This is the mail archive of the cygwin@cygwin.com mailing list for the Cygwin project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: sizes with du.exe (fileutils 4.0-3)


On Wed, Apr 18, 2001 at 07:40:52PM -0400, Jerry Heyman wrote:
>On 18 April 2001 at 17:34, Earnie Boyd <earnie_boyd@yahoo.com> wrote:
>> "(Jerry Heyman)" wrote:
>> > 
>> > Howdy,
>> > 
>> > I'm new to the list, and have the following situation:
>> > 
>> > . Product version of the Cgynus Toolkit is installed.
>> > . We noticed a problem with the du executable from fileutils 3.16
>> >   returning negative numbers on files > 2G
>> > . Downloaded the fileutils-4.0 from ftp.gnu.org and built them
>> >   results were better but only seemed to report half the actual value
>> > . Searched the Cygwin e-mail archives and found a fix for sizing problem
>> >   in fileutils-4.0-3.
>> > . Downloaded the source fileutils-4.0-3 from cygwin.com
>> >   results were much better, but the following is still occurring
>> > 
>> > $ ls -l du.exe
>> > -rwxrwxrwx   1 administ Administ   443705 Apr 13 13:55  du.exe
>
>I need to clarify this.  The du.exe that is being counted
>in each of these instances is the 4.0 version that I built
>straight from the fileutile-4.0 gnu source.
>
>>> du (3.16):
>>> $ du -b du.exe
>>> 443705  du.exe
>> > 
>> 
>> This is the correct size in bytes for the 3.16 version.
>
>the 443705 is the value returned from doing a 'du -b'
>using the 3.16 due command COUNTING the du.exe 4.0 binary
>
>>> du (4.0):
>>> $ ./du -b du.exe
>>> 222208  du.exe
>>>
>
>the 222208 is the value returned from doing a 'du -b'
>using the 4.0 binary to count itself.
> 
>> This is 1/2 the size in bytes for the 4.0 version.
>> 
>>> du (4.0-3):
>>> $ ./du403 -b du.exe
>>> 444416  du.exe
>>>
>
>the 444416 is the value returned from doing a 'du -b'
>using the 4.0-3 binary to count the bytes of the 4.0 binary.
>
>Three different versions of du counting the same du.exe
>giving three different results.  I know that 4.0 had a 
>problem - hence the use of 4.03, but I'm still confused about the
>difference in results between the 3.16 binary and the 4.0-3
>binary counting the same file.

du 4.0-3 shows the actual space used, rounding up to block sizes.

Check out the announcement for 4.0-3 in the cygwin-announce mailing list
archives for the reason for this release.

cgf

--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]