NetBSD Problem Report #47416

From  Mon Jan  7 14:38:56 2013
Return-Path: <>
Received: from ( [])
	by (Postfix) with ESMTP id 3537363ECDC
	for <>; Mon,  7 Jan 2013 14:38:56 +0000 (UTC)
Message-Id: <>
Date: Mon,  7 Jan 2013 14:38:55 +0000 (UTC)
Subject: Apache segmentation faults/Apache state sokva
X-Send-Pr-Version: www-1.0

>Number:         47416
>Category:       kern
>Synopsis:       Apache segmentation faults/Apache state sokva
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          support
>Submitter-Id:   net
>Arrival-Date:   Mon Jan 07 14:40:07 +0000 2013
>Last-Modified:  Tue May 16 06:10:00 +0000 2017
>Originator:     Al Zick
>Release:        NetBSD 6
Zick's web ventures
NetBSD 6.0 NetBSD 6.0 (FAITH-$Revision: 1.777.2.6 $) #0: Thu Jan  3 21:45:52 EST 2013 macppc
When apache processes exit, they cause a Segmentation fault because they can not free memory. It looks like the problem is a kernel problem with the socket buffers. After maybe 24 hours the server stops responding to httpd requests because apache is in a state of sokva. If I build a kernel with "options SOSEND_NO_LOAN", I don't get a state of sokva, but the system goes down after a few days. It doesn't kernel panic, but will not respond to anything, except one of the strange effects of this is that I get intermittent pings. 
Start apache loading the php module on a NetBSD 6 macppc system.



Responsible-Changed-From-To: kern-bug-people->pkg-manager
Responsible-Changed-When: Tue, 08 Sep 2015 09:07:38 +0000
misfiled PR - apache issues are not kernel bugs

Responsible-Changed-From-To: pkg-manager->kern-bug-people
Responsible-Changed-When: Fri, 05 May 2017 07:39:02 +0000
This is a kernel bug. Also have no reason to believe it was solved.

From: David Holland <>
Subject: Re: kern/47416: Apache segmentation faults/Apache state sokva
Date: Tue, 16 May 2017 06:07:59 +0000

 Not sent to gnats.


 From: Martin Husemann <>
 Subject: Re: kern/47416: Apache segmentation faults/Apache state sokva
 Date: Fri, 5 May 2017 10:35:46 +0200

 I don't think this is a kernel bug (or at least it is not clear).

 Try bumping sysctl kern.somaxkva up quite a bit and check if the "sokva"
 wait state goes away.

 The segfaults are userland and completely unrelated, it would be best
 to create two separate PRs if either problem is still reproducable.



NetBSD Home
NetBSD PR Database Search

(Contact us) $NetBSD: query-full-pr,v 1.39 2013/11/01 18:47:49 spz Exp $
$NetBSD:,v 1.8 2006/05/07 09:23:38 tsutsui Exp $
Copyright © 1994-2014 The NetBSD Foundation, Inc. ALL RIGHTS RESERVED.