bash-5.0-beta releasebash-5.0-beta

28 files changed, 655 insertions, 304 deletions

diff --git a/lib/malloc/malloc.c b/lib/malloc/malloc.c
index 88d777e5..5621adf4 100644
--- a/lib/malloc/malloc.c
+++ b/lib/malloc/malloc.c
@@ -84,6 +84,10 @@
 #include <errno.h>
 #include <stdio.h>
 
+#if defined (HAVE_MMAP)
+#include <sys/mman.h>
+#endif
+
 /* Define getpagesize () if the system does not.  */
 #ifndef HAVE_GETPAGESIZE
 #  include "getpagesize.h"
@@ -111,6 +115,7 @@
 #  define NO_VALLOC
 #endif
 
+/* SIZEOF_LONG * 4 - 2, usable bins from 1..NBUCKETS-1 */
 #define NBUCKETS	30
 
 #define ISALLOC ((char) 0xf7)	/* magic byte that implies allocation */
@@ -197,6 +202,23 @@ typedef union _malloc_guard {
 
 #define STARTBUCK	1
 
+/* Should we use mmap for large allocations? */
+#if defined (HAVE_MMAP)
+#  if !defined (MAP_ANON) && defined (MAP_ANONYMOUS)
+#    define MAP_ANON MAP_ANONYMOUS
+#  endif
+#endif
+
+#if defined (HAVE_MMAP) && defined (MAP_ANON)
+#  define USE_MMAP
+#endif
+
+#if defined (USE_MMAP)
+#  define MMAP_THRESHOLD	14	/* must be >= SPLIT_MAX, COMBINE_MAX */
+#else
+#  define MMAP_THRESHOLD	(8 * SIZEOF_LONG)
+#endif
+
 /* Flags for the internal functions. */
 #define MALLOC_WRAPPER	0x01	/* wrapper function */
 #define MALLOC_INTERNAL	0x02	/* internal function calling another */
@@ -288,6 +310,9 @@ int malloc_flags = 0;	/* future use */
 int malloc_trace = 0;	/* trace allocations and frees to stderr */
 int malloc_register = 0;	/* future use */
 
+/* Use a variable in case we want to dynamically adapt it in the future */
+int malloc_mmap_threshold = MMAP_THRESHOLD;
+
 #ifdef MALLOC_TRACE
 char _malloc_trace_buckets[NBUCKETS];
 
@@ -594,7 +619,7 @@ morecore (nu)
 
   /* Try to split a larger block here, if we're within the range of sizes
      to split. */
-  if (nu >= SPLIT_MIN)
+  if (nu >= SPLIT_MIN && nu <= malloc_mmap_threshold)
     {
       bsplit (nu);
       if (nextf[nu] != 0)
@@ -603,7 +628,7 @@ morecore (nu)
 
   /* Try to coalesce two adjacent blocks from the free list on nextf[nu - 1],
      if we can, and we're within the range of the block coalescing limits. */
-  if (nu >= COMBINE_MIN && nu < COMBINE_MAX && busy[nu - 1] == 0 && nextf[nu - 1])
+  if (nu >= COMBINE_MIN && nu < COMBINE_MAX && nu <= malloc_mmap_threshold && busy[nu - 1] == 0 && nextf[nu - 1])
     {
       bcoalesce (nu);
       if (nextf[nu] != 0)
@@ -632,6 +657,25 @@ morecore (nu)
       nblks = 1;
     }
 
+#if defined (USE_MMAP)
+  if (nu > malloc_mmap_threshold)
+    {
+      mp = (union mhead *)mmap (0, sbrk_amt, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
+      if ((void *)mp == MAP_FAILED)
+	goto morecore_done;
+      nextf[nu] = mp;
+      mp->mh_alloc = ISFREE;
+      mp->mh_index = nu;
+      CHAIN (mp) = 0;
+#ifdef MALLOC_STATS
+      _mstats.nmmap++;
+      _mstats.tmmap += sbrk_amt;
+#endif
+      goto morecore_done;
+    }
+#endif
+	
+
 #ifdef MALLOC_STATS
   _mstats.nsbrk++;
   _mstats.tsbrk += sbrk_amt;
@@ -915,6 +959,17 @@ internal_free (mem, file, line, flags)
   if (mg.i != p->mh_nbytes)
     xbotch (mem, ERR_ASSERT_FAILED, _("free: start and end chunk sizes differ"), file, line);
 
+#if defined (USE_MMAP)
+  if (nunits > malloc_mmap_threshold)
+    {
+      munmap (p, binsize (nunits));
+#if defined (MALLOC_STATS)
+      _mstats.nlesscore[nunits]++;
+#endif
+      goto free_return;
+    }
+#endif
+
 #if GLIBC21
   if (nunits >= LESSCORE_MIN && ((char *)p + binsize(nunits) == sbrk (0)))
 #else
@@ -1053,12 +1108,9 @@ internal_realloc (mem, n, file, line, flags)
   /* If ok, use the same block, just marking its size as changed.  */
   if (RIGHT_BUCKET(nbytes, nunits) || RIGHT_BUCKET(nbytes, nunits-1))
     {
-#if 0
-      m = (char *)mem + p->mh_nbytes;
-#else
       /* Compensate for increment above. */
       m -= 4;
-#endif
+
       *m++ = 0;  *m++ = 0;  *m++ = 0;  *m++ = 0;
       m = (char *)mem + (p->mh_nbytes = n);
 
@@ -1076,6 +1128,8 @@ internal_realloc (mem, n, file, line, flags)
   _mstats.nrcopy++;
 #endif
 
+  /* If we are using mmap and have mremap, we could use it here. */
+
   if ((m = internal_malloc (n, file, line, MALLOC_INTERNAL|MALLOC_NOTRACE|MALLOC_NOREG)) == 0)
     return 0;
   FASTCOPY (mem, m, tocopy);
@@ -1120,11 +1174,7 @@ internal_memalign (alignment, size, file, line, flags)
   if (((long) ptr & (alignment - 1)) == 0)
     return ptr;
   /* Otherwise, get address of byte in the block that has that alignment.  */
-#if 0
-  aligned = (char *) (((long) ptr + alignment - 1) & -alignment);
-#else
   aligned = (char *) (((long) ptr + alignment - 1) & (~alignment + 1));
-#endif
 
   /* Store a suitable indication of how to free the block,
      so that free can find the true beginning of it.  */
diff --git a/lib/malloc/mstats.h b/lib/malloc/mstats.h
index 11389ffc..dac02980 100644
--- a/lib/malloc/mstats.h
+++ b/lib/malloc/mstats.h
@@ -25,6 +25,7 @@
 
 #ifdef MALLOC_STATS
 
+/* This needs to change if the definition in malloc.c changes */
 #ifndef NBUCKETS
 #  define NBUCKETS 30
 #endif
@@ -77,6 +78,8 @@ struct _malstats {
   int nsplit[NBUCKETS];
   int tbcoalesce;
   int ncoalesce[NBUCKETS];
+  int nmmap;
+  bits32_t tmmap;
 };
 
 /* Return statistics describing allocation of blocks of size BLOCKSIZE.
@@ -98,6 +101,7 @@ struct bucket_stats {
   int nlesscore;
   int nsplit;
   int ncoalesce;
+  int nmmap;		/* currently unused */
 };
 
 extern struct bucket_stats malloc_bucket_stats __P((int));
diff --git a/lib/malloc/stats.c b/lib/malloc/stats.c
index ae555a1d..05247970 100644
--- a/lib/malloc/stats.c
+++ b/lib/malloc/stats.c
@@ -36,6 +36,8 @@
 
 extern int malloc_free_blocks __P((int));
 
+extern int malloc_mmap_threshold;
+
 extern struct _malstats _mstats;
 
 extern FILE *_imalloc_fopen __P((char *, char *, char *, char *, size_t));
@@ -103,8 +105,12 @@ _print_malloc_stats (s, fp)
   for (i = totused = totfree = 0; i < NBUCKETS; i++)
     {
       v = malloc_bucket_stats (i);
+      /* Show where the mmap threshold is; sizes greater than this use mmap to
+	 allocate and munmap to free (munmap shows up as lesscore). */
+      if (i == malloc_mmap_threshold+1)
+	fprintf (fp, "--------\n");
       if (v.nmal > 0)
-	fprintf (fp, "%8lu\t%4d\t%6d\t%5d\t%8d\t%d %5d %8d\n", (unsigned long)v.blocksize, v.nfree, v.nused, v.nmal, v.nmorecore, v.nlesscore, v.nsplit, v.ncoalesce);
+	fprintf (fp, "%8lu\t%4d\t%6d\t%5d\t%8d\t%8d %5d %8d\n", (unsigned long)v.blocksize, v.nfree, v.nused, v.nmal, v.nmorecore, v.nlesscore, v.nsplit, v.ncoalesce);
       totfree += v.nfree * v.blocksize;
       totused += v.nused * v.blocksize;
     }
@@ -115,6 +121,8 @@ _print_malloc_stats (s, fp)
 	   _mstats.nmal, _mstats.nfre, _mstats.nrealloc, _mstats.nrcopy);
   fprintf (fp, "Total sbrks: %d, total bytes via sbrk: %d\n",
   	   _mstats.nsbrk, _mstats.tsbrk);
+  fprintf (fp, "Total mmaps: %d, total bytes via mmap: %d\n",
+  	   _mstats.nmmap, _mstats.tmmap);
   fprintf (fp, "Total blocks split: %d, total block coalesces: %d\n",
   	   _mstats.tbsplit, _mstats.tbcoalesce);
 }
diff --git a/lib/readline/display.c b/lib/readline/display.c
index 75019efc..b2ed1154 100644
--- a/lib/readline/display.c
+++ b/lib/readline/display.c
@@ -2987,7 +2987,8 @@ _rl_update_final (void)
       putc (last_line[_rl_screenwidth - 1 + woff], rl_outstream);
     }
   _rl_vis_botlin = 0;
-  rl_crlf ();
+  if (botline_length > 0 || _rl_last_c_pos > 0)
+    rl_crlf ();
   fflush (rl_outstream);
   rl_display_fixed++;
 }
diff --git a/lib/readline/doc/hstech.texi b/lib/readline/doc/hstech.texi
index d28bd485..dbc21c14 100644
--- a/lib/readline/doc/hstech.texi
+++ b/lib/readline/doc/hstech.texi
@@ -403,7 +403,7 @@ to the ``normal'' terminating characters.
 Return an array of tokens parsed out of @var{string}, much as the
 shell might.  The tokens are split on the characters in the
 @var{history_word_delimiters} variable,
-and shell quoting conventions are obeyed.
+and shell quoting conventions are obeyed as described below.
 @end deftypefun
 
 @deftypefun {char *} history_arg_extract (int first, int last, const char *string)
@@ -476,8 +476,24 @@ carriage return, and @samp{=}.
 @end deftypevar
 
 @deftypevar int history_quotes_inhibit_expansion
-If non-zero, double-quoted words are not scanned for the history expansion
-character or the history comment character.  The default value is 0.
+If non-zero, the history expansion code implements shell-like quoting:
+single-quoted words are not scanned for the history expansion
+character or the history comment character, and double-quoted words may
+have history expansion performed, since single quotes are not special
+within double quotes.
+The default value is 0.
+@end deftypevar
+
+@deftypevar int history_quoting_state
+An application may set this variable to indicate that the current line
+being expanded is subject to existing quoting. If set to @samp{'}, the
+history expansion function will assume that the line is single-quoted and
+inhibit expansion until it reads an unquoted closing single quote; if set
+to @samp{"}, history expansion will assume the line is double quoted until
+it reads an unquoted closing double quote. If set to zero, the default,
+the history expansion function will assume the line is not quoted and
+treat quote characters within the line as described above.
+This is only effective if @var{history_quotes_inhibit_expansion} is set.
 @end deftypevar
 
 @deftypevar {rl_linebuf_func_t *} history_inhibit_expansion_function
diff --git a/lib/readline/doc/hsuser.texi b/lib/readline/doc/hsuser.texi
index d21f65d0..521ccc70 100644
--- a/lib/readline/doc/hsuser.texi
+++ b/lib/readline/doc/hsuser.texi
@@ -262,8 +262,8 @@ fix errors in previous commands quickly.
 @ifset BashFeatures
 History expansion is performed immediately after a complete line
 is read, before the shell breaks it into words, and is performed
-on each line individually without taking quoting on previous lines into
-account.
+on each line individually.  Bash attempts to inform the history
+expansion functions about quoting still in effect from previous lines.
 @end ifset
 
 History expansion takes place in two parts.  The first is to determine
@@ -277,9 +277,19 @@ that Bash does, so that several words
 surrounded by quotes are considered one word.
 History expansions are introduced by the appearance of the
 history expansion character, which is @samp{!} by default.
+
+History expansion implements shell-like quoting conventions:
+a backslash can be used to remove the special handling for the next character;
+single quotes enclose verbatim sequences of characters, and can be used to
+inhibit history expansion;
+and characters enclosed within double quotes may be subject to history
+expansion, since backslash can escape the history expansion character,
+but single quotes may not, since they are not treated specially within
+double quotes.
+
 @ifset BashFeatures
-Only @samp{\} and @samp{'} may be used to escape the history expansion
-character, but the history expansion character is
+When using the shell, only @samp{\} and @samp{'} may be used to escape the
+history expansion character, but the history expansion character is
 also treated as quoted if it immediately precedes the closing double quote
 in a double-quoted string. 
 @end ifset
diff --git a/lib/readline/doc/rltech.texi b/lib/readline/doc/rltech.texi
index 7d22b2cb..baf036df 100644
--- a/lib/readline/doc/rltech.texi
+++ b/lib/readline/doc/rltech.texi
@@ -846,12 +846,16 @@ Return the function invoked by @var{keyseq} in keymap @var{map}.
 If @var{map} is @code{NULL}, the current keymap is used.  If @var{type} is
 not @code{NULL}, the type of the object is returned in the @code{int} variable
 it points to (one of @code{ISFUNC}, @code{ISKMAP}, or @code{ISMACR}).
+It takes a "translated" key sequence and should not be used if the key sequence
+can include NUL.
 @end deftypefun
 
-@deftypefun {rl_command_func_t *} rl_function_of_keyseq_len (const char *keyseq, size_t len Keymap map, int *type)
+@deftypefun {rl_command_func_t *} rl_function_of_keyseq_len (const char *keyseq, size_t len, Keymap map, int *type)
 Return the function invoked by @var{keyseq} of length @var{len}
 in keymap @var{map}. Equivalent to @code{rl_function_of_keyseq} with the
 addition of the @var{len} parameter.
+It takes a "translated" key sequence and should be used if the key sequence
+can include NUL.
 @end deftypefun
 
 @deftypefun {char **} rl_invoking_keyseqs (rl_command_func_t *function)
diff --git a/lib/readline/doc/rluser.texi b/lib/readline/doc/rluser.texi
index 1c9acdcc..1acc8cda 100644
--- a/lib/readline/doc/rluser.texi
+++ b/lib/readline/doc/rluser.texi
@@ -475,6 +475,7 @@ The default value is @samp{off}.
 If set to @samp{on}, and @var{completion-ignore-case} is enabled, Readline
 treats hyphens (@samp{-}) and underscores (@samp{_}) as equivalent when
 performing case-insensitive filename matching and completion.
+The default value is @samp{off}.
 
 @item completion-prefix-display-length
 @vindex completion-prefix-display-length
@@ -802,7 +803,7 @@ Meta-Rubout: backward-kill-word
 Control-o: "> output"
 @end example
 
-In the above example, @kbd{C-u} is bound to the function
+In the example above, @kbd{C-u} is bound to the function
 @code{universal-argument},
 @kbd{M-DEL} is bound to the function @code{backward-kill-word}, and
 @kbd{C-o} is bound to run the macro
@@ -1352,7 +1353,7 @@ This function is intended to be bound to the "bracketed paste" escape
 sequence sent by some terminals, and such a binding is assigned by default.
 It allows Readline to insert the pasted text as a single unit without treating
 each character as if it had been read from the keyboard.  The characters
-are inserted as if each one was bound to @code{self-insert}) instead of
+are inserted as if each one was bound to @code{self-insert} instead of
 executing any editing commands.
 
 @item transpose-chars (C-t)
@@ -2019,10 +2020,10 @@ matches were generated.
 @item complete
 @btindex complete
 @example
-@code{complete [-abcdefgjksuv] [-o @var{comp-option}] [-DE] [-A @var{action}] [-G @var{globpat}] [-W @var{wordlist}]
-[-F @var{function}] [-C @var{command}] [-X @var{filterpat}]
+@code{complete [-abcdefgjksuv] [-o @var{comp-option}] [-DEI] [-A @var{action}] [-G @var{globpat}]
+[-W @var{wordlist}] [-F @var{function}] [-C @var{command}] [-X @var{filterpat}]
 [-P @var{prefix}] [-S @var{suffix}] @var{name} [@var{name} @dots{}]}
-@code{complete -pr [-DE] [@var{name} @dots{}]}
+@code{complete -pr [-DEI] [@var{name} @dots{}]}
 @end example
 
 Specify how arguments to each @var{name} should be completed.
@@ -2032,16 +2033,24 @@ reused as input.
 The @option{-r} option removes a completion specification for
 each @var{name}, or, if no @var{name}s are supplied, all
 completion specifications.
-The @option{-D} option indicates that the remaining options and actions should
+The @option{-D} option indicates that other supplied options and actions should
 apply to the ``default'' command completion; that is, completion attempted
 on a command for which no completion has previously been defined.
-The @option{-E} option indicates that the remaining options and actions should
+The @option{-E} option indicates that other supplied options and actions should
 apply to ``empty'' command completion; that is, completion attempted on a 
 blank line.
+The @option{-I} option indicates that other supplied options and actions should
+apply to completion on the inital non-assignment word on the line, or after a
+command delimiter such as @samp{;} or @samp{|}, which is usually command
+name completion.
+If multiple options are supplied, the @option{-D} option takes precedence
+over @option{-E}, and both take precedence over @option{-I}.
+If any of @option{-D}, @option{-E}, or @option{-I} are supplied, any other
+@var{name} arguments are ignored; these completions only apply to the case
+specified by the option.
 
 The process of applying these completion specifications when word completion
-is attempted is described above (@pxref{Programmable Completion}).  The
-@option{-D} option takes precedence over @option{-E}.
+is attempted is described above (@pxref{Programmable Completion}).
 
 Other options, if specified, have the following meanings.
 The arguments to the @option{-G}, @option{-W}, and @option{-X} options
@@ -2071,7 +2080,7 @@ Perform directory name completion if the compspec generates no matches.
 
 @item filenames
 Tell Readline that the compspec generates filenames, so it can perform any
-filename-specific processing (like adding a slash to directory names
+filename-specific processing (like adding a slash to directory names,
 quoting special characters, or suppressing trailing spaces).
 This option is intended to be used with shell functions specified
 with @option{-F}.
@@ -2226,7 +2235,7 @@ an error occurs adding a completion specification.
 @item compopt
 @btindex compopt
 @example
-@code{compopt} [-o @var{option}] [-DE] [+o @var{option}] [@var{name}]
+@code{compopt} [-o @var{option}] [-DEI] [+o @var{option}] [@var{name}]
 @end example
 Modify completion options for each @var{name} according to the
 @var{option}s, or for the currently-executing completion if no @var{name}s
@@ -2235,14 +2244,19 @@ If no @var{option}s are given, display the completion options for each
 @var{name} or the current completion.
 The possible values of @var{option} are those valid for the @code{complete}
 builtin described above.
-The @option{-D} option indicates that the remaining options should
+The @option{-D} option indicates that other supplied options should
 apply to the ``default'' command completion; that is, completion attempted
 on a command for which no completion has previously been defined.
-The @option{-E} option indicates that the remaining options should
+The @option{-E} option indicates that other supplied options should
 apply to ``empty'' command completion; that is, completion attempted on a 
 blank line.
+The @option{-I} option indicates that other supplied options should
+apply to completion on the inital non-assignment word on the line, or after a
+command delimiter such as @samp{;} or @samp{|}, which is usually command
+name completion.
 
-The @option{-D} option takes precedence over @option{-E}.
+If multiple options are supplied, the @option{-D} option takes precedence
+over @option{-E}, and both take precedence over @option{-I}
 
 The return value is true unless an invalid option is supplied, an attempt
 is made to modify the options for a @var{name} for which no completion
@@ -2289,7 +2303,7 @@ _comp_cd()
     local cur _skipdot _cdpath
     local i j k
 
-    # Tilde expansion, with side effect of expanding tilde to full pathname
+    # Tilde expansion, which also expands tilde to full pathname
     case "$2" in
     \~*)    eval cur="$2" ;;
     *)      cur=$2 ;;
diff --git a/lib/readline/doc/version.texi b/lib/readline/doc/version.texi
index 9fefcfae..79f41917 100644
--- a/lib/readline/doc/version.texi
+++ b/lib/readline/doc/version.texi
@@ -1,10 +1,10 @@
 @ignore
-Copyright (C) 1988-2017 Free Software Foundation, Inc. 
+Copyright (C) 1988-2018 Free Software Foundation, Inc. 
 @end ignore
 
-@set EDITION 7.0
-@set VERSION 7.0
-@set UPDATED 28 December 2017
-@set UPDATED-MONTH December 2017
+@set EDITION 8.0
+@set VERSION 8.0
+@set UPDATED 6 July 2018
+@set UPDATED-MONTH July 2018
 
-@set LASTCHANGE Thu Dec 28 14:44:16 EST 2017
+@set LASTCHANGE Fri Jul  6 16:25:22 MDT 2018
diff --git a/lib/readline/histexpand.c b/lib/readline/histexpand.c
index e9c333d7..92b996f9 100644
--- a/lib/readline/histexpand.c
+++ b/lib/readline/histexpand.c
@@ -1,6 +1,6 @@
 /* histexpand.c -- history expansion. */
 
-/* Copyright (C) 1989-2017 Free Software Foundation, Inc.
+/* Copyright (C) 1989-2018 Free Software Foundation, Inc.
 
    This file contains the GNU History Library (History), a set of
    routines for managing the text of previously typed lines.
@@ -107,6 +107,8 @@ char *history_word_delimiters = HISTORY_WORD_DELIMITERS;
    particular history expansion should be performed. */
 rl_linebuf_func_t *history_inhibit_expansion_function;
 
+int history_quoting_state = 0;
+
 /* **************************************************************** */
 /*								    */
 /*			History Expansion			    */
@@ -961,7 +963,22 @@ history_expand (char *hstring, char **output)
 
       /* `!' followed by one of the characters in history_no_expand_chars
 	 is NOT an expansion. */
-      for (i = dquote = squote = 0; string[i]; i++)
+      dquote = history_quoting_state == '"';
+      squote = history_quoting_state == '\'';
+
+      /* If the calling application tells us we are already reading a
+	 single-quoted string, consume the rest of the string right now
+	 and then go on. */
+      i = 0;
+      if (squote && history_quotes_inhibit_expansion)
+	{
+	  hist_string_extract_single_quoted (string, &i, 0);
+	  squote = 0;
+	  if (string[i])
+	    i++;
+	}
+
+      for ( ; string[i]; i++)
 	{
 #if defined (HANDLE_MULTIBYTE)
 	  if (MB_CUR_MAX > 1 && rl_byte_oriented == 0)
@@ -1049,7 +1066,29 @@ history_expand (char *hstring, char **output)
     }
 
   /* Extract and perform the substitution. */
-  for (passc = dquote = squote = i = j = 0; i < l; i++)
+  dquote = history_quoting_state == '"';
+  squote = history_quoting_state == '\'';
+
+  /* If the calling application tells us we are already reading a
+     single-quoted string, consume the rest of the string right now
+     and then go on. */
+  i = j = 0;
+  if (squote && history_quotes_inhibit_expansion)
+    {
+      int c;
+
+      hist_string_extract_single_quoted (string, &i, 0);
+      squote = 0;
+      for (c = 0; c < i; c++)
+	ADD_CHAR (string[c]);      
+      if (string[i])
+	{
+	  ADD_CHAR (string[i]);
+	  i++;
+	}
+    }
+
+  for (passc = 0; i < l; i++)
     {
       int qc, tchar = string[i];
 
@@ -1429,11 +1468,11 @@ history_tokenize_word (const char *string, int ind)
 	}
     }
 
-  if (member (string[i], "<>;&|$"))
+  if (member (string[i], "<>;&|"))
     {
       int peek = string[i + 1];
 
-      if (peek == string[i] && peek != '$')
+      if (peek == string[i])
 	{
 	  if (peek == '<' && string[i + 2] == '-')
 	    i++;
@@ -1456,9 +1495,8 @@ history_tokenize_word (const char *string, int ind)
 	  i += 2;
 	  return i;
 	}
-      /* XXX - separated out for later -- bash-4.2 */
-      else if ((peek == '(' && (string[i] == '>' || string[i] == '<')) || /* ) */
-	       (peek == '(' && string[i] == '$')) /*)*/
+      /* XXX - process substitution -- separated out for later -- bash-4.2 */
+      else if (peek == '(' && (string[i] == '>' || string[i] == '<')) /*)*/
 	{
 	  i += 2;
 	  delimopen = '(';
@@ -1466,34 +1504,9 @@ history_tokenize_word (const char *string, int ind)
 	  nestdelim = 1;
 	  goto get_word;
 	}
-#if 0
-      else if (peek == '\'' && string[i] == '$')
-        {
-	  i += 2;	/* XXX */
-	  return i;
-        }
-#endif
-
-      if (string[i] != '$')
-	{
-	  i++;
-	  return i;
-	}
-    }
-
-  /* same code also used for $(...)/<(...)/>(...) above */
-  if (member (string[i], "!@?+*"))
-    {
-      int peek = string[i + 1];
 
-      if (peek == '(')		/*)*/
-	{
-	  /* Shell extended globbing patterns */
-	  i += 2;
-	  delimopen = '(';
-	  delimiter = ')';	/* XXX - not perfect */
-	  nestdelim = 1;
-	}
+      i++;
+      return i;
     }
 
 get_word:
@@ -1538,6 +1551,16 @@ get_word:
 	  continue;
 	}
 
+      /* Command and process substitution; shell extended globbing patterns */
+      if (nestdelim == 0 && delimiter == 0 && member (string[i], "<>$!@?+*") && string[i+1] == '(') /*)*/
+	{
+	  i += 2;
+	  delimopen = '(';
+	  delimiter = ')';
+	  nestdelim = 1;
+	  continue;
+	}
+      
       if (delimiter == 0 && (member (string[i], history_word_delimiters)))
 	break;
 
diff --git a/lib/readline/histfile.c b/lib/readline/histfile.c
index 399bcc39..dc64bde1 100644
--- a/lib/readline/histfile.c
+++ b/lib/readline/histfile.c
@@ -1,6 +1,6 @@
 /* histfile.c - functions to manipulate the history file. */
 
-/* Copyright (C) 1989-2017 Free Software Foundation, Inc.
+/* Copyright (C) 1989-2018 Free Software Foundation, Inc.
 
    This file contains the GNU History Library (History), a set of
    routines for managing the text of previously typed lines.
@@ -396,7 +396,7 @@ read_history_range (const char *filename, int from, int to)
 	  {
 	    if (HIST_TIMESTAMP_START(line_start) == 0)
 	      {
-	      	if (last_ts == NULL && history_multiline_entries)
+	      	if (last_ts == NULL && history_length > 0 && history_multiline_entries)
 		  _hs_append_history_line (history_length - 1, line_start);
 		else
 		  add_history (line_start);
diff --git a/lib/readline/history.h b/lib/readline/history.h
index fceb99ca..cc3de29a 100644
--- a/lib/readline/history.h
+++ b/lib/readline/history.h
@@ -261,7 +261,9 @@ extern char *history_word_delimiters;
 extern char history_comment_char;
 extern char *history_no_expand_chars;
 extern char *history_search_delimiter_chars;
+
 extern int history_quotes_inhibit_expansion;
+extern int history_quoting_state;
 
 extern int history_write_timestamps;
 
diff --git a/lib/readline/kill.c b/lib/readline/kill.c
index 983a9848..9fdfc772 100644
--- a/lib/readline/kill.c
+++ b/lib/readline/kill.c
@@ -322,7 +322,7 @@ rl_unix_word_rubout (int count, int key)
 	    rl_point--;
 
 	  while (rl_point && (whitespace (rl_line_buffer[rl_point - 1]) == 0))
-	    rl_point--;
+	    rl_point--;		/* XXX - multibyte? */
 	}
 
       rl_kill_text (orig_point, rl_point);
@@ -359,7 +359,7 @@ rl_unix_filename_rubout (int count, int key)
 
 	  while (rl_point && (whitespace (c) == 0) && c != '/')
 	    {
-	      rl_point--;
+	      rl_point--;	/* XXX - multibyte? */
 	      c = rl_line_buffer[rl_point - 1];
 	    }
 	}
@@ -677,7 +677,7 @@ rl_bracketed_paste_begin (int count, int key)
   size_t len, cap;
   char *buf;
 
-  retval = 1;
+  retval = 0;
   len = 0;
   buf = xmalloc (cap = 64);
 
@@ -708,7 +708,7 @@ rl_bracketed_paste_begin (int count, int key)
       if (len == cap)
 	buf = xrealloc (buf, cap + 1);
       buf[len] = '\0';
-      retval = rl_insert_text (buf);
+      retval = rl_insert_text (buf) == len ? 0 : 1;
     }
 
   xfree (buf);
diff --git a/lib/readline/mbutil.c b/lib/readline/mbutil.c
index 65543416..17716357 100644
--- a/lib/readline/mbutil.c
+++ b/lib/readline/mbutil.c
@@ -75,6 +75,57 @@ int _rl_utf8locale = 0;
 
 #if defined(HANDLE_MULTIBYTE)
 
+/* **************************************************************** */
+/*								    */
+/*		UTF-8 specific Character Utility Functions	    */
+/*								    */
+/* **************************************************************** */
+
+/* Return the length in bytes of the possibly-multibyte character beginning
+   at S. Encoding is UTF-8. */
+static int
+_rl_utf8_mblen (const char *s, size_t n)
+{
+  unsigned char c, c1;
+
+  if (s == 0)
+    return (0);	/* no shift states */
+  if (n <= 0)
+    return (-1);
+
+  c = (unsigned char)*s;
+  if (c < 0x80)
+    return (c != 0);
+  if (c >= 0xc2)
+    {
+      c1 = (unsigned char)s[1];
+      if (c < 0xe0)
+	{
+	  if (n >= 2 && (s[1] ^ 0x80) < 0x40)
+	    return 2;
+	}
+      else if (c < 0xf0)
+	{
+	  if (n >= 3
+		&& (s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+		&& (c >= 0xe1 || c1 >= 0xa0)
+		&& (c != 0xed || c1 < 0xa0))
+	    return 3;
+	}
+      else if (c < 0xf8)
+	{
+	  if (n >= 4
+		&& (s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+	 	&& (s[3] ^ 0x80) < 0x40
+		&& (c >= 0xf1 || c1 >= 0x90)
+		&& (c < 0xf4 || (c == 0xf4 && c1 < 0x90)))
+	    return 4;
+	}
+    }
+  /* invalid or incomplete multibyte character */
+  return -1;
+}
+
 static int
 _rl_find_next_mbchar_internal (char *string, int seed, int count, int find_non_zero)
 {
@@ -228,11 +279,16 @@ _rl_get_char_len (char *src, mbstate_t *ps)
 
   /* Look at no more than MB_CUR_MAX characters */
   l = (size_t)strlen (src);
-  mb_cur_max = MB_CUR_MAX;
-  tmp = mbrlen((const char *)src, (l < mb_cur_max) ? l : mb_cur_max, ps);
+  if (_rl_utf8locale && l > 0 && UTF8_SINGLEBYTE(*src))
+    tmp = (*src != 0) ? 1 : 0;
+  else
+    {
+      mb_cur_max = MB_CUR_MAX;
+      tmp = mbrlen((const char *)src, (l < mb_cur_max) ? l : mb_cur_max, ps);
+    }
   if (tmp == (size_t)(-2))
     {
-      /* shorted to compose multibyte char */
+      /* too short to compose multibyte char */
       if (ps)
 	memset (ps, 0, sizeof(mbstate_t));
       return -2;
diff --git a/lib/readline/readline.c b/lib/readline/readline.c
index 4842a02e..678f989c 100644
--- a/lib/readline/readline.c
+++ b/lib/readline/readline.c
@@ -214,6 +214,9 @@ int _rl_eof_char = CTRL ('D');
 /* Non-zero makes this the next keystroke to read. */
 int rl_pending_input = 0;
 
+/* If non-zero when readline_internal returns, it means we found EOF */
+int _rl_eof_found = 0;
+
 /* Pointer to a useful terminal name. */
 const char *rl_terminal_name = (const char *)NULL;
 
@@ -222,7 +225,7 @@ int _rl_horizontal_scroll_mode = 0;
 
 /* Non-zero means to display an asterisk at the starts of history lines
    which have been modified. */
-int _rl_mark_modified_lines = 0;  
+int _rl_mark_modified_lines = 0;
 
 /* The style of `bell' notification preferred.  This can be set to NO_BELL,
    AUDIBLE_BELL, or VISIBLE_BELL. */
@@ -664,11 +667,9 @@ readline_internal_charloop (void)
 static char *
 readline_internal (void)
 {
-  int eof;
-
   readline_internal_setup ();
-  eof = readline_internal_charloop ();
-  return (readline_internal_teardown (eof));
+  _rl_eof_found = readline_internal_charloop ();
+  return (readline_internal_teardown (_rl_eof_found));
 }
 
 void
@@ -1055,7 +1056,7 @@ _rl_subseq_result (int r, Keymap map, int key, int got_subseq)
 	/* We probably shadowed a keymap, so keep going. */
 	r = _rl_dispatch (ANYOTHERKEY, m);
     }
-  else if (r && map[ANYOTHERKEY].function)
+  else if (r < 0 && map[ANYOTHERKEY].function)
     {
       /* We didn't match (r is probably -1), so return something to
 	 tell the caller that it should try ANYOTHERKEY for an
diff --git a/lib/readline/readline.h b/lib/readline/readline.h
index 2081b197..4e08b2e9 100644
--- a/lib/readline/readline.h
+++ b/lib/readline/readline.h
@@ -39,8 +39,8 @@ extern "C" {
 #endif
 
 /* Hex-encoded Readline version number. */
-#define RL_READLINE_VERSION	0x0700		/* Readline 7.0 */
-#define RL_VERSION_MAJOR	7
+#define RL_READLINE_VERSION	0x0800		/* Readline 8.0 */
+#define RL_VERSION_MAJOR	8
 #define RL_VERSION_MINOR	0
 
 /* Readline data structures. */
diff --git a/lib/readline/rlmbutil.h b/lib/readline/rlmbutil.h
index 8b88de40..6fd03f6a 100644
--- a/lib/readline/rlmbutil.h
+++ b/lib/readline/rlmbutil.h
@@ -173,6 +173,8 @@ _rl_wcwidth (wc)
 #endif
 
 #define UTF8_SINGLEBYTE(c)	(((c) & 0x80) == 0)
+#define UTF8_MBFIRSTCHAR(c)	(((c) & 0xc0) == 0xc0)
+#define UTF8_MBCHAR(c)		(((c) & 0xc0) == 0x80)
 
 #else /* !HANDLE_MULTIBYTE */
 
diff --git a/lib/readline/rlprivate.h b/lib/readline/rlprivate.h
index 05ee9302..69943de5 100644
--- a/lib/readline/rlprivate.h
+++ b/lib/readline/rlprivate.h
@@ -382,6 +382,7 @@ extern void _rl_set_cursor PARAMS((int, int));
 extern void _rl_fix_point PARAMS((int));
 extern int _rl_replace_text PARAMS((const char *, int, int));
 extern int _rl_forward_char_internal PARAMS((int));
+extern int _rl_backward_char_internal PARAMS((int));
 extern int _rl_insert_char PARAMS((int, int));
 extern int _rl_overwrite_char PARAMS((int, int));
 extern int _rl_overwrite_rubout PARAMS((int, int));
@@ -511,6 +512,7 @@ extern FILE *_rl_in_stream;
 extern FILE *_rl_out_stream;
 extern int _rl_last_command_was_kill;
 extern int _rl_eof_char;
+extern int _rl_eof_found;
 extern procenv_t _rl_top_level;
 extern _rl_keyseq_cxt *_rl_kscxt;
 extern int _rl_keyseq_timeout;
diff --git a/lib/readline/rltty.c b/lib/readline/rltty.c
index 6d907192..2ee36a7c 100644
--- a/lib/readline/rltty.c
+++ b/lib/readline/rltty.c
@@ -691,7 +691,11 @@ rl_deprep_terminal (void)
   tty = rl_instream ? fileno (rl_instream) : fileno (stdin);
 
   if (terminal_prepped & TPX_BRACKPASTE)
-    fprintf (rl_outstream, BRACK_PASTE_FINI);
+    {
+      fprintf (rl_outstream, BRACK_PASTE_FINI);
+      if (_rl_eof_found)
+ 	fprintf (rl_outstream, "\n");
+    }
 
   if (_rl_enable_keypad)
     _rl_control_keypad (0);
diff --git a/lib/readline/text.c b/lib/readline/text.c
index 81de2a99..cddaeebd 100644
--- a/lib/readline/text.c
+++ b/lib/readline/text.c
@@ -297,13 +297,40 @@ _rl_forward_char_internal (int count)
 #endif
 
     if (rl_end < 0)
-	rl_end = 0;
+      rl_end = 0;
 #else
   point = rl_point + count;
+#endif
+
   if (point > rl_end)
     point = rl_end;
+  return (point);
+}
+
+int
+_rl_backward_char_internal (int count)
+{
+  int point;
+
+  point = rl_point;
+#if defined (HANDLE_MULTIBYTE)
+  if (count > 0)
+    {
+      while (count > 0 && point > 0)
+	{
+	  point = _rl_find_prev_mbchar (rl_line_buffer, point, MB_FIND_NONZERO);
+	  count--;
+	}
+      if (count > 0)
+        return 0;	/* XXX - rl_ding() here? */
+    }
+#else
+  if (count > 0)
+    point -= count;
 #endif
 
+  if (point < 0)
+    point = 0;
   return (point);
 }
 
diff --git a/lib/readline/vi_mode.c b/lib/readline/vi_mode.c
index 3cb7e8c9..d6fa38e9 100644
--- a/lib/readline/vi_mode.c
+++ b/lib/readline/vi_mode.c
@@ -1,7 +1,7 @@
 /* vi_mode.c -- A vi emulation mode for Bash.
    Derived from code written by Jeff Sparkes (jsparkes@bnr.ca).  */
 
-/* Copyright (C) 1987-2017 Free Software Foundation, Inc.
+/* Copyright (C) 1987-2018 Free Software Foundation, Inc.
 
    This file is part of the GNU Readline Library (Readline), a library
    for reading lines of text with interactive input and history editing.      
@@ -63,6 +63,19 @@
 #define member(c, s) ((c) ? (char *)strchr ((s), (c)) != (char *)NULL : 0)
 #endif
 
+/* Increment START to the next character in RL_LINE_BUFFER, handling multibyte chars */
+#if defined (HANDLE_MULTIBYTE)
+#define INCREMENT_POS(start) \
+  do { \
+	if (MB_CUR_MAX == 1 || rl_byte_oriented) \
+	  start++; \
+	else \
+	  start = _rl_find_next_mbchar (rl_line_buffer, start, 1, MB_FIND_ANY); \
+  } while (0)
+#else /* !HANDLE_MULTIBYTE */
+#define INCREMENT_POS(start)    (start)++
+#endif /* !HANDLE_MULTIBYTE */
+
 /* This is global so other parts of the code can check whether the last
    command was a text modification command. */
 int _rl_vi_last_command = 'i';	/* default `.' puts you in insert mode */
@@ -118,7 +131,7 @@ static void _rl_vi_save_insert PARAMS((UNDO_LIST *));
 
 static void vi_save_insert_buffer PARAMS ((int, int));
 
-static void _rl_vi_backup PARAMS((void));
+static inline void _rl_vi_backup PARAMS((void));
 
 static int _rl_vi_arg_dispatch PARAMS((int));
 static int rl_digit_loop1 PARAMS((void));
@@ -126,6 +139,9 @@ static int rl_digit_loop1 PARAMS((void));
 static int _rl_vi_set_mark PARAMS((void));
 static int _rl_vi_goto_mark PARAMS((void));
 
+static inline int _rl_vi_advance_point PARAMS((void));
+static inline int _rl_vi_backup_point PARAMS((void));
+
 static void _rl_vi_append_forward PARAMS((int));
 
 static int _rl_vi_callback_getchar PARAMS((char *, int));
@@ -383,7 +399,7 @@ rl_vi_complete (int ignore, int key)
     {
       if (!whitespace (rl_line_buffer[rl_point + 1]))
 	rl_vi_end_word (1, 'E');
-      rl_point++;
+      _rl_vi_advance_point ();
     }
 
   if (key == '*')
@@ -451,6 +467,69 @@ rl_vi_next_word (int count, int key)
   return (0);
 }
 
+static inline int
+_rl_vi_advance_point (void)
+{
+  int point;
+
+  point = rl_point;
+  if (rl_point < rl_end)
+#if defined (HANDLE_MULTIBYTE)
+    {
+      if (MB_CUR_MAX == 1 || rl_byte_oriented)
+	rl_point++;
+      else
+	{
+	  point = rl_point;
+	  rl_point = _rl_forward_char_internal (1);
+	  if (point == rl_point || rl_point > rl_end)
+	    rl_point = rl_end;
+	}
+    }
+#else
+    rl_point++;
+#endif
+
+  return point;
+}
+
+/* Move the cursor back one character. */
+static inline void
+_rl_vi_backup (void)
+{
+  if (MB_CUR_MAX > 1 && rl_byte_oriented == 0)
+    rl_point = _rl_find_prev_mbchar (rl_line_buffer, rl_point, MB_FIND_NONZERO);
+  else
+    rl_point--;
+}
+
+/* Move the point back one character, returning the starting value and not
+   doing anything at the beginning of the line */
+static inline int
+_rl_vi_backup_point (void)
+{
+  int point;
+
+  point = rl_point;
+  if (rl_point > 0)
+#if defined (HANDLE_MULTIBYTE)
+    {
+      if (MB_CUR_MAX == 1 || rl_byte_oriented)
+	rl_point--;
+      else
+	{
+	  point = rl_point;
+	  rl_point = _rl_backward_char_internal (1);
+	  if (rl_point < 0)
+	    rl_point = 0;		/* XXX - not really necessary */
+	}
+    }
+#else
+    rl_point--;
+#endif
+  return point;
+}
+
 /* Move to the end of the ?next? word. */
 int
 rl_vi_end_word (int count, int key)
@@ -476,11 +555,11 @@ rl_vi_fWord (int count, int ignore)
     {
       /* Skip until whitespace. */
       while (!whitespace (rl_line_buffer[rl_point]) && rl_point < rl_end)
-	rl_point++;
+	_rl_vi_advance_point ();
 
       /* Now skip whitespace. */
       while (whitespace (rl_line_buffer[rl_point]) && rl_point < rl_end)
-	rl_point++;
+	_rl_vi_advance_point ();
     }
   return (0);
 }
@@ -497,12 +576,18 @@ rl_vi_bWord (int count, int ignore)
 	rl_point--;
 
       while (rl_point > 0 && whitespace (rl_line_buffer[rl_point]))
-	rl_point--;
+	_rl_vi_backup_point ();
 
       if (rl_point > 0)
 	{
-	  while (--rl_point >= 0 && !whitespace (rl_line_buffer[rl_point]));
-	  rl_point++;
+	  do
+	    _rl_vi_backup_point ();
+	  while (rl_point > 0 && !whitespace (rl_line_buffer[rl_point]));
+	  if (rl_point > 0)	/* hit whitespace */
+	    rl_point++;		
+
+	  if (rl_point < 0)
+	    rl_point = 0;
 	}
     }
   return (0);
@@ -511,28 +596,32 @@ rl_vi_bWord (int count, int ignore)
 int
 rl_vi_eWord (int count, int ignore)
 {
+  int opoint;
+
   while (count-- && rl_point < (rl_end - 1))
     {
-      if (!whitespace (rl_line_buffer[rl_point]))
-	rl_point++;
+      if (whitespace (rl_line_buffer[rl_point]) == 0)
+	_rl_vi_advance_point ();
 
       /* Move to the next non-whitespace character (to the start of the
 	 next word). */
       while (rl_point < rl_end && whitespace (rl_line_buffer[rl_point]))
-	rl_point++;
+	_rl_vi_advance_point ();
 
       if (rl_point && rl_point < rl_end)
 	{
+	  opoint = rl_point;
+
 	  /* Skip whitespace. */
 	  while (rl_point < rl_end && whitespace (rl_line_buffer[rl_point]))
-	    rl_point++;
+	    opoint = _rl_vi_advance_point ();	/* XXX - why? */
 
 	  /* Skip until whitespace. */
 	  while (rl_point < rl_end && !whitespace (rl_line_buffer[rl_point]))
-	    rl_point++;
+	    opoint = _rl_vi_advance_point ();
 
 	  /* Move back to the last character of the word. */
-	  rl_point--;
+	  rl_point = opoint;
 	}
     }
   return (0);
@@ -541,24 +630,28 @@ rl_vi_eWord (int count, int ignore)
 int
 rl_vi_fword (int count, int ignore)
 {
+  int opoint;
+
   while (count-- && rl_point < (rl_end - 1))
     {
       /* Move to white space (really non-identifer). */
       if (_rl_isident (rl_line_buffer[rl_point]))
 	{
 	  while (_rl_isident (rl_line_buffer[rl_point]) && rl_point < rl_end)
-	    rl_point++;
+	    _rl_vi_advance_point ();
 	}
       else /* if (!whitespace (rl_line_buffer[rl_point])) */
 	{
 	  while (!_rl_isident (rl_line_buffer[rl_point]) &&
 		 !whitespace (rl_line_buffer[rl_point]) && rl_point < rl_end)
-	    rl_point++;
+	    _rl_vi_advance_point ();
 	}
 
+      opoint = rl_point;
+
       /* Move past whitespace. */
       while (whitespace (rl_line_buffer[rl_point]) && rl_point < rl_end)
-	rl_point++;
+	opoint = _rl_vi_advance_point ();
     }
   return (0);
 }
@@ -566,9 +659,11 @@ rl_vi_fword (int count, int ignore)
 int
 rl_vi_bword (int count, int ignore)
 {
+  int opoint;
+
   while (count-- && rl_point > 0)
     {
-      int last_is_ident;
+      int prev_is_ident, cur_is_ident;
 
       /* If we are at the start of a word, move back to whitespace
 	 so we will go back to the start of the previous word. */
@@ -581,22 +676,35 @@ rl_vi_bword (int count, int ignore)
 	 back so we don't get messed up by the rl_point++ down there in
 	 the while loop.  Without this code, words like `l;' screw up the
 	 function. */
-      last_is_ident = _rl_isident (rl_line_buffer[rl_point - 1]);
-      if ((_rl_isident (rl_line_buffer[rl_point]) && !last_is_ident) ||
-	  (!_rl_isident (rl_line_buffer[rl_point]) && last_is_ident))
-	rl_point--;
+      cur_is_ident = _rl_isident (rl_line_buffer[rl_point]);
+      opoint = _rl_vi_backup_point ();
+      prev_is_ident = _rl_isident (rl_line_buffer[rl_point]);
+      if ((cur_is_ident && !prev_is_ident) || (!cur_is_ident && prev_is_ident))
+	;	/* leave point alone, we backed it up one character */
+      else
+	rl_point = opoint;
 
       while (rl_point > 0 && whitespace (rl_line_buffer[rl_point]))
-	rl_point--;
+	_rl_vi_backup_point ();
 
       if (rl_point > 0)
 	{
+	  opoint = rl_point;
 	  if (_rl_isident (rl_line_buffer[rl_point]))
-	    while (--rl_point >= 0 && _rl_isident (rl_line_buffer[rl_point]));
+	    do
+	      opoint = _rl_vi_backup_point ();
+	    while (rl_point > 0 && _rl_isident (rl_line_buffer[rl_point]));
 	  else
-	    while (--rl_point >= 0 && !_rl_isident (rl_line_buffer[rl_point]) &&
+	    do
+	      opoint = _rl_vi_backup_point ();
+	    while (rl_point > 0 && !_rl_isident (rl_line_buffer[rl_point]) &&
 		   !whitespace (rl_line_buffer[rl_point]));
-	  rl_point++;
+
+	  if (rl_point > 0)
+	    rl_point = opoint;
+
+	  if (rl_point < 0)
+	    rl_point = 0;
 	}
     }
   return (0);
@@ -605,23 +713,34 @@ rl_vi_bword (int count, int ignore)
 int
 rl_vi_eword (int count, int ignore)
 {
-  while (count-- && rl_point < rl_end - 1)
+  int opoint;
+
+  while (count-- && rl_point < (rl_end - 1))
     {
-      if (!whitespace (rl_line_buffer[rl_point]))
-	rl_point++;
+      if (whitespace (rl_line_buffer[rl_point]) == 0)
+	_rl_vi_advance_point ();
 
       while (rl_point < rl_end && whitespace (rl_line_buffer[rl_point]))
-	rl_point++;
+	_rl_vi_advance_point ();
 
+      opoint = rl_point;
       if (rl_point < rl_end)
 	{
 	  if (_rl_isident (rl_line_buffer[rl_point]))
-	    while (++rl_point < rl_end && _rl_isident (rl_line_buffer[rl_point]));
+	    do
+	      {
+		opoint = _rl_vi_advance_point ();
+	      }
+	    while (rl_point < rl_end && _rl_isident (rl_line_buffer[rl_point]));
 	  else
-	    while (++rl_point < rl_end && !_rl_isident (rl_line_buffer[rl_point])
+	    do
+	      {
+		opoint = _rl_vi_advance_point ();
+	      }
+	    while (rl_point < rl_end && !_rl_isident (rl_line_buffer[rl_point])
 		   && !whitespace (rl_line_buffer[rl_point]));
 	}
-      rl_point--;
+      rl_point = opoint;
     }
   return (0);
 }
@@ -637,20 +756,7 @@ rl_vi_insert_beg (int count, int key)
 static void
 _rl_vi_append_forward (int key)
 {
-  int point;
-
-  if (rl_point < rl_end)
-    {
-      if (MB_CUR_MAX == 1 || rl_byte_oriented)
-	rl_point++;
-      else
-	{
-	  point = rl_point;
-	  rl_point = _rl_forward_char_internal (1);
-	  if (point == rl_point)
-	    rl_point = rl_end;
-	}
-    }
+  _rl_vi_advance_point ();
 }
 
 int
@@ -859,7 +965,7 @@ _rl_vi_change_mbchar_case (int count)
 	  rl_begin_undo_group ();
 	  rl_vi_delete (1, 0);
 	  if (rl_point < p)	/* Did we retreat at EOL? */
-	    rl_point++;	/* XXX - should we advance more than 1 for mbchar? */
+	    _rl_vi_advance_point ();
 	  rl_insert_text (mb);
 	  rl_end_undo_group ();
 	  rl_vi_check ();
@@ -931,27 +1037,12 @@ rl_vi_put (int count, int key)
   return (0);
 }
 
-/* Move the cursor back one character. */
-static void
-_rl_vi_backup (void)
-{
-  if (MB_CUR_MAX > 1 && rl_byte_oriented == 0)
-    rl_point = _rl_find_prev_mbchar (rl_line_buffer, rl_point, MB_FIND_NONZERO);
-  else
-    rl_point--;
-}
-
 /* Move the cursor back one character if you're at the end of the line */
 int
 rl_vi_check (void)
 {
   if (rl_point && rl_point == rl_end)
-    {
-      if (MB_CUR_MAX > 1 && rl_byte_oriented == 0)
-	rl_point = _rl_find_prev_mbchar (rl_line_buffer, rl_point, MB_FIND_NONZERO);
-      else
-	rl_point--;
-    }
+    _rl_vi_backup ();
   return (0);
 }
 
@@ -962,7 +1053,10 @@ rl_vi_column (int count, int key)
   if (count > rl_end)
     rl_end_of_line (1, key);
   else
-    rl_point = count - 1;
+    {
+      rl_point = 0;
+      rl_point = _rl_forward_char_internal (count - 1);
+    }
   return (0);
 }
 
@@ -1113,7 +1207,7 @@ _rl_vi_domove_motion_cleanup (int c, _rl_vimotion_cxt *m)
      non-whitespace character, move back one (presumably to whitespace). */
   if ((_rl_to_upper (c) == 'W') && rl_point < rl_end && rl_point > rl_mark &&
       !whitespace (rl_line_buffer[rl_point]))
-    rl_point--;
+    rl_point--;		/* XXX */
 
   /* If cw or cW, back up to the end of a word, so the behaviour of ce
      or cE is the actual result.  Brute-force, no subtlety. */
@@ -1126,14 +1220,14 @@ _rl_vi_domove_motion_cleanup (int c, _rl_vimotion_cxt *m)
       /* Posix.2 says that if cw or cW moves the cursor towards the end of
 	 the line, the character under the cursor should be deleted. */
       if (rl_point == rl_mark)
-	rl_point++;
+	_rl_vi_advance_point ();
       else
 	{
 	  /* Move past the end of the word so that the kill doesn't
 	     remove the last letter of the previous word.  Only do this
 	     if we are not at the end of the line. */
 	  if (rl_point >= 0 && rl_point < (rl_end - 1) && !whitespace (rl_line_buffer[rl_point]))
-	    rl_point++;
+	    _rl_vi_advance_point ();
 	}
     }
 
@@ -1271,7 +1365,7 @@ vi_delete_dispatch (_rl_vimotion_cxt *m)
      mark. */
   if (((strchr (" l|h^0bBFT`", m->motion) == 0) && (rl_point >= m->start)) &&
       (rl_mark < rl_end))
-    rl_mark++;
+    INCREMENT_POS (rl_mark);
 
   rl_kill_text (rl_point, rl_mark);
   return (0);
@@ -1334,7 +1428,7 @@ vi_change_dispatch (_rl_vimotion_cxt *m)
      and already leave the mark at the correct location. */
   if (((strchr (" l|hwW^0bBFT`", m->motion) == 0) && (rl_point >= m->start)) &&
       (rl_mark < rl_end))
-    rl_mark++;
+    INCREMENT_POS (rl_mark);
 
   /* The cursor never moves with c[wW]. */
   if ((_rl_to_upper (m->motion) == 'W') && rl_point < m->start)
@@ -1421,7 +1515,7 @@ vi_yank_dispatch (_rl_vimotion_cxt *m)
      mark. */
   if (((strchr (" l|h^0%bBFT`", m->motion) == 0) && (rl_point >= m->start)) &&
       (rl_mark < rl_end))
-    rl_mark++;
+    INCREMENT_POS (rl_mark);
 
   rl_begin_undo_group ();
   rl_kill_text (rl_point, rl_mark);
@@ -1611,8 +1705,8 @@ rl_vi_unix_word_rubout (int count, int key)
 	    while (rl_point && vi_unix_word_boundary (rl_line_buffer[rl_point - 1]))
 	      rl_point--;
 	  else if (rl_point > 0 && vi_unix_word_boundary (rl_line_buffer[rl_point]) == 0)
-	    while (rl_point && (vi_unix_word_boundary (rl_line_buffer[rl_point - 1]) == 0))
-	      rl_point--;
+	    while (rl_point > 0 && (vi_unix_word_boundary (rl_line_buffer[rl_point - 1]) == 0))
+	      _rl_vi_backup_point ();
 	}
 
       rl_kill_text (orig_point, rl_point);
@@ -2100,7 +2194,7 @@ rl_vi_possible_completions (void)
     {
       while (rl_point < rl_end && rl_line_buffer[rl_point] != ' ' &&
 	     rl_line_buffer[rl_point] != ';')
-	rl_point++;
+	_rl_vi_advance_point ();
     }
   else if (rl_line_buffer[rl_point - 1] == ';')
     {
diff --git a/lib/sh/Makefile.in b/lib/sh/Makefile.in
index b1a086d6..06917ba4 100644
--- a/lib/sh/Makefile.in
+++ b/lib/sh/Makefile.in
@@ -92,7 +92,8 @@ CSOURCES = clktck.c clock.c getcwd.c getenv.c oslib.c setlinebuf.c \
 	   mktime.c strftime.c mbschr.c zcatfd.c zmapfd.c winsize.c eaccess.c \
 	   wcsdup.c fpurge.c zgetline.c mbscmp.c uconvert.c ufuncs.c \
 	   casemod.c dprintf.c input_avail.c mbscasecmp.c fnxform.c \
-	   strchrnul.c unicode.c wcswidth.c wcsnwidth.c shmbchar.c strdup.c
+	   strchrnul.c unicode.c wcswidth.c wcsnwidth.c shmbchar.c strdup.c \
+	   utf8.c
 
 # The header files for this library.
 HSOURCES = 
@@ -107,7 +108,7 @@ OBJECTS = clktck.o clock.o getenv.o oslib.o setlinebuf.o strnlen.o \
 	  fmtullong.o fmtumax.o zcatfd.o zmapfd.o winsize.o wcsdup.o \
 	  fpurge.o zgetline.o mbscmp.o uconvert.o ufuncs.o casemod.o \
 	  input_avail.o mbscasecmp.o fnxform.o unicode.o shmbchar.o \
-	  wcsnwidth.o ${LIBOBJS}
+	  utf8.o wcsnwidth.o ${LIBOBJS}
 
 SUPPORT = Makefile
 
@@ -200,6 +201,7 @@ tmpfile.o: tmpfile.c
 uconvert.o: uconvert.c
 ufuncs.o: ufuncs.c
 unicode.o: unicode.c
+utf8.o: utf8.c
 vprint.o: vprint.c
 wcsdup.o: wcsdup.c
 wcsnwidth.o: wcsnwidth.c
@@ -277,6 +279,7 @@ tmpfile.o: ${BUILD_DIR}/config.h
 uconvert.o: ${BUILD_DIR}/config.h
 ufuncs.o: ${BUILD_DIR}/config.h
 unicode.o: ${BUILD_DIR}/config.h
+utf8.o: ${BUILD_DIR}/config.h
 vprint.o: ${BUILD_DIR}/config.h
 wcsdup.o: ${BUILD_DIR}/config.h
 wcsnwidth.o: ${BUILD_DIR}/config.h
@@ -612,6 +615,10 @@ unicode.o: ${topdir}/bashansi.h ${BASHINCDIR}/ansi_stdlib.h
 unicode.o: ${BASHINCDIR}/stdc.h
 unicode.o: ${topdir}/xmalloc.h
 
+utf8.o: ${topdir}/bashansi.h
+utf8.o: ${BASHINCDIR}/ansi_stdlib.h
+utf8.o: ${BASHINCDIR}/shmbutil.h ${BASHINCDIR}/shmbchar.h
+
 winsize.o: ${BASHINCDIR}/stdc.h
 winsize.o: ${topdir}/xmalloc.h
 winsize.o: ${topdir}/bashtypes.h
diff --git a/lib/sh/getenv.c b/lib/sh/getenv.c
index 8b5e3406..1e682aef 100644
--- a/lib/sh/getenv.c
+++ b/lib/sh/getenv.c
@@ -69,7 +69,7 @@ getenv (name)
       if (var && exported_p (var))
 	return (value_cell (var));
     }
-  else
+  else if (environ)
     {
       register int i, len;
 
diff --git a/lib/sh/mbschr.c b/lib/sh/mbschr.c
index fdce6d5a..639962d4 100644
--- a/lib/sh/mbschr.c
+++ b/lib/sh/mbschr.c
@@ -32,13 +32,7 @@ extern int locale_utf8locale;
 
 #undef mbschr
 
-static inline char *
-utf8_mbschr (s, c)
-     const char *s;
-     int c;
-{
-  return strchr (s, c);		/* for now */
-}
+extern char *utf8_mbschr (const char *, int);	/* XXX */
 
 /* In some locales, the non-first byte of some multibyte characters have
    the same value as some ascii character.  Faced with these strings, a
diff --git a/lib/sh/mbscmp.c b/lib/sh/mbscmp.c
index aaf81f5e..c7c84435 100644
--- a/lib/sh/mbscmp.c
+++ b/lib/sh/mbscmp.c
@@ -1,6 +1,6 @@
 /* mbscmp - multibyte string comparison. */
 
-/* Copyright (C) 1995-2015 Free Software Foundation, Inc.
+/* Copyright (C) 1995-2018 Free Software Foundation, Inc.
 
    This file is part of GNU Bash, the Bourne Again SHell.
 
@@ -26,6 +26,10 @@
 #include <stddef.h>
 #include <string.h>
 
+extern int locale_utf8locale;
+
+extern int utf8_mbscmp (const char *, const char *);
+
 /* Compare MBS1 and MBS2.  */
 int
 mbscmp (mbs1, mbs2)
diff --git a/lib/sh/shmbchar.c b/lib/sh/shmbchar.c
index 80858ebf..f2f2582b 100644
--- a/lib/sh/shmbchar.c
+++ b/lib/sh/shmbchar.c
@@ -1,4 +1,4 @@
-/* Copyright (C) 2001, 2006, 2009, 2010, 2012, 2015 Free Software Foundation, Inc.
+/* Copyright (C) 2001, 2006, 2009, 2010, 2012, 2015-2018 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
@@ -45,102 +45,9 @@ const unsigned int is_basic_table [UCHAR_MAX / 32 + 1] =
 
 extern int locale_utf8locale;
 
-/* We can optimize this if we know the locale is UTF-8, but needs to handle
-   malformed byte sequences. */
-static inline size_t
-utf8_mbstrlen(s)
-     const char *s;
-{
-  size_t num = 0;
-  register unsigned char c;
-
-  while ((c = *s++))
-    /* bytes 0xc0 through 0xff are first byte of multi-byte sequence */
-    if ((c & 0xc0) != 0x80)	/* skip continuation bytes */
-      ++num;
-  return (num);
-}
-
-/* Adapted from GNU libutf8 */
-static inline int
-utf8_mblen (s, n)
-     const char *s;
-     int n;
-{
-  unsigned char c;
-
-  if (s == 0)
-    return 0;
-  else if (n == 0)
-    return -1;
-
-  c = (unsigned char) *s;
-  if (c < 0x80)
-    return (c != 0);
-  else if (c < 0xc0)
-    goto return_error;
-  else
-    {
-      const char *start = s;
-      size_t count;
-      int check_unsafe;
+extern char *utf8_mbsmbchar (const char *);
+extern int utf8_mblen (const char *, size_t);
 
-      if (c < 0xe0)
-        {
-          count = 1;
-          if (c < 0xc2)
-	    goto return_error;
-	  check_unsafe = 0;
-        }
-      else if (c < 0xf0)
-	{
-	  count = 2;
-	  check_unsafe = (c == 0xe0);
-	}
-#if SIZEOF_WCHAR_T == 4
-      else if (c < 0xf8)
-	{
-	  count = 3;
-	  check_unsafe = (c == 0xe0);
-	}
-      else if (c < 0xfc)
-	{
-	  count = 4;
-	  check_unsafe = (c == 0xf8);
-	}
-      else if (c < 0xfe)
-	{
-	  count = 5;
-	  check_unsafe = (c == 0xfc);
-	}
-#endif
-      else
-	goto return_error;
-      if (n <= count)
-	return -1;
-      s++;
-      c = (unsigned char) *s++ ^ 0x80;
-      if (c >= 0x40)
-	goto return_error;
-      if (--count > 0)
-	{
-	  if (check_unsafe && ((c >> (6 - count)) == 0))
-	    goto return_error;
-	  do
-	    {
-	      c = (unsigned char) *s++ ^ 0x80;
-	      if (c >= 0x40)
-		goto return_error;
-	    }
-	  while (--count > 0);
-	}
-      return s - start;
-    }
-return_error:
-  errno = EILSEQ;
-  return -1;
-}
-  
 /* Count the number of characters in S, counting multi-byte characters as a
    single character. */
 size_t
@@ -170,18 +77,6 @@ mbstrlen (s)
   return nc;
 }
 
-static inline char *
-utf8_mbsmbchar (str)
-     const char *str;
-{
-  register char *s;
-
-  for (s = (char *)str; *s; s++)
-    if ((*s & 0xc0) == 0x80)
-      return s;
-  return (0);
-}
-
 /* Return pointer to first multibyte char in S, or NULL if none. */
 /* XXX - if we know that the locale is UTF-8, we can just check whether or
    not any byte has the eighth bit turned on */
@@ -219,22 +114,6 @@ mbsmbchar (s)
   return 0;
 }
 
-static inline int
-utf_mbsnlen(src, srclen, maxlen)
-     const char *src;
-     size_t srclen;
-     int maxlen;
-{
-  register int sind, count;
-
-  for (sind = count = 0; src[sind] && sind <= maxlen; sind++)
-    {
-      if ((src[sind] & 0xc0) != 0x80)
-	count++;
-    }
-  return (count);
-}
-
 int
 sh_mbsnlen(src, srclen, maxlen)
      const char *src;
diff --git a/lib/sh/snprintf.c b/lib/sh/snprintf.c
index 87ca2173..6e5892ee 100644
--- a/lib/sh/snprintf.c
+++ b/lib/sh/snprintf.c
@@ -142,9 +142,11 @@ extern char *fmtullong __P((unsigned long long int, int, char *, size_t, int));
    302 / 1000 is log10 (2) rounded up;
    add one for integer division truncation;
    add one more for a minus sign if t is signed.  */
+#ifndef INT_STRLEN_BOUND
 #define INT_STRLEN_BOUND(t) \
   ((sizeof (t) * CHAR_BIT - TYPE_SIGNED (t)) * 302 / 1000 \
      + 1 + TYPE_SIGNED (t))
+#endif
 
 /* conversion flags */
 #define PF_ALTFORM	0x00001		/* # */
diff --git a/lib/sh/utf8.c b/lib/sh/utf8.c
new file mode 100644
index 00000000..d27fcf54
--- /dev/null
+++ b/lib/sh/utf8.c
@@ -0,0 +1,147 @@
+/* utf8.c - UTF-8 character handling functions */
+
+/* Copyright (C) 2018 Free Software Foundation, Inc.
+
+   This file is part of GNU Bash, the Bourne Again SHell.
+
+   Bash 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.
+
+   Bash 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 Bash.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#include <config.h>
+
+#ifdef HAVE_STDLIB_H
+#  include <stdlib.h>
+#endif
+
+#include "bashansi.h"
+#include "shmbutil.h"
+
+extern int locale_mb_cur_max;
+extern int locale_utf8locale;
+
+#if defined (HANDLE_MULTIBYTE)
+
+char *
+utf8_mbschr (s, c)
+     const char *s;
+     int c;
+{
+  return strchr (s, c);		/* for now */
+}
+
+int
+utf8_mbscmp (s1, s2)
+     const char *s1, *s2;
+{
+  /* Use the fact that the UTF-8 encoding preserves lexicographic order.  */
+  return strcmp (s1, s2);
+}
+
+char *
+utf8_mbsmbchar (str)
+     const char *str;
+{
+  register char *s;
+
+  for (s = (char *)str; *s; s++)
+    if ((*s & 0xc0) == 0x80)
+      return s;
+  return (0);
+}
+
+int
+utf8_mbsnlen(src, srclen, maxlen)
+     const char *src;
+     size_t srclen;
+     int maxlen;
+{
+  register int sind, count;
+
+  for (sind = count = 0; src[sind] && sind <= maxlen; sind++)
+    {
+      if ((src[sind] & 0xc0) != 0x80)
+	count++;
+    }
+  return (count);
+}
+
+/* Adapted from GNU gnulib */
+int
+utf8_mblen (s, n)
+     const char *s;
+     size_t n;
+{
+  unsigned char c, c1;
+
+  if (s == 0)
+    return (0);	/* no shift states */
+  if (n <= 0)
+    return (-1);
+
+  c = (unsigned char)*s;
+  if (c < 0x80)
+    return (c != 0);
+  if (c >= 0xc2)
+    {
+      c1 = (unsigned char)s[1];
+      if (c < 0xe0)
+	{
+	  if (n >= 2 && (s[1] ^ 0x80) < 0x40)
+	    return 2;
+	}
+      else if (c < 0xf0)
+	{
+	  if (n >= 3
+		&& (s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+		&& (c >= 0xe1 || c1 >= 0xa0)
+		&& (c != 0xed || c1 < 0xa0))
+	    return 3;
+	}
+      else if (c < 0xf8)
+	{
+	  if (n >= 4
+		&& (s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
+	 	&& (s[3] ^ 0x80) < 0x40
+		&& (c >= 0xf1 || c1 >= 0x90)
+		&& (c < 0xf4 || (c == 0xf4 && c1 < 0x90)))
+	    return 4;
+	}
+    }
+  /* invalid or incomplete multibyte character */
+  return -1;
+}
+
+/* We can optimize this if we know the locale is UTF-8, but needs to handle
+   malformed byte sequences. */
+size_t
+utf8_mbstrlen(s)
+     const char *s;
+{
+  size_t clen, nc;
+  int mb_cur_max;
+
+  nc = 0;
+  mb_cur_max = MB_CUR_MAX;
+  while (*s && (clen = (size_t)utf8_mblen(s, mb_cur_max)) != 0)
+    {
+      if (MB_INVALIDCH(clen))
+	clen = 1;	/* assume single byte */
+
+      s += clen;
+      nc++;
+    }
+  return nc;
+}
+
+#endif