First we must qualify the machlokes. The gemara just finished establishing that the one who houses the eiruv doesn't need to contribute to it because כולהו הכא דיירי, clearly implying that eiruv works משום דירה, as if everyone lived in the house that houses the eiruv. Yet, one line later the gemara develops a machlokes whether eiruv is משום קנין or משום דירה. Furthermore, the gemara 48b says אם אמרו דיורין להקל יאמרו דיורין להחמיר, implying once again that to be meikel we consider eiruv to be "dirah", not a kinyan.
The Gaon Yaacov approaches the sugya (according to Rashi) that all agree in concept that the result of the eiruv is to consider everyone living in the place that houses the eiruv. The machlokes in the gemara is the method to achieve that result, kinyan or dirah. Shmuel holds that a kinyan must be done so that the place which is housing the eiruv is מקנה his reshus to the others in the chatzer to consider it as if they are all living together in this house, whereas Rabba holds that without the method of kinyan, by simply placing their bread in this house it is considered to be as if they are all living in this house. Therefore, all agree that whatever method is used, the result is כולהו הכא דיירי, it is as if all are living in this house.
The Gaon Yaacov points out (in his cryptic style) a fascinating and fundamental difference between Rashi and Tosafos. A careful reading of Tosafos d.h. Eiruv, implies that the "kinyan" we are speaking about does NOT mean that the owner of the house sells rights by the receiving of the bread from the others in the chatzer (rashi). Rather, they are all makneh their reshuyos to each other to create a partnership in all the houses of the chatzer. According to this approach, there is a big conceptual difference as to how eiruv works whether it is kinyan or dirah. Shmuel who holds that it is kinyan is saying that they must have joint ownership in the entire chatzer, whereas Rabba who holds that it is dirah would still consider it as if all the people of the chatzer were living in one house.
The gemara makes it clear that even if the eiruv works as a kinyan, and money should be the ideal thing to use for a kinyan, the Rabbonon didn't want to allow it to be used even bidieved because they were afraid that people would think that money is the primary, forget that bread can be used, and since money isn't always readily available the concept of eiruv would be forgotten. Yet, in the very next line the gemara says that according to Shmuel who holds that an eiruv is משום קנין, a vessel can be used to create the kinyan. Apparently the gemara holds that although money cannot be used since it isn't as common, a vessel can be used since everyone has some kind of kli which can be used for a kinyan.
Tosafos asks that according to Shmuel that Eiruv works as a kinyan, why when using bread do we insist on enough food for two seudos; as long as there is a שוה פרוטה worth in the bread, it should be sufficient? Tosafos supports this question by saying that even a kli is usable for an eiruv according to this opinion, even though it is not eatable. Therefore, bread worth a peruta but less than 2 meals worth should also be usable? It seems to me that in this question Tosafos is arguing with Rashi. Rashi assumed that the use of a kli was serving as a kinyan sudar, not as a kinyan kesef. Rashi seems to understand that when we disallow the use of money for eiruv, we are essentially saying that we don't allow anything to be used בתורת דמים, except for bread. Therefore, when the gemara permits a kli, it must be permitting the use of a kli as a kinyan sudar, not בתורת דמים. According to this approach, it wouldn't make sense to prove that bread that is worth a peruta should work even when it doesn't contain 2 seudos from the fact that a kli an work. When a kli is working it is not at all working בתורת דמים, rather בתורת קנין סודר, and food cannot work as a sudar. Therefore, Tosafos seems to understand that when we permit the use of a kli, we are permitting a kli to be used even בתורת דמים, therefore food should also work b'toras damim even when it doesn't contain 2 seudos worth.
Tosafos answer to this question is difficult to understand. Tosafos writes that when you specify that the eiruv is being done as a kinyan, you don't need two seudos worth, only שוה פרוטה. But when it is done סתם without any specification it requires חשיבות אוכל which essentially means that it must contain two seudos worth. Does Tosafos mean to say that even Shmuel who holds that eiruv can work as a kinyan would hold that it can also work as a dirah? That doesn't seem to be the case because the gemara insists that according to Shmuel the bread be worth a peruta, and less wouldn't work even if it is not being done explicitly b'toras kinyan. So what does Tosafos mean to say?