On Fri, 22 Nov 2024 18:31:11 GMT, Artur Barashev <abaras...@openjdk.org> wrote:
>> test/jdk/sun/security/ssl/SSLCipher/SSLSocketNoServerHelloClientShutdown.java
>> line 74:
>>
>>> 72: float timeoutFactor = Float.parseFloat(
>>> 73: System.getProperty("test.timeout.factor", "1.0"));
>>> 74: serverReadTimeout = (int) (30000 * timeoutFactor);
>>
>> My suggestion was to remove the SoTimeout on both the client and server,
>> also including the one now at line 95.
>
> This is just a fallback mechanism right now, otherwise the test can hang for
> a very long time if something goes wrong. If 30 seconds is not enough for the
> test to run then we fail.
I still think two sync mechs (one to start the server read and one to start the
client shutdown) would be more deterministic/reliable, but I'll approve if you
really want to keep it this way.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/22263#discussion_r1854776445
