Re: [bug-inetutils] inetutils 1.9.1-4: ftp "get" command occasionally dumps core,

[Glenn Golden]

--
Hi Mats,

Sounds like you've been a busy bee, chasing down segfaults!  Probably
I will not have time until this weekend to download your updated git version
and try it, but I did want to respond to some of the questions you had asked:

Mats Erik Andersson <address@hidden> [2013-09-10 13:14:04 +0200]:
> 
> If you execute this
> 
>    $ ./tests/localhost
>    localhost: charybdis
> 
> do you get your host name without a corresponding domain?
>

No, I get the domain as well: 

  $ ./tests/localhost
  localhost: localhost.localdomain

> 
> The matter you found with getaddrinfo() clearly indicates
> that at least one issue is caused beyond my control, namely
> in eglibc-2.17 in use by Arch Linux. How can we insure that
> the library is not causing other issues during your execution?
> 

Perhaps the best way to sort out the implication of Arch's eglibc is to get
Eric Belanger involved.  He's the downstream maintainer of inetutils for Arch,
so he should be able to not only answer your questions about the particulars
of Arch's eglibc version -- about which I know zero -- but may also be able to
confirm/refute whether he is able to reproduce my segfault on his Arch system.

I will post a message to Eric on the Arch bugtracker. You should be able to
view it here within a few hours or so:

    https://bugs.archlinux.org/task/35898

I'm going to encourage him to contact you directly regarding the eglibc stuff,
so you two guys can sort that out between yourselves. I don't think I can be
of much help with that.

I'll also ask Eric if he can try to reproduce my segfault on his Arch system.
If not, then perhaps there's something about my particular machines that is
relevant to the problem, though I can't imagine what it might be, since I'm
seeing identical failures on three different machines (see below).

If he is able to reproduce it, I'll of course continue to work with you to try
to track it down.

>
> The fact that wrapping ftp within valgrind actually did avoid
> one persistent segmentation fault, indicates to me that there
> might well be something foul on Arch Linux in your case. I run
> a virtualized instance of Arch Linux amd64 in VirtualBox on
> OpenSolaris as comparison. Do you use i386 or amd64?
> 

The segfault occurs for me on three different machines, all running Arch:

  1:  ThinkPad T43 (i386). This machine was fully synched (pacman -Syu)
      to the Arch repos about 2 weeks ago (but in any case had been failing
      the same way prior to that time.)

  2:  Another ThinkPad T43 running Arch, fully synched 2 days ago.

  3:  Thinkpad T-510 (Intel i7 CPU), fully synched to the repos yesterday
      (and also had been failing prior to that synch).
        

Hope this is helpful.

I'll try the git version as soon as I can.

Regards,
Glenn