do md5sums differ across architectures?

do md5sums differ across architectures?

[Kevin B. Hendricks]

Hi,

I ran md5sum on a big bz2 archive on my ppc linux box and then
ftp'd it to my Sun Solaris 7 box at work.

I checked it with md5sum on that machine and it differed.

Okay I thought, a bad packet during upload.

So I tried again, same result.

I then passed back and forth (known good archives) and the md5sums
were different.

Okay, I thought maybe an endian issue but both Solaris Sparc and PPC
Linux are big endian.

Should these sums be different?

Thanks,

Kevin

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[Ethan Benson]

[-- Attachment #1: Type: text/plain, Size: 805 bytes --]

On Thu, Feb 22, 2001 at 10:24:01AM -0500, Kevin B. Hendricks wrote:
> 
> Hi,
> 
> I ran md5sum on a big bz2 archive on my ppc linux box and then
> ftp'd it to my Sun Solaris 7 box at work.
> 
> I checked it with md5sum on that machine and it differed.
> 
> Okay I thought, a bad packet during upload.
> 
> So I tried again, same result.
> 
> I then passed back and forth (known good archives) and the md5sums
> were different.
> 
> Okay, I thought maybe an endian issue but both Solaris Sparc and PPC
> Linux are big endian.
> 
> Should these sums be different?

no, that would rather defeat the purpose of md5sum hashes.  one of
your md5sum programs is broken.  

or perhaps you ftp'd in text mode instead of binary mode.  

-- 
Ethan Benson
http://www.alaska.net/~erbenson/

[-- Attachment #2: Type: application/pgp-signature, Size: 0 bytes --]

Re: do md5sums differ across architectures?

[Kevin B. Hendricks]

Hi,

No, ftping in binary.

I am running textutils-2.0e6 rpm (the latest) from www.linuxppc.org which supplies
md5sum

I will check what I am running on Solaris.

hmm, I wonder which one is broken?

Thanks,

Kevin

On Thursday, February 22, 2001, at 10:49 AM, Ethan Benson wrote:

> On Thu, Feb 22, 2001 at 10:24:01AM -0500, Kevin B. Hendricks wrote:
> >
> > Hi,
> >
> > I ran md5sum on a big bz2 archive on my ppc linux box and then
> > ftp'd it to my Sun Solaris 7 box at work.
> >
> > I checked it with md5sum on that machine and it differed.
> >
> > Okay I thought, a bad packet during upload.
> >
> > So I tried again, same result.
> >
> > I then passed back and forth (known good archives) and the md5sums
> > were different.
> >
> > Okay, I thought maybe an endian issue but both Solaris Sparc and PPC
> > Linux are big endian.
> >
> > Should these sums be different?
>
> no, that would rather defeat the purpose of md5sum hashes.  one of
> your md5sum programs is broken.
>
> or perhaps you ftp'd in text mode instead of binary mode.
>
> --
> Ethan Benson
> http://www.alaska.net/~erbenson/
>
>

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[mod+linuxppc-embedded]

Simple test cases: on both machines do

for string in \
   "" \
   "a" \
   "abc" \
   "message digest" \
   "abcdefghijklmnopqrstuvwxyz" \
   "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789" \
   "12345678901234567890123456789012345678901234567890123456789012345678901234567890"
do
    echo -n $string | md5sum
done


The results MUST be:

   d41d8cd98f00b204e9800998ecf8427e
   0cc175b9c0f1b6a831c399e269772661
   900150983cd24fb0d6963f7d28e17f72
   f96b697d7cb7938d525a2f31aaf161d0
   c3fcd3d76192e4007dfb496cca67e13b
   d174ab98d277d9f5a5611c2c9f419d9f
   57edf4a22be3c955ac49da2e2107b67a

...respectively, or it's broken.


SEE ALSO: RFC 1321

 (e.g. http://rfc.asuka.net/rfc/rfc1321.html )


** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[Kevin B. Hendricks]

Hi,

Just to have a bit of fun, how about this (3 different systems)
Each generates a different md5sum but all generate the same cksum!
(and adding -b to md5sum makes no difference)

Darwin khendricksmac 1.2 Darwin Kernel Version 1.2: Wed Aug 30 23:32:53 PDT 2000; root:xnu/xnu-103.obj~1/RELEASE_PPC  Power Macintosh powerpc

[khendricksmac:~] kbhend% md5sum oo619b_src.tar.bz2
9c6a44ad66572c834e282eb1ec665374  oo619b_src.tar.bz2
[khendricksmac:~] kbhend% cksum oo619b_src.tar.bz2
2458761606 56057947 oo619b_src.tar.bz2


SunOS crunch 5.7 Generic_106541-04 sun4u sparc SUNW,Ultra-250

kbhend$ md5sum oo619b_src.tar.bz2
278c21a764126027cba9cb685b7e96cf  oo619b_src.tar.bz2
kbhend$ cksum oo619b_src.tar.bz2
2458761606      56057947        oo619b_src.tar.bz2


Linux localhost 2.2.17pre10-ben2 #3 Sun Aug 13 13:03:45 EDT 2000 ppc unknown

[kbhend@localhost /src2]$ md5sum oo619b_src.tar.bz2
c4a5d7bda6973f6b56e725295d1b3dfb  oo619b_src.tar.bz2
[kbhend@localhost /src2]$ cksum oo619b_src.tar.bz2
2458761606 56057947 oo619b_src.tar.bz2


I guess md5sum is not a very reliable or consistent checksum measure for moving
files between ppc linux, MacOSX, and Solaris 7 Sparc.

Back to the old cksum for me.

Kevin

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[Jeffrey Hawkins]

Kevin,


Don't know what is going on with your md5sum, but
I ran it on a LinuxPPC 2.2.17 System and a SunOS 5.6,
and got the same results ??????  The version of md5sum
on Linux was 1.22, and the version on Sun was 2.0.


Jeff

,"Kevin B. Hendricks" wrote:
>
> Hi,
>
> Just to have a bit of fun, how about this (3 different systems)
> Each generates a different md5sum but all generate the same cksum!
> (and adding -b to md5sum makes no difference)
>
> Darwin khendricksmac 1.2 Darwin Kernel Version 1.2: Wed Aug 30 23:32:53 PDT 2000; root:xnu/xnu-103.obj~1/RELEASE_PPC  Power Macintosh powerpc
>
> [khendricksmac:~] kbhend% md5sum oo619b_src.tar.bz2
> 9c6a44ad66572c834e282eb1ec665374  oo619b_src.tar.bz2
> [khendricksmac:~] kbhend% cksum oo619b_src.tar.bz2
> 2458761606 56057947 oo619b_src.tar.bz2
>
> SunOS crunch 5.7 Generic_106541-04 sun4u sparc SUNW,Ultra-250
>
> kbhend$ md5sum oo619b_src.tar.bz2
> 278c21a764126027cba9cb685b7e96cf  oo619b_src.tar.bz2
> kbhend$ cksum oo619b_src.tar.bz2
> 2458761606      56057947        oo619b_src.tar.bz2
>
> Linux localhost 2.2.17pre10-ben2 #3 Sun Aug 13 13:03:45 EDT 2000 ppc unknown
>
> [kbhend@localhost /src2]$ md5sum oo619b_src.tar.bz2
> c4a5d7bda6973f6b56e725295d1b3dfb  oo619b_src.tar.bz2
> [kbhend@localhost /src2]$ cksum oo619b_src.tar.bz2
> 2458761606 56057947 oo619b_src.tar.bz2
>
> I guess md5sum is not a very reliable or consistent checksum measure for moving
> files between ppc linux, MacOSX, and Solaris 7 Sparc.
>
> Back to the old cksum for me.
>
> Kevin
>

--
************************************************

Jeffrey Hawkins
Senior Staff Software Engineer
Motorola Wireless Data Solutions Engineering

Address: DEPT:  DQ525
         MS:    IL-02-1055A              .-.
         1301 East Algonquin Road        /v\
         Schaumburg, IL 60196           // \\
                                       /(   )\
Phone:   (847)576-7463                  ^^ ^^
FAX:     (847)576-7737

************************************************

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[Kevin B. Hendricks]

Hi,

Here is what md5sum --version says on my machine.

[kbhend@localhost kbhend]$ md5sum --version
md5sum (GNU textutils) 2.0e
Written by Ulrich Drepper.

Copyright (C) 1999 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
[kbhend@localhost kbhend]$


This appears to be newer that yours.  I will upgrade textutils again on both
my box and the solaris 7 box.  Hopefully eventually they will coincide.

I still don't know why MacOSX differs from both.  I thought it would match
one or the other so that I knew which one to upgrade.

Thanks,

Kevin


On Thursday 22 February 2001 15:09, Jeffrey Hawkins wrote:
> Kevin,
>
>
> Don't know what is going on with your md5sum, but
> I ran it on a LinuxPPC 2.2.17 System and a SunOS 5.6,
> and got the same results ??????  The version of md5sum
> on Linux was 1.22, and the version on Sun was 2.0.
>
>
> Jeff
>
> ,"Kevin B. Hendricks" wrote:
> >
> > Hi,
> >
> > Just to have a bit of fun, how about this (3 different systems)
> > Each generates a different md5sum but all generate the same cksum!
> > (and adding -b to md5sum makes no difference)
> >
> > Darwin khendricksmac 1.2 Darwin Kernel Version 1.2: Wed Aug 30 23:32:53
PDT 2000; root:xnu/xnu-103.obj~1/RELEASE_PPC  Power Macintosh powerpc
> >
> > [khendricksmac:~] kbhend% md5sum oo619b_src.tar.bz2
> > 9c6a44ad66572c834e282eb1ec665374  oo619b_src.tar.bz2
> > [khendricksmac:~] kbhend% cksum oo619b_src.tar.bz2
> > 2458761606 56057947 oo619b_src.tar.bz2
> >
> > SunOS crunch 5.7 Generic_106541-04 sun4u sparc SUNW,Ultra-250
> >
> > kbhend$ md5sum oo619b_src.tar.bz2
> > 278c21a764126027cba9cb685b7e96cf  oo619b_src.tar.bz2
> > kbhend$ cksum oo619b_src.tar.bz2
> > 2458761606      56057947        oo619b_src.tar.bz2
> >
> > Linux localhost 2.2.17pre10-ben2 #3 Sun Aug 13 13:03:45 EDT 2000 ppc
unknown
> >
> > [kbhend@localhost /src2]$ md5sum oo619b_src.tar.bz2
> > c4a5d7bda6973f6b56e725295d1b3dfb  oo619b_src.tar.bz2
> > [kbhend@localhost /src2]$ cksum oo619b_src.tar.bz2
> > 2458761606 56057947 oo619b_src.tar.bz2
> >
> > I guess md5sum is not a very reliable or consistent checksum measure for
moving
> > files between ppc linux, MacOSX, and Solaris 7 Sparc.
> >
> > Back to the old cksum for me.
> >
> > Kevin
> >
>
> --
> ************************************************
>
> Jeffrey Hawkins
> Senior Staff Software Engineer
> Motorola Wireless Data Solutions Engineering
>
> Address: DEPT:  DQ525
>          MS:    IL-02-1055A              .-.
>          1301 East Algonquin Road        /v\
>          Schaumburg, IL 60196           // \\
>                                        /(   )\
> Phone:   (847)576-7463                  ^^ ^^
> FAX:     (847)576-7737
>
> ************************************************
>

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: do md5sums differ across architectures?

[Kevin B. Hendricks]

Hi,

I upgraded my Solaris 7 to the gnu 2.0 version and now it matches ppc linux
completely.

That still leaves MacOSX out in the cold but...

Thanks for the hint.

Kevin

On Thursday 22 February 2001 16:13, Kevin B. Hendricks wrote:
> Hi,
>
> Here is what md5sum --version says on my machine.
>
> [kbhend@localhost kbhend]$ md5sum --version
> md5sum (GNU textutils) 2.0e
> Written by Ulrich Drepper.
>
> Copyright (C) 1999 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.  There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
> [kbhend@localhost kbhend]$
>
>
> This appears to be newer that yours.  I will upgrade textutils again on both
> my box and the solaris 7 box.  Hopefully eventually they will coincide.
>
> I still don't know why MacOSX differs from both.  I thought it would match
> one or the other so that I knew which one to upgrade.
>
> Thanks,
>
> Kevin
>
>
> On Thursday 22 February 2001 15:09, Jeffrey Hawkins wrote:
> > Kevin,
> >
> >
> > Don't know what is going on with your md5sum, but
> > I ran it on a LinuxPPC 2.2.17 System and a SunOS 5.6,
> > and got the same results ??????  The version of md5sum
> > on Linux was 1.22, and the version on Sun was 2.0.
> >
> >
> > Jeff
> >
> > ,"Kevin B. Hendricks" wrote:
> > >
> > > Hi,
> > >
> > > Just to have a bit of fun, how about this (3 different systems)
> > > Each generates a different md5sum but all generate the same cksum!
> > > (and adding -b to md5sum makes no difference)
> > >
> > > Darwin khendricksmac 1.2 Darwin Kernel Version 1.2: Wed Aug 30 23:32:53
> PDT 2000; root:xnu/xnu-103.obj~1/RELEASE_PPC  Power Macintosh powerpc
> > >
> > > [khendricksmac:~] kbhend% md5sum oo619b_src.tar.bz2
> > > 9c6a44ad66572c834e282eb1ec665374  oo619b_src.tar.bz2
> > > [khendricksmac:~] kbhend% cksum oo619b_src.tar.bz2
> > > 2458761606 56057947 oo619b_src.tar.bz2
> > >
> > > SunOS crunch 5.7 Generic_106541-04 sun4u sparc SUNW,Ultra-250
> > >
> > > kbhend$ md5sum oo619b_src.tar.bz2
> > > 278c21a764126027cba9cb685b7e96cf  oo619b_src.tar.bz2
> > > kbhend$ cksum oo619b_src.tar.bz2
> > > 2458761606      56057947        oo619b_src.tar.bz2
> > >
> > > Linux localhost 2.2.17pre10-ben2 #3 Sun Aug 13 13:03:45 EDT 2000 ppc
> unknown
> > >
> > > [kbhend@localhost /src2]$ md5sum oo619b_src.tar.bz2
> > > c4a5d7bda6973f6b56e725295d1b3dfb  oo619b_src.tar.bz2
> > > [kbhend@localhost /src2]$ cksum oo619b_src.tar.bz2
> > > 2458761606 56057947 oo619b_src.tar.bz2
> > >
> > > I guess md5sum is not a very reliable or consistent checksum measure for
> moving
> > > files between ppc linux, MacOSX, and Solaris 7 Sparc.
> > >
> > > Back to the old cksum for me.
> > >
> > > Kevin
> > >
> >
> > --
> > ************************************************
> >
> > Jeffrey Hawkins
> > Senior Staff Software Engineer
> > Motorola Wireless Data Solutions Engineering
> >
> > Address: DEPT:  DQ525
> >          MS:    IL-02-1055A              .-.
> >          1301 East Algonquin Road        /v\
> >          Schaumburg, IL 60196           // \\
> >                                        /(   )\
> > Phone:   (847)576-7463                  ^^ ^^
> > FAX:     (847)576-7737
> >
> > ************************************************
> >
>

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/