PGP SIGNED MESSAGE
Hash: SHA1
Hi,
since dhcp-3.0.4 there are serious problems with the size of PTIN 51 (IP
address lease time) with dhcpd running on !32bit ARCHS.
Some tests showed the following for 3.0.4 on an amd64 system running
FreeBSD:
PTIN: 51 ( 8) IP address lease time 86400 (24h)
This is for the same system running dhcpd 3.0.3:
PTIN: 51 ( 4) IP address lease time 86400 (24h)
So there is an obvious difference between the IP address lease time in
the DHCPFFER between 3.0.3 and 3.0.4 on this ARCH/S combination.
I've investigated this a little bit further and came to the following
list, verified by using sizeof(time_t) on several ARCHS and Ses
S| ARCH| sizeof(time_t)
- ++
FreeBSD| i386| 4
FreeBSD| pc98| 4
FreeBSD| alpha| 4
FreeBSD| powerpc| 4
FreeBSD| amd64| 8 !
FreeBSD| ia64| 8 !
FreeBSD| sparc64| 8 !
DragonflyBSD| i386| 4
DragonflyBSD| amd64| 4
NetBSD| i386| 4
NetBSD| amd64| 4
Linux-2.6.16.13| i386| 4
Linux-2.6.16.13| x64_64| 8 !
SF1 V4.0| alpha| 4
SF1 V5.1| alpha| 4
SS 5.9| sparc64| 4
MS-X| powerpc| 4
Unfortunately, I've no access to other ARCH/S combinations, there maybe
more with 8 bytes length of time_t.
The main fact is, that the current ISC dhcp distribution does not deal
with the different size of time_t on the various ARCH/S combinations.
I tested several clients with the server running on FreeBSD amd64 using
dhcpd version 3.0.4, to see how the clients do react on the wrong size of
the PTIN 51 (8 bytes).
I got no problems with clients running on various Linux distributions and
clients used on the *BSD systems.
Unfortunately, Windows (at least XP) seems to be very strict and ignores
the DHCPFFER completely.
As a very quick hack, i changed the "includes/cf/freebsd.h" file from
#define TIME time_t
to
#define TIME u_int32_t
and recompiled everything.
Beside some warnings during the build, because gmtime() and time() expect
64bit time_t arguments and get the 32bit u_int32_t now, everything went
fine.
I tried the Windows XP clients again and now the patched dhcpd offered the
IP address lease time with a length of 4 bytes, and XP is happy again.
I've not tested how this will affect the failover situation between
systems with 4 bytes and 8 bytes length of time_t. It may be broken too,
but I'm unsure.
The main question for me is now, should the S vendors fix their
definition of time_t (not sure what PSIX says about the size of time_t)
or should the different size be handled in th ISC dhcp distribution?
At least the RFC is very clear about the size of PTIN 51
- -snip
The time is in units of seconds, and is specified as a 32-bit
unsigned integer.
The code for this option is 51, and its length is 4.
Code Len Lease Time
| 51 | 4 | t1 | t2 | t3 | t4 |
- -snip
Any comments are much appreciated.
regards
Joerg
- --
The beginning is the most important part of the work.
-Plato
PGP SIGNATURE
Version: GnuPG v1.4.3 (FreeBSD)
dJDtAs1d3pobZxp9Ub0Gqo8=
=2d7r
PGP SIGNATURE

--
Brent L. Bates (UNIX Sys. Admin.)
M.S. 912Phone:(757) 865-1400, x204
NASA Langley Research Center FAX:(757) 865-8177
Hampton, Virginia 23681-0001
Email: B.L.BATES (AT) larc (DOT) nasa.govhttp://www.vigyan.com/~blbates/

PGP SIGNED MESSAGE
Hash: SHA1

Mon, 15 May 2006, David W. Hankins wrote:

Mon, May 15, 2006 at 08:17:21PM +0200, Joerg Pulz wrote:
>So there is an obvious difference between the IP address lease time in
>the DHCPFFER between 3.0.3 and 3.0.4 on this ARCH/S combination.
>
revision 1.202.2.31
date: 2005/10/10 16:45:39; author: dhankins; state: Exp; lines: +2 -2
- Several minor bugs, largely relating to treating 8-byte time values as
4-byte entities, have been repaired after careful review of the FreeBSD
ports collection's patch set. Thanks to the nameless entities who have
contributed to the FreeBSD ports. [ISC-bugs #13722]

Well, it appears my review of those changes wasn't complete enough.

Looks like this is a good enough of a reason for a 3.0.5 beta.

David,

i noticed the above text in the RELNTES.

If you have a patch, which you want to get tested, i have access to amd64,
ia64 and sparc64 machines running FreeBSD.
Linux on x86_64 and ia64 is also available in my pool, so i can check
these too.
Just send me any patch you have to fix this issue and i will give it a try
to make sure that you roll out a functional beta, if you want to release
one.

Thanks a lot
Joerg
- --
The beginning is the most important part of the work.
-Plato
PGP SIGNATURE
Version: GnuPG v1.4.3 (FreeBSD)

uon9Z7F/inlnCA0tScTHIWM=
=4fw3
PGP SIGNATURE

Mon, May 15, 2006 at 02:45:55PM -0700, David W. Hankins wrote:

I'm not sure atm if there will be a 3.0.5 soon or in a few weeks, but
due to the severity of this, I will likely distribute the patch to
anyone who asks.

I'm asking :-)

3.0.4 just hit our testing distribution, where it'll be causing breakage for
a larger userbase, so I'd like to get it fixed ASAP.

regards

Andrew