From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 14E2BC001DF for ; Tue, 1 Aug 2023 10:06:44 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233779AbjHAKGm (ORCPT ); Tue, 1 Aug 2023 06:06:42 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33414 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232120AbjHAKGi (ORCPT ); Tue, 1 Aug 2023 06:06:38 -0400 X-Greylist: delayed 135 seconds by postgrey-1.37 at lindbergh.monkeyblade.net; Tue, 01 Aug 2023 03:06:37 PDT Received: from p3plsmtpa14-08.prod.phx3.secureserver.net (p3plsmtpa14-08.prod.phx3.secureserver.net [97.74.135.179]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4872FD3 for ; Tue, 1 Aug 2023 03:06:37 -0700 (PDT) Received: from localhost ([82.27.99.45]) by :SMTPAUTH: with ESMTPA id QmEmqrZkLyhe0QmEnqlwKH; Tue, 01 Aug 2023 03:04:21 -0700 X-CMAE-Analysis: v=2.4 cv=abin3zkt c=1 sm=1 tr=0 ts=64c8d8a5 a=YwMIiW7BGddQzL8MrqPWMg==:117 a=YwMIiW7BGddQzL8MrqPWMg==:17 a=2pMVjWlDXWENfkYnN9AA:9 X-SECURESERVER-ACCT: atomlin@atomlin.com From: Aaron Tomlin To: peterz@infradead.org, tj@kernel.org Cc: atomlin@atomlin.com, jiangshanlai@gmail.com, linux-kernel@vger.kernel.org Subject: Re: [RFC PATCH 1/2] workqueue: Introduce PF_WQ_RESCUE_WORKER Date: Tue, 1 Aug 2023 11:04:20 +0100 Message-Id: <20230801100420.939677-1-atomlin@atomlin.com> X-Mailer: git-send-email 2.39.1 In-Reply-To: <20230729160712.GA5697@hirez.programming.kicks-ass.net> References: <20230729160712.GA5697@hirez.programming.kicks-ass.net> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-CMAE-Envelope: MS4xfBYzk98O0O1erpH2VmSPAByCKAiI/9RV6J0qSksSCo+2JivjCjYFq6OzWiih80wp9X6jVMwEjAmlPcK4ptf+FFYohnZnBdbottnUBNhEVZfOVpG5br3Q FsXJ9SMefb2gGAT1b4l10gTfP5RV/61MwiXMEs0qqtTvW4bze4EgDlicYHboCtPfJeUTmNvWJeb0ipS4AzkFVO6uGbD6oieHKTgt3TDDcUojqR9X52XUTLll O7klyK9OAtsfxckFxLuHJkuYPbYi1ukvd/VvloBrLKeg58olBiwTTzMvoEoiyF+VfBXL2w2hnTgSvhJuFiJMsg== Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org > Is the implication that PF_flags are considered ABI? We've been changing > them quite a bit over the years. Hi Peter, Tejun, I never assumed they were. In this context, one should always check the Linux kernel source code first i.e. do not assume what is exported via /proc/[PID]/stat will be stable/or consistent between releases. > Also, while we have a few spare bits atm, we used to be nearly out for a > while, and I just don't think this is sane usage of them. We don't use PF > flags just for userspace. Fair statement. Albeit, I suspect it would still be useful for user-mode to easily differentiate between a kworker and a rescuer kworker. According to create_worker(), we do make it clear the difference between a CPU-specific and unbound kworker by way of the task's name. Looking at init_rescuer() a rescuer kworker is simply given the name of its workqueue. Would you consider modifying the rescuer's task's name so it is prefixed with "kworker/r-%s" and then include the workqueue's name e.g. "kworker/r-ext4-rsv-conver" acceptable? Kind regards, -- Aaron Tomlin