Mailing List Archive: [Bug 467] SPF checking doesn't fallback to HELO with null sender

[Bug 467] SPF checking doesn't fallback to HELO with null sender

holmgren at lysator

Jul 29, 2007, 3:12 PM

Post #1 of 6 (1061 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

Magnus Holmgren <holmgren@lysator.liu.se> changed:

What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|peter@bowyer.org |tom@duncanthrax.net
Status|ASSIGNED |NEW

--- Comment #3 from Magnus Holmgren <holmgren@lysator.liu.se> 2007-07-29 23:12:03 ---
See also http://bugs.debian.org/431239.

Isn't it sufficient to abstain from calling SPF_request_set_env_from() when
spf_envelope_sender is the null string? It looks that way to me, but I haven't
got around to testing yet.

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

[Bug 467] SPF checking doesn't fallback to HELO with null sender [ In reply to ]

holmgren at lysator

Jul 29, 2007, 5:54 PM

Post #2 of 6 (1046 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

--- Comment #4 from Magnus Holmgren <holmgren@lysator.liu.se> 2007-07-30 01:54:14 ---
I found a little round tuit :-) and it works the way I though - once, I then
realised. Things get trickier when a request is reused.

Another suggestion: Call spf_init() from spf_process() on the first invocation,
so that SPF isn't initialized until it's used.

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

[Bug 467] SPF checking doesn't fallback to HELO with null sender [ In reply to ]

peter at bowyer

Jul 30, 2007, 1:45 PM

Post #3 of 6 (1047 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

--- Comment #5 from Peter Bowyer <peter@bowyer.org> 2007-07-30 21:45:34 ---
(In reply to comment #4)
> I found a little round tuit :-) and it works the way I though - once, I then
> realised. Things get trickier when a request is reused.
> Another suggestion: Call spf_init() from spf_process() on the first invocation,
> so that SPF isn't initialized until it's used.

Yes, that was presumably why Tom's code is arranged how it is - I was orginally
going to do the same as your previous suggestion. We're following the same path
:-)

Maybe we should wait until your proposed change to the underlying library is
accepted - that would allow a bit of a re-factoring of the Exim code.

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

[Bug 467] SPF checking doesn't fallback to HELO with null sender [ In reply to ]

holmgren at lysator

Jul 30, 2007, 1:52 PM

Post #4 of 6 (1037 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

--- Comment #6 from Magnus Holmgren <holmgren@lysator.liu.se> 2007-07-30 21:52:51 ---
> --- Comment #5 from Peter Bowyer <peter@bowyer.org> 2007-07-30 21:45:34
> --- (In reply to comment #4)
>
> > I found a little round tuit :-) and it works the way I though - once, I
> > then realised. Things get trickier when a request is reused.
>
> Yes, that was presumably why Tom's code is arranged how it is - I was
> orginally going to do the same as your previous suggestion. We're following
> the same path

Otherwise we can just check wether sender_address is empty and pass
sender_helo_name in place of it.

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

[Bug 467] SPF checking doesn't fallback to HELO with null sender [ In reply to ]

peter at bowyer

Jul 30, 2007, 1:55 PM

Post #5 of 6 (1038 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

--- Comment #7 from Peter Bowyer <peter@bowyer.org> 2007-07-30 21:55:25 ---
(In reply to comment #6)
> > --- Comment #5 from Peter Bowyer <peter@bowyer.org> 2007-07-30 21:45:34
> > --- (In reply to comment #4)
> >
> > > I found a little round tuit :-) and it works the way I though - once, I
> > > then realised. Things get trickier when a request is reused.
> >
> > Yes, that was presumably why Tom's code is arranged how it is - I was
> > orginally going to do the same as your previous suggestion. We're following
> > the same path
> Otherwise we can just check wether sender_address is empty and pass
> sender_helo_name in place of it.

As long as sender_helo_name is non-empty. I think I had to do something about
that case, like setting it at HELO time - can't quite remember, it was a while
back and I don't have access to the system with the source on just now.

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##

[Bug 467] SPF checking doesn't fallback to HELO with null sender [ In reply to ]

holmgren at lysator

Jul 31, 2007, 7:11 AM

Post #6 of 6 (1038 views)

------- You are receiving this mail because: -------
You are the QA contact for the bug.

http://bugs.exim.org/show_bug.cgi?id=467

--- Comment #8 from Magnus Holmgren <holmgren@lysator.liu.se> 2007-07-31 15:11:22 ---
> --- Comment #7 from Peter Bowyer <peter@bowyer.org> 2007-07-30 21:55:25
> > Otherwise we can just check wether sender_address is empty and pass
> > sender_helo_name in place of it.
>
> As long as sender_helo_name is non-empty. I think I had to do something
> about that case, like setting it at HELO time - can't quite remember, it
> was a while back and I don't have access to the system with the source on
> just now.

But if both sender_helo_name and sender_address are empty (no HELO/EHLO and
then MAIL FROM:<>) there's not much to do, is there? Exim could simulate an
SPF_RESULT_NEUTRAL or SPF_RESULT_NONE, perhaps.

Regarding "err_perm" vs "unknown (permanent error)" I think maybe it's best if
Exim provided its own string representations of the result codes, instead of
calling SPF_strresult().

--
Configure bugmail: http://bugs.exim.org/userprefs.cgi?tab=email

--
## List details at http://lists.exim.org/mailman/listinfo/exim-dev Exim details at http://www.exim.org/ ##