Re: RFR: 8352623: MultiExchange should cancel exchange impl if responseFilters throws

2025-04-05 Thread Daniel Fuchs
On Fri, 21 Mar 2025 17:29:48 GMT, Daniel Jeliński wrote: > LGTM. Should we handle non-checked exceptions as well? hmm... I believe we should. Will update. - PR Comment: https://git.openjdk.org/jdk/pull/24164#issuecomment-2744180924

Re: RFR: 8352623: MultiExchange should cancel exchange impl if responseFilters throws [v2]

2025-03-25 Thread Daniel Jeliński
On Fri, 21 Mar 2025 19:55:47 GMT, Daniel Fuchs wrote: >> When adding authClient.close() noAuthClient.close() to >> UnauthorizedTest::teardown, the test fails in jtreg timeout. >> Usage of the ReferenceTracker reveals that some streams have been left >> dangling, and this can be tracked to the M

Re: RFR: 8352623: MultiExchange should cancel exchange impl if responseFilters throws [v2]

2025-03-21 Thread Daniel Fuchs
> When adding authClient.close() noAuthClient.close() to > UnauthorizedTest::teardown, the test fails in jtreg timeout. > Usage of the ReferenceTracker reveals that some streams have been left > dangling, and this can be tracked to the MultiExchange not cancelling the > stream when responseFilte

RFR: 8352623: MultiExchange should cancel exchange impl if responseFilters throws

2025-03-21 Thread Daniel Fuchs
When adding authClient.close() noAuthClient.close() to UnauthorizedTest::teardown, the test fails in jtreg timeout. Usage of the ReferenceTracker reveals that some streams have been left dangling, and this can be tracked to the MultiExchange not cancelling the stream when responseFilters throws.

Re: RFR: 8352623: MultiExchange should cancel exchange impl if responseFilters throws

2025-03-21 Thread Daniel Jeliński
On Fri, 21 Mar 2025 16:46:56 GMT, Daniel Fuchs wrote: > When adding authClient.close() noAuthClient.close() to > UnauthorizedTest::teardown, the test fails in jtreg timeout. > Usage of the ReferenceTracker reveals that some streams have been left > dangling, and this can be tracked to the Multi