Update.

1997-10-29 21:20  Ulrich Drepper  <drepper@cygnus.com>

	* libio/strops.c (_IO_str_seekoff): If mode is zero and the read/write
	pointers are tied set mode according to current stream mode.

	* include/features.h [_GNU_SOURCE] (_POSIX_C_SOURCE): Define to
	199506L.
	Define _XOPEN_SOURCE to 500 if _POSIX_C_SOURCE is defined.
	* manual/creature.texi: Describe this.

	* manual/socket.texi: Describe connect, accept, send, sendmsg, sendto,
	recv, recvfrom, and recvmsg as cancelation points.
	* sysdeps/unix/inet/syscalls.list: Add __libc_* names for these
	functions.

1997-10-17  Andreas Jaeger  <aj@arthur.rhein-neckar.de>

	* Make-dist (try-sysdeps): Don't look for stub files anymore.

	* manual/maint.texi (Porting): Remove another reference to stub
	directory.

	* sysdeps/unix/bsd/sun/sethostid.c: Include stub version from
	generic subdir.
	* sysdeps/unix/sysv/irix4/reboot.c: Likewise.
	* sysdeps/unix/sysv/irix4/swapon.c: Likewise

1997-10-29 03:54  Ulrich Drepper  <drepper@cygnus.com>

	* resolv/nss_dns/dns-host.c: Change variable pointed to by h_errnop
	in all error cases (PR 244).

1997-10-29 00:56  Ulrich Drepper  <drepper@cygnus.com>

	* posix/glob.c: Fix handling of expressions like "*/" (PR 325).
	Optimize by using mempcpy.

1997-10-17  Andreas Jaeger  <aj@arthur.rhein-neckar.de>

	* po/Makefile ($(mo-installed)): Don't fail during installation if
	files don't exist (might happen if msgfmt doesn't exist) (PR 328).
	Suggested by Wolfram Gloger <wmglo@dent.med.uni-muenchen.de>.

1997-10-24  Andreas Jaeger  <aj@arthur.rhein-neckar.de>

	* sysdeps/generic/bits/errno.h (ENOMSG): Define it.
	Pointed out by Klaus Espenlaub
	<kespenla@hydra.informatik.uni-ulm.de> (PR libc/259).

1997-10-28 17:40  Ulrich Drepper  <drepper@cygnus.com>

	* sysdeps/libm-ieee754/s_cbrt.c: Testing the returned exponent for
	zero isn't enough to determine illegal arguments.
	* sysdeps/libm-ieee754/s_cbrtf.c: Likewise.
	* sysdeps/libm-ieee754/s_cbrtl.c: Likewise.

1997-10-28 17:14  Ulrich Drepper  <drepper@cygnus.com>

	* manual/filesys.texi (S_ISVTX): Describe that it is available with
	_BSD_SOURCE only.
	Reported by Jochen Hein <jochen.hein@delphi.central.de>.

1997-10-28 04:26  Ulrich Drepper  <drepper@cygnus.com>

	* time/tzfile.c (__tzfile_compute): Use negated value of offset for
	timezone variable.
	* time/tzset.c (tz_compute): Likewise.
	Reported by Erik Troan <ewt@redhat.com>.

1997-10-28 02:51  Ulrich Drepper  <drepper@cygnus.com>

	* manual/filesys.texi: Correct prototype in readdir_r description.
	Reported by Jim Meyering <meyering@eng.ascend.com>.

1997-10-27  Andreas Jaeger  <aj@arthur.rhein-neckar.de>

	* math/libm-test.c (cbrt_test): Add test for cbrt(0.970299).

1997-10-26 19:39  Zack Weinberg  <zack@rabi.phys.columbia.edu>

	* stdlib/l64a.c: Produce a useful result for n < 0.
	* stdlib/a64l.c: Use unsigned type for working variable.
	* manual/string.texi (general): Grammar, typo, overfull fixes.
	(strlen): Insert warning about sizeof(char *).
	(a64l, l64a): Make documentation agree with implementation.

	* libio/iofdopen.c: Use _IO_FILE_complete, not _IO_FILE_plus.
	* posix/unistd.h: Add prototypes for __pread, __pread64, __pwrite

5 files changed, 181 insertions, 100 deletions

diff --git a/manual/creature.texi b/manual/creature.texi
index 38a11c0093..03f79ede96 100644
--- a/manual/creature.texi
+++ b/manual/creature.texi
@@ -36,6 +36,11 @@ available.  If you define this macro with a value of @code{2}, then both
 the functionality from the POSIX.1 standard and the functionality from
 the POSIX.2 standard (IEEE Standard 1003.2) are made available.  This is
 in addition to the @w{ISO C} facilities.
+
+The Single Unix Specification specify that setting this macro to the
+value @code{199506L} selects all the values specified by the POSIX
+standards plus those of the Single Unix Specification, i.e., is the
+same as if @code{_XOPEN_SOURCE} is set to @code{500} (see below).
 @end defvr
 
 @comment (none)
@@ -88,7 +93,7 @@ available which are necessary for the X/Open Unix brand.
 
 If the macro @code{_XOPEN_SOURCE} has the value @math{500} this includes
 all functionality described so far plus some new definitions from the
-Single Unix specification, @w{version 2}.
+Single Unix Specification, @w{version 2}.
 @end defvr
 
 @comment (NONE)
diff --git a/manual/filesys.texi b/manual/filesys.texi
index 4cf4f99c07..c33bad892e 100644
--- a/manual/filesys.texi
+++ b/manual/filesys.texi
@@ -340,7 +340,7 @@ value.  Use @code{readdir_r} when this is critical.
 
 @comment dirent.h
 @comment GNU
-@deftypefun int readdir_r (DIR *@var{dirstream}, struct *@var{entry}, struct **@var{result})
+@deftypefun int readdir_r (DIR *@var{dirstream}, struct dirent *@var{entry}, struct dirent **@var{result})
 This function is the reentrant version of @code{readdir}.  Like
 @code{readdir} it returns the next entry from the directory.  But to
 prevent conflicts for simultaneously running threads the result is not
@@ -1783,6 +1783,10 @@ waste of the server's memory to cache them a second time.  In this use
 the sticky bit also says that the filesystem may fail to record the
 file's modification time onto disk reliably (the idea being that no-one
 cares for a swap file).
+
+This bit is only available on BSD systems (and those derived from
+them).  Therefore one has to use the @code{_BSD_SOURCE} feature select
+macro to get the definition (@pxref{Feature Test Macros}).
 @end table
 
 The actual bit values of the symbols are listed in the table above
diff --git a/manual/maint.texi b/manual/maint.texi
index 7698549bf3..6d4185b52c 100644
--- a/manual/maint.texi
+++ b/manual/maint.texi
@@ -538,11 +538,11 @@ include the file @code{<stub-tag.h>} into your file.  This causes the
 function to be listed in the installed @code{<gnu/stubs.h>}, and
 makes GNU ld warn when the function is used.
 
-Some rare functions are only useful on specific systems and aren't
-defined at all on others; these do not appear anywhere in the
-system-independent source code or makefiles (including the
-@file{generic} and @file{stub} directories), only in the
-system-dependent @file{Makefile} in the specific system's subdirectory.
+Some rare functions are only useful on specific systems and
+aren't defined at all on others; these do not appear anywhere
+in the system-independent source code or makefiles (including the
+@file{generic}), only in the system-dependent @file{Makefile} in the
+specific system's subdirectory.
 
 If you come across a file that is in one of the main source directories
 (@file{string}, @file{stdio}, etc.), and you want to write a machine- or
diff --git a/manual/socket.texi b/manual/socket.texi
index 76d7863728..45b9bbb892 100644
--- a/manual/socket.texi
+++ b/manual/socket.texi
@@ -1769,6 +1769,12 @@ completely established, will fail with @code{EALREADY}.
 The socket @var{socket} is non-blocking and already has a pending
 connection in progress (see @code{EINPROGRESS} above).
 @end table
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 @node Listening
@@ -1897,6 +1903,12 @@ The descriptor @var{socket} does not support this operation.
 @var{socket} has nonblocking mode set, and there are no pending
 connections immediately available.
 @end table
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 The @code{accept} function is not allowed for sockets using
@@ -2023,6 +2035,12 @@ case, @code{send} generates a @code{SIGPIPE} signal first; if that
 signal is ignored or blocked, or if its handler returns, then
 @code{send} fails with @code{EPIPE}.
 @end table
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 @node Receiving Data
@@ -2067,6 +2085,12 @@ The operation was interrupted by a signal before any data was read.
 @item ENOTCONN
 You never connected this socket.
 @end table
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 @node Socket Data Options
@@ -2366,6 +2390,12 @@ system on your machine usually does not know this.
 
 It is also possible for one call to @code{sendto} to report an error
 due to a problem related to a previous call.
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 @node Receiving Datagrams
@@ -2398,6 +2428,12 @@ if you are not interested in this information.
 The @var{flags} are interpreted the same way as for @code{recv}
 (@pxref{Socket Data Options}).  The return value and error conditions
 are also the same as for @code{recv}.
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 You can use plain @code{recv} (@pxref{Receiving Data}) instead of
@@ -2420,11 +2456,23 @@ you don't want to specify @var{flags} (@pxref{I/O Primitives}).
 @comment sys/socket.h
 @comment BSD
 @deftypefun int sendmsg (int @var{socket}, const struct msghdr *@var{message}, int @var{flags})
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 
 @comment sys/socket.h
 @comment BSD
 @deftypefun int recvmsg (int @var{socket}, struct msghdr *@var{message}, int @var{flags})
+
+This function is defined as a cancelation point in multi-threaded
+programs.  So one has to be prepared for this and make sure that
+possibly allocated resources (like memory, files descriptors,
+semaphores or whatever) are freed even if the thread is cancel.
+@c @xref{pthread_cleanup_push}, for a method how to do this.
 @end deftypefun
 @end ignore
 
diff --git a/manual/string.texi b/manual/string.texi
index 3d60fa4d2e..f33303b356 100644
--- a/manual/string.texi
+++ b/manual/string.texi
@@ -82,7 +82,7 @@ string.
 
 The amount of memory allocated for the character array may extend past
 the null character that normally marks the end of the string.  In this
-document, the term @dfn{allocation size} is always used to refer to the
+document, the term @dfn{allocated size} is always used to refer to the
 total amount of memory allocated for the string, while the term
 @dfn{length} refers to the number of characters up to (but not
 including) the terminating null character.
@@ -155,8 +155,8 @@ strlen ("hello, world")
 @end smallexample
 
 When applied to a character array, the @code{strlen} function returns
-the length of the string stored there, not its allocation size.  You can
-get the allocation size of the character array that holds a string using
+the length of the string stored there, not its allocated size.  You can
+get the allocated size of the character array that holds a string using
 the @code{sizeof} operator:
 
 @smallexample
@@ -166,6 +166,22 @@ sizeof (string)
 strlen (string)
     @result{} 12
 @end smallexample
+
+But beware, this will not work unless @var{string} is the character
+array itself, not a pointer to it.  For example:
+
+@smallexample
+char string[32] = "hello, world";
+char *ptr = string;
+sizeof (string)
+    @result{} 32
+sizeof (ptr)
+    @result{} 4  /* @r{(on a machine with 4 byte pointers)} */
+@end smallexample
+
+This is an easy mistake to make when you are working with functions that
+take string arguments; those arguments are always pointers, not arrays.
+
 @end deftypefun
 
 @comment string.h
@@ -397,7 +413,7 @@ is implemented to be useful in contexts where this behaviour of the
 @emph{first} written null character.
 
 This function is not part of ISO or POSIX but was found useful while
-developing GNU C Library itself.
+developing the GNU C Library itself.
 
 Its behaviour is undefined if the strings overlap.
 @end deftypefun
@@ -406,12 +422,11 @@ Its behaviour is undefined if the strings overlap.
 @comment GNU
 @deftypefn {Macro} {char *} strdupa (const char *@var{s})
 This function is similar to @code{strdup} but allocates the new string
-using @code{alloca} instead of @code{malloc}
-@pxref{Variable Size Automatic}.  This means of course the returned
-string has the same limitations as any block of memory allocated using
-@code{alloca}.
+using @code{alloca} instead of @code{malloc} (@pxref{Variable Size
+Automatic}).  This means of course the returned string has the same
+limitations as any block of memory allocated using @code{alloca}.
 
-For obvious reasons @code{strdupa} is implemented only as a macro.  I.e.,
+For obvious reasons @code{strdupa} is implemented only as a macro;
 you cannot get the address of this function.  Despite this limitation
 it is a useful function.  The following code shows a situation where
 using @code{malloc} would be a lot more expensive.
@@ -434,8 +449,7 @@ allocates the new string using @code{alloca}
 @pxref{Variable Size Automatic}.  The same advantages and limitations
 of @code{strdupa} are valid for @code{strndupa}, too.
 
-This function is implemented only as a macro which means one cannot
-get the address of it.
+This function is implemented only as a macro, just like @code{strdupa}.
 
 @code{strndupa} is only available if GNU CC is used.
 @end deftypefn
@@ -613,10 +627,10 @@ is an initial substring of @var{s2}, then @var{s1} is considered to be
 @comment BSD
 @deftypefun int strcasecmp (const char *@var{s1}, const char *@var{s2})
 This function is like @code{strcmp}, except that differences in case are
-ignored.  How uppercase and lowercase character are related is
+ignored.  How uppercase and lowercase characters are related is
 determined by the currently selected locale.  In the standard @code{"C"}
 locale the characters @"A and @"a do not match but in a locale which
-regards this characters as parts of the alphabet they do match.
+regards these characters as parts of the alphabet they do match.
 
 @code{strcasecmp} is derived from BSD.
 @end deftypefun
@@ -625,8 +639,8 @@ regards this characters as parts of the alphabet they do match.
 @comment BSD
 @deftypefun int strncasecmp (const char *@var{s1}, const char *@var{s2}, size_t @var{n})
 This function is like @code{strncmp}, except that differences in case
-are ignored.  Like for @code{strcasecmp} it is locale dependent how
-uppercase and lowercase character are related.
+are ignored.  Like @code{strcasecmp}, it is locale dependent how
+uppercase and lowercase characters are related.
 
 @code{strncasecmp} is a GNU extension.
 @end deftypefun
@@ -671,7 +685,7 @@ function.  In fact, if @var{s1} and @var{s2} contain no digits,
 
 Basically, we compare strings normally (character by character), until
 we find a digit in each string - then we enter a special comparison
-mode, where each sequence of digit is taken as a whole.  If we reach the
+mode, where each sequence of digits is taken as a whole.  If we reach the
 end of these two parts without noticing a difference, we return to the
 standard comparison mode.  There are two types of numeric parts:
 "integral" and "fractional" (those  begin with a '0'). The types
@@ -693,7 +707,7 @@ than the other one; else the comparison behaves normally.
 
 @smallexample
 strverscmp ("no digit", "no digit")
-    @result{} 0    /* @r{same behaviour as strverscmp.} */
+    @result{} 0    /* @r{same behaviour as strcmp.} */
 strverscmp ("item#99", "item#100")
     @result{} <0   /* @r{same prefix, but 99 < 100.} */
 strverscmp ("alpha1", "alpha001")
@@ -873,7 +887,8 @@ sort_strings_fast (char **array, int nstrings)
 
           /* @r{The return value is not interesting because we know}
              @r{how long the transformed string is.}  */
-          (void) strxfrm (transformed, array[i], transformed_length + 1);
+          (void) strxfrm (transformed, array[i],
+                          transformed_length + 1);
         @}
 
       temp_array[i].transformed = transformed;
@@ -1096,12 +1111,11 @@ a null pointer.
 @end deftypefun
 
 @strong{Warning:} Since @code{strtok} alters the string it is parsing,
-you always copy the string to a temporary buffer before parsing it with
-@code{strtok}.  If you allow @code{strtok} to modify a string that came
-from another part of your program, you are asking for trouble; that
-string may be part of a data structure that could be used for other
-purposes during the parsing, when alteration by @code{strtok} makes the
-data structure temporarily inaccurate.
+you should always copy the string to a temporary buffer before parsing
+it with @code{strtok}.  If you allow @code{strtok} to modify a string
+that came from another part of your program, you are asking for trouble;
+that string might be used for other purposes after @code{strtok} has
+modified it, and it would not have the expected value.
 
 The string that you are operating on might even be a constant.  Then
 when @code{strtok} tries to modify it, your program will get a fatal
@@ -1146,14 +1160,13 @@ which overcome the limitation of non-reentrancy.
 @comment string.h
 @comment POSIX
 @deftypefun {char *} strtok_r (char *@var{newstring}, const char *@var{delimiters}, char **@var{save_ptr})
-Just like @code{strtok} this function splits the string into several
-tokens which can be accessed be successive calls to @code{strtok_r}.
-The difference is that the information about the next token is not set
-up in some internal state information.  Instead the caller has to
-provide another argument @var{save_ptr} which is a pointer to a string
-pointer.  Calling @code{strtok_r} with a null pointer for
-@var{newstring} and leaving @var{save_ptr} between the calls unchanged
-does the job without limiting reentrancy.
+Just like @code{strtok}, this function splits the string into several
+tokens which can be accessed by successive calls to @code{strtok_r}.
+The difference is that the information about the next token is stored in
+the space pointed to by the third argument, @var{save_ptr}, which is a
+pointer to a string pointer.  Calling @code{strtok_r} with a null
+pointer for @var{newstring} and leaving @var{save_ptr} between the calls
+unchanged does the job without hindering reentrancy.
 
 This function is defined in POSIX-1 and can be found on many systems
 which support multi-threading.
@@ -1162,12 +1175,12 @@ which support multi-threading.
 @comment string.h
 @comment BSD
 @deftypefun {char *} strsep (char **@var{string_ptr}, const char *@var{delimiter})
-A second reentrant approach is to avoid the additional first argument.
-The initialization of the moving pointer has to be done by the user.
-Successive calls of @code{strsep} move the pointer along the tokens
-separated by @var{delimiter}, returning the address of the next token
-and updating @var{string_ptr} to point to the beginning of the next
-token.
+This function is just @code{strtok_r} with the @var{newstring} argument
+replaced by the @var{save_ptr} argument.  The initialization of the
+moving pointer has to be done by the user.  Successive calls to
+@code{strsep} move the pointer along the tokens separated by
+@var{delimiter}, returning the address of the next token and updating
+@var{string_ptr} to point to the beginning of the next token.
 
 This function was introduced in 4.3BSD and therefore is widely available.
 @end deftypefun
@@ -1204,47 +1217,30 @@ token = strsep (&running, delimiters);    /* token => NULL */
 To store or transfer binary data in environments which only support text
 one has to encode the binary data by mapping the input bytes to
 characters in the range allowed for storing or transfering.  SVID
-systems (and nowadays XPG compliant systems) have such a function in the
-C library.
+systems (and nowadays XPG compliant systems) provide minimal support for
+this task.
 
 @comment stdlib.h
 @comment XPG
 @deftypefun {char *} l64a (long int @var{n})
-This function encodes an input value with 32 bits using characters from
-the basic character set.  Groups of 6 bits are encoded using the
-following table:
-
-@multitable {xxxxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx}
-@item              @tab 0 @tab 1 @tab 2 @tab 3 @tab 4 @tab 5 @tab 6 @tab 7
-@item       0      @tab @code{.} @tab @code{/} @tab @code{0} @tab @code{1}
-                   @tab @code{2} @tab @code{3} @tab @code{4} @tab @code{5}
-@item       8      @tab @code{6} @tab @code{7} @tab @code{8} @tab @code{9}
-                   @tab @code{A} @tab @code{B} @tab @code{C} @tab @code{D}
-@item       16     @tab @code{E} @tab @code{F} @tab @code{G} @tab @code{H}
-                   @tab @code{I} @tab @code{J} @tab @code{K} @tab @code{L}
-@item       24     @tab @code{M} @tab @code{N} @tab @code{O} @tab @code{P}
-                   @tab @code{Q} @tab @code{R} @tab @code{S} @tab @code{T}
-@item       32     @tab @code{U} @tab @code{V} @tab @code{W} @tab @code{X}
-                   @tab @code{Y} @tab @code{Z} @tab @code{a} @tab @code{b}
-@item       40     @tab @code{c} @tab @code{d} @tab @code{e} @tab @code{f}
-                   @tab @code{g} @tab @code{h} @tab @code{i} @tab @code{j}
-@item       48     @tab @code{k} @tab @code{l} @tab @code{m} @tab @code{n}
-                   @tab @code{o} @tab @code{p} @tab @code{q} @tab @code{r}
-@item       56     @tab @code{s} @tab @code{t} @tab @code{u} @tab @code{v}
-                   @tab @code{w} @tab @code{x} @tab @code{y} @tab @code{z}
-@end multitable
-
-The function returns a pointer to a static buffer which contains the
-string representing of the encoding of @var{n}.  To encoded a series of
-bytes the use should append the new string to the destination buffer.
-@emph{Warning:} Since a static buffer is used this function should not
+This function encodes a 32-bit input value using characters from the
+basic character set.  It returns a pointer to a 6 character buffer which
+contains an encoded version of @var{n}.  To encode a series of bytes the
+user must copy the returned string to a destination buffer.  It returns
+the empty string if @var{n} is zero, which is somewhat bizarre but
+mandated by the standard.@*
+@strong{Warning:} Since a static buffer is used this function should not
 be used in multi-threaded programs.  There is no thread-safe alternative
-to this function in the C library.
-@end deftypefun
+to this function in the C library.@*
+@strong{Compatibility Note:} The XPG standard states that the return
+value of @code{l64a} is undefined if @var{n} is negative.  In the GNU
+implementation, @code{l64a} treats its argument as unsigned, so it will
+return a sensible encoding for any nonzero @var{n}; however, portable
+programs should not rely on this.
 
-Alone the @code{l64a} function is not usable.  To encode arbitrary
-sequences of bytes one needs some more code and this could look like
-this:
+To encode a large buffer @code{l64a} must be called in a loop, once for
+each 32-bit word of the buffer.  For example, one could do something
+like this:
 
 @smallexample
 char *
@@ -1256,8 +1252,10 @@ encode (const void *buf, size_t len)
   char *cp = out;
 
   /* @r{Encode the length.} */
-  memcpy (cp, l64a (len), 6);
-  cp += 6;
+  /* @r{Using `htonl' is necessary so that the data can be}
+     @r{decoded even on machines with different byte order.} */
+
+  cp = mempcpy (cp, l64a (htonl (len)), 6);
 
   while (len > 3)
     @{
@@ -1266,10 +1264,12 @@ encode (const void *buf, size_t len)
       n = (n << 8) | *in++;
       n = (n << 8) | *in++;
       len -= 4;
-      /* @r{Using `htonl' is necessary so that the data can be}
-         @r{decoded even on machines with different byte order.} */
-      memcpy (cp, l64a (htonl (n)), 6);
-      cp += 6;
+      if (n)
+        cp = mempcpy (cp, l64a (htonl (n)), 6);
+      else
+            /* @r{`l64a' returns the empty string for n==0, so we }
+               @r{must generate its encoding (}"......"@r{) by hand.} */
+        cp = stpcpy (cp, "......");
     @}
   if (len > 0)
     @{
@@ -1289,9 +1289,9 @@ encode (const void *buf, size_t len)
 @end smallexample
 
 It is strange that the library does not provide the complete
-functionality needed but so be it.  There are some other encoding
-methods which are much more widely used (UU encoding, Base64 encoding).
-Generally, it is better to use one of these encodings.
+functionality needed but so be it.
+
+@end deftypefun
 
 To decode data produced with @code{l64a} the following function should be
 used.
@@ -1300,19 +1300,43 @@ used.
 @comment XPG
 @deftypefun {long int} a64l (const char *@var{string})
 The parameter @var{string} should contain a string which was produced by
-a call to @code{l64a}.  The function processes the next 6 characters and
-decodes the characters it finds according to the table above.
-Characters not in the conversion table are simply ignored.  This is
-useful for breaking the information in lines in which case the end of
-line characters are simply ignored.
-
-The decoded number is returned at the end as a @code{long int} value.
-Consecutive calls to this function are possible but the caller must make
-sure the buffer pointer is update after each call to @code{a64l} since
-this function does not modify the buffer pointer.  Every call consumes 6
-characters.
+a call to @code{l64a}.  The function processes at least 6 characters of
+this string, and decodes the characters it finds according to the table
+below.  It stops decoding when it finds a character not in the table,
+rather like @code{atoi}; if you have a buffer which has been broken into
+lines, you must be careful to skip over the end-of-line characters.
+
+The decoded number is returned as a @code{long int} value.
 @end deftypefun
 
+The @code{l64a} and @code{a64l} functions use a base 64 encoding, in
+which each character of an encoded string represents six bits of an
+input word.  These symbols are used for the base 64 digits:
+
+@multitable {xxxxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx} {xxx}
+@item              @tab 0 @tab 1 @tab 2 @tab 3 @tab 4 @tab 5 @tab 6 @tab 7
+@item       0      @tab @code{.} @tab @code{/} @tab @code{0} @tab @code{1}
+                   @tab @code{2} @tab @code{3} @tab @code{4} @tab @code{5}
+@item       8      @tab @code{6} @tab @code{7} @tab @code{8} @tab @code{9}
+                   @tab @code{A} @tab @code{B} @tab @code{C} @tab @code{D}
+@item       16     @tab @code{E} @tab @code{F} @tab @code{G} @tab @code{H}
+                   @tab @code{I} @tab @code{J} @tab @code{K} @tab @code{L}
+@item       24     @tab @code{M} @tab @code{N} @tab @code{O} @tab @code{P}
+                   @tab @code{Q} @tab @code{R} @tab @code{S} @tab @code{T}
+@item       32     @tab @code{U} @tab @code{V} @tab @code{W} @tab @code{X}
+                   @tab @code{Y} @tab @code{Z} @tab @code{a} @tab @code{b}
+@item       40     @tab @code{c} @tab @code{d} @tab @code{e} @tab @code{f}
+                   @tab @code{g} @tab @code{h} @tab @code{i} @tab @code{j}
+@item       48     @tab @code{k} @tab @code{l} @tab @code{m} @tab @code{n}
+                   @tab @code{o} @tab @code{p} @tab @code{q} @tab @code{r}
+@item       56     @tab @code{s} @tab @code{t} @tab @code{u} @tab @code{v}
+                   @tab @code{w} @tab @code{x} @tab @code{y} @tab @code{z}
+@end multitable
+
+This encoding scheme is not standard.  There are some other encoding
+methods which are much more widely used (UU encoding, MIME encoding).
+Generally, it is better to use one of these encodings.
+
 @node Argz and Envz Vectors
 @section Argz and Envz Vectors