Oliver Cowdery corrected scribe 2’s text here in 𝓟 by inserting the subject pronoun he before was filled in verse 22. The question is whether Oliver added the he in verse 22 to make the text agree with the original manuscript (no longer extant here) or whether Oliver’s he was an attempt to edit the text, perhaps because the two immediately following clauses have the pronoun he: “and he blessed Amulek and his house and he gave thanks unto God”. But Oliver did not change the following instance (in verse 23) of “after he had eat and was filled” by adding he there, yet in that case the immediately following clause has the pronoun he (“he saith unto Amulek”). Ultimately, the 1830 typesetter ignored Oliver’s supralinearly inserted he in verse 22, setting simply “and was filled” there (as well as in verse 23).
Elsewhere, after referring to eating, the text has two instances of “and was/were filled”—and in both cases the subject is ellipted before the conjoined predicate:
Further, the King James Bible is full of examples of the form “did eat and were filled”—and without any stated subject before the predicate “were filled”:
Most probably, the 1830 typesetter was influenced by the familiarity of the language from the synoptic Gospels (especially the stories of the feeding of the five thousand and the four thousand) when he decided to ignore Oliver Cowdery’s extra he. Yet in Alma 8:22 the text actually reads “Alma ate bread”, not “Alma did eat”; thus the expected “and was filled” is not as automatic as one might suppose.
Ultimately, there would not have been any strong motivation for Oliver Cowdery to consciously emend “and was filled” in Alma 8:22 to the unexpected “and he was filled”. The familiar biblical usage and the language in Alma 8:23 led scribe 2 of 𝓟 and the 1830 typesetter to omit the subject in “and was filled”. The most reasonable solution in Alma 8:22 is that 𝓞 itself read “and he was filled”. Thus the critical text will accept Oliver’s correction here as the original reading.
Summary: Restore Oliver Cowdery’s inserted he in Alma 8:22, the corrected reading in 𝓟 and the probable reading in 𝓞 (no longer extant here).