Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 28 Dec 1998 01:22:33 +0300
From:      Dmitrij Tejblum <dima@tejblum.dnttm.rssi.ru>
To:        garman@earthling.net
Cc:        current@FreeBSD.ORG
Subject:   Re: PR filed on dying daemons bug 
Message-ID:  <199812272222.BAA00733@tejblum.dnttm.rssi.ru>
In-Reply-To: Your message of "Sun, 27 Dec 1998 01:10:03 EST." <199812270610.WAA15223@hub.freebsd.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
> I upgraded to a recent -current (as of two days ago) and found the
> dying daemons bug still exists :(
> 
> I filed a PR on the subject since I couldn't find another which already
> addressed the issue.  It's kern/9195.  I'm more than willing to give
> people accounts, test patches, etc for the cause of trying to fix this
> bug, as myself as well as many others are still experiencing it on a
> regular basis.

I once suggested this patch:

--- sys/vm/swap_pager.c	Mon Dec 28 00:21:12 1998
+++ sys/vm/swap_pager.c	Mon Dec 28 00:21:52 1998
@@ -1130,4 +1130,5 @@ swap_pager_getpages(object, m, count, re
 		m[reqpage]->object->last_read = m[count-1]->pindex;
 
+#if 0
 		/*
 		 * If we're out of swap space, then attempt to free
@@ -1144,4 +1145,5 @@ swap_pager_getpages(object, m, count, re
 				m[0]->pindex + paging_offset, count);
 		}
+#endif
 
 	} else {

I remember 1 positive response on it, and 0 negative responses. Feel 
free to try this patch.

As a side note, IMO the code (that the patch ifdef out) is not a very 
good idea, whether it is related to the "daemon dying" bug or not. There 
is usually some amount of allocated but unused swap space (20-50%, I 
think), and this space apparently left allocated but unused forever, 
and the code free used swap space instead....

Dima



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199812272222.BAA00733>