[Bug libfortran/25835] Segfault or Bad Address error on unformatted sequential READ

jvdelisle at gcc dot gnu dot org gcc-bugzilla@gcc.gnu.org
Sun Jan 22 07:45:00 GMT 2006



------- Comment #3 from jvdelisle at gcc dot gnu dot org  2006-01-22 07:45 -------
In the following case, the second read is failing to read anything and gives no
error, giving the impression that it is working if you don't look at the data. 
Works for data size of 2044 or less.  This bug is ugly.  Works fine if you get
rid of the first read, which leads me to believe we are not cleaning up
correctly after an end error condition. (I have been down more than one trail
on this already so who knows)  :)

       integer data(2045)
       data=5
       open(unit=11,form='unformatted')
       write(11)data
       read(11,end=        1000 )data
       call abort()
 1000  continue
       backspace 11
       backspace 11
       data = 0
       read(11)data
       print *, data
       close(11)
       end


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25835



More information about the Gcc-bugs mailing list