From mboxrd@z Thu Jan 1 00:00:00 1970 From: Loic Dachary Subject: gf-complete thread safe initialization Date: Tue, 08 Apr 2014 19:22:51 +0200 Message-ID: <5344306B.8030708@dachary.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="qf1LUv3oQqQKPxJkFq5ClMoVkC9DwVpOQ" Return-path: Received: from smtp.dmail.dachary.org ([91.121.254.229]:44211 "EHLO smtp.dmail.dachary.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756079AbaDHRW5 (ORCPT ); Tue, 8 Apr 2014 13:22:57 -0400 Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Kevin Greenan Cc: Ceph Development This is an OpenPGP/MIME signed message (RFC 4880 and 3156) --qf1LUv3oQqQKPxJkFq5ClMoVkC9DwVpOQ Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Hi Kevin, The following patch is initializing the array in a lock protected / not t= hread sensitive part of code.=20 https://github.com/ceph/ceph/pull/1621/files Unless I'm mistaken, it ensures that it never happens automagically. The = first few tests that were run indicate it's not more broken than it was b= efore. We'll have to wait a few days to be sure that it fixes the issue. Do you think that's a sensible way to avoid thread safety problem in the = context of ceph ? Or are there other issues I don't see ?=20 Cheers --=20 Lo=EFc Dachary, Artisan Logiciel Libre --qf1LUv3oQqQKPxJkFq5ClMoVkC9DwVpOQ Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iEYEARECAAYFAlNEMGsACgkQ8dLMyEl6F23p+QCfZauLhMxTKfN2saUSU9YozTuQ +ckAn1ztmyb0nVyGkEdrNFYU8I4hk51t =DxA4 -----END PGP SIGNATURE----- --qf1LUv3oQqQKPxJkFq5ClMoVkC9DwVpOQ--