NetBSD Problem Report #49153

From gson@gson.org  Tue Aug 26 06:48:20 2014
Return-Path: <gson@gson.org>
Received: from mail.netbsd.org (mail.netbsd.org [149.20.53.66])
	(using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits))
	(Client CN "mail.netbsd.org", Issuer "Postmaster NetBSD.org" (verified OK))
	by mollari.NetBSD.org (Postfix) with ESMTPS id 6E363AF4B7
	for <gnats-bugs@gnats.NetBSD.org>; Tue, 26 Aug 2014 06:48:20 +0000 (UTC)
Message-Id: <20140826064810.1371675E2E@guava.gson.org>
Date: Tue, 26 Aug 2014 09:48:10 +0300 (EEST)
From: gson@gson.org (Andreas Gustafsson)
Reply-To: gson@gson.org (Andreas Gustafsson)
To: gnats-bugs@gnats.NetBSD.org
Subject: Recent regression of nfs_autoload, modautoload test cases
X-Send-Pr-Version: 3.95

>Number:         49153
>Category:       bin
>Synopsis:       Recent regression of nfs_autoload, modautoload test cases
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    pgoyette
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Aug 26 06:50:00 +0000 2014
>Closed-Date:    Fri Jan 22 10:16:05 +0000 2016
>Last-Modified:  Fri Jan 22 10:16:05 +0000 2016
>Originator:     Andreas Gustafsson
>Release:        NetBSD-current
>Organization:

>Environment:
System: NetBSD
Architecture: i386
Machine: i386
>Description:

Two test cases recently started failing:

  lib/librumphijack/t_tcpip/nfs_autoload
  rump/modautoload/t_modautoload/modautoload

Log output from the failing tests is at:

  http://releng.netbsd.org/b5reports/i386/build/2014.08.25.14.11.51/test.html#lib_librumphijack_t_tcpip_nfs_autoload
  http://releng.netbsd.org/b5reports/i386/build/2014.08.25.14.11.51/test.html#rump_modautoload_t_modautoload_modautoload

The failures appeared during a period of 47 hours when the tree did
not build, from CVS date 2014.08.23.15.02.04 until
2014.08.25.14.11.51.  During this time, 395 new file revisions were
committed by 21 developers (apb, christos, dholland, jmcneill,
jnemeth, joerg, justin, martin, matt, maxv, mlelstv, mrg, msaitoh,
nonaka, palle, pooka, riastradh, rmind, skrll, tsutsui, wiz).

>How-To-Repeat:

Run the ATF tests.

>Fix:

>Release-Note:

>Audit-Trail:

Responsible-Changed-From-To: bin-bug-people->pgoyette
Responsible-Changed-By: pgoyette@NetBSD.org
Responsible-Changed-When: Sun, 27 Dec 2015 11:23:41 +0000
Responsible-Changed-Why:
I've fixed 1/2 of this, and working on the other half.


State-Changed-From-To: open->analyzed
State-Changed-By: pgoyette@NetBSD.org
State-Changed-When: Sun, 27 Dec 2015 11:23:41 +0000
State-Changed-Why:
Problem is understood.


From: Paul Goyette <paul@whooppee.com>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: bin/49153: Recent regression of nfs_autoload, modautoload test
 cases
Date: Sun, 27 Dec 2015 19:22:15 +0800 (PHT)

 The modautoload portion of this PR should be resolved by rev 1.4 of 
 src/tests/rump/modautoload (committed earlier today by myself).

 I'm looking into the nfs_autoload failure - it is likely a result of the 
 same issue (rump-kernels are currently defaulting to having module 
 auto-load disabled).

Responsible-Changed-From-To: pgoyette->bin-bug-people
Responsible-Changed-By: pgoyette@NetBSD.org
Responsible-Changed-When: Mon, 28 Dec 2015 06:10:40 +0000
Responsible-Changed-Why:
There is more needed on this PR than simply enabling module-autoload.  I
don't have enough rump-clue to make further forward progress.


State-Changed-From-To: analyzed->open
State-Changed-By: pgoyette@NetBSD.org
State-Changed-When: Mon, 28 Dec 2015 06:10:40 +0000
State-Changed-Why:
Problem is bigger than previous expectation.


From: "Paul Goyette" <pgoyette@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/49153 CVS commit: src/tests/lib/librumphijack
Date: Tue, 29 Dec 2015 07:17:19 +0000

 Module Name:	src
 Committed By:	pgoyette
 Date:		Tue Dec 29 07:17:19 UTC 2015

 Modified Files:
 	src/tests/lib/librumphijack: t_tcpip.sh

 Log Message:
 Since we're expecting to autoload modules here, make sure that the
 kernel (or rump-server) has autoload enabled.

 This should finish fixing PR bin/49153

 XXX The test still does not run successfully, due to another bug that
 XXX was recently introduced.  This second bug is being worked on.


 To generate a diff of this commit:
 cvs rdiff -u -r1.15 -r1.16 src/tests/lib/librumphijack/t_tcpip.sh

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

Responsible-Changed-From-To: bin-bug-people->pgoyette
Responsible-Changed-By: pgoyette@NetBSD.org
Responsible-Changed-When: Tue, 29 Dec 2015 07:36:20 +0000
Responsible-Changed-Why:
I'm fixing both this original bug and a subsequent bug that was hidden
by this first one.


State-Changed-From-To: open->analyzed
State-Changed-By: pgoyette@NetBSD.org
State-Changed-When: Tue, 29 Dec 2015 07:36:20 +0000
State-Changed-Why:
Bothoriginal and new bugs are understood.


From: "Paul Goyette" <pgoyette@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/49153 CVS commit: src/sys/rump/librump/rumpkern
Date: Tue, 29 Dec 2015 10:22:06 +0000

 Module Name:	src
 Committed By:	pgoyette
 Date:		Tue Dec 29 10:22:06 UTC 2015

 Modified Files:
 	src/sys/rump/librump/rumpkern: emul.c

 Log Message:
 Now that the table of auto-loadable syscalls is per-emulation, make sure
 that the rump-kernel has its own list.  Otherwise, missing syscalls won't
 trigger a module auto-load.

 This commit finishes the work to get tests/lib/librumphijack/t_tcpip
 nfs_autoload test case working again after 16 months of failures!  (see
 PR bin/49153).


 To generate a diff of this commit:
 cvs rdiff -u -r1.173 -r1.174 src/sys/rump/librump/rumpkern/emul.c

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

From: Andreas Gustafsson <gson@gson.org>
To: gnats-bugs@NetBSD.org
Cc: pgoyette@NetBSD.org, John Nemeth <jnemeth@CornerstoneService.ca>, Antti Kantee <pooka@NetBSD.org>
Subject: Re: PR/49153 CVS commit: src/sys/rump/librump/rumpkern
Date: Fri, 22 Jan 2016 12:06:10 +0200

 There has been some further discussion about this issue in private
 email, which I'm summarizing here for the record.

 On the one hand, it seems to me that re-enabling autoloading by
 default in rump kernels would be a more correct fix than changing
 the tests to cope with it being disabled by default, since the
 commit of kern_module.c 1.98 disabled it unintentionally
 (this was confirmed by jnemeth).

 If we want this, it can be done by adding

     	CPPFLAGS+= -DMODULAR_DEFAULT_AUTOLOAD

 to src/sys/rump/librump/rumpkern/Makefile.rumpkern.

 On the other hand, pooka has come to question the utility of
 autoloading in rump kernels, and is considering removing it
 altogether.  If it is going to be removed, enabling it by default
 seems pointless.

 In any case, the test failures this PR was filed to report no longer
 occur, so I will close the PR.
 -- 
 Andreas Gustafsson, gson@gson.org

State-Changed-From-To: analyzed->closed
State-Changed-By: gson@NetBSD.org
State-Changed-When: Fri, 22 Jan 2016 10:16:05 +0000
State-Changed-Why:
The test cases no longer fail.


>Unformatted:

Home	PR Database Search