Re: [PATCH] nss_dns: Do not replace root domain with empty string

[Florian Weimer <fweimer at redhat dot com>]

* Carlos O'Donell:

> On 4/8/19 7:18 AM, Florian Weimer wrote:
>> The purpose of the bp[0] == '.' check is unclear.  Only the root domain
>> starts with '.'.  An empty string is not a valid domain name, so the
>> subsequent res_dnok check fails.
>
> Was the intent to *cause* a failure if a root domain was unpacked?
>
> Are you certain res_dnok() fails with the empty string? It looks to me
> like ns_name_pton() might work, and so would printable_string(), but
> the result is obviously garbage, and may fail later.
>
> Logically the code in question doesn't make any sense from first
> principles, therefore I think the deletion is warranted, and that
> we'll fix any fallout.

Hmm, you are right, res_dnok et al. accept it, which is something we
cannot change for backwards compatibility reasons.  Odd.  There is even
a test for it.

I'm going to use this as the commit message then:

    nss_dns: Do not replace root domain with empty string
    
    The purpose of the bp[0] == '.' check is unclear.  Only the root domain
    starts with '.'.  The empty string is accepted as a domain name in many
    places, denoting the root, but using it implicitly is confusing.

Still okay?

Thansk,
Florian