In this situation, the "survival days" for all individuals is the difference between the last observation date and the date of entry into the study. To that extent, you are correct in your calculation for Patient B. Remember, however, that all you know is that Patient B survived at least that long. Thus that is a right-censored observation.
What's important is to include for each patient, along with those "survival day" values, an indicator of whether there was an event at the end of that time period (Patient A) or if the individual had not yet experienced the event (Patients B and C). In the R survival package, that's done with a
Surv() object that, in its simplest form, combines the survival time with the event indicator.