[Bug #1190] java.io.RandomAccessFile.skipBytes() seeks past EOF

[nobody]

=================== BUG #1190: LATEST MODIFICATIONS ==================
http://savannah.gnu.org/bugs/?func=detailbug&bug_id=1190&group_id=85

Changes by: Brian Jones <address@hidden>
Date: 2003-Jan-26 10:49 (US/Eastern)

            What     | Removed                   | Added
---------------------------------------------------------------------------
          Resolution | None                      | Works for me
         Assigned to | None                      | cbj
              Status | Open                      | Closed


------------------ Additional Follow-up Comments ----------------------------
The Mauve test case has changed and I have not reproduced this.  I did however 
locate an error related to skipBytes() testing in the new test case and I'll 
check that in tomorrow.  The checks for whether to seek and how far to seek are 
done in the Java code btw.



=================== BUG #1190: FULL BUG SNAPSHOT ===================


Submitted by: crawley                 Project: classpath                    
Submitted on: 2002-Sep-12 12:16
Severity:  5 - Major                  Resolution:  Works for me             
Assigned to:  cbj                     Status:  Closed                       
Platform Version:  Red Hat 7.x        

Summary:  java.io.RandomAccessFile.skipBytes() seeks past EOF

Original Submission:  The Sun javadocs imply that skipBytes() should not skip
past the EOF of a file, and that's the way that JDK 1.4
behaves.  But the Classpath version will seek of into 
the distance.

This bug is exercised by Mauve testcase
    gnu.testlet.java.io.RandomAccessFile.raf.java
which goes into an infinite loop.

The bug is in the native code.  The lseek(2) syscall
can seek beyond the EOF position on Linux.  A 3rd lseek
call is needed to test for this, and a 4th to correct it.


Follow-up Comments
*******************

-------------------------------------------------------
Date: 2003-Jan-26 10:49             By: cbj
The Mauve test case has changed and I have not reproduced this.  I did however 
locate an error related to skipBytes() testing in the new test case and I'll 
check that in tomorrow.  The checks for whether to seek and how far to seek are 
done in the Java code btw.


CC list is empty


No files currently attached


For detailed info, follow this link:
http://savannah.gnu.org/bugs/?func=detailbug&bug_id=1190&group_id=85