From bf2006a@yahoo.com  Thu Mar 19 11:55:40 2009
Return-Path: <bf2006a@yahoo.com>
Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34])
	by hub.freebsd.org (Postfix) with ESMTP id 357B41065675
	for <bug-followup@FreeBSD.org>; Thu, 19 Mar 2009 11:55:40 +0000 (UTC)
	(envelope-from bf2006a@yahoo.com)
Received: from web39104.mail.mud.yahoo.com (web39104.mail.mud.yahoo.com [209.191.87.32])
	by mx1.freebsd.org (Postfix) with SMTP id EB98D8FC1E
	for <bug-followup@FreeBSD.org>; Thu, 19 Mar 2009 11:55:39 +0000 (UTC)
	(envelope-from bf2006a@yahoo.com)
Received: (qmail 32139 invoked by uid 60001); 19 Mar 2009 11:55:39 -0000
Received: from [145.100.100.148] by web39104.mail.mud.yahoo.com via HTTP; Thu, 19 Mar 2009 04:55:39 PDT
Message-Id: <544918.31673.qm@web39104.mail.mud.yahoo.com>
Date: Thu, 19 Mar 2009 04:55:39 -0700 (PDT)
From: bf <bf2006a@yahoo.com>
To: jhb@FreeBSD.org, ed@FreeBSD.org, kib@FreeBSD.org, marcus@FreeBSD.org,
  alc@FreeBSD.org, linimon@lonesome.com, bug-followup@FreeBSD.org,
  "O. Hartmann" <ohartman@zedat.fu-berlin.de>
Subject: Re: PRs kern/132620 and ports/132542

>Number:         132807
>Category:       kern
>Synopsis:       Re: PRs kern/132620 and ports/132542
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    gavin
>State:          closed
>Quarter:        
>Keywords:       
>Date-Required:  
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Thu Mar 19 12:00:04 UTC 2009
>Closed-Date:    Thu Apr 16 09:40:42 UTC 2009
>Last-Modified:  Thu Apr 16 09:40:42 UTC 2009
>Originator:     
>Release:        
>Organization:
>Environment:
>Description:
 --- On Wed, 3/18/09, John Baldwin <jhb@freebsd.org> wrote:
 
 > From: John Baldwin <jhb@freebsd.org>
 > Subject: Re: recent regressions on 8-CURRENT
 > To: "bf" <bf2006a@yahoo.com>
 > Date: Wednesday, March 18, 2009, 4:48 PM
 >=20
 > -----Inline Attachment Follows-----
 >=20
 > On Wednesday 18 March 2009 4:11:29
 > pm bf wrote:
 > > --- On Wed, 3/18/09, John Baldwin <jhb@freebsd.org>
 > wrote:
 > > > From: John Baldwin <jhb@freebsd.org>
 > > > Subject: Re: recent regressions on 8-CURRENT
 > > > To: "bf" <bf2006a@yahoo.com>
 > > > Date: Wednesday, March 18, 2009, 1:58 PM
 > > >=20
 > > > -----Inline Attachment Follows-----
 > > >=20
 > > > On Wednesday 18 March 2009 1:40:08
 > > > pm bf wrote:
 > > > >=20
 > > > > I'm facing an elusive problem, described
 > in:
 > > > >=20
 > > > > http://www.FreeBSD.org/cgi/query-pr.cgi?pr=3Dkern/132620
 > > > >=20
 > > > > that has only surfaced in 8-CURRENT in the
 > last few
 > > > weeks, and
 > > > > is crippling my -CURRENT systems.=A0 Any
 > ideas
 > > > about how to=20
 > > > > determine what may be the problem?=A0 Or is
 > my only
 > > > alternative
 > > > > to systematically try different revisions of
 > -CURRENT
 > > > until I
 > > > > find the changeset where things first went
 > awry?
 > > >=20
 > > > I have no idea what the problem is.=A0 Are the
 > two
 > > > instances of 'src/usr.sbin/'=20
 > > > becoming 'srcr.sbin/' typos in your transcripts
 > or is that
 > > > the actual path=20
 > > > reported by make when it fails?
 > > >=20
 > > > --=20
 > > > John Baldwin
 > > >=20
 > >=20
 > > I should mention that my original message is here:
 > >=20
 > > http://lists.freebsd.org/pipermail/freebsd-current/2009-March/004445.ht=
 ml
 > >=20
 > > with more examples.=A0 And besides my amd64 and
 > i386 systems,
 > > at least one other person has noticed the problem.
 > >=20
 > > It's strange -- even in the consistently reproducible
 > cases,
 > > there is a seemingly variable behavior.=A0 For
 > example, in
 > > the devel/icu case I cited in my last follow-up
 > message to the
 > > PR, make successfully processes the command:
 > >=20
 > > cd ../.. \
 > > && CONFIG_FILES=3Dtools/genrb/derb.1
 > CONFIG_HEADERS=3D /bin/sh ./config.status
 > >=20
 > > yielding the output:
 > >=20
 > > config.status: creating tools/genrb/derb.1
 > >=20
 > > but then dies on:
 > >=20
 > > cd ../.. \
 > > && CONFIG_FILES=3Dtools/genrb/genrb.1
 > CONFIG_HEADERS=3D /bin/sh ./config.status
 > >=20
 > > giving:
 > >=20
 > > ./config.status: Can't open ./config.status: No such
 > file or directory
 > >=20
 > >=20
 > > Note that the output gives "config.status" for the
 > successful
 > > command, but "./config.status" for the failed
 > command.=A0 The
 > > paths are usually corrupted around "." or "/"
 > characters,
 > > although this may be sheer coincidence.=A0 In the
 > period where
 > > the change seems to have occurred, das@ made his C
 > library
 > > changes, you and alc@ made some changes to the vm
 > system and
 > > pmap, and you and kib@ changed some vfs stuff and
 > sysctl.
 > > Nothing else seems likely to have caused this
 > problem.
 >=20
 > Perhaps revert svn changes 189697 and 189696?
 >=20
 > --=20
 > John Baldwin
 >=20
 
 
 No, John, you seem to be innocent ... at least in this case.=20
 The offending code seems to be marcus@ and kib@'s new
 VOP_VPTOCNP(9), either on it's own or in conjunction with other
 recent changes.  After reverting r189539-40 and r189622, the
 build error for devel/icu,  which is the most consistently
 reproducible of the errors I have observed lately, no longer
 occurs on amd64 or i386, with otherwise up-to-date kernels and
 worlds.  Nor have any other of the problems resurfaced after
 building ports on a tmpfs /tmp, which seems to be more
 susceptible to these errors than ufs.  I haven't determined the
 exact mechanism of the error.
 
 Regards,
            b.=0A=0A=0A      
>How-To-Repeat:
>Fix:
>Release-Note:
>Audit-Trail:
State-Changed-From-To: open->closed 
State-Changed-By: gavin 
State-Changed-When: Thu Apr 16 09:39:58 UTC 2009 
State-Changed-Why:  
Misfiled followup for two PRs, both of which have since been committed 
and closed 


Responsible-Changed-From-To: gnats-admin->gavin 
Responsible-Changed-By: gavin 
Responsible-Changed-When: Thu Apr 16 09:39:58 UTC 2009 
Responsible-Changed-Why:  


http://www.freebsd.org/cgi/query-pr.cgi?pr=132807 
>Unformatted:
