Fiber*: cleanup scheduling, fix keepalive

Both FiberSpawn and FiberPool share similar main loops, the
only difference being the handling of connection acceptance.
So move the scheduler into it's own function for consistency.

We'll also correctly implement keepalive timeout so clients
get disconnected at the right time.

1 files changed, 42 insertions, 0 deletions

diff --git a/t/t0010-keepalive-timeout-effective.sh b/t/t0010-keepalive-timeout-effective.sh
new file mode 100755
index 0000000..9d4d651
--- /dev/null
+++ b/t/t0010-keepalive-timeout-effective.sh
@@ -0,0 +1,42 @@
+#!/bin/sh
+. ./test-lib.sh
+t_plan 6 "keepalive_timeout tests for $model"
+
+t_begin "setup and start" && {
+        rainbows_setup
+        rainbows -D env.ru -c $unicorn_config
+        rainbows_wait_start
+}
+
+t_begin 'check server up' && {
+        curl -sSf http://$listen/
+}
+
+t_begin "send keepalive response that does not expect close" && {
+        req='GET / HTTP/1.1\r\nHost: example.com\r\n\r\n'
+        t0=$(date +%s)
+        (
+                cat $fifo > $tmp &
+                printf "$req"
+                wait
+                date +%s > $ok
+        ) | socat - TCP:$listen > $fifo
+        now="$(cat $ok)"
+        elapsed=$(( $now - $t0 ))
+        t_info "elapsed=$elapsed (expecting >=5s)"
+        test $elapsed -ge 5
+}
+
+t_begin 'keepalive not unreasonably long' && {
+        test $elapsed -lt 15
+}
+
+t_begin "killing succeeds" && {
+        kill $rainbows_pid
+}
+
+t_begin "check stderr" && {
+        check_stderr
+}
+
+t_done