RFR: 8263779: SSLEngine reports NEED_WRAP continuously without producing any further output [v2]

[Bradford Wetmore]

On Wed, 28 Apr 2021 00:12:44 GMT, Xue-Lei Andrew Fan <xuelei at openjdk.org> wrote:

>> src/java.base/share/classes/sun/security/ssl/SSLSocketImpl.java line 1694:
>> 
>>> 1692:         if (cause instanceof SocketException) {
>>> 1693:             try {
>>> 1694:                 throw conContext.fatal(alert, cause);
>> 
>> Why did you change to a throw here?  Isn't everything from fatal() going to be thrown anyway?
>
> The fatal() will always throw an exception.  We could use the method without throw.  In some places, IDE or compiler could complain because it cannot detect that the fatal() method will throw and the program ends here.  In the past, we use fatal() method without throw, later we change the fatal() return value form "void" to an exception so that throw could be used to mitigate the IDE or compiler complaints.
> 
> It is not really necessary to make this code change here.  I just want to make the coding style consistent for the fatal() calling.

Interesting, ok.  I don't recall seeing such situations with the IDE I often use.  (Netbeans)

-------------

PR: https://git.openjdk.java.net/jdk/pull/3292