Fix PR libfortran/39667

From-SVN: r147004

gcc/testsuite/ChangeLog

+2009-04-30  Janne Blomqvist  <jb@gcc.gnu.org>
+
+	PR libfortran/39667
+	* gfortran.dg/advance_1.f90: Don't require target fd_truncate.
+	* gfortran.dg/advance_4.f90: Likewise
+	* gfortran.dg/advance_5.f90: Likewise
+	* gfortran.dg/append_1.f90: Likewise
+	* gfortran.dg/backslash_1.f90: Likewise
+	* gfortran.dg/backslash_2.f90: Likewise
+	* gfortran.dg/backslash_3.f: Likewise
+	* gfortran.dg/backspace_10.f90: Likewise
+	* gfortran.dg/backspace_3.f: Likewise
+	* gfortran.dg/backspace_4.f: Likewise
+	* gfortran.dg/backspace_5.f: Likewise
+	* gfortran.dg/backspace_8.f: Likewise
+	* gfortran.dg/backspace_9.f: Likewise
+	* gfortran.dg/complex_write.f90: Likewise
+	* gfortran.dg/convert_implied_open.f90: Likewise
+	* gfortran.dg/dollar_edit_descriptor_1.f: Likewise
+	* gfortran.dg/dos_eol.f: Likewise
+	* gfortran.dg/empty_format_1.f90: Likewise
+	* gfortran.dg/endfile.f: Likewise
+	* gfortran.dg/eof_1.f90: Likewise
+	* gfortran.dg/eor_1.f90: Likewise
+	* gfortran.dg/eor_handling_1.f90: Likewise
+	* gfortran.dg/eor_handling_2.f90: Likewise
+	* gfortran.dg/eor_handling_3.f90: Likewise
+	* gfortran.dg/eor_handling_4.f90: Likewise
+	* gfortran.dg/eor_handling_5.f90: Likewise
+	* gfortran.dg/error_recovery_5.f90: Likewise
+	* gfortran.dg/f2003_inquire_1.f03: Likewise
+	* gfortran.dg/f2003_io_4.f03: Likewise
+	* gfortran.dg/f2003_io_5.f03: Likewise
+	* gfortran.dg/f2003_io_7.f03: Likewise
+	* gfortran.dg/fgetc_1.f90: Likewise
+	* gfortran.dg/fgetc_2.f90: Likewise
+	* gfortran.dg/flush_1.f90: Likewise
+	* gfortran.dg/fmt_exhaust.f90: Likewise
+	* gfortran.dg/fmt_huge.f90: Likewise
+	* gfortran.dg/fmt_read.f90: Likewise
+	* gfortran.dg/fmt_t_1.f90: Likewise
+	* gfortran.dg/fmt_t_2.f90: Likewise
+	* gfortran.dg/fmt_t_3.f90: Likewise
+	* gfortran.dg/fmt_t_4.f90: Likewise
+	* gfortran.dg/fmt_t_5.f90: Likewise
+	* gfortran.dg/fmt_t_7.f: Likewise
+	* gfortran.dg/fseek.f90: Likewise
+	* gfortran.dg/ftell_1.f90: Likewise
+	* gfortran.dg/ftell_2.f90: Likewise
+	* gfortran.dg/func_derived_3.f90: Likewise
+	* gfortran.dg/inquire_9.f90: Likewise
+	* gfortran.dg/iostat_1.f90: Likewise
+	* gfortran.dg/iostat_2.f90: Likewise
+	* gfortran.dg/list_read_1.f90: Likewise
+	* gfortran.dg/list_read_4.f90: Likewise
+	* gfortran.dg/list_read_5.f90: Likewise
+	* gfortran.dg/list_read_7.f90: Likewise
+	* gfortran.dg/list_read_8.f90: Likewise
+	* gfortran.dg/list_read_9.f90: Likewise
+	* gfortran.dg/namelist_13.f90: Likewise
+	* gfortran.dg/namelist_14.f90: Likewise
+	* gfortran.dg/namelist_15.f90: Likewise
+	* gfortran.dg/namelist_16.f90: Likewise
+	* gfortran.dg/namelist_17.f90: Likewise
+	* gfortran.dg/namelist_18.f90: Likewise
+	* gfortran.dg/namelist_19.f90: Likewise
+	* gfortran.dg/namelist_20.f90: Likewise
+	* gfortran.dg/namelist_24.f90: Likewise
+	* gfortran.dg/namelist_26.f90: Likewise
+	* gfortran.dg/namelist_27.f90: Likewise
+	* gfortran.dg/namelist_28.f90: Likewise
+	* gfortran.dg/namelist_37.f90: Likewise
+	* gfortran.dg/namelist_38.f90: Likewise
+	* gfortran.dg/namelist_39.f90: Likewise
+	* gfortran.dg/namelist_40.f90: Likewise
+	* gfortran.dg/namelist_43.f90: Likewise
+	* gfortran.dg/namelist_44.f90: Likewise
+	* gfortran.dg/namelist_45.f90: Likewise
+	* gfortran.dg/namelist_46.f90: Likewise
+	* gfortran.dg/namelist_47.f90: Likewise
+	* gfortran.dg/namelist_48.f90: Likewise
+	* gfortran.dg/namelist_49.f90: Likewise
+	* gfortran.dg/namelist_50.f90: Likewise
+	* gfortran.dg/namelist_51.f90: Likewise
+	* gfortran.dg/namelist_52.f90: Likewise
+	* gfortran.dg/namelist_56.f90: Likewise
+	* gfortran.dg/namelist_char_only.f90: Likewise
+	* gfortran.dg/namelist_use.f90: Likewise
+	* gfortran.dg/namelist_use_only.f90: Likewise
+	* gfortran.dg/noadv_size.f90: Likewise
+	* gfortran.dg/open_access_append_1.f90: Likewise
+	* gfortran.dg/pad_no.f90: Likewise
+	* gfortran.dg/pr12884.f: Likewise
+	* gfortran.dg/pr17090.f90: Likewise
+	* gfortran.dg/pr17285.f90: Likewise
+	* gfortran.dg/pr17286.f90: Likewise
+	* gfortran.dg/pr18122.f90: Likewise
+	* gfortran.dg/pr18210.f90: Likewise
+	* gfortran.dg/pr18392.f90: Likewise
+	* gfortran.dg/pr19155.f: Likewise
+	* gfortran.dg/pr19216.f: Likewise
+	* gfortran.dg/pr19467.f90: Likewise
+	* gfortran.dg/pr19657.f: Likewise
+	* gfortran.dg/pr20257.f90: Likewise
+	* gfortran.dg/read_bad_advance.f90: Likewise
+	* gfortran.dg/read_eof_2.f90: Likewise
+	* gfortran.dg/read_eof_4.f90: Likewise
+	* gfortran.dg/read_many_1.f: Likewise
+	* gfortran.dg/read_noadvance.f90: Likewise
+	* gfortran.dg/read_repeat.f90: Likewise
+	* gfortran.dg/read_size_noadvance.f90: Likewise
+	* gfortran.dg/read_x_past.f: Likewise
+	* gfortran.dg/record_marker_1.f90: Likewise
+	* gfortran.dg/record_marker_3.f90: Likewise
+	* gfortran.dg/rewind_1.f90: Likewise
+	* gfortran.dg/runtime_warning_1.f90: Likewise
+	* gfortran.dg/shape_3.f90: Likewise
+	* gfortran.dg/slash_1.f90: Likewise
+	* gfortran.dg/stat_1.f90: Likewise
+	* gfortran.dg/stat_2.f90: Likewise
+	* gfortran.dg/streamio_11.f90: Likewise
+	* gfortran.dg/streamio_3.f90: Likewise
+	* gfortran.dg/streamio_4.f90: Likewise
+	* gfortran.dg/streamio_9.f90: Likewise
+	* gfortran.dg/tl_editing.f90: Likewise
+	* gfortran.dg/unf_io_convert_1.f90: Likewise
+	* gfortran.dg/unf_io_convert_2.f90: Likewise
+	* gfortran.dg/unf_read_corrupted_2.f90: Likewise
+	* gfortran.dg/unf_short_record_1.f90: Likewise
+	* gfortran.dg/utf8_1.f03: Likewise
+	* gfortran.dg/utf8_2.f03: Likewise
+	* gfortran.dg/widechar_IO_1.f90: Likewise
+	* gfortran.dg/write_check3.f90: Likewise
+	* gfortran.dg/write_rewind_2.f: Likewise
+	* gfortran.dg/x_slash_2.f: Likewise
+	
 2009-04-29  Richard Guenther  <rguenther@suse.de>
 
 	PR target/39943

gcc/testsuite/gfortran.dg/advance_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25463 Check that advance='no' works correctly.
 ! Derived from example given in PR by Thomas Koenig
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org> 

gcc/testsuite/gfortran.dg/advance_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31207 Last record truncated for read after short write
 program main
   character(10) :: answer

gcc/testsuite/gfortran.dg/advance_5.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31207 Last record truncated for read after short write.
 character(len=20) :: b
 ! write something no advance

gcc/testsuite/gfortran.dg/append_1.f90

 ! PR libfortran/21471
 ! Testing POSITION="APPEND"
 !
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
       subroutine failed
         close (10,status='delete')
         call abort

gcc/testsuite/gfortran.dg/backslash_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
       character(len=4) a
       open (10, status='scratch')
       write (10,'(A)') '1\n2'

gcc/testsuite/gfortran.dg/backslash_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fbackslash" }
       integer :: i, e
       open (10, status='scratch')

gcc/testsuite/gfortran.dg/backslash_3.f

-C { dg-do run { target fd_truncate } }
+C { dg-do run }
 C { dg-options "-fbackslash" }
 C PR fortran/30278
       program a

gcc/testsuite/gfortran.dg/backspace_10.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33307 I/O read/positioning problem - in BACKSPACE
 ! Test case devloped from test in PR by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 program gfcbug69b

gcc/testsuite/gfortran.dg/backspace_3.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25598 Error on repeated backspaces.
 ! Derived from example given in PR by Dale Ranta
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org> 

gcc/testsuite/gfortran.dg/backspace_4.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25598 Error on repeated backspaces.
 ! Derived from example given in PR by Dale Ranta
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org> 

gcc/testsuite/gfortran.dg/backspace_5.f

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! PR26464 File I/O error related to buffering and BACKSPACE
 ! Test case derived from case by Dale Ranta.
 ! Submitted  by Jerry DeLisle  <jvdelisle@gcc.gnu.org>

gcc/testsuite/gfortran.dg/backspace_8.f

-C { dg-do run { target fd_truncate } }
+C { dg-do run }
 C PR libfortran/31618 - backspace after an error didn't work.
       program main
       character*78 msg

gcc/testsuite/gfortran.dg/backspace_9.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR32235 incorrectly position text file after backspace
 ! Test case from PR, prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
       program main

gcc/testsuite/gfortran.dg/complex_write.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr 19071
 ! test case provided by
 !   Thomas.Koenig@online.de

gcc/testsuite/gfortran.dg/convert_implied_open.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fconvert=swap" }
 ! PR 26735 - implied open didn't use to honor -fconvert
 program main

gcc/testsuite/gfortran.dg/dollar_edit_descriptor_1.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-w" }
 ! PR libfortran/20006
       character*5 c

gcc/testsuite/gfortran.dg/dos_eol.f

 ! PR libfortran/19678 and PR libfortran/19679
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
       integer i, j
       
       open (10,status='scratch')

gcc/testsuite/gfortran.dg/empty_format_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 17709
 ! We weren't resetting the internal EOR flag correctly, so the second read
 ! wasn't advancing to the next line.

gcc/testsuite/gfortran.dg/endfile.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25550 file data corrupted after reading end of file.
 ! Derived from example given in PR from Dale Ranta.
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org> 

gcc/testsuite/gfortran.dg/eof_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Program to test for proper EOF errors when reading past the end of a file.
 ! We used to get this wrong when a formatted read followed a list formatted
 ! read.

gcc/testsuite/gfortran.dg/eor_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 19451: The test for advance='NO' with eor used to be reversed.
 program main
   character*2 c

gcc/testsuite/gfortran.dg/eor_handling_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 17992:  Reading an empty file should yield zero with pad='YES'
 ! (which is the default).
 ! Test case supplied by milan@cmm.ki.si.

gcc/testsuite/gfortran.dg/eor_handling_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 19568:  Don't read across end of line when the format is longer
 !            than the line length and pad='yes' (default)
 program main

gcc/testsuite/gfortran.dg/eor_handling_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 19595:  Handle end-of-record condition with pad=yes (default)
 program main
   integer i1, i2

gcc/testsuite/gfortran.dg/eor_handling_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 20092, 20131:  Handle end-of-record condition with pad=yes (default)
 ! for standard input.  This test case only really tests  anything if,
 ! by changing unit 5, you get to manipulate the standard input.

gcc/testsuite/gfortran.dg/eor_handling_5.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 20661: Handle non-advancing I/O with iostat
 ! Test case by Walt Brainerd, The Fortran Company
 

gcc/testsuite/gfortran.dg/error_recovery_5.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR34411 hang-up during read of non-expected input
 ! Test case derived from that given in PR
 ! Prior to patch, the do loop was infinite, limits set in this one

gcc/testsuite/gfortran.dg/f2003_inquire_1.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-std=gnu" }
 character(25) :: sround, ssign, sasynchronous, sdecimal, sencoding
 integer :: vsize, vid

gcc/testsuite/gfortran.dg/f2003_io_4.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Test case prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 ! Test of decimal= feature
 

gcc/testsuite/gfortran.dg/f2003_io_5.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Test case prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 ! Test of decimal="comma" in namelist and complex
 integer :: i

gcc/testsuite/gfortran.dg/f2003_io_7.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Test case prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 ! Test of sign=, decimal=, and blank= .
 program iotests

gcc/testsuite/gfortran.dg/fgetc_1.f90

 ! Testcase for the FGETC and FPUTC intrinsics
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   character(len=5) s
   integer st
 

gcc/testsuite/gfortran.dg/fgetc_2.f90

 ! Testcase for the FGETC and FPUTC intrinsics
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   character(len=5) s
   integer st
 

gcc/testsuite/gfortran.dg/flush_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 22390 Implement flush statement
 program flush_1
 

gcc/testsuite/gfortran.dg/fmt_exhaust.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR27304 Test running out of data descriptors with data remaining.
 ! Derived from case in PR.  Submitted by Jerry DeLisle <jvdelisle@gcc.gnu.org>.
       program test

gcc/testsuite/gfortran.dg/fmt_huge.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR32446 printing big numbers in F0.1 format.
 ! This segfaulted before the patch.
       open (10, status="scratch")

gcc/testsuite/gfortran.dg/fmt_read.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr18398, missing data on sequential formatted reads
 ! test contributed by Thomas.Koenig@online.de
       open(7,status='scratch')

gcc/testsuite/gfortran.dg/fmt_t_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
       integer nrow, vec(15)
       open (10, status="scratch")
       write (10, fmt='(a)') '001    1 2 3 4 5 6'

gcc/testsuite/gfortran.dg/fmt_t_2.f90

 ! { dg-options "" }
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr24699, handle end-of-record on READ with T format
 ! test contributed by Jerry DeLisle <jvdelisle@gcc.gnu.org>
       character*132 :: foost1, foost2, foost3

gcc/testsuite/gfortran.dg/fmt_t_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31051 bug with x and t format descriptors.
 ! Test case prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org> from PR.
 program t

gcc/testsuite/gfortran.dg/fmt_t_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31199, test case from PR report.
        program write_write
        character(len=20) :: a,b,c

gcc/testsuite/gfortran.dg/fmt_t_5.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR32678 GFortan works incorrectly when writing with FORMAT Tx
 ! Before patch, NULLs were inserted in output.
 ! Test case from reporter enhanced to detect this problem.

gcc/testsuite/gfortran.dg/fmt_t_7.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR34974 null bytes when reverse-tabbing long records
 ! Test case prepared by Jerry DeLisle <jvdelisle@gcc.gnu.org>
        program test

gcc/testsuite/gfortran.dg/fseek.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 
 PROGRAM test_fseek
   INTEGER, PARAMETER :: SEEK_SET = 0, SEEK_CUR = 1, SEEK_END = 2, fd=10

gcc/testsuite/gfortran.dg/ftell_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   integer(kind=8) o, o2
 
   open (10, status="scratch")

gcc/testsuite/gfortran.dg/ftell_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   integer(kind=8) o
   open (10, status="scratch")
   if (ftell(10) /= 0) call abort

gcc/testsuite/gfortran.dg/func_derived_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! This tests the "virtual fix" for PR19561, where pointers to derived
 ! types were not generating correct code.  This testcase is based on
 ! the original PR example.  This example not only tests the

gcc/testsuite/gfortran.dg/inquire_9.f90

 ! PR fortran/24774
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   logical :: l
   l = .true.
   inquire (file='inquire_9 file that should not exist', exist=l)

gcc/testsuite/gfortran.dg/iostat_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 23598 - The iostat variable wasn't reset if the previous
 ! I/O library call had an error.
 program main

gcc/testsuite/gfortran.dg/iostat_2.f90

 ! PR libfortran/23784
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   integer i
   close(10, status="whatever", iostat=i) ! { dg-warning "STATUS specifier in CLOSE statement.*has invalid value" }
   if (i == 0) call abort()

gcc/testsuite/gfortran.dg/list_read_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Program to test terminators in list-directed input
 program list_read_1
   character(len=5) :: s

gcc/testsuite/gfortran.dg/list_read_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Test of gfortran list directed read> check delimiters are correctly
 ! treated.  Written in f77 so that g77 will run for comparison.
 !

gcc/testsuite/gfortran.dg/list_read_5.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25307 Check handling of end-of-file conditions for list directed reads.
 ! Prepared by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 program pr25307

gcc/testsuite/gfortran.dg/list_read_7.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33400 Formatted read fails if line ends without line break
 ! Test case modified from that in PR by <jvdelisle@gcc.gnu.org>
 integer, parameter :: fgsl_strmax = 128

gcc/testsuite/gfortran.dg/list_read_8.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR34676 IO error delayed
 ! Test case from PR modified by <jvdelisle@gcc.gnu.org>
 implicit none

gcc/testsuite/gfortran.dg/list_read_9.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr37083 formatted read of line without trailing new-line fails
 real :: a, b, c
 open(unit=10,file="atest",access='stream',form='unformatted',&

gcc/testsuite/gfortran.dg/namelist_13.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests simple derived types.
 ! Provided by Paul Thomas - pault@gcc.gnu.org
 

gcc/testsuite/gfortran.dg/namelist_14.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests various combinations of intrinsic types, derived types, arrays,
 ! dummy arguments and common to check nml_get_addr_expr in trans-io.c.
 ! See comments below for selection.

gcc/testsuite/gfortran.dg/namelist_15.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests arrays of derived types containing derived type arrays whose
 ! components are character arrays - exercises object name parser in
 ! list_read.c. Checks that namelist output can be reread. 

gcc/testsuite/gfortran.dg/namelist_16.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests namelist on complex variables
 ! provided by Paul Thomas - pault@gcc.gnu.org
 program namelist_16

gcc/testsuite/gfortran.dg/namelist_17.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests namelist on logical variables
 ! provided by Paul Thomas - pault@gcc.gnu.org
 

gcc/testsuite/gfortran.dg/namelist_18.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests character delimiters for namelist write 
 ! provided by Paul Thomas - pault@gcc.gnu.org
 

gcc/testsuite/gfortran.dg/namelist_19.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Test namelist error trapping.
 ! provided by Paul Thomas - pault@gcc.gnu.org
 

gcc/testsuite/gfortran.dg/namelist_20.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! Tests namelist io for an explicit shape array with negative bounds
 ! provided by Paul Thomas - pault@gcc.gnu.org
 

gcc/testsuite/gfortran.dg/namelist_24.f90

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 !{ dg-options -std=gnu }
 ! Tests namelist read when more data is provided then specified by 
 ! array qualifier in list.

gcc/testsuite/gfortran.dg/namelist_26.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR30918 Failure to skip commented out NAMELIST
 ! Before the patch, this read the commented out namelist and iuse would
 ! equal 2 when done.  Test case from PR.

gcc/testsuite/gfortran.dg/namelist_27.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31052 Bad IOSTAT values when readings NAMELISTs past EOF.
 ! Patch derived from PR, submitted by Jerry DeLisle <jvdelisle@gcc.gnu.org>
 program gfcbug61

gcc/testsuite/gfortran.dg/namelist_28.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31052 Bad IOSTAT values when readings NAMELISTs past EOF.
 ! Patch derived from PR, submitted by Jerry DeLisle <jvdelisle@gcc.gnu.org>
 program gfcbug61

gcc/testsuite/gfortran.dg/namelist_37.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33039 Read NAMELIST:  reads wrong namelist name
 ! Test case from PR modified by Jerry DeLisle <jvdelisle@gcc.gnu.org>
 PROGRAM namelist

gcc/testsuite/gfortran.dg/namelist_38.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33253 namelist: reading back a string, also fixed writing with delimiters.
 ! Test case modified from that of the PR by
 ! Jerry DeLisle  <jvdelisle@gcc.gnu.org>

gcc/testsuite/gfortran.dg/namelist_39.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33421 and PR33253 Weird quotation of namelist output of character arrays
 ! Test case from Toon Moone, adapted by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 

gcc/testsuite/gfortran.dg/namelist_40.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR33672 Additional runtime checks needed for namelist reads
 ! Submitted by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 

gcc/testsuite/gfortran.dg/namelist_43.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-mieee" { target alpha*-*-* sh*-*-* } }
 ! { dg-skip-if "NaN not supported" { spu-*-* } { "*" } { "" } }
 !

gcc/testsuite/gfortran.dg/namelist_44.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 !
 ! PR fortran/34530
 !

gcc/testsuite/gfortran.dg/namelist_45.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR35617 read namelist error with '!'
 program test
   character(len=128) :: mhdpath

gcc/testsuite/gfortran.dg/namelist_46.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR35627 Namelist read problem with short logical followed by read real
 program test
   implicit none

gcc/testsuite/gfortran.dg/namelist_47.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 
 module nml_47
   type             ::  mt

gcc/testsuite/gfortran.dg/namelist_48.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fbackslash" }
 ! PR36538 namelist failure with tabs preceding object name
       program check1

gcc/testsuite/gfortran.dg/namelist_49.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fbackslash" }
 ! PR36546 Namelist error with tab following a comma and newline
       program check1

gcc/testsuite/gfortran.dg/namelist_50.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR36657 Namelist string constant immediately followed by comment
 program gfcbug79
   implicit none

gcc/testsuite/gfortran.dg/namelist_51.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR36676 Namelist comment problems
 ! test case from PR, reduced by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 program mem_nml

gcc/testsuite/gfortran.dg/namelist_52.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR36582 Namelist I/O error: Bogus "Cannot match namelist object"
 ! Test case derived from PR.
 module mod1

gcc/testsuite/gfortran.dg/namelist_56.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR37707 Namelist read of array of derived type incorrect
 ! Test case from Tobias Burnus
   IMPLICIT NONE

gcc/testsuite/gfortran.dg/namelist_char_only.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-O0" }
 ! Test patch for PR24416.f90 - a used to come back from the read with var
 ! prepended.

gcc/testsuite/gfortran.dg/namelist_use.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! This tests the fix for PR22010, where namelists were not being written to
 ! and read back from modules.  It has two namelists: one that is USE
 ! associated and another that is concatenated by USE and host association.

gcc/testsuite/gfortran.dg/namelist_use_only.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! This tests the fix for PR22010, where namelists were not being written to
 ! and read back from modules.  It checks that namelists from modules that are
 ! selected by an ONLY declaration work correctly, even when the variables in

gcc/testsuite/gfortran.dg/noadv_size.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 20774: Handle size parameter for non-advancing I/O correctly
 program main
   open(77,status='scratch')

gcc/testsuite/gfortran.dg/open_access_append_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Testcase for the GNU extension OPEN(...,ACCESS="APPEND")
   open (10,file="foo")
   close (10,status="delete")

gcc/testsuite/gfortran.dg/pad_no.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Test correct operation for pad='no'.
 program main
   character(len=1) line(2)

gcc/testsuite/gfortran.dg/pr12884.f

-c { dg-do run { target fd_truncate } }
+c { dg-do run }
 c pr 12884
 c test namelist with input file containg / before namelist. Also checks
 c non-standard use of $ instead of &

gcc/testsuite/gfortran.dg/pr17090.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr 17090 Runtime I/O error
 ! bdavis9659@comcast.net 
 ! 9/12/2004

gcc/testsuite/gfortran.dg/pr17285.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr 17285
 ! Test that namelist can read its own output.
 ! At the same time, check arrays and different terminations

gcc/testsuite/gfortran.dg/pr17286.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR17286
 ! Namelist read failed when spaces exist between the '=' and the numbers
 ! This is a libgfortran bug

gcc/testsuite/gfortran.dg/pr18122.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! test namelist with scalars and arrays.
 ! Based on example provided by thomas.koenig@online.de
 

gcc/testsuite/gfortran.dg/pr18210.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Names in upper case and object names starting column 2
 ! Based on example provided by thomas.koenig@online.de
 

gcc/testsuite/gfortran.dg/pr18392.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr 18392
 ! test namelist with derived types
 ! Based on example provided by thomas.koenig@online.de

gcc/testsuite/gfortran.dg/pr19155.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 !
 ! PR libfortran/19155
 ! We accept 'E+00' as a valid real number. The standard says it is not,

gcc/testsuite/gfortran.dg/pr19216.f

 ! PR libfortran/19216
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
        integer dat(3), i, j
        data dat / 3,2,1 /
 

gcc/testsuite/gfortran.dg/pr19467.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr 19467
 ! test namelist with character arrays
 ! Based on example provided by paulthomas2@wanadoo.fr

gcc/testsuite/gfortran.dg/pr19657.f

-c { dg-do run { target fd_truncate } }
+c { dg-do run }
 c pr 19657
 c test namelist not skipped if ending with logical.
 c Based on example provided by fuyuki@ccsr.u-tokyo.ac.jp

gcc/testsuite/gfortran.dg/pr20257.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-skip-if "Too big for local store" { spu-*-* } { "*" } { "" } }
   integer,parameter :: n = 10000
   real(8) array(10000)

gcc/testsuite/gfortran.dg/read_bad_advance.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR27138 Failure to advance line on bad list directed read.
 ! Submitted by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
       program test

gcc/testsuite/gfortran.dg/read_eof_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25835 Check that reading from a file that is at end-of-file does not
 ! segfault or give error.  Test case derived from example in PR from Dale Ranta.
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org>

gcc/testsuite/gfortran.dg/read_eof_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 27575 and PR 30009: This test checks the error checking for end
 ! of file condition.
 ! Derived from test case in PR.

gcc/testsuite/gfortran.dg/read_many_1.f

-!{ dg-do run { target fd_truncate } }
+!{ dg-do run }
 ! PR26423 Large file I/O error related to buffering
 ! Test case derived from case by Dale Ranta.
 ! Submitted  by Jerry DeLisle  <jvdelisle@gcc.gnu.org> 

gcc/testsuite/gfortran.dg/read_noadvance.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! pr24719, non-advancing read should read more than one line
 ! test contributed by jerry delisle <jvdelisle@gcc.gnu.org>
       implicit none

gcc/testsuite/gfortran.dg/read_repeat.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR39528 repeated entries not read when using list-directed input.
 ! Test case derived from reporters example.
 program rread

gcc/testsuite/gfortran.dg/read_size_noadvance.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR26890 Test for use of SIZE variable in IO list.
 ! Test case from Paul Thomas.
 ! Submitted by Jerry DeLisle <jvdelisle@gcc.gnu.org>

gcc/testsuite/gfortran.dg/read_x_past.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options -w }
 ! PR 26661 : Test reading X's past file end with no LF or CR.
 ! PR 26880 : Tests that rewind clears the gfc_unit read_bad flag.

gcc/testsuite/gfortran.dg/record_marker_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-frecord-marker=4" }
 
 program main

gcc/testsuite/gfortran.dg/record_marker_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-frecord-marker=8" }
 
 program main

gcc/testsuite/gfortran.dg/rewind_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Check that rewind doesn't delete a file.
 ! Writing to the file truncates it at the end of the current record.  Out
 ! IO library was defering the actual truncation until the file was rewound.

gcc/testsuite/gfortran.dg/runtime_warning_1.f90

@@ -3,7 +3,7 @@
 ! Contributor Francois-Xavier Coudert  <coudert@clipper.ens.fr>
 !
 ! { dg-options "-pedantic" }
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 !
      character*5 c
      open (42,status='scratch')

gcc/testsuite/gfortran.dg/shape_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 34980 - we got a segfault for calling shape
 !            with a scalar.
 program main

gcc/testsuite/gfortran.dg/slash_1.f90

 ! PR libfortran/22170
-! { dg-do run { target fd_truncate } }
+! { dg-do run }
   integer i
   open (10,status='scratch')
   write (10,'(A,2/,A)') '12', '17'

gcc/testsuite/gfortran.dg/stat_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-skip-if "" { *-*-mingw* spu-*-* } { "*" } { "" } }
 ! { dg-options "-std=gnu" }
   character(len=*), parameter :: f = "testfile"

gcc/testsuite/gfortran.dg/stat_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-skip-if "" { *-*-mingw* spu-*-* } { "*" } { "" } }
 ! { dg-options "-std=gnu" }
   character(len=*), parameter :: f = "testfile"

gcc/testsuite/gfortran.dg/streamio_11.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR29277 Stream IO test 11, tests formatted form.
 ! Contributed by Tobias Burnas.
 program stream_test

gcc/testsuite/gfortran.dg/streamio_3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25828 Stream IO test 3, tests read_x and inquire.
 ! Contributed by Jerry DeLisle <jvdelisle@verizon.net>.
 program streamio_3

gcc/testsuite/gfortran.dg/streamio_4.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR25828 Stream IO test 4, Tests string read and writes, single byte.
 ! Verifies buffering is working correctly and position="append"
 ! Contributed by Jerry DeLisle <jvdelisle@verizon.net>.

gcc/testsuite/gfortran.dg/streamio_9.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR29053 Stream IO test 9.
 ! Contributed by Jerry DeLisle <jvdelisle@gcc.gnu.org>.
 ! Test case derived from that given in PR by Steve Kargl.

gcc/testsuite/gfortran.dg/tl_editing.f90

-! { dg-do run { target fd_truncate } }     
+! { dg-do run }     
 ! Test of fix to bug triggered by NIST fm908.for.
 ! Left tabbing, followed by X or T-tabbing to the right would
 ! cause spaces to be overwritten on output data.

gcc/testsuite/gfortran.dg/unf_io_convert_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-pedantic" }
 !  This test verifies the most basic sequential unformatted I/O
 !  with convert="swap".

gcc/testsuite/gfortran.dg/unf_io_convert_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 program main
   complex(kind=4) :: c
   real(kind=4) :: a(2)

gcc/testsuite/gfortran.dg/unf_read_corrupted_2.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR31880 silent data corruption in gfortran read statement
 ! Test from PR.
       program r3

gcc/testsuite/gfortran.dg/unf_short_record_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 29627 - partial reads of unformatted records
 program main
   character a(3)

gcc/testsuite/gfortran.dg/utf8_1.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fbackslash" }
 ! Contributed by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 program test1

gcc/testsuite/gfortran.dg/utf8_2.f03

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! { dg-options "-fbackslash" }
 ! Contributed by Tobias Burnus
 program test2

gcc/testsuite/gfortran.dg/widechar_IO_1.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! Wide chracter I/O test 1, formatted and mixed kind
 ! Test case developed by Jerry DeLisle <jvdelisle@gcc.gnu.org>
 program test1

gcc/testsuite/gfortran.dg/write_check3.f90

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR29936 Missed constraint on RECL=specifier in unformatted sequential WRITE
 ! Submitted by Jerry DeLisle  <jvdelisle@gcc.gnu.org>
 program us_recl

gcc/testsuite/gfortran.dg/write_rewind_2.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 26499  Test write with rewind sequences to make sure buffering and
 ! end-of-file conditions are handled correctly.  Derived from test case by Dale
 ! Ranta.  Submitted by Jerry DeLisle <jvdelisle@gcc.gnu.org>.

gcc/testsuite/gfortran.dg/x_slash_2.f

-! { dg-do run { target fd_truncate } }
+! { dg-do run }
 ! PR 34887 - reverse tabs followed by a slash used to confuse I/O.
       program main
       character(len=2) :: b, a

libgfortran/ChangeLog

+2009-04-30  Janne Blomqvist  <jb@gcc.gnu.org>
+
+	PR libfortran/39667
+	* io/file_pos.c (st_rewind): Don't truncate or flush.
+	* io/intrinsics.c (fgetc): Flush if switching mode.
+	(fputc): Likewise.
+
 2009-04-18  Janne Blomqvist  <jb@gcc.gnu.org>
 
 	PR libfortran/39782

libgfortran/io/file_pos.c

@@ -341,26 +341,8 @@ st_rewind (st_parameter_filepos *fpp)
 
 	  u->previous_nonadvancing_write = 0;
 
-	  /* Flush the buffers.  If we have been writing to the file, the last
-	       written record is the last record in the file, so truncate the
-	       file now.  Reset to read mode so two consecutive rewind
-	       statements do not delete the file contents.  */
-	  if (u->mode == WRITING)
-	    {
-	      /* unit_truncate takes care of flushing.  */
-	      unit_truncate (u, stell (u->s), &fpp->common);
-	      /* .. but we still need to reset since we're going to seek.  */
 	  fbuf_reset (u);
-	    }
-          else
-            {
-              /* Make sure buffers are reset.  */
-              if (u->flags.form == FORM_FORMATTED)
-                fbuf_reset (u);
-              sflush (u->s);
-            }              
 
-	  u->mode = READING;
 	  u->last_record = 0;
 
 	  if (sseek (u->s, 0, SEEK_SET) < 0)

libgfortran/io/intrinsics.c

@@ -46,6 +46,13 @@ PREFIX(fgetc) (const int * unit, char * c, gfc_charlen_type c_len)
   if (u == NULL)
     return -1;
 
+  fbuf_reset (u);
+  if (u->mode == WRITING)
+    {
+      sflush (u->s);
+      u->mode = READING;
+    }
+
   memset (c, ' ', c_len);
   ret = sread (u->s, c, 1);
   unlock_unit (u);
@@ -118,6 +125,13 @@ PREFIX(fputc) (const int * unit, char * c,
   if (u == NULL)
     return -1;
 
+  fbuf_reset (u);
+  if (u->mode == READING)
+    {
+      sflush (u->s);
+      u->mode = WRITING;
+    }
+
   s = swrite (u->s, c, 1);
   unlock_unit (u);
   if (s < 0)