nixpkgs/pkgs/development/libraries/pcre/fixes.patch

3539 lines
172 KiB
Diff
Raw Normal View History

Index: pcre_jit_compile.c
===================================================================
--- pcre_jit_compile.c (revision 1554)
+++ pcre_jit_compile.c (working copy)
@@ -1064,6 +1064,7 @@
pcre_uchar *end = NULL;
int private_data_ptr = *private_data_start;
int space, size, bracketlen;
+BOOL repeat_check = TRUE;
while (cc < ccend)
{
@@ -1071,9 +1072,10 @@
size = 0;
bracketlen = 0;
if (private_data_ptr > SLJIT_MAX_LOCAL_SIZE)
- return;
+ break;
- if (*cc == OP_ONCE || *cc == OP_ONCE_NC || *cc == OP_BRA || *cc == OP_CBRA || *cc == OP_COND)
+ if (repeat_check && (*cc == OP_ONCE || *cc == OP_ONCE_NC || *cc == OP_BRA || *cc == OP_CBRA || *cc == OP_COND))
+ {
if (detect_repeat(common, cc))
{
/* These brackets are converted to repeats, so no global
@@ -1081,6 +1083,8 @@
if (cc >= end)
end = bracketend(cc);
}
+ }
+ repeat_check = TRUE;
switch(*cc)
{
@@ -1136,6 +1140,13 @@
bracketlen = 1 + LINK_SIZE + IMM2_SIZE;
break;
+ case OP_BRAZERO:
+ case OP_BRAMINZERO:
+ case OP_BRAPOSZERO:
+ repeat_check = FALSE;
+ size = 1;
+ break;
+
CASE_ITERATOR_PRIVATE_DATA_1
space = 1;
size = -2;
@@ -1162,12 +1173,17 @@
size = 1;
break;
- CASE_ITERATOR_TYPE_PRIVATE_DATA_2B
+ case OP_TYPEUPTO:
if (cc[1 + IMM2_SIZE] != OP_ANYNL && cc[1 + IMM2_SIZE] != OP_EXTUNI)
space = 2;
size = 1 + IMM2_SIZE;
break;
+ case OP_TYPEMINUPTO:
+ space = 2;
+ size = 1 + IMM2_SIZE;
+ break;
+
case OP_CLASS:
case OP_NCLASS:
size += 1 + 32 / sizeof(pcre_uchar);
@@ -1316,6 +1332,13 @@
cc += 1 + LINK_SIZE + IMM2_SIZE;
break;
+ case OP_THEN:
+ stack_restore = TRUE;
+ if (common->control_head_ptr != 0)
+ *needs_control_head = TRUE;
+ cc ++;
+ break;
+
default:
stack_restore = TRUE;
/* Fall through. */
@@ -2220,6 +2243,7 @@
SLJIT_ASSERT_STOP();
break;
}
+ SLJIT_ASSERT(current > (sljit_sw*)current[-1]);
current = (sljit_sw*)current[-1];
}
return -1;
@@ -3209,7 +3233,7 @@
bytes[0] = len;
}
-static int scan_prefix(compiler_common *common, pcre_uchar *cc, pcre_uint32 *chars, pcre_uint8 *bytes, int max_chars)
+static int scan_prefix(compiler_common *common, pcre_uchar *cc, pcre_uint32 *chars, pcre_uint8 *bytes, int max_chars, pcre_uint32 *rec_count)
{
/* Recursive function, which scans prefix literals. */
BOOL last, any, caseless;
@@ -3227,9 +3251,14 @@
repeat = 1;
while (TRUE)
{
+ if (*rec_count == 0)
+ return 0;
+ (*rec_count)--;
+
last = TRUE;
any = FALSE;
caseless = FALSE;
+
switch (*cc)
{
case OP_CHARI:
@@ -3291,7 +3320,7 @@
#ifdef SUPPORT_UTF
if (common->utf && HAS_EXTRALEN(*cc)) len += GET_EXTRALEN(*cc);
#endif
- max_chars = scan_prefix(common, cc + len, chars, bytes, max_chars);
+ max_chars = scan_prefix(common, cc + len, chars, bytes, max_chars, rec_count);
if (max_chars == 0)
return consumed;
last = FALSE;
@@ -3314,7 +3343,7 @@
alternative = cc + GET(cc, 1);
while (*alternative == OP_ALT)
{
- max_chars = scan_prefix(common, alternative + 1 + LINK_SIZE, chars, bytes, max_chars);
+ max_chars = scan_prefix(common, alternative + 1 + LINK_SIZE, chars, bytes, max_chars, rec_count);
if (max_chars == 0)
return consumed;
alternative += GET(alternative, 1);
@@ -3556,6 +3585,7 @@
int range_right = -1, range_len = 3 - 1;
sljit_ub *update_table = NULL;
BOOL in_range;
+pcre_uint32 rec_count;
for (i = 0; i < MAX_N_CHARS; i++)
{
@@ -3564,7 +3594,8 @@
bytes[i * MAX_N_BYTES] = 0;
}
-max = scan_prefix(common, common->start, chars, bytes, MAX_N_CHARS);
+rec_count = 10000;
+max = scan_prefix(common, common->start, chars, bytes, MAX_N_CHARS, &rec_count);
if (max <= 1)
return FALSE;
@@ -4311,8 +4342,10 @@
case 4:
if ((ranges[1] - ranges[0]) == (ranges[3] - ranges[2])
&& (ranges[0] | (ranges[2] - ranges[0])) == ranges[2]
+ && (ranges[1] & (ranges[2] - ranges[0])) == 0
&& is_powerof2(ranges[2] - ranges[0]))
{
+ SLJIT_ASSERT((ranges[0] & (ranges[2] - ranges[0])) == 0 && (ranges[2] & ranges[3] & (ranges[2] - ranges[0])) != 0);
OP2(SLJIT_OR, TMP1, 0, TMP1, 0, SLJIT_IMM, ranges[2] - ranges[0]);
if (ranges[2] + 1 != ranges[3])
{
@@ -4900,9 +4933,10 @@
if (!check_class_ranges(common, (const pcre_uint8 *)cc, FALSE, TRUE, list))
{
#ifdef COMPILE_PCRE8
- SLJIT_ASSERT(common->utf);
+ jump = NULL;
+ if (common->utf)
#endif
- jump = CMP(SLJIT_GREATER, TMP1, 0, SLJIT_IMM, 255);
+ jump = CMP(SLJIT_GREATER, TMP1, 0, SLJIT_IMM, 255);
OP2(SLJIT_AND, TMP2, 0, TMP1, 0, SLJIT_IMM, 0x7);
OP2(SLJIT_LSHR, TMP1, 0, TMP1, 0, SLJIT_IMM, 3);
@@ -4911,7 +4945,10 @@
OP2(SLJIT_AND | SLJIT_SET_E, SLJIT_UNUSED, 0, TMP1, 0, TMP2, 0);
add_jump(compiler, list, JUMP(SLJIT_NOT_ZERO));
- JUMPHERE(jump);
+#ifdef COMPILE_PCRE8
+ if (common->utf)
+#endif
+ JUMPHERE(jump);
}
OP1(SLJIT_MOV, TMP1, 0, TMP3, 0);
@@ -7665,6 +7702,10 @@
OP1(SLJIT_MOV, SLJIT_MEM1(STACK_TOP), STACK(0), STR_PTR, 0);
}
+ /* Even if the match is empty, we need to reset the control head. */
+ if (needs_control_head)
+ OP1(SLJIT_MOV, SLJIT_MEM1(SLJIT_SP), common->control_head_ptr, SLJIT_MEM1(STACK_TOP), STACK(stack));
+
if (opcode == OP_SBRAPOS || opcode == OP_SCBRAPOS)
add_jump(compiler, &emptymatch, CMP(SLJIT_EQUAL, TMP1, 0, STR_PTR, 0));
@@ -7692,6 +7733,10 @@
OP1(SLJIT_MOV, SLJIT_MEM1(TMP2), (framesize + 1) * sizeof(sljit_sw), STR_PTR, 0);
}
+ /* Even if the match is empty, we need to reset the control head. */
+ if (needs_control_head)
+ OP1(SLJIT_MOV, SLJIT_MEM1(SLJIT_SP), common->control_head_ptr, SLJIT_MEM1(STACK_TOP), STACK(stack));
+
if (opcode == OP_SBRAPOS || opcode == OP_SCBRAPOS)
add_jump(compiler, &emptymatch, CMP(SLJIT_EQUAL, TMP1, 0, STR_PTR, 0));
@@ -7704,9 +7749,6 @@
}
}
- if (needs_control_head)
- OP1(SLJIT_MOV, SLJIT_MEM1(SLJIT_SP), common->control_head_ptr, SLJIT_MEM1(STACK_TOP), STACK(stack));
-
JUMPTO(SLJIT_JUMP, loop);
flush_stubs(common);
@@ -8441,8 +8483,7 @@
OP1(SLJIT_MOV, SLJIT_MEM1(STACK_TOP), STACK(1), STR_PTR, 0);
}
BACKTRACK_AS(braminzero_backtrack)->matchingpath = LABEL();
- if (cc[1] > OP_ASSERTBACK_NOT)
- count_match(common);
+ count_match(common);
break;
case OP_ONCE:
@@ -9624,7 +9665,7 @@
DEFINE_COMPILER;
pcre_uchar *cc = common->start + common->currententry->start;
pcre_uchar *ccbegin = cc + 1 + LINK_SIZE + (*cc == OP_BRA ? 0 : IMM2_SIZE);
-pcre_uchar *ccend = bracketend(cc);
+pcre_uchar *ccend = bracketend(cc) - (1 + LINK_SIZE);
BOOL needs_control_head;
int framesize = get_framesize(common, cc, NULL, TRUE, &needs_control_head);
int private_data_size = get_private_data_copy_length(common, ccbegin, ccend, needs_control_head);
@@ -9648,6 +9689,7 @@
sljit_emit_fast_enter(compiler, TMP2, 0);
allocate_stack(common, private_data_size + framesize + alternativesize);
+count_match(common);
OP1(SLJIT_MOV, SLJIT_MEM1(STACK_TOP), STACK(private_data_size + framesize + alternativesize - 1), TMP2, 0);
copy_private_data(common, ccbegin, ccend, TRUE, private_data_size + framesize + alternativesize, framesize + alternativesize, needs_control_head);
if (needs_control_head)
@@ -9992,6 +10034,7 @@
OP1(SLJIT_MOV_UI, TMP1, 0, SLJIT_MEM1(TMP1), SLJIT_OFFSETOF(jit_arguments, limit_match));
OP1(SLJIT_MOV, STACK_TOP, 0, SLJIT_MEM1(TMP2), SLJIT_OFFSETOF(struct sljit_stack, base));
OP1(SLJIT_MOV, STACK_LIMIT, 0, SLJIT_MEM1(TMP2), SLJIT_OFFSETOF(struct sljit_stack, limit));
+OP2(SLJIT_ADD, TMP1, 0, TMP1, 0, SLJIT_IMM, 1);
OP1(SLJIT_MOV, SLJIT_MEM1(SLJIT_SP), LIMIT_MATCH, TMP1, 0);
if (mode == JIT_PARTIAL_SOFT_COMPILE)
Index: RunGrepTest
===================================================================
--- RunGrepTest (revision 1554)
+++ RunGrepTest (working copy)
@@ -512,6 +512,14 @@
(cd $srcdir; $valgrind $pcregrep --line-offsets '(?<=\Ka)' $builddir/testtemp1grep) >>testtrygrep 2>&1
echo "RC=$?" >>testtrygrep
+echo "---------------------------- Test 108 ------------------------------" >>testtrygrep
+(cd $srcdir; $valgrind $pcregrep -lq PATTERN ./testdata/grepinput ./testdata/grepinputx) >>testtrygrep
+echo "RC=$?" >>testtrygrep
+
+echo "---------------------------- Test 109 -----------------------------" >>testtrygrep
+(cd $srcdir; $valgrind $pcregrep -cq lazy ./testdata/grepinput*) >>testtrygrep
+echo "RC=$?" >>testtrygrep
+
# Now compare the results.
$cf $srcdir/testdata/grepoutput testtrygrep
Index: pcre_compile.c
===================================================================
--- pcre_compile.c (revision 1554)
+++ pcre_compile.c (working copy)
@@ -174,7 +174,7 @@
-ESC_Z, CHAR_LEFT_SQUARE_BRACKET,
CHAR_BACKSLASH, CHAR_RIGHT_SQUARE_BRACKET,
CHAR_CIRCUMFLEX_ACCENT, CHAR_UNDERSCORE,
- CHAR_GRAVE_ACCENT, 7,
+ CHAR_GRAVE_ACCENT, ESC_a,
-ESC_b, 0,
-ESC_d, ESC_e,
ESC_f, 0,
@@ -202,9 +202,9 @@
/* 68 */ 0, 0, '|', ',', '%', '_', '>', '?',
/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0,
/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"',
-/* 80 */ 0, 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0,
+/* 80 */ 0, ESC_a, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0,
/* 88 */-ESC_h, 0, 0, '{', 0, 0, 0, 0,
-/* 90 */ 0, 0, -ESC_k, 'l', 0, ESC_n, 0, -ESC_p,
+/* 90 */ 0, 0, -ESC_k, 0, 0, ESC_n, 0, -ESC_p,
/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0,
/* A0 */ 0, '~', -ESC_s, ESC_tee, 0,-ESC_v, -ESC_w, 0,
/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0,
@@ -219,6 +219,12 @@
/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0,
/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0
};
+
+/* We also need a table of characters that may follow \c in an EBCDIC
+environment for characters 0-31. */
+
+static unsigned char ebcdic_escape_c[] = "@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_";
+
#endif
@@ -458,7 +464,7 @@
"range out of order in character class\0"
"nothing to repeat\0"
/* 10 */
- "operand of unlimited repeat could match the empty string\0" /** DEAD **/
+ "internal error: invalid forward reference offset\0"
"internal error: unexpected repeat\0"
"unrecognized character after (? or (?-\0"
"POSIX named classes are supported only within a class\0"
@@ -527,7 +533,11 @@
"different names for subpatterns of the same number are not allowed\0"
"(*MARK) must have an argument\0"
"this version of PCRE is not compiled with Unicode property support\0"
+#ifndef EBCDIC
"\\c must be followed by an ASCII character\0"
+#else
+ "\\c must be followed by a letter or one of [\\]^_?\0"
+#endif
"\\k is not followed by a braced, angle-bracketed, or quoted name\0"
/* 70 */
"internal error: unknown opcode in find_fixedlength()\0"
@@ -1425,7 +1435,16 @@
c ^= 0x40;
#else /* EBCDIC coding */
if (c >= CHAR_a && c <= CHAR_z) c += 64;
- c ^= 0xC0;
+ if (c == CHAR_QUESTION_MARK)
+ c = ('\\' == 188 && '`' == 74)? 0x5f : 0xff;
+ else
+ {
+ for (i = 0; i < 32; i++)
+ {
+ if (c == ebcdic_escape_c[i]) break;
+ }
+ if (i < 32) c = i; else *errorcodeptr = ERR68;
+ }
#endif
break;
@@ -1799,7 +1818,7 @@
case OP_ASSERTBACK:
case OP_ASSERTBACK_NOT:
do cc += GET(cc, 1); while (*cc == OP_ALT);
- cc += PRIV(OP_lengths)[*cc];
+ cc += 1 + LINK_SIZE;
break;
/* Skip over things that don't match chars */
@@ -2487,7 +2506,7 @@
if (c == OP_BRA || c == OP_BRAPOS ||
c == OP_CBRA || c == OP_CBRAPOS ||
c == OP_ONCE || c == OP_ONCE_NC ||
- c == OP_COND)
+ c == OP_COND || c == OP_SCOND)
{
BOOL empty_branch;
if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */
@@ -3886,11 +3905,11 @@
The problem in trying to be exactly like Perl is in the handling of escapes. We
have to be sure that [abc[:x\]pqr] is *not* treated as containing a POSIX
class, but [abc[:x\]pqr:]] is (so that an error can be generated). The code
-below handles the special case of \], but does not try to do any other escape
-processing. This makes it different from Perl for cases such as [:l\ower:]
-where Perl recognizes it as the POSIX class "lower" but PCRE does not recognize
-"l\ower". This is a lesser evil than not diagnosing bad classes when Perl does,
-I think.
+below handles the special cases \\ and \], but does not try to do any other
+escape processing. This makes it different from Perl for cases such as
+[:l\ower:] where Perl recognizes it as the POSIX class "lower" but PCRE does
+not recognize "l\ower". This is a lesser evil than not diagnosing bad classes
+when Perl does, I think.
A user pointed out that PCRE was rejecting [:a[:digit:]] whereas Perl was not.
It seems that the appearance of a nested POSIX class supersedes an apparent
@@ -3917,21 +3936,16 @@
terminator = *(++ptr); /* compiler warns about "non-constant" initializer. */
for (++ptr; *ptr != CHAR_NULL; ptr++)
{
- if (*ptr == CHAR_BACKSLASH && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
+ if (*ptr == CHAR_BACKSLASH &&
+ (ptr[1] == CHAR_RIGHT_SQUARE_BRACKET ||
+ ptr[1] == CHAR_BACKSLASH))
ptr++;
- else if (*ptr == CHAR_RIGHT_SQUARE_BRACKET) return FALSE;
- else
+ else if ((*ptr == CHAR_LEFT_SQUARE_BRACKET && ptr[1] == terminator) ||
+ *ptr == CHAR_RIGHT_SQUARE_BRACKET) return FALSE;
+ else if (*ptr == terminator && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
{
- if (*ptr == terminator && ptr[1] == CHAR_RIGHT_SQUARE_BRACKET)
- {
- *endptr = ptr;
- return TRUE;
- }
- if (*ptr == CHAR_LEFT_SQUARE_BRACKET &&
- (ptr[1] == CHAR_COLON || ptr[1] == CHAR_DOT ||
- ptr[1] == CHAR_EQUALS_SIGN) &&
- check_posix_syntax(ptr, endptr))
- return FALSE;
+ *endptr = ptr;
+ return TRUE;
}
}
return FALSE;
@@ -3985,11 +3999,12 @@
is called, the partially compiled regex must be temporarily terminated with
OP_END.
-This function has been extended with the possibility of forward references for
-recursions and subroutine calls. It must also check the list of such references
-for the group we are dealing with. If it finds that one of the recursions in
-the current group is on this list, it adjusts the offset in the list, not the
-value in the reference (which is a group number).
+This function has been extended to cope with forward references for recursions
+and subroutine calls. It must check the list of such references for the
+group we are dealing with. If it finds that one of the recursions in the
+current group is on this list, it does not adjust the value in the reference
+(which is a group number). After the group has been scanned, all the offsets in
+the forward reference list for the group are adjusted.
Arguments:
group points to the start of the group
@@ -4005,29 +4020,21 @@
adjust_recurse(pcre_uchar *group, int adjust, BOOL utf, compile_data *cd,
size_t save_hwm_offset)
{
+int offset;
+pcre_uchar *hc;
pcre_uchar *ptr = group;
while ((ptr = (pcre_uchar *)find_recurse(ptr, utf)) != NULL)
{
- int offset;
- pcre_uchar *hc;
-
- /* See if this recursion is on the forward reference list. If so, adjust the
- reference. */
-
for (hc = (pcre_uchar *)cd->start_workspace + save_hwm_offset; hc < cd->hwm;
hc += LINK_SIZE)
{
offset = (int)GET(hc, 0);
- if (cd->start_code + offset == ptr + 1)
- {
- PUT(hc, 0, offset + adjust);
- break;
- }
+ if (cd->start_code + offset == ptr + 1) break;
}
- /* Otherwise, adjust the recursion offset if it's after the start of this
- group. */
+ /* If we have not found this recursion on the forward reference list, adjust
+ the recursion's offset if it's after the start of this group. */
if (hc >= cd->hwm)
{
@@ -4037,6 +4044,15 @@
ptr += 1 + LINK_SIZE;
}
+
+/* Now adjust all forward reference offsets for the group. */
+
+for (hc = (pcre_uchar *)cd->start_workspace + save_hwm_offset; hc < cd->hwm;
+ hc += LINK_SIZE)
+ {
+ offset = (int)GET(hc, 0);
+ PUT(hc, 0, offset + adjust);
+ }
}
@@ -4465,7 +4481,7 @@
const pcre_uchar *nestptr = NULL;
pcre_uchar *previous = NULL;
pcre_uchar *previous_callout = NULL;
-size_t save_hwm_offset = 0;
+size_t item_hwm_offset = 0;
pcre_uint8 classbits[32];
/* We can fish out the UTF-8 setting once and for all into a BOOL, but we
@@ -4623,8 +4639,7 @@
/* In the real compile phase, just check the workspace used by the forward
reference list. */
- else if (cd->hwm > cd->start_workspace + cd->workspace_size -
- WORK_SIZE_SAFETY_MARGIN)
+ else if (cd->hwm > cd->start_workspace + cd->workspace_size)
{
*errorcodeptr = ERR52;
goto FAILED;
@@ -4767,6 +4782,7 @@
zeroreqchar = reqchar;
zeroreqcharflags = reqcharflags;
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
*code++ = ((options & PCRE_DOTALL) != 0)? OP_ALLANY: OP_ANY;
break;
@@ -4818,6 +4834,7 @@
/* Handle a real character class. */
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
/* PCRE supports POSIX class stuff inside a class. Perl gives an error if
they are encountered at the top level, so we'll do that too. */
@@ -5195,9 +5212,9 @@
cd, PRIV(vspace_list));
continue;
-#ifdef SUPPORT_UCP
case ESC_p:
case ESC_P:
+#ifdef SUPPORT_UCP
{
BOOL negated;
unsigned int ptype = 0, pdata = 0;
@@ -5211,6 +5228,9 @@
class_has_8bitchar--; /* Undo! */
continue;
}
+#else
+ *errorcodeptr = ERR45;
+ goto FAILED;
#endif
/* Unrecognized escapes are faulted if PCRE is running in its
strict mode. By default, for compatibility with Perl, they are
@@ -5930,7 +5950,7 @@
{
register int i;
int len = (int)(code - previous);
- size_t base_hwm_offset = save_hwm_offset;
+ size_t base_hwm_offset = item_hwm_offset;
pcre_uchar *bralink = NULL;
pcre_uchar *brazeroptr = NULL;
@@ -5985,7 +6005,7 @@
if (repeat_max <= 1) /* Covers 0, 1, and unlimited */
{
*code = OP_END;
- adjust_recurse(previous, 1, utf, cd, save_hwm_offset);
+ adjust_recurse(previous, 1, utf, cd, item_hwm_offset);
memmove(previous + 1, previous, IN_UCHARS(len));
code++;
if (repeat_max == 0)
@@ -6009,7 +6029,7 @@
{
int offset;
*code = OP_END;
- adjust_recurse(previous, 2 + LINK_SIZE, utf, cd, save_hwm_offset);
+ adjust_recurse(previous, 2 + LINK_SIZE, utf, cd, item_hwm_offset);
memmove(previous + 2 + LINK_SIZE, previous, IN_UCHARS(len));
code += 2 + LINK_SIZE;
*previous++ = OP_BRAZERO + repeat_type;
@@ -6254,6 +6274,12 @@
while (*scode == OP_ALT);
}
+ /* A conditional group with only one branch has an implicit empty
+ alternative branch. */
+
+ if (*bracode == OP_COND && bracode[GET(bracode,1)] != OP_ALT)
+ *bracode = OP_SCOND;
+
/* Handle possessive quantifiers. */
if (possessive_quantifier)
@@ -6267,11 +6293,11 @@
{
int nlen = (int)(code - bracode);
*code = OP_END;
- adjust_recurse(bracode, 1 + LINK_SIZE, utf, cd, save_hwm_offset);
+ adjust_recurse(bracode, 1 + LINK_SIZE, utf, cd, item_hwm_offset);
memmove(bracode + 1 + LINK_SIZE, bracode, IN_UCHARS(nlen));
code += 1 + LINK_SIZE;
nlen += 1 + LINK_SIZE;
- *bracode = OP_BRAPOS;
+ *bracode = (*bracode == OP_COND)? OP_BRAPOS : OP_SBRAPOS;
*code++ = OP_KETRPOS;
PUTINC(code, 0, nlen);
PUT(bracode, 1, nlen);
@@ -6401,7 +6427,7 @@
else
{
*code = OP_END;
- adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, save_hwm_offset);
+ adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, item_hwm_offset);
memmove(tempcode + 1 + LINK_SIZE, tempcode, IN_UCHARS(len));
code += 1 + LINK_SIZE;
len += 1 + LINK_SIZE;
@@ -6450,7 +6476,7 @@
default:
*code = OP_END;
- adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, save_hwm_offset);
+ adjust_recurse(tempcode, 1 + LINK_SIZE, utf, cd, item_hwm_offset);
memmove(tempcode + 1 + LINK_SIZE, tempcode, IN_UCHARS(len));
code += 1 + LINK_SIZE;
len += 1 + LINK_SIZE;
@@ -6623,7 +6649,7 @@
newoptions = options;
skipbytes = 0;
bravalue = OP_CBRA;
- save_hwm_offset = cd->hwm - cd->start_workspace;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
reset_bracount = FALSE;
/* Deal with the extended parentheses; all are introduced by '?', and the
@@ -6641,6 +6667,7 @@
/* ------------------------------------------------------------ */
case CHAR_VERTICAL_LINE: /* Reset capture count for each branch */
reset_bracount = TRUE;
+ cd->dupgroups = TRUE; /* Record (?| encountered */
/* Fall through */
/* ------------------------------------------------------------ */
@@ -6741,6 +6768,12 @@
{
while (IS_DIGIT(*ptr))
{
+ if (recno > INT_MAX / 10 - 1) /* Integer overflow */
+ {
+ while (IS_DIGIT(*ptr)) ptr++;
+ *errorcodeptr = ERR61;
+ goto FAILED;
+ }
recno = recno * 10 + (int)(*ptr - CHAR_0);
ptr++;
}
@@ -6769,7 +6802,7 @@
ptr++;
}
namelen = (int)(ptr - name);
- if (lengthptr != NULL) *lengthptr += IMM2_SIZE;
+ if (lengthptr != NULL) skipbytes += IMM2_SIZE;
}
/* Check the terminator */
@@ -6875,6 +6908,11 @@
*errorcodeptr = ERR15;
goto FAILED;
}
+ if (recno > INT_MAX / 10 - 1) /* Integer overflow */
+ {
+ *errorcodeptr = ERR61;
+ goto FAILED;
+ }
recno = recno * 10 + name[i] - CHAR_0;
}
if (recno == 0) recno = RREF_ANY;
@@ -7151,7 +7189,8 @@
if (lengthptr != NULL)
{
named_group *ng;
-
+ recno = 0;
+
if (namelen == 0)
{
*errorcodeptr = ERR62;
@@ -7168,20 +7207,6 @@
goto FAILED;
}
- /* The name table does not exist in the first pass; instead we must
- scan the list of names encountered so far in order to get the
- number. If the name is not found, set the value to 0 for a forward
- reference. */
-
- ng = cd->named_groups;
- for (i = 0; i < cd->names_found; i++, ng++)
- {
- if (namelen == ng->length &&
- STRNCMP_UC_UC(name, ng->name, namelen) == 0)
- break;
- }
- recno = (i < cd->names_found)? ng->number : 0;
-
/* Count named back references. */
if (!is_recurse) cd->namedrefcount++;
@@ -7191,6 +7216,56 @@
16-bit data item. */
*lengthptr += IMM2_SIZE;
+
+ /* If this is a forward reference and we are within a (?|...) group,
+ the reference may end up as the number of a group which we are
+ currently inside, that is, it could be a recursive reference. In the
+ real compile this will be picked up and the reference wrapped with
+ OP_ONCE to make it atomic, so we must space in case this occurs. */
+
+ /* In fact, this can happen for a non-forward reference because
+ another group with the same number might be created later. This
+ issue is fixed "properly" in PCRE2. As PCRE1 is now in maintenance
+ only mode, we finesse the bug by allowing more memory always. */
+
+ *lengthptr += 2 + 2*LINK_SIZE;
+
+ /* It is even worse than that. The current reference may be to an
+ existing named group with a different number (so apparently not
+ recursive) but which later on is also attached to a group with the
+ current number. This can only happen if $(| has been previous
+ encountered. In that case, we allow yet more memory, just in case.
+ (Again, this is fixed "properly" in PCRE2. */
+
+ if (cd->dupgroups) *lengthptr += 4 + 4*LINK_SIZE;
+
+ /* Otherwise, check for recursion here. The name table does not exist
+ in the first pass; instead we must scan the list of names encountered
+ so far in order to get the number. If the name is not found, leave
+ the value of recno as 0 for a forward reference. */
+
+ else
+ {
+ ng = cd->named_groups;
+ for (i = 0; i < cd->names_found; i++, ng++)
+ {
+ if (namelen == ng->length &&
+ STRNCMP_UC_UC(name, ng->name, namelen) == 0)
+ {
+ open_capitem *oc;
+ recno = ng->number;
+ if (is_recurse) break;
+ for (oc = cd->open_caps; oc != NULL; oc = oc->next)
+ {
+ if (oc->number == recno)
+ {
+ oc->flag = TRUE;
+ break;
+ }
+ }
+ }
+ }
+ }
}
/* In the real compile, search the name table. We check the name
@@ -7237,8 +7312,6 @@
for (i++; i < cd->names_found; i++)
{
if (STRCMP_UC_UC(slot + IMM2_SIZE, cslot + IMM2_SIZE) != 0) break;
-
-
count++;
cslot += cd->name_entry_size;
}
@@ -7247,6 +7320,7 @@
{
if (firstcharflags == REQ_UNSET) firstcharflags = REQ_NONE;
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
*code++ = ((options & PCRE_CASELESS) != 0)? OP_DNREFI : OP_DNREF;
PUT2INC(code, 0, index);
PUT2INC(code, 0, count);
@@ -7284,9 +7358,14 @@
/* ------------------------------------------------------------ */
- case CHAR_R: /* Recursion */
- ptr++; /* Same as (?0) */
- /* Fall through */
+ case CHAR_R: /* Recursion, same as (?0) */
+ recno = 0;
+ if (*(++ptr) != CHAR_RIGHT_PARENTHESIS)
+ {
+ *errorcodeptr = ERR29;
+ goto FAILED;
+ }
+ goto HANDLE_RECURSION;
/* ------------------------------------------------------------ */
@@ -7323,7 +7402,15 @@
recno = 0;
while(IS_DIGIT(*ptr))
+ {
+ if (recno > INT_MAX / 10 - 1) /* Integer overflow */
+ {
+ while (IS_DIGIT(*ptr)) ptr++;
+ *errorcodeptr = ERR61;
+ goto FAILED;
+ }
recno = recno * 10 + *ptr++ - CHAR_0;
+ }
if (*ptr != (pcre_uchar)terminator)
{
@@ -7360,6 +7447,7 @@
HANDLE_RECURSION:
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
called = cd->start_code;
/* When we are actually compiling, find the bracket that is being
@@ -7561,7 +7649,11 @@
previous = NULL;
cd->iscondassert = FALSE;
}
- else previous = code;
+ else
+ {
+ previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
+ }
*code = bravalue;
tempcode = code;
@@ -7809,7 +7901,7 @@
const pcre_uchar *p;
pcre_uint32 cf;
- save_hwm_offset = cd->hwm - cd->start_workspace; /* Normally this is set when '(' is read */
+ item_hwm_offset = cd->hwm - cd->start_workspace; /* Normally this is set when '(' is read */
terminator = (*(++ptr) == CHAR_LESS_THAN_SIGN)?
CHAR_GREATER_THAN_SIGN : CHAR_APOSTROPHE;
@@ -7838,7 +7930,7 @@
if (*p != (pcre_uchar)terminator)
{
*errorcodeptr = ERR57;
- break;
+ goto FAILED;
}
ptr++;
goto HANDLE_NUMERICAL_RECURSION;
@@ -7853,7 +7945,7 @@
ptr[1] != CHAR_APOSTROPHE && ptr[1] != CHAR_LEFT_CURLY_BRACKET))
{
*errorcodeptr = ERR69;
- break;
+ goto FAILED;
}
is_recurse = FALSE;
terminator = (*(++ptr) == CHAR_LESS_THAN_SIGN)?
@@ -7877,6 +7969,7 @@
HANDLE_REFERENCE:
if (firstcharflags == REQ_UNSET) firstcharflags = REQ_NONE;
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
*code++ = ((options & PCRE_CASELESS) != 0)? OP_REFI : OP_REF;
PUT2INC(code, 0, recno);
cd->backref_map |= (recno < 32)? (1 << recno) : 1;
@@ -7906,6 +7999,7 @@
if (!get_ucp(&ptr, &negated, &ptype, &pdata, errorcodeptr))
goto FAILED;
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
*code++ = ((escape == ESC_p) != negated)? OP_PROP : OP_NOTPROP;
*code++ = ptype;
*code++ = pdata;
@@ -7946,6 +8040,7 @@
{
previous = (escape > ESC_b && escape < ESC_Z)? code : NULL;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
*code++ = (!utf && escape == ESC_C)? OP_ALLANY : escape;
}
}
@@ -7989,6 +8084,7 @@
ONE_CHAR:
previous = code;
+ item_hwm_offset = cd->hwm - cd->start_workspace;
/* For caseless UTF-8 mode when UCP support is available, check whether
this character has more than one other case. If so, generate a special
@@ -9164,6 +9260,7 @@
cd->name_entry_size = 0;
cd->name_table = NULL;
cd->dupnames = FALSE;
+cd->dupgroups = FALSE;
cd->namedrefcount = 0;
cd->start_code = cworkspace;
cd->hwm = cworkspace;
@@ -9198,7 +9295,7 @@
DPRINTF(("end pre-compile: length=%d workspace=%d\n", length,
(int)(cd->hwm - cworkspace)));
-
+
if (length > MAX_PATTERN_SIZE)
{
errorcode = ERR20;
@@ -9336,6 +9433,16 @@
int offset, recno;
cd->hwm -= LINK_SIZE;
offset = GET(cd->hwm, 0);
+
+ /* Check that the hwm handling hasn't gone wrong. This whole area is
+ rewritten in PCRE2 because there are some obscure cases. */
+
+ if (offset == 0 || codestart[offset-1] != OP_RECURSE)
+ {
+ errorcode = ERR10;
+ break;
+ }
+
recno = GET(codestart, offset);
if (recno != prev_recno)
{
@@ -9366,7 +9473,7 @@
"const" attribute if the cast (pcre_uchar *)codestart is used directly in the
function call. */
-if ((options & PCRE_NO_AUTO_POSSESS) == 0)
+if (errorcode == 0 && (options & PCRE_NO_AUTO_POSSESS) == 0)
{
pcre_uchar *temp = (pcre_uchar *)codestart;
auto_possessify(temp, utf, cd);
@@ -9380,7 +9487,7 @@
exceptional ones forgo this. We scan the pattern to check that they are fixed
length, and set their lengths. */
-if (cd->check_lookbehind)
+if (errorcode == 0 && cd->check_lookbehind)
{
pcre_uchar *cc = (pcre_uchar *)codestart;
@@ -9593,4 +9700,3 @@
}
/* End of pcre_compile.c */
-
Index: ChangeLog
===================================================================
--- ChangeLog (revision 1554)
+++ ChangeLog (working copy)
@@ -1,6 +1,162 @@
ChangeLog for PCRE
------------------
+Note that the PCRE 8.xx series (PCRE1) is now in a bugfix-only state. All
+development is happening in the PCRE2 10.xx series.
+
+Version 8.38 xx-xxx-xxxx
+------------------------
+
+1. If a group that contained a recursive back reference also contained a
+ forward reference subroutine call followed by a non-forward-reference
+ subroutine call, for example /.((?2)(?R)\1)()/, pcre2_compile() failed to
+ compile correct code, leading to undefined behaviour or an internally
+ detected error. This bug was discovered by the LLVM fuzzer.
+
+2. Quantification of certain items (e.g. atomic back references) could cause
+ incorrect code to be compiled when recursive forward references were
+ involved. For example, in this pattern: /(?1)()((((((\1++))\x85)+)|))/.
+ This bug was discovered by the LLVM fuzzer.
+
+3. A repeated conditional group whose condition was a reference by name caused
+ a buffer overflow if there was more than one group with the given name.
+ This bug was discovered by the LLVM fuzzer.
+
+4. A recursive back reference by name within a group that had the same name as
+ another group caused a buffer overflow. For example:
+ /(?J)(?'d'(?'d'\g{d}))/. This bug was discovered by the LLVM fuzzer.
+
+5. A forward reference by name to a group whose number is the same as the
+ current group, for example in this pattern: /(?|(\k'Pm')|(?'Pm'))/, caused
+ a buffer overflow at compile time. This bug was discovered by the LLVM
+ fuzzer.
+
+6. A lookbehind assertion within a set of mutually recursive subpatterns could
+ provoke a buffer overflow. This bug was discovered by the LLVM fuzzer.
+
+7. Another buffer overflow bug involved duplicate named groups with a
+ reference between their definition, with a group that reset capture
+ numbers, for example: /(?J:(?|(?'R')(\k'R')|((?'R'))))/. This has been
+ fixed by always allowing for more memory, even if not needed. (A proper fix
+ is implemented in PCRE2, but it involves more refactoring.)
+
+8. There was no check for integer overflow in subroutine calls such as (?123).
+
+9. The table entry for \l in EBCDIC environments was incorrect, leading to its
+ being treated as a literal 'l' instead of causing an error.
+
+10. There was a buffer overflow if pcre_exec() was called with an ovector of
+ size 1. This bug was found by american fuzzy lop.
+
+11. If a non-capturing group containing a conditional group that could match
+ an empty string was repeated, it was not identified as matching an empty
+ string itself. For example: /^(?:(?(1)x|)+)+$()/.
+
+12. In an EBCDIC environment, pcretest was mishandling the escape sequences
+ \a and \e in test subject lines.
+
+13. In an EBCDIC environment, \a in a pattern was converted to the ASCII
+ instead of the EBCDIC value.
+
+14. The handling of \c in an EBCDIC environment has been revised so that it is
+ now compatible with the specification in Perl's perlebcdic page.
+
+15. The EBCDIC character 0x41 is a non-breaking space, equivalent to 0xa0 in
+ ASCII/Unicode. This has now been added to the list of characters that are
+ recognized as white space in EBCDIC.
+
+16. When PCRE was compiled without UCP support, the use of \p and \P gave an
+ error (correctly) when used outside a class, but did not give an error
+ within a class.
+
+17. \h within a class was incorrectly compiled in EBCDIC environments.
+
+18. A pattern with an unmatched closing parenthesis that contained a backward
+ assertion which itself contained a forward reference caused buffer
+ overflow. And example pattern is: /(?=di(?<=(?1))|(?=(.))))/.
+
+19. JIT should return with error when the compiled pattern requires more stack
+ space than the maximum.
+
+20. A possessively repeated conditional group that could match an empty string,
+ for example, /(?(R))*+/, was incorrectly compiled.
+
+21. Fix infinite recursion in the JIT compiler when certain patterns such as
+ /(?:|a|){100}x/ are analysed.
+
+22. Some patterns with character classes involving [: and \\ were incorrectly
+ compiled and could cause reading from uninitialized memory or an incorrect
+ error diagnosis.
+
+23. Pathological patterns containing many nested occurrences of [: caused
+ pcre_compile() to run for a very long time.
+
+24. A conditional group with only one branch has an implicit empty alternative
+ branch and must therefore be treated as potentially matching an empty
+ string.
+
+25. If (?R was followed by - or + incorrect behaviour happened instead of a
+ diagnostic.
+
+26. Arrange to give up on finding the minimum matching length for overly
+ complex patterns.
+
+27. Similar to (4) above: in a pattern with duplicated named groups and an
+ occurrence of (?| it is possible for an apparently non-recursive back
+ reference to become recursive if a later named group with the relevant
+ number is encountered. This could lead to a buffer overflow. Wen Guanxing
+ from Venustech ADLAB discovered this bug.
+
+28. If pcregrep was given the -q option with -c or -l, or when handling a
+ binary file, it incorrectly wrote output to stdout.
+
+29. The JIT compiler did not restore the control verb head in case of *THEN
+ control verbs. This issue was found by Karl Skomski with a custom LLVM
+ fuzzer.
+
+30. Error messages for syntax errors following \g and \k were giving inaccurate
+ offsets in the pattern.
+
+31. Added a check for integer overflow in conditions (?(<digits>) and
+ (?(R<digits>). This omission was discovered by Karl Skomski with the LLVM
+ fuzzer.
+
+32. Handling recursive references such as (?2) when the reference is to a group
+ later in the pattern uses code that is very hacked about and error-prone.
+ It has been re-written for PCRE2. Here in PCRE1, a check has been added to
+ give an internal error if it is obvious that compiling has gone wrong.
+
+33. The JIT compiler should not check repeats after a {0,1} repeat byte code.
+ This issue was found by Karl Skomski with a custom LLVM fuzzer.
+
+34. The JIT compiler should restore the control chain for empty possessive
+ repeats. This issue was found by Karl Skomski with a custom LLVM fuzzer.
+
+35. Match limit check added to JIT recursion. This issue was found by Karl
+ Skomski with a custom LLVM fuzzer.
+
+36. Yet another case similar to 27 above has been circumvented by an
+ unconditional allocation of extra memory. This issue is fixed "properly" in
+ PCRE2 by refactoring the way references are handled. Wen Guanxing
+ from Venustech ADLAB discovered this bug.
+
+37. Fix two assertion fails in JIT. These issues were found by Karl Skomski
+ with a custom LLVM fuzzer.
+
+38. Fixed a corner case of range optimization in JIT.
+
+39. An incorrect error "overran compiling workspace" was given if there were
+ exactly enough group forward references such that the last one extended
+ into the workspace safety margin. The next one would have expanded the
+ workspace. The test for overflow was not including the safety margin.
+
+40. A match limit issue is fixed in JIT which was found by Karl Skomski
+ with a custom LLVM fuzzer.
+
+41. Remove the use of /dev/null in testdata/testinput2, because it doesn't
+ work under Windows. (Why has it taken so long for anyone to notice?)
+
+
Version 8.37 28-April-2015
--------------------------
Index: pcretest.c
===================================================================
--- pcretest.c (revision 1554)
+++ pcretest.c (working copy)
@@ -4621,9 +4621,9 @@
else switch ((c = *p++))
{
- case 'a': c = 7; break;
+ case 'a': c = CHAR_BEL; break;
case 'b': c = '\b'; break;
- case 'e': c = 27; break;
+ case 'e': c = CHAR_ESC; break;
case 'f': c = '\f'; break;
case 'n': c = '\n'; break;
case 'r': c = '\r'; break;
Index: sljit/sljitLir.c
===================================================================
--- sljit/sljitLir.c (revision 1554)
+++ sljit/sljitLir.c (working copy)
@@ -845,8 +845,8 @@
}
static SLJIT_CONST char* op0_names[] = {
- (char*)"breakpoint", (char*)"nop",
- (char*)"lumul", (char*)"lsmul", (char*)"ludiv", (char*)"lsdiv",
+ (char*)"breakpoint", (char*)"nop", (char*)"lumul", (char*)"lsmul",
+ (char*)"udivmod", (char*)"sdivmod", (char*)"udivi", (char*)"sdivi"
};
static SLJIT_CONST char* op1_names[] = {
@@ -1036,7 +1036,7 @@
{
#if (defined SLJIT_ARGUMENT_CHECKS && SLJIT_ARGUMENT_CHECKS)
CHECK_ARGUMENT((op >= SLJIT_BREAKPOINT && op <= SLJIT_LSMUL)
- || ((op & ~SLJIT_INT_OP) >= SLJIT_LUDIV && (op & ~SLJIT_INT_OP) <= SLJIT_LSDIV));
+ || ((op & ~SLJIT_INT_OP) >= SLJIT_UDIVMOD && (op & ~SLJIT_INT_OP) <= SLJIT_SDIVI));
CHECK_ARGUMENT(op < SLJIT_LUMUL || compiler->scratches >= 2);
#endif
#if (defined SLJIT_VERBOSE && SLJIT_VERBOSE)
@@ -1447,6 +1447,8 @@
static SLJIT_INLINE CHECK_RETURN_TYPE check_sljit_get_local_base(struct sljit_compiler *compiler, sljit_si dst, sljit_sw dstw, sljit_sw offset)
{
+ SLJIT_UNUSED_ARG(offset);
+
#if (defined SLJIT_ARGUMENT_CHECKS && SLJIT_ARGUMENT_CHECKS)
FUNCTION_CHECK_DST(dst, dstw);
#endif
@@ -1462,6 +1464,8 @@
static SLJIT_INLINE CHECK_RETURN_TYPE check_sljit_emit_const(struct sljit_compiler *compiler, sljit_si dst, sljit_sw dstw, sljit_sw init_value)
{
+ SLJIT_UNUSED_ARG(init_value);
+
#if (defined SLJIT_ARGUMENT_CHECKS && SLJIT_ARGUMENT_CHECKS)
FUNCTION_CHECK_DST(dst, dstw);
#endif
Index: sljit/sljitNativeSPARC_common.c
===================================================================
--- sljit/sljitNativeSPARC_common.c (revision 1554)
+++ sljit/sljitNativeSPARC_common.c (working copy)
@@ -777,20 +777,25 @@
#else
#error "Implementation required"
#endif
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+ SLJIT_COMPILE_ASSERT((SLJIT_UDIVMOD & 0x2) == 0 && SLJIT_UDIVI - 0x2 == SLJIT_UDIVMOD, bad_div_opcode_assignments);
#if (defined SLJIT_CONFIG_SPARC_32 && SLJIT_CONFIG_SPARC_32)
- if (op == SLJIT_LUDIV)
+ if ((op | 0x2) == SLJIT_UDIVI)
FAIL_IF(push_inst(compiler, WRY | S1(0), MOVABLE_INS));
else {
FAIL_IF(push_inst(compiler, SRA | D(TMP_REG1) | S1(SLJIT_R0) | IMM(31), DR(TMP_REG1)));
FAIL_IF(push_inst(compiler, WRY | S1(TMP_REG1), MOVABLE_INS));
}
- FAIL_IF(push_inst(compiler, OR | D(TMP_REG2) | S1(0) | S2(SLJIT_R0), DR(TMP_REG2)));
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? UDIV : SDIV) | D(SLJIT_R0) | S1(SLJIT_R0) | S2(SLJIT_R1), DR(SLJIT_R0)));
+ if (op <= SLJIT_SDIVMOD)
+ FAIL_IF(push_inst(compiler, OR | D(TMP_REG2) | S1(0) | S2(SLJIT_R0), DR(TMP_REG2)));
+ FAIL_IF(push_inst(compiler, ((op | 0x2) == SLJIT_UDIVI ? UDIV : SDIV) | D(SLJIT_R0) | S1(SLJIT_R0) | S2(SLJIT_R1), DR(SLJIT_R0)));
+ if (op >= SLJIT_UDIVI)
+ return SLJIT_SUCCESS;
FAIL_IF(push_inst(compiler, SMUL | D(SLJIT_R1) | S1(SLJIT_R0) | S2(SLJIT_R1), DR(SLJIT_R1)));
- FAIL_IF(push_inst(compiler, SUB | D(SLJIT_R1) | S1(TMP_REG2) | S2(SLJIT_R1), DR(SLJIT_R1)));
- return SLJIT_SUCCESS;
+ return push_inst(compiler, SUB | D(SLJIT_R1) | S1(TMP_REG2) | S2(SLJIT_R1), DR(SLJIT_R1));
#else
#error "Implementation required"
#endif
Index: sljit/sljitNativeMIPS_common.c
===================================================================
--- sljit/sljitNativeMIPS_common.c (revision 1554)
+++ sljit/sljitNativeMIPS_common.c (working copy)
@@ -1053,8 +1053,11 @@
#endif
FAIL_IF(push_inst(compiler, MFLO | D(SLJIT_R0), DR(SLJIT_R0)));
return push_inst(compiler, MFHI | D(SLJIT_R1), DR(SLJIT_R1));
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+ SLJIT_COMPILE_ASSERT((SLJIT_UDIVMOD & 0x2) == 0 && SLJIT_UDIVI - 0x2 == SLJIT_UDIVMOD, bad_div_opcode_assignments);
#if !(defined SLJIT_MIPS_R1 && SLJIT_MIPS_R1)
FAIL_IF(push_inst(compiler, NOP, UNMOVABLE_INS));
FAIL_IF(push_inst(compiler, NOP, UNMOVABLE_INS));
@@ -1062,15 +1065,15 @@
#if (defined SLJIT_CONFIG_MIPS_64 && SLJIT_CONFIG_MIPS_64)
if (int_op)
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DIVU : DIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
+ FAIL_IF(push_inst(compiler, ((op | 0x2) == SLJIT_UDIVI ? DIVU : DIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
else
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DDIVU : DDIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
+ FAIL_IF(push_inst(compiler, ((op | 0x2) == SLJIT_UDIVI ? DDIVU : DDIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
#else
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DIVU : DIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
+ FAIL_IF(push_inst(compiler, ((op | 0x2) == SLJIT_UDIVI ? DIVU : DIV) | S(SLJIT_R0) | T(SLJIT_R1), MOVABLE_INS));
#endif
FAIL_IF(push_inst(compiler, MFLO | D(SLJIT_R0), DR(SLJIT_R0)));
- return push_inst(compiler, MFHI | D(SLJIT_R1), DR(SLJIT_R1));
+ return (op >= SLJIT_UDIVI) ? SLJIT_SUCCESS : push_inst(compiler, MFHI | D(SLJIT_R1), DR(SLJIT_R1));
}
return SLJIT_SUCCESS;
Index: sljit/sljitNativeARM_32.c
===================================================================
--- sljit/sljitNativeARM_32.c (revision 1554)
+++ sljit/sljitNativeARM_32.c (working copy)
@@ -1833,18 +1833,33 @@
| (reg_map[SLJIT_R0] << 8)
| reg_map[TMP_REG1]);
#endif
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
- if (compiler->scratches >= 3)
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+ SLJIT_COMPILE_ASSERT((SLJIT_UDIVMOD & 0x2) == 0 && SLJIT_UDIVI - 0x2 == SLJIT_UDIVMOD, bad_div_opcode_assignments);
+ SLJIT_COMPILE_ASSERT(reg_map[2] == 1 && reg_map[3] == 2, bad_register_mapping);
+
+ if ((op >= SLJIT_UDIVI) && (compiler->scratches >= 3)) {
FAIL_IF(push_inst(compiler, 0xe52d2008 /* str r2, [sp, #-8]! */));
+ FAIL_IF(push_inst(compiler, 0xe58d1004 /* str r1, [sp, #4] */));
+ }
+ else if ((op >= SLJIT_UDIVI) || (compiler->scratches >= 3))
+ FAIL_IF(push_inst(compiler, 0xe52d0008 | (op >= SLJIT_UDIVI ? 0x1000 : 0x2000) /* str r1/r2, [sp, #-8]! */));
+
#if defined(__GNUC__)
FAIL_IF(sljit_emit_ijump(compiler, SLJIT_FAST_CALL, SLJIT_IMM,
- (op == SLJIT_LUDIV ? SLJIT_FUNC_OFFSET(__aeabi_uidivmod) : SLJIT_FUNC_OFFSET(__aeabi_idivmod))));
+ ((op | 0x2) == SLJIT_UDIVI ? SLJIT_FUNC_OFFSET(__aeabi_uidivmod) : SLJIT_FUNC_OFFSET(__aeabi_idivmod))));
#else
#error "Software divmod functions are needed"
#endif
- if (compiler->scratches >= 3)
- return push_inst(compiler, 0xe49d2008 /* ldr r2, [sp], #8 */);
+
+ if ((op >= SLJIT_UDIVI) && (compiler->scratches >= 3)) {
+ FAIL_IF(push_inst(compiler, 0xe59d1004 /* ldr r1, [sp, #4] */));
+ FAIL_IF(push_inst(compiler, 0xe49d2008 /* ldr r2, [sp], #8 */));
+ }
+ else if ((op >= SLJIT_UDIVI) || (compiler->scratches >= 3))
+ return push_inst(compiler, 0xe49d0008 | (op >= SLJIT_UDIVI ? 0x1000 : 0x2000) /* ldr r1/r2, [sp], #8 */);
return SLJIT_SUCCESS;
}
Index: sljit/sljitLir.h
===================================================================
--- sljit/sljitLir.h (revision 1554)
+++ sljit/sljitLir.h (working copy)
@@ -687,7 +687,7 @@
#define SLJIT_OP0_BASE 0
/* Flags: - (never set any flags)
- Note: breakpoint instruction is not supported by all architectures (namely ppc)
+ Note: breakpoint instruction is not supported by all architectures (e.g. ppc)
It falls back to SLJIT_NOP in those cases. */
#define SLJIT_BREAKPOINT (SLJIT_OP0_BASE + 0)
/* Flags: - (never set any flags)
@@ -696,24 +696,42 @@
#define SLJIT_NOP (SLJIT_OP0_BASE + 1)
/* Flags: - (may destroy flags)
Unsigned multiplication of SLJIT_R0 and SLJIT_R1.
- Result goes to SLJIT_R1:SLJIT_R0 (high:low) word */
+ Result is placed into SLJIT_R1:SLJIT_R0 (high:low) word */
#define SLJIT_LUMUL (SLJIT_OP0_BASE + 2)
/* Flags: - (may destroy flags)
Signed multiplication of SLJIT_R0 and SLJIT_R1.
- Result goes to SLJIT_R1:SLJIT_R0 (high:low) word */
+ Result is placed into SLJIT_R1:SLJIT_R0 (high:low) word */
#define SLJIT_LSMUL (SLJIT_OP0_BASE + 3)
/* Flags: I - (may destroy flags)
Unsigned divide of the value in SLJIT_R0 by the value in SLJIT_R1.
- The result is placed in SLJIT_R0 and the remainder goes to SLJIT_R1.
- Note: if SLJIT_R1 contains 0, the behaviour is undefined. */
-#define SLJIT_LUDIV (SLJIT_OP0_BASE + 4)
-#define SLJIT_ILUDIV (SLJIT_LUDIV | SLJIT_INT_OP)
+ The result is placed into SLJIT_R0 and the remainder into SLJIT_R1.
+ Note: if SLJIT_R1 is 0, the behaviour is undefined. */
+#define SLJIT_UDIVMOD (SLJIT_OP0_BASE + 4)
+#define SLJIT_IUDIVMOD (SLJIT_UDIVMOD | SLJIT_INT_OP)
/* Flags: I - (may destroy flags)
Signed divide of the value in SLJIT_R0 by the value in SLJIT_R1.
- The result is placed in SLJIT_R0 and the remainder goes to SLJIT_R1.
- Note: if SLJIT_R1 contains 0, the behaviour is undefined. */
-#define SLJIT_LSDIV (SLJIT_OP0_BASE + 5)
-#define SLJIT_ILSDIV (SLJIT_LSDIV | SLJIT_INT_OP)
+ The result is placed into SLJIT_R0 and the remainder into SLJIT_R1.
+ Note: if SLJIT_R1 is 0, the behaviour is undefined.
+ Note: if SLJIT_R1 is -1 and SLJIT_R0 is integer min (0x800..00),
+ the behaviour is undefined. */
+#define SLJIT_SDIVMOD (SLJIT_OP0_BASE + 5)
+#define SLJIT_ISDIVMOD (SLJIT_SDIVMOD | SLJIT_INT_OP)
+/* Flags: I - (may destroy flags)
+ Unsigned divide of the value in SLJIT_R0 by the value in SLJIT_R1.
+ The result is placed into SLJIT_R0. SLJIT_R1 preserves its value.
+ Note: if SLJIT_R1 is 0, the behaviour is undefined.
+ Note: SLJIT_SDIV is single precision divide. */
+#define SLJIT_UDIVI (SLJIT_OP0_BASE + 6)
+#define SLJIT_IUDIVI (SLJIT_UDIVI | SLJIT_INT_OP)
+/* Flags: I - (may destroy flags)
+ Signed divide of the value in SLJIT_R0 by the value in SLJIT_R1.
+ The result is placed into SLJIT_R0. SLJIT_R1 preserves its value.
+ Note: if SLJIT_R1 is 0, the behaviour is undefined.
+ Note: if SLJIT_R1 is -1 and SLJIT_R0 is integer min (0x800..00),
+ the behaviour is undefined.
+ Note: SLJIT_SDIV is single precision divide. */
+#define SLJIT_SDIVI (SLJIT_OP0_BASE + 7)
+#define SLJIT_ISDIVI (SLJIT_SDIVI | SLJIT_INT_OP)
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op0(struct sljit_compiler *compiler, sljit_si op);
@@ -851,34 +869,6 @@
sljit_si src1, sljit_sw src1w,
sljit_si src2, sljit_sw src2w);
-/* The following function is a helper function for sljit_emit_op_custom.
- It returns with the real machine register index ( >=0 ) of any SLJIT_R,
- SLJIT_S and SLJIT_SP registers.
-
- Note: it returns with -1 for virtual registers (only on x86-32). */
-
-SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_get_register_index(sljit_si reg);
-
-/* The following function is a helper function for sljit_emit_op_custom.
- It returns with the real machine register index of any SLJIT_FLOAT register.
-
- Note: the index is always an even number on ARM (except ARM-64), MIPS, and SPARC. */
-
-SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_get_float_register_index(sljit_si reg);
-
-/* Any instruction can be inserted into the instruction stream by
- sljit_emit_op_custom. It has a similar purpose as inline assembly.
- The size parameter must match to the instruction size of the target
- architecture:
-
- x86: 0 < size <= 15. The instruction argument can be byte aligned.
- Thumb2: if size == 2, the instruction argument must be 2 byte aligned.
- if size == 4, the instruction argument must be 4 byte aligned.
- Otherwise: size must be 4 and instruction argument must be 4 byte aligned. */
-
-SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op_custom(struct sljit_compiler *compiler,
- void *instruction, sljit_si size);
-
/* Returns with non-zero if fpu is available. */
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_is_fpu_available(void);
@@ -1196,4 +1186,64 @@
#endif /* !(defined SLJIT_INDIRECT_CALL && SLJIT_INDIRECT_CALL) */
+/* --------------------------------------------------------------------- */
+/* CPU specific functions */
+/* --------------------------------------------------------------------- */
+
+/* The following function is a helper function for sljit_emit_op_custom.
+ It returns with the real machine register index ( >=0 ) of any SLJIT_R,
+ SLJIT_S and SLJIT_SP registers.
+
+ Note: it returns with -1 for virtual registers (only on x86-32). */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_get_register_index(sljit_si reg);
+
+/* The following function is a helper function for sljit_emit_op_custom.
+ It returns with the real machine register index of any SLJIT_FLOAT register.
+
+ Note: the index is always an even number on ARM (except ARM-64), MIPS, and SPARC. */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_get_float_register_index(sljit_si reg);
+
+/* Any instruction can be inserted into the instruction stream by
+ sljit_emit_op_custom. It has a similar purpose as inline assembly.
+ The size parameter must match to the instruction size of the target
+ architecture:
+
+ x86: 0 < size <= 15. The instruction argument can be byte aligned.
+ Thumb2: if size == 2, the instruction argument must be 2 byte aligned.
+ if size == 4, the instruction argument must be 4 byte aligned.
+ Otherwise: size must be 4 and instruction argument must be 4 byte aligned. */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op_custom(struct sljit_compiler *compiler,
+ void *instruction, sljit_si size);
+
+#if (defined SLJIT_CONFIG_X86 && SLJIT_CONFIG_X86)
+
+/* Returns with non-zero if sse2 is available. */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_is_sse2_available(void);
+
+/* Returns with non-zero if cmov instruction is available. */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_is_cmov_available(void);
+
+/* Emit a conditional mov instruction on x86 CPUs. This instruction
+ moves src to destination, if the condition is satisfied. Unlike
+ other arithmetic instructions, destination must be a register.
+ Before such instructions are emitted, cmov support should be
+ checked by sljit_x86_is_cmov_available function.
+ type must be between SLJIT_EQUAL and SLJIT_S_ORDERED
+ dst_reg must be a valid register and it can be combined
+ with SLJIT_INT_OP to perform 32 bit arithmetic
+ Flags: I - (never set any flags)
+ */
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_emit_cmov(struct sljit_compiler *compiler,
+ sljit_si type,
+ sljit_si dst_reg,
+ sljit_si src, sljit_sw srcw);
+
+#endif
+
#endif /* _SLJIT_LIR_H_ */
Index: sljit/sljitNativeARM_64.c
===================================================================
--- sljit/sljitNativeARM_64.c (revision 1554)
+++ sljit/sljitNativeARM_64.c (working copy)
@@ -1087,7 +1087,8 @@
saved_regs_size += sizeof(sljit_sw);
}
local_size -= saved_regs_size + SLJIT_LOCALS_OFFSET;
- FAIL_IF(push_inst(compiler, SUBI | RD(TMP_SP) | RN(TMP_SP) | (saved_regs_size << 10)));
+ if (saved_regs_size > 0)
+ FAIL_IF(push_inst(compiler, SUBI | RD(TMP_SP) | RN(TMP_SP) | (saved_regs_size << 10)));
}
tmp = saveds < SLJIT_NUMBER_OF_SAVED_REGISTERS ? (SLJIT_S0 + 1 - saveds) : SLJIT_FIRST_SAVED_REG;
@@ -1094,7 +1095,12 @@
prev = -1;
for (i = SLJIT_S0; i >= tmp; i--) {
if (prev == -1) {
- prev = i;
+ if (!(offs & (1 << 15))) {
+ prev = i;
+ continue;
+ }
+ FAIL_IF(push_inst(compiler, STRI | RT(i) | RN(TMP_SP) | (offs >> 5)));
+ offs += 1 << 15;
continue;
}
FAIL_IF(push_inst(compiler, STP | RT(prev) | RT2(i) | RN(TMP_SP) | offs));
@@ -1104,7 +1110,12 @@
for (i = scratches; i >= SLJIT_FIRST_SAVED_REG; i--) {
if (prev == -1) {
- prev = i;
+ if (!(offs & (1 << 15))) {
+ prev = i;
+ continue;
+ }
+ FAIL_IF(push_inst(compiler, STRI | RT(i) | RN(TMP_SP) | (offs >> 5)));
+ offs += 1 << 15;
continue;
}
FAIL_IF(push_inst(compiler, STP | RT(prev) | RT2(i) | RN(TMP_SP) | offs));
@@ -1112,8 +1123,7 @@
prev = -1;
}
- if (prev != -1)
- FAIL_IF(push_inst(compiler, STRI | RT(prev) | RN(TMP_SP) | (offs >> 5)));
+ SLJIT_ASSERT(prev == -1);
if (compiler->local_size > (63 * sizeof(sljit_sw))) {
/* The local_size is already adjusted by the saved registers. */
@@ -1188,7 +1198,12 @@
prev = -1;
for (i = SLJIT_S0; i >= tmp; i--) {
if (prev == -1) {
- prev = i;
+ if (!(offs & (1 << 15))) {
+ prev = i;
+ continue;
+ }
+ FAIL_IF(push_inst(compiler, LDRI | RT(i) | RN(TMP_SP) | (offs >> 5)));
+ offs += 1 << 15;
continue;
}
FAIL_IF(push_inst(compiler, LDP | RT(prev) | RT2(i) | RN(TMP_SP) | offs));
@@ -1198,7 +1213,12 @@
for (i = compiler->scratches; i >= SLJIT_FIRST_SAVED_REG; i--) {
if (prev == -1) {
- prev = i;
+ if (!(offs & (1 << 15))) {
+ prev = i;
+ continue;
+ }
+ FAIL_IF(push_inst(compiler, LDRI | RT(i) | RN(TMP_SP) | (offs >> 5)));
+ offs += 1 << 15;
continue;
}
FAIL_IF(push_inst(compiler, LDP | RT(prev) | RT2(i) | RN(TMP_SP) | offs));
@@ -1206,13 +1226,12 @@
prev = -1;
}
- if (prev != -1)
- FAIL_IF(push_inst(compiler, LDRI | RT(prev) | RN(TMP_SP) | (offs >> 5)));
+ SLJIT_ASSERT(prev == -1);
if (compiler->local_size <= (63 * sizeof(sljit_sw))) {
FAIL_IF(push_inst(compiler, LDP_PST | 29 | RT2(TMP_LR)
| RN(TMP_SP) | (((local_size >> 3) & 0x7f) << 15)));
- } else {
+ } else if (saved_regs_size > 0) {
FAIL_IF(push_inst(compiler, ADDI | RD(TMP_SP) | RN(TMP_SP) | (saved_regs_size << 10)));
}
@@ -1242,12 +1261,15 @@
FAIL_IF(push_inst(compiler, ORR | RD(TMP_REG1) | RN(TMP_ZERO) | RM(SLJIT_R0)));
FAIL_IF(push_inst(compiler, MADD | RD(SLJIT_R0) | RN(SLJIT_R0) | RM(SLJIT_R1) | RT2(TMP_ZERO)));
return push_inst(compiler, (op == SLJIT_LUMUL ? UMULH : SMULH) | RD(SLJIT_R1) | RN(TMP_REG1) | RM(SLJIT_R1));
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
FAIL_IF(push_inst(compiler, (ORR ^ inv_bits) | RD(TMP_REG1) | RN(TMP_ZERO) | RM(SLJIT_R0)));
- FAIL_IF(push_inst(compiler, ((op == SLJIT_LUDIV ? UDIV : SDIV) ^ inv_bits) | RD(SLJIT_R0) | RN(SLJIT_R0) | RM(SLJIT_R1)));
+ FAIL_IF(push_inst(compiler, ((op == SLJIT_UDIVMOD ? UDIV : SDIV) ^ inv_bits) | RD(SLJIT_R0) | RN(SLJIT_R0) | RM(SLJIT_R1)));
FAIL_IF(push_inst(compiler, (MADD ^ inv_bits) | RD(SLJIT_R1) | RN(SLJIT_R0) | RM(SLJIT_R1) | RT2(TMP_ZERO)));
return push_inst(compiler, (SUB ^ inv_bits) | RD(SLJIT_R1) | RN(TMP_REG1) | RM(SLJIT_R1));
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+ return push_inst(compiler, ((op == SLJIT_UDIVI ? UDIV : SDIV) ^ inv_bits) | RD(SLJIT_R0) | RN(SLJIT_R0) | RM(SLJIT_R1));
}
return SLJIT_SUCCESS;
Index: sljit/sljitNativeARM_T2_32.c
===================================================================
--- sljit/sljitNativeARM_T2_32.c (revision 1554)
+++ sljit/sljitNativeARM_T2_32.c (working copy)
@@ -1239,6 +1239,9 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op0(struct sljit_compiler *compiler, sljit_si op)
{
+ sljit_sw saved_reg_list[3];
+ sljit_sw saved_reg_count;
+
CHECK_ERROR();
CHECK(check_sljit_emit_op0(compiler, op));
@@ -1255,24 +1258,53 @@
| (reg_map[SLJIT_R0] << 12)
| (reg_map[SLJIT_R0] << 16)
| reg_map[SLJIT_R1]);
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
- if (compiler->scratches >= 4) {
- FAIL_IF(push_inst32(compiler, 0xf84d2d04 /* str r2, [sp, #-4]! */));
- FAIL_IF(push_inst32(compiler, 0xf84dcd04 /* str ip, [sp, #-4]! */));
- } else if (compiler->scratches >= 3)
- FAIL_IF(push_inst32(compiler, 0xf84d2d08 /* str r2, [sp, #-8]! */));
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+ SLJIT_COMPILE_ASSERT((SLJIT_UDIVMOD & 0x2) == 0 && SLJIT_UDIVI - 0x2 == SLJIT_UDIVMOD, bad_div_opcode_assignments);
+ SLJIT_COMPILE_ASSERT(reg_map[2] == 1 && reg_map[3] == 2 && reg_map[4] == 12, bad_register_mapping);
+
+ saved_reg_count = 0;
+ if (compiler->scratches >= 4)
+ saved_reg_list[saved_reg_count++] = 12;
+ if (compiler->scratches >= 3)
+ saved_reg_list[saved_reg_count++] = 2;
+ if (op >= SLJIT_UDIVI)
+ saved_reg_list[saved_reg_count++] = 1;
+
+ if (saved_reg_count > 0) {
+ FAIL_IF(push_inst32(compiler, 0xf84d0d00 | (saved_reg_count >= 3 ? 16 : 8)
+ | (saved_reg_list[0] << 12) /* str rX, [sp, #-8/-16]! */));
+ if (saved_reg_count >= 2) {
+ SLJIT_ASSERT(saved_reg_list[1] < 8);
+ FAIL_IF(push_inst16(compiler, 0x9001 | (saved_reg_list[1] << 8) /* str rX, [sp, #4] */));
+ }
+ if (saved_reg_count >= 3) {
+ SLJIT_ASSERT(saved_reg_list[2] < 8);
+ FAIL_IF(push_inst16(compiler, 0x9002 | (saved_reg_list[2] << 8) /* str rX, [sp, #8] */));
+ }
+ }
+
#if defined(__GNUC__)
FAIL_IF(sljit_emit_ijump(compiler, SLJIT_FAST_CALL, SLJIT_IMM,
- (op == SLJIT_LUDIV ? SLJIT_FUNC_OFFSET(__aeabi_uidivmod) : SLJIT_FUNC_OFFSET(__aeabi_idivmod))));
+ ((op | 0x2) == SLJIT_UDIVI ? SLJIT_FUNC_OFFSET(__aeabi_uidivmod) : SLJIT_FUNC_OFFSET(__aeabi_idivmod))));
#else
#error "Software divmod functions are needed"
#endif
- if (compiler->scratches >= 4) {
- FAIL_IF(push_inst32(compiler, 0xf85dcb04 /* ldr ip, [sp], #4 */));
- return push_inst32(compiler, 0xf85d2b04 /* ldr r2, [sp], #4 */);
- } else if (compiler->scratches >= 3)
- return push_inst32(compiler, 0xf85d2b08 /* ldr r2, [sp], #8 */);
+
+ if (saved_reg_count > 0) {
+ if (saved_reg_count >= 3) {
+ SLJIT_ASSERT(saved_reg_list[2] < 8);
+ FAIL_IF(push_inst16(compiler, 0x9802 | (saved_reg_list[2] << 8) /* ldr rX, [sp, #8] */));
+ }
+ if (saved_reg_count >= 2) {
+ SLJIT_ASSERT(saved_reg_list[1] < 8);
+ FAIL_IF(push_inst16(compiler, 0x9801 | (saved_reg_list[1] << 8) /* ldr rX, [sp, #4] */));
+ }
+ return push_inst32(compiler, 0xf85d0b00 | (saved_reg_count >= 3 ? 16 : 8)
+ | (saved_reg_list[0] << 12) /* ldr rX, [sp], #8/16 */);
+ }
return SLJIT_SUCCESS;
}
Index: sljit/sljitNativePPC_common.c
===================================================================
--- sljit/sljitNativePPC_common.c (revision 1554)
+++ sljit/sljitNativePPC_common.c (working copy)
@@ -1267,22 +1267,23 @@
FAIL_IF(push_inst(compiler, MULLW | D(SLJIT_R0) | A(TMP_REG1) | B(SLJIT_R1)));
return push_inst(compiler, (op == SLJIT_LUMUL ? MULHWU : MULHW) | D(SLJIT_R1) | A(TMP_REG1) | B(SLJIT_R1));
#endif
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
FAIL_IF(push_inst(compiler, OR | S(SLJIT_R0) | A(TMP_REG1) | B(SLJIT_R0)));
#if (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64)
- if (int_op) {
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DIVWU : DIVW) | D(SLJIT_R0) | A(TMP_REG1) | B(SLJIT_R1)));
- FAIL_IF(push_inst(compiler, MULLW | D(SLJIT_R1) | A(SLJIT_R0) | B(SLJIT_R1)));
- } else {
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DIVDU : DIVD) | D(SLJIT_R0) | A(TMP_REG1) | B(SLJIT_R1)));
- FAIL_IF(push_inst(compiler, MULLD | D(SLJIT_R1) | A(SLJIT_R0) | B(SLJIT_R1)));
- }
- return push_inst(compiler, SUBF | D(SLJIT_R1) | A(SLJIT_R1) | B(TMP_REG1));
+ FAIL_IF(push_inst(compiler, (int_op ? (op == SLJIT_UDIVMOD ? DIVWU : DIVW) : (op == SLJIT_UDIVMOD ? DIVDU : DIVD)) | D(SLJIT_R0) | A(SLJIT_R0) | B(SLJIT_R1)));
+ FAIL_IF(push_inst(compiler, (int_op ? MULLW : MULLD) | D(SLJIT_R1) | A(SLJIT_R0) | B(SLJIT_R1)));
#else
- FAIL_IF(push_inst(compiler, (op == SLJIT_LUDIV ? DIVWU : DIVW) | D(SLJIT_R0) | A(TMP_REG1) | B(SLJIT_R1)));
+ FAIL_IF(push_inst(compiler, (op == SLJIT_UDIVMOD ? DIVWU : DIVW) | D(SLJIT_R0) | A(SLJIT_R0) | B(SLJIT_R1)));
FAIL_IF(push_inst(compiler, MULLW | D(SLJIT_R1) | A(SLJIT_R0) | B(SLJIT_R1)));
+#endif
return push_inst(compiler, SUBF | D(SLJIT_R1) | A(SLJIT_R1) | B(TMP_REG1));
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
+#if (defined SLJIT_CONFIG_PPC_64 && SLJIT_CONFIG_PPC_64)
+ return push_inst(compiler, (int_op ? (op == SLJIT_UDIVI ? DIVWU : DIVW) : (op == SLJIT_UDIVI ? DIVDU : DIVD)) | D(SLJIT_R0) | A(SLJIT_R0) | B(SLJIT_R1));
+#else
+ return push_inst(compiler, (op == SLJIT_UDIVI ? DIVWU : DIVW) | D(SLJIT_R0) | A(SLJIT_R0) | B(SLJIT_R1));
#endif
}
Index: sljit/sljitNativeX86_common.c
===================================================================
--- sljit/sljitNativeX86_common.c (revision 1554)
+++ sljit/sljitNativeX86_common.c (working copy)
@@ -273,7 +273,9 @@
#endif
static sljit_si cpu_has_cmov = -1;
-#if defined(_MSC_VER) && _MSC_VER >= 1400
+#ifdef _WIN32_WCE
+#include <cmnintrin.h>
+#elif defined(_MSC_VER) && _MSC_VER >= 1400
#include <intrin.h>
#endif
@@ -742,8 +744,10 @@
break;
case SLJIT_LUMUL:
case SLJIT_LSMUL:
- case SLJIT_LUDIV:
- case SLJIT_LSDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_SDIVMOD:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
compiler->flags_saved = 0;
#if (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
#ifdef _WIN64
@@ -761,9 +765,10 @@
#endif
compiler->mode32 = op & SLJIT_INT_OP;
#endif
+ SLJIT_COMPILE_ASSERT((SLJIT_UDIVMOD & 0x2) == 0 && SLJIT_UDIVI - 0x2 == SLJIT_UDIVMOD, bad_div_opcode_assignments);
op = GET_OPCODE(op);
- if (op == SLJIT_LUDIV) {
+ if ((op | 0x2) == SLJIT_UDIVI) {
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) || defined(_WIN64)
EMIT_MOV(compiler, TMP_REG1, 0, SLJIT_R1, 0);
inst = emit_x86_instruction(compiler, 1, SLJIT_R1, 0, SLJIT_R1, 0);
@@ -774,7 +779,7 @@
*inst = XOR_r_rm;
}
- if (op == SLJIT_LSDIV) {
+ if ((op | 0x2) == SLJIT_SDIVI) {
#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32) || defined(_WIN64)
EMIT_MOV(compiler, TMP_REG1, 0, SLJIT_R1, 0);
#endif
@@ -805,10 +810,10 @@
FAIL_IF(!inst);
INC_SIZE(2);
*inst++ = GROUP_F7;
- *inst = MOD_REG | ((op >= SLJIT_LUDIV) ? reg_map[TMP_REG1] : reg_map[SLJIT_R1]);
+ *inst = MOD_REG | ((op >= SLJIT_UDIVMOD) ? reg_map[TMP_REG1] : reg_map[SLJIT_R1]);
#else
#ifdef _WIN64
- size = (!compiler->mode32 || op >= SLJIT_LUDIV) ? 3 : 2;
+ size = (!compiler->mode32 || op >= SLJIT_UDIVMOD) ? 3 : 2;
#else
size = (!compiler->mode32) ? 3 : 2;
#endif
@@ -817,11 +822,11 @@
INC_SIZE(size);
#ifdef _WIN64
if (!compiler->mode32)
- *inst++ = REX_W | ((op >= SLJIT_LUDIV) ? REX_B : 0);
- else if (op >= SLJIT_LUDIV)
+ *inst++ = REX_W | ((op >= SLJIT_UDIVMOD) ? REX_B : 0);
+ else if (op >= SLJIT_UDIVMOD)
*inst++ = REX_B;
*inst++ = GROUP_F7;
- *inst = MOD_REG | ((op >= SLJIT_LUDIV) ? reg_lmap[TMP_REG1] : reg_lmap[SLJIT_R1]);
+ *inst = MOD_REG | ((op >= SLJIT_UDIVMOD) ? reg_lmap[TMP_REG1] : reg_lmap[SLJIT_R1]);
#else
if (!compiler->mode32)
*inst++ = REX_W;
@@ -836,15 +841,21 @@
case SLJIT_LSMUL:
*inst |= IMUL;
break;
- case SLJIT_LUDIV:
+ case SLJIT_UDIVMOD:
+ case SLJIT_UDIVI:
*inst |= DIV;
break;
- case SLJIT_LSDIV:
+ case SLJIT_SDIVMOD:
+ case SLJIT_SDIVI:
*inst |= IDIV;
break;
}
#if (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64) && !defined(_WIN64)
- EMIT_MOV(compiler, SLJIT_R1, 0, TMP_REG1, 0);
+ if (op <= SLJIT_SDIVMOD)
+ EMIT_MOV(compiler, SLJIT_R1, 0, TMP_REG1, 0);
+#else
+ if (op >= SLJIT_UDIVI)
+ EMIT_MOV(compiler, SLJIT_R1, 0, TMP_REG1, 0);
#endif
break;
}
@@ -1905,60 +1916,62 @@
return SLJIT_SUCCESS;
}
- if (FAST_IS_REG(src1)) {
+ if (!(src1 & SLJIT_IMM)) {
if (src2 & SLJIT_IMM) {
#if (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
if (IS_HALFWORD(src2w) || compiler->mode32) {
- inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src2w, src1, 0);
+ inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src2w, src1, src1w);
FAIL_IF(!inst);
*inst = GROUP_F7;
}
else {
FAIL_IF(emit_load_imm64(compiler, TMP_REG2, src2w));
- inst = emit_x86_instruction(compiler, 1, TMP_REG2, 0, src1, 0);
+ inst = emit_x86_instruction(compiler, 1, TMP_REG2, 0, src1, src1w);
FAIL_IF(!inst);
*inst = TEST_rm_r;
}
#else
- inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src2w, src1, 0);
+ inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src2w, src1, src1w);
FAIL_IF(!inst);
*inst = GROUP_F7;
#endif
+ return SLJIT_SUCCESS;
}
- else {
+ else if (FAST_IS_REG(src1)) {
inst = emit_x86_instruction(compiler, 1, src1, 0, src2, src2w);
FAIL_IF(!inst);
*inst = TEST_rm_r;
+ return SLJIT_SUCCESS;
}
- return SLJIT_SUCCESS;
}
- if (FAST_IS_REG(src2)) {
+ if (!(src2 & SLJIT_IMM)) {
if (src1 & SLJIT_IMM) {
#if (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
if (IS_HALFWORD(src1w) || compiler->mode32) {
- inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src1w, src2, 0);
+ inst = emit_x86_instruction(compiler, 1, SLJIT_IMM, src1w, src2, src2w);
FAIL_IF(!inst);
*inst = GROUP_F7;
}
else {
FAIL_IF(emit_load_imm64(compiler, TMP_REG2, src1w));
- inst = emit_x86_instruction(compiler, 1, TMP_REG2, 0, src2, 0);
+ inst = emit_x86_instruction(compiler, 1, TMP_REG2, 0, src2, src2w);
FAIL_IF(!inst);
*inst = TEST_rm_r;
}
#else
- inst = emit_x86_instruction(compiler, 1, src1, src1w, src2, 0);
+ inst = emit_x86_instruction(compiler, 1, src1, src1w, src2, src2w);
FAIL_IF(!inst);
*inst = GROUP_F7;
#endif
+ return SLJIT_SUCCESS;
}
- else {
+ else if (FAST_IS_REG(src2)) {
inst = emit_x86_instruction(compiler, 1, src2, 0, src1, src1w);
FAIL_IF(!inst);
*inst = TEST_rm_r;
+ return SLJIT_SUCCESS;
}
- return SLJIT_SUCCESS;
}
EMIT_MOV(compiler, TMP_REG1, 0, src1, src1w);
@@ -2923,3 +2936,69 @@
{
*(sljit_sw*)addr = new_constant;
}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_is_sse2_available(void)
+{
+#if (defined SLJIT_DETECT_SSE2 && SLJIT_DETECT_SSE2)
+ if (cpu_has_sse2 == -1)
+ get_cpu_features();
+ return cpu_has_sse2;
+#else
+ return 1;
+#endif
+}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_is_cmov_available(void)
+{
+ if (cpu_has_cmov == -1)
+ get_cpu_features();
+ return cpu_has_cmov;
+}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_x86_emit_cmov(struct sljit_compiler *compiler,
+ sljit_si type,
+ sljit_si dst_reg,
+ sljit_si src, sljit_sw srcw)
+{
+ sljit_ub* inst;
+
+ CHECK_ERROR();
+#if (defined SLJIT_ARGUMENT_CHECKS && SLJIT_ARGUMENT_CHECKS)
+ CHECK_ARGUMENT(sljit_x86_is_cmov_available());
+ CHECK_ARGUMENT(!(type & ~(0xff | SLJIT_INT_OP)));
+ CHECK_ARGUMENT((type & 0xff) >= SLJIT_EQUAL && (type & 0xff) <= SLJIT_D_ORDERED);
+ CHECK_ARGUMENT(FUNCTION_CHECK_IS_REG(dst_reg & ~SLJIT_INT_OP));
+ FUNCTION_CHECK_SRC(src, srcw);
+#endif
+#if (defined SLJIT_VERBOSE && SLJIT_VERBOSE)
+ if (SLJIT_UNLIKELY(!!compiler->verbose)) {
+ fprintf(compiler->verbose, " x86_cmov%s %s%s, ",
+ !(dst_reg & SLJIT_INT_OP) ? "" : ".i",
+ JUMP_PREFIX(type), jump_names[type & 0xff]);
+ sljit_verbose_reg(compiler, dst_reg & ~SLJIT_INT_OP);
+ fprintf(compiler->verbose, ", ");
+ sljit_verbose_param(compiler, src, srcw);
+ fprintf(compiler->verbose, "\n");
+ }
+#endif
+
+ ADJUST_LOCAL_OFFSET(src, srcw);
+ CHECK_EXTRA_REGS(src, srcw, (void)0);
+
+#if (defined SLJIT_CONFIG_X86_64 && SLJIT_CONFIG_X86_64)
+ compiler->mode32 = dst_reg & SLJIT_INT_OP;
+#endif
+ dst_reg &= ~SLJIT_INT_OP;
+
+ if (SLJIT_UNLIKELY(src & SLJIT_IMM)) {
+ EMIT_MOV(compiler, TMP_REG1, 0, SLJIT_IMM, srcw);
+ src = TMP_REG1;
+ srcw = 0;
+ }
+
+ inst = emit_x86_instruction(compiler, 2, dst_reg, 0, src, srcw);
+ FAIL_IF(!inst);
+ *inst++ = GROUP_0F;
+ *inst = get_jump_code(type & 0xff) - 0x40;
+ return SLJIT_SUCCESS;
+}
Index: sljit/sljitConfigInternal.h
===================================================================
--- sljit/sljitConfigInternal.h (revision 1554)
+++ sljit/sljitConfigInternal.h (working copy)
@@ -468,8 +468,13 @@
#ifndef SLJIT_CALL
-#if (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32)
+#if (defined SLJIT_USE_CDECL_CALLING_CONVENTION && SLJIT_USE_CDECL_CALLING_CONVENTION)
+/* Force cdecl. */
+#define SLJIT_CALL
+
+#elif (defined SLJIT_CONFIG_X86_32 && SLJIT_CONFIG_X86_32)
+
#if defined(__GNUC__) && !defined(__APPLE__)
#define SLJIT_CALL __attribute__ ((fastcall))
@@ -608,6 +613,12 @@
#define SLJIT_LOCALS_OFFSET_BASE ((23 + 1) * sizeof(sljit_sw))
#endif
+#elif (defined SLJIT_CONFIG_TILEGX && SLJIT_CONFIG_TILEGX)
+
+#define SLJIT_NUMBER_OF_REGISTERS 10
+#define SLJIT_NUMBER_OF_SAVED_REGISTERS 5
+#define SLJIT_LOCALS_OFFSET_BASE 0
+
#elif (defined SLJIT_CONFIG_UNSUPPORTED && SLJIT_CONFIG_UNSUPPORTED)
#define SLJIT_NUMBER_OF_REGISTERS 0
Index: sljit/sljitConfig.h
===================================================================
--- sljit/sljitConfig.h (revision 1554)
+++ sljit/sljitConfig.h (working copy)
@@ -96,6 +96,15 @@
#define SLJIT_EXECUTABLE_ALLOCATOR 1
#endif
+/* Force cdecl calling convention even if a better calling
+ convention (e.g. fastcall) is supported by the C compiler.
+ If this option is enabled, C functions without
+ SLJIT_CALL can also be called from JIT code. */
+#ifndef SLJIT_USE_CDECL_CALLING_CONVENTION
+/* Disabled by default */
+#define SLJIT_USE_CDECL_CALLING_CONVENTION 0
+#endif
+
/* Return with error when an invalid argument is passed. */
#ifndef SLJIT_ARGUMENT_CHECKS
/* Disabled by default */
Index: sljit/sljitNativeTILEGX_64.c
===================================================================
--- sljit/sljitNativeTILEGX_64.c (revision 1554)
+++ sljit/sljitNativeTILEGX_64.c (working copy)
@@ -35,21 +35,21 @@
#define SIMM_16BIT_MIN (-0x8000)
#define SIMM_17BIT_MAX (0xffff)
#define SIMM_17BIT_MIN (-0x10000)
-#define SIMM_32BIT_MIN (-0x80000000)
#define SIMM_32BIT_MAX (0x7fffffff)
-#define SIMM_48BIT_MIN (0x800000000000L)
+#define SIMM_32BIT_MIN (-0x7fffffff - 1)
#define SIMM_48BIT_MAX (0x7fffffff0000L)
+#define SIMM_48BIT_MIN (-0x800000000000L)
#define IMM16(imm) ((imm) & 0xffff)
#define UIMM_16BIT_MAX (0xffff)
-#define TMP_REG1 (SLJIT_NO_REGISTERS + 1)
-#define TMP_REG2 (SLJIT_NO_REGISTERS + 2)
-#define TMP_REG3 (SLJIT_NO_REGISTERS + 3)
-#define ADDR_TMP (SLJIT_NO_REGISTERS + 4)
+#define TMP_REG1 (SLJIT_NUMBER_OF_REGISTERS + 2)
+#define TMP_REG2 (SLJIT_NUMBER_OF_REGISTERS + 3)
+#define TMP_REG3 (SLJIT_NUMBER_OF_REGISTERS + 4)
+#define ADDR_TMP (SLJIT_NUMBER_OF_REGISTERS + 5)
#define PIC_ADDR_REG TMP_REG2
-static SLJIT_CONST sljit_ub reg_map[SLJIT_NO_REGISTERS + 5] = {
+static SLJIT_CONST sljit_ub reg_map[SLJIT_NUMBER_OF_REGISTERS + 6] = {
63, 0, 1, 2, 3, 4, 30, 31, 32, 33, 34, 54, 5, 16, 6, 7
};
@@ -58,11 +58,6 @@
#define TMP_REG2_mapped 16
#define TMP_REG3_mapped 6
#define ADDR_TMP_mapped 7
-#define SLJIT_SAVED_REG1_mapped 30
-#define SLJIT_SAVED_REG2_mapped 31
-#define SLJIT_SAVED_REG3_mapped 32
-#define SLJIT_SAVED_EREG1_mapped 33
-#define SLJIT_SAVED_EREG2_mapped 34
/* Flags are keept in volatile registers. */
#define EQUAL_FLAG 8
@@ -399,6 +394,9 @@
#define SUB(dst, srca, srcb) \
push_3_buffer(compiler, TILEGX_OPC_SUB, dst, srca, srcb, __LINE__)
+#define MUL(dst, srca, srcb) \
+ push_3_buffer(compiler, TILEGX_OPC_MULX, dst, srca, srcb, __LINE__)
+
#define NOR(dst, srca, srcb) \
push_3_buffer(compiler, TILEGX_OPC_NOR, dst, srca, srcb, __LINE__)
@@ -547,8 +545,8 @@
const struct Format* match = NULL;
const struct Format *b = NULL;
- unsigned int i = 0;
- for (i; i < sizeof formats / sizeof formats[0]; i++) {
+ unsigned int i;
+ for (i = 0; i < sizeof formats / sizeof formats[0]; i++) {
b = &formats[i];
if ((b->pipe_mask & compatible_pipes) == b->pipe_mask) {
match = b;
@@ -625,7 +623,6 @@
static sljit_si update_buffer(struct sljit_compiler *compiler)
{
- int count;
int i;
int orig_index = inst_buf_index;
struct jit_instr inst0 = inst_buf[0];
@@ -738,8 +735,10 @@
static sljit_si flush_buffer(struct sljit_compiler *compiler)
{
- while (inst_buf_index != 0)
- update_buffer(compiler);
+ while (inst_buf_index != 0) {
+ FAIL_IF(update_buffer(compiler));
+ }
+ return SLJIT_SUCCESS;
}
static sljit_si push_4_buffer(struct sljit_compiler *compiler, tilegx_mnemonic opc, int op0, int op1, int op2, int op3, int line)
@@ -787,6 +786,7 @@
case TILEGX_OPC_ADD:
case TILEGX_OPC_AND:
case TILEGX_OPC_SUB:
+ case TILEGX_OPC_MULX:
case TILEGX_OPC_OR:
case TILEGX_OPC_XOR:
case TILEGX_OPC_NOR:
@@ -905,7 +905,6 @@
sljit_sw diff;
sljit_uw target_addr;
sljit_ins *inst;
- sljit_ins saved_inst;
if (jump->flags & SLJIT_REWRITABLE_JUMP)
return code_ptr;
@@ -1009,7 +1008,7 @@
struct sljit_const *const_;
CHECK_ERROR_PTR();
- check_sljit_generate_code(compiler);
+ CHECK_PTR(check_sljit_generate_code(compiler));
reverse_buf(compiler);
code = (sljit_ins *)SLJIT_MALLOC_EXEC(compiler->size * sizeof(sljit_ins));
@@ -1178,13 +1177,13 @@
sljit_si fscratches, sljit_si fsaveds, sljit_si local_size)
{
sljit_ins base;
- sljit_ins bundle = 0;
-
+ sljit_si i, tmp;
+
CHECK_ERROR();
- check_sljit_emit_enter(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size);
+ CHECK(check_sljit_emit_enter(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size));
set_emit_enter(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size);
- local_size += (saveds + 1) * sizeof(sljit_sw);
+ local_size += GET_SAVED_REGISTERS_SIZE(scratches, saveds, 1);
local_size = (local_size + 7) & ~7;
compiler->local_size = local_size;
@@ -1200,46 +1199,41 @@
local_size = 0;
}
+ /* Save the return address. */
FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 8));
FAIL_IF(ST_ADD(ADDR_TMP_mapped, RA, -8));
- if (saveds >= 1)
- FAIL_IF(ST_ADD(ADDR_TMP_mapped, SLJIT_SAVED_REG1_mapped, -8));
+ /* Save the S registers. */
+ tmp = saveds < SLJIT_NUMBER_OF_SAVED_REGISTERS ? (SLJIT_S0 + 1 - saveds) : SLJIT_FIRST_SAVED_REG;
+ for (i = SLJIT_S0; i >= tmp; i--) {
+ FAIL_IF(ST_ADD(ADDR_TMP_mapped, reg_map[i], -8));
+ }
- if (saveds >= 2)
- FAIL_IF(ST_ADD(ADDR_TMP_mapped, SLJIT_SAVED_REG2_mapped, -8));
+ /* Save the R registers that need to be reserved. */
+ for (i = scratches; i >= SLJIT_FIRST_SAVED_REG; i--) {
+ FAIL_IF(ST_ADD(ADDR_TMP_mapped, reg_map[i], -8));
+ }
- if (saveds >= 3)
- FAIL_IF(ST_ADD(ADDR_TMP_mapped, SLJIT_SAVED_REG3_mapped, -8));
+ /* Move the arguments to S registers. */
+ for (i = 0; i < args; i++) {
+ FAIL_IF(ADD(reg_map[SLJIT_S0 - i], i, ZERO));
+ }
- if (saveds >= 4)
- FAIL_IF(ST_ADD(ADDR_TMP_mapped, SLJIT_SAVED_EREG1_mapped, -8));
-
- if (saveds >= 5)
- FAIL_IF(ST_ADD(ADDR_TMP_mapped, SLJIT_SAVED_EREG2_mapped, -8));
-
- if (args >= 1)
- FAIL_IF(ADD(SLJIT_SAVED_REG1_mapped, 0, ZERO));
-
- if (args >= 2)
- FAIL_IF(ADD(SLJIT_SAVED_REG2_mapped, 1, ZERO));
-
- if (args >= 3)
- FAIL_IF(ADD(SLJIT_SAVED_REG3_mapped, 2, ZERO));
-
return SLJIT_SUCCESS;
}
-SLJIT_API_FUNC_ATTRIBUTE void sljit_set_context(struct sljit_compiler *compiler,
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_set_context(struct sljit_compiler *compiler,
sljit_si options, sljit_si args, sljit_si scratches, sljit_si saveds,
sljit_si fscratches, sljit_si fsaveds, sljit_si local_size)
{
- CHECK_ERROR_VOID();
- check_sljit_set_context(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size);
+ CHECK_ERROR();
+ CHECK(check_sljit_set_context(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size));
set_set_context(compiler, options, args, scratches, saveds, fscratches, fsaveds, local_size);
- local_size += (saveds + 1) * sizeof(sljit_sw);
+ local_size += GET_SAVED_REGISTERS_SIZE(scratches, saveds, 1);
compiler->local_size = (local_size + 7) & ~7;
+
+ return SLJIT_SUCCESS;
}
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_return(struct sljit_compiler *compiler, sljit_si op, sljit_si src, sljit_sw srcw)
@@ -1246,10 +1240,11 @@
{
sljit_si local_size;
sljit_ins base;
- int addr_initialized = 0;
+ sljit_si i, tmp;
+ sljit_si saveds;
CHECK_ERROR();
- check_sljit_emit_return(compiler, op, src, srcw);
+ CHECK(check_sljit_emit_return(compiler, op, src, srcw));
FAIL_IF(emit_mov_before_return(compiler, op, src, srcw));
@@ -1263,52 +1258,22 @@
local_size = 0;
}
+ /* Restore the return address. */
FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 8));
- FAIL_IF(LD(RA, ADDR_TMP_mapped));
+ FAIL_IF(LD_ADD(RA, ADDR_TMP_mapped, -8));
- if (compiler->saveds >= 5) {
- FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 48));
- addr_initialized = 1;
-
- FAIL_IF(LD_ADD(SLJIT_SAVED_EREG2_mapped, ADDR_TMP_mapped, 8));
+ /* Restore the S registers. */
+ saveds = compiler->saveds;
+ tmp = saveds < SLJIT_NUMBER_OF_SAVED_REGISTERS ? (SLJIT_S0 + 1 - saveds) : SLJIT_FIRST_SAVED_REG;
+ for (i = SLJIT_S0; i >= tmp; i--) {
+ FAIL_IF(LD_ADD(reg_map[i], ADDR_TMP_mapped, -8));
}
- if (compiler->saveds >= 4) {
- if (addr_initialized == 0) {
- FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 40));
- addr_initialized = 1;
- }
-
- FAIL_IF(LD_ADD(SLJIT_SAVED_EREG1_mapped, ADDR_TMP_mapped, 8));
+ /* Restore the R registers that need to be reserved. */
+ for (i = compiler->scratches; i >= SLJIT_FIRST_SAVED_REG; i--) {
+ FAIL_IF(LD_ADD(reg_map[i], ADDR_TMP_mapped, -8));
}
- if (compiler->saveds >= 3) {
- if (addr_initialized == 0) {
- FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 32));
- addr_initialized = 1;
- }
-
- FAIL_IF(LD_ADD(SLJIT_SAVED_REG3_mapped, ADDR_TMP_mapped, 8));
- }
-
- if (compiler->saveds >= 2) {
- if (addr_initialized == 0) {
- FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 24));
- addr_initialized = 1;
- }
-
- FAIL_IF(LD_ADD(SLJIT_SAVED_REG2_mapped, ADDR_TMP_mapped, 8));
- }
-
- if (compiler->saveds >= 1) {
- if (addr_initialized == 0) {
- FAIL_IF(ADDLI(ADDR_TMP_mapped, base, local_size - 16));
- /* addr_initialized = 1; no need to initialize as it's the last one. */
- }
-
- FAIL_IF(LD_ADD(SLJIT_SAVED_REG1_mapped, ADDR_TMP_mapped, 8));
- }
-
if (compiler->local_size <= SIMM_16BIT_MAX)
FAIL_IF(ADDLI(SLJIT_LOCALS_REG_mapped, SLJIT_LOCALS_REG_mapped, compiler->local_size));
else
@@ -1585,7 +1550,7 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_enter(struct sljit_compiler *compiler, sljit_si dst, sljit_sw dstw)
{
CHECK_ERROR();
- check_sljit_emit_fast_enter(compiler, dst, dstw);
+ CHECK(check_sljit_emit_fast_enter(compiler, dst, dstw));
ADJUST_LOCAL_OFFSET(dst, dstw);
/* For UNUSED dst. Uncommon, but possible. */
@@ -1602,7 +1567,7 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_fast_return(struct sljit_compiler *compiler, sljit_si src, sljit_sw srcw)
{
CHECK_ERROR();
- check_sljit_emit_fast_return(compiler, src, srcw);
+ CHECK(check_sljit_emit_fast_return(compiler, src, srcw));
ADJUST_LOCAL_OFFSET(src, srcw);
if (FAST_IS_REG(src))
@@ -1636,9 +1601,11 @@
if (op == SLJIT_MOV_SI)
return BFEXTS(reg_map[dst], reg_map[src2], 0, 31);
- return BFEXTU(reg_map[dst], reg_map[src2], 0, 31);
- } else if (dst != src2)
- SLJIT_ASSERT_STOP();
+ return BFEXTU(reg_map[dst], reg_map[src2], 0, 31);
+ } else if (dst != src2) {
+ SLJIT_ASSERT(src2 == 0);
+ return ADD(reg_map[dst], reg_map[src2], ZERO);
+ }
return SLJIT_SUCCESS;
@@ -1650,8 +1617,10 @@
return BFEXTS(reg_map[dst], reg_map[src2], 0, 7);
return BFEXTU(reg_map[dst], reg_map[src2], 0, 7);
- } else if (dst != src2)
- SLJIT_ASSERT_STOP();
+ } else if (dst != src2) {
+ SLJIT_ASSERT(src2 == 0);
+ return ADD(reg_map[dst], reg_map[src2], ZERO);
+ }
return SLJIT_SUCCESS;
@@ -1663,8 +1632,10 @@
return BFEXTS(reg_map[dst], reg_map[src2], 0, 15);
return BFEXTU(reg_map[dst], reg_map[src2], 0, 15);
- } else if (dst != src2)
- SLJIT_ASSERT_STOP();
+ } else if (dst != src2) {
+ SLJIT_ASSERT(src2 == 0);
+ return ADD(reg_map[dst], reg_map[src2], ZERO);
+ }
return SLJIT_SUCCESS;
@@ -1811,7 +1782,6 @@
else {
/* Rare ocasion. */
FAIL_IF(ADD(TMP_EREG2, reg_map[src1], ZERO));
-
overflow_ra = TMP_EREG2;
}
}
@@ -1903,6 +1873,17 @@
return SLJIT_SUCCESS;
+ case SLJIT_MUL:
+ if (flags & SRC2_IMM) {
+ FAIL_IF(load_immediate(compiler, TMP_REG2_mapped, src2));
+ src2 = TMP_REG2;
+ flags &= ~SRC2_IMM;
+ }
+
+ FAIL_IF(MUL(reg_map[dst], reg_map[src1], reg_map[src2]));
+
+ return SLJIT_SUCCESS;
+
#define EMIT_LOGICAL(op_imm, op_norm) \
if (flags & SRC2_IMM) { \
FAIL_IF(load_immediate(compiler, ADDR_TMP_mapped, src2)); \
@@ -1950,8 +1931,8 @@
} else { \
if (op & SLJIT_SET_E) \
FAIL_IF(push_3_buffer( \
- compiler, op_imm, reg_map[dst], reg_map[src1], \
- src2 & 0x3F, __LINE__)); \
+ compiler, op_norm, EQUAL_FLAG, reg_map[src1], \
+ reg_map[src2], __LINE__)); \
if (CHECK_FLAGS(SLJIT_SET_E)) \
FAIL_IF(push_3_buffer( \
compiler, op_norm, reg_map[dst], reg_map[src1], \
@@ -2105,9 +2086,10 @@
{
sljit_si sugg_dst_ar, dst_ar;
sljit_si flags = GET_ALL_FLAGS(op);
+ sljit_si mem_type = (op & SLJIT_INT_OP) ? (INT_DATA | SIGNED_DATA) : WORD_DATA;
CHECK_ERROR();
- check_sljit_emit_op_flags(compiler, op, dst, dstw, src, srcw, type);
+ CHECK(check_sljit_emit_op_flags(compiler, op, dst, dstw, src, srcw, type));
ADJUST_LOCAL_OFFSET(dst, dstw);
if (dst == SLJIT_UNUSED)
@@ -2114,6 +2096,8 @@
return SLJIT_SUCCESS;
op = GET_OPCODE(op);
+ if (op == SLJIT_MOV_SI || op == SLJIT_MOV_UI)
+ mem_type = INT_DATA | SIGNED_DATA;
sugg_dst_ar = reg_map[(op < SLJIT_ADD && FAST_IS_REG(dst)) ? dst : TMP_REG2];
compiler->cache_arg = 0;
@@ -2120,51 +2104,43 @@
compiler->cache_argw = 0;
if (op >= SLJIT_ADD && (src & SLJIT_MEM)) {
ADJUST_LOCAL_OFFSET(src, srcw);
- FAIL_IF(emit_op_mem2(compiler, WORD_DATA | LOAD_DATA, TMP_REG1_mapped, src, srcw, dst, dstw));
+ FAIL_IF(emit_op_mem2(compiler, mem_type | LOAD_DATA, TMP_REG1_mapped, src, srcw, dst, dstw));
src = TMP_REG1;
srcw = 0;
}
- switch (type) {
- case SLJIT_C_EQUAL:
- case SLJIT_C_NOT_EQUAL:
+ switch (type & 0xff) {
+ case SLJIT_EQUAL:
+ case SLJIT_NOT_EQUAL:
FAIL_IF(CMPLTUI(sugg_dst_ar, EQUAL_FLAG, 1));
dst_ar = sugg_dst_ar;
break;
- case SLJIT_C_LESS:
- case SLJIT_C_GREATER_EQUAL:
- case SLJIT_C_FLOAT_LESS:
- case SLJIT_C_FLOAT_GREATER_EQUAL:
+ case SLJIT_LESS:
+ case SLJIT_GREATER_EQUAL:
dst_ar = ULESS_FLAG;
break;
- case SLJIT_C_GREATER:
- case SLJIT_C_LESS_EQUAL:
- case SLJIT_C_FLOAT_GREATER:
- case SLJIT_C_FLOAT_LESS_EQUAL:
+ case SLJIT_GREATER:
+ case SLJIT_LESS_EQUAL:
dst_ar = UGREATER_FLAG;
break;
- case SLJIT_C_SIG_LESS:
- case SLJIT_C_SIG_GREATER_EQUAL:
+ case SLJIT_SIG_LESS:
+ case SLJIT_SIG_GREATER_EQUAL:
dst_ar = LESS_FLAG;
break;
- case SLJIT_C_SIG_GREATER:
- case SLJIT_C_SIG_LESS_EQUAL:
+ case SLJIT_SIG_GREATER:
+ case SLJIT_SIG_LESS_EQUAL:
dst_ar = GREATER_FLAG;
break;
- case SLJIT_C_OVERFLOW:
- case SLJIT_C_NOT_OVERFLOW:
+ case SLJIT_OVERFLOW:
+ case SLJIT_NOT_OVERFLOW:
dst_ar = OVERFLOW_FLAG;
break;
- case SLJIT_C_MUL_OVERFLOW:
- case SLJIT_C_MUL_NOT_OVERFLOW:
+ case SLJIT_MUL_OVERFLOW:
+ case SLJIT_MUL_NOT_OVERFLOW:
FAIL_IF(CMPLTUI(sugg_dst_ar, OVERFLOW_FLAG, 1));
dst_ar = sugg_dst_ar;
type ^= 0x1; /* Flip type bit for the XORI below. */
break;
- case SLJIT_C_FLOAT_EQUAL:
- case SLJIT_C_FLOAT_NOT_EQUAL:
- dst_ar = EQUAL_FLAG;
- break;
default:
SLJIT_ASSERT_STOP();
@@ -2180,11 +2156,11 @@
if (op >= SLJIT_ADD) {
if (TMP_REG2_mapped != dst_ar)
FAIL_IF(ADD(TMP_REG2_mapped, dst_ar, ZERO));
- return emit_op(compiler, op | flags, CUMULATIVE_OP | LOGICAL_OP | IMM_OP | ALT_KEEP_CACHE, dst, dstw, src, srcw, TMP_REG2, 0);
+ return emit_op(compiler, op | flags, mem_type | CUMULATIVE_OP | LOGICAL_OP | IMM_OP | ALT_KEEP_CACHE, dst, dstw, src, srcw, TMP_REG2, 0);
}
if (dst & SLJIT_MEM)
- return emit_op_mem(compiler, WORD_DATA, dst_ar, dst, dstw);
+ return emit_op_mem(compiler, mem_type, dst_ar, dst, dstw);
if (sugg_dst_ar != dst_ar)
return ADD(sugg_dst_ar, dst_ar, ZERO);
@@ -2194,7 +2170,7 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op0(struct sljit_compiler *compiler, sljit_si op) {
CHECK_ERROR();
- check_sljit_emit_op0(compiler, op);
+ CHECK(check_sljit_emit_op0(compiler, op));
op = GET_OPCODE(op);
switch (op) {
@@ -2204,10 +2180,10 @@
case SLJIT_BREAKPOINT:
return PI(BPT);
- case SLJIT_UMUL:
- case SLJIT_SMUL:
- case SLJIT_UDIV:
- case SLJIT_SDIV:
+ case SLJIT_LUMUL:
+ case SLJIT_LSMUL:
+ case SLJIT_UDIVI:
+ case SLJIT_SDIVI:
SLJIT_ASSERT_STOP();
}
@@ -2217,7 +2193,7 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op1(struct sljit_compiler *compiler, sljit_si op, sljit_si dst, sljit_sw dstw, sljit_si src, sljit_sw srcw)
{
CHECK_ERROR();
- check_sljit_emit_op1(compiler, op, dst, dstw, src, srcw);
+ CHECK(check_sljit_emit_op1(compiler, op, dst, dstw, src, srcw));
ADJUST_LOCAL_OFFSET(dst, dstw);
ADJUST_LOCAL_OFFSET(src, srcw);
@@ -2273,7 +2249,7 @@
return emit_op(compiler, SLJIT_SUB | GET_ALL_FLAGS(op), IMM_OP, dst, dstw, SLJIT_IMM, 0, src, srcw);
case SLJIT_CLZ:
- return emit_op(compiler, op, 0, dst, dstw, TMP_REG1, 0, src, srcw);
+ return emit_op(compiler, op, (op & SLJIT_INT_OP) ? INT_DATA : WORD_DATA, dst, dstw, TMP_REG1, 0, src, srcw);
}
return SLJIT_SUCCESS;
@@ -2282,7 +2258,7 @@
SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op2(struct sljit_compiler *compiler, sljit_si op, sljit_si dst, sljit_sw dstw, sljit_si src1, sljit_sw src1w, sljit_si src2, sljit_sw src2w)
{
CHECK_ERROR();
- check_sljit_emit_op2(compiler, op, dst, dstw, src1, src1w, src2, src2w);
+ CHECK(check_sljit_emit_op2(compiler, op, dst, dstw, src1, src1w, src2, src2w));
ADJUST_LOCAL_OFFSET(dst, dstw);
ADJUST_LOCAL_OFFSET(src1, src1w);
ADJUST_LOCAL_OFFSET(src2, src2w);
@@ -2325,7 +2301,7 @@
flush_buffer(compiler);
CHECK_ERROR_PTR();
- check_sljit_emit_label(compiler);
+ CHECK_PTR(check_sljit_emit_label(compiler));
if (compiler->last_label && compiler->last_label->size == compiler->size)
return compiler->last_label;
@@ -2344,7 +2320,7 @@
flush_buffer(compiler);
CHECK_ERROR();
- check_sljit_emit_ijump(compiler, type, src, srcw);
+ CHECK(check_sljit_emit_ijump(compiler, type, src, srcw));
ADJUST_LOCAL_OFFSET(src, srcw);
if (FAST_IS_REG(src)) {
@@ -2404,8 +2380,10 @@
return SLJIT_SUCCESS;
- } else if (src & SLJIT_MEM)
+ } else if (src & SLJIT_MEM) {
FAIL_IF(emit_op(compiler, SLJIT_MOV, WORD_DATA, TMP_REG2, 0, TMP_REG1, 0, src, srcw));
+ flush_buffer(compiler);
+ }
FAIL_IF(JR_SOLO(reg_map[src_r]));
@@ -2432,7 +2410,7 @@
flush_buffer(compiler);
CHECK_ERROR_PTR();
- check_sljit_emit_jump(compiler, type);
+ CHECK_PTR(check_sljit_emit_jump(compiler, type));
jump = (struct sljit_jump *)ensure_abuf(compiler, sizeof(struct sljit_jump));
PTR_FAIL_IF(!jump);
@@ -2440,48 +2418,42 @@
type &= 0xff;
switch (type) {
- case SLJIT_C_EQUAL:
- case SLJIT_C_FLOAT_NOT_EQUAL:
+ case SLJIT_EQUAL:
BR_NZ(EQUAL_FLAG);
break;
- case SLJIT_C_NOT_EQUAL:
- case SLJIT_C_FLOAT_EQUAL:
+ case SLJIT_NOT_EQUAL:
BR_Z(EQUAL_FLAG);
break;
- case SLJIT_C_LESS:
- case SLJIT_C_FLOAT_LESS:
+ case SLJIT_LESS:
BR_Z(ULESS_FLAG);
break;
- case SLJIT_C_GREATER_EQUAL:
- case SLJIT_C_FLOAT_GREATER_EQUAL:
+ case SLJIT_GREATER_EQUAL:
BR_NZ(ULESS_FLAG);
break;
- case SLJIT_C_GREATER:
- case SLJIT_C_FLOAT_GREATER:
+ case SLJIT_GREATER:
BR_Z(UGREATER_FLAG);
break;
- case SLJIT_C_LESS_EQUAL:
- case SLJIT_C_FLOAT_LESS_EQUAL:
+ case SLJIT_LESS_EQUAL:
BR_NZ(UGREATER_FLAG);
break;
- case SLJIT_C_SIG_LESS:
+ case SLJIT_SIG_LESS:
BR_Z(LESS_FLAG);
break;
- case SLJIT_C_SIG_GREATER_EQUAL:
+ case SLJIT_SIG_GREATER_EQUAL:
BR_NZ(LESS_FLAG);
break;
- case SLJIT_C_SIG_GREATER:
+ case SLJIT_SIG_GREATER:
BR_Z(GREATER_FLAG);
break;
- case SLJIT_C_SIG_LESS_EQUAL:
+ case SLJIT_SIG_LESS_EQUAL:
BR_NZ(GREATER_FLAG);
break;
- case SLJIT_C_OVERFLOW:
- case SLJIT_C_MUL_OVERFLOW:
+ case SLJIT_OVERFLOW:
+ case SLJIT_MUL_OVERFLOW:
BR_Z(OVERFLOW_FLAG);
break;
- case SLJIT_C_NOT_OVERFLOW:
- case SLJIT_C_MUL_NOT_OVERFLOW:
+ case SLJIT_NOT_OVERFLOW:
+ case SLJIT_MUL_NOT_OVERFLOW:
BR_NZ(OVERFLOW_FLAG);
break;
default:
@@ -2536,7 +2508,7 @@
flush_buffer(compiler);
CHECK_ERROR_PTR();
- check_sljit_emit_const(compiler, dst, dstw, init_value);
+ CHECK_PTR(check_sljit_emit_const(compiler, dst, dstw, init_value));
ADJUST_LOCAL_OFFSET(dst, dstw);
const_ = (struct sljit_const *)ensure_abuf(compiler, sizeof(struct sljit_const));
@@ -2572,3 +2544,18 @@
inst[3] = (inst[3] & ~(0xFFFFL << 43)) | ((new_constant & 0xFFFFL) << 43);
SLJIT_CACHE_FLUSH(inst, inst + 4);
}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_get_register_index(sljit_si reg)
+{
+ CHECK_REG_INDEX(check_sljit_get_register_index(reg));
+ return reg_map[reg];
+}
+
+SLJIT_API_FUNC_ATTRIBUTE sljit_si sljit_emit_op_custom(struct sljit_compiler *compiler,
+ void *instruction, sljit_si size)
+{
+ CHECK_ERROR();
+ CHECK(check_sljit_emit_op_custom(compiler, instruction, size));
+ return SLJIT_ERR_UNSUPPORTED;
+}
+
Index: pcregrep.c
===================================================================
--- pcregrep.c (revision 1554)
+++ pcregrep.c (working copy)
@@ -1692,9 +1692,13 @@
if (filenames == FN_NOMATCH_ONLY) return 1;
+ /* If all we want is a yes/no answer, stop now. */
+
+ if (quiet) return 0;
+
/* Just count if just counting is wanted. */
- if (count_only) count++;
+ else if (count_only) count++;
/* When handling a binary file and binary-files==binary, the "binary"
variable will be set true (it's false in all other cases). In this
@@ -1715,10 +1719,6 @@
return 0;
}
- /* Likewise, if all we want is a yes/no answer. */
-
- else if (quiet) return 0;
-
/* The --only-matching option prints just the substring that matched,
and/or one or more captured portions of it, as long as these strings are
not empty. The --file-offsets and --line-offsets options output offsets for
@@ -2089,7 +2089,7 @@
/* Print the match count if wanted */
-if (count_only)
+if (count_only && !quiet)
{
if (count > 0 || !omit_zero_count)
{
Index: pcre_study.c
===================================================================
--- pcre_study.c (revision 1554)
+++ pcre_study.c (working copy)
@@ -71,6 +71,7 @@
startcode pointer to start of the whole pattern's code
options the compiling options
recurses chain of recurse_check to catch mutual recursion
+ countptr pointer to call count (to catch over complexity)
Returns: the minimum length
-1 if \C in UTF-8 mode or (*ACCEPT) was encountered
@@ -80,7 +81,8 @@
static int
find_minlength(const REAL_PCRE *re, const pcre_uchar *code,
- const pcre_uchar *startcode, int options, recurse_check *recurses)
+ const pcre_uchar *startcode, int options, recurse_check *recurses,
+ int *countptr)
{
int length = -1;
/* PCRE_UTF16 has the same value as PCRE_UTF8. */
@@ -90,6 +92,8 @@
register int branchlength = 0;
register pcre_uchar *cc = (pcre_uchar *)code + 1 + LINK_SIZE;
+if ((*countptr)++ > 1000) return -1; /* too complex */
+
if (*code == OP_CBRA || *code == OP_SCBRA ||
*code == OP_CBRAPOS || *code == OP_SCBRAPOS) cc += IMM2_SIZE;
@@ -131,7 +135,7 @@
case OP_SBRAPOS:
case OP_ONCE:
case OP_ONCE_NC:
- d = find_minlength(re, cc, startcode, options, recurses);
+ d = find_minlength(re, cc, startcode, options, recurses, countptr);
if (d < 0) return d;
branchlength += d;
do cc += GET(cc, 1); while (*cc == OP_ALT);
@@ -415,7 +419,8 @@
int dd;
this_recurse.prev = recurses;
this_recurse.group = cs;
- dd = find_minlength(re, cs, startcode, options, &this_recurse);
+ dd = find_minlength(re, cs, startcode, options, &this_recurse,
+ countptr);
if (dd < d) d = dd;
}
}
@@ -451,7 +456,8 @@
{
this_recurse.prev = recurses;
this_recurse.group = cs;
- d = find_minlength(re, cs, startcode, options, &this_recurse);
+ d = find_minlength(re, cs, startcode, options, &this_recurse,
+ countptr);
}
}
}
@@ -514,7 +520,7 @@
this_recurse.prev = recurses;
this_recurse.group = cs;
branchlength += find_minlength(re, cs, startcode, options,
- &this_recurse);
+ &this_recurse, countptr);
}
}
cc += 1 + LINK_SIZE;
@@ -1453,6 +1459,7 @@
#endif
{
int min;
+int count = 0;
BOOL bits_set = FALSE;
pcre_uint8 start_bits[32];
PUBL(extra) *extra = NULL;
@@ -1539,7 +1546,7 @@
/* Find the minimum length of subject string. */
-switch(min = find_minlength(re, code, code, re->options, NULL))
+switch(min = find_minlength(re, code, code, re->options, NULL, &count))
{
case -2: *errorptr = "internal error: missing capturing bracket"; return NULL;
case -3: *errorptr = "internal error: opcode not recognized"; return NULL;
Index: pcre_internal.h
===================================================================
--- pcre_internal.h (revision 1554)
+++ pcre_internal.h (working copy)
@@ -984,7 +984,7 @@
#ifndef EBCDIC
#define HSPACE_LIST \
- CHAR_HT, CHAR_SPACE, 0xa0, \
+ CHAR_HT, CHAR_SPACE, CHAR_NBSP, \
0x1680, 0x180e, 0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, \
0x2006, 0x2007, 0x2008, 0x2009, 0x200A, 0x202f, 0x205f, 0x3000, \
NOTACHAR
@@ -1010,7 +1010,7 @@
#define HSPACE_BYTE_CASES \
case CHAR_HT: \
case CHAR_SPACE: \
- case 0xa0 /* NBSP */
+ case CHAR_NBSP
#define HSPACE_CASES \
HSPACE_BYTE_CASES: \
@@ -1037,11 +1037,12 @@
/* ------ EBCDIC environments ------ */
#else
-#define HSPACE_LIST CHAR_HT, CHAR_SPACE
+#define HSPACE_LIST CHAR_HT, CHAR_SPACE, CHAR_NBSP, NOTACHAR
#define HSPACE_BYTE_CASES \
case CHAR_HT: \
- case CHAR_SPACE
+ case CHAR_SPACE: \
+ case CHAR_NBSP
#define HSPACE_CASES HSPACE_BYTE_CASES
@@ -1215,6 +1216,7 @@
#define CHAR_ESC '\047'
#define CHAR_DEL '\007'
+#define CHAR_NBSP '\x41'
#define STR_ESC "\047"
#define STR_DEL "\007"
@@ -1229,6 +1231,7 @@
#define CHAR_NEL ((unsigned char)'\x85')
#define CHAR_ESC '\033'
#define CHAR_DEL '\177'
+#define CHAR_NBSP ((unsigned char)'\xa0')
#define STR_LF "\n"
#define STR_NL STR_LF
@@ -1606,6 +1609,7 @@
#define CHAR_VERTICAL_LINE '\174'
#define CHAR_RIGHT_CURLY_BRACKET '\175'
#define CHAR_TILDE '\176'
+#define CHAR_NBSP ((unsigned char)'\xa0')
#define STR_HT "\011"
#define STR_VT "\013"
@@ -1762,6 +1766,10 @@
/* Escape items that are just an encoding of a particular data value. */
+#ifndef ESC_a
+#define ESC_a CHAR_BEL
+#endif
+
#ifndef ESC_e
#define ESC_e CHAR_ESC
#endif
@@ -2446,6 +2454,7 @@
BOOL had_pruneorskip; /* (*PRUNE) or (*SKIP) encountered */
BOOL check_lookbehind; /* Lookbehinds need later checking */
BOOL dupnames; /* Duplicate names exist */
+ BOOL dupgroups; /* Duplicate groups exist: (?| found */
BOOL iscondassert; /* Next assert is a condition */
int nltype; /* Newline type */
int nllen; /* Newline string length */
Index: pcre_exec.c
===================================================================
--- pcre_exec.c (revision 1554)
+++ pcre_exec.c (working copy)
@@ -6685,7 +6685,8 @@
register int *iend = iptr - re->top_bracket;
if (iend < md->offset_vector + 2) iend = md->offset_vector + 2;
while (--iptr >= iend) *iptr = -1;
- md->offset_vector[0] = md->offset_vector[1] = -1;
+ if (offsetcount > 0) md->offset_vector[0] = -1;
+ if (offsetcount > 1) md->offset_vector[1] = -1;
}
/* Set up the first character to match, if available. The first_char value is
Index: pcre_jit_test.c
===================================================================
--- pcre_jit_test.c (revision 1554)
+++ pcre_jit_test.c (working copy)
@@ -182,6 +182,7 @@
{ CMUAP, 0, "\xf0\x90\x90\x80{2}", "\xf0\x90\x90\x80#\xf0\x90\x90\xa8\xf0\x90\x90\x80" },
{ CMUAP, 0, "\xf0\x90\x90\xa8{2}", "\xf0\x90\x90\x80#\xf0\x90\x90\xa8\xf0\x90\x90\x80" },
{ CMUAP, 0, "\xe1\xbd\xb8\xe1\xbf\xb8", "\xe1\xbf\xb8\xe1\xbd\xb8" },
+ { MA, 0, "[3-57-9]", "5" },
/* Assertions. */
{ MUA, 0, "\\b[^A]", "A_B#" },
Index: configure.ac
===================================================================
--- configure.ac (revision 1554)
+++ configure.ac (working copy)
@@ -9,9 +9,9 @@
dnl be defined as -RC2, for example. For real releases, it should be empty.
m4_define(pcre_major, [8])
-m4_define(pcre_minor, [37])
-m4_define(pcre_prerelease, [])
-m4_define(pcre_date, [2015-04-28])
+m4_define(pcre_minor, [38])
+m4_define(pcre_prerelease, [-RC1])
+m4_define(pcre_date, [2015-05-03])
# NOTE: The CMakeLists.txt file searches for the above variables in the first
# 50 lines of this file. Please update that if the variables above are moved.
Index: doc/pcrepattern.3
===================================================================
--- doc/pcrepattern.3 (revision 1554)
+++ doc/pcrepattern.3 (working copy)
@@ -1,4 +1,4 @@
-.TH PCREPATTERN 3 "08 January 2014" "PCRE 8.35"
+.TH PCREPATTERN 3 "14 June 2015" "PCRE 8.38"
.SH NAME
PCRE - Perl-compatible regular expressions
.SH "PCRE REGULAR EXPRESSION DETAILS"
@@ -308,7 +308,8 @@
in patterns in a visible manner. There is no restriction on the appearance of
non-printing characters, apart from the binary zero that terminates a pattern,
but when a pattern is being prepared by text editing, it is often easier to use
-one of the following escape sequences than the binary character it represents:
+one of the following escape sequences than the binary character it represents.
+In an ASCII or Unicode environment, these escapes are as follows:
.sp
\ea alarm, that is, the BEL character (hex 07)
\ecx "control-x", where x is any ASCII character
@@ -330,19 +331,31 @@
but \ec{ becomes hex 3B ({ is 7B), and \ec; becomes hex 7B (; is 3B). If the
data item (byte or 16-bit value) following \ec has a value greater than 127, a
compile-time error occurs. This locks out non-ASCII characters in all modes.
+.P
+When PCRE is compiled in EBCDIC mode, \ea, \ee, \ef, \en, \er, and \et
+generate the appropriate EBCDIC code values. The \ec escape is processed
+as specified for Perl in the \fBperlebcdic\fP document. The only characters
+that are allowed after \ec are A-Z, a-z, or one of @, [, \e, ], ^, _, or ?. Any
+other character provokes a compile-time error. The sequence \e@ encodes
+character code 0; the letters (in either case) encode characters 1-26 (hex 01
+to hex 1A); [, \e, ], ^, and _ encode characters 27-31 (hex 1B to hex 1F), and
+\e? becomes either 255 (hex FF) or 95 (hex 5F).
.P
-The \ec facility was designed for use with ASCII characters, but with the
-extension to Unicode it is even less useful than it once was. It is, however,
-recognized when PCRE is compiled in EBCDIC mode, where data items are always
-bytes. In this mode, all values are valid after \ec. If the next character is a
-lower case letter, it is converted to upper case. Then the 0xc0 bits of the
-byte are inverted. Thus \ecA becomes hex 01, as in ASCII (A is C1), but because
-the EBCDIC letters are disjoint, \ecZ becomes hex 29 (Z is E9), and other
-characters also generate different values.
+Thus, apart from \e?, these escapes generate the same character code values as
+they do in an ASCII environment, though the meanings of the values mostly
+differ. For example, \eG always generates code value 7, which is BEL in ASCII
+but DEL in EBCDIC.
.P
+The sequence \e? generates DEL (127, hex 7F) in an ASCII environment, but
+because 127 is not a control character in EBCDIC, Perl makes it generate the
+APC character. Unfortunately, there are several variants of EBCDIC. In most of
+them the APC character has the value 255 (hex FF), but in the one Perl calls
+POSIX-BC its value is 95 (hex 5F). If certain other characters have POSIX-BC
+values, PCRE makes \e? generate 95; otherwise it generates 255.
+.P
After \e0 up to two further octal digits are read. If there are fewer than two
-digits, just those that are present are used. Thus the sequence \e0\ex\e07
-specifies two binary zeros followed by a BEL character (code value 7). Make
+digits, just those that are present are used. Thus the sequence \e0\ex\e015
+specifies two binary zeros followed by a CR character (code value 13). Make
sure you supply two digits after the initial zero if the pattern character that
follows is itself an octal digit.
.P
@@ -3283,6 +3296,6 @@
.rs
.sp
.nf
-Last updated: 08 January 2014
-Copyright (c) 1997-2014 University of Cambridge.
+Last updated: 14 June 2015
+Copyright (c) 1997-2015 University of Cambridge.
.fi
Index: testdata/testoutput11-32
===================================================================
--- testdata/testoutput11-32 (revision 1554)
+++ testdata/testoutput11-32 (working copy)
@@ -231,7 +231,7 @@
------------------------------------------------------------------
/(?P<a>a)...(?P=a)bbb(?P>a)d/BM
-Memory allocation (code space): 125
+Memory allocation (code space): 157
------------------------------------------------------------------
0 24 Bra
2 5 CBra 1
@@ -748,4 +748,21 @@
22 End
------------------------------------------------------------------
+/.((?2)(?R)\1)()/B
+------------------------------------------------------------------
+ 0 23 Bra
+ 2 Any
+ 3 13 Once
+ 5 9 CBra 1
+ 8 18 Recurse
+ 10 0 Recurse
+ 12 \1
+ 14 9 Ket
+ 16 13 Ket
+ 18 3 CBra 2
+ 21 3 Ket
+ 23 23 Ket
+ 25 End
+------------------------------------------------------------------
+
/-- End of testinput11 --/
Index: testdata/testinputEBC
===================================================================
--- testdata/testinputEBC (revision 1554)
+++ testdata/testinputEBC (working copy)
@@ -29,13 +29,16 @@
/^A\<5C>/
A B
+ A\x41B
/-- Test \H --/
/^A\<5C>/
AB
+ A\x42B
** Fail
A B
+ A\x41B
/-- Test \R --/
Index: testdata/testoutput1
===================================================================
--- testdata/testoutput1 (revision 1554)
+++ testdata/testoutput1 (working copy)
@@ -9429,4 +9429,9 @@
0: aaaaaaaaa
1: a
+"(?|(\k'Pm')|(?'Pm'))"
+ abcd
+ 0:
+ 1:
+
/-- End of testinput1 --/
Index: testdata/testoutput2
===================================================================
--- testdata/testoutput2 (revision 1554)
+++ testdata/testoutput2 (working copy)
@@ -5614,9 +5614,9 @@
123456\P
No match
-//KF>/dev/null
-Compiled pattern written to /dev/null
-Study data written to /dev/null
+//KF>testsavedregex
+Compiled pattern written to testsavedregex
+Study data written to testsavedregex
/abc/IS>testsavedregex
Capturing subpattern count = 0
@@ -9135,10 +9135,10 @@
Failed: subpattern name expected at offset 3
/\k/
-Failed: \k is not followed by a braced, angle-bracketed, or quoted name at offset 2
+Failed: \k is not followed by a braced, angle-bracketed, or quoted name at offset 1
/\kabc/
-Failed: \k is not followed by a braced, angle-bracketed, or quoted name at offset 5
+Failed: \k is not followed by a braced, angle-bracketed, or quoted name at offset 1
/(?P=)/
Failed: subpattern name expected at offset 4
@@ -9186,7 +9186,7 @@
Failed: unknown POSIX class name at offset 3
/(^(a|b\g<-1'c))/
-Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 15
+Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 8
/^(?+1)(?<a>x|y){0}z/
xzxx
@@ -14098,10 +14098,10 @@
Failed: group name must start with a non-digit at offset 4
/\g'3gh'/
-Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 7
+Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 2
/\g<5fg>/
-Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 7
+Failed: \g is not followed by a braced, angle-bracketed, or quoted name/number or by a plain number at offset 2
/(?(<4gh>)abc)/
Failed: group name must start with a non-digit at offset 4
@@ -14423,4 +14423,146 @@
/((?2){73}(?2))((?1))/
+/.((?2)(?R)\1)()/BZ
+------------------------------------------------------------------
+ Bra
+ Any
+ Once
+ CBra 1
+ Recurse
+ Recurse
+ \1
+ Ket
+ Ket
+ CBra 2
+ Ket
+ Ket
+ End
+------------------------------------------------------------------
+
+/(?1)()((((((\1++))\x85)+)|))/
+
+/(\9*+(?2);\3++()2|)++{/
+Failed: reference to non-existent subpattern at offset 22
+
+/\V\x85\9*+((?2)\3++()2)*:2/
+Failed: reference to non-existent subpattern at offset 26
+
+/(((?(R)){0,2}) (?''((?'R')((?'R')))))/J
+
+/(((?(X)){0,2}) (?''((?'X')((?'X')))))/J
+
+/(((?(R)){0,2}) (?''((?'X')((?'R')))))/
+
+"(?J)(?'d'(?'d'\g{d}))"
+
+".*?\h.+.\.+\R*?\xd(?i)(?=!(?=b`b`b`\`b\xa9b!)`\a`bbbbbbbbbbbbb`bbbbbbbbbbbb*R\x85bbbbbbb\C?{((?2)(?))((
+\H){8(?<=(?1){29}\xa8bbbb\x16\xd\xc6^($(?<! )(\xa9H4){4}h}1)B))\x15')"
+
+"(?J:(?|(?'R')(\k'R')|((?'R'))))"
+
+/(?<=|(\,\$(?73591620449005828816)\xa8.{7}){6}\x09)/
+Failed: number is too big at offset 32
+
+//
+\O1
+Matched, but too many substrings
+
+/^(?:(?(1)x|)+)+$()/BZ
+------------------------------------------------------------------
+ Bra
+ ^
+ SBra
+ SCond
+ 1 Cond ref
+ x
+ Alt
+ KetRmax
+ KetRmax
+ $
+ CBra 1
+ Ket
+ Ket
+ End
+------------------------------------------------------------------
+
+/(?=di(?<=(?1))|(?=(.))))/
+Failed: unmatched parentheses at offset 23
+
+/(?(R))*+/BZ
+------------------------------------------------------------------
+ Bra
+ Braposzero
+ SBraPos
+ SCond
+ Cond recurse any
+ Ket
+ KetRpos
+ Ket
+ End
+------------------------------------------------------------------
+
+/[[:\\](?'abc')[a:]/
+
+"[[[.\xe8Nq\xffq\xff\xe0\x2|||::Nq\xffq\xff\xe0\x6\x2|||::[[[:[::::::[[[[[::::::::[:[[[:[:::[[[[[[[[[[[[:::::::::::::::::[[.\xe8Nq\xffq\xff\xe0\x2|||::Nq\xffq\xff\xe0\x6\x2|||::[[[:[::::::[[[[[::::::::[:[[[:[:::[[[[[[[[[[[[[[:::E[[[:[:[[:[:::[[:::E[[[:[:[[:'[:::::E[[[:[::::::[[[:[[[[[[[::E[[[:[::::::[[[:[[[[[[[[:[[::[::::[[:::::::[[:[[[[[[[:[[::[:[[:[~"
+Failed: missing terminating ] for character class at offset 353
+
+/()(?(R)0)*+/BZ
+------------------------------------------------------------------
+ Bra
+ CBra 1
+ Ket
+ Braposzero
+ SBraPos
+ SCond
+ Cond recurse any
+ 0
+ Ket
+ KetRpos
+ Ket
+ End
+------------------------------------------------------------------
+
+/(?R-:(?</
+Failed: (?R or (?[+-]digits must be followed by ) at offset 3
+
+/(?1){3918}(((((0(\k'R'))))(?J)(?'R'(?'R'\3){99})))/I
+Capturing subpattern count = 8
+Max back reference = 8
+Named capturing subpatterns:
+ R 7
+ R 8
+No options
+Duplicate name status changes
+No first char
+Need char = '0'
+
+/(?J:(?|(:(?|(?'R')(\k'R')|((?'R')))H'Rk'Rf)|s(?'R')))/
+
+/0(?0)|(1)(*THEN)(*SKIP:0)(*FAIL)/
+ 01
+No match
+
+/((?(R8000000000)))/
+Failed: number is too big at offset 16
+
+/(?(8000000000/
+Failed: number is too big at offset 13
+
+/(?:ab)?(?:ab)(?:ab)/
+ abab
+ 0: abab
+ ababab
+ 0: ababab
+ aba
+No match
+
+/((*MARK:A))++a(*SKIP:B)b/
+ aacb
+No match
+
+/(?J:(?|(:(?|(?'R')(\z(?|(?'R')(\k'R')|((?'R')))k'R')|((?'R')))H'Ak'Rf)|s(?'R')))/
+
+/(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&
+
/-- End of testinput2 --/
Index: testdata/testoutput11-16
===================================================================
--- testdata/testoutput11-16 (revision 1554)
+++ testdata/testoutput11-16 (working copy)
@@ -231,7 +231,7 @@
------------------------------------------------------------------
/(?P<a>a)...(?P=a)bbb(?P>a)d/BM
-Memory allocation (code space): 61
+Memory allocation (code space): 77
------------------------------------------------------------------
0 24 Bra
2 5 CBra 1
@@ -748,4 +748,21 @@
22 End
------------------------------------------------------------------
+/.((?2)(?R)\1)()/B
+------------------------------------------------------------------
+ 0 23 Bra
+ 2 Any
+ 3 13 Once
+ 5 9 CBra 1
+ 8 18 Recurse
+ 10 0 Recurse
+ 12 \1
+ 14 9 Ket
+ 16 13 Ket
+ 18 3 CBra 2
+ 21 3 Ket
+ 23 23 Ket
+ 25 End
+------------------------------------------------------------------
+
/-- End of testinput11 --/
Index: testdata/testoutput6
===================================================================
--- testdata/testoutput6 (revision 1554)
+++ testdata/testoutput6 (working copy)
@@ -2469,4 +2469,8 @@
Ӆ\x0aT
No match
+/[\pS#moq]/
+ =
+ 0: =
+
/-- End of testinput6 --/
Index: testdata/testinput11
===================================================================
--- testdata/testinput11 (revision 1554)
+++ testdata/testinput11 (working copy)
@@ -136,4 +136,6 @@
/((?+1)(\1))/B
+/.((?2)(?R)\1)()/B
+
/-- End of testinput11 --/
Index: testdata/testinput12
===================================================================
--- testdata/testinput12 (revision 1554)
+++ testdata/testinput12 (working copy)
@@ -8,6 +8,8 @@
/(?(?C1)(?=a)a)/S!+I
+/b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b
+
/abc/S+I>testsavedregex
<testsavedregex
@@ -95,4 +97,11 @@
/(a(?:a|b|c|d|e)b){8,16}/S++
+/(?:|a|){100}x/S++
+
+/(x(?1)){4}/S++
+
+/(.|.)*?bx/
+ aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabax
+
/-- End of testinput12 --/
Index: testdata/testoutput12
===================================================================
--- testdata/testoutput12 (revision 1554)
+++ testdata/testoutput12 (working copy)
@@ -30,6 +30,15 @@
No starting char list
JIT study was not successful
+/b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b*b
+Capturing subpattern count = 0
+May match empty string
+No options
+No first char
+No need char
+Study returned NULL
+JIT study was not successful
+
/abc/S+I>testsavedregex
Capturing subpattern count = 0
No options
@@ -184,4 +193,12 @@
/(a(?:a|b|c|d|e)b){8,16}/S++
+/(?:|a|){100}x/S++
+
+/(x(?1)){4}/S++
+
+/(.|.)*?bx/
+ aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaabax
+Error -8 (match limit exceeded)
+
/-- End of testinput12 --/
Index: testdata/testoutput11-8
===================================================================
--- testdata/testoutput11-8 (revision 1554)
+++ testdata/testoutput11-8 (working copy)
@@ -231,7 +231,7 @@
------------------------------------------------------------------
/(?P<a>a)...(?P=a)bbb(?P>a)d/BM
-Memory allocation (code space): 38
+Memory allocation (code space): 50
------------------------------------------------------------------
0 30 Bra
3 7 CBra 1
@@ -748,4 +748,21 @@
34 End
------------------------------------------------------------------
+/.((?2)(?R)\1)()/B
+------------------------------------------------------------------
+ 0 35 Bra
+ 3 Any
+ 4 20 Once
+ 7 14 CBra 1
+ 12 27 Recurse
+ 15 0 Recurse
+ 18 \1
+ 21 14 Ket
+ 24 20 Ket
+ 27 5 CBra 2
+ 32 5 Ket
+ 35 35 Ket
+ 38 End
+------------------------------------------------------------------
+
/-- End of testinput11 --/
Index: testdata/testoutputEBC
===================================================================
--- testdata/testoutputEBC (revision 1554)
+++ testdata/testoutputEBC (working copy)
@@ -41,6 +41,8 @@
/^A\<5C>/
A B
0: A\x20
+ A\x41B
+ 0: AA
/-- Test \H --/
@@ -47,10 +49,14 @@
/^A\<5C>/
AB
0: AB
+ A\x42B
+ 0: AB
** Fail
No match
A B
No match
+ A\x41B
+No match
/-- Test \R --/
Index: testdata/grepoutput
===================================================================
--- testdata/grepoutput (revision 1554)
+++ testdata/grepoutput (working copy)
@@ -751,3 +751,7 @@
2:3,1
2:4,1
RC=0
+---------------------------- Test 108 ------------------------------
+RC=0
+---------------------------- Test 109 -----------------------------
+RC=0
Index: testdata/testinput1
===================================================================
--- testdata/testinput1 (revision 1554)
+++ testdata/testinput1 (working copy)
@@ -5730,4 +5730,7 @@
"(?1)(?#?'){8}(a)"
baaaaaaaaac
+"(?|(\k'Pm')|(?'Pm'))"
+ abcd
+
/-- End of testinput1 --/
Index: testdata/testinput2
===================================================================
--- testdata/testinput2 (revision 1554)
+++ testdata/testinput2 (working copy)
@@ -1380,7 +1380,7 @@
1X
123456\P
-//KF>/dev/null
+//KF>testsavedregex
/abc/IS>testsavedregex
<testsavedregex
@@ -4152,4 +4152,67 @@
/((?2){73}(?2))((?1))/
+/.((?2)(?R)\1)()/BZ
+
+/(?1)()((((((\1++))\x85)+)|))/
+
+/(\9*+(?2);\3++()2|)++{/
+
+/\V\x85\9*+((?2)\3++()2)*:2/
+
+/(((?(R)){0,2}) (?''((?'R')((?'R')))))/J
+
+/(((?(X)){0,2}) (?''((?'X')((?'X')))))/J
+
+/(((?(R)){0,2}) (?''((?'X')((?'R')))))/
+
+"(?J)(?'d'(?'d'\g{d}))"
+
+".*?\h.+.\.+\R*?\xd(?i)(?=!(?=b`b`b`\`b\xa9b!)`\a`bbbbbbbbbbbbb`bbbbbbbbbbbb*R\x85bbbbbbb\C?{((?2)(?))((
+\H){8(?<=(?1){29}\xa8bbbb\x16\xd\xc6^($(?<! )(\xa9H4){4}h}1)B))\x15')"
+
+"(?J:(?|(?'R')(\k'R')|((?'R'))))"
+
+/(?<=|(\,\$(?73591620449005828816)\xa8.{7}){6}\x09)/
+
+//
+\O1
+
+/^(?:(?(1)x|)+)+$()/BZ
+
+/(?=di(?<=(?1))|(?=(.))))/
+
+/(?(R))*+/BZ
+
+/[[:\\](?'abc')[a:]/
+
+"[[[.\xe8Nq\xffq\xff\xe0\x2|||::Nq\xffq\xff\xe0\x6\x2|||::[[[:[::::::[[[[[::::::::[:[[[:[:::[[[[[[[[[[[[:::::::::::::::::[[.\xe8Nq\xffq\xff\xe0\x2|||::Nq\xffq\xff\xe0\x6\x2|||::[[[:[::::::[[[[[::::::::[:[[[:[:::[[[[[[[[[[[[[[:::E[[[:[:[[:[:::[[:::E[[[:[:[[:'[:::::E[[[:[::::::[[[:[[[[[[[::E[[[:[::::::[[[:[[[[[[[[:[[::[::::[[:::::::[[:[[[[[[[:[[::[:[[:[~"
+
+/()(?(R)0)*+/BZ
+
+/(?R-:(?</
+
+/(?1){3918}(((((0(\k'R'))))(?J)(?'R'(?'R'\3){99})))/I
+
+/(?J:(?|(:(?|(?'R')(\k'R')|((?'R')))H'Rk'Rf)|s(?'R')))/
+
+/0(?0)|(1)(*THEN)(*SKIP:0)(*FAIL)/
+ 01
+
+/((?(R8000000000)))/
+
+/(?(8000000000/
+
+/(?:ab)?(?:ab)(?:ab)/
+ abab
+ ababab
+ aba
+
+/((*MARK:A))++a(*SKIP:B)b/
+ aacb
+
+/(?J:(?|(:(?|(?'R')(\z(?|(?'R')(\k'R')|((?'R')))k'R')|((?'R')))H'Ak'Rf)|s(?'R')))/
+
+/(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&a)(?&
+
/-- End of testinput2 --/
Index: testdata/testinput6
===================================================================
--- testdata/testinput6 (revision 1554)
+++ testdata/testinput6 (working copy)
@@ -1502,4 +1502,7 @@
/\C\X*QT/8
Ӆ\x0aT
+/[\pS#moq]/
+ =
+
/-- End of testinput6 --/
Index: NON-AUTOTOOLS-BUILD
===================================================================
--- NON-AUTOTOOLS-BUILD (revision 1554)
+++ NON-AUTOTOOLS-BUILD (working copy)
@@ -764,9 +764,9 @@
http://www.zaconsultants.net
-There is also a mirror here:
+You may download PCRE from WWW.CBTTAPE.ORG, file 882.  Everything, source and
+executable, is in EBCDIC and native z/OS file formats and this is the
+recommended download site.
- http://www.vsoft-software.com/downloads.html
-
==========================
-Last Updated: 10 February 2015
+Last Updated: 25 June 2015