From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752930AbdKVDGa (ORCPT ); Tue, 21 Nov 2017 22:06:30 -0500 Received: from mail.cn.fujitsu.com ([183.91.158.132]:16333 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752430AbdKVDG1 (ORCPT ); Tue, 21 Nov 2017 22:06:27 -0500 X-IronPort-AV: E=Sophos;i="5.43,368,1503331200"; d="scan'208";a="30359732" Subject: Re: [PATCH_v4.1_3/3] Make core_pattern support namespace To: Aleksa Sarai , References: <1501655849-9149-1-git-send-email-caosf.fnst@cn.fujitsu.com> <1501655849-9149-4-git-send-email-caosf.fnst@cn.fujitsu.com> <8bb63f0a-d0b7-edf7-6dca-4d12641074b4@suse.de> CC: , , From: =?UTF-8?B?5pu55qCR54O9?= Message-ID: <5A14E9F7.4070305@cn.fujitsu.com> Date: Wed, 22 Nov 2017 11:07:35 +0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.1.0 MIME-Version: 1.0 In-Reply-To: <8bb63f0a-d0b7-edf7-6dca-4d12641074b4@suse.de> Content-Type: text/plain; charset="utf-8"; format=flowed Content-Transfer-Encoding: 8bit X-yoursite-MailScanner-ID: 6C8D7482D01B.A831F X-yoursite-MailScanner: Found to be clean X-yoursite-MailScanner-From: caosf.fnst@cn.fujitsu.com Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi, Aleksa Sarai: Sorry for the late replay. > what happens if you have processes in the same pidns that have different mount namespaces? We support this. The coredump file will be saved in the same mount namespace with the processes. This is implemented by patch > Just my $0.02. Thanks. Best Regards Cao ShuFeng 在 2017年08月02日 15:07, Aleksa Sarai 写道: >> Currently, each container shared one copy of coredump setting >> with the host system, if host system changed the setting, each >> running containers will be affected. >> Same story happened when container changed core_pattern, both >> host and other container will be affected. >> >> For container based on namespace design, it is good to allow >> each container keeping their own coredump setting. > > From what I can see, this is basically setting a per-pidns > core_pattern (which is hierarchically applied). I'm not sure this > actually solves the more general problem (that usermode helper > settings aren't generally namespace-aware) -- and what happens if you > have processes in the same pidns that have different mount namespaces? > > If we _had_ to do it like this I would think it makes more sense to > pin it to mountns, but I was under the impression that someone was > working on making usermode helpers play nicer with namespaces. > > Just my $0.02. >