postgres.git / summary / log / commit / refs

commit    bf2617981ceb88bdcd3cba58308d68f0c4ea3de2
Author:   Alvaro Herrera <alvherre@alvh.no-ip.org>
Date:     Sun Oct 06 02:24:50 2013 +0000

    Fix various bugs in postmaster SIGKILL processing
    
    Clamp the minimum sleep time during immediate shutdown or crash to a
    minimum of zero, not a maximum of one second.  The previous code could
    result in a negative sleep time, leading to failure in select() calls.
    
    Also, on crash recovery, reset AbortStartTime as soon as SIGKILL is sent
    or abort processing has commenced instead of waiting until the startup
    process completes.  Reset AbortStartTime as soon as SIGKILL is sent,
    too, to avoid doing that repeatedly.
    
    Per trouble report from Jeff Janes on
    CAMkU=1xd3=wFqZwwuXPWe4BQs3h1seYo8LV9JtSjW5RodoPxMg@mail.gmail.com
    
    Author: MauMau


 src/backend/postmaster/postmaster.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 01d2618ebdb..9f721b76515 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -1422,9 +1422,9 @@ DetermineSleepTime(struct timeval * timeout)
 	{
 		if (AbortStartTime > 0)
 		{
-			/* remaining time, but at least 1 second */
-			timeout->tv_sec = Min(SIGKILL_CHILDREN_AFTER_SECS -
-								  (time(NULL) - AbortStartTime), 1);
+			/* time left to abort; clamp to 0 in case it already expired */
+			timeout->tv_sec = Max(SIGKILL_CHILDREN_AFTER_SECS -
+								  (time(NULL) - AbortStartTime), 0);
 			timeout->tv_usec = 0;
 		}
 		else
@@ -1676,10 +1676,13 @@ ServerLoop(void)
 		 * Note we also do this during recovery from a process crash.
 		 */
 		if ((Shutdown >= ImmediateShutdown || (FatalError && !SendStop)) &&
+			AbortStartTime > 0 &&
 			now - AbortStartTime >= SIGKILL_CHILDREN_AFTER_SECS)
 		{
 			/* We were gentle with them before. Not anymore */
 			TerminateChildren(SIGKILL);
+			/* reset flag so we don't SIGKILL again */
+			AbortStartTime = 0;
 
 			/*
 			 * Additionally, unless we're recovering from a process crash, it's
@@ -2584,7 +2587,7 @@ reaper(SIGNAL_ARGS)
 			 * Startup succeeded, commence normal operations
 			 */
 			FatalError = false;
-			AbortStartTime = 0;
+			Assert(AbortStartTime == 0);
 			ReachedNormalRunning = true;
 			pmState = PM_RUN;
 
@@ -3544,6 +3547,8 @@ PostmasterStateMachine(void)
 		StartupPID = StartupDataBase();
 		Assert(StartupPID != 0);
 		pmState = PM_STARTUP;
+		/* crash recovery started, reset SIGKILL flag */
+		AbortStartTime = 0;
 	}
 }
 
@@ -4737,7 +4742,7 @@ sigusr1_handler(SIGNAL_ARGS)
 	{
 		/* WAL redo has started. We're out of reinitialization. */
 		FatalError = false;
-		AbortStartTime = 0;
+		Assert(AbortStartTime == 0);
 
 		/*
 		 * Crank up the background tasks.  It doesn't matter if this fails,


[parent: 2d6c0f10ef4f]