[Bug fortran/34411] New: hang-up during read of non-expected input

[Bug fortran/34411] New: hang-up during read of non-expected input

[hroch at physics dot muni dot cz]

I gets hang-up of the fragment of code during execution:

program zk
  real :: x,y
  do
     read(*,'(T7,2F9.3)',iostat=ii,end=666) x,y
     if( iostat == 0 ) write(*,*) x,y
  end do
666 continue
end program zk

with this data on input:

 NL   NX   NY  LOWBAD HIGHBAD  THRESH     AP1  PH/ADU  RNOISE
  2  765  510 1.26E+3 6.50E+4  156.96    2.00    2.30   15.00

   289  329.142  214.107   12.313   12.050   11.913   11.868   11.834   11.722  
     2038.497 99.99  0.00   0.019    0.021    0.025    0.034    0.045    0.055  

   413  360.334  245.261   12.375   11.910   11.469   11.086   10.719   10.327  
     2596.395 99.99  0.00   0.019    0.017    0.016    0.015    0.015    0.014  

   655  332.704  317.964   12.523   12.212   11.998   11.892   11.819   11.711  
     1627.586 99.99  0.00   0.005    0.005    0.006    0.007    0.009    0.011  

   360  379.769  231.226   12.709   12.422   12.195   11.941   11.644   11.324  
     2561.539 99.99  0.00   0.042    0.043    0.050    0.055    0.057    0.057  

 ------
The reason of this is read only lines with float numbers on appropriate
positions. For example, the first line to read is '289  329.142  214.107...'.

The hang-up has been verified under various recent linux distributions (Ubuntu,
Gentoo (gcc version 4.1.2), ... ), architectures (amd64, i686) and under last
night-build (gcc version 4.3.0 20071209 (experimental) [trunk revision 130717]
(GCC)).

The code is correctly compiled without any warnings as 'gfortran -Wall zk.f90'.

The fragment of the code comes from my long-term project Munipack and has been
verified by a lot of f90 and f77 compilers and it doesn't violate any fortran
norm (in my oppinion). The hang-up doesn't not appear with original g95
project, but only with the new gfortran project at least year ago.


-- 
           Summary: hang-up during read of non-expected input
           Product: gcc
           Version: 4.3.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: hroch at physics dot muni dot cz


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[burnus at gcc dot gnu dot org]

------- Comment #1 from burnus at gcc dot gnu dot org  2007-12-09 21:03 -------
Do you have an idea, Jerry? I can reproduce it with gfortran 4.1.3, 4.2.2 and
4.3.0 and as he writes, other compilers (NAG f95, g95, ifort) work.

(Note: One needs to change "ii" into "iostat" or vice versa.)

gfortran does not print anything and hangs, with other compilers, the output
is:

 0. 0.
 329.142 214.107
 0. 0.
 360.334 245.261
 0. 0.
 332.704 317.964
 0. 0.
 379.769 231.226
 0. 0.


-- 

burnus at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jvdelisle at gcc dot gnu dot
                   |                            |org, burnus at gcc dot gnu
                   |                            |dot org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2007-12-09 21:03:10
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[jvdelisle at gcc dot gnu dot org]

------- Comment #2 from jvdelisle at gcc dot gnu dot org  2007-12-09 22:31 -------
I will investigate this one.


-- 

jvdelisle at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |jvdelisle at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2007-12-09 21:03:10         |2007-12-09 22:31:33
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[jvdelisle at gcc dot gnu dot org]

------- Comment #3 from jvdelisle at gcc dot gnu dot org  2007-12-10 02:16 -------
First, you are relying on non-portable behavior:

F95 Standard   9.4.1.5 Error branch - subpart (2)->

"The position of the file specified in the input/output statement becomes
indeterminate"

So gfortran does not try to change the file position after the read and the
subsequent reads continue to read bad values and ad-infinitum.  The loop is in
your program.

Second, as a Quality of Implementation issue, and under these conditions we can
do whatever we want, I have a patch testing that moves the file to the next
record position.  :)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[burnus at gcc dot gnu dot org]

------- Comment #4 from burnus at gcc dot gnu dot org  2007-12-10 06:43 -------
Patch:
http://gcc.gnu.org/ml/gcc-patches/2007-12/msg00433.html


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[dominiq at lps dot ens dot fr]

------- Comment #5 from dominiq at lps dot ens dot fr  2007-12-10 19:49 -------
Patch of comment #4 works as advertised without regression at -m32 and -m64 
for ppc-Darwin9 and Intel-Darwin8/9.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[jvdelisle at gcc dot gnu dot org]

------- Comment #6 from jvdelisle at gcc dot gnu dot org  2007-12-11 03:50 -------
Subject: Bug 34411

Author: jvdelisle
Date: Tue Dec 11 03:49:52 2007
New Revision: 130767

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=130767
Log:
2007-12-10  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR libfortran/34411
        * io/read.c (convert_real, read_l, read_decimal, read_radix, read_f):
        Call next_record after bad read or overflow error.

Modified:
    trunk/libgfortran/ChangeLog
    trunk/libgfortran/io/read.c


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[jvdelisle at gcc dot gnu dot org]

------- Comment #7 from jvdelisle at gcc dot gnu dot org  2007-12-11 03:56 -------
Subject: Bug 34411

Author: jvdelisle
Date: Tue Dec 11 03:56:05 2007
New Revision: 130768

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=130768
Log:
2007-12-10  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

        PR fortran/34411
        * gfortran.dg/error_recovery_5.f90: New.
        * gfortran.dg/optional_dim_2.f90: Fix test, not related to this PR.

Added:
    trunk/gcc/testsuite/gfortran.dg/error_recovery_5.f90
Modified:
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/gfortran.dg/optional_dim_2.f90


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411

[Bug fortran/34411] hang-up during read of non-expected input

[jvdelisle at gcc dot gnu dot org]

------- Comment #8 from jvdelisle at gcc dot gnu dot org  2007-12-14 19:25 -------
Fixed on trunk.


-- 

jvdelisle at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34411