Sendmail 5.64+IDA: IDA Ruleset 8

Sendmail 5.64+IDA: IDA Ruleset 8

Post by Seth Roberts » Mon, 09 Jul 1990 17:23:29


I'm in the proccess of installing Sendmail 5.64 with IDA changes on a
Sun 3 using the real Bind 4.8 libraries and statically linked to SunOS
4.1 libraries (on a 4.0.3 machine... don't ask)

In any case, my question has to do with the name resolver lookup in
ruleset 8 (for those who don't happen to have ruleset 8 on them:


R<$+>$+                   $:<$1?>$2                 Add marker.
R<$=X[$+.$+]?>$+  <$1[$2.$3]>$4                     [] is ok.
R<$=X$*$~P?>$+            $:<$1$[ $2$3 $:$2$3? $]?>$4       First try resolver
R<$=X$+?$*>$+             $:<$1$(N $2 $:$2$3 $)>$4  Next domain table
R<$=X$=U?>!$+             <$1$2.UUCP>!$3                    Local UUCP host?
R<$=X$-?>!$+              $:<$1$2?$(M $2.UUCP $: $)>!$3     Known in mailertable?
R<$+?$+>$+                <$1.UUCP>$3                       Add .UUCP if known
R<$*?$*>$*                <$1$2>$3                  Remove marker

My question has to do with retries via the various permutations of the
appending of the default domain to try to see if the name was a
partial match: (i.e.  for "cs" it would look up "",
"", and "cs" (not necessarily in that order, but it
makes sense that way)) Specificly, what is an easy, portable method of
doing that (either via config file changes or via sendmail/bind
changes)?  I made a quick hack, substituting the following three lines
for the first of those three lines, but I'm sure you will all agree
that this is *very* ugly.

R<$=X$*$~P?>$+            $:<$1$[ $2$3 $:$2$3? $]?>$4       First try resolver
R<$=X$+??>$+              $:<$1$[ $ $:$2? $]?>$3  First try resolver
R<$=X$+??>$+              $:<$1$[ $ $:$2? $]?>$3      First try resolver

Any suggestions?

                                        -Seth Robertson


1. sendmail 5.64 w/ IDA freeze/debug problem

When a frozen configuration file exists, and sendmail (5.64 with
IDA enhancements) is run with the -d debug flag, the process
stops and dumps core (memory fault). e.g.:
    sendmail -d -v -bv foo
        Version 5.64+
        setoption v=
        setoption c=F
        setoption d=
        ksh: 15597 Memory fault

This doesn't happen if 1) vanilla 5.64 is used, or 2) no frozen
configuration file exists.  A preliminary investigation revealed
that malloc complains about an apparent corruption of the heap;
5.64+ seems to call malloc and free a few times more before
freeze than does 5.64.

I'm still investigating, but as there are no doubt many
differences in implementation of malloc(), I wondered if anybody
else could confirm this problem.

Please reply via email; I'll post a summary.


    Digital Television Tape Recording,   | uunet!{sonyusa,vmp}!sonyd1!bruce

    Teaneck, NJ 07666  |  Telephone: 1(201)833-5693   | FAX: 1(201)833-9279

2. How to Setfocus to tabstrip item

3. Where to get Sendmail 5.64+IDA patched for A/UX?

4. Bmatrix (?) environment in AmS-LaTeX documented in amsldoc.tex

5. Sendmail 5.64 IDA 1.3.4 Problem

6. RTF file format

7. running sendmail 5.64 + IDA on [3p]max's (and system crashes)

8. FS. CISCO ROUTERS DC 2-2511'S 1-2507

9. Sendmail 5.64 IDA 1.3.4 / DEC mail11v3 problem

10. Source for Sendmail 5.64+IDA Needed

11. Where to FTP sendmail 5.64+IDA for A/UX 2.0?

12. Problems with Sendmail 5.64+IDA on DG/UX 4.20

13. Problems with Sendmail 5.64 (w/IDA enhancements) on PC/RT (Long)