Merge branch 'Assert string ptrs in-bounds' into blead

There are a bunch of functions that are passed pointers to two positions
in a string. These commits allow you to declare these in embed.fnc, and
have an assert like (s < e) generated that is automatically placed in
the PERL_ARGS_ASSERT for the function.  Many entries in embed.fnc are
changed to use this new capability.

This is extended for functions where it should instead be assert(s <= e)
and assert(s <= cur)

Author: khwilliamson

autodoc.pl

@@ -620,7 +620,7 @@ sub check_and_add_proto_defn {
     $flags .= "n" if $flags =~ /#/;    # No threads, arguments for #ifdef
 
     my @munged_args= $args_ref->@*;
-    s/\b(?:NN|NULLOK)\b\s+//g for @munged_args;
+    s/\b(?:NN|NULLOK|[SM]PTR|EPTRQ?)\b\s+//g for @munged_args;
 
     my $flags_sans_d = $flags;
     my $docs_expected = $flags_sans_d =~ s/d//g;

embed.fnc

@@ -2963,7 +2963,6 @@ Perl_utf8_hop_back_overshoot(const U8 *s, SSize_t off,
                              const U8 * const start, SSize_t *remaining)
 {
     PERL_ARGS_ASSERT_UTF8_HOP_BACK_OVERSHOOT;
-    assert(start <= s);
     assert(off <= 0);
 
     /* Note: if we know that the input is well-formed, we can do per-word
@@ -3031,8 +3030,6 @@ Perl_utf8_hop_overshoot(const U8 *s, SSize_t off,
 {
     PERL_ARGS_ASSERT_UTF8_HOP_OVERSHOOT;
 
-    assert(start <= s && s <= end);
-
     if (off >= 0) {
         return utf8_hop_forward_overshoot(s, off, end, remaining);
     }

inline.h

@@ -627,7 +627,6 @@ STATIC U8 *
 S_find_span_end(U8 * s, const U8 * send, const U8 span_byte)
 {
     PERL_ARGS_ASSERT_FIND_SPAN_END;
-    assert(send >= s);
 
     /* Returns the position of the first byte in the sequence between 's' and
      * 'send-1' inclusive that isn't 'span_byte'; returns 'send' if none found.
@@ -700,7 +699,6 @@ STATIC U8 *
 S_find_next_masked(U8 * s, const U8 * send, const U8 byte, const U8 mask)
 {
     PERL_ARGS_ASSERT_FIND_NEXT_MASKED;
-    assert(send >= s);
     assert((byte & mask) == byte);
 
     /* Returns the position of the first byte in the sequence between 's'
@@ -777,7 +775,6 @@ STATIC U8 *
 S_find_span_end_mask(U8 * s, const U8 * send, const U8 span_byte, const U8 mask)
 {
     PERL_ARGS_ASSERT_FIND_SPAN_END_MASK;
-    assert(send >= s);
     assert((span_byte & mask) == span_byte);
 
     /* Returns the position of the first byte in the sequence between 's' and

proto.h

@@ -1887,7 +1887,6 @@ STATIC void
 S_incline(pTHX_ const char *s, const char *end)
 {
     PERL_ARGS_ASSERT_INCLINE;
-    assert(end >= s);
 
     const char *t;
     const char *n;
@@ -2855,7 +2854,6 @@ Perl_get_and_check_backslash_N_name(pTHX_ const char* s,
                                           const char ** error_msg)
 {
     PERL_ARGS_ASSERT_GET_AND_CHECK_BACKSLASH_N_NAME;
-    assert(e >= s);
     assert(s > (char *) 3);
 
     /* <s> points to first character of interior of \N{}, <e> to one beyond the

regen/embed.pl

@@ -725,7 +725,6 @@ STRLEN
 Perl_is_utf8_char_helper_(const U8 * const s, const U8 * e, const U32 flags)
 {
     PERL_ARGS_ASSERT_IS_UTF8_CHAR_HELPER_;
-    assert(e > s);
     assert(0 == (flags & ~UTF8_DISALLOW_ILLEGAL_INTERCHANGE));
 
     SSize_t len, full_len;
@@ -755,6 +754,7 @@ Perl_is_utf8_char_helper_(const U8 * const s, const U8 * e, const U32 flags)
      *  determined with just the first one or two bytes.
      *
      */
+
     full_len = UTF8SKIP(s);
 
     len = e - s;
@@ -840,7 +840,6 @@ Perl_is_utf8_FF_helper_(const U8 * const s0, const U8 * const e,
                         const bool require_partial)
 {
     PERL_ARGS_ASSERT_IS_UTF8_FF_HELPER_;
-    assert(s0 < e);
     assert(*s0 == I8_TO_NATIVE_UTF8(0xFF));
 
     /* This is called to determine if the UTF-8 sequence starting at s0 and
@@ -4245,7 +4244,6 @@ S_turkic_fc(pTHX_ const U8 * const p, const U8 * const e,
                         U8 * ustrp, STRLEN *lenp)
 {
     PERL_ARGS_ASSERT_TURKIC_FC;
-    assert(e > p);
 
     /* Returns 0 if the foldcase of the input UTF-8 encoded sequence from
      * p0..e-1 according to Turkic rules is the same as for non-Turkic.
@@ -4280,7 +4278,6 @@ S_turkic_lc(pTHX_ const U8 * const p0, const U8 * const e,
                         U8 * ustrp, STRLEN *lenp)
 {
     PERL_ARGS_ASSERT_TURKIC_LC;
-    assert(e > p0);
 
     /* Returns 0 if the lowercase of the input UTF-8 encoded sequence from
      * p0..e-1 according to Turkic rules is the same as for non-Turkic.
@@ -4326,7 +4323,6 @@ S_turkic_uc(pTHX_ const U8 * const p, const U8 * const e,
                         U8 * ustrp, STRLEN *lenp)
 {
     PERL_ARGS_ASSERT_TURKIC_UC;
-    assert(e > p);
 
     /* Returns 0 if the upper or title-case of the input UTF-8 encoded sequence
      * from p0..e-1 according to Turkic rules is the same as for non-Turkic.

regexec.c

@@ -619,9 +619,6 @@ Perl_delimcpy_no_escape(char *to, const char *to_end,
     ptrdiff_t to_len = to_end - to;
     SSize_t copy_len;
 
-    assert(from_len >= 0);
-    assert(to_len >= 0);
-
     /* Look for the first delimiter in the source */
     delim_pos = (const char *) memchr(from, delim, from_len);
 
@@ -718,8 +715,6 @@ Perl_delimcpy(char *to, const char *to_end,
               const int delim, I32 *retlen)
 {
     PERL_ARGS_ASSERT_DELIMCPY;
-    assert(from_end >= from);
-    assert(to_end >= to);
 
     const char * const orig_to = to;
     ptrdiff_t copy_len = 0;
@@ -1100,7 +1095,6 @@ char *
 Perl_fbm_instr(pTHX_ unsigned char *big, unsigned char *bigend, SV *littlestr, U32 flags)
 {
     PERL_ARGS_ASSERT_FBM_INSTR;
-    assert(bigend >= big);
 
     unsigned char *s;
     STRLEN l;