On Fri, 18 Sep 2026 12:04:48 GMT, Lee Jiwon <[email protected]> wrote:
>> Before this change, Sockets.supportedOptions(Class<?>) used a hard-coded >> option map created during class initialization. The reported set could >> differ from the options supported by an actual socket instance. >> >> This change creates an instance of the requested socket type for each call >> and returns its supportedOptions() result. The temporary socket is closed >> after use, and jdk.net.Sockets does not cache the result. The hard-coded >> option map and helper code used only to build that map are removed. >> >> Testing: >> >> * make test TEST=test/jdk/jdk/net/Sockets/SupportedOptions.java >> * make test TEST=test/jdk/jdk/net/Sockets >> * make test TEST=jdk_net — 1,082 tests passed >> >> --------- >> - [x] I confirm that I make this contribution in accordance with the >> [OpenJDK Interim AI Policy](https://openjdk.org/legal/ai). > > Lee Jiwon has updated the pull request incrementally with one additional > commit since the last revision: > > 8392528: Use UncheckedIOException for socket creation failures test/jdk/jdk/net/Sockets/SupportedOptions.java line 62: > 60: void invalidSocketTypes() { > 61: assertThrows(IllegalArgumentException.class, () -> > Sockets.supportedOptions(null)); > 62: assertThrows(IllegalArgumentException.class, () -> > Sockets.supportedOptions(String.class)); Sockets.supportedOptions doesn't specify how it handles null to testing long standing behavior is okay. test/jdk/jdk/net/Sockets/SupportedOptions.java line 66: > 64: > 65: @Test > 66: void serverSocketSupportsIpTos() throws Exception { The test is now a unit test for Sockets.supportedOptions so I think this one looks a bit out of places. Is it needed? It looks like SocketOpotion/OptionTest.java tests setting IP_TOS on ServerSOcket. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/32943#discussion_r4047834896 PR Review Comment: https://git.openjdk.org/jdk/pull/32943#discussion_r4047872764
