libio fflush() bug

[Thorsten Kukuk <kukuk@weber-eb.uni-paderborn.de>]

Hello,

I have found a big problem. If you do fseek/fread/fflush, the file pointer
is set to the end of the file. If you don't call fseek, the pointer is set
correct to the next byte after the last read one.

I have attached a test program and a patch against libio/fileops.c. Could
somebody, who knows more about the internal, check, if my patch is correct ?

  Thorsten

--
Thorsten Kukuk  kukuk@vt.uni-paderborn.de
                http://www-vt.uni-paderborn.de/~kukuk
Linux is like a Vorlon.  It is incredibly powerful, gives terse,
cryptic answers and has a lot of things going on in the background.
#include <stdio.h>

int
main (void)
{
  FILE *fp;
  long entry = 0;
  int before, after;

  fp = fopen ("fflush_bug.c", "r+");

  fseek (fp, 0, SEEK_SET);

  fread (&entry, 4, 1, fp);

  before = ftell (fp);

  fflush (fp);

  after = ftell (fp);

  fclose (fp);

  printf ("ftell() before fflush(): %d, after fflush(): %d\n", before, after);

  if (before == after)
    return 0;
  else
    return 1;
}
1998-08-07  Thorsten Kukuk  <kukuk@vt.uni-paderborn.de>

	* libio/fileops.c (_IO_file_sync): Fix file pointer position.

diff -u glibc-2.0.95/libio/fileops.c libc-work/libio/fileops.c
--- glibc-2.0.95/libio/fileops.c	Thu Jul 23 07:43:58 1998
+++ libc-work/libio/fileops.c	Fri Aug  7 19:01:40 1998
@@ -437,7 +437,7 @@
   /*    char* ptr = cur_ptr(); */
   if (fp->_IO_write_ptr > fp->_IO_write_base)
     if (_IO_do_flush(fp)) return EOF;
-  delta = fp->_IO_read_ptr - fp->_IO_read_end;
+  delta = fp->_IO_read_base - fp->_IO_read_ptr;
   if (delta != 0)
     {
 #ifdef TODO
@@ -446,7 +446,7 @@
 #endif
       _IO_off64_t new_pos = _IO_SYSSEEK (fp, delta, 1);
       if (new_pos != (_IO_off64_t) EOF)
-	fp->_IO_read_end = fp->_IO_read_ptr;
+	fp->_IO_read_base = fp->_IO_read_ptr;
 #ifdef ESPIPE
       else if (errno == ESPIPE)
 	; /* Ignore error from unseekable devices. */