> According to https://www.ordabok.is/index.asp the dative
> sg of leikur is leik so that would explain it. I guess, it
> must be what Gordon calls an i-stem noun, and so declines
> like staðr and bekkr (his examples), where the dative
> singular is the same as the accusative.

On the other hand, the comment at the end of

(http://skaldic.abdn.ac.uk/m.php?p=verse&i=3190&v=i)

says that <-leikr> is a masculine a-stem, with regular
dative singular <-leiki> ‘unless it is either an a-stem
without a dat. ending [...], or an earlier i-stem (cf. Goth.
<laikins> m. acc. pl., i-stem)’. And it appears that the
dative singular of Faroese <leikur> is <leiki>.

It appears that it’s not certain whether the PGmc. etymon
was an a-stem (*laika-) or an i-stem (*laiki-); my guess is
that the ON noun began as an a-stem and was at some point
assimilated to the i-stems. And I’ve now thought to check
CV, which for some reason I didn’t do before: ‘dat. leiki is
freq. in poetry and old prose, but mod. leik; plur. in old
usage always leikar, even in late MSS. such as Fb. iii. 405;
but in mod. usage leikir, acc. leiki’.

At any rate, it appears that both <leiki> and <leik> are
possible for the ON dative singular, depending on the date
and manuscript.

Brian