Date: Thu, 14 Nov 2013 12:34:08 +0100 From: Carlos Jacobo Puga Medina <cjpugmed@gmail.com> To: freebsd-hackers@freebsd.org, cperciva@freebsd.org Subject: Re: Automated submission of kernel panic reports: sysutils/panicmail Message-ID: <CAHHLbRPmz-TqzH21vHf_O6UGbs55zLAs_3%2BfZ99dokO0zebs9g@mail.gmail.com> In-Reply-To: <CAHHLbROX9opc6YwhMnXR3xDW9FKpzyAFDtG065Hiv%2BHDOT5j4g@mail.gmail.com> References: <CAHHLbRNv2T2-arUXn7WDFHDOAOTQvVqe0nniOr5sVPcgob4QJQ@mail.gmail.com> <CAHHLbRNGwP56Qq_FLZGvN7hORJtE1SUVAZKpJty_cN4%2B7mEEUA@mail.gmail.com> <CAHHLbROX9opc6YwhMnXR3xDW9FKpzyAFDtG065Hiv%2BHDOT5j4g@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Well, I realized that savecore(8) doesn't work as it should. It's clear that after apply the patch, it changes adequately the order of scripts as follow: % rcorder /etc/rc.d/* | less /etc/rc.d/sysctl /etc/rc.d/hostid /etc/rc.d/zvol /etc/rc.d/dumpon /etc/rc.d/ddb /etc/rc.d/initrandom /etc/rc.d/geli /etc/rc.d/gbde /etc/rc.d/ccd /etc/rc.d/savecore /etc/rc.d/encswap /etc/rc.d/swap1 <snip> % swapinfo Device 1K-blocks Used Avail Capacity /dev/ada0p3.eli 3907584 0 3907584 0% Also, /dev/dumpdev is pointing to /dev/ada0p3 % ll /dev/dumpdev lrwxr-xr-x 1 root wheel 11 13 nov 23:51 /dev/dumpdev@ -> /dev/ada0p3 I created a crash dump using "reboot -d" but I see these messages related to savecore at boot stage: savecore: reboot savecore: unable to write to bounds file: Read-only file system savecore: info.0: Read-only file system savecore: unsaved dumps found but not saved No crash dumps in /var/crash. Any suggestions? --CJPM 2013/11/7 Carlos Jacobo Puga Medina <cjpugmed@gmail.com> > At last, I realized how to solved this inconvenient. Digging a bit I found > what I needed. All is summarized in the pr124747 which describes perfectly > this problem and attached a patch to fix it. > > I add the modified patch according to current rc.d scripts included. > > diff -ur encswap.orig encswap > --- encswap.orig 2013-11-06 23:31:43.000000000 +0100 > +++ encswap 2013-11-07 00:14:03.000000000 +0100 > @@ -3,8 +3,8 @@ > # $FreeBSD: release/9.2.0/etc/rc.d/encswap 180563 2008-07-16 19:22:48Z > dougb $ > # > > -# PROVIDE: disks > -# REQUIRE: initrandom > +# PROVIDE: encswap > +# REQUIRE: initrandom disks > # KEYWORD: nojail > > . /etc/rc.subr > diff -ur savecore.orig savecore > --- savecore.orig 2013-11-06 23:32:21.000000000 +0100 > +++ savecore 2013-11-07 00:14:18.000000000 +0100 > @@ -4,7 +4,7 @@ > # > > # PROVIDE: savecore > -# REQUIRE: dumpon ddb syslogd > +# BEFORE: encswap > # KEYWORD: nojail > > . /etc/rc.subr > diff -ur swap1.orig swap1 > --- swap1.orig 2013-11-06 23:32:28.000000000 +0100 > +++ swap1 2013-11-07 00:14:29.000000000 +0100 > @@ -4,7 +4,7 @@ > # > > # PROVIDE: localswap > -# REQUIRE: disks > +# REQUIRE: encswap disks > # KEYWORD: nojail shutdown > > . /etc/rc.subr > > I tested it and now it works like a charm. > > > 2013/11/5 Carlos Jacobo Puga Medina <cjpugmed@gmail.com> > >> Add missing thread related to Pawel's proposal [1] >> >> [1] >> http://lists.freebsd.org/pipermail/freebsd-fs/2010-September/009256.html >> >> >> 2013/11/5 Carlos Jacobo Puga Medina <cjpugmed@gmail.com> >> >>> Seems that rc.d scripts order doesn't allow geli-encrypted swap and >>> dumpdev to cooperate together. For this reason, I wonder what happens in >>> such cases, is not possible to use panicmail unless Pawel's proposal [1] >>> which argued to obtain and save crash info in memory somewhere before we >>> configure swap and copy it to /var/crash/ once we mount it. >>> >>> Is sustainable this proposal or is there yet another alternative to >>> consider for those who use an encrypted GELI swap? >>> >>> --CJPM >>> >> >> >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAHHLbRPmz-TqzH21vHf_O6UGbs55zLAs_3%2BfZ99dokO0zebs9g>