Collin Funk wrote: > only one failure for test-nonblocking-socket.sh. But I assume that this > is not an actual bug since we intentionally disable this test in CI. Is > it just the guess for SOCKET_DATA_BLOCK_SIZE's definition doesn't match > the real kernel buffer size?
It depends on the messages that you see in the .log file. Each of these messages has a comment, in the source code, indicating what is the likely cause. Bruno
