According to both dictionaries, from SUM (gal₄ 'Scham, Vulva' [IK 301]). Cf. also maḳlālu, a name for female genitals, SB [CAD 251]; according to [AHw 607], 'Geringes', vulva (i.e. interpeted as a form derived from ḳalālu 'to be thin, small'), which seems rather a folk etymology

    Related with metathesis and a meaning shift?

    According to Leslau, a reconstructed form of AMH ḳʷǝla (OLD AMH ḳʷǝlḥa) borrowed, in its turn, from CUSH

    Note the ʔVn- prefix

    On -r- <*-l- see [LGur XLVIII]

    Cf. ARB ḥlḳ 'avoir la verge écorchée et rouge à la suite du coit' [BK 1 481] (related with metathesis and a meaning shift?); cf. also MGR ḳǝlwä 'testicles' [LGur 70] (note that in MGR dialects ḥ is normally preserved).

    Cf. ETH forms with somewhat unusual phonetic development: SEL ūnḳulābčä [LGur 70], WOL mʷanḳulabče 'testicles' [ibid. 411]; MSQ ǝnḳulalčä do. [ibid. 70] (on -č < -t see [ibid. LVIII-LXII]; on "w becoming b through pseudo-correction or false reconstruction" see [ibid. LXXXII]).

    Note HAR kūr 'testicles' [LHar 93]; according to Leslau, from CUSH, which is probable in view of irregular k- and -r.

    Cf. SOQ míḳliz 'penis' [LS 375]

    See Abdokov 1983, 123.

    The etymological relation of PAT *ħʷǝ and PAK *χʷǝ is quite transparent (cf., e.g., the same combination of meanings in the reflexes of the related PL *χ:Iol), and the attempts to relate PAK *χʷǝ to the verb *χʷǝ- 'to ripen, to become' (Яковлев 1948, 180, Shagirov 1977,2,112) are obviously folk-etymological. Ub. χʷǝ (compared with PAK *χʷǝ by Abdokov 1973, 73) may be an Ad. loan (which is the idea of T. Esenč, see Vogt 1963, 209, and of Shagirov 1977,2,112), but may well also be genuinely related (the phonetic correspondences are quite regular).