cppfiles.c (initialize_input_buffer): New function.

1999-05-10 18:21 -0400  Zack Weinberg  <zack@rabi.columbia.edu>
	* cppfiles.c (initialize_input_buffer): New function.
	(finclude): Call it, if pfile->input_buffer is NULL.  Accept
	any character device as an input file.
	(read_and_prescan): Use pfile->input_buffer and
	pfile->input_speccase.
	* cppinit.c (cpp_cleanup): Free pfile->input_buffer and
	pfile->input_speccase.
	* cpplib.h (cpp_reader): Add input_buffer, input_speccase, and
	input_buffer_len members.  Use memcpy in CPP_PUTS_Q.
	* cppmain.c: Buffer output in the token_buffer; throttle
	number of calls to fwrite; check for errors from fwrite.

From-SVN: r26869

gcc/cppfiles.c

@@ -45,6 +45,9 @@ static long read_and_prescan		PROTO ((cpp_reader *, cpp_buffer *,
 						int, size_t));
 static struct file_name_list *actual_directory PROTO ((cpp_reader *, char *));
 
+static void initialize_input_buffer	PROTO ((cpp_reader *, int,
+						struct stat *));
+
 #if 0
 static void hack_vms_include_specification PROTO ((char *));
 #endif
@@ -648,10 +651,12 @@ finclude (pfile, fd, ihash)
 	}
     }
   else if (S_ISFIFO (st.st_mode) || S_ISSOCK (st.st_mode)
+	   /* Permit any kind of character device: the sensible ones are
+	      ttys and /dev/null, but weeding out the others is too hard.  */
+	   || S_ISCHR (st.st_mode)
 	   /* Some 4.x (x<4) derivatives have a bug that makes fstat() of a
 	      socket or pipe return a stat struct with most fields zeroed.  */
-	   || (st.st_mode == 0 && st.st_nlink == 0 && st.st_size == 0)
+	   || (st.st_mode == 0 && st.st_nlink == 0 && st.st_size == 0))
-	   || (S_ISCHR (st.st_mode) && isatty (fd)))
     {
       /* Cannot get its file size before reading.  4k is a decent
          first guess. */
@@ -663,6 +668,9 @@ finclude (pfile, fd, ihash)
       goto fail;
     }
 
+  if (pfile->input_buffer == NULL)
+    initialize_input_buffer (pfile, fd, &st);
+
   /* Read the file, converting end-of-line characters and trigraphs
      (if enabled). */
   fp->ihash = ihash;
@@ -758,8 +766,7 @@ actual_directory (pfile, fname)
   return x;
 }
 
-/* Almost but not quite the same as adjust_position in cpplib.c.
+/* Determine the current line and column.  Used only by read_and_prescan. */
-   Used only by read_and_prescan. */
 static void
 find_position (start, limit, linep, colp)
      U_CHAR *start;
@@ -804,9 +811,14 @@ find_position (start, limit, linep, colp)
    If your file has more than one kind of end-of-line marker, you
    will get messed-up line numbering.  */
 
-#ifndef PIPE_BUF
+/* Table of characters that can't be handled in the inner loop.
-#define PIPE_BUF 4096
+   Keep these contiguous to optimize the performance of the code generated
-#endif
+   for the switch that uses them.  */
+#define SPECCASE_EMPTY     0
+#define SPECCASE_NUL       1
+#define SPECCASE_CR        2
+#define SPECCASE_BACKSLASH 3
+#define SPECCASE_QUESTION  4
 
 static long
 read_and_prescan (pfile, fp, desc, len)
@@ -818,45 +830,24 @@ read_and_prescan (pfile, fp, desc, len)
   U_CHAR *buf = (U_CHAR *) xmalloc (len);
   U_CHAR *ip, *op, *line_base;
   U_CHAR *ibase;
+  U_CHAR *speccase = pfile->input_speccase;
   unsigned long line;
   unsigned int deferred_newlines;
   int count;
   size_t offset;
-  /* PIPE_BUF bytes of buffer proper, 2 to detect running off the end
-     without address arithmetic all the time, and 2 for pushback in
-     the case there's a potential trigraph or end-of-line digraph at
-     the end of a block. */
-  U_CHAR intermed[PIPE_BUF + 2 + 2];
-
-  /* Table of characters that can't be handled in the inner loop.
-     Keep these contiguous to optimize the performance of the code generated
-     for the switch that uses them.  */
-#define SPECCASE_EMPTY     0
-#define SPECCASE_NUL       1
-#define SPECCASE_CR        2
-#define SPECCASE_BACKSLASH 3
-#define SPECCASE_QUESTION  4
-  U_CHAR speccase[256];
 
   offset = 0;
   op = buf;
   line_base = buf;
   line = 1;
-  ibase = intermed + 2;
+  ibase = pfile->input_buffer + 2;
   deferred_newlines = 0;
 
-  memset (speccase, SPECCASE_EMPTY, sizeof (speccase));
-  speccase['\0'] = SPECCASE_NUL;
-  speccase['\r'] = SPECCASE_CR;
-  speccase['\\'] = SPECCASE_BACKSLASH;
-  if (CPP_OPTIONS (pfile)->trigraphs || CPP_OPTIONS (pfile)->warn_trigraphs)
-    speccase['?'] = SPECCASE_QUESTION;
-
   for (;;)
     {
     read_next:
 
-      count = read (desc, intermed + 2, PIPE_BUF);
+      count = read (desc, pfile->input_buffer + 2, pfile->input_buffer_len);
       if (count < 0)
 	goto error;
       else if (count == 0)
@@ -864,7 +855,7 @@ read_and_prescan (pfile, fp, desc, len)
 
       offset += count;
       ip = ibase;
-      ibase = intermed + 2;
+      ibase = pfile->input_buffer + 2;
       ibase[count] = ibase[count+1] = '\0';
 
       if (offset > len)
@@ -924,8 +915,7 @@ read_and_prescan (pfile, fp, desc, len)
 		ip++;
 	      else if (*ip == '\0')
 		{
-		  --ibase;
+		  *--ibase = '\r';
-		  intermed[1] = '\r';
 		  goto read_next;
 		}
 	      else if (ip[-2] == '\n')
@@ -941,8 +931,7 @@ read_and_prescan (pfile, fp, desc, len)
 		 and come back next pass. */
 	      if (*ip == '\0')
 		{
-		  --ibase;
+		  *--ibase = '\\';
-		  intermed[1] = '\\';
 		  goto read_next;
 		}
 	      else if (*ip == '\n')
@@ -965,9 +954,8 @@ read_and_prescan (pfile, fp, desc, len)
 		  if (*ip == '\n') ip++;
 		  else if (*ip == '\0')
 		    {
-		      ibase -= 2;
+		      *--ibase = '\r';
-		      intermed[0] = '\\';
+		      *--ibase = '\\';
-		      intermed[1] = '\r';
 		      goto read_next;
 		    }
 		  else if (*ip == '\r' || *ip == '\t' || *ip == ' ')
@@ -991,8 +979,7 @@ read_and_prescan (pfile, fp, desc, len)
 		d = ip[0];
 		if (d == '\0')
 		  {
-		    --ibase;
+		    *--ibase = '?';
-		    intermed[1] = '?';
 		    goto read_next;
 		  }
 		if (d != '?')
@@ -1003,8 +990,8 @@ read_and_prescan (pfile, fp, desc, len)
 		d = ip[1];
 		if (d == '\0')
 		  {
-		    ibase -= 2;
+		    *--ibase = '?';
-		    intermed[0] = intermed[1] = '?';
+		    *--ibase = '?';
 		    goto read_next;
 		  }
 		if (!trigraph_table[d])
@@ -1047,7 +1034,7 @@ read_and_prescan (pfile, fp, desc, len)
      This may be any of:  ?? ? \ \r \n \\r \\n.
      \r must become \n, \\r or \\n must become \r.
      We know we have space already. */
-  if (ibase == intermed)
+  if (ibase == pfile->input_buffer)
     {
       if (*ibase == '?')
 	{
@@ -1057,7 +1044,7 @@ read_and_prescan (pfile, fp, desc, len)
       else
 	*op++ = '\r';
     }
-  else if (ibase == intermed + 1)
+  else if (ibase == pfile->input_buffer + 1)
     {
       if (*ibase == '\r')
 	*op++ = '\n';
@@ -1095,6 +1082,67 @@ read_and_prescan (pfile, fp, desc, len)
   return -1;
 }
 
+/* Initialize the `input_buffer' and `input_speccase' tables.
+   These are only used by read_and_prescan, but they're large and
+   somewhat expensive to set up, so we want them allocated once for
+   the duration of the cpp run.  */
+
+static void
+initialize_input_buffer (pfile, fd, st)
+     cpp_reader *pfile;
+     int fd;
+     struct stat *st;
+{
+  long pipe_buf;
+  U_CHAR *tmp;
+
+  /* Table of characters that cannot be handled by the
+     read_and_prescan inner loop.  The number of non-EMPTY entries
+     should be as small as humanly possible.  */
+
+  tmp = xmalloc (1 << CHAR_BIT);
+  memset (tmp, SPECCASE_EMPTY, 1 << CHAR_BIT);
+  tmp['\0'] = SPECCASE_NUL;
+  tmp['\r'] = SPECCASE_CR;
+  tmp['\\'] = SPECCASE_BACKSLASH;
+  if (CPP_OPTIONS (pfile)->trigraphs || CPP_OPTIONS (pfile)->warn_trigraphs)
+    tmp['?'] = SPECCASE_QUESTION;
+
+  pfile->input_speccase = tmp;
+
+  /* Determine the appropriate size for the input buffer.  Normal C
+     source files are smaller than eight K.  If we are reading a pipe,
+     we want to make sure the input buffer is bigger than the kernel's
+     pipe buffer.  */
+  pipe_buf = -1;
+
+  if (! S_ISREG (st->st_mode))
+    {
+#ifdef _PC_PIPE_BUF
+      pipe_buf = fpathconf (fd, _PC_PIPE_BUF);
+#endif
+      if (pipe_buf == -1)
+	{
+#ifdef PIPE_BUF
+	  pipe_buf = PIPE_BUF;
+#else
+	  pipe_buf = 8192;
+#endif
+	}
+    }
+
+  if (pipe_buf < 8192)
+    pipe_buf = 8192;
+  /* PIPE_BUF bytes of buffer proper, 2 to detect running off the end
+     without address arithmetic all the time, and 2 for pushback in
+     the case there's a potential trigraph or end-of-line digraph at
+     the end of a block. */
+
+  tmp = xmalloc (pipe_buf + 2 + 2);
+  pfile->input_buffer = tmp;
+  pfile->input_buffer_len = pipe_buf;
+}
+
 /* Add output to `deps_buffer' for the -M switch.
    STRING points to the text to be output.
    SPACER is ':' for targets, ' ' for dependencies, zero for text

gcc/cppinit.c

@@ -569,6 +569,14 @@ cpp_cleanup (pfile)
       pfile->deps_allocated_size = 0;
     }
 
+  if (pfile->input_buffer)
+    {
+      free (pfile->input_buffer);
+      free (pfile->input_speccase);
+      pfile->input_buffer = pfile->input_speccase = NULL;
+      pfile->input_buffer_len = 0;
+    }
+
   while (pfile->if_stack)
     {
       IF_STACK_FRAME *temp = pfile->if_stack;

gcc/cpplib.h

@@ -241,6 +241,13 @@ struct cpp_reader
 
   /* Number of bytes since the last newline.  */
   int deps_column;
+
+  /* A buffer and a table, used only by read_and_prescan (in cppfiles.c)
+     which are allocated once per cpp_reader object to keep them off the
+     stack and avoid setup costs.  */
+  U_CHAR *input_buffer;
+  U_CHAR *input_speccase;
+  size_t input_buffer_len;
 };
 
 #define CPP_FATAL_LIMIT 1000
@@ -269,7 +276,7 @@ struct cpp_reader
 /* Append string STR (of length N) to PFILE's output buffer.
    Assume there is enough space. */
 #define CPP_PUTS_Q(PFILE, STR, N) \
-  (bcopy (STR, (PFILE)->limit, (N)), (PFILE)->limit += (N))
+  (memcpy ((PFILE)->limit, STR, (N)), (PFILE)->limit += (N))
 /* Append string STR (of length N) to PFILE's output buffer.  Make space. */
 #define CPP_PUTS(PFILE, STR, N) CPP_RESERVE(PFILE, N), CPP_PUTS_Q(PFILE, STR,N)
 /* Append character CH to PFILE's output buffer.  Assume sufficient space. */

gcc/cppmain.c

@@ -46,6 +46,7 @@ main (argc, argv)
   char *p;
   int argi = 1;  /* Next argument to handle.  */
   struct cpp_options *opts = &options;
+  enum cpp_token kind;
 
   p = argv[0] + strlen (argv[0]);
   while (p != argv[0] && p[-1] != '/') --p;
@@ -80,21 +81,30 @@ main (argc, argv)
   else if (! freopen (opts->out_fname, "w", stdout))
     cpp_pfatal_with_name (&parse_in, opts->out_fname);
 
-  for (;;)
+  do
     {
-      enum cpp_token kind;
-      if (! opts->no_output)
-	{
-	  fwrite (parse_in.token_buffer, 1, CPP_WRITTEN (&parse_in), stdout);
-	}
-      CPP_SET_WRITTEN (&parse_in, 0);
       kind = cpp_get_token (&parse_in);
-      if (kind == CPP_EOF)
+      if (CPP_WRITTEN (&parse_in) >= BUFSIZ || kind == CPP_EOF)
-	break;
+	{
+	  if (! opts->no_output)
+	    {
+	      size_t rem, count = CPP_WRITTEN (&parse_in);
+
+	      rem = fwrite (parse_in.token_buffer, 1, count, stdout);
+	      if (rem < count)
+		/* Write error. */
+		cpp_pfatal_with_name (&parse_in, opts->out_fname);
+	    }
+
+	  CPP_SET_WRITTEN (&parse_in, 0);
+	}
     }
+  while (kind != CPP_EOF);
 
   cpp_finish (&parse_in);
-  fwrite (parse_in.token_buffer, 1, CPP_WRITTEN (&parse_in), stdout);
+  if (fwrite (parse_in.token_buffer, 1, CPP_WRITTEN (&parse_in), stdout)
+      < CPP_WRITTEN (&parse_in))
+    cpp_pfatal_with_name (&parse_in, opts->out_fname);
 
   if (parse_in.errors)
     exit (FATAL_EXIT_CODE);