From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx2.suse.de ([195.135.220.15]:53754 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2409957AbgJPPY1 (ORCPT ); Fri, 16 Oct 2020 11:24:27 -0400 Date: Fri, 16 Oct 2020 17:24:19 +0200 From: Petr Tesarik Subject: Re: Is OS_INFO_VMCOREINFO unimplemented? Message-ID: <20201016172419.3abfdeda@ezekiel.suse.cz> In-Reply-To: <20201016161125.2fcf494d@ibm> References: <20201013145303.6f99fe89@ezekiel.suse.cz> <20201016161125.2fcf494d@ibm> MIME-Version: 1.0 Content-Type: multipart/signed; boundary="Sig_/J0NdDn94vZ/rqqA9iOHb9hf"; protocol="application/pgp-signature"; micalg=pgp-sha256 List-ID: To: Philipp Rudo Cc: linux-s390@vger.kernel.org, Michael Holzheu --Sig_/J0NdDn94vZ/rqqA9iOHb9hf Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable Hi Philipp, On Fri, 16 Oct 2020 16:11:25 +0200 Philipp Rudo wrote: > Hi Petr, >=20 > sorry for the late reply. No problem. ;-) > It's an interface for non-Linux systems for the stand-alone kdump. >=20 > But that's all I'm sure of. I'm afraid only Michael knows the full history > behind the implementation. Unfortunately he left IBM ~2 years ago so this= piece > of knowledge is lost... >=20 > My theory is that originally it was planned to use this mechanism for the > "normal" kdump as well. But for kdump common code "corrupts" the vmcorein= fo by > adding the CRASHTIME shortly before kexec'ing the crash kernel. So the cr= ash > kernel would refuse to load the os_info anyway and thus it is never set. Sure, the checksum would have to be recalculated after setting CRASHTIME. B= ut that's perfectly possible. > Hope this helps you at least a little Yes, to some extent. The reason I asked was that I also implemented parsing= of OS_INFO_VMCOREINFO in libkdumpfile a few years ago, but it has no test = coverage. So, I looked around a bit and to my surprise all dump files conta= ined a NULL pointer there, which looked somewhat suspicious. Anyway, if nobody knows for certain, then my plan is to add the necessary c= ode to the Linux kernel. Patch coming soon on the mailing list. ;-) Thanks, Petr T > Philipp >=20 >=20 > On Tue, 13 Oct 2020 14:53:03 +0200 > Petr Tesarik wrote: >=20 > > Hi all, > >=20 > > I've been looking into kernel crash dump analysis for some time now, > > and I've noticed that none of my sample dumps for z/Architecture sets > > OS_INFO_VMCOREINFO. > >=20 > > Commit 4857d4bbe9821c8d732cb84455e18e12b3d79add suggests that the > > "os_info" structure should contain vmcoreinfo, but it seems that the > > corresponding fields are always zero, and a quick grep for > > OS_INFO_VMCOREINFO finds only code that tries to read this entry in the > > panic kernel, but no code that would initialize it in the old (crashed) > > kernel. > >=20 > > In short, the panic kernel always prints an informational message that > > entry 0 is not available and falls back to get_vmcoreinfo_old(). > >=20 > > Is this a bug, or is this interface used by a non-Linux operating > > system that I'm not aware of? > >=20 > > TIA, > > Petr Tesarik > > SUSE HW Enablement =20 --Sig_/J0NdDn94vZ/rqqA9iOHb9hf Content-Type: application/pgp-signature Content-Description: =?unknown-8bit?q?Digit=C3=A1ln=C3=AD?= podpis OpenPGP -----BEGIN PGP SIGNATURE----- iQEzBAEBCAAdFiEEHl2YIZkIo5VO2MxYqlA7ya4PR6cFAl+JuyMACgkQqlA7ya4P R6fZagf/UHB0waAsfcrsiS//sAu6FX4k+7elra51TeeYl2YvFeDL9wpCa47GkjC+ sNDcksMRFm0KsND/mShSKFhUJ7Va6OMD1g+GqlmLEeJTNn2LyzIBcmcYRtJRd+jU nhPXebUWoWX3T91w8eLHumxPnan8eZt64XiMgje6Fd+3YQHDupCgqHDFSAAlV7RE oxEqZfjsJpFERts5+4u9b4OLtOLqB8NFqQrcDW0R8/cVrapINoGQWXDpEvCjud9v Dhb57HmsJdHHIRJiyB8V37sVjs5Z5gmi5VHL3Xwf16Zs6Urm9iydTgVJCxD077Iy EB6sRCLrjxHVRegdQaF0oioJ1w0IVw== =t6Be -----END PGP SIGNATURE----- --Sig_/J0NdDn94vZ/rqqA9iOHb9hf--