Hi Milan, Your test need the corresponding "TcpTimeout.dns" file to run successfully, I believe you forgot to add with your patch. please check the existing tests in the same folder if you need any additional information.
Thanks, Vyom On Wed, Aug 21, 2019 at 10:48 PM Milan Mimica <milan.mim...@gmail.com> wrote: > Hi Pavel > > Updated the patch with the jtreg test. > The test hangs when the fix is not applied. I don't know why > main/timeout=20 does not work for me. > > On Tue, 20 Aug 2019 at 00:08, Pavel Rappo <pavel.ra...@oracle.com> wrote: > > > > Thanks for doing that. I've only skimmed through the patch and I’d > recommend that no matter if the changes are adequate or not there should be > a test [1]: > > > > "...A unit test, written for the jtreg test harness, that validates your > change. The test should fail when run against a build without the change > and succeed when run against a build with the change. Unit tests aren't > always practical, especially for changes such as performance improvements > or fixes to bugs that are highly platform-dependent, but if practical then > a unit test is required." > > > > Please consider adding it to the patch while the changes are being > (preliminary) reviewed. > > > > -Pavel > > > > > ------------------------------------------------------------------------------- > > [1] https://openjdk.java.net/contribute/ > > > > > On 19 Aug 2019, at 17:20, Milan Mimica <milan.mim...@gmail.com> wrote: > > > > > > Hello list > > > > > > Please find the attached patch that fixes JDK-8228580. > > > > > > It works the similar way UDP timeout does: calculate the initial > > > timeout from retry attempt, and account for duration of every blocking > > > call on the TCP socket. > > > > > > I am listed as Author[1] on "jdk" project. > > > > > > [1] https://openjdk.java.net/census#mmimica > > > > > > > > > -- > > > Milan Mimica > > > <JDK-8228580.patch> > > > > > -- > Milan Mimica > -- Thanks, Vyom
