There's still something I don't quite get yet: If my app is a server and it waits for clients to connect, then the first operation of the socket will be a read by the server. So if my server app makes a select call on the socket, and I/O is detected on the socket, does it matter whether or not the data is for the SSL layer or the application layer? Does PR_Read handle that? Why would PR_Read block if there's I/O waiting to be read???
-- POC "John Gardiner Myers" <[EMAIL PROTECTED]> wrote in message ajgb8r$98j$[EMAIL PROTECTED]">news:ajgb8r$98j$[EMAIL PROTECTED]... > > > POC wrote: > > Are you saying there needs to be more to the above logic in order to > > use NSPR calls after a select call? > > Yes. Sometimes a PR_Recv() on a PRFileDesc results in a write() on the > underlying socket. Sometimes when that write() returns EWOULDBLOCK, the > only way the connection will make further progress would be for the > application to wait until the underlying socket becomes writeable and > then call PR_Recv() (or, as it happens, PR_Send()). >
