From 3773798aa6fa752e4210506ebbfba6857863ca75 Mon Sep 17 00:00:00 2001
From: Tom Callaway <spot@fedoraproject.org>
Date: Sat, 12 Jul 2014 11:34:09 -0400
Subject: [PATCH 1/8] fix license handling

---
 grep.spec | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/grep.spec b/grep.spec
index 5f38f17..c1de47c 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.20
-Release: 2%{?dist}
+Release: 3%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -76,7 +76,9 @@ fi
 
 %files -f %{name}.lang
 %defattr(-,root,root,-)
-%doc ABOUT-NLS AUTHORS THANKS TODO NEWS README ChangeLog COPYING
+%doc ABOUT-NLS AUTHORS THANKS TODO NEWS README ChangeLog
+%{!?_licensedir:%global license %%doc}
+%license COPYING
 
 %{_bindir}/*
 %config(noreplace) %{_sysconfdir}/profile.d/colorgrep.*sh
@@ -85,6 +87,9 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Sat Jul 12 2014 Tom Callaway <spot@fedoraproject.org> - 2.20-3
+- fix license handling .
+
 * Sat Jun 07 2014 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 2.20-2
 - Rebuilt for https://fedoraproject.org/wiki/Fedora_21_Mass_Rebuild
 

From 3740ef625b6fb2e39c351b6037ca9ad0aefdfe4d Mon Sep 17 00:00:00 2001
From: Peter Robinson <pbrobinson@fedoraproject.org>
Date: Sat, 16 Aug 2014 18:59:10 +0000
Subject: [PATCH 2/8] - Rebuilt for
 https://fedoraproject.org/wiki/Fedora_21_22_Mass_Rebuild

---
 grep.spec | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/grep.spec b/grep.spec
index c1de47c..a4798b0 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.20
-Release: 3%{?dist}
+Release: 4%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -87,6 +87,9 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Sat Aug 16 2014 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 2.20-4
+- Rebuilt for https://fedoraproject.org/wiki/Fedora_21_22_Mass_Rebuild
+
 * Sat Jul 12 2014 Tom Callaway <spot@fedoraproject.org> - 2.20-3
 - fix license handling .
 

From 9469ab4bec98694b06fcb3b0d4fcaa6235f64a7c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
Date: Tue, 11 Nov 2014 16:41:57 +0100
Subject: [PATCH 3/8] Fixed invalid UTF-8 byte sequence error in PCRE mode

  (by pcre-invalid-utf8-fix patch)
  Resolves: rhbz#1161832
---
 grep-2.20-pcre-invalid-utf8-fix.patch | 136 ++++++++++++++++++++++++++
 grep.spec                             |  10 +-
 2 files changed, 145 insertions(+), 1 deletion(-)
 create mode 100644 grep-2.20-pcre-invalid-utf8-fix.patch

diff --git a/grep-2.20-pcre-invalid-utf8-fix.patch b/grep-2.20-pcre-invalid-utf8-fix.patch
new file mode 100644
index 0000000..5f7530f
--- /dev/null
+++ b/grep-2.20-pcre-invalid-utf8-fix.patch
@@ -0,0 +1,136 @@
+diff --git a/src/pcresearch.c b/src/pcresearch.c
+index 820dd00..11df488 100644
+--- a/src/pcresearch.c
++++ b/src/pcresearch.c
+@@ -136,34 +136,42 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+ #else
+   /* This array must have at least two elements; everything after that
+      is just for performance improvement in pcre_exec.  */
+-  int sub[300];
++  enum { nsub = 300 };
++  int sub[nsub];
+ 
+-  const char *line_buf, *line_end, *line_next;
++  char const *p = start_ptr ? start_ptr : buf;
++  int options = p == buf || p[-1] == eolbyte ? 0 : PCRE_NOTBOL;
++  char const *line_start = buf;
+   int e = PCRE_ERROR_NOMATCH;
+-  ptrdiff_t start_ofs = start_ptr ? start_ptr - buf : 0;
++  char const *line_end;
+ 
+   /* PCRE can't limit the matching to single lines, therefore we have to
+      match each line in the buffer separately.  */
+-  for (line_next = buf;
+-       e == PCRE_ERROR_NOMATCH && line_next < buf + size;
+-       start_ofs -= line_next - line_buf)
++  for (; p < buf + size; p = line_start = line_end + 1)
+     {
+-      line_buf = line_next;
+-      line_end = memchr (line_buf, eolbyte, (buf + size) - line_buf);
+-      if (line_end == NULL)
+-        line_next = line_end = buf + size;
+-      else
+-        line_next = line_end + 1;
+-
+-      if (start_ptr && start_ptr >= line_end)
+-        continue;
++      line_end = memchr (p, eolbyte, buf + size - p);
+ 
+-      if (INT_MAX < line_end - line_buf)
++      if (INT_MAX < line_end - p)
+         error (EXIT_TROUBLE, 0, _("exceeded PCRE's line length limit"));
+ 
+-      e = pcre_exec (cre, extra, line_buf, line_end - line_buf,
+-                     start_ofs < 0 ? 0 : start_ofs, 0,
+-                     sub, sizeof sub / sizeof *sub);
++      /* Treat encoding-error bytes as data that cannot match.  */
++      for (;;)
++        {
++          e = pcre_exec (cre, extra, p, line_end - p, 0, options, sub, nsub);
++          if (e != PCRE_ERROR_BADUTF8)
++            break;
++          e = pcre_exec (cre, extra, p, sub[0], 0,
++                         options | PCRE_NO_UTF8_CHECK | PCRE_NOTEOL,
++                         sub, nsub);
++          if (e != PCRE_ERROR_NOMATCH)
++            break;
++          p += sub[0] + 1;
++          options = PCRE_NOTBOL;
++        }
++
++      if (e != PCRE_ERROR_NOMATCH)
++        break;
++      options = 0;
+     }
+ 
+   if (e <= 0)
+@@ -180,10 +188,6 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+           error (EXIT_TROUBLE, 0,
+                  _("exceeded PCRE's backtracking limit"));
+ 
+-        case PCRE_ERROR_BADUTF8:
+-          error (EXIT_TROUBLE, 0,
+-                 _("invalid UTF-8 byte sequence in input"));
+-
+         default:
+           /* For now, we lump all remaining PCRE failures into this basket.
+              If anyone cares to provide sample grep usage that can trigger
+@@ -197,25 +201,8 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+     }
+   else
+     {
+-      /* Narrow down to the line we've found.  */
+-      char const *beg = line_buf + sub[0];
+-      char const *end = line_buf + sub[1];
+-      char const *buflim = buf + size;
+-      char eol = eolbyte;
+-      if (!start_ptr)
+-        {
+-          /* FIXME: The case when '\n' is not found indicates a bug:
+-             Since grep is line oriented, the match should never contain
+-             a newline, so there _must_ be a newline following.
+-           */
+-          if (!(end = memchr (end, eol, buflim - end)))
+-            end = buflim;
+-          else
+-            end++;
+-          while (buf < beg && beg[-1] != eol)
+-            --beg;
+-        }
+-
++      char const *beg = start_ptr ? p + sub[0] : line_start;
++      char const *end = start_ptr ? p + sub[1] : line_end + 1;
+       *match_size = end - beg;
+       return beg - buf;
+     }
+diff --git a/tests/pcre-infloop b/tests/pcre-infloop
+index 1b33e72..b92f8e1 100755
+--- a/tests/pcre-infloop
++++ b/tests/pcre-infloop
+@@ -28,6 +28,6 @@ printf 'a\201b\r' > in || framework_failure_
+ fail=0
+ 
+ LC_ALL=en_US.UTF-8 timeout 3 grep -P 'a.?..b' in
+-test $? = 2 || fail_ "libpcre's match function appears to infloop"
++test $? = 1 || fail_ "libpcre's match function appears to infloop"
+ 
+ Exit $fail
+diff --git a/tests/pcre-invalid-utf8-input b/tests/pcre-invalid-utf8-input
+index 913e8ee..9da4b18 100755
+--- a/tests/pcre-invalid-utf8-input
++++ b/tests/pcre-invalid-utf8-input
+@@ -13,9 +13,12 @@ require_en_utf8_locale_
+ 
+ fail=0
+ 
+-printf 'j\202\nj\n' > in || framework_failure_
++printf 'j\202j\nj\nk\202\n' > in || framework_failure_
+ 
+ LC_ALL=en_US.UTF-8 grep -P j in
+-test $? -eq 2 || fail=1
++test $? -eq 0 || fail=1
++
++LC_ALL=en_US.UTF-8 grep -P 'k$' in
++test $? -eq 1 || fail=1
+ 
+ Exit $fail
diff --git a/grep.spec b/grep.spec
index a4798b0..6870319 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.20
-Release: 4%{?dist}
+Release: 5%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -14,6 +14,8 @@ Source3: GREP_COLORS
 Patch0: grep-2.20-man-fix-gs.patch
 # upstream ticket 39445
 Patch1: grep-2.20-help-align.patch
+# backported from upstream
+Patch2: grep-2.20-pcre-invalid-utf8-fix.patch
 URL: http://www.gnu.org/software/grep/
 Requires(post): /sbin/install-info
 Requires(preun): /sbin/install-info
@@ -34,6 +36,7 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
 %setup -q
 %patch0 -p1 -b .man-fix-gs
 %patch1 -p1 -b .help-align
+%patch2 -p1 -b .pcre-invalid-utf8-fix
 
 %build
 %global BUILD_FLAGS $RPM_OPT_FLAGS
@@ -87,6 +90,11 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Tue Nov 11 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.20-5
+- Fixed invalid UTF-8 byte sequence error in PCRE mode
+  (by pcre-invalid-utf8-fix patch)
+  Resolves: rhbz#1161832
+
 * Sat Aug 16 2014 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 2.20-4
 - Rebuilt for https://fedoraproject.org/wiki/Fedora_21_22_Mass_Rebuild
 

From 57348fea032fc5950a11835346b6d81072462d63 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
Date: Fri, 14 Nov 2014 17:35:02 +0100
Subject: [PATCH 4/8] Backported more PCRE fixes (by pcre-backported-fixes
 patch)

- Dropped pcre-invalid-utf8-fix patch, handled by pcre-backported-fixes patch
---
 grep-2.20-pcre-backported-fixes.patch | 389 ++++++++++++++++++++++++++
 grep-2.20-pcre-invalid-utf8-fix.patch | 136 ---------
 grep.spec                             |  10 +-
 3 files changed, 396 insertions(+), 139 deletions(-)
 create mode 100644 grep-2.20-pcre-backported-fixes.patch
 delete mode 100644 grep-2.20-pcre-invalid-utf8-fix.patch

diff --git a/grep-2.20-pcre-backported-fixes.patch b/grep-2.20-pcre-backported-fixes.patch
new file mode 100644
index 0000000..4a9dbcd
--- /dev/null
+++ b/grep-2.20-pcre-backported-fixes.patch
@@ -0,0 +1,389 @@
+diff --git a/src/grep.h b/src/grep.h
+index 4935872..729c906 100644
+--- a/src/grep.h
++++ b/src/grep.h
+@@ -27,4 +27,19 @@ extern int match_words;		/* -w */
+ extern int match_lines;		/* -x */
+ extern unsigned char eolbyte;	/* -z */
+ 
++/* An enum textbin describes the file's type, inferred from data read
++   before the first line is selected for output.  */
++enum textbin
++  {
++    /* Binary, as it contains null bytes and the -z option is not in effect,
++       or it contains encoding errors.  */
++    TEXTBIN_BINARY = -1,
++
++    /* Not known yet.  Only text has been seen so far.  */
++    TEXTBIN_UNKNOWN = 0,
++
++    /* Text.  */
++    TEXTBIN_TEXT = 1
++  };
++
+ #endif
+diff --git a/src/pcresearch.c b/src/pcresearch.c
+index 820dd00..9938ffc 100644
+--- a/src/pcresearch.c
++++ b/src/pcresearch.c
+@@ -33,13 +33,19 @@ static pcre *cre;
+ /* Additional information about the pattern.  */
+ static pcre_extra *extra;
+ 
+-# ifdef PCRE_STUDY_JIT_COMPILE
+-static pcre_jit_stack *jit_stack;
+-# else
++# ifndef PCRE_STUDY_JIT_COMPILE
+ #  define PCRE_STUDY_JIT_COMPILE 0
+ # endif
+ #endif
+ 
++/* Table, indexed by ! (flag & PCRE_NOTBOL), of whether the empty
++   string matches when that flag is used.  */
++static int empty_match[2];
++
++/* This must be at least 2; everything after that is for performance
++   in pcre_exec.  */
++enum { NSUB = 300 };
++
+ void
+ Pcompile (char const *pattern, size_t size)
+ {
+@@ -52,13 +58,17 @@ Pcompile (char const *pattern, size_t size)
+   char const *ep;
+   char *re = xnmalloc (4, size + 7);
+   int flags = (PCRE_MULTILINE
+-               | (match_icase ? PCRE_CASELESS : 0)
+-               | (using_utf8 () ? PCRE_UTF8 : 0));
++               | (match_icase ? PCRE_CASELESS : 0));
+   char const *patlim = pattern + size;
+   char *n = re;
+   char const *p;
+   char const *pnul;
+ 
++  if (using_utf8 ())
++    flags |= PCRE_UTF8;
++  else if (MB_CUR_MAX != 1)
++    error (EXIT_TROUBLE, 0, _("-P supports only unibyte and UTF-8 locales"));
++
+   /* FIXME: Remove these restrictions.  */
+   if (memchr (pattern, '\n', size))
+     error (EXIT_TROUBLE, 0, _("the -P option only supports a single pattern"));
+@@ -114,14 +124,20 @@ Pcompile (char const *pattern, size_t size)
+       /* A 32K stack is allocated for the machine code by default, which
+          can grow to 512K if necessary. Since JIT uses far less memory
+          than the interpreter, this should be enough in practice.  */
+-      jit_stack = pcre_jit_stack_alloc (32 * 1024, 512 * 1024);
++      pcre_jit_stack *jit_stack = pcre_jit_stack_alloc (32 * 1024, 512 * 1024);
+       if (!jit_stack)
+         error (EXIT_TROUBLE, 0,
+                _("failed to allocate memory for the PCRE JIT stack"));
+       pcre_assign_jit_stack (extra, NULL, jit_stack);
+     }
++
+ # endif
+   free (re);
++
++  int sub[NSUB];
++  empty_match[false] = pcre_exec (cre, extra, "", 0, 0,
++                                  PCRE_NOTBOL, sub, NSUB);
++  empty_match[true] = pcre_exec (cre, extra, "", 0, 0, 0, sub, NSUB);
+ #endif /* HAVE_LIBPCRE */
+ }
+ 
+@@ -134,36 +150,110 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+   error (EXIT_TROUBLE, 0, _("internal error"));
+   return -1;
+ #else
+-  /* This array must have at least two elements; everything after that
+-     is just for performance improvement in pcre_exec.  */
+-  int sub[300];
+-
+-  const char *line_buf, *line_end, *line_next;
++  int sub[NSUB];
++  char const *p = start_ptr ? start_ptr : buf;
++  bool bol = p[-1] == eolbyte;
++  char const *line_start = buf;
+   int e = PCRE_ERROR_NOMATCH;
+-  ptrdiff_t start_ofs = start_ptr ? start_ptr - buf : 0;
++  char const *line_end;
+ 
+-  /* PCRE can't limit the matching to single lines, therefore we have to
+-     match each line in the buffer separately.  */
+-  for (line_next = buf;
+-       e == PCRE_ERROR_NOMATCH && line_next < buf + size;
+-       start_ofs -= line_next - line_buf)
++  /* If the input type is unknown, the caller is still testing the
++     input, which means the current buffer cannot contain encoding
++     errors and a multiline search is typically more efficient.
++     Otherwise, a single-line search is typically faster, so that
++     pcre_exec doesn't waste time validating the entire input
++     buffer.  */
++  bool multiline = TEXTBIN_UNKNOWN;
++
++  for (; p < buf + size; p = line_start = line_end + 1)
+     {
+-      line_buf = line_next;
+-      line_end = memchr (line_buf, eolbyte, (buf + size) - line_buf);
+-      if (line_end == NULL)
+-        line_next = line_end = buf + size;
+-      else
+-        line_next = line_end + 1;
++      bool too_big;
+ 
+-      if (start_ptr && start_ptr >= line_end)
+-        continue;
++      if (multiline)
++        {
++          size_t pcre_size_max = MIN (INT_MAX, SIZE_MAX - 1);
++          size_t scan_size = MIN (pcre_size_max + 1, buf + size - p);
++          line_end = memrchr (p, eolbyte, scan_size);
++          too_big = ! line_end;
++        }
++      else
++        {
++          line_end = memchr (p, eolbyte, buf + size - p);
++          too_big = INT_MAX < line_end - p;
++        }
+ 
+-      if (INT_MAX < line_end - line_buf)
++      if (too_big)
+         error (EXIT_TROUBLE, 0, _("exceeded PCRE's line length limit"));
+ 
+-      e = pcre_exec (cre, extra, line_buf, line_end - line_buf,
+-                     start_ofs < 0 ? 0 : start_ofs, 0,
+-                     sub, sizeof sub / sizeof *sub);
++      for (;;)
++        {
++          /* Skip past bytes that are easily determined to be encoding
++             errors, treating them as data that cannot match.  This is
++             faster than having pcre_exec check them.  */
++          while (mbclen_cache[to_uchar (*p)] == (size_t) -1)
++            {
++              p++;
++              bol = false;
++            }
++
++          /* Check for an empty match; this is faster than letting
++             pcre_exec do it.  */
++          int search_bytes = line_end - p;
++          if (search_bytes == 0)
++            {
++              sub[0] = sub[1] = 0;
++              e = empty_match[bol];
++              break;
++            }
++
++          int options = 0;
++          if (!bol)
++            options |= PCRE_NOTBOL;
++          if (multiline)
++            options |= PCRE_NO_UTF8_CHECK;
++
++          e = pcre_exec (cre, extra, p, search_bytes, 0,
++                         options, sub, NSUB);
++          if (e != PCRE_ERROR_BADUTF8)
++            {
++              if (0 < e && multiline && sub[1] - sub[0] != 0)
++                {
++                  char const *nl = memchr (p + sub[0], eolbyte,
++                                           sub[1] - sub[0]);
++                  if (nl)
++                    {
++                      /* This match crosses a line boundary; reject it.  */
++                      p += sub[0];
++                      line_end = nl;
++                      continue;
++                    }
++                }
++              break;
++            }
++          int valid_bytes = sub[0];
++
++          /* Try to match the string before the encoding error.
++             Again, handle the empty-match case specially, for speed.  */
++          if (valid_bytes == 0)
++            {
++              sub[1] = 0;
++              e = empty_match[bol];
++            }
++          else
++            e = pcre_exec (cre, extra, p, valid_bytes, 0,
++                           options | PCRE_NO_UTF8_CHECK | PCRE_NOTEOL,
++                           sub, NSUB);
++          if (e != PCRE_ERROR_NOMATCH)
++            break;
++
++          /* Treat the encoding error as data that cannot match.  */
++          p += valid_bytes + 1;
++          bol = false;
++        }
++
++      if (e != PCRE_ERROR_NOMATCH)
++        break;
++      bol = true;
+     }
+ 
+   if (e <= 0)
+@@ -171,7 +261,7 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+       switch (e)
+         {
+         case PCRE_ERROR_NOMATCH:
+-          return -1;
++          break;
+ 
+         case PCRE_ERROR_NOMEMORY:
+           error (EXIT_TROUBLE, 0, _("memory exhausted"));
+@@ -180,10 +270,6 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+           error (EXIT_TROUBLE, 0,
+                  _("exceeded PCRE's backtracking limit"));
+ 
+-        case PCRE_ERROR_BADUTF8:
+-          error (EXIT_TROUBLE, 0,
+-                 _("invalid UTF-8 byte sequence in input"));
+-
+         default:
+           /* For now, we lump all remaining PCRE failures into this basket.
+              If anyone cares to provide sample grep usage that can trigger
+@@ -192,30 +278,33 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
+           error (EXIT_TROUBLE, 0, _("internal PCRE error: %d"), e);
+         }
+ 
+-      /* NOTREACHED */
+       return -1;
+     }
+   else
+     {
+-      /* Narrow down to the line we've found.  */
+-      char const *beg = line_buf + sub[0];
+-      char const *end = line_buf + sub[1];
+-      char const *buflim = buf + size;
+-      char eol = eolbyte;
+-      if (!start_ptr)
++      char const *matchbeg = p + sub[0];
++      char const *matchend = p + sub[1];
++      char const *beg;
++      char const *end;
++      if (start_ptr)
+         {
+-          /* FIXME: The case when '\n' is not found indicates a bug:
+-             Since grep is line oriented, the match should never contain
+-             a newline, so there _must_ be a newline following.
+-           */
+-          if (!(end = memchr (end, eol, buflim - end)))
+-            end = buflim;
+-          else
+-            end++;
+-          while (buf < beg && beg[-1] != eol)
+-            --beg;
++          beg = matchbeg;
++          end = matchend;
++        }
++      else if (multiline)
++        {
++          char const *prev_nl = memrchr (line_start - 1, eolbyte,
++                                         matchbeg - (line_start - 1));
++          char const *next_nl = memchr (matchend, eolbyte,
++                                        line_end + 1 - matchend);
++          beg = prev_nl + 1;
++          end = next_nl + 1;
++        }
++      else
++        {
++          beg = line_start;
++          end = line_end + 1;
+         }
+-
+       *match_size = end - beg;
+       return beg - buf;
+     }
+diff --git a/src/search.h b/src/search.h
+index 14877bc..e671bea 100644
+--- a/src/search.h
++++ b/src/search.h
+@@ -45,6 +45,7 @@ extern void kwsinit (kwset_t *);
+ 
+ extern char *mbtoupper (char const *, size_t *, mb_len_map_t **);
+ extern void build_mbclen_cache (void);
++extern size_t mbclen_cache[];
+ extern ptrdiff_t mb_goback (char const **, char const *, char const *);
+ extern wint_t mb_prev_wc (char const *, char const *, char const *);
+ extern wint_t mb_next_wc (char const *, char const *);
+diff --git a/src/searchutils.c b/src/searchutils.c
+index 5eb9a12..aba9335 100644
+--- a/src/searchutils.c
++++ b/src/searchutils.c
+@@ -22,7 +22,7 @@
+ 
+ #define NCHAR (UCHAR_MAX + 1)
+ 
+-static size_t mbclen_cache[NCHAR];
++size_t mbclen_cache[NCHAR];
+ 
+ void
+ kwsinit (kwset_t *kwset)
+diff --git a/tests/pcre-infloop b/tests/pcre-infloop
+index 1b33e72..8054844 100755
+--- a/tests/pcre-infloop
++++ b/tests/pcre-infloop
+@@ -18,16 +18,16 @@
+ # along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ 
+ . "${srcdir=.}/init.sh"; path_prepend_ ../src
+-require_pcre_
+ require_timeout_
+ require_en_utf8_locale_
+ require_compiled_in_MB_support
++LC_ALL=en_US.UTF-8 require_pcre_
+ 
+ printf 'a\201b\r' > in || framework_failure_
+ 
+ fail=0
+ 
+ LC_ALL=en_US.UTF-8 timeout 3 grep -P 'a.?..b' in
+-test $? = 2 || fail_ "libpcre's match function appears to infloop"
++test $? = 1 || fail_ "libpcre's match function appears to infloop"
+ 
+ Exit $fail
+diff --git a/tests/pcre-invalid-utf8-input b/tests/pcre-invalid-utf8-input
+index 913e8ee..abcc7e8 100755
+--- a/tests/pcre-invalid-utf8-input
++++ b/tests/pcre-invalid-utf8-input
+@@ -8,14 +8,19 @@
+ # notice and this notice are preserved.
+ 
+ . "${srcdir=.}/init.sh"; path_prepend_ ../src
+-require_pcre_
++require_timeout_
+ require_en_utf8_locale_
++require_compiled_in_MB_support
++LC_ALL=en_US.UTF-8 require_pcre_
+ 
+ fail=0
+ 
+-printf 'j\202\nj\n' > in || framework_failure_
++printf 'j\202j\nj\nk\202\n' > in || framework_failure_
+ 
+-LC_ALL=en_US.UTF-8 grep -P j in
+-test $? -eq 2 || fail=1
++LC_ALL=en_US.UTF-8 timeout 3 grep -P j in
++test $? -eq 0 || fail=1
++
++LC_ALL=en_US.UTF-8 timeout 3 grep -P 'k$' in
++test $? -eq 1 || fail=1
+ 
+ Exit $fail
+diff --git a/tests/pcre-utf8 b/tests/pcre-utf8
+index 41676f4..2dda116 100755
+--- a/tests/pcre-utf8
++++ b/tests/pcre-utf8
+@@ -8,8 +8,8 @@
+ # notice and this notice are preserved.
+ 
+ . "${srcdir=.}/init.sh"; path_prepend_ ../src
+-require_pcre_
+ require_en_utf8_locale_
++LC_ALL=en_US.UTF-8 require_pcre_
+ 
+ fail=0
+ 
diff --git a/grep-2.20-pcre-invalid-utf8-fix.patch b/grep-2.20-pcre-invalid-utf8-fix.patch
deleted file mode 100644
index 5f7530f..0000000
--- a/grep-2.20-pcre-invalid-utf8-fix.patch
+++ /dev/null
@@ -1,136 +0,0 @@
-diff --git a/src/pcresearch.c b/src/pcresearch.c
-index 820dd00..11df488 100644
---- a/src/pcresearch.c
-+++ b/src/pcresearch.c
-@@ -136,34 +136,42 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
- #else
-   /* This array must have at least two elements; everything after that
-      is just for performance improvement in pcre_exec.  */
--  int sub[300];
-+  enum { nsub = 300 };
-+  int sub[nsub];
- 
--  const char *line_buf, *line_end, *line_next;
-+  char const *p = start_ptr ? start_ptr : buf;
-+  int options = p == buf || p[-1] == eolbyte ? 0 : PCRE_NOTBOL;
-+  char const *line_start = buf;
-   int e = PCRE_ERROR_NOMATCH;
--  ptrdiff_t start_ofs = start_ptr ? start_ptr - buf : 0;
-+  char const *line_end;
- 
-   /* PCRE can't limit the matching to single lines, therefore we have to
-      match each line in the buffer separately.  */
--  for (line_next = buf;
--       e == PCRE_ERROR_NOMATCH && line_next < buf + size;
--       start_ofs -= line_next - line_buf)
-+  for (; p < buf + size; p = line_start = line_end + 1)
-     {
--      line_buf = line_next;
--      line_end = memchr (line_buf, eolbyte, (buf + size) - line_buf);
--      if (line_end == NULL)
--        line_next = line_end = buf + size;
--      else
--        line_next = line_end + 1;
--
--      if (start_ptr && start_ptr >= line_end)
--        continue;
-+      line_end = memchr (p, eolbyte, buf + size - p);
- 
--      if (INT_MAX < line_end - line_buf)
-+      if (INT_MAX < line_end - p)
-         error (EXIT_TROUBLE, 0, _("exceeded PCRE's line length limit"));
- 
--      e = pcre_exec (cre, extra, line_buf, line_end - line_buf,
--                     start_ofs < 0 ? 0 : start_ofs, 0,
--                     sub, sizeof sub / sizeof *sub);
-+      /* Treat encoding-error bytes as data that cannot match.  */
-+      for (;;)
-+        {
-+          e = pcre_exec (cre, extra, p, line_end - p, 0, options, sub, nsub);
-+          if (e != PCRE_ERROR_BADUTF8)
-+            break;
-+          e = pcre_exec (cre, extra, p, sub[0], 0,
-+                         options | PCRE_NO_UTF8_CHECK | PCRE_NOTEOL,
-+                         sub, nsub);
-+          if (e != PCRE_ERROR_NOMATCH)
-+            break;
-+          p += sub[0] + 1;
-+          options = PCRE_NOTBOL;
-+        }
-+
-+      if (e != PCRE_ERROR_NOMATCH)
-+        break;
-+      options = 0;
-     }
- 
-   if (e <= 0)
-@@ -180,10 +188,6 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-           error (EXIT_TROUBLE, 0,
-                  _("exceeded PCRE's backtracking limit"));
- 
--        case PCRE_ERROR_BADUTF8:
--          error (EXIT_TROUBLE, 0,
--                 _("invalid UTF-8 byte sequence in input"));
--
-         default:
-           /* For now, we lump all remaining PCRE failures into this basket.
-              If anyone cares to provide sample grep usage that can trigger
-@@ -197,25 +201,8 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-     }
-   else
-     {
--      /* Narrow down to the line we've found.  */
--      char const *beg = line_buf + sub[0];
--      char const *end = line_buf + sub[1];
--      char const *buflim = buf + size;
--      char eol = eolbyte;
--      if (!start_ptr)
--        {
--          /* FIXME: The case when '\n' is not found indicates a bug:
--             Since grep is line oriented, the match should never contain
--             a newline, so there _must_ be a newline following.
--           */
--          if (!(end = memchr (end, eol, buflim - end)))
--            end = buflim;
--          else
--            end++;
--          while (buf < beg && beg[-1] != eol)
--            --beg;
--        }
--
-+      char const *beg = start_ptr ? p + sub[0] : line_start;
-+      char const *end = start_ptr ? p + sub[1] : line_end + 1;
-       *match_size = end - beg;
-       return beg - buf;
-     }
-diff --git a/tests/pcre-infloop b/tests/pcre-infloop
-index 1b33e72..b92f8e1 100755
---- a/tests/pcre-infloop
-+++ b/tests/pcre-infloop
-@@ -28,6 +28,6 @@ printf 'a\201b\r' > in || framework_failure_
- fail=0
- 
- LC_ALL=en_US.UTF-8 timeout 3 grep -P 'a.?..b' in
--test $? = 2 || fail_ "libpcre's match function appears to infloop"
-+test $? = 1 || fail_ "libpcre's match function appears to infloop"
- 
- Exit $fail
-diff --git a/tests/pcre-invalid-utf8-input b/tests/pcre-invalid-utf8-input
-index 913e8ee..9da4b18 100755
---- a/tests/pcre-invalid-utf8-input
-+++ b/tests/pcre-invalid-utf8-input
-@@ -13,9 +13,12 @@ require_en_utf8_locale_
- 
- fail=0
- 
--printf 'j\202\nj\n' > in || framework_failure_
-+printf 'j\202j\nj\nk\202\n' > in || framework_failure_
- 
- LC_ALL=en_US.UTF-8 grep -P j in
--test $? -eq 2 || fail=1
-+test $? -eq 0 || fail=1
-+
-+LC_ALL=en_US.UTF-8 grep -P 'k$' in
-+test $? -eq 1 || fail=1
- 
- Exit $fail
diff --git a/grep.spec b/grep.spec
index 6870319..f227ad1 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.20
-Release: 5%{?dist}
+Release: 6%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -15,7 +15,7 @@ Patch0: grep-2.20-man-fix-gs.patch
 # upstream ticket 39445
 Patch1: grep-2.20-help-align.patch
 # backported from upstream
-Patch2: grep-2.20-pcre-invalid-utf8-fix.patch
+Patch2: grep-2.20-pcre-backported-fixes.patch
 URL: http://www.gnu.org/software/grep/
 Requires(post): /sbin/install-info
 Requires(preun): /sbin/install-info
@@ -36,7 +36,7 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
 %setup -q
 %patch0 -p1 -b .man-fix-gs
 %patch1 -p1 -b .help-align
-%patch2 -p1 -b .pcre-invalid-utf8-fix
+%patch2 -p1 -b .pcre-backported-fixes
 
 %build
 %global BUILD_FLAGS $RPM_OPT_FLAGS
@@ -90,6 +90,10 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Fri Nov 14 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.20-6
+- Backported more PCRE fixes (by pcre-backported-fixes patch)
+- Dropped pcre-invalid-utf8-fix patch, handled by pcre-backported-fixes patch
+
 * Tue Nov 11 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.20-5
 - Fixed invalid UTF-8 byte sequence error in PCRE mode
   (by pcre-invalid-utf8-fix patch)

From 4397c94fae30391a503f9404ed93141e6455a570 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
Date: Tue, 25 Nov 2014 11:16:00 +0100
Subject: [PATCH 5/8] New version

  Resolves: rhbz#1167657
- De-fuzzified patches
- Dropped pcre-backported-fixes patch (not needed)
---
 grep-2.20-pcre-backported-fixes.patch         | 389 ------------------
 ...-align.patch => grep-2.21-help-align.patch |   9 +-
 ...fix-gs.patch => grep-2.21-man-fix-gs.patch |  12 +-
 grep.spec                                     |  17 +-
 sources                                       |   2 +-
 5 files changed, 22 insertions(+), 407 deletions(-)
 delete mode 100644 grep-2.20-pcre-backported-fixes.patch
 rename grep-2.20-help-align.patch => grep-2.21-help-align.patch (92%)
 rename grep-2.20-man-fix-gs.patch => grep-2.21-man-fix-gs.patch (89%)

diff --git a/grep-2.20-pcre-backported-fixes.patch b/grep-2.20-pcre-backported-fixes.patch
deleted file mode 100644
index 4a9dbcd..0000000
--- a/grep-2.20-pcre-backported-fixes.patch
+++ /dev/null
@@ -1,389 +0,0 @@
-diff --git a/src/grep.h b/src/grep.h
-index 4935872..729c906 100644
---- a/src/grep.h
-+++ b/src/grep.h
-@@ -27,4 +27,19 @@ extern int match_words;		/* -w */
- extern int match_lines;		/* -x */
- extern unsigned char eolbyte;	/* -z */
- 
-+/* An enum textbin describes the file's type, inferred from data read
-+   before the first line is selected for output.  */
-+enum textbin
-+  {
-+    /* Binary, as it contains null bytes and the -z option is not in effect,
-+       or it contains encoding errors.  */
-+    TEXTBIN_BINARY = -1,
-+
-+    /* Not known yet.  Only text has been seen so far.  */
-+    TEXTBIN_UNKNOWN = 0,
-+
-+    /* Text.  */
-+    TEXTBIN_TEXT = 1
-+  };
-+
- #endif
-diff --git a/src/pcresearch.c b/src/pcresearch.c
-index 820dd00..9938ffc 100644
---- a/src/pcresearch.c
-+++ b/src/pcresearch.c
-@@ -33,13 +33,19 @@ static pcre *cre;
- /* Additional information about the pattern.  */
- static pcre_extra *extra;
- 
--# ifdef PCRE_STUDY_JIT_COMPILE
--static pcre_jit_stack *jit_stack;
--# else
-+# ifndef PCRE_STUDY_JIT_COMPILE
- #  define PCRE_STUDY_JIT_COMPILE 0
- # endif
- #endif
- 
-+/* Table, indexed by ! (flag & PCRE_NOTBOL), of whether the empty
-+   string matches when that flag is used.  */
-+static int empty_match[2];
-+
-+/* This must be at least 2; everything after that is for performance
-+   in pcre_exec.  */
-+enum { NSUB = 300 };
-+
- void
- Pcompile (char const *pattern, size_t size)
- {
-@@ -52,13 +58,17 @@ Pcompile (char const *pattern, size_t size)
-   char const *ep;
-   char *re = xnmalloc (4, size + 7);
-   int flags = (PCRE_MULTILINE
--               | (match_icase ? PCRE_CASELESS : 0)
--               | (using_utf8 () ? PCRE_UTF8 : 0));
-+               | (match_icase ? PCRE_CASELESS : 0));
-   char const *patlim = pattern + size;
-   char *n = re;
-   char const *p;
-   char const *pnul;
- 
-+  if (using_utf8 ())
-+    flags |= PCRE_UTF8;
-+  else if (MB_CUR_MAX != 1)
-+    error (EXIT_TROUBLE, 0, _("-P supports only unibyte and UTF-8 locales"));
-+
-   /* FIXME: Remove these restrictions.  */
-   if (memchr (pattern, '\n', size))
-     error (EXIT_TROUBLE, 0, _("the -P option only supports a single pattern"));
-@@ -114,14 +124,20 @@ Pcompile (char const *pattern, size_t size)
-       /* A 32K stack is allocated for the machine code by default, which
-          can grow to 512K if necessary. Since JIT uses far less memory
-          than the interpreter, this should be enough in practice.  */
--      jit_stack = pcre_jit_stack_alloc (32 * 1024, 512 * 1024);
-+      pcre_jit_stack *jit_stack = pcre_jit_stack_alloc (32 * 1024, 512 * 1024);
-       if (!jit_stack)
-         error (EXIT_TROUBLE, 0,
-                _("failed to allocate memory for the PCRE JIT stack"));
-       pcre_assign_jit_stack (extra, NULL, jit_stack);
-     }
-+
- # endif
-   free (re);
-+
-+  int sub[NSUB];
-+  empty_match[false] = pcre_exec (cre, extra, "", 0, 0,
-+                                  PCRE_NOTBOL, sub, NSUB);
-+  empty_match[true] = pcre_exec (cre, extra, "", 0, 0, 0, sub, NSUB);
- #endif /* HAVE_LIBPCRE */
- }
- 
-@@ -134,36 +150,110 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-   error (EXIT_TROUBLE, 0, _("internal error"));
-   return -1;
- #else
--  /* This array must have at least two elements; everything after that
--     is just for performance improvement in pcre_exec.  */
--  int sub[300];
--
--  const char *line_buf, *line_end, *line_next;
-+  int sub[NSUB];
-+  char const *p = start_ptr ? start_ptr : buf;
-+  bool bol = p[-1] == eolbyte;
-+  char const *line_start = buf;
-   int e = PCRE_ERROR_NOMATCH;
--  ptrdiff_t start_ofs = start_ptr ? start_ptr - buf : 0;
-+  char const *line_end;
- 
--  /* PCRE can't limit the matching to single lines, therefore we have to
--     match each line in the buffer separately.  */
--  for (line_next = buf;
--       e == PCRE_ERROR_NOMATCH && line_next < buf + size;
--       start_ofs -= line_next - line_buf)
-+  /* If the input type is unknown, the caller is still testing the
-+     input, which means the current buffer cannot contain encoding
-+     errors and a multiline search is typically more efficient.
-+     Otherwise, a single-line search is typically faster, so that
-+     pcre_exec doesn't waste time validating the entire input
-+     buffer.  */
-+  bool multiline = TEXTBIN_UNKNOWN;
-+
-+  for (; p < buf + size; p = line_start = line_end + 1)
-     {
--      line_buf = line_next;
--      line_end = memchr (line_buf, eolbyte, (buf + size) - line_buf);
--      if (line_end == NULL)
--        line_next = line_end = buf + size;
--      else
--        line_next = line_end + 1;
-+      bool too_big;
- 
--      if (start_ptr && start_ptr >= line_end)
--        continue;
-+      if (multiline)
-+        {
-+          size_t pcre_size_max = MIN (INT_MAX, SIZE_MAX - 1);
-+          size_t scan_size = MIN (pcre_size_max + 1, buf + size - p);
-+          line_end = memrchr (p, eolbyte, scan_size);
-+          too_big = ! line_end;
-+        }
-+      else
-+        {
-+          line_end = memchr (p, eolbyte, buf + size - p);
-+          too_big = INT_MAX < line_end - p;
-+        }
- 
--      if (INT_MAX < line_end - line_buf)
-+      if (too_big)
-         error (EXIT_TROUBLE, 0, _("exceeded PCRE's line length limit"));
- 
--      e = pcre_exec (cre, extra, line_buf, line_end - line_buf,
--                     start_ofs < 0 ? 0 : start_ofs, 0,
--                     sub, sizeof sub / sizeof *sub);
-+      for (;;)
-+        {
-+          /* Skip past bytes that are easily determined to be encoding
-+             errors, treating them as data that cannot match.  This is
-+             faster than having pcre_exec check them.  */
-+          while (mbclen_cache[to_uchar (*p)] == (size_t) -1)
-+            {
-+              p++;
-+              bol = false;
-+            }
-+
-+          /* Check for an empty match; this is faster than letting
-+             pcre_exec do it.  */
-+          int search_bytes = line_end - p;
-+          if (search_bytes == 0)
-+            {
-+              sub[0] = sub[1] = 0;
-+              e = empty_match[bol];
-+              break;
-+            }
-+
-+          int options = 0;
-+          if (!bol)
-+            options |= PCRE_NOTBOL;
-+          if (multiline)
-+            options |= PCRE_NO_UTF8_CHECK;
-+
-+          e = pcre_exec (cre, extra, p, search_bytes, 0,
-+                         options, sub, NSUB);
-+          if (e != PCRE_ERROR_BADUTF8)
-+            {
-+              if (0 < e && multiline && sub[1] - sub[0] != 0)
-+                {
-+                  char const *nl = memchr (p + sub[0], eolbyte,
-+                                           sub[1] - sub[0]);
-+                  if (nl)
-+                    {
-+                      /* This match crosses a line boundary; reject it.  */
-+                      p += sub[0];
-+                      line_end = nl;
-+                      continue;
-+                    }
-+                }
-+              break;
-+            }
-+          int valid_bytes = sub[0];
-+
-+          /* Try to match the string before the encoding error.
-+             Again, handle the empty-match case specially, for speed.  */
-+          if (valid_bytes == 0)
-+            {
-+              sub[1] = 0;
-+              e = empty_match[bol];
-+            }
-+          else
-+            e = pcre_exec (cre, extra, p, valid_bytes, 0,
-+                           options | PCRE_NO_UTF8_CHECK | PCRE_NOTEOL,
-+                           sub, NSUB);
-+          if (e != PCRE_ERROR_NOMATCH)
-+            break;
-+
-+          /* Treat the encoding error as data that cannot match.  */
-+          p += valid_bytes + 1;
-+          bol = false;
-+        }
-+
-+      if (e != PCRE_ERROR_NOMATCH)
-+        break;
-+      bol = true;
-     }
- 
-   if (e <= 0)
-@@ -171,7 +261,7 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-       switch (e)
-         {
-         case PCRE_ERROR_NOMATCH:
--          return -1;
-+          break;
- 
-         case PCRE_ERROR_NOMEMORY:
-           error (EXIT_TROUBLE, 0, _("memory exhausted"));
-@@ -180,10 +270,6 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-           error (EXIT_TROUBLE, 0,
-                  _("exceeded PCRE's backtracking limit"));
- 
--        case PCRE_ERROR_BADUTF8:
--          error (EXIT_TROUBLE, 0,
--                 _("invalid UTF-8 byte sequence in input"));
--
-         default:
-           /* For now, we lump all remaining PCRE failures into this basket.
-              If anyone cares to provide sample grep usage that can trigger
-@@ -192,30 +278,33 @@ Pexecute (char const *buf, size_t size, size_t *match_size,
-           error (EXIT_TROUBLE, 0, _("internal PCRE error: %d"), e);
-         }
- 
--      /* NOTREACHED */
-       return -1;
-     }
-   else
-     {
--      /* Narrow down to the line we've found.  */
--      char const *beg = line_buf + sub[0];
--      char const *end = line_buf + sub[1];
--      char const *buflim = buf + size;
--      char eol = eolbyte;
--      if (!start_ptr)
-+      char const *matchbeg = p + sub[0];
-+      char const *matchend = p + sub[1];
-+      char const *beg;
-+      char const *end;
-+      if (start_ptr)
-         {
--          /* FIXME: The case when '\n' is not found indicates a bug:
--             Since grep is line oriented, the match should never contain
--             a newline, so there _must_ be a newline following.
--           */
--          if (!(end = memchr (end, eol, buflim - end)))
--            end = buflim;
--          else
--            end++;
--          while (buf < beg && beg[-1] != eol)
--            --beg;
-+          beg = matchbeg;
-+          end = matchend;
-+        }
-+      else if (multiline)
-+        {
-+          char const *prev_nl = memrchr (line_start - 1, eolbyte,
-+                                         matchbeg - (line_start - 1));
-+          char const *next_nl = memchr (matchend, eolbyte,
-+                                        line_end + 1 - matchend);
-+          beg = prev_nl + 1;
-+          end = next_nl + 1;
-+        }
-+      else
-+        {
-+          beg = line_start;
-+          end = line_end + 1;
-         }
--
-       *match_size = end - beg;
-       return beg - buf;
-     }
-diff --git a/src/search.h b/src/search.h
-index 14877bc..e671bea 100644
---- a/src/search.h
-+++ b/src/search.h
-@@ -45,6 +45,7 @@ extern void kwsinit (kwset_t *);
- 
- extern char *mbtoupper (char const *, size_t *, mb_len_map_t **);
- extern void build_mbclen_cache (void);
-+extern size_t mbclen_cache[];
- extern ptrdiff_t mb_goback (char const **, char const *, char const *);
- extern wint_t mb_prev_wc (char const *, char const *, char const *);
- extern wint_t mb_next_wc (char const *, char const *);
-diff --git a/src/searchutils.c b/src/searchutils.c
-index 5eb9a12..aba9335 100644
---- a/src/searchutils.c
-+++ b/src/searchutils.c
-@@ -22,7 +22,7 @@
- 
- #define NCHAR (UCHAR_MAX + 1)
- 
--static size_t mbclen_cache[NCHAR];
-+size_t mbclen_cache[NCHAR];
- 
- void
- kwsinit (kwset_t *kwset)
-diff --git a/tests/pcre-infloop b/tests/pcre-infloop
-index 1b33e72..8054844 100755
---- a/tests/pcre-infloop
-+++ b/tests/pcre-infloop
-@@ -18,16 +18,16 @@
- # along with this program.  If not, see <http://www.gnu.org/licenses/>.
- 
- . "${srcdir=.}/init.sh"; path_prepend_ ../src
--require_pcre_
- require_timeout_
- require_en_utf8_locale_
- require_compiled_in_MB_support
-+LC_ALL=en_US.UTF-8 require_pcre_
- 
- printf 'a\201b\r' > in || framework_failure_
- 
- fail=0
- 
- LC_ALL=en_US.UTF-8 timeout 3 grep -P 'a.?..b' in
--test $? = 2 || fail_ "libpcre's match function appears to infloop"
-+test $? = 1 || fail_ "libpcre's match function appears to infloop"
- 
- Exit $fail
-diff --git a/tests/pcre-invalid-utf8-input b/tests/pcre-invalid-utf8-input
-index 913e8ee..abcc7e8 100755
---- a/tests/pcre-invalid-utf8-input
-+++ b/tests/pcre-invalid-utf8-input
-@@ -8,14 +8,19 @@
- # notice and this notice are preserved.
- 
- . "${srcdir=.}/init.sh"; path_prepend_ ../src
--require_pcre_
-+require_timeout_
- require_en_utf8_locale_
-+require_compiled_in_MB_support
-+LC_ALL=en_US.UTF-8 require_pcre_
- 
- fail=0
- 
--printf 'j\202\nj\n' > in || framework_failure_
-+printf 'j\202j\nj\nk\202\n' > in || framework_failure_
- 
--LC_ALL=en_US.UTF-8 grep -P j in
--test $? -eq 2 || fail=1
-+LC_ALL=en_US.UTF-8 timeout 3 grep -P j in
-+test $? -eq 0 || fail=1
-+
-+LC_ALL=en_US.UTF-8 timeout 3 grep -P 'k$' in
-+test $? -eq 1 || fail=1
- 
- Exit $fail
-diff --git a/tests/pcre-utf8 b/tests/pcre-utf8
-index 41676f4..2dda116 100755
---- a/tests/pcre-utf8
-+++ b/tests/pcre-utf8
-@@ -8,8 +8,8 @@
- # notice and this notice are preserved.
- 
- . "${srcdir=.}/init.sh"; path_prepend_ ../src
--require_pcre_
- require_en_utf8_locale_
-+LC_ALL=en_US.UTF-8 require_pcre_
- 
- fail=0
- 
diff --git a/grep-2.20-help-align.patch b/grep-2.21-help-align.patch
similarity index 92%
rename from grep-2.20-help-align.patch
rename to grep-2.21-help-align.patch
index cef6311..56f24d6 100644
--- a/grep-2.20-help-align.patch
+++ b/grep-2.21-help-align.patch
@@ -1,8 +1,8 @@
 diff --git a/src/grep.c b/src/grep.c
-index 0fcc272..2208a4e 100644
+index e3461a7..50a9868 100644
 --- a/src/grep.c
 +++ b/src/grep.c
-@@ -1579,16 +1579,19 @@ Output control:\n\
+@@ -1757,17 +1757,20 @@ Output control:\n\
    -D, --devices=ACTION      how to handle devices, FIFOs and sockets;\n\
                              ACTION is 'read' or 'skip'\n\
    -r, --recursive           like --directories=recurse\n\
@@ -12,11 +12,12 @@ index 0fcc272..2208a4e 100644
  "));
        printf (_("\
 -      --include=FILE_PATTERN  search only files that match FILE_PATTERN\n\
--      --exclude=FILE_PATTERN  skip files and directories matching FILE_PATTERN\n\
+-      --exclude=FILE_PATTERN  skip files and directories matching\
 +      --include=FILE_PATTERN\n\
 +                            search only files that match FILE_PATTERN\n\
 +      --exclude=FILE_PATTERN\n\
-+                            skip files and directories matching FILE_PATTERN\n\
++                            skip files and directories matching\
+  FILE_PATTERN\n\
        --exclude-from=FILE   skip files matching any file pattern from FILE\n\
 -      --exclude-dir=PATTERN  directories that match PATTERN will be skipped.\n\
 +      --exclude-dir=PATTERN directories that match PATTERN will be skipped.\n\
diff --git a/grep-2.20-man-fix-gs.patch b/grep-2.21-man-fix-gs.patch
similarity index 89%
rename from grep-2.20-man-fix-gs.patch
rename to grep-2.21-man-fix-gs.patch
index db3dd24..65f41bf 100644
--- a/grep-2.20-man-fix-gs.patch
+++ b/grep-2.21-man-fix-gs.patch
@@ -1,8 +1,8 @@
 diff --git a/doc/grep.in.1 b/doc/grep.in.1
-index 58a6c0e..3e6a8cf 100644
+index b6362ee..5a1e3ea 100644
 --- a/doc/grep.in.1
 +++ b/doc/grep.in.1
-@@ -377,7 +377,7 @@ Print
+@@ -314,7 +314,7 @@ Print
  .I NUM
  lines of trailing context after matching lines.
  Places a line containing a group separator
@@ -11,7 +11,7 @@ index 58a6c0e..3e6a8cf 100644
  between contiguous groups of matches.
  With the
  .B \-o
-@@ -390,7 +390,7 @@ Print
+@@ -327,7 +327,7 @@ Print
  .I NUM
  lines of leading context before matching lines.
  Places a line containing a group separator
@@ -20,7 +20,7 @@ index 58a6c0e..3e6a8cf 100644
  between contiguous groups of matches.
  With the
  .B \-o
-@@ -403,13 +403,24 @@ Print
+@@ -340,13 +340,24 @@ Print
  .I NUM
  lines of output context.
  Places a line containing a group separator
@@ -47,10 +47,10 @@ index 58a6c0e..3e6a8cf 100644
  .TP
  .BR \-a ", " \-\^\-text
 diff --git a/src/grep.c b/src/grep.c
-index 7c0f8a8..0fcc272 100644
+index 8dbf86e..e3461a7 100644
 --- a/src/grep.c
 +++ b/src/grep.c
-@@ -1602,6 +1602,8 @@ Context control:\n\
+@@ -1781,6 +1781,8 @@ Context control:\n\
  "));
        printf (_("\
    -NUM                      same as --context=NUM\n\
diff --git a/grep.spec b/grep.spec
index f227ad1..441dae3 100644
--- a/grep.spec
+++ b/grep.spec
@@ -2,8 +2,8 @@
 
 Summary: Pattern matching utilities
 Name: grep
-Version: 2.20
-Release: 6%{?dist}
+Version: 2.21
+Release: 1%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -11,11 +11,9 @@ Source1: colorgrep.sh
 Source2: colorgrep.csh
 Source3: GREP_COLORS
 # upstream ticket 39444
-Patch0: grep-2.20-man-fix-gs.patch
+Patch0: grep-2.21-man-fix-gs.patch
 # upstream ticket 39445
-Patch1: grep-2.20-help-align.patch
-# backported from upstream
-Patch2: grep-2.20-pcre-backported-fixes.patch
+Patch1: grep-2.21-help-align.patch
 URL: http://www.gnu.org/software/grep/
 Requires(post): /sbin/install-info
 Requires(preun): /sbin/install-info
@@ -36,7 +34,6 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
 %setup -q
 %patch0 -p1 -b .man-fix-gs
 %patch1 -p1 -b .help-align
-%patch2 -p1 -b .pcre-backported-fixes
 
 %build
 %global BUILD_FLAGS $RPM_OPT_FLAGS
@@ -90,6 +87,12 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Tue Nov 25 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21-1
+- New version
+  Resolves: rhbz#1167657
+- De-fuzzified patches
+- Dropped pcre-backported-fixes patch (not needed)
+
 * Fri Nov 14 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.20-6
 - Backported more PCRE fixes (by pcre-backported-fixes patch)
 - Dropped pcre-invalid-utf8-fix patch, handled by pcre-backported-fixes patch
diff --git a/sources b/sources
index fe1f529..d9d5863 100644
--- a/sources
+++ b/sources
@@ -1 +1 @@
-2cbea44a4f1548aee20b9ff2d3076908  grep-2.20.tar.xz
+43c48064d6409862b8a850db83c8038a  grep-2.21.tar.xz

From 9ce00030a5baad778f208ccf9ea92f58f101508f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
Date: Tue, 20 Jan 2015 15:30:25 +0100
Subject: [PATCH 6/8] Fixed buffer overrun for grep -F

  Resolves: rhbz#1183653
---
 grep-2.21-buf-overrun-fix.patch | 146 ++++++++++++++++++++++++++++++++
 grep.spec                       |  11 ++-
 2 files changed, 156 insertions(+), 1 deletion(-)
 create mode 100644 grep-2.21-buf-overrun-fix.patch

diff --git a/grep-2.21-buf-overrun-fix.patch b/grep-2.21-buf-overrun-fix.patch
new file mode 100644
index 0000000..661a3ff
--- /dev/null
+++ b/grep-2.21-buf-overrun-fix.patch
@@ -0,0 +1,146 @@
+From 83a95bd8c8561875b948cadd417c653dbe7ef2e2 Mon Sep 17 00:00:00 2001
+From: Yuliy Pisetsky <ypisetsky@fb.com>
+Date: Thu, 01 Jan 2015 23:36:55 +0000
+Subject: grep -F: fix a heap buffer (read) overrun
+
+grep's read buffer is often filled to its full size, except when
+reading the final buffer of a file.  In that case, the number of
+bytes read may be far less than the size of the buffer.  However, for
+certain unusual pattern/text combinations, grep -F would mistakenly
+examine bytes in that uninitialized region of memory when searching
+for a match.  With carefully chosen inputs, one can cause grep -F to
+read beyond the end of that buffer altogether.  This problem arose via
+commit v2.18-90-g73893ff with the introduction of a more efficient
+heuristic using what is now the memchr_kwset function. The use of
+that function in bmexec_trans could leave TP much larger than EP,
+and the subsequent call to bm_delta2_search would mistakenly access
+beyond end of the main input read buffer.
+
+* src/kwset.c (bmexec_trans): When TP reaches or exceeds EP,
+do not call bm_delta2_search.
+* tests/kwset-abuse: New file.
+* tests/Makefile.am (TESTS): Add it.
+* THANKS.in: Update.
+* NEWS (Bug fixes): Mention it.
+
+Prior to this patch, this command would trigger a UMR:
+
+  printf %0360db 0 | valgrind src/grep -F $(printf %019dXb 0)
+
+  Use of uninitialised value of size 8
+     at 0x4142BE: bmexec_trans (kwset.c:657)
+     by 0x4143CA: bmexec (kwset.c:678)
+     by 0x414973: kwsexec (kwset.c:848)
+     by 0x414DC4: Fexecute (kwsearch.c:128)
+     by 0x404E2E: grepbuf (grep.c:1238)
+     by 0x4054BF: grep (grep.c:1417)
+     by 0x405CEB: grepdesc (grep.c:1645)
+     by 0x405EC1: grep_command_line_arg (grep.c:1692)
+     by 0x4077D4: main (grep.c:2570)
+
+See the accompanying test for how to trigger the heap buffer overrun.
+
+Thanks to Nima Aghdaii for testing and finding numerous
+ways to break early iterations of this patch.
+---
+--- a/THANKS.in
++++ b/THANKS.in
+@@ -62,6 +62,7 @@ Michael Aichlmayr                   mikla@nx.com
+ Miles Bader                         miles@ccs.mt.nec.co.jp
+ Mirraz Mirraz                       mirraz1@rambler.ru
+ Nelson H. F. Beebe                  beebe@math.utah.edu
++Nima Aghdaii                        naghdaii@fb.com
+ Olaf Kirch                          okir@ns.lst.de
+ Paul Kimoto                         kimoto@spacenet.tn.cornell.edu
+ Péter Radics                        mitchnull@gmail.com
+diff --git a/src/kwset.c b/src/kwset.c
+index 6d21893..998dbfe 100644
+--- a/src/kwset.c
++++ b/src/kwset.c
+@@ -643,6 +643,8 @@ bmexec_trans (kwset_t kwset, char const *text, size_t size)
+                     if (! tp)
+                       return -1;
+                     tp++;
++                    if (ep <= tp)
++                      break;
+                   }
+               }
+           }
+diff --git a/tests/Makefile.am b/tests/Makefile.am
+index 217a731..2f69835 100644
+--- a/tests/Makefile.am
++++ b/tests/Makefile.am
+@@ -72,6 +72,7 @@ TESTS =						\
+   inconsistent-range				\
+   invalid-multibyte-infloop			\
+   khadafy					\
++  kwset-abuse					\
+   long-line-vs-2GiB-read			\
+   match-lines					\
+   max-count-overread				\
+diff --git a/tests/Makefile.in b/tests/Makefile.in
+index e40a070..9ecafe7 100644
+--- a/tests/Makefile.in
++++ b/tests/Makefile.in
+@@ -1376,6 +1376,7 @@ TESTS = \
+   inconsistent-range				\
+   invalid-multibyte-infloop			\
+   khadafy					\
++  kwset-abuse					\
+   long-line-vs-2GiB-read			\
+   match-lines					\
+   max-count-overread				\
+@@ -2030,6 +2031,13 @@
+ 	$(am__check_pre) $(LOG_DRIVER) --test-name "$$f" \
+ 	--log-file $$b.log --trs-file $$b.trs \
+ 	$(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
++	"$$tst" $(AM_TESTS_FD_REDIRECT)
++kwset-abuse.log: kwset-abuse
++	@p='kwset-abuse'; \
++	b='kwset-abuse'; \
++	$(am__check_pre) $(LOG_DRIVER) --test-name "$$f" \
++	--log-file $$b.log --trs-file $$b.trs \
++	$(am__common_driver_flags) $(AM_LOG_DRIVER_FLAGS) $(LOG_DRIVER_FLAGS) -- $(LOG_COMPILE) \
+ 	"$$tst" $(AM_TESTS_FD_REDIRECT)
+ long-line-vs-2GiB-read.log: long-line-vs-2GiB-read
+ 	@p='long-line-vs-2GiB-read'; \
+diff --git a/tests/kwset-abuse b/tests/kwset-abuse
+new file mode 100755
+index 0000000..6d8ec0c
+--- a/dev/null
++++ b/tests/kwset-abuse
+@@ -0,0 +1,32 @@
++#! /bin/sh
++# Evoke a segfault in a hard-to-reach code path of kwset.c.
++# This bug affected grep versions 2.19 through 2.21.
++#
++# Copyright (C) 2015 Free Software Foundation, Inc.
++#
++# This program is free software: you can redistribute it and/or modify
++# it under the terms of the GNU General Public License as published by
++# the Free Software Foundation, either version 3 of the License, or
++# (at your option) any later version.
++
++# This program is distributed in the hope that it will be useful,
++# but WITHOUT ANY WARRANTY; without even the implied warranty of
++# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
++# GNU General Public License for more details.
++
++# You should have received a copy of the GNU General Public License
++# along with this program.  If not, see <http://www.gnu.org/licenses/>.
++
++. "${srcdir=.}/init.sh"; path_prepend_ ../src
++
++fail=0
++
++# This test case chooses a haystack of size 260,000, since prodding
++# with gdb showed a reallocation slightly larger than that in fillbuf.
++# To reach the buggy code, the needle must have length < 1/11 that of
++# the haystack, and 10,000 is a nice round number that fits the bill.
++printf '%0260000dXy\n' 0 | grep -F $(printf %010000dy 0)
++
++test $? = 1 || fail=1
++
++Exit $fail
+--
+cgit v0.9.0.2
diff --git a/grep.spec b/grep.spec
index 441dae3..26cad67 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.21
-Release: 1%{?dist}
+Release: 2%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -14,6 +14,8 @@ Source3: GREP_COLORS
 Patch0: grep-2.21-man-fix-gs.patch
 # upstream ticket 39445
 Patch1: grep-2.21-help-align.patch
+# fix buffer overrun for grep -F, rhbz#1183653
+Patch2: grep-2.21-buf-overrun-fix.patch
 URL: http://www.gnu.org/software/grep/
 Requires(post): /sbin/install-info
 Requires(preun): /sbin/install-info
@@ -34,6 +36,9 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
 %setup -q
 %patch0 -p1 -b .man-fix-gs
 %patch1 -p1 -b .help-align
+%patch2 -p1 -b .buf-overrun-fix
+
+chmod 755 tests/kwset-abuse
 
 %build
 %global BUILD_FLAGS $RPM_OPT_FLAGS
@@ -87,6 +92,10 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Tue Jan 20 2015 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21-2
+- Fixed buffer overrun for grep -F
+  Resolves: rhbz#1183653
+
 * Tue Nov 25 2014 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21-1
 - New version
   Resolves: rhbz#1167657

From da0d7afa9eee3bb7bfecedd506cfe04039e5f19b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaroslav=20=C5=A0karvada?= <jskarvad@redhat.com>
Date: Tue, 7 Apr 2015 11:16:45 +0200
Subject: [PATCH 7/8] Documented change in behaviour of recurse option

  Resolves: rhbz#1178305
---
 grep-2.21-recurse-behaviour-change-doc.patch | 24 ++++++++++++++++++++
 grep.spec                                    | 10 +++++++-
 2 files changed, 33 insertions(+), 1 deletion(-)
 create mode 100644 grep-2.21-recurse-behaviour-change-doc.patch

diff --git a/grep-2.21-recurse-behaviour-change-doc.patch b/grep-2.21-recurse-behaviour-change-doc.patch
new file mode 100644
index 0000000..52a3550
--- /dev/null
+++ b/grep-2.21-recurse-behaviour-change-doc.patch
@@ -0,0 +1,24 @@
+diff --git a/doc/grep.in.1 b/doc/grep.in.1
+index 5a1e3ea..3f633ea 100644
+--- a/doc/grep.in.1
++++ b/doc/grep.in.1
+@@ -478,6 +478,7 @@ Search only files whose base name matches
+ .BR \-r ", " \-\^\-recursive
+ Read all files under each directory, recursively,
+ following symbolic links only if they are on the command line.
++Note that if no file operand is given, grep searches the working directory.
+ This is equivalent to the
+ .B "\-d recurse"
+ option.
+diff --git a/doc/grep.texi b/doc/grep.texi
+index da9a1be..63016bd 100644
+--- a/doc/grep.texi
++++ b/doc/grep.texi
+@@ -698,6 +698,7 @@ For each directory operand,
+ read and process all files in that directory, recursively.
+ Follow symbolic links on the command line, but skip symlinks
+ that are encountered recursively.
++Note that if no file operand is given, grep searches the working directory.
+ This is the same as the @samp{--directories=recurse} option.
+ 
+ @item -R
diff --git a/grep.spec b/grep.spec
index 26cad67..af8c0b8 100644
--- a/grep.spec
+++ b/grep.spec
@@ -3,7 +3,7 @@
 Summary: Pattern matching utilities
 Name: grep
 Version: 2.21
-Release: 2%{?dist}
+Release: 3%{?dist}
 License: GPLv3+
 Group: Applications/Text
 Source: ftp://ftp.gnu.org/pub/gnu/grep/grep-%{version}.tar.xz
@@ -16,6 +16,9 @@ Patch0: grep-2.21-man-fix-gs.patch
 Patch1: grep-2.21-help-align.patch
 # fix buffer overrun for grep -F, rhbz#1183653
 Patch2: grep-2.21-buf-overrun-fix.patch
+# backported from upstream
+# http://git.savannah.gnu.org/cgit/grep.git/commit/?id=c8b9364d5900a40809827aee6cc53705073278f6
+Patch3: grep-2.21-recurse-behaviour-change-doc.patch
 URL: http://www.gnu.org/software/grep/
 Requires(post): /sbin/install-info
 Requires(preun): /sbin/install-info
@@ -37,6 +40,7 @@ GNU grep is needed by many scripts, so it shall be installed on every system.
 %patch0 -p1 -b .man-fix-gs
 %patch1 -p1 -b .help-align
 %patch2 -p1 -b .buf-overrun-fix
+%patch3 -p1 -b .recurse-behaviour-change-doc
 
 chmod 755 tests/kwset-abuse
 
@@ -92,6 +96,10 @@ fi
 %{_mandir}/*/*
 
 %changelog
+* Tue Apr  7 2015 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21-3
+- Documented change in behaviour of recurse option
+  Resolves: rhbz#1178305
+
 * Tue Jan 20 2015 Jaroslav Škarvada <jskarvad@redhat.com> - 2.21-2
 - Fixed buffer overrun for grep -F
   Resolves: rhbz#1183653

From 401e77224387663f2dddfa6bd49bedc01476f348 Mon Sep 17 00:00:00 2001
From: Jaromir Capik <jcapik@redhat.com>
Date: Thu, 11 Jun 2015 15:51:11 +0200
Subject: [PATCH 8/8] Adding STAGE1 bootstrap recipe

---
 STAGE1-grep | 7 +++++++
 1 file changed, 7 insertions(+)
 create mode 100644 STAGE1-grep

diff --git a/STAGE1-grep b/STAGE1-grep
new file mode 100644
index 0000000..f80f9ca
--- /dev/null
+++ b/STAGE1-grep
@@ -0,0 +1,7 @@
+srpm $1
+mcd $BUILDDIR/$1
+$SRC/${1}-*/configure $TCONFIGARGS
+notparallel
+test -d tools/gnulib/lib && make $J V=1 -C tools/gnulib/lib
+make $J V=1
+make $J install DESTDIR=${ROOTFS}