As I read Hawting on “The Origin of Jedda”, I’m pleased to learn that Yanbu‘ is attested by ancient writers as Iambia/Lambia; see Ptolemy’s map (B3). 
That is to say, both Yathrib and its port Yanbu‘ are known to ancient writers; but neither Mecca nor its port Jedda are so attested.
From analogy with Macoraba, which was meant to be Mecca, some scholars have argued that Pliny’s ‘portus Mochorbae’ (port of *Mochorba?) was Jedda;
but Pliny, citing Juba, puts this port in the south-east of the Peninsula, which seems to rule it out. (Compare ‘Dabanegoris regio’: iandavidmorris.com/mecca-before-i…)
Aloys Sprenger thought Jedda was the Gadda mentioned by Late Antique sources, but this too is in the wrong place.
Stephanus Byzantinus says: “Gadda, a town of Arabia [according to] Glaukos, [book] 4. [Grammatically] feminine and neuter. Ethnonym: ‘Gaddēnos’.”
This Glaukos, otherwise unknown, is cited a few places by Stephanus; always it seems for towns in the northwest of Arabia, roughly Nabataea.
This is confirmed by the Tabula Peutingeriana, a putatively Roman map existing in medieval copies: Gadda is put 13 miles north of Amman (‘Philadelfia’). tabula-peutingeriana.de/tabula.html?se…
Meanwhile the Notitia Dignitatum, dating around 400 CE, says that “native horse archers” (equites sagittari indigenae) were posted at Gadda;
again, this indicates a northwestern location in the Arabian Peninsula, where the Roman state was active.
(Given how awful I rate Sprenger’s work on Macoraba and Mariaba Baramalacum, I am not at all surprised to see this lazy take on Gadda.)
Gadda’s true location isn’t known precisely, though an identification with Qal‘at Zarqā has been suggested. It is almost certainly in Jordan.
