(In reply to Miroslav Lachman from comment #0)
> I am using dns/fastresolve for years, but it no longer works on FreeBSD 10.x
> amd64. It always segfaults.
> 
> # dns-terror -p 1000 -d /var/tmp/ip2host.db -o < /tmp/access.log
> Segmentation fault (core dumped)

Thank you, Miroslav, for all your valuable input.
I use this on a 9-STABLE box, without trouble(s). But given that
you are using this on a 10.x box, I just installed a fresh copy
on my 11-CURRENT dev box, and tested it with the exact command
you reported having used it with. My experience was as expected.
I had no problems.

As to the errors emitted during build time. My experience was
that dns/adns emitted far worse error messages, than dns/fastresolve.
But I would agree that *any* errors should idealy be delt with.

Perhaps you're dealing with symbol clash. Are your external
dependencies in sync? What versions are you using?

I'm working with:
adns-1.5.0
p5-BerkeleyDB-0.55
fastresolve-2.10_5

Working Copy Root Path: /usr/ports
URL: svn://svn.freebsd.org/ports/head
Repository Root: svn://svn.freebsd.org/ports
Revision: 406193

Thanks again, for your feedback, Miroslav.

--Chris

It is strange. I already asked few other people on Czech FreeBSD mailinglist and fastresolve doen't work for them too.
http://www.freebsd.cz/listserv/archive/users-l/2016q1/029182.html

I have this problem for about one year on all 10.x installations. I previously used it on 8.4 without any problems, but after upgrade to 10.1 or 10.2 it stops (after recompilation from fresh ports tree). The behavior is the same on all our webservers (15+ machines). Some of them were upgraded from older releases, some of them freshly installed from 10.2 ISO media.
Then I migrated from individual ports tree to centralized private pkg repo with poudriere.
All machines have kernel and userland in sync, all machines have installed fastresolve from the same repo where packages are built in 10.2-RELEASE-p8 amd64 environment:

# poudriere jail -l
JAILNAME   VERSION          ARCH  METHOD TIMESTAMP           PATH
10_1_amd64 10.1-RELEASE-p22 amd64 ftp    2015-10-03 19:21:59 /vol0/poudriere/jails/10_1_amd64
10_2_amd64 10.2-RELEASE-p8  amd64 ftp    2016-01-07 10:34:52 /vol0/poudriere/jails/10_2_amd64

Did you tested fastresolve package from official repo or did you compiled it from ports tree?

Info from one of our machines

# uname -srmi
FreeBSD 10.2-RELEASE-p10 amd64 GENERIC

# pkg info -d -r fastresolve
fastresolve-2.10_5
Depends on     :
        adns-1.5.0
        p5-BerkeleyDB-0.55
        perl5-5.20.3_8
        db5-5.3.28_3


adns (adnslogres command) works without problem

I am really out of ideas what to try next to make fastresolve work on 10.2 amd64 machines.

Thank you for your reply anyway.

(In reply to Miroslav Lachman from comment #2)

Thanks for your quick reply, Miroslav.

I have done some more testing, and I can *confirm*
there are problems. If I feed dns-terror a file that *only*
has IP addresses, one per line. I receive the following error:

Bus error (core dumped)

So, there is more investigation to do. I'll see if I can trace
it down.

> Did you tested fastresolve package from official repo or did
> you compiled it from ports tree?

I build it from the ports tree.

Thanks again, Miroslav!

--Chris

(In reply to Chris Hutchinson from comment #3)
> (In reply to Miroslav Lachman from comment #2)
> 
> Thanks for your quick reply, Miroslav.
> 
> I have done some more testing, and I can *confirm*
> there are problems. If I feed dns-terror a file that *only*
> has IP addresses, one per line. I receive the following error:
> 
> Bus error (core dumped)
> 
> So, there is more investigation to do. I'll see if I can trace
> it down.
> 
> > Did you tested fastresolve package from official repo or did
> > you compiled it from ports tree?
> 
> I build it from the ports tree.
> 
> Thanks again, Miroslav!
> 
> --Chris

OK. I've narrowed it down to an issue with dns/adns. I've
confirmed it works with dns/adns-1.42. I'll attempt to find
out at which version it no longer works. But even at
version 1.42, it (adns) chokes on a stabdard FreeBSD
resolv.conf(5) file.

But frankly this is *really* a dns/adns issue. :-/

--Chris

Hmm I am completely lost, because adns (adnslogres) works for me and I assume it uses the same way to read resolv.conf as library used by dns-terror.

Well, lets allow me to publish results of this night research.

Note the ud2a instruction on the end of disassembly of _ZN10BoolStringaSERKS_ function. The ud2a is instruction guaranteed by Intel to be invalid.

Some compilers use it to fire intentional "Illegal instruction" abort at runtime in the case they can't compile c++ code correctly, but the cause should generate warning only at compile time.

The function _ZN10BoolStringaSERKS_ come from following source code function:

 ----------------------
  BoolString &operator=(const BoolString &other) {
    str = other.str;
    flag = other.flag;
  }
 ----------------------

The compilation of such code produce
1. ./BoolString.h:59:3: warning: control reaches end of non-void function 
2. ud2a instruction in the code

Lets patch the code to be correct:

 ----------------------
  BoolString &operator=(const BoolString &other) {
    str = other.str;
    flag = other.flag;
    return *this;
  }
 ----------------------

Compiler no longer produce a warning - and no longer emit ud2a into code.

I consider it solution of the reported bug.

 ............

Note that there are other warnings during compilation as well, so there may be severe issues with the resulting code.

warning: conversion from string literal to 'char *' is deprecated
is fired at
dns-terror.cc:346:34
dns-terror.cc:373:11
dns-terror.cc:431:27
dns-terror.cc:581

and 
format specifies type 'int' but the argument has type 'size_type' (aka 'unsigned long')
is fired at
dns-terror.cc:706:33

I didn't analyzed them. They may be "warnings safe to be ignored" or may be warnings fatal for proper function of resulting code.

 ..............

Some additional data. Analyzed on:

 -------------
OS: 10.2-RELEASE-p12 amd64

c++ -v
FreeBSD clang version 3.4.1 (tags/RELEASE_34/dot1-final 208032) 20140512
Target: x86_64-unknown-freebsd10.2
Thread model: posix
Selected GCC installation:

Command to compile:
c++ -DPACKAGE=\"fastresolve\" -DVERSION=\"2.10\" -DHAVE_FGETLN=1 -DHAVE_LIBZ=1 -DHAVE_LIBADNS=1  -I. -I. -g -O0 -pipe -fstack-protector -fno-strict-aliasing  -I/usr/local/include -I/usr/local/include/db6 -c dns-terror.cc

Command to link:
c++ -g -O0 -pipe -fstack-protector -fno-strict-aliasing  -I/usr/local/include -I/usr/local/include/db6 -L/usr/local/lib -L/usr/local/lib/db6 -fstack-protector -o dns-terror  dns-terror.o getline.o fgetln.o DatedStringDb.o  -ladns -lz  -ldb_cxx-6.1

(In reply to Dan Lukes from comment #6)
> Well, lets allow me to publish results of this night research.
> 
Thanks for the feedback, Dan. I too considered the warnings
emitted, and determined that it *might* be the version/brand
of c++. But given that dns/fastresolve hasn't changed versions,
yet the demends; dns/adns && databases/p5-BerkeleyDB *have*.
I felt they were good first candidates. I have already submitted
a patch for databases/p5-BerkeleyDB, as a result. But then
ran out of time to further investigate. I've already determined
that it's a string related issue. So it's either a change in
dns/adns, or the version/brand of c++ that causes it. I'll
have time today to completely sort it out.

Thanks again, Dan!

--Chris

OK Dan. Further investigation reveals (at least for now)
that your suggestion will prevent the fatal crash -- thanks!

Miroslav, and others; please see pr
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=207628
for an svn diff to patch your current version. Or wait until
that diff gets committed, and svn up your current ports tree.

Thanks!

--Chris

Comment 9 commit-hook 2016-03-02 07:58:29 UTC

A commit references this bug:

Author: pi
Date: Wed Mar  2 07:58:13 UTC 2016
New revision: 409915
URL: https://svnweb.freebsd.org/changeset/ports/409915

Log:
  dns/fastresolve: fix segfault

  PR:		207589, 207628
  Submitted by:	Miroslav Lachman <000.fbsd@quip.cz>, Dan Lukes <dan@obluda.cz>, Chris Hutchinsin <portmaster@bsdforge.com> (maintainer)

Changes:
  head/dns/fastresolve/Makefile
  head/dns/fastresolve/files/patch-dns-terror_BoolString.h
  head/dns/fastresolve/files/patch-dns-terror__DatedStringDb.cc
  head/dns/fastresolve/files/patch-dns-terror__dns-terror.cc
  head/dns/fastresolve/files/patch-dns-terror_getline.c

Comment 10 Kurt Jaeger 2016-03-02 08:07:54 UTC

Committed, thanks.