From owner-freebsd-current Thu Jul 19 20:57:10 2001 Delivered-To: freebsd-current@freebsd.org Received: from wint.itfs.nsk.su (wint.itfs.nsk.su [212.20.32.43]) by hub.freebsd.org (Postfix) with ESMTP id C439437B405 for ; Thu, 19 Jul 2001 20:57:04 -0700 (PDT) (envelope-from nnd@wint.itfs.nsk.su) Received: (from nnd@localhost) by wint.itfs.nsk.su (8.11.4/8.11.4) id f6K3v2518785; Fri, 20 Jul 2001 10:57:02 +0700 (NOVST) (envelope-from nnd) Date: Fri, 20 Jul 2001 10:57:02 +0700 (NOVST) Message-Id: <200107200357.f6K3v2518785@wint.itfs.nsk.su> From: nnd@mail.nsk.ru To: current@freebsd.org Subject: Re: -current kernel hangs? In-Reply-To: User-Agent: tin/1.5.9-20010630 ("Blue Water") (UNIX) (FreeBSD/5.0-CURRENT (i386)) Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG In article you wrote: > On Thu, 19 Jul 2001, David Wolfskill wrote: > >> >Date: Thu, 19 Jul 2001 02:48:10 -1000 (HST) >> >From: Vincent Poy >> >> > With a July 18, 2001 sources, it seems like the kernel hands at >> >the entropy harvesting stage.... ctrl-t shows: >> >> >load : 098 cmd : sycctl 51 [running] 4.51u 210.37s 0% 172k >> >> > It will just sit there forever until ctrl-c is hit. Anyone knows >> >what's wrong? Thanks. >> >> This was discussed (to some extent) aboust a week & a half ago in -current. >> It seems (pointed out by Alexander Leidinger ) >> that -- for some of us, at least -- "sysctl -a" (used in the entropy >> harvesting by /etc/rc -- fails to terminate. >> >> Also (again, for me) "sysctl -N -a" outputs a (non-terminating) stream of >> >> net.inet.accf.373 >> >> lines using my (slightly customized) kernel, but with GENERIC, it exhibits >> similar behavior, but outputs a stream of >> >> net.inet.accf.372 >> >> lines. >> >> >> Unfortunately, I'm not sufficiently clueful to figure out how the notion >> of what sysctls exist on the system could get so confused as to do that. >> >> Given the behavior, though, my current (wild) guess is that some code is >> stomping on a data structure in a somewhat configuration-dependent way. >> >> Clues would be quite welcome.... > > Interesting... I just didn't see anyone mention it this week so I > thought it was fixed already but is there a work around this so ctrl-c > doesn't have to be hit on each reboot? I have some kind of "workaround" (see the patch). In my config without this patch 'sysctl -a' hungs at net.inet.accf and with this patch it successfully run up to the end. I can not see what is wrong with the lines disabled by the patch. N.Dudorov Index: sys/kern/uipc_accf.c =================================================================== RCS file: /scratch/CVS/src/sys/kern/uipc_accf.c,v retrieving revision 1.6 diff -b -u -r1.6 uipc_accf.c --- sys/kern/uipc_accf.c 2001/06/01 21:47:34 1.6 +++ sys/kern/uipc_accf.c 2001/07/19 05:12:01 @@ -49,10 +49,12 @@ static int unloadable = 0; +#if 0 SYSCTL_DECL(_net_inet); /* XXX: some header should do this for me */ SYSCTL_NODE(_net_inet, OID_AUTO, accf, CTLFLAG_RW, 0, "Accept filters"); SYSCTL_INT(_net_inet_accf, OID_AUTO, unloadable, CTLFLAG_RW, &unloadable, 0, "Allow unload of accept filters (not recommended)"); +#endif /* * must be passed a malloc'd structure so we don't explode if the kld To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message