. The default size of internal buffer used by pcregrep can be set by, for
example:
. The default size (in bytes) of the internal buffer used by pcregrep can be
set by, for example:
--with-pcregrep-bufsize=50K
--with-pcregrep-bufsize=51200
The default value is 20K.
The value must be a plain integer. The default is 20480.
. It is possible to compile pcretest so that it links with the libreadline
or libedit libraries, by specifying, respectively,
@ -987,4 +988,4 @@ pcre_xxx, one with the name pcre16_xx, and a third with the name pcre32_xxx.
Philip Hazel
Email local part: ph10
Email domain: cam.ac.uk
Last updated: 05 November 2013
Last updated: 17 January 2014

216
ext/pcre/pcrelib/config.h
View file

@ -24,37 +24,41 @@
#endif
/* Exclude these below definitions when building within PHP */
#ifndef ZEND_API
/* config.h. Generated from config.h.in by configure. */
/* config.h.in. Generated from configure.ac by autoheader. */
/* PCRE is written in Standard C, but there are a few non-standard things it
can cope with, allowing it to run on SunOS4 and other "close to standard"
systems.
In environments that support the facilities, config.h.in is converted by
"configure", or config-cmake.h.in is converted by CMake, into config.h. If you
are going to build PCRE "by hand" without using "configure" or CMake, you
should copy the distributed config.h.generic to config.h, and then edit the
macro definitions to be the way you need them. You must then add
-DHAVE_CONFIG_H to all of your compile commands, so that config.h is included
at the start of every source.
In environments that support the GNU autotools, config.h.in is converted into
config.h by the "configure" script. In environments that use CMake,
config-cmake.in is converted into config.h. If you are going to build PCRE "by
hand" without using "configure" or CMake, you should copy the distributed
config.h.generic to config.h, and edit the macro definitions to be the way you
need them. You must then add -DHAVE_CONFIG_H to all of your compile commands,
so that config.h is included at the start of every source.
Alternatively, you can avoid editing by using -D on the compiler command line
to set the macro values. In this case, you do not have to set -DHAVE_CONFIG_H.
to set the macro values. In this case, you do not have to set -DHAVE_CONFIG_H,
but if you do, default values will be taken from config.h for non-boolean
macros that are not defined on the command line.
PCRE uses memmove() if HAVE_MEMMOVE is set to 1; otherwise it uses bcopy() if
HAVE_BCOPY is set to 1. If your system has neither bcopy() nor memmove(), set
them both to 0; an emulation function will be used. */
Boolean macros such as HAVE_STDLIB_H and SUPPORT_PCRE8 should either be defined
(conventionally to 1) for TRUE, and not defined at all for FALSE. All such
macros are listed as a commented #undef in config.h.generic. Macros such as
MATCH_LIMIT, whose actual value is relevant, have defaults defined, but are
surrounded by #ifndef/#endif lines so that the value can be overridden by -D.
PCRE uses memmove() if HAVE_MEMMOVE is defined; otherwise it uses bcopy() if
HAVE_BCOPY is defined. If your system has neither bcopy() nor memmove(), make
sure both macros are undefined; an emulation function will then be used. */
/* By default, the \R escape sequence matches any Unicode line ending
character or sequence of characters. If BSR_ANYCRLF is defined (to any
value), this is changed so that backslash-R matches only CR, LF, or CRLF.
The build-time default can be overridden by the user of PCRE at runtime. */
#undef BSR_ANYCRLF
/* #undef BSR_ANYCRLF */
/* If you are compiling for a system that uses EBCDIC instead of ASCII
character codes, define this macro to any value. You must also edit the
@ -64,113 +68,80 @@ them both to 0; an emulation function will be used. */
strings are in EBCDIC. If you do not define this macro, PCRE will assume
input strings are ASCII or UTF-8/16/32 Unicode. It is not possible to build
a version of PCRE that supports both EBCDIC and UTF-8/16/32. */
#undef EBCDIC
/* #undef EBCDIC */
/* In an EBCDIC environment, define this macro to any value to arrange for the
NL character to be 0x25 instead of the default 0x15. NL plays the role that
LF does in an ASCII/Unicode environment. The value must also be set in the
NEWLINE macro below. On systems that can use "configure" or CMake to set
EBCDIC_NL25, the adjustment of NEWLINE is automatic. */
#undef EBCDIC_NL25
/* #undef EBCDIC_NL25 */
/* Define to 1 if you have the `bcopy' function. */
#ifndef HAVE_BCOPY
#define HAVE_BCOPY 1
#endif
/* #undef HAVE_BCOPY */
/* Define to 1 if you have the <bits/type_traits.h> header file. */
/* #undef HAVE_BITS_TYPE_TRAITS_H */
/* Define to 1 if you have the <bzlib.h> header file. */
#ifndef HAVE_BZLIB_H
#define HAVE_BZLIB_H 1
#endif
/* #undef HAVE_BZLIB_H */
/* Define to 1 if you have the <dirent.h> header file. */
#ifndef HAVE_DIRENT_H
#define HAVE_DIRENT_H 1
#endif
/* #undef HAVE_DIRENT_H */
/* Define to 1 if you have the <dlfcn.h> header file. */
#ifndef HAVE_DLFCN_H
#define HAVE_DLFCN_H 1
#endif
/* #undef HAVE_DLFCN_H */
/* Define to 1 if you have the <editline/readline.h> header file. */
/*#undef HAVE_EDITLINE_READLINE_H*/
/* #undef HAVE_EDITLINE_READLINE_H */
/* Define to 1 if you have the <edit/readline/readline.h> header file. */
/* #undef HAVE_EDIT_READLINE_READLINE_H */
/* Define to 1 if you have the <inttypes.h> header file. */
#ifndef HAVE_INTTYPES_H
#define HAVE_INTTYPES_H 1
#endif
/* #undef HAVE_INTTYPES_H */
/* Define to 1 if you have the <limits.h> header file. */
#ifndef HAVE_LIMITS_H
#define HAVE_LIMITS_H 1
#endif
/* #undef HAVE_LIMITS_H */
/* Define to 1 if the system has the type `long long'. */
#ifndef HAVE_LONG_LONG
#define HAVE_LONG_LONG 1
#endif
/* #undef HAVE_LONG_LONG */
/* Define to 1 if you have the `memmove' function. */
#ifndef HAVE_MEMMOVE
#define HAVE_MEMMOVE 1
#endif
/* #undef HAVE_MEMMOVE */
/* Define to 1 if you have the <memory.h> header file. */
#ifndef HAVE_MEMORY_H
#define HAVE_MEMORY_H 1
#endif
/* #undef HAVE_MEMORY_H */
/* Define if you have POSIX threads libraries and header files. */
#undef HAVE_PTHREAD
/* #undef HAVE_PTHREAD */
/* Have PTHREAD_PRIO_INHERIT. */
#undef HAVE_PTHREAD_PRIO_INHERIT
/* #undef HAVE_PTHREAD_PRIO_INHERIT */
/* Define to 1 if you have the <readline/history.h> header file. */
#ifndef HAVE_READLINE_HISTORY_H
#define HAVE_READLINE_HISTORY_H 1
#endif
/* #undef HAVE_READLINE_HISTORY_H */
/* Define to 1 if you have the <readline/readline.h> header file. */
#ifndef HAVE_READLINE_READLINE_H
#define HAVE_READLINE_READLINE_H 1
#endif
/* #undef HAVE_READLINE_READLINE_H */
/* Define to 1 if you have the <stdint.h> header file. */
#ifndef HAVE_STDINT_H
#define HAVE_STDINT_H 1
#endif
/* #undef HAVE_STDINT_H */
/* Define to 1 if you have the <stdlib.h> header file. */
#ifndef HAVE_STDLIB_H
#define HAVE_STDLIB_H 1
#endif
/* #undef HAVE_STDLIB_H */
/* Define to 1 if you have the `strerror' function. */
#ifndef HAVE_STRERROR
#define HAVE_STRERROR 1
#endif
/* #undef HAVE_STRERROR */
/* Define to 1 if you have the <string> header file. */
#ifndef HAVE_STRING
#define HAVE_STRING 1
#endif
/* #undef HAVE_STRING */
/* Define to 1 if you have the <strings.h> header file. */
#ifndef HAVE_STRINGS_H
#define HAVE_STRINGS_H 1
#endif
/* #undef HAVE_STRINGS_H */
/* Define to 1 if you have the <string.h> header file. */
#ifndef HAVE_STRING_H
#define HAVE_STRING_H 1
#endif
/* #undef HAVE_STRING_H */
/* Define to 1 if you have `strtoimax'. */
/* #undef HAVE_STRTOIMAX */
@ -179,63 +150,47 @@ them both to 0; an emulation function will be used. */
/* #undef HAVE_STRTOLL */
/* Define to 1 if you have `strtoq'. */
#ifndef HAVE_STRTOQ
#define HAVE_STRTOQ 1
#endif
/* #undef HAVE_STRTOQ */
/* Define to 1 if you have the <sys/stat.h> header file. */
#ifndef HAVE_SYS_STAT_H
#define HAVE_SYS_STAT_H 1
#endif
/* #undef HAVE_SYS_STAT_H */
/* Define to 1 if you have the <sys/types.h> header file. */
#ifndef HAVE_SYS_TYPES_H
#define HAVE_SYS_TYPES_H 1
#endif
/* #undef HAVE_SYS_TYPES_H */
/* Define to 1 if you have the <type_traits.h> header file. */
/* #undef HAVE_TYPE_TRAITS_H */
/* Define to 1 if you have the <unistd.h> header file. */
#ifndef HAVE_UNISTD_H
#define HAVE_UNISTD_H 1
#endif
/* #undef HAVE_UNISTD_H */
/* Define to 1 if the system has the type `unsigned long long'. */
#ifndef HAVE_UNSIGNED_LONG_LONG
#define HAVE_UNSIGNED_LONG_LONG 1
#endif
/* #undef HAVE_UNSIGNED_LONG_LONG */
/* Define to 1 or 0, depending whether the compiler supports simple visibility
declarations. */
/* Define to 1 if the compiler supports simple visibility declarations. */
/* #undef HAVE_VISIBILITY */
/* Define to 1 if you have the <windows.h> header file. */
/* #undef HAVE_WINDOWS_H */
/* Define to 1 if you have the <zlib.h> header file. */
#ifndef HAVE_ZLIB_H
#define HAVE_ZLIB_H 1
#endif
/* #undef HAVE_ZLIB_H */
/* Define to 1 if you have `_strtoi64'. */
/* #undef HAVE__STRTOI64 */
/* Exclude these above definitions when building within PHP */
#endif
/* The value of LINK_SIZE determines the number of bytes used to store links
as offsets within the compiled regex. The default is 2, which allows for
compiled patterns up to 64K long. This covers the vast majority of cases.
However, PCRE can also be compiled to use 3 or 4 bytes instead. This allows
for longer patterns in extreme cases. On systems that support it,
"configure" can be used to override this default. */
for longer patterns in extreme cases. */
#ifndef LINK_SIZE
#define LINK_SIZE 2
#endif
/* Define to the sub-directory in which libtool stores uninstalled libraries.
*/
/* This is ignored unless you are using libtool. */
#ifndef LT_OBJDIR
#define LT_OBJDIR ".libs/"
#endif
@ -245,8 +200,7 @@ them both to 0; an emulation function will be used. */
pcre_exec(). There is a runtime interface for setting a different limit.
The limit exists in order to catch runaway regular expressions that take
for ever to determine that they do not match. The default is set very large
so that it does not accidentally catch legitimate cases. On systems that
support it, "configure" can be used to override this default default. */
so that it does not accidentally catch legitimate cases. */
#ifndef MATCH_LIMIT
#define MATCH_LIMIT 10000000
#endif
@ -258,8 +212,7 @@ them both to 0; an emulation function will be used. */
used. The value of MATCH_LIMIT_RECURSION applies only to recursive calls of
match(). To have any useful effect, it must be less than the value of
MATCH_LIMIT. The default is to use the same value as MATCH_LIMIT. There is
a runtime method for setting a different limit. On systems that support it,
"configure" can be used to override the default. */
a runtime method for setting a different limit. */
#ifndef MATCH_LIMIT_RECURSION
#define MATCH_LIMIT_RECURSION MATCH_LIMIT
#endif
@ -290,9 +243,6 @@ them both to 0; an emulation function will be used. */
#define NEWLINE 10
#endif
/* Define to 1 if your C compiler doesn't accept -c and -o together. */
/* #undef NO_MINUS_C_MINUS_O */
/* PCRE uses recursive function calls to handle backtracking while matching.
This can sometimes be a problem on systems that have stacks of limited
size. Define NO_RECURSE to any value to get a version that doesn't use
@ -302,8 +252,6 @@ them both to 0; an emulation function will be used. */
*/
/* #undef NO_RECURSE */
#define PARENS_NEST_LIMIT 250
/* Name of package */
#define PACKAGE "pcre"
@ -314,7 +262,7 @@ them both to 0; an emulation function will be used. */
#define PACKAGE_NAME "PCRE"
/* Define to the full name and version of this package. */
#define PACKAGE_STRING "PCRE 8.32"
#define PACKAGE_STRING "PCRE 8.35"
/* Define to the one symbol short name of this package. */
#define PACKAGE_TARNAME "pcre"
@ -323,33 +271,23 @@ them both to 0; an emulation function will be used. */
#define PACKAGE_URL ""
/* Define to the version of this package. */
#define PACKAGE_VERSION "8.32"
#define PACKAGE_VERSION "8.35"
/* to make a symbol visible */
/* #undef PCRECPP_EXP_DECL */
/* to make a symbol visible */
/* #undef PCRECPP_EXP_DEFN */
/* The value of PARENS_NEST_LIMIT specifies the maximum depth of nested
parentheses (of any kind) in a pattern. This limits the amount of system
stack that is used while compiling a pattern. */
#ifndef PARENS_NEST_LIMIT
#define PARENS_NEST_LIMIT 250
#endif
/* The value of PCREGREP_BUFSIZE determines the size of buffer used by
pcregrep to hold parts of the file it is searching. This is also the
minimum value. The actual amount of memory used by pcregrep is three times
this number, because it allows for the buffering of "before" and "after"
lines. */
/* #undef PCREGREP_BUFSIZE */
/* to make a symbol visible */
/* #undef PCREPOSIX_EXP_DECL */
/* to make a symbol visible */
/* #undef PCREPOSIX_EXP_DEFN */
/* to make a symbol visible */
/* #undef PCRE_EXP_DATA_DEFN */
/* to make a symbol visible */
/* #undef PCRE_EXP_DECL */
#ifndef PCREGREP_BUFSIZE
#define PCREGREP_BUFSIZE 20480
#endif
/* If you are compiling for a system other than a Unix-like system or
Win32, and it needs some magic to be inserted before the definition
@ -381,23 +319,17 @@ them both to 0; an emulation function will be used. */
/* #undef PTHREAD_CREATE_JOINABLE */
/* Define to 1 if you have the ANSI C header files. */
#ifndef STDC_HEADERS
#define STDC_HEADERS 1
#endif
/* Define to allow pcretest and pcregrep to be linked with gcov, so that they
are able to generate code coverage reports. */
#undef SUPPORT_GCOV
/* #undef STDC_HEADERS */
/* Define to any value to enable support for Just-In-Time compiling. */
#undef SUPPORT_JIT
#define SUPPORT_JIT
/* Define to any value to allow pcregrep to be linked with libbz2, so that it
is able to handle .bz2 files. */
/* #undef SUPPORT_LIBBZ2 */
/* Define to any value to allow pcretest to be linked with libedit. */
#undef SUPPORT_LIBEDIT
/* #undef SUPPORT_LIBEDIT */
/* Define to any value to allow pcretest to be linked with libreadline. */
/* #undef SUPPORT_LIBREADLINE */
@ -416,24 +348,22 @@ them both to 0; an emulation function will be used. */
/* #undef SUPPORT_PCRE8 */
/* Define to any value to enable JIT support in pcregrep. */
/* #undef SUPPORT_PCREGREP_JIT */
#define SUPPORT_PCREGREP_JIT
/* Define to enable support for Unicode properties */
/* Define to any value to enable support for Unicode properties. */
/* #undef SUPPORT_UCP */
/* Define to any value to enable support for the UTF-8/16/32 Unicode encoding.
This will work even in an EBCDIC environment, but it is incompatible with
the EBCDIC macro. That is, PCRE can support *either* EBCDIC code *or*
ASCII/UTF-8/16/32, but not both at once. */
/* #undef SUPPORT_UTF8 */
/* #undef SUPPORT_UTF */
/* Valgrind support to find invalid memory reads. */
/* Define to any value for valgrind support to find invalid memory reads. */
/* #undef SUPPORT_VALGRIND */
/* Version number of package */
#ifndef VERSION
#define VERSION "8.34"
#endif
#define VERSION "8.35"
/* Define to empty if `const' does not conform to ANSI C. */
/* #undef const */

2
ext/pcre/pcrelib/dftables.c
View file

@ -43,9 +43,7 @@ character tables for PCRE. The tables are built according to the current
locale. Now that pcre_maketables is a function visible to the outside world, we
make use of its code from here in order to be consistent. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include <ctype.h>
#include <stdio.h>

1323
ext/pcre/pcrelib/doc/pcre.txt
View file

File diff suppressed because it is too large Load diff

12
ext/pcre/pcrelib/pcre.h
View file

@ -5,7 +5,7 @@
/* This is the public header file for the PCRE library, to be #included by
applications that call the PCRE functions.
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -42,9 +42,9 @@ POSSIBILITY OF SUCH DAMAGE.
/* The current PCRE version information. */
#define PCRE_MAJOR 8
#define PCRE_MINOR 34
#define PCRE_MINOR 35
#define PCRE_PRERELEASE
#define PCRE_DATE 2013-12-15
#define PCRE_DATE 2014-04-04
/* When an application links to a PCRE DLL in Windows, the symbols that are
imported have to be identified as such. When building PCRE, the appropriate
@ -491,36 +491,42 @@ PCRE_EXP_DECL void (*pcre_free)(void *);
PCRE_EXP_DECL void *(*pcre_stack_malloc)(size_t);
PCRE_EXP_DECL void (*pcre_stack_free)(void *);
PCRE_EXP_DECL int (*pcre_callout)(pcre_callout_block *);
PCRE_EXP_DECL int (*pcre_stack_guard)(void);
PCRE_EXP_DECL void *(*pcre16_malloc)(size_t);
PCRE_EXP_DECL void (*pcre16_free)(void *);
PCRE_EXP_DECL void *(*pcre16_stack_malloc)(size_t);
PCRE_EXP_DECL void (*pcre16_stack_free)(void *);
PCRE_EXP_DECL int (*pcre16_callout)(pcre16_callout_block *);
PCRE_EXP_DECL int (*pcre16_stack_guard)(void);
PCRE_EXP_DECL void *(*pcre32_malloc)(size_t);
PCRE_EXP_DECL void (*pcre32_free)(void *);
PCRE_EXP_DECL void *(*pcre32_stack_malloc)(size_t);
PCRE_EXP_DECL void (*pcre32_stack_free)(void *);
PCRE_EXP_DECL int (*pcre32_callout)(pcre32_callout_block *);
PCRE_EXP_DECL int (*pcre32_stack_guard)(void);
#else /* VPCOMPAT */
PCRE_EXP_DECL void *pcre_malloc(size_t);
PCRE_EXP_DECL void pcre_free(void *);
PCRE_EXP_DECL void *pcre_stack_malloc(size_t);
PCRE_EXP_DECL void pcre_stack_free(void *);
PCRE_EXP_DECL int pcre_callout(pcre_callout_block *);
PCRE_EXP_DECL int pcre_stack_guard(void);
PCRE_EXP_DECL void *pcre16_malloc(size_t);
PCRE_EXP_DECL void pcre16_free(void *);
PCRE_EXP_DECL void *pcre16_stack_malloc(size_t);
PCRE_EXP_DECL void pcre16_stack_free(void *);
PCRE_EXP_DECL int pcre16_callout(pcre16_callout_block *);
PCRE_EXP_DECL int pcre16_stack_guard(void);
PCRE_EXP_DECL void *pcre32_malloc(size_t);
PCRE_EXP_DECL void pcre32_free(void *);
PCRE_EXP_DECL void *pcre32_stack_malloc(size_t);
PCRE_EXP_DECL void pcre32_stack_free(void *);
PCRE_EXP_DECL int pcre32_callout(pcre32_callout_block *);
PCRE_EXP_DECL int pcre32_stack_guard(void);
#endif /* VPCOMPAT */
/* User defined callback which provides a stack just before the match starts. */

2
ext/pcre/pcrelib/pcre_chartables.c
View file

@ -20,9 +20,7 @@ and dead code stripping is activated. This leads to link errors. Pulling in the
header ensures that the array gets flagged as "someone outside this compilation
unit might reference this" and so it will always be supplied to the linker. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

141
ext/pcre/pcrelib/pcre_compile.c
View file

@ -6,7 +6,7 @@
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
supporting internal functions that are not used by other modules. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#define NLBLOCK cd /* Block containing newline information */
#define PSSTART start_pattern /* Field containing processed string start */
@ -547,6 +545,8 @@ static const char error_texts[] =
"parentheses are too deeply nested\0"
"invalid range in character class\0"
"group name must start with a non-digit\0"
/* 85 */
"parentheses are too deeply nested (stack check)\0"
;
/* Table to identify digits and hex digits. This is used when compiling
@ -3070,8 +3070,11 @@ const pcre_uint32 *chr_ptr;
const pcre_uint32 *ochr_ptr;
const pcre_uint32 *list_ptr;
const pcre_uchar *next_code;
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
const pcre_uchar *xclass_flags;
#endif
const pcre_uint8 *class_bitset;
const pcre_uint32 *set1, *set2, *set_end;
const pcre_uint8 *set1, *set2, *set_end;
pcre_uint32 chr;
BOOL accepted, invert_bits;
@ -3202,12 +3205,12 @@ for(;;)
if (base_list[0] == OP_CLASS)
#endif
{
set1 = (pcre_uint32 *)(base_end - base_list[2]);
set1 = (pcre_uint8 *)(base_end - base_list[2]);
list_ptr = list;
}
else
{
set1 = (pcre_uint32 *)(code - list[2]);
set1 = (pcre_uint8 *)(code - list[2]);
list_ptr = base_list;
}
@ -3216,41 +3219,53 @@ for(;;)
{
case OP_CLASS:
case OP_NCLASS:
set2 = (pcre_uint32 *)
set2 = (pcre_uint8 *)
((list_ptr == list ? code : base_end) - list_ptr[2]);
break;
/* OP_XCLASS cannot be supported here, because its bitset
is not necessarily complete. E.g: [a-\0x{200}] is stored
as a character range, and the appropriate bits are not set. */
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
case OP_XCLASS:
xclass_flags = (list_ptr == list ? code : base_end) - list_ptr[2] + LINK_SIZE;
if ((*xclass_flags & XCL_HASPROP) != 0) return FALSE;
if ((*xclass_flags & XCL_MAP) == 0)
{
/* No bits are set for characters < 256. */
if (list[1] == 0) return TRUE;
/* Might be an empty repeat. */
continue;
}
set2 = (pcre_uint8 *)(xclass_flags + 1);
break;
#endif
case OP_NOT_DIGIT:
invert_bits = TRUE;
/* Fall through */
invert_bits = TRUE;
/* Fall through */
case OP_DIGIT:
set2 = (pcre_uint32 *)(cd->cbits + cbit_digit);
break;
set2 = (pcre_uint8 *)(cd->cbits + cbit_digit);
break;
case OP_NOT_WHITESPACE:
invert_bits = TRUE;
/* Fall through */
invert_bits = TRUE;
/* Fall through */
case OP_WHITESPACE:
set2 = (pcre_uint32 *)(cd->cbits + cbit_space);
break;
set2 = (pcre_uint8 *)(cd->cbits + cbit_space);
break;
case OP_NOT_WORDCHAR:
invert_bits = TRUE;
/* Fall through */
invert_bits = TRUE;
/* Fall through */
case OP_WORDCHAR:
set2 = (pcre_uint32 *)(cd->cbits + cbit_word);
break;
set2 = (pcre_uint8 *)(cd->cbits + cbit_word);
break;
default:
return FALSE;
}
/* Compare 4 bytes to improve speed. */
set_end = set1 + (32 / 4);
/* Because the sets are unaligned, we need
to perform byte comparison here. */
set_end = set1 + 32;
if (invert_bits)
{
do
@ -3551,7 +3566,9 @@ for(;;)
if (list[1] == 0) return TRUE;
}
return FALSE;
/* Control never reaches here. There used to be a fail-save return FALSE; here,
but some compilers complain about an unreachable statement. */
}
@ -4062,12 +4079,16 @@ for (c = *cptr; c <= d; c++)
if (c > d) return -1; /* Reached end of range */
/* Found a character that has a single other case. Search for the end of the
range, which is either the end of the input range, or a character that has zero
or more than one other cases. */
*ocptr = othercase;
next = othercase + 1;
for (++c; c <= d; c++)
{
if (UCD_OTHERCASE(c) != next) break;
if ((co = UCD_CASESET(c)) != 0 || UCD_OTHERCASE(c) != next) break;
next++;
}
@ -4105,6 +4126,7 @@ add_to_class(pcre_uint8 *classbits, pcre_uchar **uchardptr, int options,
compile_data *cd, pcre_uint32 start, pcre_uint32 end)
{
pcre_uint32 c;
pcre_uint32 classbits_end = (end <= 0xff ? end : 0xff);
int n8 = 0;
/* If caseless matching is required, scan the range and process alternate
@ -4148,7 +4170,7 @@ if ((options & PCRE_CASELESS) != 0)
/* Not UTF-mode, or no UCP */
for (c = start; c <= end && c < 256; c++)
for (c = start; c <= classbits_end; c++)
{
SETBIT(classbits, cd->fcc[c]);
n8++;
@ -4173,22 +4195,21 @@ in all cases. */
#endif /* COMPILE_PCRE[8|16] */
/* If all characters are less than 256, use the bit map. Otherwise use extra
data. */
/* Use the bitmap for characters < 256. Otherwise use extra data.*/
if (end < 0x100)
for (c = start; c <= classbits_end; c++)
{
for (c = start; c <= end; c++)
{
n8++;
SETBIT(classbits, c);
}
/* Regardless of start, c will always be <= 255. */
SETBIT(classbits, c);
n8++;
}
else
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
if (start <= 0xff) start = 0xff + 1;
if (end >= start)
{
pcre_uchar *uchardata = *uchardptr;
#ifdef SUPPORT_UTF
if ((options & PCRE_UTF8) != 0) /* All UTFs use the same flag bit */
{
@ -4228,6 +4249,7 @@ else
*uchardptr = uchardata; /* Updata extra data pointer */
}
#endif /* SUPPORT_UTF || !COMPILE_PCRE8 */
return n8; /* Number of 8-bit characters */
}
@ -4449,6 +4471,9 @@ for (;; ptr++)
BOOL reset_bracount;
int class_has_8bitchar;
int class_one_char;
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
BOOL xclass_has_prop;
#endif
int newoptions;
int recno;
int refsign;
@ -4783,13 +4808,26 @@ for (;; ptr++)
should_flip_negation = FALSE;
/* Extended class (xclass) will be used when characters > 255
might match. */
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
xclass = FALSE;
class_uchardata = code + LINK_SIZE + 2; /* For XCLASS items */
class_uchardata_base = class_uchardata; /* Save the start */
#endif
/* For optimization purposes, we track some properties of the class:
class_has_8bitchar will be non-zero if the class contains at least one <
256 character; class_one_char will be 1 if the class contains just one
character. */
character; xclass_has_prop will be TRUE if unicode property checks
are present in the class. */
class_has_8bitchar = 0;
class_one_char = 0;
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
xclass_has_prop = FALSE;
#endif
/* Initialize the 32-char bit map to all zeros. We build the map in a
temporary bit of memory, in case the class contains fewer than two
@ -4798,12 +4836,6 @@ for (;; ptr++)
memset(classbits, 0, 32 * sizeof(pcre_uint8));
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
xclass = FALSE;
class_uchardata = code + LINK_SIZE + 2; /* For XCLASS items */
class_uchardata_base = class_uchardata; /* Save the start */
#endif
/* Process characters until ] is reached. By writing this as a "do" it
means that an initial ] is taken as a data character. At the start of the
loop, c contains the first byte of the character. */
@ -4927,6 +4959,7 @@ for (;; ptr++)
*class_uchardata++ = local_negate? XCL_NOTPROP : XCL_PROP;
*class_uchardata++ = ptype;
*class_uchardata++ = 0;
xclass_has_prop = TRUE;
ptr = tempptr + 1;
continue;
@ -5109,6 +5142,7 @@ for (;; ptr++)
XCL_PROP : XCL_NOTPROP;
*class_uchardata++ = ptype;
*class_uchardata++ = pdata;
xclass_has_prop = TRUE;
class_has_8bitchar--; /* Undo! */
continue;
}
@ -5403,6 +5437,7 @@ for (;; ptr++)
*code++ = OP_XCLASS;
code += LINK_SIZE;
*code = negate_class? XCL_NOT:0;
if (xclass_has_prop) *code |= XCL_HASPROP;
/* If the map is required, move up the extra data to make room for it;
otherwise just move the code pointer to the end of the extra data. */
@ -5412,6 +5447,8 @@ for (;; ptr++)
*code++ |= XCL_MAP;
memmove(code + (32 / sizeof(pcre_uchar)), code,
IN_UCHARS(class_uchardata - code));
if (negate_class && !xclass_has_prop)
for (c = 0; c < 32; c++) classbits[c] = ~classbits[c];
memcpy(code, classbits, 32);
code = class_uchardata + (32 / sizeof(pcre_uchar));
}
@ -6580,7 +6617,10 @@ for (;; ptr++)
code[1+LINK_SIZE] = OP_CREF;
skipbytes = 1+IMM2_SIZE;
refsign = -1;
refsign = -1; /* => not a number */
namelen = -1; /* => not a name; must set to avoid warning */
name = NULL; /* Always set to avoid warning */
recno = 0; /* Always set to avoid warning */
/* Check for a test for recursion in a named group. */
@ -6617,7 +6657,6 @@ for (;; ptr++)
if (refsign >= 0)
{
recno = 0;
while (IS_DIGIT(*ptr))
{
recno = recno * 10 + (int)(*ptr - CHAR_0);
@ -7994,6 +8033,16 @@ unsigned int orig_bracount;
unsigned int max_bracount;
branch_chain bc;
/* If set, call the external function that checks for stack availability. */
if (PUBL(stack_guard) != NULL && PUBL(stack_guard)())
{
*errorcodeptr= ERR85;
return FALSE;
}
/* Miscellaneous initialization */
bc.outer = bcptr;
bc.current_branch = code;

2
ext/pcre/pcrelib/pcre_config.c
View file

@ -41,9 +41,7 @@ POSSIBILITY OF SUCH DAMAGE.
/* This module contains the external function pcre_config(). */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
/* Keep the original link size. */
static int real_link_size = LINK_SIZE;

104
ext/pcre/pcrelib/pcre_exec.c
View file

@ -6,7 +6,7 @@
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -41,9 +41,7 @@ POSSIBILITY OF SUCH DAMAGE.
pattern matching using an NFA algorithm, trying to mimic Perl as closely as
possible. There are also some static supporting functions. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#define NLBLOCK md /* Block containing newline information */
#define PSSTART start_subject /* Field containing processed string start */
@ -134,7 +132,7 @@ pcre_uint32 c;
BOOL utf = md->utf;
if (is_subject && length > md->end_subject - p) length = md->end_subject - p;
while (length-- > 0)
if (isprint(c = RAWUCHARINCTEST(p))) printf("%c", (char)c); else printf("\\x{%02x}", c);
if (isprint(c = UCHAR21INCTEST(p))) printf("%c", (char)c); else printf("\\x{%02x}", c);
}
#endif
@ -237,8 +235,8 @@ if (caseless)
{
pcre_uint32 cc, cp;
if (eptr >= md->end_subject) return -2; /* Partial match */
cc = RAWUCHARTEST(eptr);
cp = RAWUCHARTEST(p);
cc = UCHAR21TEST(eptr);
cp = UCHAR21TEST(p);
if (TABLE_GET(cp, md->lcc, cp) != TABLE_GET(cc, md->lcc, cc)) return -1;
p++;
eptr++;
@ -254,7 +252,7 @@ else
while (length-- > 0)
{
if (eptr >= md->end_subject) return -2; /* Partial match */
if (RAWUCHARINCTEST(p) != RAWUCHARINCTEST(eptr)) return -1;
if (UCHAR21INCTEST(p) != UCHAR21INCTEST(eptr)) return -1;
}
}
@ -2103,7 +2101,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
UCHAR21TEST(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -2147,7 +2145,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
UCHAR21TEST(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -2290,7 +2288,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHARTEST(eptr) == NLBLOCK->nl[0])
UCHAR21TEST(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -2444,7 +2442,7 @@ for (;;)
{
SCHECK_PARTIAL();
}
else if (RAWUCHARTEST(eptr) == CHAR_LF) eptr++;
else if (UCHAR21TEST(eptr) == CHAR_LF) eptr++;
break;
case CHAR_LF:
@ -2691,16 +2689,22 @@ for (;;)
pcre_uchar *slot = md->name_table + GET2(ecode, 1) * md->name_entry_size;
ecode += 1 + 2*IMM2_SIZE;
/* Setting the default length first and initializing 'offset' avoids
compiler warnings in the REF_REPEAT code. */
length = (md->jscript_compat)? 0 : -1;
offset = 0;
while (count-- > 0)
{
offset = GET2(slot, 0) << 1;
if (offset < offset_top && md->offset_vector[offset] >= 0) break;
if (offset < offset_top && md->offset_vector[offset] >= 0)
{
length = md->offset_vector[offset+1] - md->offset_vector[offset];
break;
}
slot += md->name_entry_size;
}
if (count < 0)
length = (md->jscript_compat)? 0 : -1;
else
length = md->offset_vector[offset+1] - md->offset_vector[offset];
}
goto REF_REPEAT;
@ -3212,7 +3216,7 @@ for (;;)
CHECK_PARTIAL(); /* Not SCHECK_PARTIAL() */
RRETURN(MATCH_NOMATCH);
}
while (length-- > 0) if (*ecode++ != RAWUCHARINC(eptr)) RRETURN(MATCH_NOMATCH);
while (length-- > 0) if (*ecode++ != UCHAR21INC(eptr)) RRETURN(MATCH_NOMATCH);
}
else
#endif
@ -3252,7 +3256,7 @@ for (;;)
if (fc < 128)
{
pcre_uint32 cc = RAWUCHAR(eptr);
pcre_uint32 cc = UCHAR21(eptr);
if (md->lcc[fc] != TABLE_GET(cc, md->lcc, cc)) RRETURN(MATCH_NOMATCH);
ecode++;
eptr++;
@ -3521,7 +3525,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHARTEST(eptr);
cc = UCHAR21TEST(eptr);
if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
eptr++;
}
@ -3539,7 +3543,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHARTEST(eptr);
cc = UCHAR21TEST(eptr);
if (fc != cc && foc != cc) RRETURN(MATCH_NOMATCH);
eptr++;
}
@ -3556,7 +3560,7 @@ for (;;)
SCHECK_PARTIAL();
break;
}
cc = RAWUCHARTEST(eptr);
cc = UCHAR21TEST(eptr);
if (fc != cc && foc != cc) break;
eptr++;
}
@ -3583,7 +3587,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
if (fc != UCHAR21INCTEST(eptr)) RRETURN(MATCH_NOMATCH);
}
if (min == max) continue;
@ -3600,7 +3604,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
if (fc != RAWUCHARINCTEST(eptr)) RRETURN(MATCH_NOMATCH);
if (fc != UCHAR21INCTEST(eptr)) RRETURN(MATCH_NOMATCH);
}
/* Control never gets here */
}
@ -3614,7 +3618,7 @@ for (;;)
SCHECK_PARTIAL();
break;
}
if (fc != RAWUCHARTEST(eptr)) break;
if (fc != UCHAR21TEST(eptr)) break;
eptr++;
}
if (possessive) continue; /* No backtracking */
@ -4369,7 +4373,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHAR(eptr) == NLBLOCK->nl[0])
UCHAR21(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -4411,7 +4415,7 @@ for (;;)
default: RRETURN(MATCH_NOMATCH);
case CHAR_CR:
if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
if (eptr < md->end_subject && UCHAR21(eptr) == CHAR_LF) eptr++;
break;
case CHAR_LF:
@ -4521,7 +4525,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHAR(eptr);
cc = UCHAR21(eptr);
if (cc >= 128 || (md->ctypes[cc] & ctype_digit) == 0)
RRETURN(MATCH_NOMATCH);
eptr++;
@ -4538,7 +4542,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHAR(eptr);
cc = UCHAR21(eptr);
if (cc < 128 && (md->ctypes[cc] & ctype_space) != 0)
RRETURN(MATCH_NOMATCH);
eptr++;
@ -4555,7 +4559,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHAR(eptr);
cc = UCHAR21(eptr);
if (cc >= 128 || (md->ctypes[cc] & ctype_space) == 0)
RRETURN(MATCH_NOMATCH);
eptr++;
@ -4572,7 +4576,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHAR(eptr);
cc = UCHAR21(eptr);
if (cc < 128 && (md->ctypes[cc] & ctype_word) != 0)
RRETURN(MATCH_NOMATCH);
eptr++;
@ -4589,7 +4593,7 @@ for (;;)
SCHECK_PARTIAL();
RRETURN(MATCH_NOMATCH);
}
cc = RAWUCHAR(eptr);
cc = UCHAR21(eptr);
if (cc >= 128 || (md->ctypes[cc] & ctype_word) == 0)
RRETURN(MATCH_NOMATCH);
eptr++;
@ -5150,7 +5154,7 @@ for (;;)
{
default: RRETURN(MATCH_NOMATCH);
case CHAR_CR:
if (eptr < md->end_subject && RAWUCHAR(eptr) == CHAR_LF) eptr++;
if (eptr < md->end_subject && UCHAR21(eptr) == CHAR_LF) eptr++;
break;
case CHAR_LF:
@ -5689,7 +5693,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHAR(eptr) == NLBLOCK->nl[0])
UCHAR21(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -5715,7 +5719,7 @@ for (;;)
eptr + 1 >= md->end_subject &&
NLBLOCK->nltype == NLTYPE_FIXED &&
NLBLOCK->nllen == 2 &&
RAWUCHAR(eptr) == NLBLOCK->nl[0])
UCHAR21(eptr) == NLBLOCK->nl[0])
{
md->hitend = TRUE;
if (md->partial > 1) RRETURN(PCRE_ERROR_PARTIAL);
@ -5772,7 +5776,7 @@ for (;;)
if (c == CHAR_CR)
{
if (++eptr >= md->end_subject) break;
if (RAWUCHAR(eptr) == CHAR_LF) eptr++;
if (UCHAR21(eptr) == CHAR_LF) eptr++;
}
else
{
@ -5935,8 +5939,8 @@ for (;;)
if (rrc != MATCH_NOMATCH) RRETURN(rrc);
eptr--;
BACKCHAR(eptr);
if (ctype == OP_ANYNL && eptr > pp && RAWUCHAR(eptr) == CHAR_NL &&
RAWUCHAR(eptr - 1) == CHAR_CR) eptr--;
if (ctype == OP_ANYNL && eptr > pp && UCHAR21(eptr) == CHAR_NL &&
UCHAR21(eptr - 1) == CHAR_CR) eptr--;
}
}
else
@ -6783,10 +6787,10 @@ for(;;)
if (first_char != first_char2)
while (start_match < end_subject &&
(smc = RAWUCHARTEST(start_match)) != first_char && smc != first_char2)
(smc = UCHAR21TEST(start_match)) != first_char && smc != first_char2)
start_match++;
else
while (start_match < end_subject && RAWUCHARTEST(start_match) != first_char)
while (start_match < end_subject && UCHAR21TEST(start_match) != first_char)
start_match++;
}
@ -6818,7 +6822,7 @@ for(;;)
if (start_match[-1] == CHAR_CR &&
(md->nltype == NLTYPE_ANY || md->nltype == NLTYPE_ANYCRLF) &&
start_match < end_subject &&
RAWUCHARTEST(start_match) == CHAR_NL)
UCHAR21TEST(start_match) == CHAR_NL)
start_match++;
}
}
@ -6829,22 +6833,12 @@ for(;;)
{
while (start_match < end_subject)
{
register pcre_uint32 c = RAWUCHARTEST(start_match);
register pcre_uint32 c = UCHAR21TEST(start_match);
#ifndef COMPILE_PCRE8
if (c > 255) c = 255;
#endif
if ((start_bits[c/8] & (1 << (c&7))) == 0)
{
start_match++;
#if defined SUPPORT_UTF && defined COMPILE_PCRE8
/* In non 8-bit mode, the iteration will stop for
characters > 255 at the beginning or not stop at all. */
if (utf)
ACROSSCHAR(start_match < end_subject, *start_match,
start_match++);
#endif
}
else break;
if ((start_bits[c/8] & (1 << (c&7))) != 0) break;
start_match++;
}
}
} /* Starting optimizations */
@ -6897,7 +6891,7 @@ for(;;)
{
while (p < end_subject)
{
register pcre_uint32 pp = RAWUCHARINCTEST(p);
register pcre_uint32 pp = UCHAR21INCTEST(p);
if (pp == req_char || pp == req_char2) { p--; break; }
}
}
@ -6905,7 +6899,7 @@ for(;;)
{
while (p < end_subject)
{
if (RAWUCHARINCTEST(p) == req_char) { p--; break; }
if (UCHAR21INCTEST(p) == req_char) { p--; break; }
}
}

2
ext/pcre/pcrelib/pcre_fullinfo.c
View file

@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
information about a compiled pattern. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

2
ext/pcre/pcrelib/pcre_get.c
View file

@ -43,9 +43,7 @@ from the subject string after a regex match has succeeded. The original idea
for these functions came from Scott Wimer. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

6
ext/pcre/pcrelib/pcre_globals.c
View file

@ -6,7 +6,7 @@
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2012 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -52,9 +52,7 @@ a local function is used.
Also, when compiling for Virtual Pascal, things are done differently, and
global variables are not used. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"
@ -72,6 +70,7 @@ PCRE_EXP_DATA_DEFN void (*PUBL(free))(void *) = LocalPcreFree;
PCRE_EXP_DATA_DEFN void *(*PUBL(stack_malloc))(size_t) = LocalPcreMalloc;
PCRE_EXP_DATA_DEFN void (*PUBL(stack_free))(void *) = LocalPcreFree;
PCRE_EXP_DATA_DEFN int (*PUBL(callout))(PUBL(callout_block) *) = NULL;
PCRE_EXP_DATA_DEFN int (*PUBL(stack_guard))(void) = NULL;
#elif !defined VPCOMPAT
PCRE_EXP_DATA_DEFN void *(*PUBL(malloc))(size_t) = malloc;
@ -79,6 +78,7 @@ PCRE_EXP_DATA_DEFN void (*PUBL(free))(void *) = free;
PCRE_EXP_DATA_DEFN void *(*PUBL(stack_malloc))(size_t) = malloc;
PCRE_EXP_DATA_DEFN void (*PUBL(stack_free))(void *) = free;
PCRE_EXP_DATA_DEFN int (*PUBL(callout))(PUBL(callout_block) *) = NULL;
PCRE_EXP_DATA_DEFN int (*PUBL(stack_guard))(void) = NULL;
#endif
/* End of pcre_globals.c */

118
ext/pcre/pcrelib/pcre_internal.h
View file

@ -7,7 +7,7 @@
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2013 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -316,8 +316,8 @@ start/end of string field names are. */
&(NLBLOCK->nllen), utf)) \
: \
((p) <= NLBLOCK->PSEND - NLBLOCK->nllen && \
RAWUCHARTEST(p) == NLBLOCK->nl[0] && \
(NLBLOCK->nllen == 1 || RAWUCHARTEST(p+1) == NLBLOCK->nl[1]) \
UCHAR21TEST(p) == NLBLOCK->nl[0] && \
(NLBLOCK->nllen == 1 || UCHAR21TEST(p+1) == NLBLOCK->nl[1]) \
) \
)
@ -330,8 +330,8 @@ start/end of string field names are. */
&(NLBLOCK->nllen), utf)) \
: \
((p) >= NLBLOCK->PSSTART + NLBLOCK->nllen && \
RAWUCHARTEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] && \
(NLBLOCK->nllen == 1 || RAWUCHARTEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
UCHAR21TEST(p - NLBLOCK->nllen) == NLBLOCK->nl[0] && \
(NLBLOCK->nllen == 1 || UCHAR21TEST(p - NLBLOCK->nllen + 1) == NLBLOCK->nl[1]) \
) \
)
@ -582,12 +582,27 @@ changed in future to be a fixed number of bytes or to depend on LINK_SIZE. */
#define MAX_MARK ((1u << 8) - 1)
#endif
/* There is a proposed future special "UTF-21" mode, in which only the lowest
21 bits of a 32-bit character are interpreted as UTF, with the remaining 11
high-order bits available to the application for other uses. In preparation for
the future implementation of this mode, there are macros that load a data item
and, if in this special mode, mask it to 21 bits. These macros all have names
starting with UCHAR21. In all other modes, including the normal 32-bit
library, the macros all have the same simple definitions. When the new mode is
implemented, it is expected that these definitions will be varied appropriately
using #ifdef when compiling the library that supports the special mode. */
#define UCHAR21(eptr) (*(eptr))
#define UCHAR21TEST(eptr) (*(eptr))
#define UCHAR21INC(eptr) (*(eptr)++)
#define UCHAR21INCTEST(eptr) (*(eptr)++)
/* When UTF encoding is being used, a character is no longer just a single
byte. The macros for character handling generate simple sequences when used in
character-mode, and more complicated ones for UTF characters. GETCHARLENTEST
and other macros are not used when UTF is not supported, so they are not
defined. To make sure they can never even appear when UTF support is omitted,
we don't even define them. */
byte in 8-bit mode or a single short in 16-bit mode. The macros for character
handling generate simple sequences when used in the basic mode, and more
complicated ones for UTF characters. GETCHARLENTEST and other macros are not
used when UTF is not supported. To make sure they can never even appear when
UTF support is omitted, we don't even define them. */
#ifndef SUPPORT_UTF
@ -600,10 +615,6 @@ we don't even define them. */
#define GETCHARINC(c, eptr) c = *eptr++;
#define GETCHARINCTEST(c, eptr) c = *eptr++;
#define GETCHARLEN(c, eptr, len) c = *eptr;
#define RAWUCHAR(eptr) (*(eptr))
#define RAWUCHARINC(eptr) (*(eptr)++)
#define RAWUCHARTEST(eptr) (*(eptr))
#define RAWUCHARINCTEST(eptr) (*(eptr)++)
/* #define GETCHARLENTEST(c, eptr, len) */
/* #define BACKCHAR(eptr) */
/* #define FORWARDCHAR(eptr) */
@ -776,30 +787,6 @@ do not know if we are in UTF-8 mode. */
c = *eptr; \
if (utf && c >= 0xc0) GETUTF8LEN(c, eptr, len);
/* Returns the next uchar, not advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHAR(eptr) \
(*(eptr))
/* Returns the next uchar, advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHARINC(eptr) \
(*((eptr)++))
/* Returns the next uchar, testing for UTF mode, and not advancing the
pointer. */
#define RAWUCHARTEST(eptr) \
(*(eptr))
/* Returns the next uchar, testing for UTF mode, advancing the
pointer. */
#define RAWUCHARINCTEST(eptr) \
(*((eptr)++))
/* If the pointer is not at the start of a character, move it back until
it is. This is called only in UTF-8 mode - we don't put a test within the macro
because almost all calls are already within a block of UTF-8 only code. */
@ -895,30 +882,6 @@ we do not know if we are in UTF-16 mode. */
c = *eptr; \
if (utf && (c & 0xfc00) == 0xd800) GETUTF16LEN(c, eptr, len);
/* Returns the next uchar, not advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHAR(eptr) \
(*(eptr))
/* Returns the next uchar, advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHARINC(eptr) \
(*((eptr)++))
/* Returns the next uchar, testing for UTF mode, and not advancing the
pointer. */
#define RAWUCHARTEST(eptr) \
(*(eptr))
/* Returns the next uchar, testing for UTF mode, advancing the
pointer. */
#define RAWUCHARINCTEST(eptr) \
(*((eptr)++))
/* If the pointer is not at the start of a character, move it back until
it is. This is called only in UTF-16 mode - we don't put a test within the
macro because almost all calls are already within a block of UTF-16 only
@ -980,30 +943,6 @@ This is called when we do not know if we are in UTF-32 mode. */
#define GETCHARLENTEST(c, eptr, len) \
GETCHARTEST(c, eptr)
/* Returns the next uchar, not advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHAR(eptr) \
(*(eptr))
/* Returns the next uchar, advancing the pointer. This is called when
we know we are in UTF mode. */
#define RAWUCHARINC(eptr) \
(*((eptr)++))
/* Returns the next uchar, testing for UTF mode, and not advancing the
pointer. */
#define RAWUCHARTEST(eptr) \
(*(eptr))
/* Returns the next uchar, testing for UTF mode, advancing the
pointer. */
#define RAWUCHARINCTEST(eptr) \
(*((eptr)++))
/* If the pointer is not at the start of a character, move it back until
it is. This is called only in UTF-32 mode - we don't put a test within the
macro because almost all calls are already within a block of UTF-32 only
@ -1874,8 +1813,9 @@ table. */
/* Flag bits and data types for the extended class (OP_XCLASS) for classes that
contain characters with values greater than 255. */
#define XCL_NOT 0x01 /* Flag: this is a negative class */
#define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
#define XCL_NOT 0x01 /* Flag: this is a negative class */
#define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
#define XCL_HASPROP 0x04 /* Flag: property checks are present. */
#define XCL_END 0 /* Marks end of individual items */
#define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
@ -2341,7 +2281,7 @@ enum { ERR0, ERR1, ERR2, ERR3, ERR4, ERR5, ERR6, ERR7, ERR8, ERR9,
ERR50, ERR51, ERR52, ERR53, ERR54, ERR55, ERR56, ERR57, ERR58, ERR59,
ERR60, ERR61, ERR62, ERR63, ERR64, ERR65, ERR66, ERR67, ERR68, ERR69,
ERR70, ERR71, ERR72, ERR73, ERR74, ERR75, ERR76, ERR77, ERR78, ERR79,
ERR80, ERR81, ERR82, ERR83, ERR84, ERRCOUNT };
ERR80, ERR81, ERR82, ERR83, ERR84, ERR85, ERRCOUNT };
/* JIT compiling modes. The function list is indexed by them. */

10699
ext/pcre/pcrelib/pcre_jit_compile.c Normal file
View file

File diff suppressed because it is too large Load diff

2
ext/pcre/pcrelib/pcre_maketables.c
View file

@ -45,9 +45,7 @@ compilation of dftables.c, in which case the macro DFTABLES is defined. */
#ifndef DFTABLES
# ifdef HAVE_CONFIG_H
# include "config.h"
# endif
# include "pcre_internal.h"
#endif

2
ext/pcre/pcrelib/pcre_newline.c
View file

@ -47,9 +47,7 @@ and NLTYPE_ANY. The full list of Unicode newline characters is taken from
http://unicode.org/unicode/reports/tr18/. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

2
ext/pcre/pcrelib/pcre_ord2utf8.c
View file

@ -41,9 +41,7 @@ POSSIBILITY OF SUCH DAMAGE.
/* This file contains a private PCRE function that converts an ordinal
character value into a UTF8 string. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#define COMPILE_PCRE8

832
ext/pcre/pcrelib/pcre_printint.c Normal file
View file

@ -0,0 +1,832 @@
/*************************************************
* Perl-Compatible Regular Expressions *
*************************************************/
/* PCRE is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2012 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* 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.
* Neither the name of the University of Cambridge nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 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 OWNER 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 module contains a PCRE private debugging function for printing out the
internal form of a compiled regular expression, along with some supporting
local functions. This source file is used in two places:
(1) It is #included by pcre_compile.c when it is compiled in debugging mode
(PCRE_DEBUG defined in pcre_internal.h). It is not included in production
compiles. In this case PCRE_INCLUDED is defined.
(2) It is also compiled separately and linked with pcretest.c, which can be
asked to print out a compiled regex for debugging purposes. */
#ifndef PCRE_INCLUDED
#include "config.h"
/* For pcretest program. */
#define PRIV(name) name
/* We have to include pcre_internal.h because we need the internal info for
displaying the results of pcre_study() and we also need to know about the
internal macros, structures, and other internal data values; pcretest has
"inside information" compared to a program that strictly follows the PCRE API.
Although pcre_internal.h does itself include pcre.h, we explicitly include it
here before pcre_internal.h so that the PCRE_EXP_xxx macros get set
appropriately for an application, not for building PCRE. */
#include "pcre.h"
#include "pcre_internal.h"
/* These are the funtions that are contained within. It doesn't seem worth
having a separate .h file just for this. */
#endif /* PCRE_INCLUDED */
#ifdef PCRE_INCLUDED
static /* Keep the following function as private. */
#endif
#if defined COMPILE_PCRE8
void pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths);
#elif defined COMPILE_PCRE16
void pcre16_printint(pcre *external_re, FILE *f, BOOL print_lengths);
#elif defined COMPILE_PCRE32
void pcre32_printint(pcre *external_re, FILE *f, BOOL print_lengths);
#endif
/* Macro that decides whether a character should be output as a literal or in
hexadecimal. We don't use isprint() because that can vary from system to system
(even without the use of locales) and we want the output always to be the same,
for testing purposes. */
#ifdef EBCDIC
#define PRINTABLE(c) ((c) >= 64 && (c) < 255)
#else
#define PRINTABLE(c) ((c) >= 32 && (c) < 127)
#endif
/* The table of operator names. */
static const char *priv_OP_names[] = { OP_NAME_LIST };
/* This table of operator lengths is not actually used by the working code,
but its size is needed for a check that ensures it is the correct size for the
number of opcodes (thus catching update omissions). */
static const pcre_uint8 priv_OP_lengths[] = { OP_LENGTHS };
/*************************************************
* Print single- or multi-byte character *
*************************************************/
static unsigned int
print_char(FILE *f, pcre_uchar *ptr, BOOL utf)
{
pcre_uint32 c = *ptr;
#ifndef SUPPORT_UTF
(void)utf; /* Avoid compiler warning */
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
else fprintf(f, "\\x{%x}", c);
return 0;
#else
#if defined COMPILE_PCRE8
if (!utf || (c & 0xc0) != 0xc0)
{
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
else if (c < 0x80) fprintf(f, "\\x%02x", c);
else fprintf(f, "\\x{%02x}", c);
return 0;
}
else
{
int i;
int a = PRIV(utf8_table4)[c & 0x3f]; /* Number of additional bytes */
int s = 6*a;
c = (c & PRIV(utf8_table3)[a]) << s;
for (i = 1; i <= a; i++)
{
/* This is a check for malformed UTF-8; it should only occur if the sanity
check has been turned off. Rather than swallow random bytes, just stop if
we hit a bad one. Print it with \X instead of \x as an indication. */
if ((ptr[i] & 0xc0) != 0x80)
{
fprintf(f, "\\X{%x}", c);
return i - 1;
}
/* The byte is OK */
s -= 6;
c |= (ptr[i] & 0x3f) << s;
}
fprintf(f, "\\x{%x}", c);
return a;
}
#elif defined COMPILE_PCRE16
if (!utf || (c & 0xfc00) != 0xd800)
{
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
else fprintf(f, "\\x{%02x}", c);
return 0;
}
else
{
/* This is a check for malformed UTF-16; it should only occur if the sanity
check has been turned off. Rather than swallow a low surrogate, just stop if
we hit a bad one. Print it with \X instead of \x as an indication. */
if ((ptr[1] & 0xfc00) != 0xdc00)
{
fprintf(f, "\\X{%x}", c);
return 0;
}
c = (((c & 0x3ff) << 10) | (ptr[1] & 0x3ff)) + 0x10000;
fprintf(f, "\\x{%x}", c);
return 1;
}
#elif defined COMPILE_PCRE32
if (!utf || (c & 0xfffff800u) != 0xd800u)
{
if (PRINTABLE(c)) fprintf(f, "%c", (char)c);
else if (c <= 0x80) fprintf(f, "\\x%02x", c);
else fprintf(f, "\\x{%x}", c);
return 0;
}
else
{
/* This is a check for malformed UTF-32; it should only occur if the sanity
check has been turned off. Rather than swallow a surrogate, just stop if
we hit one. Print it with \X instead of \x as an indication. */
fprintf(f, "\\X{%x}", c);
return 0;
}
#endif /* COMPILE_PCRE[8|16|32] */
#endif /* SUPPORT_UTF */
}
/*************************************************
* Print uchar string (regardless of utf) *
*************************************************/
static void
print_puchar(FILE *f, PCRE_PUCHAR ptr)
{
while (*ptr != '\0')
{
register pcre_uint32 c = *ptr++;
if (PRINTABLE(c)) fprintf(f, "%c", c); else fprintf(f, "\\x{%x}", c);
}
}
/*************************************************
* Find Unicode property name *
*************************************************/
static const char *
get_ucpname(unsigned int ptype, unsigned int pvalue)
{
#ifdef SUPPORT_UCP
int i;
for (i = PRIV(utt_size) - 1; i >= 0; i--)
{
if (ptype == PRIV(utt)[i].type && pvalue == PRIV(utt)[i].value) break;
}
return (i >= 0)? PRIV(utt_names) + PRIV(utt)[i].name_offset : "??";
#else
/* It gets harder and harder to shut off unwanted compiler warnings. */
ptype = ptype * pvalue;
return (ptype == pvalue)? "??" : "??";
#endif
}
/*************************************************
* Print Unicode property value *
*************************************************/
/* "Normal" properties can be printed from tables. The PT_CLIST property is a
pseudo-property that contains a pointer to a list of case-equivalent
characters. This is used only when UCP support is available and UTF mode is
selected. It should never occur otherwise, but just in case it does, have
something ready to print. */
static void
print_prop(FILE *f, pcre_uchar *code, const char *before, const char *after)
{
if (code[1] != PT_CLIST)
{
fprintf(f, "%s%s %s%s", before, priv_OP_names[*code], get_ucpname(code[1],
code[2]), after);
}
else
{
const char *not = (*code == OP_PROP)? "" : "not ";
#ifndef SUPPORT_UCP
fprintf(f, "%s%sclist %d%s", before, not, code[2], after);
#else
const pcre_uint32 *p = PRIV(ucd_caseless_sets) + code[2];
fprintf (f, "%s%sclist", before, not);
while (*p < NOTACHAR) fprintf(f, " %04x", *p++);
fprintf(f, "%s", after);
#endif
}
}
/*************************************************
* Print compiled regex *
*************************************************/
/* Make this function work for a regex with integers either byte order.
However, we assume that what we are passed is a compiled regex. The
print_lengths flag controls whether offsets and lengths of items are printed.
They can be turned off from pcretest so that automatic tests on bytecode can be
written that do not depend on the value of LINK_SIZE. */
#ifdef PCRE_INCLUDED
static /* Keep the following function as private. */
#endif
#if defined COMPILE_PCRE8
void
pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths)
#elif defined COMPILE_PCRE16
void
pcre16_printint(pcre *external_re, FILE *f, BOOL print_lengths)
#elif defined COMPILE_PCRE32
void
pcre32_printint(pcre *external_re, FILE *f, BOOL print_lengths)
#endif
{
REAL_PCRE *re = (REAL_PCRE *)external_re;
pcre_uchar *codestart, *code;
BOOL utf;
unsigned int options = re->options;
int offset = re->name_table_offset;
int count = re->name_count;
int size = re->name_entry_size;
if (re->magic_number != MAGIC_NUMBER)
{
offset = ((offset << 8) & 0xff00) | ((offset >> 8) & 0xff);
count = ((count << 8) & 0xff00) | ((count >> 8) & 0xff);
size = ((size << 8) & 0xff00) | ((size >> 8) & 0xff);
options = ((options << 24) & 0xff000000) |
((options << 8) & 0x00ff0000) |
((options >> 8) & 0x0000ff00) |
((options >> 24) & 0x000000ff);
}
code = codestart = (pcre_uchar *)re + offset + count * size;
/* PCRE_UTF(16|32) have the same value as PCRE_UTF8. */
utf = (options & PCRE_UTF8) != 0;
for(;;)
{
pcre_uchar *ccode;
const char *flag = " ";
pcre_uint32 c;
unsigned int extra = 0;
if (print_lengths)
fprintf(f, "%3d ", (int)(code - codestart));
else
fprintf(f, " ");
switch(*code)
{
/* ========================================================================== */
/* These cases are never obeyed. This is a fudge that causes a compile-
time error if the vectors OP_names or OP_lengths, which are indexed
by opcode, are not the correct length. It seems to be the only way to do
such a check at compile time, as the sizeof() operator does not work in
the C preprocessor. */
case OP_TABLE_LENGTH:
case OP_TABLE_LENGTH +
((sizeof(priv_OP_names)/sizeof(const char *) == OP_TABLE_LENGTH) &&
(sizeof(priv_OP_lengths) == OP_TABLE_LENGTH)):
break;
/* ========================================================================== */
case OP_END:
fprintf(f, " %s\n", priv_OP_names[*code]);
fprintf(f, "------------------------------------------------------------------\n");
return;
case OP_CHAR:
fprintf(f, " ");
do
{
code++;
code += 1 + print_char(f, code, utf);
}
while (*code == OP_CHAR);
fprintf(f, "\n");
continue;
case OP_CHARI:
fprintf(f, " /i ");
do
{
code++;
code += 1 + print_char(f, code, utf);
}
while (*code == OP_CHARI);
fprintf(f, "\n");
continue;
case OP_CBRA:
case OP_CBRAPOS:
case OP_SCBRA:
case OP_SCBRAPOS:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s %d", priv_OP_names[*code], GET2(code, 1+LINK_SIZE));
break;
case OP_BRA:
case OP_BRAPOS:
case OP_SBRA:
case OP_SBRAPOS:
case OP_KETRMAX:
case OP_KETRMIN:
case OP_KETRPOS:
case OP_ALT:
case OP_KET:
case OP_ASSERT:
case OP_ASSERT_NOT:
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
case OP_ONCE:
case OP_ONCE_NC:
case OP_COND:
case OP_SCOND:
case OP_REVERSE:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s", priv_OP_names[*code]);
break;
case OP_CLOSE:
fprintf(f, " %s %d", priv_OP_names[*code], GET2(code, 1));
break;
case OP_CREF:
fprintf(f, "%3d %s", GET2(code,1), priv_OP_names[*code]);
break;
case OP_DNCREF:
{
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
IMM2_SIZE;
fprintf(f, " %s Cond ref <", flag);
print_puchar(f, entry);
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
}
break;
case OP_RREF:
c = GET2(code, 1);
if (c == RREF_ANY)
fprintf(f, " Cond recurse any");
else
fprintf(f, " Cond recurse %d", c);
break;
case OP_DNRREF:
{
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
IMM2_SIZE;
fprintf(f, " %s Cond recurse <", flag);
print_puchar(f, entry);
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
}
break;
case OP_DEF:
fprintf(f, " Cond def");
break;
case OP_STARI:
case OP_MINSTARI:
case OP_POSSTARI:
case OP_PLUSI:
case OP_MINPLUSI:
case OP_POSPLUSI:
case OP_QUERYI:
case OP_MINQUERYI:
case OP_POSQUERYI:
flag = "/i";
/* Fall through */
case OP_STAR:
case OP_MINSTAR:
case OP_POSSTAR:
case OP_PLUS:
case OP_MINPLUS:
case OP_POSPLUS:
case OP_QUERY:
case OP_MINQUERY:
case OP_POSQUERY:
case OP_TYPESTAR:
case OP_TYPEMINSTAR:
case OP_TYPEPOSSTAR:
case OP_TYPEPLUS:
case OP_TYPEMINPLUS:
case OP_TYPEPOSPLUS:
case OP_TYPEQUERY:
case OP_TYPEMINQUERY:
case OP_TYPEPOSQUERY:
fprintf(f, " %s ", flag);
if (*code >= OP_TYPESTAR)
{
if (code[1] == OP_PROP || code[1] == OP_NOTPROP)
{
print_prop(f, code + 1, "", " ");
extra = 2;
}
else fprintf(f, "%s", priv_OP_names[code[1]]);
}
else extra = print_char(f, code+1, utf);
fprintf(f, "%s", priv_OP_names[*code]);
break;
case OP_EXACTI:
case OP_UPTOI:
case OP_MINUPTOI:
case OP_POSUPTOI:
flag = "/i";
/* Fall through */
case OP_EXACT:
case OP_UPTO:
case OP_MINUPTO:
case OP_POSUPTO:
fprintf(f, " %s ", flag);
extra = print_char(f, code + 1 + IMM2_SIZE, utf);
fprintf(f, "{");
if (*code != OP_EXACT && *code != OP_EXACTI) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_MINUPTO || *code == OP_MINUPTOI) fprintf(f, "?");
else if (*code == OP_POSUPTO || *code == OP_POSUPTOI) fprintf(f, "+");
break;
case OP_TYPEEXACT:
case OP_TYPEUPTO:
case OP_TYPEMINUPTO:
case OP_TYPEPOSUPTO:
if (code[1 + IMM2_SIZE] == OP_PROP || code[1 + IMM2_SIZE] == OP_NOTPROP)
{
print_prop(f, code + IMM2_SIZE + 1, " ", " ");
extra = 2;
}
else fprintf(f, " %s", priv_OP_names[code[1 + IMM2_SIZE]]);
fprintf(f, "{");
if (*code != OP_TYPEEXACT) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_TYPEMINUPTO) fprintf(f, "?");
else if (*code == OP_TYPEPOSUPTO) fprintf(f, "+");
break;
case OP_NOTI:
flag = "/i";
/* Fall through */
case OP_NOT:
fprintf(f, " %s [^", flag);
extra = print_char(f, code + 1, utf);
fprintf(f, "]");
break;
case OP_NOTSTARI:
case OP_NOTMINSTARI:
case OP_NOTPOSSTARI:
case OP_NOTPLUSI:
case OP_NOTMINPLUSI:
case OP_NOTPOSPLUSI:
case OP_NOTQUERYI:
case OP_NOTMINQUERYI:
case OP_NOTPOSQUERYI:
flag = "/i";
/* Fall through */
case OP_NOTSTAR:
case OP_NOTMINSTAR:
case OP_NOTPOSSTAR:
case OP_NOTPLUS:
case OP_NOTMINPLUS:
case OP_NOTPOSPLUS:
case OP_NOTQUERY:
case OP_NOTMINQUERY:
case OP_NOTPOSQUERY:
fprintf(f, " %s [^", flag);
extra = print_char(f, code + 1, utf);
fprintf(f, "]%s", priv_OP_names[*code]);
break;
case OP_NOTEXACTI:
case OP_NOTUPTOI:
case OP_NOTMINUPTOI:
case OP_NOTPOSUPTOI:
flag = "/i";
/* Fall through */
case OP_NOTEXACT:
case OP_NOTUPTO:
case OP_NOTMINUPTO:
case OP_NOTPOSUPTO:
fprintf(f, " %s [^", flag);
extra = print_char(f, code + 1 + IMM2_SIZE, utf);
fprintf(f, "]{");
if (*code != OP_NOTEXACT && *code != OP_NOTEXACTI) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_NOTMINUPTO || *code == OP_NOTMINUPTOI) fprintf(f, "?");
else
if (*code == OP_NOTPOSUPTO || *code == OP_NOTPOSUPTOI) fprintf(f, "+");
break;
case OP_RECURSE:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s", priv_OP_names[*code]);
break;
case OP_REFI:
flag = "/i";
/* Fall through */
case OP_REF:
fprintf(f, " %s \\%d", flag, GET2(code,1));
ccode = code + priv_OP_lengths[*code];
goto CLASS_REF_REPEAT;
case OP_DNREFI:
flag = "/i";
/* Fall through */
case OP_DNREF:
{
pcre_uchar *entry = (pcre_uchar *)re + offset + (GET2(code, 1) * size) +
IMM2_SIZE;
fprintf(f, " %s \\k<", flag);
print_puchar(f, entry);
fprintf(f, ">%d", GET2(code, 1 + IMM2_SIZE));
}
ccode = code + priv_OP_lengths[*code];
goto CLASS_REF_REPEAT;
case OP_CALLOUT:
fprintf(f, " %s %d %d %d", priv_OP_names[*code], code[1], GET(code,2),
GET(code, 2 + LINK_SIZE));
break;
case OP_PROP:
case OP_NOTPROP:
print_prop(f, code, " ", "");
break;
/* OP_XCLASS cannot occur in 8-bit, non-UTF mode. However, there's no harm
in having this code always here, and it makes it less messy without all
those #ifdefs. */
case OP_CLASS:
case OP_NCLASS:
case OP_XCLASS:
{
int i;
unsigned int min, max;
BOOL printmap;
BOOL invertmap = FALSE;
pcre_uint8 *map;
pcre_uint8 inverted_map[32];
fprintf(f, " [");
if (*code == OP_XCLASS)
{
extra = GET(code, 1);
ccode = code + LINK_SIZE + 1;
printmap = (*ccode & XCL_MAP) != 0;
if ((*ccode & XCL_NOT) != 0)
{
invertmap = (*ccode & XCL_HASPROP) == 0;
fprintf(f, "^");
}
ccode++;
}
else
{
printmap = TRUE;
ccode = code + 1;
}
/* Print a bit map */
if (printmap)
{
map = (pcre_uint8 *)ccode;
if (invertmap)
{
for (i = 0; i < 32; i++) inverted_map[i] = ~map[i];
map = inverted_map;
}
for (i = 0; i < 256; i++)
{
if ((map[i/8] & (1 << (i&7))) != 0)
{
int j;
for (j = i+1; j < 256; j++)
if ((map[j/8] & (1 << (j&7))) == 0) break;
if (i == '-' || i == ']') fprintf(f, "\\");
if (PRINTABLE(i)) fprintf(f, "%c", i);
else fprintf(f, "\\x%02x", i);
if (--j > i)
{
if (j != i + 1) fprintf(f, "-");
if (j == '-' || j == ']') fprintf(f, "\\");
if (PRINTABLE(j)) fprintf(f, "%c", j);
else fprintf(f, "\\x%02x", j);
}
i = j;
}
}
ccode += 32 / sizeof(pcre_uchar);
}
/* For an XCLASS there is always some additional data */
if (*code == OP_XCLASS)
{
pcre_uchar ch;
while ((ch = *ccode++) != XCL_END)
{
BOOL not = FALSE;
const char *notch = "";
switch(ch)
{
case XCL_NOTPROP:
not = TRUE;
notch = "^";
/* Fall through */
case XCL_PROP:
{
unsigned int ptype = *ccode++;
unsigned int pvalue = *ccode++;
switch(ptype)
{
case PT_PXGRAPH:
fprintf(f, "[:%sgraph:]", notch);
break;
case PT_PXPRINT:
fprintf(f, "[:%sprint:]", notch);
break;
case PT_PXPUNCT:
fprintf(f, "[:%spunct:]", notch);
break;
default:
fprintf(f, "\\%c{%s}", (not? 'P':'p'),
get_ucpname(ptype, pvalue));
break;
}
}
break;
default:
ccode += 1 + print_char(f, ccode, utf);
if (ch == XCL_RANGE)
{
fprintf(f, "-");
ccode += 1 + print_char(f, ccode, utf);
}
break;
}
}
}
/* Indicate a non-UTF class which was created by negation */
fprintf(f, "]%s", (*code == OP_NCLASS)? " (neg)" : "");
/* Handle repeats after a class or a back reference */
CLASS_REF_REPEAT:
switch(*ccode)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRPLUS:
case OP_CRMINPLUS:
case OP_CRQUERY:
case OP_CRMINQUERY:
case OP_CRPOSSTAR:
case OP_CRPOSPLUS:
case OP_CRPOSQUERY:
fprintf(f, "%s", priv_OP_names[*ccode]);
extra += priv_OP_lengths[*ccode];
break;
case OP_CRRANGE:
case OP_CRMINRANGE:
case OP_CRPOSRANGE:
min = GET2(ccode,1);
max = GET2(ccode,1 + IMM2_SIZE);
if (max == 0) fprintf(f, "{%u,}", min);
else fprintf(f, "{%u,%u}", min, max);
if (*ccode == OP_CRMINRANGE) fprintf(f, "?");
else if (*ccode == OP_CRPOSRANGE) fprintf(f, "+");
extra += priv_OP_lengths[*ccode];
break;
/* Do nothing if it's not a repeat; this code stops picky compilers
warning about the lack of a default code path. */
default:
break;
}
}
break;
case OP_MARK:
case OP_PRUNE_ARG:
case OP_SKIP_ARG:
case OP_THEN_ARG:
fprintf(f, " %s ", priv_OP_names[*code]);
print_puchar(f, code + 2);
extra += code[1];
break;
case OP_THEN:
fprintf(f, " %s", priv_OP_names[*code]);
break;
case OP_CIRCM:
case OP_DOLLM:
flag = "/m";
/* Fall through */
/* Anything else is just an item with no data, but possibly a flag. */
default:
fprintf(f, " %s %s", flag, priv_OP_names[*code]);
break;
}
code += priv_OP_lengths[*code] + extra;
fprintf(f, "\n");
}
}
/* End of pcre_printint.src */

572
ext/pcre/pcrelib/pcre_printint.src
View file

@ -1,572 +0,0 @@
/*************************************************
* Perl-Compatible Regular Expressions *
*************************************************/
/* PCRE is a library of functions to support regular expressions whose syntax
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2010 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
* Redistributions of source code must retain the above copyright notice,
this list of conditions and the following disclaimer.
* 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.
* Neither the name of the University of Cambridge nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 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 OWNER 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 module contains a PCRE private debugging function for printing out the
internal form of a compiled regular expression, along with some supporting
local functions. This source file is used in two places:
(1) It is #included by pcre_compile.c when it is compiled in debugging mode
(PCRE_DEBUG defined in pcre_internal.h). It is not included in production
compiles.
(2) It is always #included by pcretest.c, which can be asked to print out a
compiled regex for debugging purposes. */
/* Macro that decides whether a character should be output as a literal or in
hexadecimal. We don't use isprint() because that can vary from system to system
(even without the use of locales) and we want the output always to be the same,
for testing purposes. This macro is used in pcretest as well as in this file. */
#ifdef EBCDIC
#define PRINTABLE(c) ((c) >= 64 && (c) < 255)
#else
#define PRINTABLE(c) ((c) >= 32 && (c) < 127)
#endif
/* The table of operator names. */
static const char *OP_names[] = { OP_NAME_LIST };
/*************************************************
* Print single- or multi-byte character *
*************************************************/
static int
print_char(FILE *f, uschar *ptr, BOOL utf8)
{
int c = *ptr;
#ifndef SUPPORT_UTF8
utf8 = utf8; /* Avoid compiler warning */
if (PRINTABLE(c)) fprintf(f, "%c", c); else fprintf(f, "\\x%02x", c);
return 0;
#else
if (!utf8 || (c & 0xc0) != 0xc0)
{
if (PRINTABLE(c)) fprintf(f, "%c", c); else fprintf(f, "\\x%02x", c);
return 0;
}
else
{
int i;
int a = _pcre_utf8_table4[c & 0x3f]; /* Number of additional bytes */
int s = 6*a;
c = (c & _pcre_utf8_table3[a]) << s;
for (i = 1; i <= a; i++)
{
/* This is a check for malformed UTF-8; it should only occur if the sanity
check has been turned off. Rather than swallow random bytes, just stop if
we hit a bad one. Print it with \X instead of \x as an indication. */
if ((ptr[i] & 0xc0) != 0x80)
{
fprintf(f, "\\X{%x}", c);
return i - 1;
}
/* The byte is OK */
s -= 6;
c |= (ptr[i] & 0x3f) << s;
}
if (c < 128) fprintf(f, "\\x%02x", c); else fprintf(f, "\\x{%x}", c);
return a;
}
#endif
}
/*************************************************
* Find Unicode property name *
*************************************************/
static const char *
get_ucpname(int ptype, int pvalue)
{
#ifdef SUPPORT_UCP
int i;
for (i = _pcre_utt_size - 1; i >= 0; i--)
{
if (ptype == _pcre_utt[i].type && pvalue == _pcre_utt[i].value) break;
}
return (i >= 0)? _pcre_utt_names + _pcre_utt[i].name_offset : "??";
#else
/* It gets harder and harder to shut off unwanted compiler warnings. */
ptype = ptype * pvalue;
return (ptype == pvalue)? "??" : "??";
#endif
}
/*************************************************
* Print compiled regex *
*************************************************/
/* Make this function work for a regex with integers either byte order.
However, we assume that what we are passed is a compiled regex. The
print_lengths flag controls whether offsets and lengths of items are printed.
They can be turned off from pcretest so that automatic tests on bytecode can be
written that do not depend on the value of LINK_SIZE. */
static void
pcre_printint(pcre *external_re, FILE *f, BOOL print_lengths)
{
real_pcre *re = (real_pcre *)external_re;
uschar *codestart, *code;
BOOL utf8;
unsigned int options = re->options;
int offset = re->name_table_offset;
int count = re->name_count;
int size = re->name_entry_size;
if (re->magic_number != MAGIC_NUMBER)
{
offset = ((offset << 8) & 0xff00) | ((offset >> 8) & 0xff);
count = ((count << 8) & 0xff00) | ((count >> 8) & 0xff);
size = ((size << 8) & 0xff00) | ((size >> 8) & 0xff);
options = ((options << 24) & 0xff000000) |
((options << 8) & 0x00ff0000) |
((options >> 8) & 0x0000ff00) |
((options >> 24) & 0x000000ff);
}
code = codestart = (uschar *)re + offset + count * size;
utf8 = (options & PCRE_UTF8) != 0;
for(;;)
{
uschar *ccode;
int c;
int extra = 0;
if (print_lengths)
fprintf(f, "%3d ", (int)(code - codestart));
else
fprintf(f, " ");
switch(*code)
{
/* ========================================================================== */
/* These cases are never obeyed. This is a fudge that causes a compile-
time error if the vectors OP_names or _pcre_OP_lengths, which are indexed
by opcode, are not the correct length. It seems to be the only way to do
such a check at compile time, as the sizeof() operator does not work in
the C preprocessor. We do this while compiling pcretest, because that
#includes pcre_tables.c, which holds _pcre_OP_lengths. We can't do this
when building pcre_compile.c with PCRE_DEBUG set, because it doesn't then
know the size of _pcre_OP_lengths. */
#ifdef COMPILING_PCRETEST
case OP_TABLE_LENGTH:
case OP_TABLE_LENGTH +
((sizeof(OP_names)/sizeof(const char *) == OP_TABLE_LENGTH) &&
(sizeof(_pcre_OP_lengths) == OP_TABLE_LENGTH)):
break;
#endif
/* ========================================================================== */
case OP_END:
fprintf(f, " %s\n", OP_names[*code]);
fprintf(f, "------------------------------------------------------------------\n");
return;
case OP_OPT:
fprintf(f, " %.2x %s", code[1], OP_names[*code]);
break;
case OP_CHAR:
fprintf(f, " ");
do
{
code++;
code += 1 + print_char(f, code, utf8);
}
while (*code == OP_CHAR);
fprintf(f, "\n");
continue;
case OP_CHARNC:
fprintf(f, " NC ");
do
{
code++;
code += 1 + print_char(f, code, utf8);
}
while (*code == OP_CHARNC);
fprintf(f, "\n");
continue;
case OP_CBRA:
case OP_SCBRA:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s %d", OP_names[*code], GET2(code, 1+LINK_SIZE));
break;
case OP_BRA:
case OP_SBRA:
case OP_KETRMAX:
case OP_KETRMIN:
case OP_ALT:
case OP_KET:
case OP_ASSERT:
case OP_ASSERT_NOT:
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
case OP_ONCE:
case OP_COND:
case OP_SCOND:
case OP_REVERSE:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s", OP_names[*code]);
break;
case OP_CLOSE:
fprintf(f, " %s %d", OP_names[*code], GET2(code, 1));
break;
case OP_CREF:
case OP_NCREF:
fprintf(f, "%3d %s", GET2(code,1), OP_names[*code]);
break;
case OP_RREF:
c = GET2(code, 1);
if (c == RREF_ANY)
fprintf(f, " Cond recurse any");
else
fprintf(f, " Cond recurse %d", c);
break;
case OP_NRREF:
c = GET2(code, 1);
if (c == RREF_ANY)
fprintf(f, " Cond nrecurse any");
else
fprintf(f, " Cond nrecurse %d", c);
break;
case OP_DEF:
fprintf(f, " Cond def");
break;
case OP_STAR:
case OP_MINSTAR:
case OP_POSSTAR:
case OP_PLUS:
case OP_MINPLUS:
case OP_POSPLUS:
case OP_QUERY:
case OP_MINQUERY:
case OP_POSQUERY:
case OP_TYPESTAR:
case OP_TYPEMINSTAR:
case OP_TYPEPOSSTAR:
case OP_TYPEPLUS:
case OP_TYPEMINPLUS:
case OP_TYPEPOSPLUS:
case OP_TYPEQUERY:
case OP_TYPEMINQUERY:
case OP_TYPEPOSQUERY:
fprintf(f, " ");
if (*code >= OP_TYPESTAR)
{
fprintf(f, "%s", OP_names[code[1]]);
if (code[1] == OP_PROP || code[1] == OP_NOTPROP)
{
fprintf(f, " %s ", get_ucpname(code[2], code[3]));
extra = 2;
}
}
else extra = print_char(f, code+1, utf8);
fprintf(f, "%s", OP_names[*code]);
break;
case OP_EXACT:
case OP_UPTO:
case OP_MINUPTO:
case OP_POSUPTO:
fprintf(f, " ");
extra = print_char(f, code+3, utf8);
fprintf(f, "{");
if (*code != OP_EXACT) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_MINUPTO) fprintf(f, "?");
else if (*code == OP_POSUPTO) fprintf(f, "+");
break;
case OP_TYPEEXACT:
case OP_TYPEUPTO:
case OP_TYPEMINUPTO:
case OP_TYPEPOSUPTO:
fprintf(f, " %s", OP_names[code[3]]);
if (code[3] == OP_PROP || code[3] == OP_NOTPROP)
{
fprintf(f, " %s ", get_ucpname(code[4], code[5]));
extra = 2;
}
fprintf(f, "{");
if (*code != OP_TYPEEXACT) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_TYPEMINUPTO) fprintf(f, "?");
else if (*code == OP_TYPEPOSUPTO) fprintf(f, "+");
break;
case OP_NOT:
c = code[1];
if (PRINTABLE(c)) fprintf(f, " [^%c]", c);
else fprintf(f, " [^\\x%02x]", c);
break;
case OP_NOTSTAR:
case OP_NOTMINSTAR:
case OP_NOTPOSSTAR:
case OP_NOTPLUS:
case OP_NOTMINPLUS:
case OP_NOTPOSPLUS:
case OP_NOTQUERY:
case OP_NOTMINQUERY:
case OP_NOTPOSQUERY:
c = code[1];
if (PRINTABLE(c)) fprintf(f, " [^%c]", c);
else fprintf(f, " [^\\x%02x]", c);
fprintf(f, "%s", OP_names[*code]);
break;
case OP_NOTEXACT:
case OP_NOTUPTO:
case OP_NOTMINUPTO:
case OP_NOTPOSUPTO:
c = code[3];
if (PRINTABLE(c)) fprintf(f, " [^%c]{", c);
else fprintf(f, " [^\\x%02x]{", c);
if (*code != OP_NOTEXACT) fprintf(f, "0,");
fprintf(f, "%d}", GET2(code,1));
if (*code == OP_NOTMINUPTO) fprintf(f, "?");
else if (*code == OP_NOTPOSUPTO) fprintf(f, "+");
break;
case OP_RECURSE:
if (print_lengths) fprintf(f, "%3d ", GET(code, 1));
else fprintf(f, " ");
fprintf(f, "%s", OP_names[*code]);
break;
case OP_REF:
fprintf(f, " \\%d", GET2(code,1));
ccode = code + _pcre_OP_lengths[*code];
goto CLASS_REF_REPEAT;
case OP_CALLOUT:
fprintf(f, " %s %d %d %d", OP_names[*code], code[1], GET(code,2),
GET(code, 2 + LINK_SIZE));
break;
case OP_PROP:
case OP_NOTPROP:
fprintf(f, " %s %s", OP_names[*code], get_ucpname(code[1], code[2]));
break;
/* OP_XCLASS can only occur in UTF-8 mode. However, there's no harm in
having this code always here, and it makes it less messy without all those
#ifdefs. */
case OP_CLASS:
case OP_NCLASS:
case OP_XCLASS:
{
int i, min, max;
BOOL printmap;
fprintf(f, " [");
if (*code == OP_XCLASS)
{
extra = GET(code, 1);
ccode = code + LINK_SIZE + 1;
printmap = (*ccode & XCL_MAP) != 0;
if ((*ccode++ & XCL_NOT) != 0) fprintf(f, "^");
}
else
{
printmap = TRUE;
ccode = code + 1;
}
/* Print a bit map */
if (printmap)
{
for (i = 0; i < 256; i++)
{
if ((ccode[i/8] & (1 << (i&7))) != 0)
{
int j;
for (j = i+1; j < 256; j++)
if ((ccode[j/8] & (1 << (j&7))) == 0) break;
if (i == '-' || i == ']') fprintf(f, "\\");
if (PRINTABLE(i)) fprintf(f, "%c", i);
else fprintf(f, "\\x%02x", i);
if (--j > i)
{
if (j != i + 1) fprintf(f, "-");
if (j == '-' || j == ']') fprintf(f, "\\");
if (PRINTABLE(j)) fprintf(f, "%c", j);
else fprintf(f, "\\x%02x", j);
}
i = j;
}
}
ccode += 32;
}
/* For an XCLASS there is always some additional data */
if (*code == OP_XCLASS)
{
int ch;
while ((ch = *ccode++) != XCL_END)
{
if (ch == XCL_PROP)
{
int ptype = *ccode++;
int pvalue = *ccode++;
fprintf(f, "\\p{%s}", get_ucpname(ptype, pvalue));
}
else if (ch == XCL_NOTPROP)
{
int ptype = *ccode++;
int pvalue = *ccode++;
fprintf(f, "\\P{%s}", get_ucpname(ptype, pvalue));
}
else
{
ccode += 1 + print_char(f, ccode, TRUE);
if (ch == XCL_RANGE)
{
fprintf(f, "-");
ccode += 1 + print_char(f, ccode, TRUE);
}
}
}
}
/* Indicate a non-UTF8 class which was created by negation */
fprintf(f, "]%s", (*code == OP_NCLASS)? " (neg)" : "");
/* Handle repeats after a class or a back reference */
CLASS_REF_REPEAT:
switch(*ccode)
{
case OP_CRSTAR:
case OP_CRMINSTAR:
case OP_CRPLUS:
case OP_CRMINPLUS:
case OP_CRQUERY:
case OP_CRMINQUERY:
fprintf(f, "%s", OP_names[*ccode]);
extra += _pcre_OP_lengths[*ccode];
break;
case OP_CRRANGE:
case OP_CRMINRANGE:
min = GET2(ccode,1);
max = GET2(ccode,3);
if (max == 0) fprintf(f, "{%d,}", min);
else fprintf(f, "{%d,%d}", min, max);
if (*ccode == OP_CRMINRANGE) fprintf(f, "?");
extra += _pcre_OP_lengths[*ccode];
break;
/* Do nothing if it's not a repeat; this code stops picky compilers
warning about the lack of a default code path. */
default:
break;
}
}
break;
case OP_MARK:
case OP_PRUNE_ARG:
case OP_SKIP_ARG:
fprintf(f, " %s %s", OP_names[*code], code + 2);
extra += code[1];
break;
case OP_THEN:
if (print_lengths)
fprintf(f, " %s %d", OP_names[*code], GET(code, 1));
else
fprintf(f, " %s", OP_names[*code]);
break;
case OP_THEN_ARG:
if (print_lengths)
fprintf(f, " %s %d %s", OP_names[*code], GET(code, 1),
code + 2 + LINK_SIZE);
else
fprintf(f, " %s %s", OP_names[*code], code + 2 + LINK_SIZE);
extra += code[1+LINK_SIZE];
break;
/* Anything else is just an item with no data*/
default:
fprintf(f, " %s", OP_names[*code]);
break;
}
code += _pcre_OP_lengths[*code] + extra;
fprintf(f, "\n");
}
}
/* End of pcre_printint.src */

2
ext/pcre/pcrelib/pcre_refcount.c
View file

@ -44,9 +44,7 @@ pattern data block. This might be helpful in applications where the block is
shared by different users. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

72
ext/pcre/pcrelib/pcre_study.c
View file

@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
supporting functions. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"
@ -879,9 +877,6 @@ do
case OP_SOM:
case OP_THEN:
case OP_THEN_ARG:
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
case OP_XCLASS:
#endif
return SSB_FAIL;
/* We can ignore word boundary tests. */
@ -1257,6 +1252,16 @@ do
with a value >= 0xc4 is a potentially valid starter because it starts a
character with a value > 255. */
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
case OP_XCLASS:
if ((tcode[1 + LINK_SIZE] & XCL_HASPROP) != 0)
return SSB_FAIL;
/* All bits are set. */
if ((tcode[1 + LINK_SIZE] & XCL_MAP) == 0 && (tcode[1 + LINK_SIZE] & XCL_NOT) != 0)
return SSB_FAIL;
#endif
/* Fall through */
case OP_NCLASS:
#if defined SUPPORT_UTF && defined COMPILE_PCRE8
if (utf)
@ -1273,8 +1278,21 @@ do
case OP_CLASS:
{
pcre_uint8 *map;
tcode++;
map = (pcre_uint8 *)tcode;
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
map = NULL;
if (*tcode == OP_XCLASS)
{
if ((tcode[1 + LINK_SIZE] & XCL_MAP) != 0)
map = (pcre_uint8 *)(tcode + 1 + LINK_SIZE + 1);
tcode += GET(tcode, 1);
}
else
#endif
{
tcode++;
map = (pcre_uint8 *)tcode;
tcode += 32 / sizeof(pcre_uchar);
}
/* In UTF-8 mode, the bits in a bit map correspond to character
values, not to byte values. However, the bit map we are constructing is
@ -1282,31 +1300,35 @@ do
value is > 127. In fact, there are only two possible starting bytes for
characters in the range 128 - 255. */
#if defined SUPPORT_UTF && defined COMPILE_PCRE8
if (utf)
{
for (c = 0; c < 16; c++) start_bits[c] |= map[c];
for (c = 128; c < 256; c++)
{
if ((map[c/8] && (1 << (c&7))) != 0)
{
int d = (c >> 6) | 0xc0; /* Set bit for this starter */
start_bits[d/8] |= (1 << (d&7)); /* and then skip on to the */
c = (c & 0xc0) + 0x40 - 1; /* next relevant character. */
}
}
}
else
#if defined SUPPORT_UTF || !defined COMPILE_PCRE8
if (map != NULL)
#endif
{
/* In non-UTF-8 mode, the two bit maps are completely compatible. */
for (c = 0; c < 32; c++) start_bits[c] |= map[c];
#if defined SUPPORT_UTF && defined COMPILE_PCRE8
if (utf)
{
for (c = 0; c < 16; c++) start_bits[c] |= map[c];
for (c = 128; c < 256; c++)
{
if ((map[c/8] && (1 << (c&7))) != 0)
{
int d = (c >> 6) | 0xc0; /* Set bit for this starter */
start_bits[d/8] |= (1 << (d&7)); /* and then skip on to the */
c = (c & 0xc0) + 0x40 - 1; /* next relevant character. */
}
}
}
else
#endif
{
/* In non-UTF-8 mode, the two bit maps are completely compatible. */
for (c = 0; c < 32; c++) start_bits[c] |= map[c];
}
}
/* Advance past the bit map, and act on what follows. For a zero
minimum repeat, continue; otherwise stop processing. */
tcode += 32 / sizeof(pcre_uchar);
switch (*tcode)
{
case OP_CRSTAR:

2
ext/pcre/pcrelib/pcre_tables.c
View file

@ -45,9 +45,7 @@ uses macros to change their names from _pcre_xxx to xxxx, thereby avoiding name
clashes with the library. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

2
ext/pcre/pcrelib/pcre_ucd.c
View file

@ -10,9 +10,7 @@ needed. */
#ifndef PCRE_INCLUDED
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

2
ext/pcre/pcrelib/pcre_valid_utf8.c
View file

@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
strings. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

2
ext/pcre/pcrelib/pcre_version.c
View file

@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
string that identifies the PCRE version that is in use. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"

7
ext/pcre/pcrelib/pcre_xclass.c
View file

@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
class. It is used by both pcre_exec() and pcre_def_exec(). */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
#include "pcre_internal.h"
@ -81,6 +79,11 @@ additional data. */
if (c < 256)
{
if ((*data & XCL_HASPROP) == 0)
{
if ((*data & XCL_MAP) == 0) return negated;
return (((pcre_uint8 *)(data + 1))[c/8] & (1 << (c&7))) != 0;
}
if ((*data & XCL_MAP) != 0 &&
(((pcre_uint8 *)(data + 1))[c/8] & (1 << (c&7))) != 0)
return !negated; /* char found */

8
ext/pcre/pcrelib/pcreposix.c
View file

@ -6,7 +6,7 @@
and semantics are as close as possible to those of the Perl 5 language.
Written by Philip Hazel
Copyright (c) 1997-2012 University of Cambridge
Copyright (c) 1997-2014 University of Cambridge
-----------------------------------------------------------------------------
Redistribution and use in source and binary forms, with or without
@ -42,9 +42,7 @@ POSSIBILITY OF SUCH DAMAGE.
functions. */
#ifdef HAVE_CONFIG_H
#include "config.h"
#endif
/* Ensure that the PCREPOSIX_EXP_xxx macros are set appropriately for
@ -170,7 +168,9 @@ static const int eint[] = {
REG_BADPAT, /* missing opening brace after \o */
REG_BADPAT, /* parentheses too deeply nested */
REG_BADPAT, /* invalid range in character class */
REG_BADPAT /* group name must start with a non-digit */
REG_BADPAT, /* group name must start with a non-digit */
/* 85 */
REG_BADPAT /* parentheses too deeply nested (stack check) */
};
/* Table of texts corresponding to POSIX error codes */

120
ext/pcre/pcrelib/sljit/sljitConfig.h Normal file
View file

@ -0,0 +1,120 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
#ifndef _SLJIT_CONFIG_H_
#define _SLJIT_CONFIG_H_
/* --------------------------------------------------------------------- */
/* Custom defines */
/* --------------------------------------------------------------------- */
/* Put your custom defines here. This empty section will never change
which helps maintaining patches (with diff / patch utilities). */
/* --------------------------------------------------------------------- */
/* Architecture */
/* --------------------------------------------------------------------- */
/* Architecture selection. */
/* #define SLJIT_CONFIG_X86_32 1 */
/* #define SLJIT_CONFIG_X86_64 1 */
/* #define SLJIT_CONFIG_ARM_V5 1 */
/* #define SLJIT_CONFIG_ARM_V7 1 */
/* #define SLJIT_CONFIG_ARM_THUMB2 1 */
/* #define SLJIT_CONFIG_ARM_64 1 */
/* #define SLJIT_CONFIG_PPC_32 1 */
/* #define SLJIT_CONFIG_PPC_64 1 */
/* #define SLJIT_CONFIG_MIPS_32 1 */
/* #define SLJIT_CONFIG_MIPS_64 1 */
/* #define SLJIT_CONFIG_SPARC_32 1 */
/* #define SLJIT_CONFIG_TILEGX 1 */
/* #define SLJIT_CONFIG_AUTO 1 */
/* #define SLJIT_CONFIG_UNSUPPORTED 1 */
/* --------------------------------------------------------------------- */
/* Utilities */
/* --------------------------------------------------------------------- */
/* Useful for thread-safe compiling of global functions. */
#ifndef SLJIT_UTIL_GLOBAL_LOCK
/* Enabled by default */
#define SLJIT_UTIL_GLOBAL_LOCK 1
#endif
/* Implements a stack like data structure (by using mmap / VirtualAlloc). */
#ifndef SLJIT_UTIL_STACK
/* Enabled by default */
#define SLJIT_UTIL_STACK 1
#endif
/* Single threaded application. Does not require any locks. */
#ifndef SLJIT_SINGLE_THREADED
/* Disabled by default. */
#define SLJIT_SINGLE_THREADED 0
#endif
/* --------------------------------------------------------------------- */
/* Configuration */
/* --------------------------------------------------------------------- */
/* If SLJIT_STD_MACROS_DEFINED is not defined, the application should
define SLJIT_MALLOC, SLJIT_FREE, SLJIT_MEMMOVE, and NULL. */
#ifndef SLJIT_STD_MACROS_DEFINED
/* Disabled by default. */
#define SLJIT_STD_MACROS_DEFINED 0
#endif
/* Executable code allocation:
If SLJIT_EXECUTABLE_ALLOCATOR is not defined, the application should
define both SLJIT_MALLOC_EXEC and SLJIT_FREE_EXEC. */
#ifndef SLJIT_EXECUTABLE_ALLOCATOR
/* Enabled by default. */
#define SLJIT_EXECUTABLE_ALLOCATOR 1
#endif
/* Debug checks (assertions, etc.). */
#ifndef SLJIT_DEBUG
/* Enabled by default */
#define SLJIT_DEBUG 1
#endif
/* Verbose operations */
#ifndef SLJIT_VERBOSE
/* Enabled by default */
#define SLJIT_VERBOSE 1
#endif
/*
SLJIT_IS_FPU_AVAILABLE
The availability of the FPU can be controlled by SLJIT_IS_FPU_AVAILABLE.
zero value - FPU is NOT present.
nonzero value - FPU is present.
*/
/* For further configurations, see the beginning of sljitConfigInternal.h */
#endif

523
ext/pcre/pcrelib/sljit/sljitConfigInternal.h Normal file
View file

@ -0,0 +1,523 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
#ifndef _SLJIT_CONFIG_INTERNAL_H_
#define _SLJIT_CONFIG_INTERNAL_H_
/*
SLJIT defines the following macros depending on the target architecture:
Feature detection (boolean) macros:
SLJIT_32BIT_ARCHITECTURE : 32 bit architecture
SLJIT_64BIT_ARCHITECTURE : 64 bit architecture
SLJIT_WORD_SHIFT : the shift required to apply when accessing a sljit_sw/sljit_uw array by index
SLJIT_DOUBLE_SHIFT : the shift required to apply when accessing a double array by index
SLJIT_LITTLE_ENDIAN : little endian architecture
SLJIT_BIG_ENDIAN : big endian architecture
SLJIT_UNALIGNED : allows unaligned memory accesses for non-fpu operations (only!)
SLJIT_INDIRECT_CALL : see SLJIT_FUNC_OFFSET() for more information
SLJIT_RETURN_ADDRESS_OFFSET : a return instruction always adds this offset to the return address
Types and useful macros:
sljit_sb, sljit_ub : signed and unsigned 8 bit byte
sljit_sh, sljit_uh : signed and unsigned 16 bit half-word (short) type
sljit_si, sljit_ui : signed and unsigned 32 bit integer type
sljit_sw, sljit_uw : signed and unsigned machine word, enough to store a pointer
sljit_p : unsgined pointer value (usually the same as sljit_uw, but
some 64 bit ABIs may use 32 bit pointers)
sljit_s : single precision floating point value
sljit_d : double precision floating point value
SLJIT_CALL : C calling convention define for both calling JIT form C and C callbacks for JIT
SLJIT_W(number) : defining 64 bit constants on 64 bit architectures (compiler independent helper)
*/
#if !((defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) \
|| (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) \
|| (defined SLJIT_CONFIG_ARM_V5 && SLJIT_CONFIG_ARM_V5) \
|| (defined SLJIT_CONFIG_ARM_V7 && SLJIT_CONFIG_ARM_V7) \
|| (defined SLJIT_CONFIG_ARM_THUMB2 && SLJIT_CONFIG_ARM_THUMB2) \
|| (defined SLJIT_CONFIG_ARM_64 && SLJIT_CONFIG_ARM_64) \
|| (defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) \
|| (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64) \
|| (defined SLJIT_CONFIG_MIPS_32 && SLJIT_CONFIG_MIPS_32) \
|| (defined SLJIT_CONFIG_MIPS_64 && SLJIT_CONFIG_MIPS_64) \
|| (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32) \
|| (defined SLJIT_CONFIG_TILEGX && SLJIT_CONFIG_TILEGX) \
|| (defined SLJIT_CONFIG_AUTO && SLJIT_CONFIG_AUTO) \
|| (defined SLJIT_CONFIG_UNSUPPORTED && SLJIT_CONFIG_UNSUPPORTED))
#error "An architecture must be selected"
#endif
/* Sanity check. */
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) \
+ (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) \
+ (defined SLJIT_CONFIG_ARM_V5 && SLJIT_CONFIG_ARM_V5) \
+ (defined SLJIT_CONFIG_ARM_V7 && SLJIT_CONFIG_ARM_V7) \
+ (defined SLJIT_CONFIG_ARM_THUMB2 && SLJIT_CONFIG_ARM_THUMB2) \
+ (defined SLJIT_CONFIG_ARM_64 && SLJIT_CONFIG_ARM_64) \
+ (defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) \
+ (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64) \
+ (defined SLJIT_CONFIG_TILEGX && SLJIT_CONFIG_TILEGX) \
+ (defined SLJIT_CONFIG_MIPS_32 && SLJIT_CONFIG_MIPS_32) \
+ (defined SLJIT_CONFIG_MIPS_64 && SLJIT_CONFIG_MIPS_64) \
+ (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32) \
+ (defined SLJIT_CONFIG_AUTO && SLJIT_CONFIG_AUTO) \
+ (defined SLJIT_CONFIG_UNSUPPORTED && SLJIT_CONFIG_UNSUPPORTED) >= 2
#error "Multiple architectures are selected"
#endif
/* Auto select option (requires compiler support) */
#if (defined SLJIT_CONFIG_AUTO && SLJIT_CONFIG_AUTO)
#ifndef _WIN32
#if defined(__i386__) || defined(__i386)
#define SLJIT_CONFIG_X86_32 1
#elif defined(__x86_64__)
#define SLJIT_CONFIG_X86_64 1
#elif defined(__arm__) || defined(__ARM__)
#ifdef __thumb2__
#define SLJIT_CONFIG_ARM_THUMB2 1
#elif defined(__ARM_ARCH_7__) || defined(__ARM_ARCH_7A__) || defined(__ARM_ARCH_7R__)
#define SLJIT_CONFIG_ARM_V7 1
#else
#define SLJIT_CONFIG_ARM_V5 1
#endif
#elif defined (__aarch64__)
#define SLJIT_CONFIG_ARM_64 1
#elif defined(__ppc64__) || defined(__powerpc64__) || defined(_ARCH_PPC64) || (defined(_POWER) && defined(__64BIT__))
#define SLJIT_CONFIG_PPC_64 1
#elif defined(__ppc__) || defined(__powerpc__) || defined(_ARCH_PPC) || defined(_ARCH_PWR) || defined(_ARCH_PWR2) || defined(_POWER)
#define SLJIT_CONFIG_PPC_32 1
#elif defined(__mips__) && !defined(_LP64)
#define SLJIT_CONFIG_MIPS_32 1
#elif defined(__mips64)
#define SLJIT_CONFIG_MIPS_64 1
#elif defined(__sparc__) || defined(__sparc)
#define SLJIT_CONFIG_SPARC_32 1
#elif defined(__tilegx__)
#define SLJIT_CONFIG_TILEGX 1
#else
/* Unsupported architecture */
#define SLJIT_CONFIG_UNSUPPORTED 1
#endif
#else /* !_WIN32 */
#if defined(_M_X64) || defined(__x86_64__)
#define SLJIT_CONFIG_X86_64 1
#elif defined(_ARM_)
#define SLJIT_CONFIG_ARM_V5 1
#else
#define SLJIT_CONFIG_X86_32 1
#endif
#endif /* !WIN32 */
#endif /* SLJIT_CONFIG_AUTO */
#if (defined SLJIT_CONFIG_UNSUPPORTED && SLJIT_CONFIG_UNSUPPORTED)
#undef SLJIT_EXECUTABLE_ALLOCATOR
#endif
#if !(defined SLJIT_STD_MACROS_DEFINED && SLJIT_STD_MACROS_DEFINED)
/* These libraries are needed for the macros below. */
#include <stdlib.h>
#include <string.h>
#endif /* STD_MACROS_DEFINED */
/* General macros:
Note: SLJIT is designed to be independent from them as possible.
In release mode (SLJIT_DEBUG is not defined) only the following macros are needed:
*/
#ifndef SLJIT_MALLOC
#define SLJIT_MALLOC(size) malloc(size)
#endif
#ifndef SLJIT_FREE
#define SLJIT_FREE(ptr) free(ptr)
#endif
#ifndef SLJIT_MEMMOVE
#define SLJIT_MEMMOVE(dest, src, len) memmove(dest, src, len)
#endif
#ifndef SLJIT_ZEROMEM
#define SLJIT_ZEROMEM(dest, len) memset(dest, 0, len)
#endif
#if !defined(SLJIT_LIKELY) && !defined(SLJIT_UNLIKELY)
#if defined(__GNUC__) && (__GNUC__ >= 3)
#define SLJIT_LIKELY(x) __builtin_expect((x), 1)
#define SLJIT_UNLIKELY(x) __builtin_expect((x), 0)
#else
#define SLJIT_LIKELY(x) (x)
#define SLJIT_UNLIKELY(x) (x)
#endif
#endif /* !defined(SLJIT_LIKELY) && !defined(SLJIT_UNLIKELY) */
#ifndef SLJIT_INLINE
/* Inline functions. Some old compilers do not support them. */
#if defined(__SUNPRO_C) && __SUNPRO_C <= 0x510
#define SLJIT_INLINE
#else
#define SLJIT_INLINE __inline
#endif
#endif /* !SLJIT_INLINE */
#ifndef SLJIT_CONST
/* Const variables. */
#define SLJIT_CONST const
#endif
#ifndef SLJIT_UNUSED_ARG
/* Unused arguments. */
#define SLJIT_UNUSED_ARG(arg) (void)arg
#endif
#if (defined SLJIT_CONFIG_STATIC && SLJIT_CONFIG_STATIC)
/* Static ABI functions. For all-in-one programs. */
#if defined(__GNUC__)
/* Disable unused warnings in gcc. */
#define SLJIT_API_FUNC_ATTRIBUTE static __attribute__((unused))
#else
#define SLJIT_API_FUNC_ATTRIBUTE static
#endif
#else
#define SLJIT_API_FUNC_ATTRIBUTE
#endif /* (defined SLJIT_CONFIG_STATIC && SLJIT_CONFIG_STATIC) */
#ifndef SLJIT_CACHE_FLUSH
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) || (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
/* Not required to implement on archs with unified caches. */
#define SLJIT_CACHE_FLUSH(from, to)
#elif defined __APPLE__
/* Supported by all macs since Mac OS 10.5.
However, it does not work on non-jailbroken iOS devices,
although the compilation is successful. */
#define SLJIT_CACHE_FLUSH(from, to) \
sys_icache_invalidate((char*)(from), (char*)(to) - (char*)(from))
#elif defined __ANDROID__
/* Android lacks __clear_cache; instead, cacheflush should be used. */
#define SLJIT_CACHE_FLUSH(from, to) \
cacheflush((long)(from), (long)(to), 0)
#elif (defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) || (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64)
/* The __clear_cache() implementation of GCC is a dummy function on PowerPC. */
#define SLJIT_CACHE_FLUSH(from, to) \
ppc_cache_flush((from), (to))
#elif (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32)
/* The __clear_cache() implementation of GCC is a dummy function on Sparc. */
#define SLJIT_CACHE_FLUSH(from, to) \
sparc_cache_flush((from), (to))
#else
/* Calls __ARM_NR_cacheflush on ARM-Linux. */
#define SLJIT_CACHE_FLUSH(from, to) \
__clear_cache((char*)(from), (char*)(to))
#endif
#endif /* !SLJIT_CACHE_FLUSH */
/* 8 bit byte type. */
typedef unsigned char sljit_ub;
typedef signed char sljit_sb;
/* 16 bit half-word type. */
typedef unsigned short int sljit_uh;
typedef signed short int sljit_sh;
/* 32 bit integer type. */
typedef unsigned int sljit_ui;
typedef signed int sljit_si;
/* Machine word type. Can encapsulate a pointer.
32 bit for 32 bit machines.
64 bit for 64 bit machines. */
#if (defined SLJIT_CONFIG_UNSUPPORTED && SLJIT_CONFIG_UNSUPPORTED)
/* Just to have something. */
#define SLJIT_WORD_SHIFT 0
typedef unsigned long int sljit_uw;
typedef long int sljit_sw;
#elif !(defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) \
&& !(defined SLJIT_CONFIG_ARM_64 && SLJIT_CONFIG_ARM_64) \
&& !(defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64) \
&& !(defined SLJIT_CONFIG_MIPS_64 && SLJIT_CONFIG_MIPS_64) \
&& !(defined SLJIT_CONFIG_TILEGX && SLJIT_CONFIG_TILEGX)
#define SLJIT_32BIT_ARCHITECTURE 1
#define SLJIT_WORD_SHIFT 2
typedef unsigned int sljit_uw;
typedef int sljit_sw;
#else
#define SLJIT_64BIT_ARCHITECTURE 1
#define SLJIT_WORD_SHIFT 3
#ifdef _WIN32
typedef unsigned __int64 sljit_uw;
typedef __int64 sljit_sw;
#else
typedef unsigned long int sljit_uw;
typedef long int sljit_sw;
#endif
#endif
typedef sljit_uw sljit_p;
/* Floating point types. */
typedef float sljit_s;
typedef double sljit_d;
/* Shift for pointer sized data. */
#define SLJIT_POINTER_SHIFT SLJIT_WORD_SHIFT
/* Shift for double precision sized data. */
#define SLJIT_DOUBLE_SHIFT 3
#ifndef SLJIT_W
/* Defining long constants. */
#if (defined SLJIT_64BIT_ARCHITECTURE && SLJIT_64BIT_ARCHITECTURE)
#define SLJIT_W(w) (w##ll)
#else
#define SLJIT_W(w) (w)
#endif
#endif /* !SLJIT_W */
#ifndef SLJIT_CALL
/* ABI (Application Binary Interface) types. */
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32)
#if defined(__GNUC__) && !defined(__APPLE__)
#define SLJIT_CALL __attribute__ ((fastcall))
#define SLJIT_X86_32_FASTCALL 1
#elif defined(_MSC_VER)
#define SLJIT_CALL __fastcall
#define SLJIT_X86_32_FASTCALL 1
#elif defined(__BORLANDC__)
#define SLJIT_CALL __msfastcall
#define SLJIT_X86_32_FASTCALL 1
#else /* Unknown compiler. */
/* The cdecl attribute is the default. */
#define SLJIT_CALL
#endif
#else /* Non x86-32 architectures. */
#define SLJIT_CALL
#endif /* SLJIT_CONFIG_X86_32 */
#endif /* !SLJIT_CALL */
#if !defined(SLJIT_BIG_ENDIAN) && !defined(SLJIT_LITTLE_ENDIAN)
/* These macros are useful for the applications. */
#if (defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) \
|| (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64)
#ifdef __LITTLE_ENDIAN__
#define SLJIT_LITTLE_ENDIAN 1
#else
#define SLJIT_BIG_ENDIAN 1
#endif
#elif (defined SLJIT_CONFIG_MIPS_32 && SLJIT_CONFIG_MIPS_32) \
|| (defined SLJIT_CONFIG_MIPS_64 && SLJIT_CONFIG_MIPS_64)
#ifdef __MIPSEL__
#define SLJIT_LITTLE_ENDIAN 1
#else
#define SLJIT_BIG_ENDIAN 1
#endif
#elif (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32)
#define SLJIT_BIG_ENDIAN 1
#else
#define SLJIT_LITTLE_ENDIAN 1
#endif
#endif /* !defined(SLJIT_BIG_ENDIAN) && !defined(SLJIT_LITTLE_ENDIAN) */
/* Sanity check. */
#if (defined SLJIT_BIG_ENDIAN && SLJIT_BIG_ENDIAN) && (defined SLJIT_LITTLE_ENDIAN && SLJIT_LITTLE_ENDIAN)
#error "Exactly one endianness must be selected"
#endif
#if !(defined SLJIT_BIG_ENDIAN && SLJIT_BIG_ENDIAN) && !(defined SLJIT_LITTLE_ENDIAN && SLJIT_LITTLE_ENDIAN)
#error "Exactly one endianness must be selected"
#endif
#ifndef SLJIT_INDIRECT_CALL
#if ((defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64) && (defined SLJIT_BIG_ENDIAN && SLJIT_BIG_ENDIAN)) \
|| ((defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) && defined _AIX)
/* It seems certain ppc compilers use an indirect addressing for functions
which makes things complicated. */
#define SLJIT_INDIRECT_CALL 1
#endif
#endif /* SLJIT_INDIRECT_CALL */
#ifndef SLJIT_RETURN_ADDRESS_OFFSET
#if (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32)
#define SLJIT_RETURN_ADDRESS_OFFSET 8
#else
#define SLJIT_RETURN_ADDRESS_OFFSET 0
#endif
#endif /* SLJIT_RETURN_ADDRESS_OFFSET */
#ifndef SLJIT_SSE2
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) || (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
/* Turn on SSE2 support on x86. */
#define SLJIT_SSE2 1
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32)
/* Auto detect SSE2 support using CPUID.
On 64 bit x86 cpus, sse2 must be present. */
#define SLJIT_DETECT_SSE2 1
#endif
#endif /* (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) || (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) */
#endif /* !SLJIT_SSE2 */
#ifndef SLJIT_UNALIGNED
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) \
|| (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) \
|| (defined SLJIT_CONFIG_ARM_V7 && SLJIT_CONFIG_ARM_V7) \
|| (defined SLJIT_CONFIG_ARM_THUMB2 && SLJIT_CONFIG_ARM_THUMB2) \
|| (defined SLJIT_CONFIG_ARM_64 && SLJIT_CONFIG_ARM_64) \
|| (defined SLJIT_CONFIG_PPC_32 && SLJIT_CONFIG_PPC_32) \
|| (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64)
#define SLJIT_UNALIGNED 1
#endif
#endif /* !SLJIT_UNALIGNED */
#if (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR)
SLJIT_API_FUNC_ATTRIBUTE void* sljit_malloc_exec(sljit_uw size);
SLJIT_API_FUNC_ATTRIBUTE void sljit_free_exec(void* ptr);
SLJIT_API_FUNC_ATTRIBUTE void sljit_free_unused_memory_exec(void);
#define SLJIT_MALLOC_EXEC(size) sljit_malloc_exec(size)
#define SLJIT_FREE_EXEC(ptr) sljit_free_exec(ptr)
#endif
#if (defined SLJIT_VERBOSE && SLJIT_VERBOSE)
#include <stdio.h>
#endif
#if (defined SLJIT_DEBUG && SLJIT_DEBUG)
#if !defined(SLJIT_ASSERT) || !defined(SLJIT_ASSERT_STOP)
/* SLJIT_HALT_PROCESS must halt the process. */
#ifndef SLJIT_HALT_PROCESS
#include <stdlib.h>
#define SLJIT_HALT_PROCESS() \
abort();
#endif /* !SLJIT_HALT_PROCESS */
#include <stdio.h>
#endif /* !SLJIT_ASSERT || !SLJIT_ASSERT_STOP */
/* Feel free to redefine these two macros. */
#ifndef SLJIT_ASSERT
#define SLJIT_ASSERT(x) \
do { \
if (SLJIT_UNLIKELY(!(x))) { \
printf("Assertion failed at " __FILE__ ":%d\n", __LINE__); \
SLJIT_HALT_PROCESS(); \
} \
} while (0)
#endif /* !SLJIT_ASSERT */
#ifndef SLJIT_ASSERT_STOP
#define SLJIT_ASSERT_STOP() \
do { \
printf("Should never been reached " __FILE__ ":%d\n", __LINE__); \
SLJIT_HALT_PROCESS(); \
} while (0)
#endif /* !SLJIT_ASSERT_STOP */
#else /* (defined SLJIT_DEBUG && SLJIT_DEBUG) */
/* Forcing empty, but valid statements. */
#undef SLJIT_ASSERT
#undef SLJIT_ASSERT_STOP
#define SLJIT_ASSERT(x) \
do { } while (0)
#define SLJIT_ASSERT_STOP() \
do { } while (0)
#endif /* (defined SLJIT_DEBUG && SLJIT_DEBUG) */
#ifndef SLJIT_COMPILE_ASSERT
/* Should be improved eventually. */
#define SLJIT_COMPILE_ASSERT(x, description) \
SLJIT_ASSERT(x)
#endif /* !SLJIT_COMPILE_ASSERT */
#endif

312
ext/pcre/pcrelib/sljit/sljitExecAllocator.c Normal file
View file

@ -0,0 +1,312 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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
/*
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
*/
#ifdef _WIN32
static SLJIT_INLINE void* alloc_chunk(sljit_uw size)
{
return VirtualAlloc(NULL, size, MEM_COMMIT | MEM_RESERVE, PAGE_EXECUTE_READWRITE);
}
static SLJIT_INLINE void free_chunk(void* chunk, sljit_uw size)
{
SLJIT_UNUSED_ARG(size);
VirtualFree(chunk, 0, MEM_RELEASE);
}
#else
static SLJIT_INLINE void* alloc_chunk(sljit_uw size)
{
void* retval;
#ifdef MAP_ANON
retval = mmap(NULL, size, PROT_READ | PROT_WRITE | PROT_EXEC, MAP_PRIVATE | MAP_ANON, -1, 0);
#else
if (dev_zero < 0) {
if (open_dev_zero())
return NULL;
}
retval = mmap(NULL, size, PROT_READ | PROT_WRITE | PROT_EXEC, MAP_PRIVATE, dev_zero, 0);
#endif
return (retval != MAP_FAILED) ? retval : NULL;
}
static SLJIT_INLINE void free_chunk(void* chunk, sljit_uw size)
{
munmap(chunk, size);
}
#endif
/* --------------------------------------------------------------------- */
/* Common functions */
/* --------------------------------------------------------------------- */
#define CHUNK_MASK (~(CHUNK_SIZE - 1))
struct block_header {
sljit_uw size;
sljit_uw prev_size;
};
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_ub*)base) + offset))
#define AS_FREE_BLOCK(base, offset) \
((struct free_block*)(((sljit_ub*)base) + offset))
#define MEM_START(base) ((void*)(((sljit_ub*)base) + sizeof(struct block_header)))
#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 = 0;
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 block_header *header;
struct block_header *next_header;
struct free_block *free_block;
sljit_uw chunk_size;
allocator_grab_lock();
if (size < sizeof(struct free_block))
size = sizeof(struct free_block);
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;
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 = (size + sizeof(struct block_header) + CHUNK_SIZE - 1) & CHUNK_MASK;
header = (struct block_header*)alloc_chunk(chunk_size);
if (!header) {
allocator_release_lock();
return NULL;
}
chunk_size -= sizeof(struct block_header);
total_size += chunk_size;
header->prev_size = 0;
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;
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;
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));
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();
}

1838
ext/pcre/pcrelib/sljit/sljitLir.c Normal file
View file

File diff suppressed because it is too large Load diff

1001
ext/pcre/pcrelib/sljit/sljitLir.h Normal file
View file

File diff suppressed because it is too large Load diff

2524
ext/pcre/pcrelib/sljit/sljitNativeARM_32.c Normal file
View file

File diff suppressed because it is too large Load diff

1902
ext/pcre/pcrelib/sljit/sljitNativeARM_64.c Normal file
View file

File diff suppressed because it is too large Load diff

2007
ext/pcre/pcrelib/sljit/sljitNativeARM_T2_32.c Normal file
View file

File diff suppressed because it is too large Load diff

366
ext/pcre/pcrelib/sljit/sljitNativeMIPS_32.c Normal file
View file

@ -0,0 +1,366 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* mips 32-bit arch dependent functions. */
static sljit_si load_immediate(struct sljit_compiler *compiler, sljit_si dst_ar, sljit_sw imm)
{
if (!(imm & ~0xffff))
return push_inst(compiler, ORI | SA(0) | TA(dst_ar) | IMM(imm), dst_ar);
if (imm < 0 && imm >= SIMM_MIN)
return push_inst(compiler, ADDIU | SA(0) | TA(dst_ar) | IMM(imm), dst_ar);
FAIL_IF(push_inst(compiler, LUI | TA(dst_ar) | IMM(imm >> 16), dst_ar));
return (imm & 0xffff) ? push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(imm), dst_ar) : SLJIT_SUCCESS;
}
#define EMIT_LOGICAL(op_imm, op_norm) \
if (flags & SRC2_IMM) { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_imm | S(src1) | TA(EQUAL_FLAG) | IMM(src2), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_imm | S(src1) | T(dst) | IMM(src2), DR(dst))); \
} \
else { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_norm | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_norm | S(src1) | T(src2) | D(dst), DR(dst))); \
}
#define EMIT_SHIFT(op_imm, op_v) \
if (flags & SRC2_IMM) { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_imm | T(src1) | DA(EQUAL_FLAG) | SH_IMM(src2), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_imm | T(src1) | D(dst) | SH_IMM(src2), DR(dst))); \
} \
else { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_v | S(src2) | T(src1) | DA(EQUAL_FLAG), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_v | S(src2) | T(src1) | D(dst), DR(dst))); \
}
static SLJIT_INLINE sljit_si emit_single_op(struct sljit_compiler *compiler, sljit_si op, sljit_si flags,
sljit_si dst, sljit_si src1, sljit_sw src2)
{
switch (GET_OPCODE(op)) {
case SLJIT_MOV:
case SLJIT_MOV_UI:
case SLJIT_MOV_SI:
case SLJIT_MOV_P:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if (dst != src2)
return push_inst(compiler, ADDU | S(src2) | TA(0) | D(dst), DR(dst));
return SLJIT_SUCCESS;
case SLJIT_MOV_UB:
case SLJIT_MOV_SB:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SB) {
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
return push_inst(compiler, SEB | T(src2) | D(dst), DR(dst));
#else
FAIL_IF(push_inst(compiler, SLL | T(src2) | D(dst) | SH_IMM(24), DR(dst)));
return push_inst(compiler, SRA | T(dst) | D(dst) | SH_IMM(24), DR(dst));
#endif
}
return push_inst(compiler, ANDI | S(src2) | T(dst) | IMM(0xff), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_MOV_UH:
case SLJIT_MOV_SH:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SH) {
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
return push_inst(compiler, SEH | T(src2) | D(dst), DR(dst));
#else
FAIL_IF(push_inst(compiler, SLL | T(src2) | D(dst) | SH_IMM(16), DR(dst)));
return push_inst(compiler, SRA | T(dst) | D(dst) | SH_IMM(16), DR(dst));
#endif
}
return push_inst(compiler, ANDI | S(src2) | T(dst) | IMM(0xffff), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_NOT:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, NOR | S(src2) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, NOR | S(src2) | T(src2) | D(dst), DR(dst)));
return SLJIT_SUCCESS;
case SLJIT_CLZ:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, CLZ | S(src2) | TA(EQUAL_FLAG) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, CLZ | S(src2) | T(dst) | D(dst), DR(dst)));
#else
if (SLJIT_UNLIKELY(flags & UNUSED_DEST)) {
FAIL_IF(push_inst(compiler, SRL | T(src2) | DA(EQUAL_FLAG) | SH_IMM(31), EQUAL_FLAG));
return push_inst(compiler, XORI | SA(EQUAL_FLAG) | TA(EQUAL_FLAG) | IMM(1), EQUAL_FLAG);
}
/* Nearly all instructions are unmovable in the following sequence. */
FAIL_IF(push_inst(compiler, ADDU | S(src2) | TA(0) | D(TMP_REG1), DR(TMP_REG1)));
/* Check zero. */
FAIL_IF(push_inst(compiler, BEQ | S(TMP_REG1) | TA(0) | IMM(5), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, ORI | SA(0) | T(dst) | IMM(32), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | T(dst) | IMM(-1), DR(dst)));
/* Loop for searching the highest bit. */
FAIL_IF(push_inst(compiler, ADDIU | S(dst) | T(dst) | IMM(1), DR(dst)));
FAIL_IF(push_inst(compiler, BGEZ | S(TMP_REG1) | IMM(-2), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, SLL | T(TMP_REG1) | D(TMP_REG1) | SH_IMM(1), UNMOVABLE_INS));
if (op & SLJIT_SET_E)
return push_inst(compiler, ADDU | S(dst) | TA(0) | DA(EQUAL_FLAG), EQUAL_FLAG);
#endif
return SLJIT_SUCCESS;
case SLJIT_ADD:
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_O) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
else
FAIL_IF(push_inst(compiler, NOR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
}
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | TA(EQUAL_FLAG) | IMM(src2), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O)) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, ORI | S(src1) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
else {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
FAIL_IF(push_inst(compiler, OR | S(src1) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
}
}
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | T(dst) | IMM(src2), DR(dst)));
}
else {
if (op & SLJIT_SET_O)
FAIL_IF(push_inst(compiler, XOR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, ADDU | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src2) | DA(ULESS_FLAG), ULESS_FLAG));
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, ADDU | S(src1) | T(src2) | D(dst), DR(dst)));
}
/* a + b >= a | b (otherwise, the carry should be set to 1). */
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
if (!(op & SLJIT_SET_O))
return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SLL | TA(ULESS_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | SA(TMP_EREG1) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | S(dst) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
return push_inst(compiler, SLL | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG);
case SLJIT_ADDC:
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_C) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, ORI | S(src1) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
else {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
FAIL_IF(push_inst(compiler, OR | S(src1) | TA(TMP_EREG1) | DA(TMP_EREG1), TMP_EREG1));
}
}
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | T(dst) | IMM(src2), DR(dst)));
} else {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, ADDU | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(TMP_EREG1) | DA(TMP_EREG1), TMP_EREG1));
FAIL_IF(push_inst(compiler, ADDU | S(dst) | TA(ULESS_FLAG) | D(dst), DR(dst)));
if (!(op & SLJIT_SET_C))
return SLJIT_SUCCESS;
/* Set ULESS_FLAG (dst == 0) && (ULESS_FLAG == 1). */
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
/* Set carry flag. */
return push_inst(compiler, OR | SA(ULESS_FLAG) | TA(TMP_EREG1) | DA(ULESS_FLAG), ULESS_FLAG);
case SLJIT_SUB:
if ((flags & SRC2_IMM) && ((op & (SLJIT_SET_U | SLJIT_SET_S)) || src2 == SIMM_MIN)) {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | T(TMP_REG2) | IMM(src2), DR(TMP_REG2)));
src2 = TMP_REG2;
flags &= ~SRC2_IMM;
}
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_O) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
else
FAIL_IF(push_inst(compiler, NOR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
}
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | TA(EQUAL_FLAG) | IMM(-src2), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTIU | S(src1) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | T(dst) | IMM(-src2), DR(dst)));
}
else {
if (op & SLJIT_SET_O)
FAIL_IF(push_inst(compiler, XOR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SUBU | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (op & (SLJIT_SET_U | SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTU | S(src1) | T(src2) | DA(ULESS_FLAG), ULESS_FLAG));
if (op & SLJIT_SET_U)
FAIL_IF(push_inst(compiler, SLTU | S(src2) | T(src1) | DA(UGREATER_FLAG), UGREATER_FLAG));
if (op & SLJIT_SET_S) {
FAIL_IF(push_inst(compiler, SLT | S(src1) | T(src2) | DA(LESS_FLAG), LESS_FLAG));
FAIL_IF(push_inst(compiler, SLT | S(src2) | T(src1) | DA(GREATER_FLAG), GREATER_FLAG));
}
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E | SLJIT_SET_U | SLJIT_SET_S | SLJIT_SET_C))
FAIL_IF(push_inst(compiler, SUBU | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (!(op & SLJIT_SET_O))
return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SLL | TA(ULESS_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | SA(TMP_EREG1) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | S(dst) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
return push_inst(compiler, SRL | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG);
case SLJIT_SUBC:
if ((flags & SRC2_IMM) && src2 == SIMM_MIN) {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | T(TMP_REG2) | IMM(src2), DR(TMP_REG2)));
src2 = TMP_REG2;
flags &= ~SRC2_IMM;
}
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTIU | S(src1) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, ADDIU | S(src1) | T(dst) | IMM(-src2), DR(dst)));
}
else {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, SUBU | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(TMP_EREG2), TMP_EREG2));
FAIL_IF(push_inst(compiler, SUBU | S(dst) | TA(ULESS_FLAG) | D(dst), DR(dst)));
return (op & SLJIT_SET_C) ? push_inst(compiler, OR | SA(TMP_EREG1) | TA(TMP_EREG2) | DA(ULESS_FLAG), ULESS_FLAG) : SLJIT_SUCCESS;
case SLJIT_MUL:
SLJIT_ASSERT(!(flags & SRC2_IMM));
if (!(op & SLJIT_SET_O)) {
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
return push_inst(compiler, MUL | S(src1) | T(src2) | D(dst), DR(dst));
#else
FAIL_IF(push_inst(compiler, MULT | S(src1) | T(src2), MOVABLE_INS));
return push_inst(compiler, MFLO | D(dst), DR(dst));
#endif
}
FAIL_IF(push_inst(compiler, MULT | S(src1) | T(src2), MOVABLE_INS));
FAIL_IF(push_inst(compiler, MFHI | DA(TMP_EREG1), TMP_EREG1));
FAIL_IF(push_inst(compiler, MFLO | D(dst), DR(dst)));
FAIL_IF(push_inst(compiler, SRA | T(dst) | DA(TMP_EREG2) | SH_IMM(31), TMP_EREG2));
return push_inst(compiler, SUBU | SA(TMP_EREG1) | TA(TMP_EREG2) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG);
case SLJIT_AND:
EMIT_LOGICAL(ANDI, AND);
return SLJIT_SUCCESS;
case SLJIT_OR:
EMIT_LOGICAL(ORI, OR);
return SLJIT_SUCCESS;
case SLJIT_XOR:
EMIT_LOGICAL(XORI, XOR);
return SLJIT_SUCCESS;
case SLJIT_SHL:
EMIT_SHIFT(SLL, SLLV);
return SLJIT_SUCCESS;
case SLJIT_LSHR:
EMIT_SHIFT(SRL, SRLV);
return SLJIT_SUCCESS;
case SLJIT_ASHR:
EMIT_SHIFT(SRA, SRAV);
return SLJIT_SUCCESS;
}
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
}
static SLJIT_INLINE sljit_si emit_const(struct sljit_compiler *compiler, sljit_si dst, sljit_sw init_value)
{
FAIL_IF(push_inst(compiler, LUI | T(dst) | IMM(init_value >> 16), DR(dst)));
return push_inst(compiler, ORI | S(dst) | T(dst) | IMM(init_value), DR(dst));
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_jump_addr(sljit_uw addr, sljit_uw new_addr)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_addr >> 16) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | (new_addr & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_const(sljit_uw addr, sljit_sw new_constant)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_constant >> 16) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | (new_constant & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}

469
ext/pcre/pcrelib/sljit/sljitNativeMIPS_64.c Normal file
View file

@ -0,0 +1,469 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* mips 64-bit arch dependent functions. */
static sljit_si load_immediate(struct sljit_compiler *compiler, sljit_si dst_ar, sljit_sw imm)
{
sljit_si shift = 32;
sljit_si shift2;
sljit_si inv = 0;
sljit_ins ins;
sljit_uw uimm;
if (!(imm & ~0xffff))
return push_inst(compiler, ORI | SA(0) | TA(dst_ar) | IMM(imm), dst_ar);
if (imm < 0 && imm >= SIMM_MIN)
return push_inst(compiler, ADDIU | SA(0) | TA(dst_ar) | IMM(imm), dst_ar);
if (imm <= 0x7fffffffl && imm >= -0x80000000l) {
FAIL_IF(push_inst(compiler, LUI | TA(dst_ar) | IMM(imm >> 16), dst_ar));
return (imm & 0xffff) ? push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(imm), dst_ar) : SLJIT_SUCCESS;
}
/* Zero extended number. */
uimm = imm;
if (imm < 0) {
uimm = ~imm;
inv = 1;
}
while (!(uimm & 0xff00000000000000l)) {
shift -= 8;
uimm <<= 8;
}
if (!(uimm & 0xf000000000000000l)) {
shift -= 4;
uimm <<= 4;
}
if (!(uimm & 0xc000000000000000l)) {
shift -= 2;
uimm <<= 2;
}
if ((sljit_sw)uimm < 0) {
uimm >>= 1;
shift += 1;
}
SLJIT_ASSERT(((uimm & 0xc000000000000000l) == 0x4000000000000000l) && (shift > 0) && (shift <= 32));
if (inv)
uimm = ~uimm;
FAIL_IF(push_inst(compiler, LUI | TA(dst_ar) | IMM(uimm >> 48), dst_ar));
if (uimm & 0x0000ffff00000000l)
FAIL_IF(push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(uimm >> 32), dst_ar));
imm &= (1l << shift) - 1;
if (!(imm & ~0xffff)) {
ins = (shift == 32) ? DSLL32 : DSLL;
if (shift < 32)
ins |= SH_IMM(shift);
FAIL_IF(push_inst(compiler, ins | TA(dst_ar) | DA(dst_ar), dst_ar));
return !(imm & 0xffff) ? SLJIT_SUCCESS : push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(imm), dst_ar);
}
/* Double shifts needs to be performed. */
uimm <<= 32;
shift2 = shift - 16;
while (!(uimm & 0xf000000000000000l)) {
shift2 -= 4;
uimm <<= 4;
}
if (!(uimm & 0xc000000000000000l)) {
shift2 -= 2;
uimm <<= 2;
}
if (!(uimm & 0x8000000000000000l)) {
shift2--;
uimm <<= 1;
}
SLJIT_ASSERT((uimm & 0x8000000000000000l) && (shift2 > 0) && (shift2 <= 16));
FAIL_IF(push_inst(compiler, DSLL | TA(dst_ar) | DA(dst_ar) | SH_IMM(shift - shift2), dst_ar));
FAIL_IF(push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(uimm >> 48), dst_ar));
FAIL_IF(push_inst(compiler, DSLL | TA(dst_ar) | DA(dst_ar) | SH_IMM(shift2), dst_ar));
imm &= (1l << shift2) - 1;
return !(imm & 0xffff) ? SLJIT_SUCCESS : push_inst(compiler, ORI | SA(dst_ar) | TA(dst_ar) | IMM(imm), dst_ar);
}
#define SELECT_OP(a, b) \
(!(op & SLJIT_INT_OP) ? a : b)
#define EMIT_LOGICAL(op_imm, op_norm) \
if (flags & SRC2_IMM) { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_imm | S(src1) | TA(EQUAL_FLAG) | IMM(src2), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_imm | S(src1) | T(dst) | IMM(src2), DR(dst))); \
} \
else { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, op_norm | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, op_norm | S(src1) | T(src2) | D(dst), DR(dst))); \
}
#define EMIT_SHIFT(op_dimm, op_dimm32, op_imm, op_dv, op_v) \
if (flags & SRC2_IMM) { \
if (src2 >= 32) { \
SLJIT_ASSERT(!(op & SLJIT_INT_OP)); \
ins = op_dimm32; \
src2 -= 32; \
} \
else \
ins = (op & SLJIT_INT_OP) ? op_imm : op_dimm; \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, ins | T(src1) | DA(EQUAL_FLAG) | SH_IMM(src2), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, ins | T(src1) | D(dst) | SH_IMM(src2), DR(dst))); \
} \
else { \
ins = (op & SLJIT_INT_OP) ? op_v : op_dv; \
if (op & SLJIT_SET_E) \
FAIL_IF(push_inst(compiler, ins | S(src2) | T(src1) | DA(EQUAL_FLAG), EQUAL_FLAG)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_inst(compiler, ins | S(src2) | T(src1) | D(dst), DR(dst))); \
}
static SLJIT_INLINE sljit_si emit_single_op(struct sljit_compiler *compiler, sljit_si op, sljit_si flags,
sljit_si dst, sljit_si src1, sljit_sw src2)
{
sljit_ins ins;
switch (GET_OPCODE(op)) {
case SLJIT_MOV:
case SLJIT_MOV_P:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if (dst != src2)
return push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(src2) | TA(0) | D(dst), DR(dst));
return SLJIT_SUCCESS;
case SLJIT_MOV_UB:
case SLJIT_MOV_SB:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SB) {
FAIL_IF(push_inst(compiler, DSLL32 | T(src2) | D(dst) | SH_IMM(24), DR(dst)));
return push_inst(compiler, DSRA32 | T(dst) | D(dst) | SH_IMM(24), DR(dst));
}
return push_inst(compiler, ANDI | S(src2) | T(dst) | IMM(0xff), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_MOV_UH:
case SLJIT_MOV_SH:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SH) {
FAIL_IF(push_inst(compiler, DSLL32 | T(src2) | D(dst) | SH_IMM(16), DR(dst)));
return push_inst(compiler, DSRA32 | T(dst) | D(dst) | SH_IMM(16), DR(dst));
}
return push_inst(compiler, ANDI | S(src2) | T(dst) | IMM(0xffff), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_MOV_UI:
SLJIT_ASSERT(!(op & SLJIT_INT_OP));
FAIL_IF(push_inst(compiler, DSLL32 | T(src2) | D(dst) | SH_IMM(0), DR(dst)));
return push_inst(compiler, DSRL32 | T(dst) | D(dst) | SH_IMM(0), DR(dst));
case SLJIT_MOV_SI:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
return push_inst(compiler, SLL | T(src2) | D(dst) | SH_IMM(0), DR(dst));
case SLJIT_NOT:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, NOR | S(src2) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, NOR | S(src2) | T(src2) | D(dst), DR(dst)));
return SLJIT_SUCCESS;
case SLJIT_CLZ:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SELECT_OP(DCLZ, CLZ) | S(src2) | TA(EQUAL_FLAG) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, SELECT_OP(DCLZ, CLZ) | S(src2) | T(dst) | D(dst), DR(dst)));
#else
if (SLJIT_UNLIKELY(flags & UNUSED_DEST)) {
FAIL_IF(push_inst(compiler, SELECT_OP(DSRL32, SRL) | T(src2) | DA(EQUAL_FLAG) | SH_IMM(31), EQUAL_FLAG));
return push_inst(compiler, XORI | SA(EQUAL_FLAG) | TA(EQUAL_FLAG) | IMM(1), EQUAL_FLAG);
}
/* Nearly all instructions are unmovable in the following sequence. */
FAIL_IF(push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(src2) | TA(0) | D(TMP_REG1), DR(TMP_REG1)));
/* Check zero. */
FAIL_IF(push_inst(compiler, BEQ | S(TMP_REG1) | TA(0) | IMM(5), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, ORI | SA(0) | T(dst) | IMM((op & SLJIT_INT_OP) ? 32 : 64), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | SA(0) | T(dst) | IMM(-1), DR(dst)));
/* Loop for searching the highest bit. */
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(dst) | T(dst) | IMM(1), DR(dst)));
FAIL_IF(push_inst(compiler, BGEZ | S(TMP_REG1) | IMM(-2), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, SELECT_OP(DSLL, SLL) | T(TMP_REG1) | D(TMP_REG1) | SH_IMM(1), UNMOVABLE_INS));
if (op & SLJIT_SET_E)
return push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(dst) | TA(0) | DA(EQUAL_FLAG), EQUAL_FLAG);
#endif
return SLJIT_SUCCESS;
case SLJIT_ADD:
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_O) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
else
FAIL_IF(push_inst(compiler, NOR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
}
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | TA(EQUAL_FLAG) | IMM(src2), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O)) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, ORI | S(src1) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
else {
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | SA(0) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
FAIL_IF(push_inst(compiler, OR | S(src1) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
}
}
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | T(dst) | IMM(src2), DR(dst)));
}
else {
if (op & SLJIT_SET_O)
FAIL_IF(push_inst(compiler, XOR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src2) | DA(ULESS_FLAG), ULESS_FLAG));
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(src1) | T(src2) | D(dst), DR(dst)));
}
/* a + b >= a | b (otherwise, the carry should be set to 1). */
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
if (!(op & SLJIT_SET_O))
return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SELECT_OP(DSLL32, SLL) | TA(ULESS_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | SA(TMP_EREG1) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | S(dst) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
return push_inst(compiler, SELECT_OP(DSRL32, SLL) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG);
case SLJIT_ADDC:
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_C) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, ORI | S(src1) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
else {
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | SA(0) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
FAIL_IF(push_inst(compiler, OR | S(src1) | TA(TMP_EREG1) | DA(TMP_EREG1), TMP_EREG1));
}
}
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | T(dst) | IMM(src2), DR(dst)));
} else {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(TMP_EREG1) | DA(TMP_EREG1), TMP_EREG1));
FAIL_IF(push_inst(compiler, SELECT_OP(DADDU, ADDU) | S(dst) | TA(ULESS_FLAG) | D(dst), DR(dst)));
if (!(op & SLJIT_SET_C))
return SLJIT_SUCCESS;
/* Set ULESS_FLAG (dst == 0) && (ULESS_FLAG == 1). */
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(ULESS_FLAG), ULESS_FLAG));
/* Set carry flag. */
return push_inst(compiler, OR | SA(ULESS_FLAG) | TA(TMP_EREG1) | DA(ULESS_FLAG), ULESS_FLAG);
case SLJIT_SUB:
if ((flags & SRC2_IMM) && ((op & (SLJIT_SET_U | SLJIT_SET_S)) || src2 == SIMM_MIN)) {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | T(TMP_REG2) | IMM(src2), DR(TMP_REG2)));
src2 = TMP_REG2;
flags &= ~SRC2_IMM;
}
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_O) {
if (src2 >= 0)
FAIL_IF(push_inst(compiler, OR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
else
FAIL_IF(push_inst(compiler, NOR | S(src1) | T(src1) | DA(TMP_EREG1), TMP_EREG1));
}
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | TA(EQUAL_FLAG) | IMM(-src2), EQUAL_FLAG));
if (op & (SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTIU | S(src1) | TA(ULESS_FLAG) | IMM(src2), ULESS_FLAG));
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E))
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | T(dst) | IMM(-src2), DR(dst)));
}
else {
if (op & SLJIT_SET_O)
FAIL_IF(push_inst(compiler, XOR | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
if (op & SLJIT_SET_E)
FAIL_IF(push_inst(compiler, SELECT_OP(DSUBU, SUBU) | S(src1) | T(src2) | DA(EQUAL_FLAG), EQUAL_FLAG));
if (op & (SLJIT_SET_U | SLJIT_SET_C | SLJIT_SET_O))
FAIL_IF(push_inst(compiler, SLTU | S(src1) | T(src2) | DA(ULESS_FLAG), ULESS_FLAG));
if (op & SLJIT_SET_U)
FAIL_IF(push_inst(compiler, SLTU | S(src2) | T(src1) | DA(UGREATER_FLAG), UGREATER_FLAG));
if (op & SLJIT_SET_S) {
FAIL_IF(push_inst(compiler, SLT | S(src1) | T(src2) | DA(LESS_FLAG), LESS_FLAG));
FAIL_IF(push_inst(compiler, SLT | S(src2) | T(src1) | DA(GREATER_FLAG), GREATER_FLAG));
}
/* dst may be the same as src1 or src2. */
if (CHECK_FLAGS(SLJIT_SET_E | SLJIT_SET_U | SLJIT_SET_S | SLJIT_SET_C))
FAIL_IF(push_inst(compiler, SELECT_OP(DSUBU, SUBU) | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (!(op & SLJIT_SET_O))
return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SELECT_OP(DSLL32, SLL) | TA(ULESS_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | SA(TMP_EREG1) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
FAIL_IF(push_inst(compiler, XOR | S(dst) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG));
return push_inst(compiler, SELECT_OP(DSRL32, SRL) | TA(OVERFLOW_FLAG) | DA(OVERFLOW_FLAG) | SH_IMM(31), OVERFLOW_FLAG);
case SLJIT_SUBC:
if ((flags & SRC2_IMM) && src2 == SIMM_MIN) {
FAIL_IF(push_inst(compiler, ADDIU | SA(0) | T(TMP_REG2) | IMM(src2), DR(TMP_REG2)));
src2 = TMP_REG2;
flags &= ~SRC2_IMM;
}
if (flags & SRC2_IMM) {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTIU | S(src1) | TA(TMP_EREG1) | IMM(src2), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, SELECT_OP(DADDIU, ADDIU) | S(src1) | T(dst) | IMM(-src2), DR(dst)));
}
else {
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(src1) | T(src2) | DA(TMP_EREG1), TMP_EREG1));
/* dst may be the same as src1 or src2. */
FAIL_IF(push_inst(compiler, SELECT_OP(DSUBU, SUBU) | S(src1) | T(src2) | D(dst), DR(dst)));
}
if (op & SLJIT_SET_C)
FAIL_IF(push_inst(compiler, SLTU | S(dst) | TA(ULESS_FLAG) | DA(TMP_EREG2), TMP_EREG2));
FAIL_IF(push_inst(compiler, SELECT_OP(DSUBU, SUBU) | S(dst) | TA(ULESS_FLAG) | D(dst), DR(dst)));
return (op & SLJIT_SET_C) ? push_inst(compiler, OR | SA(TMP_EREG1) | TA(TMP_EREG2) | DA(ULESS_FLAG), ULESS_FLAG) : SLJIT_SUCCESS;
case SLJIT_MUL:
SLJIT_ASSERT(!(flags & SRC2_IMM));
if (!(op & SLJIT_SET_O)) {
#if (defined SLJIT_MIPS_32_64 && SLJIT_MIPS_32_64)
if (op & SLJIT_INT_OP)
return push_inst(compiler, MUL | S(src1) | T(src2) | D(dst), DR(dst));
FAIL_IF(push_inst(compiler, DMULT | S(src1) | T(src2), MOVABLE_INS));
return push_inst(compiler, MFLO | D(dst), DR(dst));
#else
FAIL_IF(push_inst(compiler, SELECT_OP(DMULT, MULT) | S(src1) | T(src2), MOVABLE_INS));
return push_inst(compiler, MFLO | D(dst), DR(dst));
#endif
}
FAIL_IF(push_inst(compiler, SELECT_OP(DMULT, MULT) | S(src1) | T(src2), MOVABLE_INS));
FAIL_IF(push_inst(compiler, MFHI | DA(TMP_EREG1), TMP_EREG1));
FAIL_IF(push_inst(compiler, MFLO | D(dst), DR(dst)));
FAIL_IF(push_inst(compiler, SELECT_OP(DSRA32, SRA) | T(dst) | DA(TMP_EREG2) | SH_IMM(31), TMP_EREG2));
return push_inst(compiler, SELECT_OP(DSUBU, SUBU) | SA(TMP_EREG1) | TA(TMP_EREG2) | DA(OVERFLOW_FLAG), OVERFLOW_FLAG);
case SLJIT_AND:
EMIT_LOGICAL(ANDI, AND);
return SLJIT_SUCCESS;
case SLJIT_OR:
EMIT_LOGICAL(ORI, OR);
return SLJIT_SUCCESS;
case SLJIT_XOR:
EMIT_LOGICAL(XORI, XOR);
return SLJIT_SUCCESS;
case SLJIT_SHL:
EMIT_SHIFT(DSLL, DSLL32, SLL, DSLLV, SLLV);
return SLJIT_SUCCESS;
case SLJIT_LSHR:
EMIT_SHIFT(DSRL, DSRL32, SRL, DSRLV, SRLV);
return SLJIT_SUCCESS;
case SLJIT_ASHR:
EMIT_SHIFT(DSRA, DSRA32, SRA, DSRAV, SRAV);
return SLJIT_SUCCESS;
}
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
}
static SLJIT_INLINE sljit_si emit_const(struct sljit_compiler *compiler, sljit_si dst, sljit_sw init_value)
{
FAIL_IF(push_inst(compiler, LUI | T(dst) | IMM(init_value >> 48), DR(dst)));
FAIL_IF(push_inst(compiler, ORI | S(dst) | T(dst) | IMM(init_value >> 32), DR(dst)));
FAIL_IF(push_inst(compiler, DSLL | T(dst) | D(dst) | SH_IMM(16), DR(dst)));
FAIL_IF(push_inst(compiler, ORI | S(dst) | T(dst) | IMM(init_value >> 16), DR(dst)));
FAIL_IF(push_inst(compiler, DSLL | T(dst) | D(dst) | SH_IMM(16), DR(dst)));
return push_inst(compiler, ORI | S(dst) | T(dst) | IMM(init_value), DR(dst));
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_jump_addr(sljit_uw addr, sljit_uw new_addr)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_addr >> 48) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | ((new_addr >> 32) & 0xffff);
inst[3] = (inst[3] & 0xffff0000) | ((new_addr >> 16) & 0xffff);
inst[5] = (inst[5] & 0xffff0000) | (new_addr & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 6);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_const(sljit_uw addr, sljit_sw new_constant)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_constant >> 48) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | ((new_constant >> 32) & 0xffff);
inst[3] = (inst[3] & 0xffff0000) | ((new_constant >> 16) & 0xffff);
inst[5] = (inst[5] & 0xffff0000) | (new_constant & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 6);
}

2042
ext/pcre/pcrelib/sljit/sljitNativeMIPS_common.c Normal file
View file

File diff suppressed because it is too large Load diff

269
ext/pcre/pcrelib/sljit/sljitNativePPC_32.c Normal file
View file

@ -0,0 +1,269 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* ppc 32-bit arch dependent functions. */
static sljit_si load_immediate(struct sljit_compiler *compiler, sljit_si reg, sljit_sw imm)
{
if (imm <= SIMM_MAX && imm >= SIMM_MIN)
return push_inst(compiler, ADDI | D(reg) | A(0) | IMM(imm));
if (!(imm & ~0xffff))
return push_inst(compiler, ORI | S(TMP_ZERO) | A(reg) | IMM(imm));
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(imm >> 16)));
return (imm & 0xffff) ? push_inst(compiler, ORI | S(reg) | A(reg) | IMM(imm)) : SLJIT_SUCCESS;
}
#define INS_CLEAR_LEFT(dst, src, from) \
(RLWINM | S(src) | A(dst) | ((from) << 6) | (31 << 1))
static SLJIT_INLINE sljit_si emit_single_op(struct sljit_compiler *compiler, sljit_si op, sljit_si flags,
sljit_si dst, sljit_si src1, sljit_si src2)
{
switch (op) {
case SLJIT_MOV:
case SLJIT_MOV_UI:
case SLJIT_MOV_SI:
case SLJIT_MOV_P:
SLJIT_ASSERT(src1 == TMP_REG1);
if (dst != src2)
return push_inst(compiler, OR | S(src2) | A(dst) | B(src2));
return SLJIT_SUCCESS;
case SLJIT_MOV_UB:
case SLJIT_MOV_SB:
SLJIT_ASSERT(src1 == TMP_REG1);
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SB)
return push_inst(compiler, EXTSB | S(src2) | A(dst));
return push_inst(compiler, INS_CLEAR_LEFT(dst, src2, 24));
}
else if ((flags & REG_DEST) && op == SLJIT_MOV_SB)
return push_inst(compiler, EXTSB | S(src2) | A(dst));
else {
SLJIT_ASSERT(dst == src2);
}
return SLJIT_SUCCESS;
case SLJIT_MOV_UH:
case SLJIT_MOV_SH:
SLJIT_ASSERT(src1 == TMP_REG1);
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SH)
return push_inst(compiler, EXTSH | S(src2) | A(dst));
return push_inst(compiler, INS_CLEAR_LEFT(dst, src2, 16));
}
else {
SLJIT_ASSERT(dst == src2);
}
return SLJIT_SUCCESS;
case SLJIT_NOT:
SLJIT_ASSERT(src1 == TMP_REG1);
return push_inst(compiler, NOR | RC(flags) | S(src2) | A(dst) | B(src2));
case SLJIT_NEG:
SLJIT_ASSERT(src1 == TMP_REG1);
return push_inst(compiler, NEG | OERC(flags) | D(dst) | A(src2));
case SLJIT_CLZ:
SLJIT_ASSERT(src1 == TMP_REG1);
return push_inst(compiler, CNTLZW | RC(flags) | S(src2) | A(dst));
case SLJIT_ADD:
if (flags & ALT_FORM1) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ADDI | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM2) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ADDIS | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ADDIC | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM4) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
FAIL_IF(push_inst(compiler, ADDI | D(dst) | A(src1) | (compiler->imm & 0xffff)));
return push_inst(compiler, ADDIS | D(dst) | A(dst) | (((compiler->imm >> 16) & 0xffff) + ((compiler->imm >> 15) & 0x1)));
}
if (!(flags & ALT_SET_FLAGS))
return push_inst(compiler, ADD | D(dst) | A(src1) | B(src2));
return push_inst(compiler, ADDC | OERC(ALT_SET_FLAGS) | D(dst) | A(src1) | B(src2));
case SLJIT_ADDC:
if (flags & ALT_FORM1) {
FAIL_IF(push_inst(compiler, MFXER | D(0)));
FAIL_IF(push_inst(compiler, ADDE | D(dst) | A(src1) | B(src2)));
return push_inst(compiler, MTXER | S(0));
}
return push_inst(compiler, ADDE | D(dst) | A(src1) | B(src2));
case SLJIT_SUB:
if (flags & ALT_FORM1) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, SUBFIC | D(dst) | A(src1) | compiler->imm);
}
if (flags & (ALT_FORM2 | ALT_FORM3)) {
SLJIT_ASSERT(src2 == TMP_REG2);
if (flags & ALT_FORM2)
FAIL_IF(push_inst(compiler, CMPI | CRD(0) | A(src1) | compiler->imm));
if (flags & ALT_FORM3)
return push_inst(compiler, CMPLI | CRD(4) | A(src1) | compiler->imm);
return SLJIT_SUCCESS;
}
if (flags & (ALT_FORM4 | ALT_FORM5)) {
if (flags & ALT_FORM4)
FAIL_IF(push_inst(compiler, CMPL | CRD(4) | A(src1) | B(src2)));
if (flags & ALT_FORM5)
FAIL_IF(push_inst(compiler, CMP | CRD(0) | A(src1) | B(src2)));
return SLJIT_SUCCESS;
}
if (!(flags & ALT_SET_FLAGS))
return push_inst(compiler, SUBF | D(dst) | A(src2) | B(src1));
if (flags & ALT_FORM6)
FAIL_IF(push_inst(compiler, CMPL | CRD(4) | A(src1) | B(src2)));
return push_inst(compiler, SUBFC | OERC(ALT_SET_FLAGS) | D(dst) | A(src2) | B(src1));
case SLJIT_SUBC:
if (flags & ALT_FORM1) {
FAIL_IF(push_inst(compiler, MFXER | D(0)));
FAIL_IF(push_inst(compiler, SUBFE | D(dst) | A(src2) | B(src1)));
return push_inst(compiler, MTXER | S(0));
}
return push_inst(compiler, SUBFE | D(dst) | A(src2) | B(src1));
case SLJIT_MUL:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, MULLI | D(dst) | A(src1) | compiler->imm);
}
return push_inst(compiler, MULLW | OERC(flags) | D(dst) | A(src2) | B(src1));
case SLJIT_AND:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ANDI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ANDIS | S(src1) | A(dst) | compiler->imm);
}
return push_inst(compiler, AND | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_OR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ORI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ORIS | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
FAIL_IF(push_inst(compiler, ORI | S(src1) | A(dst) | IMM(compiler->imm)));
return push_inst(compiler, ORIS | S(dst) | A(dst) | IMM(compiler->imm >> 16));
}
return push_inst(compiler, OR | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_XOR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, XORI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, XORIS | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
FAIL_IF(push_inst(compiler, XORI | S(src1) | A(dst) | IMM(compiler->imm)));
return push_inst(compiler, XORIS | S(dst) | A(dst) | IMM(compiler->imm >> 16));
}
return push_inst(compiler, XOR | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_SHL:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
compiler->imm &= 0x1f;
return push_inst(compiler, RLWINM | RC(flags) | S(src1) | A(dst) | (compiler->imm << 11) | ((31 - compiler->imm) << 1));
}
return push_inst(compiler, SLW | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_LSHR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
compiler->imm &= 0x1f;
return push_inst(compiler, RLWINM | RC(flags) | S(src1) | A(dst) | (((32 - compiler->imm) & 0x1f) << 11) | (compiler->imm << 6) | (31 << 1));
}
return push_inst(compiler, SRW | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_ASHR:
if (flags & ALT_FORM3)
FAIL_IF(push_inst(compiler, MFXER | D(0)));
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
compiler->imm &= 0x1f;
FAIL_IF(push_inst(compiler, SRAWI | RC(flags) | S(src1) | A(dst) | (compiler->imm << 11)));
}
else
FAIL_IF(push_inst(compiler, SRAW | RC(flags) | S(src1) | A(dst) | B(src2)));
return (flags & ALT_FORM3) ? push_inst(compiler, MTXER | S(0)) : SLJIT_SUCCESS;
}
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
}
static SLJIT_INLINE sljit_si emit_const(struct sljit_compiler *compiler, sljit_si reg, sljit_sw init_value)
{
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(init_value >> 16)));
return push_inst(compiler, ORI | S(reg) | A(reg) | IMM(init_value));
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_jump_addr(sljit_uw addr, sljit_uw new_addr)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_addr >> 16) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | (new_addr & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_const(sljit_uw addr, sljit_sw new_constant)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_constant >> 16) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | (new_constant & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}

421
ext/pcre/pcrelib/sljit/sljitNativePPC_64.c Normal file
View file

@ -0,0 +1,421 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* ppc 64-bit arch dependent functions. */
#if defined(__GNUC__) || (defined(__IBM_GCC_ASM) && __IBM_GCC_ASM)
#define ASM_SLJIT_CLZ(src, dst) \
__asm__ volatile ( "cntlzd %0, %1" : "=r"(dst) : "r"(src) )
#elif defined(__xlc__)
#error "Please enable GCC syntax for inline assembly statements"
#else
#error "Must implement count leading zeroes"
#endif
#define RLDI(dst, src, sh, mb, type) \
(HI(30) | S(src) | A(dst) | ((type) << 2) | (((sh) & 0x1f) << 11) | (((sh) & 0x20) >> 4) | (((mb) & 0x1f) << 6) | ((mb) & 0x20))
#define PUSH_RLDICR(reg, shift) \
push_inst(compiler, RLDI(reg, reg, 63 - shift, shift, 1))
static sljit_si load_immediate(struct sljit_compiler *compiler, sljit_si reg, sljit_sw imm)
{
sljit_uw tmp;
sljit_uw shift;
sljit_uw tmp2;
sljit_uw shift2;
if (imm <= SIMM_MAX && imm >= SIMM_MIN)
return push_inst(compiler, ADDI | D(reg) | A(0) | IMM(imm));
if (!(imm & ~0xffff))
return push_inst(compiler, ORI | S(TMP_ZERO) | A(reg) | IMM(imm));
if (imm <= 0x7fffffffl && imm >= -0x80000000l) {
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(imm >> 16)));
return (imm & 0xffff) ? push_inst(compiler, ORI | S(reg) | A(reg) | IMM(imm)) : SLJIT_SUCCESS;
}
/* Count leading zeroes. */
tmp = (imm >= 0) ? imm : ~imm;
ASM_SLJIT_CLZ(tmp, shift);
SLJIT_ASSERT(shift > 0);
shift--;
tmp = (imm << shift);
if ((tmp & ~0xffff000000000000ul) == 0) {
FAIL_IF(push_inst(compiler, ADDI | D(reg) | A(0) | IMM(tmp >> 48)));
shift += 15;
return PUSH_RLDICR(reg, shift);
}
if ((tmp & ~0xffffffff00000000ul) == 0) {
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(tmp >> 48)));
FAIL_IF(push_inst(compiler, ORI | S(reg) | A(reg) | IMM(tmp >> 32)));
shift += 31;
return PUSH_RLDICR(reg, shift);
}
/* Cut out the 16 bit from immediate. */
shift += 15;
tmp2 = imm & ((1ul << (63 - shift)) - 1);
if (tmp2 <= 0xffff) {
FAIL_IF(push_inst(compiler, ADDI | D(reg) | A(0) | IMM(tmp >> 48)));
FAIL_IF(PUSH_RLDICR(reg, shift));
return push_inst(compiler, ORI | S(reg) | A(reg) | tmp2);
}
if (tmp2 <= 0xffffffff) {
FAIL_IF(push_inst(compiler, ADDI | D(reg) | A(0) | IMM(tmp >> 48)));
FAIL_IF(PUSH_RLDICR(reg, shift));
FAIL_IF(push_inst(compiler, ORIS | S(reg) | A(reg) | (tmp2 >> 16)));
return (imm & 0xffff) ? push_inst(compiler, ORI | S(reg) | A(reg) | IMM(tmp2)) : SLJIT_SUCCESS;
}
ASM_SLJIT_CLZ(tmp2, shift2);
tmp2 <<= shift2;
if ((tmp2 & ~0xffff000000000000ul) == 0) {
FAIL_IF(push_inst(compiler, ADDI | D(reg) | A(0) | IMM(tmp >> 48)));
shift2 += 15;
shift += (63 - shift2);
FAIL_IF(PUSH_RLDICR(reg, shift));
FAIL_IF(push_inst(compiler, ORI | S(reg) | A(reg) | (tmp2 >> 48)));
return PUSH_RLDICR(reg, shift2);
}
/* The general version. */
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(imm >> 48)));
FAIL_IF(push_inst(compiler, ORI | S(reg) | A(reg) | IMM(imm >> 32)));
FAIL_IF(PUSH_RLDICR(reg, 31));
FAIL_IF(push_inst(compiler, ORIS | S(reg) | A(reg) | IMM(imm >> 16)));
return push_inst(compiler, ORI | S(reg) | A(reg) | IMM(imm));
}
/* Simplified mnemonics: clrldi. */
#define INS_CLEAR_LEFT(dst, src, from) \
(RLDICL | S(src) | A(dst) | ((from) << 6) | (1 << 5))
/* Sign extension for integer operations. */
#define UN_EXTS() \
if ((flags & (ALT_SIGN_EXT | REG2_SOURCE)) == (ALT_SIGN_EXT | REG2_SOURCE)) { \
FAIL_IF(push_inst(compiler, EXTSW | S(src2) | A(TMP_REG2))); \
src2 = TMP_REG2; \
}
#define BIN_EXTS() \
if (flags & ALT_SIGN_EXT) { \
if (flags & REG1_SOURCE) { \
FAIL_IF(push_inst(compiler, EXTSW | S(src1) | A(TMP_REG1))); \
src1 = TMP_REG1; \
} \
if (flags & REG2_SOURCE) { \
FAIL_IF(push_inst(compiler, EXTSW | S(src2) | A(TMP_REG2))); \
src2 = TMP_REG2; \
} \
}
#define BIN_IMM_EXTS() \
if ((flags & (ALT_SIGN_EXT | REG1_SOURCE)) == (ALT_SIGN_EXT | REG1_SOURCE)) { \
FAIL_IF(push_inst(compiler, EXTSW | S(src1) | A(TMP_REG1))); \
src1 = TMP_REG1; \
}
static SLJIT_INLINE sljit_si emit_single_op(struct sljit_compiler *compiler, sljit_si op, sljit_si flags,
sljit_si dst, sljit_si src1, sljit_si src2)
{
switch (op) {
case SLJIT_MOV:
case SLJIT_MOV_P:
SLJIT_ASSERT(src1 == TMP_REG1);
if (dst != src2)
return push_inst(compiler, OR | S(src2) | A(dst) | B(src2));
return SLJIT_SUCCESS;
case SLJIT_MOV_UI:
case SLJIT_MOV_SI:
SLJIT_ASSERT(src1 == TMP_REG1);
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SI)
return push_inst(compiler, EXTSW | S(src2) | A(dst));
return push_inst(compiler, INS_CLEAR_LEFT(dst, src2, 0));
}
else {
SLJIT_ASSERT(dst == src2);
}
return SLJIT_SUCCESS;
case SLJIT_MOV_UB:
case SLJIT_MOV_SB:
SLJIT_ASSERT(src1 == TMP_REG1);
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SB)
return push_inst(compiler, EXTSB | S(src2) | A(dst));
return push_inst(compiler, INS_CLEAR_LEFT(dst, src2, 24));
}
else if ((flags & REG_DEST) && op == SLJIT_MOV_SB)
return push_inst(compiler, EXTSB | S(src2) | A(dst));
else {
SLJIT_ASSERT(dst == src2);
}
return SLJIT_SUCCESS;
case SLJIT_MOV_UH:
case SLJIT_MOV_SH:
SLJIT_ASSERT(src1 == TMP_REG1);
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_SH)
return push_inst(compiler, EXTSH | S(src2) | A(dst));
return push_inst(compiler, INS_CLEAR_LEFT(dst, src2, 16));
}
else {
SLJIT_ASSERT(dst == src2);
}
return SLJIT_SUCCESS;
case SLJIT_NOT:
SLJIT_ASSERT(src1 == TMP_REG1);
UN_EXTS();
return push_inst(compiler, NOR | RC(flags) | S(src2) | A(dst) | B(src2));
case SLJIT_NEG:
SLJIT_ASSERT(src1 == TMP_REG1);
UN_EXTS();
return push_inst(compiler, NEG | OERC(flags) | D(dst) | A(src2));
case SLJIT_CLZ:
SLJIT_ASSERT(src1 == TMP_REG1);
if (flags & ALT_FORM1)
return push_inst(compiler, CNTLZW | RC(flags) | S(src2) | A(dst));
return push_inst(compiler, CNTLZD | RC(flags) | S(src2) | A(dst));
case SLJIT_ADD:
if (flags & ALT_FORM1) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ADDI | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM2) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ADDIS | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
BIN_IMM_EXTS();
return push_inst(compiler, ADDIC | D(dst) | A(src1) | compiler->imm);
}
if (flags & ALT_FORM4) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
FAIL_IF(push_inst(compiler, ADDI | D(dst) | A(src1) | (compiler->imm & 0xffff)));
return push_inst(compiler, ADDIS | D(dst) | A(dst) | (((compiler->imm >> 16) & 0xffff) + ((compiler->imm >> 15) & 0x1)));
}
if (!(flags & ALT_SET_FLAGS))
return push_inst(compiler, ADD | D(dst) | A(src1) | B(src2));
BIN_EXTS();
return push_inst(compiler, ADDC | OERC(ALT_SET_FLAGS) | D(dst) | A(src1) | B(src2));
case SLJIT_ADDC:
if (flags & ALT_FORM1) {
FAIL_IF(push_inst(compiler, MFXER | D(0)));
FAIL_IF(push_inst(compiler, ADDE | D(dst) | A(src1) | B(src2)));
return push_inst(compiler, MTXER | S(0));
}
BIN_EXTS();
return push_inst(compiler, ADDE | D(dst) | A(src1) | B(src2));
case SLJIT_SUB:
if (flags & ALT_FORM1) {
/* Flags does not set: BIN_IMM_EXTS unnecessary. */
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, SUBFIC | D(dst) | A(src1) | compiler->imm);
}
if (flags & (ALT_FORM2 | ALT_FORM3)) {
SLJIT_ASSERT(src2 == TMP_REG2);
if (flags & ALT_FORM2)
FAIL_IF(push_inst(compiler, CMPI | CRD(0 | ((flags & ALT_SIGN_EXT) ? 0 : 1)) | A(src1) | compiler->imm));
if (flags & ALT_FORM3)
return push_inst(compiler, CMPLI | CRD(4 | ((flags & ALT_SIGN_EXT) ? 0 : 1)) | A(src1) | compiler->imm);
return SLJIT_SUCCESS;
}
if (flags & (ALT_FORM4 | ALT_FORM5)) {
if (flags & ALT_FORM4)
FAIL_IF(push_inst(compiler, CMPL | CRD(4 | ((flags & ALT_SIGN_EXT) ? 0 : 1)) | A(src1) | B(src2)));
if (flags & ALT_FORM5)
return push_inst(compiler, CMP | CRD(0 | ((flags & ALT_SIGN_EXT) ? 0 : 1)) | A(src1) | B(src2));
return SLJIT_SUCCESS;
}
if (!(flags & ALT_SET_FLAGS))
return push_inst(compiler, SUBF | D(dst) | A(src2) | B(src1));
BIN_EXTS();
if (flags & ALT_FORM6)
FAIL_IF(push_inst(compiler, CMPL | CRD(4 | ((flags & ALT_SIGN_EXT) ? 0 : 1)) | A(src1) | B(src2)));
return push_inst(compiler, SUBFC | OERC(ALT_SET_FLAGS) | D(dst) | A(src2) | B(src1));
case SLJIT_SUBC:
if (flags & ALT_FORM1) {
FAIL_IF(push_inst(compiler, MFXER | D(0)));
FAIL_IF(push_inst(compiler, SUBFE | D(dst) | A(src2) | B(src1)));
return push_inst(compiler, MTXER | S(0));
}
BIN_EXTS();
return push_inst(compiler, SUBFE | D(dst) | A(src2) | B(src1));
case SLJIT_MUL:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, MULLI | D(dst) | A(src1) | compiler->imm);
}
BIN_EXTS();
if (flags & ALT_FORM2)
return push_inst(compiler, MULLW | OERC(flags) | D(dst) | A(src2) | B(src1));
return push_inst(compiler, MULLD | OERC(flags) | D(dst) | A(src2) | B(src1));
case SLJIT_AND:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ANDI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ANDIS | S(src1) | A(dst) | compiler->imm);
}
return push_inst(compiler, AND | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_OR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ORI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, ORIS | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
FAIL_IF(push_inst(compiler, ORI | S(src1) | A(dst) | IMM(compiler->imm)));
return push_inst(compiler, ORIS | S(dst) | A(dst) | IMM(compiler->imm >> 16));
}
return push_inst(compiler, OR | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_XOR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, XORI | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM2) {
SLJIT_ASSERT(src2 == TMP_REG2);
return push_inst(compiler, XORIS | S(src1) | A(dst) | compiler->imm);
}
if (flags & ALT_FORM3) {
SLJIT_ASSERT(src2 == TMP_REG2);
FAIL_IF(push_inst(compiler, XORI | S(src1) | A(dst) | IMM(compiler->imm)));
return push_inst(compiler, XORIS | S(dst) | A(dst) | IMM(compiler->imm >> 16));
}
return push_inst(compiler, XOR | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_SHL:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
if (flags & ALT_FORM2) {
compiler->imm &= 0x1f;
return push_inst(compiler, RLWINM | RC(flags) | S(src1) | A(dst) | (compiler->imm << 11) | ((31 - compiler->imm) << 1));
}
else {
compiler->imm &= 0x3f;
return push_inst(compiler, RLDI(dst, src1, compiler->imm, 63 - compiler->imm, 1) | RC(flags));
}
}
return push_inst(compiler, ((flags & ALT_FORM2) ? SLW : SLD) | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_LSHR:
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
if (flags & ALT_FORM2) {
compiler->imm &= 0x1f;
return push_inst(compiler, RLWINM | RC(flags) | S(src1) | A(dst) | (((32 - compiler->imm) & 0x1f) << 11) | (compiler->imm << 6) | (31 << 1));
}
else {
compiler->imm &= 0x3f;
return push_inst(compiler, RLDI(dst, src1, 64 - compiler->imm, compiler->imm, 0) | RC(flags));
}
}
return push_inst(compiler, ((flags & ALT_FORM2) ? SRW : SRD) | RC(flags) | S(src1) | A(dst) | B(src2));
case SLJIT_ASHR:
if (flags & ALT_FORM3)
FAIL_IF(push_inst(compiler, MFXER | D(0)));
if (flags & ALT_FORM1) {
SLJIT_ASSERT(src2 == TMP_REG2);
if (flags & ALT_FORM2) {
compiler->imm &= 0x1f;
FAIL_IF(push_inst(compiler, SRAWI | RC(flags) | S(src1) | A(dst) | (compiler->imm << 11)));
}
else {
compiler->imm &= 0x3f;
FAIL_IF(push_inst(compiler, SRADI | RC(flags) | S(src1) | A(dst) | ((compiler->imm & 0x1f) << 11) | ((compiler->imm & 0x20) >> 4)));
}
}
else
FAIL_IF(push_inst(compiler, ((flags & ALT_FORM2) ? SRAW : SRAD) | RC(flags) | S(src1) | A(dst) | B(src2)));
return (flags & ALT_FORM3) ? push_inst(compiler, MTXER | S(0)) : SLJIT_SUCCESS;
}
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
}
static SLJIT_INLINE sljit_si emit_const(struct sljit_compiler *compiler, sljit_si reg, sljit_sw init_value)
{
FAIL_IF(push_inst(compiler, ADDIS | D(reg) | A(0) | IMM(init_value >> 48)));
FAIL_IF(push_inst(compiler, ORI | S(reg) | A(reg) | IMM(init_value >> 32)));
FAIL_IF(PUSH_RLDICR(reg, 31));
FAIL_IF(push_inst(compiler, ORIS | S(reg) | A(reg) | IMM(init_value >> 16)));
return push_inst(compiler, ORI | S(reg) | A(reg) | IMM(init_value));
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_jump_addr(sljit_uw addr, sljit_uw new_addr)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_addr >> 48) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | ((new_addr >> 32) & 0xffff);
inst[3] = (inst[3] & 0xffff0000) | ((new_addr >> 16) & 0xffff);
inst[4] = (inst[4] & 0xffff0000) | (new_addr & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 5);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_const(sljit_uw addr, sljit_sw new_constant)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffff0000) | ((new_constant >> 48) & 0xffff);
inst[1] = (inst[1] & 0xffff0000) | ((new_constant >> 32) & 0xffff);
inst[3] = (inst[3] & 0xffff0000) | ((new_constant >> 16) & 0xffff);
inst[4] = (inst[4] & 0xffff0000) | (new_constant & 0xffff);
SLJIT_CACHE_FLUSH(inst, inst + 5);
}

2188
ext/pcre/pcrelib/sljit/sljitNativePPC_common.c Normal file
View file

File diff suppressed because it is too large Load diff

164
ext/pcre/pcrelib/sljit/sljitNativeSPARC_32.c Normal file
View file

@ -0,0 +1,164 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
static sljit_si load_immediate(struct sljit_compiler *compiler, sljit_si dst, sljit_sw imm)
{
if (imm <= SIMM_MAX && imm >= SIMM_MIN)
return push_inst(compiler, OR | D(dst) | S1(0) | IMM(imm), DR(dst));
FAIL_IF(push_inst(compiler, SETHI | D(dst) | ((imm >> 10) & 0x3fffff), DR(dst)));
return (imm & 0x3ff) ? push_inst(compiler, OR | D(dst) | S1(dst) | IMM_ARG | (imm & 0x3ff), DR(dst)) : SLJIT_SUCCESS;
}
#define ARG2(flags, src2) ((flags & SRC2_IMM) ? IMM(src2) : S2(src2))
static SLJIT_INLINE sljit_si emit_single_op(struct sljit_compiler *compiler, sljit_si op, sljit_si flags,
sljit_si dst, sljit_si src1, sljit_sw src2)
{
SLJIT_COMPILE_ASSERT(ICC_IS_SET == SET_FLAGS, icc_is_set_and_set_flags_must_be_the_same);
switch (op) {
case SLJIT_MOV:
case SLJIT_MOV_UI:
case SLJIT_MOV_SI:
case SLJIT_MOV_P:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if (dst != src2)
return push_inst(compiler, OR | D(dst) | S1(0) | S2(src2), DR(dst));
return SLJIT_SUCCESS;
case SLJIT_MOV_UB:
case SLJIT_MOV_SB:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
if (op == SLJIT_MOV_UB)
return push_inst(compiler, AND | D(dst) | S1(src2) | IMM(0xff), DR(dst));
FAIL_IF(push_inst(compiler, SLL | D(dst) | S1(src2) | IMM(24), DR(dst)));
return push_inst(compiler, SRA | D(dst) | S1(dst) | IMM(24), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_MOV_UH:
case SLJIT_MOV_SH:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
if ((flags & (REG_DEST | REG2_SOURCE)) == (REG_DEST | REG2_SOURCE)) {
FAIL_IF(push_inst(compiler, SLL | D(dst) | S1(src2) | IMM(16), DR(dst)));
return push_inst(compiler, (op == SLJIT_MOV_SH ? SRA : SRL) | D(dst) | S1(dst) | IMM(16), DR(dst));
}
else if (dst != src2)
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
case SLJIT_NOT:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
return push_inst(compiler, XNOR | (flags & SET_FLAGS) | D(dst) | S1(0) | S2(src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_CLZ:
SLJIT_ASSERT(src1 == TMP_REG1 && !(flags & SRC2_IMM));
/* sparc 32 does not support SLJIT_KEEP_FLAGS. Not sure I can fix this. */
FAIL_IF(push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(src2) | S2(0), SET_FLAGS));
FAIL_IF(push_inst(compiler, OR | D(TMP_REG1) | S1(0) | S2(src2), DR(TMP_REG1)));
FAIL_IF(push_inst(compiler, BICC | DA(0x1) | (7 & DISP_MASK), UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, OR | (flags & SET_FLAGS) | D(dst) | S1(0) | IMM(32), UNMOVABLE_INS | (flags & SET_FLAGS)));
FAIL_IF(push_inst(compiler, OR | D(dst) | S1(0) | IMM(-1), DR(dst)));
/* Loop. */
FAIL_IF(push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(TMP_REG1) | S2(0), SET_FLAGS));
FAIL_IF(push_inst(compiler, SLL | D(TMP_REG1) | S1(TMP_REG1) | IMM(1), DR(TMP_REG1)));
FAIL_IF(push_inst(compiler, BICC | DA(0xe) | (-2 & DISP_MASK), UNMOVABLE_INS));
return push_inst(compiler, ADD | (flags & SET_FLAGS) | D(dst) | S1(dst) | IMM(1), UNMOVABLE_INS | (flags & SET_FLAGS));
case SLJIT_ADD:
return push_inst(compiler, ADD | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_ADDC:
return push_inst(compiler, ADDC | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_SUB:
return push_inst(compiler, SUB | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_SUBC:
return push_inst(compiler, SUBC | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_MUL:
FAIL_IF(push_inst(compiler, SMUL | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst)));
if (!(flags & SET_FLAGS))
return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SRA | D(TMP_REG1) | S1(dst) | IMM(31), DR(TMP_REG1)));
FAIL_IF(push_inst(compiler, RDY | D(TMP_REG4), DR(TMP_REG4)));
return push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(TMP_REG1) | S2(TMP_REG4), MOVABLE_INS | SET_FLAGS);
case SLJIT_AND:
return push_inst(compiler, AND | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_OR:
return push_inst(compiler, OR | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_XOR:
return push_inst(compiler, XOR | (flags & SET_FLAGS) | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst) | (flags & SET_FLAGS));
case SLJIT_SHL:
FAIL_IF(push_inst(compiler, SLL | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst)));
return !(flags & SET_FLAGS) ? SLJIT_SUCCESS : push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(dst) | S2(0), SET_FLAGS);
case SLJIT_LSHR:
FAIL_IF(push_inst(compiler, SRL | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst)));
return !(flags & SET_FLAGS) ? SLJIT_SUCCESS : push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(dst) | S2(0), SET_FLAGS);
case SLJIT_ASHR:
FAIL_IF(push_inst(compiler, SRA | D(dst) | S1(src1) | ARG2(flags, src2), DR(dst)));
return !(flags & SET_FLAGS) ? SLJIT_SUCCESS : push_inst(compiler, SUB | SET_FLAGS | D(0) | S1(dst) | S2(0), SET_FLAGS);
}
SLJIT_ASSERT_STOP();
return SLJIT_SUCCESS;
}
static SLJIT_INLINE sljit_si emit_const(struct sljit_compiler *compiler, sljit_si dst, sljit_sw init_value)
{
FAIL_IF(push_inst(compiler, SETHI | D(dst) | ((init_value >> 10) & 0x3fffff), DR(dst)));
return push_inst(compiler, OR | D(dst) | S1(dst) | IMM_ARG | (init_value & 0x3ff), DR(dst));
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_jump_addr(sljit_uw addr, sljit_uw new_addr)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffc00000) | ((new_addr >> 10) & 0x3fffff);
inst[1] = (inst[1] & 0xfffffc00) | (new_addr & 0x3ff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_const(sljit_uw addr, sljit_sw new_constant)
{
sljit_ins *inst = (sljit_ins*)addr;
inst[0] = (inst[0] & 0xffc00000) | ((new_constant >> 10) & 0x3fffff);
inst[1] = (inst[1] & 0xfffffc00) | (new_constant & 0x3ff);
SLJIT_CACHE_FLUSH(inst, inst + 2);
}

1383
ext/pcre/pcrelib/sljit/sljitNativeSPARC_common.c Normal file
View file

File diff suppressed because it is too large Load diff

10159
ext/pcre/pcrelib/sljit/sljitNativeTILEGX-encoder.c Normal file
View file

File diff suppressed because it is too large Load diff

2580
ext/pcre/pcrelib/sljit/sljitNativeTILEGX_64.c Normal file
View file

File diff suppressed because it is too large Load diff

557
ext/pcre/pcrelib/sljit/sljitNativeX86_32.c Normal file
View file

@ -0,0 +1,557 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* x86 32-bit arch dependent functions. */
static sljit_si emit_do_imm(struct sljit_compiler *compiler, sljit_ub opcode, sljit_sw imm)
{
sljit_ub *inst;
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1 + sizeof(sljit_sw));
FAIL_IF(!inst);
INC_SIZE(1 + sizeof(sljit_sw));
*inst++ = opcode;
*(sljit_sw*)inst = imm;
return SLJIT_SUCCESS;
}
static sljit_ub* generate_far_jump_code(struct sljit_jump *jump, sljit_ub *code_ptr, sljit_si type)
{
if (type == SLJIT_JUMP) {
*code_ptr++ = JMP_i32;
jump->addr++;
}
else if (type >= SLJIT_FAST_CALL) {
*code_ptr++ = CALL_i32;
jump->addr++;
}
else {
*code_ptr++ = GROUP_0F;
*code_ptr++ = get_jump_code(type);
jump->addr += 2;
}
if (jump->flags & JUMP_LABEL)
jump->flags |= PATCH_MW;
else
*(sljit_sw*)code_ptr = jump->u.target - (jump->addr + 4);
code_ptr += 4;
return code_ptr;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_enter(struct sljit_compiler *compiler, sljit_si args, sljit_si scratches, sljit_si saveds, sljit_si local_size)
{
sljit_si size;
sljit_si locals_offset;
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_enter(compiler, args, scratches, saveds, local_size);
compiler->scratches = scratches;
compiler->saveds = saveds;
compiler->args = args;
compiler->flags_saved = 0;
#if (defined SLJIT_DEBUG && SLJIT_DEBUG)
compiler->logical_local_size = local_size;
#endif
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
size = 1 + (saveds <= 3 ? saveds : 3) + (args > 0 ? (args * 2) : 0) + (args > 2 ? 2 : 0);
#else
size = 1 + (saveds <= 3 ? saveds : 3) + (args > 0 ? (2 + args * 3) : 0);
#endif
inst = (sljit_ub*)ensure_buf(compiler, 1 + size);
FAIL_IF(!inst);
INC_SIZE(size);
PUSH_REG(reg_map[TMP_REG1]);
#if !(defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
if (args > 0) {
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[TMP_REG1] << 3) | 0x4 /* esp */;
}
#endif
if (saveds > 2)
PUSH_REG(reg_map[SLJIT_SAVED_REG3]);
if (saveds > 1)
PUSH_REG(reg_map[SLJIT_SAVED_REG2]);
if (saveds > 0)
PUSH_REG(reg_map[SLJIT_SAVED_REG1]);
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
if (args > 0) {
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG1] << 3) | reg_map[SLJIT_SCRATCH_REG3];
}
if (args > 1) {
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG2] << 3) | reg_map[SLJIT_SCRATCH_REG2];
}
if (args > 2) {
*inst++ = MOV_r_rm;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SAVED_REG3] << 3) | 0x4 /* esp */;
*inst++ = 0x24;
*inst++ = sizeof(sljit_sw) * (3 + 2); /* saveds >= 3 as well. */
}
#else
if (args > 0) {
*inst++ = MOV_r_rm;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SAVED_REG1] << 3) | reg_map[TMP_REG1];
*inst++ = sizeof(sljit_sw) * 2;
}
if (args > 1) {
*inst++ = MOV_r_rm;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SAVED_REG2] << 3) | reg_map[TMP_REG1];
*inst++ = sizeof(sljit_sw) * 3;
}
if (args > 2) {
*inst++ = MOV_r_rm;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SAVED_REG3] << 3) | reg_map[TMP_REG1];
*inst++ = sizeof(sljit_sw) * 4;
}
#endif
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
locals_offset = 2 * sizeof(sljit_uw);
#else
SLJIT_COMPILE_ASSERT(FIXED_LOCALS_OFFSET >= 2 * sizeof(sljit_uw), require_at_least_two_words);
locals_offset = FIXED_LOCALS_OFFSET;
#endif
compiler->scratches_start = locals_offset;
if (scratches > 3)
locals_offset += (scratches - 3) * sizeof(sljit_uw);
compiler->saveds_start = locals_offset;
if (saveds > 3)
locals_offset += (saveds - 3) * sizeof(sljit_uw);
compiler->locals_offset = locals_offset;
#if defined(__APPLE__)
saveds = (2 + (saveds <= 3 ? saveds : 3)) * sizeof(sljit_uw);
local_size = ((locals_offset + saveds + local_size + 15) & ~15) - saveds;
#else
local_size = locals_offset + ((local_size + sizeof(sljit_uw) - 1) & ~(sizeof(sljit_uw) - 1));
#endif
compiler->local_size = local_size;
#ifdef _WIN32
if (local_size > 1024) {
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
FAIL_IF(emit_do_imm(compiler, MOV_r_i32 + reg_map[SLJIT_SCRATCH_REG1], local_size));
#else
local_size -= FIXED_LOCALS_OFFSET;
FAIL_IF(emit_do_imm(compiler, MOV_r_i32 + reg_map[SLJIT_SCRATCH_REG1], local_size));
FAIL_IF(emit_non_cum_binary(compiler, SUB_r_rm, SUB_rm_r, SUB, SUB_EAX_i32,
SLJIT_LOCALS_REG, 0, SLJIT_LOCALS_REG, 0, SLJIT_IMM, FIXED_LOCALS_OFFSET));
#endif
FAIL_IF(sljit_emit_ijump(compiler, SLJIT_CALL1, SLJIT_IMM, SLJIT_FUNC_OFFSET(sljit_grow_stack)));
}
#endif
SLJIT_ASSERT(local_size > 0);
return emit_non_cum_binary(compiler, SUB_r_rm, SUB_rm_r, SUB, SUB_EAX_i32,
SLJIT_LOCALS_REG, 0, SLJIT_LOCALS_REG, 0, SLJIT_IMM, local_size);
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_context(struct sljit_compiler *compiler, sljit_si args, sljit_si scratches, sljit_si saveds, sljit_si local_size)
{
sljit_si locals_offset;
CHECK_ERROR_VOID();
check_sljit_set_context(compiler, args, scratches, saveds, local_size);
compiler->scratches = scratches;
compiler->saveds = saveds;
compiler->args = args;
#if (defined SLJIT_DEBUG && SLJIT_DEBUG)
compiler->logical_local_size = local_size;
#endif
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
locals_offset = 2 * sizeof(sljit_uw);
#else
locals_offset = FIXED_LOCALS_OFFSET;
#endif
compiler->scratches_start = locals_offset;
if (scratches > 3)
locals_offset += (scratches - 3) * sizeof(sljit_uw);
compiler->saveds_start = locals_offset;
if (saveds > 3)
locals_offset += (saveds - 3) * sizeof(sljit_uw);
compiler->locals_offset = locals_offset;
#if defined(__APPLE__)
saveds = (2 + (saveds <= 3 ? saveds : 3)) * sizeof(sljit_uw);
compiler->local_size = ((locals_offset + saveds + local_size + 15) & ~15) - saveds;
#else
compiler->local_size = locals_offset + ((local_size + sizeof(sljit_uw) - 1) & ~(sizeof(sljit_uw) - 1));
#endif
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_return(struct sljit_compiler *compiler, sljit_si op, sljit_si src, sljit_sw srcw)
{
sljit_si size;
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_return(compiler, op, src, srcw);
SLJIT_ASSERT(compiler->args >= 0);
compiler->flags_saved = 0;
FAIL_IF(emit_mov_before_return(compiler, op, src, srcw));
SLJIT_ASSERT(compiler->local_size > 0);
FAIL_IF(emit_cum_binary(compiler, ADD_r_rm, ADD_rm_r, ADD, ADD_EAX_i32,
SLJIT_LOCALS_REG, 0, SLJIT_LOCALS_REG, 0, SLJIT_IMM, compiler->local_size));
size = 2 + (compiler->saveds <= 3 ? compiler->saveds : 3);
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
if (compiler->args > 2)
size += 2;
#else
if (compiler->args > 0)
size += 2;
#endif
inst = (sljit_ub*)ensure_buf(compiler, 1 + size);
FAIL_IF(!inst);
INC_SIZE(size);
if (compiler->saveds > 0)
POP_REG(reg_map[SLJIT_SAVED_REG1]);
if (compiler->saveds > 1)
POP_REG(reg_map[SLJIT_SAVED_REG2]);
if (compiler->saveds > 2)
POP_REG(reg_map[SLJIT_SAVED_REG3]);
POP_REG(reg_map[TMP_REG1]);
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
if (compiler->args > 2)
RET_I16(sizeof(sljit_sw));
else
RET();
#else
RET();
#endif
return SLJIT_SUCCESS;
}
/* --------------------------------------------------------------------- */
/* Operators */
/* --------------------------------------------------------------------- */
/* Size contains the flags as well. */
static sljit_ub* emit_x86_instruction(struct sljit_compiler *compiler, sljit_si size,
/* The register or immediate operand. */
sljit_si a, sljit_sw imma,
/* The general operand (not immediate). */
sljit_si b, sljit_sw immb)
{
sljit_ub *inst;
sljit_ub *buf_ptr;
sljit_si flags = size & ~0xf;
sljit_si inst_size;
/* Both cannot be switched on. */
SLJIT_ASSERT((flags & (EX86_BIN_INS | EX86_SHIFT_INS)) != (EX86_BIN_INS | EX86_SHIFT_INS));
/* Size flags not allowed for typed instructions. */
SLJIT_ASSERT(!(flags & (EX86_BIN_INS | EX86_SHIFT_INS)) || (flags & (EX86_BYTE_ARG | EX86_HALF_ARG)) == 0);
/* Both size flags cannot be switched on. */
SLJIT_ASSERT((flags & (EX86_BYTE_ARG | EX86_HALF_ARG)) != (EX86_BYTE_ARG | EX86_HALF_ARG));
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
/* SSE2 and immediate is not possible. */
SLJIT_ASSERT(!(a & SLJIT_IMM) || !(flags & EX86_SSE2));
SLJIT_ASSERT((flags & (EX86_PREF_F2 | EX86_PREF_F3)) != (EX86_PREF_F2 | EX86_PREF_F3)
&& (flags & (EX86_PREF_F2 | EX86_PREF_66)) != (EX86_PREF_F2 | EX86_PREF_66)
&& (flags & (EX86_PREF_F3 | EX86_PREF_66)) != (EX86_PREF_F3 | EX86_PREF_66));
#endif
size &= 0xf;
inst_size = size;
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
if (flags & (EX86_PREF_F2 | EX86_PREF_F3))
inst_size++;
#endif
if (flags & EX86_PREF_66)
inst_size++;
/* Calculate size of b. */
inst_size += 1; /* mod r/m byte. */
if (b & SLJIT_MEM) {
if ((b & REG_MASK) == SLJIT_UNUSED)
inst_size += sizeof(sljit_sw);
else if (immb != 0 && !(b & OFFS_REG_MASK)) {
/* Immediate operand. */
if (immb <= 127 && immb >= -128)
inst_size += sizeof(sljit_sb);
else
inst_size += sizeof(sljit_sw);
}
if ((b & REG_MASK) == SLJIT_LOCALS_REG && !(b & OFFS_REG_MASK))
b |= TO_OFFS_REG(SLJIT_LOCALS_REG);
if ((b & OFFS_REG_MASK) != SLJIT_UNUSED)
inst_size += 1; /* SIB byte. */
}
/* Calculate size of a. */
if (a & SLJIT_IMM) {
if (flags & EX86_BIN_INS) {
if (imma <= 127 && imma >= -128) {
inst_size += 1;
flags |= EX86_BYTE_ARG;
} else
inst_size += 4;
}
else if (flags & EX86_SHIFT_INS) {
imma &= 0x1f;
if (imma != 1) {
inst_size ++;
flags |= EX86_BYTE_ARG;
}
} else if (flags & EX86_BYTE_ARG)
inst_size++;
else if (flags & EX86_HALF_ARG)
inst_size += sizeof(short);
else
inst_size += sizeof(sljit_sw);
}
else
SLJIT_ASSERT(!(flags & EX86_SHIFT_INS) || a == SLJIT_PREF_SHIFT_REG);
inst = (sljit_ub*)ensure_buf(compiler, 1 + inst_size);
PTR_FAIL_IF(!inst);
/* Encoding the byte. */
INC_SIZE(inst_size);
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
if (flags & EX86_PREF_F2)
*inst++ = 0xf2;
if (flags & EX86_PREF_F3)
*inst++ = 0xf3;
#endif
if (flags & EX86_PREF_66)
*inst++ = 0x66;
buf_ptr = inst + size;
/* Encode mod/rm byte. */
if (!(flags & EX86_SHIFT_INS)) {
if ((flags & EX86_BIN_INS) && (a & SLJIT_IMM))
*inst = (flags & EX86_BYTE_ARG) ? GROUP_BINARY_83 : GROUP_BINARY_81;
if ((a & SLJIT_IMM) || (a == 0))
*buf_ptr = 0;
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
else if (!(flags & EX86_SSE2))
*buf_ptr = reg_map[a] << 3;
else
*buf_ptr = a << 3;
#else
else
*buf_ptr = reg_map[a] << 3;
#endif
}
else {
if (a & SLJIT_IMM) {
if (imma == 1)
*inst = GROUP_SHIFT_1;
else
*inst = GROUP_SHIFT_N;
} else
*inst = GROUP_SHIFT_CL;
*buf_ptr = 0;
}
if (!(b & SLJIT_MEM))
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
*buf_ptr++ |= MOD_REG + ((!(flags & EX86_SSE2)) ? reg_map[b] : b);
#else
*buf_ptr++ |= MOD_REG + reg_map[b];
#endif
else if ((b & REG_MASK) != SLJIT_UNUSED) {
if ((b & OFFS_REG_MASK) == SLJIT_UNUSED || (b & OFFS_REG_MASK) == TO_OFFS_REG(SLJIT_LOCALS_REG)) {
if (immb != 0) {
if (immb <= 127 && immb >= -128)
*buf_ptr |= 0x40;
else
*buf_ptr |= 0x80;
}
if ((b & OFFS_REG_MASK) == SLJIT_UNUSED)
*buf_ptr++ |= reg_map[b & REG_MASK];
else {
*buf_ptr++ |= 0x04;
*buf_ptr++ = reg_map[b & REG_MASK] | (reg_map[OFFS_REG(b)] << 3);
}
if (immb != 0) {
if (immb <= 127 && immb >= -128)
*buf_ptr++ = immb; /* 8 bit displacement. */
else {
*(sljit_sw*)buf_ptr = immb; /* 32 bit displacement. */
buf_ptr += sizeof(sljit_sw);
}
}
}
else {
*buf_ptr++ |= 0x04;
*buf_ptr++ = reg_map[b & REG_MASK] | (reg_map[OFFS_REG(b)] << 3) | (immb << 6);
}
}
else {
*buf_ptr++ |= 0x05;
*(sljit_sw*)buf_ptr = immb; /* 32 bit displacement. */
buf_ptr += sizeof(sljit_sw);
}
if (a & SLJIT_IMM) {
if (flags & EX86_BYTE_ARG)
*buf_ptr = imma;
else if (flags & EX86_HALF_ARG)
*(short*)buf_ptr = imma;
else if (!(flags & EX86_SHIFT_INS))
*(sljit_sw*)buf_ptr = imma;
}
return !(flags & EX86_SHIFT_INS) ? inst : (inst + 1);
}
/* --------------------------------------------------------------------- */
/* Call / return instructions */
/* --------------------------------------------------------------------- */
static SLJIT_INLINE sljit_si call_with_args(struct sljit_compiler *compiler, sljit_si type)
{
sljit_ub *inst;
#if (defined SLJIT_X86_32_FASTCALL && SLJIT_X86_32_FASTCALL)
inst = (sljit_ub*)ensure_buf(compiler, type >= SLJIT_CALL3 ? 1 + 2 + 1 : 1 + 2);
FAIL_IF(!inst);
INC_SIZE(type >= SLJIT_CALL3 ? 2 + 1 : 2);
if (type >= SLJIT_CALL3)
PUSH_REG(reg_map[SLJIT_SCRATCH_REG3]);
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SCRATCH_REG3] << 3) | reg_map[SLJIT_SCRATCH_REG1];
#else
inst = (sljit_ub*)ensure_buf(compiler, 1 + 4 * (type - SLJIT_CALL0));
FAIL_IF(!inst);
INC_SIZE(4 * (type - SLJIT_CALL0));
*inst++ = MOV_rm_r;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SCRATCH_REG1] << 3) | 0x4 /* SIB */;
*inst++ = (0x4 /* none*/ << 3) | reg_map[SLJIT_LOCALS_REG];
*inst++ = 0;
if (type >= SLJIT_CALL2) {
*inst++ = MOV_rm_r;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SCRATCH_REG2] << 3) | 0x4 /* SIB */;
*inst++ = (0x4 /* none*/ << 3) | reg_map[SLJIT_LOCALS_REG];
*inst++ = sizeof(sljit_sw);
}
if (type >= SLJIT_CALL3) {
*inst++ = MOV_rm_r;
*inst++ = MOD_DISP8 | (reg_map[SLJIT_SCRATCH_REG3] << 3) | 0x4 /* SIB */;
*inst++ = (0x4 /* none*/ << 3) | reg_map[SLJIT_LOCALS_REG];
*inst++ = 2 * sizeof(sljit_sw);
}
#endif
return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_enter(struct sljit_compiler *compiler, sljit_si dst, sljit_sw dstw)
{
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_fast_enter(compiler, dst, dstw);
ADJUST_LOCAL_OFFSET(dst, dstw);
CHECK_EXTRA_REGS(dst, dstw, (void)0);
/* For UNUSED dst. Uncommon, but possible. */
if (dst == SLJIT_UNUSED)
dst = TMP_REG1;
if (FAST_IS_REG(dst)) {
/* Unused dest is possible here. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1);
POP_REG(reg_map[dst]);
return SLJIT_SUCCESS;
}
/* Memory. */
inst = emit_x86_instruction(compiler, 1, 0, 0, dst, dstw);
FAIL_IF(!inst);
*inst++ = POP_rm;
return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_return(struct sljit_compiler *compiler, sljit_si src, sljit_sw srcw)
{
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_fast_return(compiler, src, srcw);
ADJUST_LOCAL_OFFSET(src, srcw);
CHECK_EXTRA_REGS(src, srcw, (void)0);
if (FAST_IS_REG(src)) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1 + 1);
PUSH_REG(reg_map[src]);
}
else if (src & SLJIT_MEM) {
inst = emit_x86_instruction(compiler, 1, 0, 0, src, srcw);
FAIL_IF(!inst);
*inst++ = GROUP_FF;
*inst |= PUSH_rm;
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1);
}
else {
/* SLJIT_IMM. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 5 + 1);
FAIL_IF(!inst);
INC_SIZE(5 + 1);
*inst++ = PUSH_i32;
*(sljit_sw*)inst = srcw;
inst += sizeof(sljit_sw);
}
RET();
return SLJIT_SUCCESS;
}

810
ext/pcre/pcrelib/sljit/sljitNativeX86_64.c Normal file
View file

@ -0,0 +1,810 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* x86 64-bit arch dependent functions. */
static sljit_si emit_load_imm64(struct sljit_compiler *compiler, sljit_si reg, sljit_sw imm)
{
sljit_ub *inst;
inst = (sljit_ub*)ensure_buf(compiler, 1 + 2 + sizeof(sljit_sw));
FAIL_IF(!inst);
INC_SIZE(2 + sizeof(sljit_sw));
*inst++ = REX_W | ((reg_map[reg] <= 7) ? 0 : REX_B);
*inst++ = MOV_r_i32 + (reg_map[reg] & 0x7);
*(sljit_sw*)inst = imm;
return SLJIT_SUCCESS;
}
static sljit_ub* generate_far_jump_code(struct sljit_jump *jump, sljit_ub *code_ptr, sljit_si type)
{
if (type < SLJIT_JUMP) {
/* Invert type. */
*code_ptr++ = get_jump_code(type ^ 0x1) - 0x10;
*code_ptr++ = 10 + 3;
}
SLJIT_COMPILE_ASSERT(reg_map[TMP_REG3] == 9, tmp3_is_9_first);
*code_ptr++ = REX_W | REX_B;
*code_ptr++ = MOV_r_i32 + 1;
jump->addr = (sljit_uw)code_ptr;
if (jump->flags & JUMP_LABEL)
jump->flags |= PATCH_MD;
else
*(sljit_sw*)code_ptr = jump->u.target;
code_ptr += sizeof(sljit_sw);
*code_ptr++ = REX_B;
*code_ptr++ = GROUP_FF;
*code_ptr++ = (type >= SLJIT_FAST_CALL) ? (MOD_REG | CALL_rm | 1) : (MOD_REG | JMP_rm | 1);
return code_ptr;
}
static sljit_ub* generate_fixed_jump(sljit_ub *code_ptr, sljit_sw addr, sljit_si type)
{
sljit_sw delta = addr - ((sljit_sw)code_ptr + 1 + sizeof(sljit_si));
if (delta <= HALFWORD_MAX && delta >= HALFWORD_MIN) {
*code_ptr++ = (type == 2) ? CALL_i32 : JMP_i32;
*(sljit_sw*)code_ptr = delta;
}
else {
SLJIT_COMPILE_ASSERT(reg_map[TMP_REG3] == 9, tmp3_is_9_second);
*code_ptr++ = REX_W | REX_B;
*code_ptr++ = MOV_r_i32 + 1;
*(sljit_sw*)code_ptr = addr;
code_ptr += sizeof(sljit_sw);
*code_ptr++ = REX_B;
*code_ptr++ = GROUP_FF;
*code_ptr++ = (type == 2) ? (MOD_REG | CALL_rm | 1) : (MOD_REG | JMP_rm | 1);
}
return code_ptr;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_enter(struct sljit_compiler *compiler, sljit_si args, sljit_si scratches, sljit_si saveds, sljit_si local_size)
{
sljit_si size, pushed_size;
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_enter(compiler, args, scratches, saveds, local_size);
compiler->scratches = scratches;
compiler->saveds = saveds;
compiler->flags_saved = 0;
#if (defined SLJIT_DEBUG && SLJIT_DEBUG)
compiler->logical_local_size = local_size;
#endif
size = saveds;
/* Including the return address saved by the call instruction. */
pushed_size = (saveds + 1) * sizeof(sljit_sw);
#ifndef _WIN64
if (saveds >= 2)
size += saveds - 1;
#else
if (saveds >= 4)
size += saveds - 3;
if (scratches >= 5) {
size += (5 - 4) * 2;
pushed_size += sizeof(sljit_sw);
}
#endif
size += args * 3;
if (size > 0) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + size);
FAIL_IF(!inst);
INC_SIZE(size);
if (saveds >= 5) {
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_EREG2] >= 8, saved_ereg2_is_hireg);
*inst++ = REX_B;
PUSH_REG(reg_lmap[SLJIT_SAVED_EREG2]);
}
if (saveds >= 4) {
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_EREG1] >= 8, saved_ereg1_is_hireg);
*inst++ = REX_B;
PUSH_REG(reg_lmap[SLJIT_SAVED_EREG1]);
}
if (saveds >= 3) {
#ifndef _WIN64
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_REG3] >= 8, saved_reg3_is_hireg);
*inst++ = REX_B;
#else
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_REG3] < 8, saved_reg3_is_loreg);
#endif
PUSH_REG(reg_lmap[SLJIT_SAVED_REG3]);
}
if (saveds >= 2) {
#ifndef _WIN64
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_REG2] >= 8, saved_reg2_is_hireg);
*inst++ = REX_B;
#else
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_REG2] < 8, saved_reg2_is_loreg);
#endif
PUSH_REG(reg_lmap[SLJIT_SAVED_REG2]);
}
if (saveds >= 1) {
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SAVED_REG1] < 8, saved_reg1_is_loreg);
PUSH_REG(reg_lmap[SLJIT_SAVED_REG1]);
}
#ifdef _WIN64
if (scratches >= 5) {
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_TEMPORARY_EREG2] >= 8, temporary_ereg2_is_hireg);
*inst++ = REX_B;
PUSH_REG(reg_lmap[SLJIT_TEMPORARY_EREG2]);
}
#endif
#ifndef _WIN64
if (args > 0) {
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG1] << 3) | 0x7 /* rdi */;
}
if (args > 1) {
*inst++ = REX_W | REX_R;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_lmap[SLJIT_SAVED_REG2] << 3) | 0x6 /* rsi */;
}
if (args > 2) {
*inst++ = REX_W | REX_R;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_lmap[SLJIT_SAVED_REG3] << 3) | 0x2 /* rdx */;
}
#else
if (args > 0) {
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG1] << 3) | 0x1 /* rcx */;
}
if (args > 1) {
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG2] << 3) | 0x2 /* rdx */;
}
if (args > 2) {
*inst++ = REX_W | REX_B;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (reg_map[SLJIT_SAVED_REG3] << 3) | 0x0 /* r8 */;
}
#endif
}
local_size = ((local_size + FIXED_LOCALS_OFFSET + pushed_size + 16 - 1) & ~(16 - 1)) - pushed_size;
compiler->local_size = local_size;
#ifdef _WIN64
if (local_size > 1024) {
/* Allocate stack for the callback, which grows the stack. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 4 + (3 + sizeof(sljit_si)));
FAIL_IF(!inst);
INC_SIZE(4 + (3 + sizeof(sljit_si)));
*inst++ = REX_W;
*inst++ = GROUP_BINARY_83;
*inst++ = MOD_REG | SUB | 4;
/* Pushed size must be divisible by 8. */
SLJIT_ASSERT(!(pushed_size & 0x7));
if (pushed_size & 0x8) {
*inst++ = 5 * sizeof(sljit_sw);
local_size -= 5 * sizeof(sljit_sw);
} else {
*inst++ = 4 * sizeof(sljit_sw);
local_size -= 4 * sizeof(sljit_sw);
}
/* Second instruction */
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SCRATCH_REG1] < 8, temporary_reg1_is_loreg);
*inst++ = REX_W;
*inst++ = MOV_rm_i32;
*inst++ = MOD_REG | reg_lmap[SLJIT_SCRATCH_REG1];
*(sljit_si*)inst = local_size;
#if (defined SLJIT_VERBOSE && SLJIT_VERBOSE) || (defined SLJIT_DEBUG && SLJIT_DEBUG)
compiler->skip_checks = 1;
#endif
FAIL_IF(sljit_emit_ijump(compiler, SLJIT_CALL1, SLJIT_IMM, SLJIT_FUNC_OFFSET(sljit_grow_stack)));
}
#endif
SLJIT_ASSERT(local_size > 0);
if (local_size <= 127) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 4);
FAIL_IF(!inst);
INC_SIZE(4);
*inst++ = REX_W;
*inst++ = GROUP_BINARY_83;
*inst++ = MOD_REG | SUB | 4;
*inst++ = local_size;
}
else {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 7);
FAIL_IF(!inst);
INC_SIZE(7);
*inst++ = REX_W;
*inst++ = GROUP_BINARY_81;
*inst++ = MOD_REG | SUB | 4;
*(sljit_si*)inst = local_size;
inst += sizeof(sljit_si);
}
#ifdef _WIN64
/* Save xmm6 with MOVAPS instruction. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 5);
FAIL_IF(!inst);
INC_SIZE(5);
*inst++ = GROUP_0F;
*(sljit_si*)inst = 0x20247429;
#endif
return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE void sljit_set_context(struct sljit_compiler *compiler, sljit_si args, sljit_si scratches, sljit_si saveds, sljit_si local_size)
{
sljit_si pushed_size;
CHECK_ERROR_VOID();
check_sljit_set_context(compiler, args, scratches, saveds, local_size);
compiler->scratches = scratches;
compiler->saveds = saveds;
#if (defined SLJIT_DEBUG && SLJIT_DEBUG)
compiler->logical_local_size = local_size;
#endif
/* Including the return address saved by the call instruction. */
pushed_size = (saveds + 1) * sizeof(sljit_sw);
#ifdef _WIN64
if (scratches >= 5)
pushed_size += sizeof(sljit_sw);
#endif
compiler->local_size = ((local_size + FIXED_LOCALS_OFFSET + pushed_size + 16 - 1) & ~(16 - 1)) - pushed_size;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_return(struct sljit_compiler *compiler, sljit_si op, sljit_si src, sljit_sw srcw)
{
sljit_si size;
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_return(compiler, op, src, srcw);
compiler->flags_saved = 0;
FAIL_IF(emit_mov_before_return(compiler, op, src, srcw));
#ifdef _WIN64
/* Restore xmm6 with MOVAPS instruction. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 5);
FAIL_IF(!inst);
INC_SIZE(5);
*inst++ = GROUP_0F;
*(sljit_si*)inst = 0x20247428;
#endif
SLJIT_ASSERT(compiler->local_size > 0);
if (compiler->local_size <= 127) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 4);
FAIL_IF(!inst);
INC_SIZE(4);
*inst++ = REX_W;
*inst++ = GROUP_BINARY_83;
*inst++ = MOD_REG | ADD | 4;
*inst = compiler->local_size;
}
else {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 7);
FAIL_IF(!inst);
INC_SIZE(7);
*inst++ = REX_W;
*inst++ = GROUP_BINARY_81;
*inst++ = MOD_REG | ADD | 4;
*(sljit_si*)inst = compiler->local_size;
}
size = 1 + compiler->saveds;
#ifndef _WIN64
if (compiler->saveds >= 2)
size += compiler->saveds - 1;
#else
if (compiler->saveds >= 4)
size += compiler->saveds - 3;
if (compiler->scratches >= 5)
size += (5 - 4) * 2;
#endif
inst = (sljit_ub*)ensure_buf(compiler, 1 + size);
FAIL_IF(!inst);
INC_SIZE(size);
#ifdef _WIN64
if (compiler->scratches >= 5) {
*inst++ = REX_B;
POP_REG(reg_lmap[SLJIT_TEMPORARY_EREG2]);
}
#endif
if (compiler->saveds >= 1)
POP_REG(reg_map[SLJIT_SAVED_REG1]);
if (compiler->saveds >= 2) {
#ifndef _WIN64
*inst++ = REX_B;
#endif
POP_REG(reg_lmap[SLJIT_SAVED_REG2]);
}
if (compiler->saveds >= 3) {
#ifndef _WIN64
*inst++ = REX_B;
#endif
POP_REG(reg_lmap[SLJIT_SAVED_REG3]);
}
if (compiler->saveds >= 4) {
*inst++ = REX_B;
POP_REG(reg_lmap[SLJIT_SAVED_EREG1]);
}
if (compiler->saveds >= 5) {
*inst++ = REX_B;
POP_REG(reg_lmap[SLJIT_SAVED_EREG2]);
}
RET();
return SLJIT_SUCCESS;
}
/* --------------------------------------------------------------------- */
/* Operators */
/* --------------------------------------------------------------------- */
static sljit_si emit_do_imm32(struct sljit_compiler *compiler, sljit_ub rex, sljit_ub opcode, sljit_sw imm)
{
sljit_ub *inst;
sljit_si length = 1 + (rex ? 1 : 0) + sizeof(sljit_si);
inst = (sljit_ub*)ensure_buf(compiler, 1 + length);
FAIL_IF(!inst);
INC_SIZE(length);
if (rex)
*inst++ = rex;
*inst++ = opcode;
*(sljit_si*)inst = imm;
return SLJIT_SUCCESS;
}
static sljit_ub* emit_x86_instruction(struct sljit_compiler *compiler, sljit_si size,
/* The register or immediate operand. */
sljit_si a, sljit_sw imma,
/* The general operand (not immediate). */
sljit_si b, sljit_sw immb)
{
sljit_ub *inst;
sljit_ub *buf_ptr;
sljit_ub rex = 0;
sljit_si flags = size & ~0xf;
sljit_si inst_size;
/* The immediate operand must be 32 bit. */
SLJIT_ASSERT(!(a & SLJIT_IMM) || compiler->mode32 || IS_HALFWORD(imma));
/* Both cannot be switched on. */
SLJIT_ASSERT((flags & (EX86_BIN_INS | EX86_SHIFT_INS)) != (EX86_BIN_INS | EX86_SHIFT_INS));
/* Size flags not allowed for typed instructions. */
SLJIT_ASSERT(!(flags & (EX86_BIN_INS | EX86_SHIFT_INS)) || (flags & (EX86_BYTE_ARG | EX86_HALF_ARG)) == 0);
/* Both size flags cannot be switched on. */
SLJIT_ASSERT((flags & (EX86_BYTE_ARG | EX86_HALF_ARG)) != (EX86_BYTE_ARG | EX86_HALF_ARG));
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
/* SSE2 and immediate is not possible. */
SLJIT_ASSERT(!(a & SLJIT_IMM) || !(flags & EX86_SSE2));
SLJIT_ASSERT((flags & (EX86_PREF_F2 | EX86_PREF_F3)) != (EX86_PREF_F2 | EX86_PREF_F3)
&& (flags & (EX86_PREF_F2 | EX86_PREF_66)) != (EX86_PREF_F2 | EX86_PREF_66)
&& (flags & (EX86_PREF_F3 | EX86_PREF_66)) != (EX86_PREF_F3 | EX86_PREF_66));
#endif
size &= 0xf;
inst_size = size;
if ((b & SLJIT_MEM) && !(b & OFFS_REG_MASK) && NOT_HALFWORD(immb)) {
if (emit_load_imm64(compiler, TMP_REG3, immb))
return NULL;
immb = 0;
if (b & REG_MASK)
b |= TO_OFFS_REG(TMP_REG3);
else
b |= TMP_REG3;
}
if (!compiler->mode32 && !(flags & EX86_NO_REXW))
rex |= REX_W;
else if (flags & EX86_REX)
rex |= REX;
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
if (flags & (EX86_PREF_F2 | EX86_PREF_F3))
inst_size++;
#endif
if (flags & EX86_PREF_66)
inst_size++;
/* Calculate size of b. */
inst_size += 1; /* mod r/m byte. */
if (b & SLJIT_MEM) {
if ((b & REG_MASK) == SLJIT_UNUSED)
inst_size += 1 + sizeof(sljit_si); /* SIB byte required to avoid RIP based addressing. */
else {
if (reg_map[b & REG_MASK] >= 8)
rex |= REX_B;
if (immb != 0 && !(b & OFFS_REG_MASK)) {
/* Immediate operand. */
if (immb <= 127 && immb >= -128)
inst_size += sizeof(sljit_sb);
else
inst_size += sizeof(sljit_si);
}
}
if ((b & REG_MASK) == SLJIT_LOCALS_REG && !(b & OFFS_REG_MASK))
b |= TO_OFFS_REG(SLJIT_LOCALS_REG);
if ((b & OFFS_REG_MASK) != SLJIT_UNUSED) {
inst_size += 1; /* SIB byte. */
if (reg_map[OFFS_REG(b)] >= 8)
rex |= REX_X;
}
}
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
else if (!(flags & EX86_SSE2) && reg_map[b] >= 8)
rex |= REX_B;
#else
else if (reg_map[b] >= 8)
rex |= REX_B;
#endif
if (a & SLJIT_IMM) {
if (flags & EX86_BIN_INS) {
if (imma <= 127 && imma >= -128) {
inst_size += 1;
flags |= EX86_BYTE_ARG;
} else
inst_size += 4;
}
else if (flags & EX86_SHIFT_INS) {
imma &= compiler->mode32 ? 0x1f : 0x3f;
if (imma != 1) {
inst_size ++;
flags |= EX86_BYTE_ARG;
}
} else if (flags & EX86_BYTE_ARG)
inst_size++;
else if (flags & EX86_HALF_ARG)
inst_size += sizeof(short);
else
inst_size += sizeof(sljit_si);
}
else {
SLJIT_ASSERT(!(flags & EX86_SHIFT_INS) || a == SLJIT_PREF_SHIFT_REG);
/* reg_map[SLJIT_PREF_SHIFT_REG] is less than 8. */
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
if (!(flags & EX86_SSE2) && reg_map[a] >= 8)
rex |= REX_R;
#else
if (reg_map[a] >= 8)
rex |= REX_R;
#endif
}
if (rex)
inst_size++;
inst = (sljit_ub*)ensure_buf(compiler, 1 + inst_size);
PTR_FAIL_IF(!inst);
/* Encoding the byte. */
INC_SIZE(inst_size);
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
if (flags & EX86_PREF_F2)
*inst++ = 0xf2;
if (flags & EX86_PREF_F3)
*inst++ = 0xf3;
#endif
if (flags & EX86_PREF_66)
*inst++ = 0x66;
if (rex)
*inst++ = rex;
buf_ptr = inst + size;
/* Encode mod/rm byte. */
if (!(flags & EX86_SHIFT_INS)) {
if ((flags & EX86_BIN_INS) && (a & SLJIT_IMM))
*inst = (flags & EX86_BYTE_ARG) ? GROUP_BINARY_83 : GROUP_BINARY_81;
if ((a & SLJIT_IMM) || (a == 0))
*buf_ptr = 0;
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
else if (!(flags & EX86_SSE2))
*buf_ptr = reg_lmap[a] << 3;
else
*buf_ptr = a << 3;
#else
else
*buf_ptr = reg_lmap[a] << 3;
#endif
}
else {
if (a & SLJIT_IMM) {
if (imma == 1)
*inst = GROUP_SHIFT_1;
else
*inst = GROUP_SHIFT_N;
} else
*inst = GROUP_SHIFT_CL;
*buf_ptr = 0;
}
if (!(b & SLJIT_MEM))
#if (defined SLJIT_SSE2 && SLJIT_SSE2)
*buf_ptr++ |= MOD_REG + ((!(flags & EX86_SSE2)) ? reg_lmap[b] : b);
#else
*buf_ptr++ |= MOD_REG + reg_lmap[b];
#endif
else if ((b & REG_MASK) != SLJIT_UNUSED) {
if ((b & OFFS_REG_MASK) == SLJIT_UNUSED || (b & OFFS_REG_MASK) == TO_OFFS_REG(SLJIT_LOCALS_REG)) {
if (immb != 0) {
if (immb <= 127 && immb >= -128)
*buf_ptr |= 0x40;
else
*buf_ptr |= 0x80;
}
if ((b & OFFS_REG_MASK) == SLJIT_UNUSED)
*buf_ptr++ |= reg_lmap[b & REG_MASK];
else {
*buf_ptr++ |= 0x04;
*buf_ptr++ = reg_lmap[b & REG_MASK] | (reg_lmap[OFFS_REG(b)] << 3);
}
if (immb != 0) {
if (immb <= 127 && immb >= -128)
*buf_ptr++ = immb; /* 8 bit displacement. */
else {
*(sljit_si*)buf_ptr = immb; /* 32 bit displacement. */
buf_ptr += sizeof(sljit_si);
}
}
}
else {
*buf_ptr++ |= 0x04;
*buf_ptr++ = reg_lmap[b & REG_MASK] | (reg_lmap[OFFS_REG(b)] << 3) | (immb << 6);
}
}
else {
*buf_ptr++ |= 0x04;
*buf_ptr++ = 0x25;
*(sljit_si*)buf_ptr = immb; /* 32 bit displacement. */
buf_ptr += sizeof(sljit_si);
}
if (a & SLJIT_IMM) {
if (flags & EX86_BYTE_ARG)
*buf_ptr = imma;
else if (flags & EX86_HALF_ARG)
*(short*)buf_ptr = imma;
else if (!(flags & EX86_SHIFT_INS))
*(sljit_si*)buf_ptr = imma;
}
return !(flags & EX86_SHIFT_INS) ? inst : (inst + 1);
}
/* --------------------------------------------------------------------- */
/* Call / return instructions */
/* --------------------------------------------------------------------- */
static SLJIT_INLINE sljit_si call_with_args(struct sljit_compiler *compiler, sljit_si type)
{
sljit_ub *inst;
#ifndef _WIN64
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SCRATCH_REG2] == 6 && reg_map[SLJIT_SCRATCH_REG1] < 8 && reg_map[SLJIT_SCRATCH_REG3] < 8, args_registers);
inst = (sljit_ub*)ensure_buf(compiler, 1 + ((type < SLJIT_CALL3) ? 3 : 6));
FAIL_IF(!inst);
INC_SIZE((type < SLJIT_CALL3) ? 3 : 6);
if (type >= SLJIT_CALL3) {
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (0x2 /* rdx */ << 3) | reg_lmap[SLJIT_SCRATCH_REG3];
}
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (0x7 /* rdi */ << 3) | reg_lmap[SLJIT_SCRATCH_REG1];
#else
SLJIT_COMPILE_ASSERT(reg_map[SLJIT_SCRATCH_REG2] == 2 && reg_map[SLJIT_SCRATCH_REG1] < 8 && reg_map[SLJIT_SCRATCH_REG3] < 8, args_registers);
inst = (sljit_ub*)ensure_buf(compiler, 1 + ((type < SLJIT_CALL3) ? 3 : 6));
FAIL_IF(!inst);
INC_SIZE((type < SLJIT_CALL3) ? 3 : 6);
if (type >= SLJIT_CALL3) {
*inst++ = REX_W | REX_R;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (0x0 /* r8 */ << 3) | reg_lmap[SLJIT_SCRATCH_REG3];
}
*inst++ = REX_W;
*inst++ = MOV_r_rm;
*inst++ = MOD_REG | (0x1 /* rcx */ << 3) | reg_lmap[SLJIT_SCRATCH_REG1];
#endif
return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_enter(struct sljit_compiler *compiler, sljit_si dst, sljit_sw dstw)
{
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_fast_enter(compiler, dst, dstw);
ADJUST_LOCAL_OFFSET(dst, dstw);
/* For UNUSED dst. Uncommon, but possible. */
if (dst == SLJIT_UNUSED)
dst = TMP_REG1;
if (FAST_IS_REG(dst)) {
if (reg_map[dst] < 8) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1);
POP_REG(reg_lmap[dst]);
return SLJIT_SUCCESS;
}
inst = (sljit_ub*)ensure_buf(compiler, 1 + 2);
FAIL_IF(!inst);
INC_SIZE(2);
*inst++ = REX_B;
POP_REG(reg_lmap[dst]);
return SLJIT_SUCCESS;
}
/* REX_W is not necessary (src is not immediate). */
compiler->mode32 = 1;
inst = emit_x86_instruction(compiler, 1, 0, 0, dst, dstw);
FAIL_IF(!inst);
*inst++ = POP_rm;
return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_return(struct sljit_compiler *compiler, sljit_si src, sljit_sw srcw)
{
sljit_ub *inst;
CHECK_ERROR();
check_sljit_emit_fast_return(compiler, src, srcw);
ADJUST_LOCAL_OFFSET(src, srcw);
if ((src & SLJIT_IMM) && NOT_HALFWORD(srcw)) {
FAIL_IF(emit_load_imm64(compiler, TMP_REG1, srcw));
src = TMP_REG1;
}
if (FAST_IS_REG(src)) {
if (reg_map[src] < 8) {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1 + 1);
PUSH_REG(reg_lmap[src]);
}
else {
inst = (sljit_ub*)ensure_buf(compiler, 1 + 2 + 1);
FAIL_IF(!inst);
INC_SIZE(2 + 1);
*inst++ = REX_B;
PUSH_REG(reg_lmap[src]);
}
}
else if (src & SLJIT_MEM) {
/* REX_W is not necessary (src is not immediate). */
compiler->mode32 = 1;
inst = emit_x86_instruction(compiler, 1, 0, 0, src, srcw);
FAIL_IF(!inst);
*inst++ = GROUP_FF;
*inst |= PUSH_rm;
inst = (sljit_ub*)ensure_buf(compiler, 1 + 1);
FAIL_IF(!inst);
INC_SIZE(1);
}
else {
SLJIT_ASSERT(IS_HALFWORD(srcw));
/* SLJIT_IMM. */
inst = (sljit_ub*)ensure_buf(compiler, 1 + 5 + 1);
FAIL_IF(!inst);
INC_SIZE(5 + 1);
*inst++ = PUSH_i32;
*(sljit_si*)inst = srcw;
inst += sizeof(sljit_si);
}
RET();
return SLJIT_SUCCESS;
}
/* --------------------------------------------------------------------- */
/* Extend input */
/* --------------------------------------------------------------------- */
static sljit_si emit_mov_int(struct sljit_compiler *compiler, sljit_si sign,
sljit_si dst, sljit_sw dstw,
sljit_si src, sljit_sw srcw)
{
sljit_ub* inst;
sljit_si dst_r;
compiler->mode32 = 0;
if (dst == SLJIT_UNUSED && !(src & SLJIT_MEM))
return SLJIT_SUCCESS; /* Empty instruction. */
if (src & SLJIT_IMM) {
if (FAST_IS_REG(dst)) {
if (sign || ((sljit_uw)srcw <= 0x7fffffff)) {
inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, (sljit_sw)(sljit_si)srcw, dst, dstw);
FAIL_IF(!inst);
*inst = MOV_rm_i32;
return SLJIT_SUCCESS;
}
return emit_load_imm64(compiler, dst, srcw);
}
compiler->mode32 = 1;
inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, (sljit_sw)(sljit_si)srcw, dst, dstw);
FAIL_IF(!inst);
*inst = MOV_rm_i32;
compiler->mode32 = 0;
return SLJIT_SUCCESS;
}
dst_r = FAST_IS_REG(dst) ? dst : TMP_REG1;
if ((dst & SLJIT_MEM) && FAST_IS_REG(src))
dst_r = src;
else {
if (sign) {
inst = emit_x86_instruction(compiler, 1, dst_r, 0, src, srcw);
FAIL_IF(!inst);
*inst++ = MOVSXD_r_rm;
} else {
compiler->mode32 = 1;
FAIL_IF(emit_mov(compiler, dst_r, 0, src, srcw));
compiler->mode32 = 0;
}
}
if (dst & SLJIT_MEM) {
compiler->mode32 = 1;
inst = emit_x86_instruction(compiler, 1, dst_r, 0, dst, dstw);
FAIL_IF(!inst);
*inst = MOV_rm_r;
compiler->mode32 = 0;
}
return SLJIT_SUCCESS;
}

2883
ext/pcre/pcrelib/sljit/sljitNativeX86_common.c Normal file
View file

File diff suppressed because it is too large Load diff

332
ext/pcre/pcrelib/sljit/sljitUtils.c Normal file
View file

@ -0,0 +1,332 @@
/*
* Stack-less Just-In-Time compiler
*
* Copyright 2009-2012 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.
*/
/* ------------------------------------------------------------------------ */
/* Locks */
/* ------------------------------------------------------------------------ */
#if (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR) || (defined SLJIT_UTIL_GLOBAL_LOCK && SLJIT_UTIL_GLOBAL_LOCK)
#if (defined SLJIT_SINGLE_THREADED && SLJIT_SINGLE_THREADED)
#if (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR)
static SLJIT_INLINE void allocator_grab_lock(void)
{
/* Always successful. */
}
static SLJIT_INLINE void allocator_release_lock(void)
{
/* Always successful. */
}
#endif /* SLJIT_EXECUTABLE_ALLOCATOR */
#if (defined SLJIT_UTIL_GLOBAL_LOCK && SLJIT_UTIL_GLOBAL_LOCK)
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_grab_lock(void)
{
/* Always successful. */
}
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_release_lock(void)
{
/* Always successful. */
}
#endif /* SLJIT_UTIL_GLOBAL_LOCK */
#elif defined(_WIN32) /* SLJIT_SINGLE_THREADED */
#include "windows.h"
#if (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR)
static HANDLE allocator_mutex = 0;
static SLJIT_INLINE void allocator_grab_lock(void)
{
/* No idea what to do if an error occures. Static mutexes should never fail... */
if (!allocator_mutex)
allocator_mutex = CreateMutex(NULL, TRUE, NULL);
else
WaitForSingleObject(allocator_mutex, INFINITE);
}
static SLJIT_INLINE void allocator_release_lock(void)
{
ReleaseMutex(allocator_mutex);
}
#endif /* SLJIT_EXECUTABLE_ALLOCATOR */
#if (defined SLJIT_UTIL_GLOBAL_LOCK && SLJIT_UTIL_GLOBAL_LOCK)
static HANDLE global_mutex = 0;
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_grab_lock(void)
{
/* No idea what to do if an error occures. Static mutexes should never fail... */
if (!global_mutex)
global_mutex = CreateMutex(NULL, TRUE, NULL);
else
WaitForSingleObject(global_mutex, INFINITE);
}
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_release_lock(void)
{
ReleaseMutex(global_mutex);
}
#endif /* SLJIT_UTIL_GLOBAL_LOCK */
#else /* _WIN32 */
#if (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR)
#include <pthread.h>
static pthread_mutex_t allocator_mutex = PTHREAD_MUTEX_INITIALIZER;
static SLJIT_INLINE void allocator_grab_lock(void)
{
pthread_mutex_lock(&allocator_mutex);
}
static SLJIT_INLINE void allocator_release_lock(void)
{
pthread_mutex_unlock(&allocator_mutex);
}
#endif /* SLJIT_EXECUTABLE_ALLOCATOR */
#if (defined SLJIT_UTIL_GLOBAL_LOCK && SLJIT_UTIL_GLOBAL_LOCK)
#include <pthread.h>
static pthread_mutex_t global_mutex = PTHREAD_MUTEX_INITIALIZER;
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_grab_lock(void)
{
pthread_mutex_lock(&global_mutex);
}
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_release_lock(void)
{
pthread_mutex_unlock(&global_mutex);
}
#endif /* SLJIT_UTIL_GLOBAL_LOCK */
#endif /* _WIN32 */
/* ------------------------------------------------------------------------ */
/* Stack */
/* ------------------------------------------------------------------------ */
#if (defined SLJIT_UTIL_STACK && SLJIT_UTIL_STACK) || (defined SLJIT_EXECUTABLE_ALLOCATOR && SLJIT_EXECUTABLE_ALLOCATOR)
#ifdef _WIN32
#include "windows.h"
#else
/* Provides mmap function. */
#include <sys/mman.h>
/* For detecting the page size. */
#include <unistd.h>
#ifndef MAP_ANON
#include <fcntl.h>
/* Some old systems does not have MAP_ANON. */
static sljit_si dev_zero = -1;
#if (defined SLJIT_SINGLE_THREADED && SLJIT_SINGLE_THREADED)
static SLJIT_INLINE sljit_si open_dev_zero(void)
{
dev_zero = open("/dev/zero", O_RDWR);
return dev_zero < 0;
}
#else /* SLJIT_SINGLE_THREADED */
#include <pthread.h>
static pthread_mutex_t dev_zero_mutex = PTHREAD_MUTEX_INITIALIZER;
static SLJIT_INLINE sljit_si open_dev_zero(void)
{
pthread_mutex_lock(&dev_zero_mutex);
dev_zero = open("/dev/zero", O_RDWR);
pthread_mutex_unlock(&dev_zero_mutex);
return dev_zero < 0;
}
#endif /* SLJIT_SINGLE_THREADED */
#endif
#endif
#endif /* SLJIT_UTIL_STACK || SLJIT_EXECUTABLE_ALLOCATOR */
#if (defined SLJIT_UTIL_STACK && SLJIT_UTIL_STACK)
/* Planning to make it even more clever in the future. */
static sljit_sw sljit_page_align = 0;
SLJIT_API_FUNC_ATTRIBUTE struct sljit_stack* SLJIT_CALL sljit_allocate_stack(sljit_uw limit, sljit_uw max_limit)
{
struct sljit_stack *stack;
union {
void *ptr;
sljit_uw uw;
} base;
#ifdef _WIN32
SYSTEM_INFO si;
#endif
if (limit > max_limit || limit < 1)
return NULL;
#ifdef _WIN32
if (!sljit_page_align) {
GetSystemInfo(&si);
sljit_page_align = si.dwPageSize - 1;
}
#else
if (!sljit_page_align) {
sljit_page_align = sysconf(_SC_PAGESIZE);
/* Should never happen. */
if (sljit_page_align < 0)
sljit_page_align = 4096;
sljit_page_align--;
}
#endif
/* Align limit and max_limit. */
max_limit = (max_limit + sljit_page_align) & ~sljit_page_align;
stack = (struct sljit_stack*)SLJIT_MALLOC(sizeof(struct sljit_stack));
if (!stack)
return NULL;
#ifdef _WIN32
base.ptr = VirtualAlloc(NULL, max_limit, MEM_RESERVE, PAGE_READWRITE);
if (!base.ptr) {
SLJIT_FREE(stack);
return NULL;
}
stack->base = base.uw;
stack->limit = stack->base;
stack->max_limit = stack->base + max_limit;
if (sljit_stack_resize(stack, stack->base + limit)) {
sljit_free_stack(stack);
return NULL;
}
#else
#ifdef MAP_ANON
base.ptr = mmap(NULL, max_limit, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0);
#else
if (dev_zero < 0) {
if (open_dev_zero()) {
SLJIT_FREE(stack);
return NULL;
}
}
base.ptr = mmap(NULL, max_limit, PROT_READ | PROT_WRITE, MAP_PRIVATE, dev_zero, 0);
#endif
if (base.ptr == MAP_FAILED) {
SLJIT_FREE(stack);
return NULL;
}
stack->base = base.uw;
stack->limit = stack->base + limit;
stack->max_limit = stack->base + max_limit;
#endif
stack->top = stack->base;
return stack;
}
#undef PAGE_ALIGN
SLJIT_API_FUNC_ATTRIBUTE void SLJIT_CALL sljit_free_stack(struct sljit_stack* stack)
{
#ifdef _WIN32
VirtualFree((void*)stack->base, 0, MEM_RELEASE);
#else
munmap((void*)stack->base, stack->max_limit - stack->base);
#endif
SLJIT_FREE(stack);
}
SLJIT_API_FUNC_ATTRIBUTE sljit_sw SLJIT_CALL sljit_stack_resize(struct sljit_stack* stack, sljit_uw new_limit)
{
sljit_uw aligned_old_limit;
sljit_uw aligned_new_limit;
if ((new_limit > stack->max_limit) || (new_limit < stack->base))
return -1;
#ifdef _WIN32
aligned_new_limit = (new_limit + sljit_page_align) & ~sljit_page_align;
aligned_old_limit = (stack->limit + sljit_page_align) & ~sljit_page_align;
if (aligned_new_limit != aligned_old_limit) {
if (aligned_new_limit > aligned_old_limit) {
if (!VirtualAlloc((void*)aligned_old_limit, aligned_new_limit - aligned_old_limit, MEM_COMMIT, PAGE_READWRITE))
return -1;
}
else {
if (!VirtualFree((void*)aligned_new_limit, aligned_old_limit - aligned_new_limit, MEM_DECOMMIT))
return -1;
}
}
stack->limit = new_limit;
return 0;
#else
if (new_limit >= stack->limit) {
stack->limit = new_limit;
return 0;
}
aligned_new_limit = (new_limit + sljit_page_align) & ~sljit_page_align;
aligned_old_limit = (stack->limit + sljit_page_align) & ~sljit_page_align;
/* If madvise is available, we release the unnecessary space. */
#if defined(MADV_DONTNEED)
if (aligned_new_limit < aligned_old_limit)
madvise((void*)aligned_new_limit, aligned_old_limit - aligned_new_limit, MADV_DONTNEED);
#elif defined(POSIX_MADV_DONTNEED)
if (aligned_new_limit < aligned_old_limit)
posix_madvise((void*)aligned_new_limit, aligned_old_limit - aligned_new_limit, POSIX_MADV_DONTNEED);
#endif
stack->limit = new_limit;
return 0;
#endif
}
#endif /* SLJIT_UTIL_STACK */
#endif

2
ext/pcre/pcrelib/testdata/greppatN4 vendored
View file

@ -1,2 +1,2 @@
xxx
xxx
jkl

BIN
ext/pcre/pcrelib/testdata/saved16BE-1 vendored
View file

Binary file not shown.

BIN
ext/pcre/pcrelib/testdata/saved16LE-1 vendored
View file

Binary file not shown.

BIN
ext/pcre/pcrelib/testdata/saved32BE-1 vendored
View file

Binary file not shown.

BIN
ext/pcre/pcrelib/testdata/saved32LE-1 vendored
View file

Binary file not shown.

2
ext/pcre/pcrelib/testdata/testinput18 vendored
View file

@ -207,7 +207,7 @@ correctly, but that messes up comparisons). --/
CDBABC
\x{2000}ABC
/\R*A/SI8
/\R*A/SI8<bsr_unicode>
CDBABC
\x{2028}A

17
ext/pcre/pcrelib/testdata/testinput2 vendored
View file

@ -907,6 +907,9 @@
/\U/I
/a{1,3}b/U
ab
/[/I
/[a-/I
@ -4045,4 +4048,18 @@ backtracking verbs. --/
/[a[:<:]] should give error/
/(?=ab\K)/+
abcd
/abcd/f<lf>
xx\nxabcd
/ -- Test stack check external calls --/
/(((((a)))))/Q0
/(((((a)))))/Q1
/(((((a)))))/Q
/-- End of testinput2 --/

2
ext/pcre/pcrelib/testdata/testinput25 vendored
View file

@ -1,6 +1,6 @@
/-- Tests for the 32-bit library only */
< forbid 8w
< forbid 8W
/-- Check maximum character size --/

11
ext/pcre/pcrelib/testdata/testinput3 vendored
View file

@ -1,7 +1,10 @@
/-- This set of tests checks local-specific features, using the fr_FR locale.
It is not Perl-compatible. There is different version called wintestinput3
f or use on Windows, where the locale is called "french". --/
/-- This set of tests checks local-specific features, using the "fr_FR" locale.
It is not Perl-compatible. When run via RunTest, the locale is edited to
be whichever of "fr_FR", "french", or "fr" is found to exist. There is
different version of this file called wintestinput3 for use on Windows,
where the locale is called "french" and the tests are run using
RunTest.bat. --/
< forbid 8W
/^[\w]+/

6
ext/pcre/pcrelib/testdata/testinput4 vendored
View file

@ -716,4 +716,10 @@
/^a+[a\x{200}]/8
aa
/^.\B.\B./8
\x{10123}\x{10124}\x{10125}
/^#[^\x{ffff}]#[^\x{ffff}]#[^\x{ffff}]#/8
#\x{10000}#\x{100}#\x{10ffff}#
/-- End of testinput4 --/

2
ext/pcre/pcrelib/testdata/testinput5 vendored
View file

@ -788,4 +788,6 @@
/^a+[a\x{200}]/8BZ
aa
/[b-d\x{200}-\x{250}]*[ae-h]?#[\x{200}-\x{250}]{0,8}[\x00-\xff]*#[\x{200}-\x{250}]+[a-z]/8BZ
/-- End of testinput5 --/

9
ext/pcre/pcrelib/testdata/testinput6 vendored
View file

@ -1484,4 +1484,13 @@
\x{a1}\x{a7}
\x{37e}
/[RST]+/8iW
Ss\x{17f}
/[R-T]+/8iW
Ss\x{17f}
/[q-u]+/8iW
Ss\x{17f}
/-- End of testinput6 --/

6
ext/pcre/pcrelib/testdata/testinput7 vendored
View file

@ -829,4 +829,10 @@ of case for anything other than the ASCII letters. --/
/\d+\s{0,5}=\s*\S?=\w{0,4}\W*/8WBZ
/[RST]+/8iWBZ
/[R-T]+/8iWBZ
/[Q-U]+/8iWBZ
/-- End of testinput7 --/

8
ext/pcre/pcrelib/testdata/testoutput12 vendored
View file

@ -8,7 +8,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 3
No set of starting bytes
No starting char list
JIT study was successful
/(?(?C1)(?=a)a)/S+I
@ -27,7 +27,7 @@ No options
No first char
No need char
Subject length lower bound = -1
No set of starting bytes
No starting char list
JIT study was not successful
/abc/S+I>testsavedregex
@ -36,7 +36,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 3
No set of starting bytes
No starting char list
JIT study was successful
Compiled pattern written to testsavedregex
Study data written to testsavedregex
@ -165,7 +165,7 @@ No options
First char = 'a'
Need char = 'd'
Subject length lower bound = 4
No set of starting bytes
No starting char list
JIT study was successful
/(*NO_START_OPT)a(*:m)b/KS++

2
ext/pcre/pcrelib/testdata/testoutput13 vendored
View file

@ -8,7 +8,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 3
No set of starting bytes
No starting char list
JIT support is not available in this version of PCRE
/a*/SI

12
ext/pcre/pcrelib/testdata/testoutput14 vendored
View file

@ -361,7 +361,7 @@ Options: extended
No first char
No need char
Subject length lower bound = 3
Starting byte set: \x09 \x20 ! " # $ % & ' ( * + - / 0 1 2 3 4 5 6 7 8
Starting chars: \x09 \x20 ! " # $ % & ' ( * + - / 0 1 2 3 4 5 6 7 8
9 = ? A B C D E F G H I J K L M N O P Q R S T U V W X Y Z ^ _ ` a b c d e
f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f
@ -388,7 +388,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xa0
Starting chars: \x09 \x20 \xa0
/\H/SI
Capturing subpattern count = 0
@ -396,7 +396,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\v/SI
Capturing subpattern count = 0
@ -404,7 +404,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85
Starting chars: \x0a \x0b \x0c \x0d \x85
/\V/SI
Capturing subpattern count = 0
@ -412,7 +412,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\R/SI
Capturing subpattern count = 0
@ -420,7 +420,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85
Starting chars: \x0a \x0b \x0c \x0d \x85
/[\h]/BZ
------------------------------------------------------------------

36
ext/pcre/pcrelib/testdata/testoutput15 vendored
View file

@ -481,7 +481,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
\x0b \x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19
\x1a \x1b \x1c \x1d \x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4
5 6 7 8 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y
@ -519,7 +519,7 @@ Options: utf
First char = \x{c4}
Need char = \x{80}
Subject length lower bound = 3
No set of starting bytes
No starting char list
\x{100}\x{100}\x{100}\x{100\x{100}
0: \x{100}\x{100}\x{100}
@ -539,7 +539,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xc4
Starting chars: x \xc4
/(\x{100}*a|x)/8SDZ
------------------------------------------------------------------
@ -558,7 +558,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xc4
Starting chars: a x \xc4
/(\x{100}{0,2}a|x)/8SDZ
------------------------------------------------------------------
@ -577,7 +577,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xc4
Starting chars: a x \xc4
/(\x{100}{1,2}a|x)/8SDZ
------------------------------------------------------------------
@ -597,7 +597,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xc4
Starting chars: x \xc4
/\x{100}/8DZ
------------------------------------------------------------------
@ -799,7 +799,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xc2 \xe1 \xe2 \xe3
Starting chars: \x09 \x20 \xc2 \xe1 \xe2 \xe3
ABC\x{09}
0: \x{09}
ABC\x{20}
@ -825,7 +825,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \xc2 \xe2
Starting chars: \x0a \x0b \x0c \x0d \xc2 \xe2
ABC\x{0a}
0: \x{0a}
ABC\x{0b}
@ -845,7 +845,7 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 1
Starting byte set: \x09 \x20 A \xc2 \xe1 \xe2 \xe3
Starting chars: \x09 \x20 A \xc2 \xe1 \xe2 \xe3
CDBABC
0: A
@ -855,7 +855,7 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 2
Starting byte set: \x0a \x0b \x0c \x0d \xc2 \xe2
Starting chars: \x0a \x0b \x0c \x0d \xc2 \xe2
/\s?xxx\s/8SI
Capturing subpattern count = 0
@ -863,7 +863,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 4
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20 x
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20 x
/\sxxx\s/I8ST1
Capturing subpattern count = 0
@ -871,7 +871,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 5
Starting byte set: \x09 \x0a \x0c \x0d \x20 \xc2
Starting chars: \x09 \x0a \x0c \x0d \x20 \xc2
AB\x{85}xxx\x{a0}XYZ
0: \x{85}xxx\x{a0}
AB\x{a0}xxx\x{85}XYZ
@ -883,7 +883,7 @@ Options: utf
No first char
Need char = ' '
Subject length lower bound = 3
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
\x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a \x1b \x1c \x1d
\x1e \x1f ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e
@ -917,7 +917,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \xe1
Starting chars: \xe1
/\x{1234}+?/iS8I
Capturing subpattern count = 0
@ -925,7 +925,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \xe1
Starting chars: \xe1
/\x{1234}++/iS8I
Capturing subpattern count = 0
@ -933,7 +933,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \xe1
Starting chars: \xe1
/\x{1234}{2}/iS8I
Capturing subpattern count = 0
@ -941,7 +941,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 2
Starting byte set: \xe1
Starting chars: \xe1
/[^\x{c4}]/8DZ
------------------------------------------------------------------
@ -974,7 +974,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \xc2 \xe2
Starting chars: \x0a \x0b \x0c \x0d \xc2 \xe2
/\777/8DZ
------------------------------------------------------------------

8
ext/pcre/pcrelib/testdata/testoutput16 vendored
View file

@ -64,7 +64,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 17
Starting byte set: \xd0 \xd1
Starting chars: \xd0 \xd1
\x{401}\x{420}\x{421}\x{422}\x{423}\x{424}\x{425}\x{426}\x{427}\x{428}\x{429}\x{42a}\x{42b}\x{42c}\x{42d}\x{42e}\x{42f}
0: \x{401}\x{420}\x{421}\x{422}\x{423}\x{424}\x{425}\x{426}\x{427}\x{428}\x{429}\x{42a}\x{42b}\x{42c}\x{42d}\x{42e}\x{42f}
\x{451}\x{440}\x{441}\x{442}\x{443}\x{444}\x{445}\x{446}\x{447}\x{448}\x{449}\x{44a}\x{44b}\x{44c}\x{44d}\x{44e}\x{44f}
@ -92,7 +92,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xa0
Starting chars: \x09 \x20 \xa0
/\v/SI
Capturing subpattern count = 0
@ -100,7 +100,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85
Starting chars: \x0a \x0b \x0c \x0d \x85
/\R/SI
Capturing subpattern count = 0
@ -108,7 +108,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85
Starting chars: \x0a \x0b \x0c \x0d \x85
/[[:blank:]]/WBZ
------------------------------------------------------------------

28
ext/pcre/pcrelib/testdata/testoutput17 vendored
View file

@ -228,7 +228,7 @@ Options: extended
No first char
No need char
Subject length lower bound = 3
Starting byte set: \x09 \x20 ! " # $ % & ' ( * + - / 0 1 2 3 4 5 6 7 8
Starting chars: \x09 \x20 ! " # $ % & ' ( * + - / 0 1 2 3 4 5 6 7 8
9 = ? A B C D E F G H I J K L M N O P Q R S T U V W X Y Z ^ _ ` a b c d e
f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f \xff
@ -274,7 +274,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xa0 \xff
Starting chars: \x09 \x20 \xa0 \xff
\x{1681}\x{200b}\x{1680}\x{2000}\x{202f}\x{3000}
0: \x{1680}\x{2000}\x{202f}\x{3000}
\x{3001}\x{2fff}\x{200a}\xa0\x{2000}
@ -292,7 +292,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x09 \x20 \xa0 \xff
\x{1681}\x{200b}\x{1680}\x{2000}\x{202f}\x{3000}
0: \x{1680}\x{2000}\x{202f}\x{3000}
\x{3001}\x{2fff}\x{200a}\xa0\x{2000}
@ -304,7 +304,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
\x{1680}\x{180e}\x{167f}\x{1681}\x{180d}\x{180f}
0: \x{167f}\x{1681}\x{180d}\x{180f}
\x{2000}\x{200a}\x{1fff}\x{200b}
@ -330,7 +330,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
\x{2027}\x{2030}\x{2028}\x{2029}
0: \x{2028}\x{2029}
\x09\x0e\x84\x86\x85\x0a\x0b\x0c\x0d
@ -348,7 +348,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
\x{2027}\x{2030}\x{2028}\x{2029}
0: \x{2028}\x{2029}
\x09\x0e\x84\x86\x85\x0a\x0b\x0c\x0d
@ -360,7 +360,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
\x{2028}\x{2029}\x{2027}\x{2030}
0: \x{2027}\x{2030}
\x85\x0a\x0b\x0c\x0d\x09\x0e\x84\x86
@ -378,7 +378,7 @@ Options: bsr_unicode
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
\x{2027}\x{2030}\x{2028}\x{2029}
0: \x{2028}\x{2029}
\x09\x0e\x84\x86\x85\x0a\x0b\x0c\x0d
@ -534,18 +534,18 @@ MK: 0123456789ABCDEF0123456789ABCDEF0123456789ABCDEF0123456789ABCDEF0123456789AB
------------------------------------------------------------------
Bra
a*
[b-\x{200}]?+
[b-\xff\x{100}-\x{200}]?+
a#
a*+
[b-\x{200}]?
[b-\xff\x{100}-\x{200}]?
b#
[a-f]*
[g-\x{200}]*+
[a-f]*+
[g-\xff\x{100}-\x{200}]*+
#
[g-\x{200}]*
[g-\xff\x{100}-\x{200}]*+
[a-c]*+
#
[g-\x{200}]*
[g-\xff\x{100}-\x{200}]*
[a-h]*+
Ket
End

42
ext/pcre/pcrelib/testdata/testoutput18-16 vendored
View file

@ -339,7 +339,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
\x0b \x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19
\x1a \x1b \x1c \x1d \x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4
5 6 7 8 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y
@ -378,7 +378,7 @@ Options: utf
First char = \x{100}
Need char = \x{100}
Subject length lower bound = 3
No set of starting bytes
No starting char list
\x{100}\x{100}\x{100}\x{100\x{100}
0: \x{100}\x{100}\x{100}
@ -398,7 +398,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xff
Starting chars: x \xff
/(\x{100}*a|x)/8SDZ
------------------------------------------------------------------
@ -417,7 +417,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xff
Starting chars: a x \xff
/(\x{100}{0,2}a|x)/8SDZ
------------------------------------------------------------------
@ -436,7 +436,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xff
Starting chars: a x \xff
/(\x{100}{1,2}a|x)/8SDZ
------------------------------------------------------------------
@ -456,7 +456,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xff
Starting chars: x \xff
/\x{100}/8DZ
------------------------------------------------------------------
@ -666,7 +666,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xa0 \xff
Starting chars: \x09 \x20 \xa0 \xff
ABC\x{09}
0: \x{09}
ABC\x{20}
@ -692,7 +692,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
ABC\x{0a}
0: \x{0a}
ABC\x{0b}
@ -712,19 +712,19 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 1
Starting byte set: \x09 \x20 A \xa0 \xff
Starting chars: \x09 \x20 A \xa0 \xff
CDBABC
0: A
\x{2000}ABC
0: \x{2000}A
/\R*A/SI8
/\R*A/SI8<bsr_unicode>
Capturing subpattern count = 0
Options: utf
Options: bsr_unicode utf
No first char
Need char = 'A'
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d A \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d A \x85 \xff
CDBABC
0: A
\x{2028}A
@ -736,7 +736,7 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 2
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
/\s?xxx\s/8SI
Capturing subpattern count = 0
@ -744,7 +744,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 4
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20 x
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20 x
/\sxxx\s/I8ST1
Capturing subpattern count = 0
@ -752,7 +752,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 5
Starting byte set: \x09 \x0a \x0c \x0d \x20 \x85 \xa0
Starting chars: \x09 \x0a \x0c \x0d \x20 \x85 \xa0
AB\x{85}xxx\x{a0}XYZ
0: \x{85}xxx\x{a0}
AB\x{a0}xxx\x{85}XYZ
@ -764,7 +764,7 @@ Options: utf
No first char
Need char = ' '
Subject length lower bound = 3
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
\x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a \x1b \x1c \x1d
\x1e \x1f ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e
@ -803,7 +803,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}+?/iS8I
Capturing subpattern count = 0
@ -811,7 +811,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}++/iS8I
Capturing subpattern count = 0
@ -819,7 +819,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}{2}/iS8I
Capturing subpattern count = 0
@ -827,7 +827,7 @@ Options: caseless utf
First char = \x{1234}
Need char = \x{1234}
Subject length lower bound = 2
No set of starting bytes
No starting char list
/[^\x{c4}]/8DZ
------------------------------------------------------------------
@ -860,7 +860,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
/-- Check bad offset --/

42
ext/pcre/pcrelib/testdata/testoutput18-32 vendored
View file

@ -337,7 +337,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
\x0b \x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19
\x1a \x1b \x1c \x1d \x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4
5 6 7 8 9 : ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y
@ -376,7 +376,7 @@ Options: utf
First char = \x{100}
Need char = \x{100}
Subject length lower bound = 3
No set of starting bytes
No starting char list
\x{100}\x{100}\x{100}\x{100\x{100}
0: \x{100}\x{100}\x{100}
@ -396,7 +396,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xff
Starting chars: x \xff
/(\x{100}*a|x)/8SDZ
------------------------------------------------------------------
@ -415,7 +415,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xff
Starting chars: a x \xff
/(\x{100}{0,2}a|x)/8SDZ
------------------------------------------------------------------
@ -434,7 +434,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: a x \xff
Starting chars: a x \xff
/(\x{100}{1,2}a|x)/8SDZ
------------------------------------------------------------------
@ -454,7 +454,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: x \xff
Starting chars: x \xff
/\x{100}/8DZ
------------------------------------------------------------------
@ -663,7 +663,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x20 \xa0 \xff
Starting chars: \x09 \x20 \xa0 \xff
ABC\x{09}
0: \x{09}
ABC\x{20}
@ -689,7 +689,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
ABC\x{0a}
0: \x{0a}
ABC\x{0b}
@ -709,19 +709,19 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 1
Starting byte set: \x09 \x20 A \xa0 \xff
Starting chars: \x09 \x20 A \xa0 \xff
CDBABC
0: A
\x{2000}ABC
0: \x{2000}A
/\R*A/SI8
/\R*A/SI8<bsr_unicode>
Capturing subpattern count = 0
Options: utf
Options: bsr_unicode utf
No first char
Need char = 'A'
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d A \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d A \x85 \xff
CDBABC
0: A
\x{2028}A
@ -733,7 +733,7 @@ Options: utf
No first char
Need char = 'A'
Subject length lower bound = 2
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
/\s?xxx\s/8SI
Capturing subpattern count = 0
@ -741,7 +741,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 4
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20 x
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20 x
/\sxxx\s/I8ST1
Capturing subpattern count = 0
@ -749,7 +749,7 @@ Options: utf
No first char
Need char = 'x'
Subject length lower bound = 5
Starting byte set: \x09 \x0a \x0c \x0d \x20 \x85 \xa0
Starting chars: \x09 \x0a \x0c \x0d \x20 \x85 \xa0
AB\x{85}xxx\x{a0}XYZ
0: \x{85}xxx\x{a0}
AB\x{a0}xxx\x{85}XYZ
@ -761,7 +761,7 @@ Options: utf
No first char
Need char = ' '
Subject length lower bound = 3
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0b \x0e
\x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a \x1b \x1c \x1d
\x1e \x1f ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = > ? @
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e
@ -800,7 +800,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}+?/iS8I
Capturing subpattern count = 0
@ -808,7 +808,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}++/iS8I
Capturing subpattern count = 0
@ -816,7 +816,7 @@ Options: caseless utf
First char = \x{1234}
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/\x{1234}{2}/iS8I
Capturing subpattern count = 0
@ -824,7 +824,7 @@ Options: caseless utf
First char = \x{1234}
Need char = \x{1234}
Subject length lower bound = 2
No set of starting bytes
No starting char list
/[^\x{c4}]/8DZ
------------------------------------------------------------------
@ -857,7 +857,7 @@ Options: utf
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x0a \x0b \x0c \x0d \x85 \xff
Starting chars: \x0a \x0b \x0c \x0d \x85 \xff
/-- Check bad offset --/

2
ext/pcre/pcrelib/testdata/testoutput19 vendored
View file

@ -55,7 +55,7 @@ Options: caseless utf
First char = \x{401} (caseless)
Need char = \x{42f} (caseless)
Subject length lower bound = 17
No set of starting bytes
No starting char list
\x{401}\x{420}\x{421}\x{422}\x{423}\x{424}\x{425}\x{426}\x{427}\x{428}\x{429}\x{42a}\x{42b}\x{42c}\x{42d}\x{42e}\x{42f}
0: \x{401}\x{420}\x{421}\x{422}\x{423}\x{424}\x{425}\x{426}\x{427}\x{428}\x{429}\x{42a}\x{42b}\x{42c}\x{42d}\x{42e}\x{42f}
\x{451}\x{440}\x{441}\x{442}\x{443}\x{444}\x{445}\x{446}\x{447}\x{448}\x{449}\x{44a}\x{44b}\x{44c}\x{44d}\x{44e}\x{44f}

192
ext/pcre/pcrelib/testdata/testoutput2 vendored
View file

@ -178,7 +178,7 @@ No options
No first char
No need char
Subject length lower bound = 3
Starting byte set: c d e
Starting chars: c d e
this sentence eventually mentions a cat
0: cat
this sentences rambles on and on for a while and then reaches elephant
@ -190,7 +190,7 @@ Options: caseless
No first char
No need char
Subject length lower bound = 3
Starting byte set: C D E c d e
Starting chars: C D E c d e
this sentence eventually mentions a CAT cat
0: CAT
this sentences rambles on and on for a while to elephant ElePhant
@ -202,7 +202,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/(a|[^\dZ])/IS
Capturing subpattern count = 1
@ -210,7 +210,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0a
\x0b \x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19
\x1a \x1b \x1c \x1d \x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / : ; < = >
? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y [ \ ] ^ _ ` a b c d
@ -231,7 +231,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20 a b
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20 a b
/(ab\2)/
Failed: reference to non-existent subpattern at offset 6
@ -512,7 +512,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/(?i)[abcd]/IS
Capturing subpattern count = 0
@ -520,7 +520,7 @@ Options: caseless
No first char
No need char
Subject length lower bound = 1
Starting byte set: A B C D a b c d
Starting chars: A B C D a b c d
/(?m)[xy]|(b|c)/IS
Capturing subpattern count = 1
@ -528,7 +528,7 @@ Options: multiline
No first char
No need char
Subject length lower bound = 1
Starting byte set: b c x y
Starting chars: b c x y
/(^a|^b)/Im
Capturing subpattern count = 1
@ -591,7 +591,7 @@ No options
First char = 'b' (caseless)
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/(a*b|(?i:c*(?-i)d))/IS
Capturing subpattern count = 1
@ -599,7 +599,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: C a b c d
Starting chars: C a b c d
/a$/I
Capturing subpattern count = 0
@ -666,7 +666,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
/(?<!foo)(alpha|omega)/IS
Capturing subpattern count = 1
@ -675,7 +675,7 @@ No options
No first char
Need char = 'a'
Subject length lower bound = 5
Starting byte set: a o
Starting chars: a o
/(?!alphabet)[ab]/IS
Capturing subpattern count = 0
@ -683,7 +683,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
/(?<=foo\n)^bar/Im
Capturing subpattern count = 0
@ -1642,7 +1642,7 @@ Options: anchored
No first char
Need char = 'd'
Subject length lower bound = 4
No set of starting bytes
No starting char list
/\( # ( at start
(?: # Non-capturing bracket
@ -1875,7 +1875,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Starting chars: A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
_ a b c d e f g h i j k l m n o p q r s t u v w x y z
/^[[:ascii:]]/DZ
@ -1937,7 +1937,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20
/^[[:cntrl:]]/DZ
------------------------------------------------------------------
@ -3178,6 +3178,10 @@ Failed: PCRE does not support \L, \l, \N{name}, \U, or \u at offset 1
/\U/I
Failed: PCRE does not support \L, \l, \N{name}, \U, or \u at offset 1
/a{1,3}b/U
ab
0: ab
/[/I
Failed: missing terminating ] for character class at offset 1
@ -3434,7 +3438,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
/[^a]/I
Capturing subpattern count = 0
@ -3454,7 +3458,7 @@ No options
No first char
Need char = '6'
Subject length lower bound = 4
Starting byte set: 0 1 2 3 4 5 6 7 8 9
Starting chars: 0 1 2 3 4 5 6 7 8 9
/a^b/I
Capturing subpattern count = 0
@ -3488,7 +3492,7 @@ Options: caseless
No first char
No need char
Subject length lower bound = 1
Starting byte set: A B a b
Starting chars: A B a b
/[ab](?i)cd/IS
Capturing subpattern count = 0
@ -3496,7 +3500,7 @@ No options
No first char
Need char = 'd' (caseless)
Subject length lower bound = 3
Starting byte set: a b
Starting chars: a b
/abc(?C)def/I
Capturing subpattern count = 0
@ -3537,7 +3541,7 @@ No options
No first char
Need char = 'f'
Subject length lower bound = 7
Starting byte set: 0 1 2 3 4 5 6 7 8 9
Starting chars: 0 1 2 3 4 5 6 7 8 9
1234abcdef
--->1234abcdef
1 ^ \d
@ -3856,7 +3860,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
/(?R)/I
Failed: recursive call could loop indefinitely at offset 3
@ -4637,7 +4641,7 @@ Options: caseless
No first char
Need char = 'g' (caseless)
Subject length lower bound = 8
No set of starting bytes
No starting char list
Baby Bjorn Active Carrier - With free SHIPPING!!
0: Baby Bjorn Active Carrier - With free SHIPPING!!
1: Baby Bjorn Active Carrier - With free SHIPPING!!
@ -4656,7 +4660,7 @@ No options
No first char
Need char = 'b'
Subject length lower bound = 1
No set of starting bytes
No starting char list
/(a|b)*.?c/ISDZ
------------------------------------------------------------------
@ -4677,7 +4681,7 @@ No options
No first char
Need char = 'c'
Subject length lower bound = 1
No set of starting bytes
No starting char list
/abc(?C255)de(?C)f/DZ
------------------------------------------------------------------
@ -4750,7 +4754,7 @@ Options:
No first char
Need char = 'b'
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
ab
--->ab
+0 ^ a*
@ -4893,7 +4897,7 @@ Options:
No first char
Need char = 'x'
Subject length lower bound = 4
Starting byte set: a d
Starting chars: a d
abcx
--->abcx
+0 ^ (abc|def)
@ -5127,7 +5131,7 @@ Options:
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b x
Starting chars: a b x
Note: that { does NOT introduce a quantifier
--->Note: that { does NOT introduce a quantifier
+0 ^ ([ab]{,4}c|xy)
@ -5607,7 +5611,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 3
No set of starting bytes
No starting char list
Compiled pattern written to testsavedregex
Study data written to testsavedregex
<testsavedregex
@ -5642,7 +5646,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 3
No set of starting bytes
No starting char list
Compiled pattern written to testsavedregex
Study data written to testsavedregex
<testsavedregex
@ -5677,7 +5681,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
Compiled pattern written to testsavedregex
Study data written to testsavedregex
<testsavedregex
@ -5716,7 +5720,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b
Starting chars: a b
Compiled pattern written to testsavedregex
Study data written to testsavedregex
<testsavedregex
@ -6431,7 +6435,7 @@ No options
No first char
Need char = ','
Subject length lower bound = 1
Starting byte set: \x09 \x0a \x0b \x0c \x0d \x20 ,
Starting chars: \x09 \x0a \x0b \x0c \x0d \x20 ,
\x0b,\x0b
0: \x0b,\x0b
\x0c,\x0d
@ -6738,7 +6742,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: C a b c d
Starting chars: C a b c d
/()[ab]xyz/IS
Capturing subpattern count = 1
@ -6746,7 +6750,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b
Starting chars: a b
/(|)[ab]xyz/IS
Capturing subpattern count = 1
@ -6754,7 +6758,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b
Starting chars: a b
/(|c)[ab]xyz/IS
Capturing subpattern count = 1
@ -6762,7 +6766,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b c
Starting chars: a b c
/(|c?)[ab]xyz/IS
Capturing subpattern count = 1
@ -6770,7 +6774,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b c
Starting chars: a b c
/(d?|c?)[ab]xyz/IS
Capturing subpattern count = 1
@ -6778,7 +6782,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b c d
Starting chars: a b c d
/(d?|c)[ab]xyz/IS
Capturing subpattern count = 1
@ -6786,7 +6790,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 4
Starting byte set: a b c d
Starting chars: a b c d
/^a*b\d/DZ
------------------------------------------------------------------
@ -6879,7 +6883,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/(a+|b*)[cd]/IS
Capturing subpattern count = 1
@ -6887,7 +6891,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/(a*|b+)[cd]/IS
Capturing subpattern count = 1
@ -6895,7 +6899,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/(a+|b+)[cd]/IS
Capturing subpattern count = 1
@ -6903,7 +6907,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((
((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((
@ -9307,7 +9311,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: x y z
Starting chars: x y z
/(?(?=.*b)b|^)/CI
Capturing subpattern count = 0
@ -10096,7 +10100,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/(a|bc)\1{2,3}/SI
Capturing subpattern count = 1
@ -10105,7 +10109,7 @@ No options
No first char
No need char
Subject length lower bound = 3
Starting byte set: a b
Starting chars: a b
/(a|bc)(?1)/SI
Capturing subpattern count = 1
@ -10113,7 +10117,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/(a|b\1)(a|b\1)/SI
Capturing subpattern count = 2
@ -10122,7 +10126,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/(a|b\1){2}/SI
Capturing subpattern count = 1
@ -10131,7 +10135,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/(a|bbbb\1)(a|bbbb\1)/SI
Capturing subpattern count = 2
@ -10140,7 +10144,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/(a|bbbb\1){2}/SI
Capturing subpattern count = 1
@ -10149,7 +10153,7 @@ No options
No first char
No need char
Subject length lower bound = 2
Starting byte set: a b
Starting chars: a b
/^From +([^ ]+) +[a-zA-Z][a-zA-Z][a-zA-Z] +[a-zA-Z][a-zA-Z][a-zA-Z] +[0-9]?[0-9] +[0-9][0-9]:[0-9][0-9]/SI
Capturing subpattern count = 1
@ -10157,7 +10161,7 @@ Options: anchored
No first char
Need char = ':'
Subject length lower bound = 22
No set of starting bytes
No starting char list
/<tr([\w\W\s\d][^<>]{0,})><TD([\w\W\s\d][^<>]{0,})>([\d]{0,}\.)(.*)((<BR>([\w\W\s\d][^<>]{0,})|[\s]{0,}))<\/a><\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><TD([\w\W\s\d][^<>]{0,})>([\w\W\s\d][^<>]{0,})<\/TD><\/TR>/isIS
Capturing subpattern count = 11
@ -10165,7 +10169,7 @@ Options: caseless dotall
First char = '<'
Need char = '>'
Subject length lower bound = 47
No set of starting bytes
No starting char list
"(?>.*/)foo"SI
Capturing subpattern count = 0
@ -10173,7 +10177,7 @@ No options
No first char
Need char = 'o'
Subject length lower bound = 4
No set of starting bytes
No starting char list
/(?(?=[^a-z]+[a-z]) \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} ) /xSI
Capturing subpattern count = 0
@ -10181,7 +10185,7 @@ Options: extended
No first char
Need char = '-'
Subject length lower bound = 8
No set of starting bytes
No starting char list
/(?:(?:(?:(?:(?:(?:(?:(?:(?:(a|b|c))))))))))/iSI
Capturing subpattern count = 1
@ -10189,7 +10193,7 @@ Options: caseless
No first char
No need char
Subject length lower bound = 1
Starting byte set: A B C a b c
Starting chars: A B C a b c
/(?:c|d)(?:)(?:aaaaaaaa(?:)(?:bbbbbbbb)(?:bbbbbbbb(?:))(?:bbbbbbbb(?:)(?:bbbbbbbb)))/SI
Capturing subpattern count = 0
@ -10197,7 +10201,7 @@ No options
No first char
Need char = 'b'
Subject length lower bound = 41
Starting byte set: c d
Starting chars: c d
/<a[\s]+href[\s]*=[\s]* # find <a href=
([\"\'])? # find single or double quote
@ -10210,7 +10214,7 @@ Options: caseless extended dotall
First char = '<'
Need char = '='
Subject length lower bound = 9
No set of starting bytes
No starting char list
/^(?!:) # colon disallowed at start
(?: # start of item
@ -10226,7 +10230,7 @@ Options: anchored caseless extended
No first char
Need char = ':'
Subject length lower bound = 2
No set of starting bytes
No starting char list
/(?|(?<a>A)|(?<a>B))/I
Capturing subpattern count = 1
@ -10450,7 +10454,7 @@ Options:
No first char
Need char = 'a'
Subject length lower bound = 1
No set of starting bytes
No starting char list
cat
0: a
1:
@ -10464,7 +10468,7 @@ No options
No first char
Need char = 'a'
Subject length lower bound = 3
No set of starting bytes
No starting char list
cat
No match
@ -10476,7 +10480,7 @@ No options
First char = 'i'
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
i
0: i
@ -10486,7 +10490,7 @@ No options
No first char
Need char = 'i'
Subject length lower bound = 1
Starting byte set: i
Starting chars: i
ia
0: ia
1:
@ -11080,7 +11084,7 @@ No options
First char = 'a'
Need char = '4'
Subject length lower bound = 5
No set of starting bytes
No starting char list
/([abc])++1234/SI
Capturing subpattern count = 1
@ -11088,7 +11092,7 @@ No options
No first char
Need char = '4'
Subject length lower bound = 5
Starting byte set: a b c
Starting chars: a b c
/(?<=(abc)+)X/
Failed: lookbehind assertion is not fixed length at offset 10
@ -11369,7 +11373,7 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/(a(?2)|b)(b(?1)|a)(?:(?1)|(?2))/SI
Capturing subpattern count = 2
@ -11377,7 +11381,7 @@ No options
No first char
No need char
Subject length lower bound = 3
Starting byte set: a b
Starting chars: a b
/(a(?2)|b)(b(?1)|a)(?1)(?2)/SI
Capturing subpattern count = 2
@ -11385,7 +11389,7 @@ No options
No first char
No need char
Subject length lower bound = 4
Starting byte set: a b
Starting chars: a b
/(abc)(?1)/SI
Capturing subpattern count = 1
@ -11393,7 +11397,7 @@ No options
First char = 'a'
Need char = 'c'
Subject length lower bound = 6
No set of starting bytes
No starting char list
/^(?>a)++/
aa\M
@ -11711,7 +11715,7 @@ No options
First char = 't'
Need char = 't'
Subject length lower bound = 18
No set of starting bytes
No starting char list
/\btype\b\W*?\btext\b\W*?\bjavascript\b|\burl\b\W*?\bshell:|<input\b.*?\btype\b\W*?\bimage\b|\bonkeyup\b\W*?\=/IS
Capturing subpattern count = 0
@ -11720,7 +11724,7 @@ No options
No first char
No need char
Subject length lower bound = 8
Starting byte set: < o t u
Starting chars: < o t u
/a(*SKIP)c|b(*ACCEPT)|/+S!I
Capturing subpattern count = 0
@ -11729,7 +11733,7 @@ No options
No first char
No need char
Subject length lower bound = -1
No set of starting bytes
No starting char list
a
0:
0+
@ -11740,7 +11744,7 @@ No options
No first char
No need char
Subject length lower bound = -1
Starting byte set: a b x
Starting chars: a b x
ax
0: x
@ -12436,7 +12440,7 @@ No options
No first char
No need char
Subject length lower bound = -1
No set of starting bytes
No starting char list
/(?:(a)+(?C1)bb|aa(?C2)b)/
aab\C+
@ -12722,7 +12726,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 2
Starting byte set: a z
Starting chars: a z
aaaaaaaaaaaaaz
Error -21 (recursion limit exceeded)
aaaaaaaaaaaaaz\Q1000
@ -12735,7 +12739,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 2
Starting byte set: a z
Starting chars: a z
aaaaaaaaaaaaaz
Error -21 (recursion limit exceeded)
@ -12746,7 +12750,7 @@ No options
No first char
Need char = 'z'
Subject length lower bound = 2
Starting byte set: a z
Starting chars: a z
aaaaaaaaaaaaaz
No match
aaaaaaaaaaaaaz\Q10
@ -12790,7 +12794,7 @@ Options: dupnames
First char = 'a'
Need char = 'z'
Subject length lower bound = 5
No set of starting bytes
No starting char list
/a*[bcd]/BZ
------------------------------------------------------------------
@ -13902,7 +13906,7 @@ No options
No first char
Need char = 'd'
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/[a-c]+d/DZS
------------------------------------------------------------------
@ -13917,7 +13921,7 @@ No options
No first char
Need char = 'd'
Subject length lower bound = 2
Starting byte set: a b c
Starting chars: a b c
/[a-c]?d/DZS
------------------------------------------------------------------
@ -13932,7 +13936,7 @@ No options
No first char
Need char = 'd'
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/[a-c]{4,6}d/DZS
------------------------------------------------------------------
@ -13947,7 +13951,7 @@ No options
No first char
Need char = 'd'
Subject length lower bound = 5
Starting byte set: a b c
Starting chars: a b c
/[a-c]{0,6}d/DZS
------------------------------------------------------------------
@ -13962,7 +13966,7 @@ No options
No first char
Need char = 'd'
Subject length lower bound = 1
Starting byte set: a b c d
Starting chars: a b c d
/-- End of special auto-possessive tests --/
@ -14125,4 +14129,24 @@ No match
/[a[:<:]] should give error/
Failed: unknown POSIX class name at offset 4
/(?=ab\K)/+
abcd
Start of matched string is beyond its end - displaying from end to start.
0: ab
0+ abcd
/abcd/f<lf>
xx\nxabcd
No match
/ -- Test stack check external calls --/
/(((((a)))))/Q0
/(((((a)))))/Q1
Failed: parentheses are too deeply nested (stack check) at offset 0
/(((((a)))))/Q
** Missing 0 or 1 after /Q
/-- End of testinput2 --/

4
ext/pcre/pcrelib/testdata/testoutput21-16 vendored
View file

@ -50,7 +50,7 @@ Options: anchored extended
No first char
No need char
Subject length lower bound = 6
No set of starting bytes
No starting char list
<!testsaved16BE-1
Compiled pattern loaded from testsaved16BE-1
@ -83,7 +83,7 @@ Options: anchored extended
No first char
No need char
Subject length lower bound = 6
No set of starting bytes
No starting char list
<!testsaved32LE-1
Compiled pattern loaded from testsaved32LE-1

4
ext/pcre/pcrelib/testdata/testoutput21-32 vendored
View file

@ -62,7 +62,7 @@ Options: anchored extended
No first char
No need char
Subject length lower bound = 6
No set of starting bytes
No starting char list
<!testsaved32BE-1
Compiled pattern loaded from testsaved32BE-1
@ -95,6 +95,6 @@ Options: anchored extended
No first char
No need char
Subject length lower bound = 6
No set of starting bytes
No starting char list
/-- End of testinput21 --/

4
ext/pcre/pcrelib/testdata/testoutput22-16 vendored
View file

@ -37,7 +37,7 @@ Options: extended utf
No first char
No need char
Subject length lower bound = 2
No set of starting bytes
No starting char list
<!testsaved16BE-2
Compiled pattern loaded from testsaved16BE-2
@ -64,7 +64,7 @@ Options: extended utf
No first char
No need char
Subject length lower bound = 2
No set of starting bytes
No starting char list
<!testsaved32LE-2
Compiled pattern loaded from testsaved32LE-2

4
ext/pcre/pcrelib/testdata/testoutput22-32 vendored
View file

@ -49,7 +49,7 @@ Options: extended utf
No first char
No need char
Subject length lower bound = 2
No set of starting bytes
No starting char list
<!testsaved32BE-2
Compiled pattern loaded from testsaved32BE-2
@ -76,6 +76,6 @@ Options: extended utf
No first char
No need char
Subject length lower bound = 2
No set of starting bytes
No starting char list
/-- End of testinput22 --/

34
ext/pcre/pcrelib/testdata/testoutput23 vendored
View file

@ -18,7 +18,7 @@ Failed: character value in \x{} or \o{} is too large at offset 8
/[\H]/BZSI
------------------------------------------------------------------
Bra
[\x00-\x08\x0a-\x1f!-\x9f\x{a1}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{ffff}]
[\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{ffff}]
Ket
End
------------------------------------------------------------------
@ -27,12 +27,25 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0a \x0b
\x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a
\x1b \x1c \x1d \x1e \x1f ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9
: ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^
_ ` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f \x80
\x81 \x82 \x83 \x84 \x85 \x86 \x87 \x88 \x89 \x8a \x8b \x8c \x8d \x8e \x8f
\x90 \x91 \x92 \x93 \x94 \x95 \x96 \x97 \x98 \x99 \x9a \x9b \x9c \x9d \x9e
\x9f \xa1 \xa2 \xa3 \xa4 \xa5 \xa6 \xa7 \xa8 \xa9 \xaa \xab \xac \xad \xae
\xaf \xb0 \xb1 \xb2 \xb3 \xb4 \xb5 \xb6 \xb7 \xb8 \xb9 \xba \xbb \xbc \xbd
\xbe \xbf \xc0 \xc1 \xc2 \xc3 \xc4 \xc5 \xc6 \xc7 \xc8 \xc9 \xca \xcb \xcc
\xcd \xce \xcf \xd0 \xd1 \xd2 \xd3 \xd4 \xd5 \xd6 \xd7 \xd8 \xd9 \xda \xdb
\xdc \xdd \xde \xdf \xe0 \xe1 \xe2 \xe3 \xe4 \xe5 \xe6 \xe7 \xe8 \xe9 \xea
\xeb \xec \xed \xee \xef \xf0 \xf1 \xf2 \xf3 \xf4 \xf5 \xf6 \xf7 \xf8 \xf9
\xfa \xfb \xfc \xfd \xfe \xff
/[\V]/BZSI
------------------------------------------------------------------
Bra
[\x00-\x09\x0e-\x84\x{86}-\x{2027}\x{202a}-\x{ffff}]
[\x00-\x09\x0e-\x84\x86-\xff\x{100}-\x{2027}\x{202a}-\x{ffff}]
Ket
End
------------------------------------------------------------------
@ -41,6 +54,19 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0e
\x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a \x1b \x1c \x1d
\x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = >
? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c
d e f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f \x80 \x81 \x82
\x83 \x84 \x86 \x87 \x88 \x89 \x8a \x8b \x8c \x8d \x8e \x8f \x90 \x91 \x92
\x93 \x94 \x95 \x96 \x97 \x98 \x99 \x9a \x9b \x9c \x9d \x9e \x9f \xa0 \xa1
\xa2 \xa3 \xa4 \xa5 \xa6 \xa7 \xa8 \xa9 \xaa \xab \xac \xad \xae \xaf \xb0
\xb1 \xb2 \xb3 \xb4 \xb5 \xb6 \xb7 \xb8 \xb9 \xba \xbb \xbc \xbd \xbe \xbf
\xc0 \xc1 \xc2 \xc3 \xc4 \xc5 \xc6 \xc7 \xc8 \xc9 \xca \xcb \xcc \xcd \xce
\xcf \xd0 \xd1 \xd2 \xd3 \xd4 \xd5 \xd6 \xd7 \xd8 \xd9 \xda \xdb \xdc \xdd
\xde \xdf \xe0 \xe1 \xe2 \xe3 \xe4 \xe5 \xe6 \xe7 \xe8 \xe9 \xea \xeb \xec
\xed \xee \xef \xf0 \xf1 \xf2 \xf3 \xf4 \xf5 \xf6 \xf7 \xf8 \xf9 \xfa \xfb
\xfc \xfd \xfe \xff
/-- End of testinput23 --/

36
ext/pcre/pcrelib/testdata/testoutput25 vendored
View file

@ -1,6 +1,6 @@
/-- Tests for the 32-bit library only */
< forbid 8w
< forbid 8W
/-- Check maximum character size --/
@ -65,7 +65,7 @@ Need char = \x{800000}
/[\H]/BZSI
------------------------------------------------------------------
Bra
[\x00-\x08\x0a-\x1f!-\x9f\x{a1}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{ffffffff}]
[\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{ffffffff}]
Ket
End
------------------------------------------------------------------
@ -74,12 +74,25 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x0a \x0b
\x0c \x0d \x0e \x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a
\x1b \x1c \x1d \x1e \x1f ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9
: ; < = > ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^
_ ` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f \x80
\x81 \x82 \x83 \x84 \x85 \x86 \x87 \x88 \x89 \x8a \x8b \x8c \x8d \x8e \x8f
\x90 \x91 \x92 \x93 \x94 \x95 \x96 \x97 \x98 \x99 \x9a \x9b \x9c \x9d \x9e
\x9f \xa1 \xa2 \xa3 \xa4 \xa5 \xa6 \xa7 \xa8 \xa9 \xaa \xab \xac \xad \xae
\xaf \xb0 \xb1 \xb2 \xb3 \xb4 \xb5 \xb6 \xb7 \xb8 \xb9 \xba \xbb \xbc \xbd
\xbe \xbf \xc0 \xc1 \xc2 \xc3 \xc4 \xc5 \xc6 \xc7 \xc8 \xc9 \xca \xcb \xcc
\xcd \xce \xcf \xd0 \xd1 \xd2 \xd3 \xd4 \xd5 \xd6 \xd7 \xd8 \xd9 \xda \xdb
\xdc \xdd \xde \xdf \xe0 \xe1 \xe2 \xe3 \xe4 \xe5 \xe6 \xe7 \xe8 \xe9 \xea
\xeb \xec \xed \xee \xef \xf0 \xf1 \xf2 \xf3 \xf4 \xf5 \xf6 \xf7 \xf8 \xf9
\xfa \xfb \xfc \xfd \xfe \xff
/[\V]/BZSI
------------------------------------------------------------------
Bra
[\x00-\x09\x0e-\x84\x{86}-\x{2027}\x{202a}-\x{ffffffff}]
[\x00-\x09\x0e-\x84\x86-\xff\x{100}-\x{2027}\x{202a}-\x{ffffffff}]
Ket
End
------------------------------------------------------------------
@ -88,6 +101,19 @@ No options
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
Starting chars: \x00 \x01 \x02 \x03 \x04 \x05 \x06 \x07 \x08 \x09 \x0e
\x0f \x10 \x11 \x12 \x13 \x14 \x15 \x16 \x17 \x18 \x19 \x1a \x1b \x1c \x1d
\x1e \x1f \x20 ! " # $ % & ' ( ) * + , - . / 0 1 2 3 4 5 6 7 8 9 : ; < = >
? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c
d e f g h i j k l m n o p q r s t u v w x y z { | } ~ \x7f \x80 \x81 \x82
\x83 \x84 \x86 \x87 \x88 \x89 \x8a \x8b \x8c \x8d \x8e \x8f \x90 \x91 \x92
\x93 \x94 \x95 \x96 \x97 \x98 \x99 \x9a \x9b \x9c \x9d \x9e \x9f \xa0 \xa1
\xa2 \xa3 \xa4 \xa5 \xa6 \xa7 \xa8 \xa9 \xaa \xab \xac \xad \xae \xaf \xb0
\xb1 \xb2 \xb3 \xb4 \xb5 \xb6 \xb7 \xb8 \xb9 \xba \xbb \xbc \xbd \xbe \xbf
\xc0 \xc1 \xc2 \xc3 \xc4 \xc5 \xc6 \xc7 \xc8 \xc9 \xca \xcb \xcc \xcd \xce
\xcf \xd0 \xd1 \xd2 \xd3 \xd4 \xd5 \xd6 \xd7 \xd8 \xd9 \xda \xdb \xdc \xdd
\xde \xdf \xe0 \xe1 \xe2 \xe3 \xe4 \xe5 \xe6 \xe7 \xe8 \xe9 \xea \xeb \xec
\xed \xee \xef \xf0 \xf1 \xf2 \xf3 \xf4 \xf5 \xf6 \xf7 \xf8 \xf9 \xfa \xfb
\xfc \xfd \xfe \xff
/-- End of testinput25 --/

15
ext/pcre/pcrelib/testdata/testoutput3 vendored
View file

@ -1,7 +1,10 @@
/-- This set of tests checks local-specific features, using the fr_FR locale.
It is not Perl-compatible. There is different version called wintestinput3
f or use on Windows, where the locale is called "french". --/
/-- This set of tests checks local-specific features, using the "fr_FR" locale.
It is not Perl-compatible. When run via RunTest, the locale is edited to
be whichever of "fr_FR", "french", or "fr" is found to exist. There is
different version of this file called wintestinput3 for use on Windows,
where the locale is called "french" and the tests are run using
RunTest.bat. --/
< forbid 8W
/^[\w]+/
@ -90,7 +93,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Starting chars: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Q R S T U V W X Y Z _ a b c d e f g h i j k l m n o p q r s t u v w x y z
/\w/ISLfr_FR
@ -99,7 +102,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Starting chars: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Q R S T U V W X Y Z _ a b c d e f g h i j k l m n o p q r s t u v w x y z
ª µ º À Á Â Ã Ä Å Æ Ç È É Ê Ë Ì Í Î Ï Ð Ñ Ò Ó Ô Õ Ö Ø Ù Ú Û Ü Ý Þ ß à á â
ã ä å æ ç è é ê ë ì í î ï ð ñ ò ó ô õ ö ø ù ú û ü ý þ ÿ

8
ext/pcre/pcrelib/testdata/testoutput4 vendored
View file

@ -1263,4 +1263,12 @@ No match
aa
0: aa
/^.\B.\B./8
\x{10123}\x{10124}\x{10125}
0: \x{10123}\x{10124}\x{10125}
/^#[^\x{ffff}]#[^\x{ffff}]#[^\x{ffff}]#/8
#\x{10000}#\x{100}#\x{10ffff}#
0: #\x{10000}#\x{100}#\x{10ffff}#
/-- End of testinput4 --/

33
ext/pcre/pcrelib/testdata/testoutput5 vendored
View file

@ -270,7 +270,7 @@ No match
/[z-\x{100}]/8DZ
------------------------------------------------------------------
Bra
[z-\x{100}]
[z-\xff\x{100}]
Ket
End
------------------------------------------------------------------
@ -812,7 +812,7 @@ No match
/[\H]/8BZ
------------------------------------------------------------------
Bra
[\x00-\x08\x0a-\x1f!-\x9f\x{a1}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}]
[\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}]
Ket
End
------------------------------------------------------------------
@ -820,7 +820,7 @@ No match
/[\V]/8BZ
------------------------------------------------------------------
Bra
[\x00-\x09\x0e-\x84\x{86}-\x{2027}\x{202a}-\x{10ffff}]
[\x00-\x09\x0e-\x84\x86-\xff\x{100}-\x{2027}\x{202a}-\x{10ffff}]
Ket
End
------------------------------------------------------------------
@ -1536,7 +1536,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/[^\x{1234}]+?/iS8I
Capturing subpattern count = 0
@ -1544,7 +1544,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/[^\x{1234}]++/iS8I
Capturing subpattern count = 0
@ -1552,7 +1552,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 1
No set of starting bytes
No starting char list
/[^\x{1234}]{2}/iS8I
Capturing subpattern count = 0
@ -1560,7 +1560,7 @@ Options: caseless utf
No first char
No need char
Subject length lower bound = 2
No set of starting bytes
No starting char list
//<bsr_anycrlf><bsr_unicode>
Failed: inconsistent NEWLINE options at offset 0
@ -1620,7 +1620,7 @@ Failed: disallowed Unicode code point (>= 0xd800 && <= 0xdfff) at offset 7
/[\H\x{d7ff}]+/8BZ
------------------------------------------------------------------
Bra
[\x00-\x08\x0a-\x1f!-\x9f\x{a1}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}\x{d7ff}]++
[\x00-\x08\x0a-\x1f!-\x9f\xa1-\xff\x{100}-\x{167f}\x{1681}-\x{180d}\x{180f}-\x{1fff}\x{200b}-\x{202e}\x{2030}-\x{205e}\x{2060}-\x{2fff}\x{3001}-\x{10ffff}\x{d7ff}]++
Ket
End
------------------------------------------------------------------
@ -1660,7 +1660,7 @@ Failed: disallowed Unicode code point (>= 0xd800 && <= 0xdfff) at offset 7
/[\V\x{d7ff}]+/8BZ
------------------------------------------------------------------
Bra
[\x00-\x09\x0e-\x84\x{86}-\x{2027}\x{202a}-\x{10ffff}\x{d7ff}]++
[\x00-\x09\x0e-\x84\x86-\xff\x{100}-\x{2027}\x{202a}-\x{10ffff}\x{d7ff}]++
Ket
End
------------------------------------------------------------------
@ -1882,4 +1882,19 @@ Failed: disallowed Unicode code point (>= 0xd800 && <= 0xdfff) at offset 5
aa
0: aa
/[b-d\x{200}-\x{250}]*[ae-h]?#[\x{200}-\x{250}]{0,8}[\x00-\xff]*#[\x{200}-\x{250}]+[a-z]/8BZ
------------------------------------------------------------------
Bra
[b-d\x{200}-\x{250}]*+
[ae-h]?+
#
[\x{200}-\x{250}]{0,8}+
[\x00-\xff]*
#
[\x{200}-\x{250}]++
[a-z]
Ket
End
------------------------------------------------------------------
/-- End of testinput5 --/

12
ext/pcre/pcrelib/testdata/testoutput6 vendored
View file

@ -2445,4 +2445,16 @@ No match
\x{37e}
No match
/[RST]+/8iW
Ss\x{17f}
0: Ss\x{17f}
/[R-T]+/8iW
Ss\x{17f}
0: Ss\x{17f}
/[q-u]+/8iW
Ss\x{17f}
0: Ss\x{17f}
/-- End of testinput6 --/

28
ext/pcre/pcrelib/testdata/testoutput7 vendored
View file

@ -124,7 +124,7 @@ No match
/[z-\x{100}]/8iDZ
------------------------------------------------------------------
Bra
[Z\x{39c}\x{3bc}\x{1e9e}\x{178}z-\x{101}]
[Zz-\xff\x{39c}\x{3bc}\x{212b}\x{1e9e}\x{212b}\x{178}\x{100}-\x{101}]
Ket
End
------------------------------------------------------------------
@ -162,7 +162,7 @@ No match
/[z-\x{100}]/8DZi
------------------------------------------------------------------
Bra
[Z\x{39c}\x{3bc}\x{1e9e}\x{178}z-\x{101}]
[Zz-\xff\x{39c}\x{3bc}\x{212b}\x{1e9e}\x{212b}\x{178}\x{100}-\x{101}]
Ket
End
------------------------------------------------------------------
@ -2263,4 +2263,28 @@ No match
End
------------------------------------------------------------------
/[RST]+/8iWBZ
------------------------------------------------------------------
Bra
[R-Tr-t\x{17f}]++
Ket
End
------------------------------------------------------------------
/[R-T]+/8iWBZ
------------------------------------------------------------------
Bra
[R-Tr-t\x{17f}]++
Ket
End
------------------------------------------------------------------
/[Q-U]+/8iWBZ
------------------------------------------------------------------
Bra
[Q-Uq-u\x{17f}]++
Ket
End
------------------------------------------------------------------
/-- End of testinput7 --/

2
ext/pcre/pcrelib/testdata/testoutput8 vendored
View file

@ -7232,7 +7232,7 @@ No options
No first char
No need char
Subject length lower bound = 3
Starting byte set: a d x
Starting chars: a d x
terhjk;abcdaadsfe
0: abc
the quick xyz brown fox

4
ext/pcre/pcrelib/testdata/wintestoutput3 vendored
View file

@ -84,7 +84,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Starting chars: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Q R S T U V W X Y Z _ a b c d e f g h i j k l m n o p q r s t u v w x y z
/\w/ISLfrench
@ -93,7 +93,7 @@ No options
No first char
No need char
Subject length lower bound = 1
Starting byte set: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Starting chars: 0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P
Q R S T U V W X Y Z _ a b c d e f g h i j k l m n o p q r s t u v w x y z
ƒ Š Œ Ž š œ ž Ÿ ª ² ³ µ ¹ º À Á Â Ã Ä Å Æ Ç È É Ê Ë Ì Í Î Ï Ð Ñ Ò Ó Ô Õ Ö
Ø Ù Ú Û Ü Ý Þ ß à á â ã ä å æ ç è é ê ë ì í î ï ð ñ ò ó ô õ ö ø ù ú û ü ý

1
ext/pcre/tests/006.phpt
View file

@ -1,6 +1,7 @@
--TEST--
preg_replace() with array of failing regular expressions
--INI--
pcre.jit=0
pcre.backtrack_limit=100000
--FILE--
<?php

1
ext/pcre/tests/recursion_limit.phpt
View file

@ -7,6 +7,7 @@ if (@preg_match_all('/\p{N}/', '0123456789', $dummy) === false) {
}
?>
--INI--
pcre.jit=0
pcre.recursion_limit=2
--FILE--
<?php