The matter is whether four words in the Greek should or should not be included. The Greek phrase in question is: "kai apo melissiou keriou" which is literally, "and from a beehive a honecomb".
The UBS5 ranks the omission of the phrase as {B} meaning they were fairly confident (but not certain) that it should be omitted. The evidence cited includes:
OMISSION: P75, 01, 02, 03, 05, 019, 032, 579, it(d, e), syr(s), cop(sa, bo-pt), Clemment, Origen, Cyril(1/2), Ps-Athanasius, Ausgustine(1/4). Compiled GNT: WH, NA4, NA27, NA28, UBS4, UBS5, Souter, NIV, THGNT, SBL.
INCLUSION: 037, 044, f1, 28, 33, 180, 205, 565, 579, 700, 892, 1006, 1010, 1071, 1241, 1292, 1342, 1424, 1505, 07, 013, 022, some lectionaries, it(b, q), syr(c, p, h), cop(bo-pt), eth, Justin(??), Amphilochius, Epiphanius, Cyril(1/2), 07(*), 038, f13, 157, 1243, L253, some italas, + a few more. Compiled GNT: Majority Text, Byzantine text, F35, Orthodox-Patriarchal Text, Textus Receptus. Both Jerome's Vulgate (400 AD) and the Clementine Vulgate text (1592) have "et favum mellis" = and honeycomb.
Thus, the earliest MSS omit and later MSS (from about 400 onwards) appear to add it. However, omitting this text in some MSS continued well into the high middle ages.
This another of those textual differences that are interesting but does not affect any teaching of Scripture; that is, whether it is included or omitted does not appear to matter.
Lastly, since most modern translations use NA28/UBS5, the phrase is omitted. The obvious exceptions are those that follow the Textus Receptus, namely the KJV and its progeny, where the phrase is included.