From owner-svn-src-user@freebsd.org Mon Nov 23 15:04:25 2015 Return-Path: Delivered-To: svn-src-user@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id 63A36A3381A for ; Mon, 23 Nov 2015 15:04:25 +0000 (UTC) (envelope-from pho@FreeBSD.org) Received: from repo.freebsd.org (repo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:0]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by mx1.freebsd.org (Postfix) with ESMTPS id 17D4F106E; Mon, 23 Nov 2015 15:04:25 +0000 (UTC) (envelope-from pho@FreeBSD.org) Received: from repo.freebsd.org ([127.0.1.37]) by repo.freebsd.org (8.15.2/8.15.2) with ESMTP id tANF4OhZ003691; Mon, 23 Nov 2015 15:04:24 GMT (envelope-from pho@FreeBSD.org) Received: (from pho@localhost) by repo.freebsd.org (8.15.2/8.15.2/Submit) id tANF4Owq003690; Mon, 23 Nov 2015 15:04:24 GMT (envelope-from pho@FreeBSD.org) Message-Id: <201511231504.tANF4Owq003690@repo.freebsd.org> X-Authentication-Warning: repo.freebsd.org: pho set sender to pho@FreeBSD.org using -f From: Peter Holm Date: Mon, 23 Nov 2015 15:04:24 +0000 (UTC) To: src-committers@freebsd.org, svn-src-user@freebsd.org Subject: svn commit: r291208 - user/pho/stress2/misc X-SVN-Group: user MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-BeenThere: svn-src-user@freebsd.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: "SVN commit messages for the experimental " user" src tree" List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 23 Nov 2015 15:04:25 -0000 Author: pho Date: Mon Nov 23 15:04:24 2015 New Revision: 291208 URL: https://svnweb.freebsd.org/changeset/base/291208 Log: Capped runtime, added findings and cleaned up the shell script a bit. Sponsored by: EMC / Isilon storage division Modified: user/pho/stress2/misc/suj23.sh Modified: user/pho/stress2/misc/suj23.sh ============================================================================== --- user/pho/stress2/misc/suj23.sh Mon Nov 23 14:06:21 2015 (r291207) +++ user/pho/stress2/misc/suj23.sh Mon Nov 23 15:04:24 2015 (r291208) @@ -31,6 +31,12 @@ # Scenario from kern/159971 # bstg0003.c by Kirk Russell +# panic: ino 0xc84c9b00(0x3C8209) 65554, 32780 != 65570 +# https://people.freebsd.org/~pho/stress/log/suj23.txt + +# panic: first_unlinked_inodedep: prev != next. inodedep = 0xcadf9e00 +# https://people.freebsd.org/~pho/stress/log/jeff091.txt + [ `id -u ` -ne 0 ] && echo "Must be root!" && exit 1 . ../default.cfg @@ -41,21 +47,21 @@ sed '1,/^EOF/d' < $here/$0 > suj23.c mycc -o suj23 -Wall -Wextra -O2 suj23.c rm -f suj23.c -mount | grep "${mntpoint}" | grep -q md${mdstart} && umount ${mntpoint} -mdconfig -l | grep -q md${mdstart} && mdconfig -d -u ${mdstart} +mount | grep "on $mntpoint " | grep -q md$mdstart && umount $mntpoint +[ -c /dev/md$mdstart ] && mdconfig -d -u $mdstart -mdconfig -a -t swap -s 1g -u ${mdstart} -bsdlabel -w md${mdstart} auto -newfs -j md${mdstart}${part} > /dev/null +mdconfig -a -t swap -s 1g -u $mdstart +bsdlabel -w md$mdstart auto +newfs -j md${mdstart}$part > /dev/null mount /dev/md${mdstart}$part $mntpoint chmod 777 $mntpoint su $testuser -c '/tmp/suj23' -while mount | grep -q ${mntpoint}; do - umount ${mntpoint} || sleep 1 +while mount | grep -q "on $mntpoint "; do + umount $mntpoint || sleep 1 done -mdconfig -d -u ${mdstart} +mdconfig -d -u $mdstart rm -f /tmp/suj23 exit 0 EOF @@ -81,10 +87,13 @@ EOF #include #include #include +#include #include #include #include +#define RUNTIME 600 + static char *bstg_pathstore[] = { "/mnt/111/z", "/mnt/111/aaaa", @@ -382,6 +391,7 @@ dosync() int main() { + time_t start; unsigned x; int i, status; void (*funcs[]) () = { @@ -413,10 +423,13 @@ main() /* give child a new seed for the pathname selection */ srand(x); + start = time(NULL); for (i = 0; i < 1000; i++) { /* each child will start looping at different * function */ (*funcs[x++ % 16]) (); + if (time(NULL) - start > RUNTIME) + break; } /* we never expect this code to run */ _exit(1);