regex: don't assume uint64_t or uint32_t

This avoids -Werror=overflow errors for 32-bit systems in
the 64-bit case.  Problem reported by Joseph Myers in:
https://sourceware.org/ml/libc-alpha/2017-11/msg00694.html
Also, when this code is used in Gnulib it ports to platforms
that lack uint64_t and uint32_t.  The C standard doesn't guarantee
them, and on some 32-bit compilers there is no uint64_t.
Problem reported by Gianluigi Tiesi in:
http://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00154.html
* posix/regcomp.c (init_word_char): Don't assume that the types
uint64_t and uint32_t exist.  Adapted from Gnulib patch
2012-05-27T06:40:00!eggert@cs.ucla.edu.  See:
https://git.savannah.gnu.org/cgit/gnulib.git/commit/?id=252b52457da7887667c036d18cc5169777615bb0

diff --git a/ChangeLog b/ChangeLog
index 3688c7f62470c33590c2b015a637d1b42a74da16..50da3df0b98cf783b0baba8ea4d8706bdf093db0 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,19 @@
+2017-11-20  Paul Eggert  <eggert@cs.ucla.edu>
+
+       regex: don't assume uint64_t or uint32_t
+       This avoids -Werror=overflow errors for 32-bit systems in
+       the 64-bit case.  Problem reported by Joseph Myers in:
+       https://sourceware.org/ml/libc-alpha/2017-11/msg00694.html
+       Also, when this code is used in Gnulib it ports to platforms
+       that lack uint64_t and uint32_t.  The C standard doesn't guarantee
+       them, and on some 32-bit compilers there is no uint64_t.
+       Problem reported by Gianluigi Tiesi in:
+       http://lists.gnu.org/archive/html/bug-gnulib/2012-03/msg00154.html
+       * posix/regcomp.c (init_word_char): Don't assume that the types
+       uint64_t and uint32_t exist.  Adapted from Gnulib patch
+       2012-05-27T06:40:00!eggert@cs.ucla.edu.  See:
+       https://git.savannah.gnu.org/cgit/gnulib.git/commit/?id=252b52457da7887667c036d18cc5169777615bb0
+
 2017-11-20  Siddhesh Poyarekar  <siddhesh@sourceware.org>
 
        * sysdeps/aarch64/memset-reg.h: New file.

diff --git a/posix/regcomp.c b/posix/regcomp.c
index 871ae2ffab5ffc57a9d3b594c42feec5cdd5a4f6..520596b51b3e3d203a8c5fbd1e5cb235d9493944 100644 (file)
--- a/posix/regcomp.c
+++ b/posix/regcomp.c
@@ -925,26 +925,26 @@ init_word_char (re_dfa_t *dfa)
   int ch = 0;
   if (BE (dfa->map_notascii == 0, 1))
     {
-      if (sizeof (dfa->word_char[0]) == 8)
-       {
-          /* The extra temporaries here avoid "implicitly truncated"
-             warnings in the case when this is dead code, i.e. 32-bit.  */
-          const uint64_t wc0 = UINT64_C (0x03ff000000000000);
-          const uint64_t wc1 = UINT64_C (0x07fffffe87fffffe);
-         dfa->word_char[0] = wc0;
-         dfa->word_char[1] = wc1;
+      bitset_word_t bits0 = 0x00000000;
+      bitset_word_t bits1 = 0x03ff0000;
+      bitset_word_t bits2 = 0x87fffffe;
+      bitset_word_t bits3 = 0x07fffffe;
+      if (BITSET_WORD_BITS == 64)
+       {
+         dfa->word_char[0] = bits1 << 31 << 1 | bits0;
+         dfa->word_char[1] = bits3 << 31 << 1 | bits2;
          i = 2;
        }
-      else if (sizeof (dfa->word_char[0]) == 4)
+      else if (BITSET_WORD_BITS == 32)
        {
-         dfa->word_char[0] = UINT32_C (0x00000000);
-         dfa->word_char[1] = UINT32_C (0x03ff0000);
-         dfa->word_char[2] = UINT32_C (0x87fffffe);
-         dfa->word_char[3] = UINT32_C (0x07fffffe);
+         dfa->word_char[0] = bits0;
+         dfa->word_char[1] = bits1;
+         dfa->word_char[2] = bits2;
+         dfa->word_char[3] = bits3;
          i = 4;
        }
       else
-       abort ();
+        goto general_case;
       ch = 128;
 
       if (BE (dfa->is_utf8, 1))
@@ -954,6 +954,7 @@ init_word_char (re_dfa_t *dfa)
        }
     }
 
+ general_case:
   for (; i < BITSET_WORDS; ++i)
     for (int j = 0; j < BITSET_WORD_BITS; ++j, ++ch)
       if (isalnum (ch) || ch == '_')