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 X-Spam-Level: X-Spam-Status: No, score=-0.8 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SPF_PASS,URIBL_BLOCKED autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 20920C10F05 for ; Wed, 20 Mar 2019 11:33:59 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id D2BCB2186A for ; Wed, 20 Mar 2019 11:33:58 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (1024-bit key) header.d=joelfernandes.org header.i=@joelfernandes.org header.b="mRAEamYd" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727988AbfCTLd5 (ORCPT ); Wed, 20 Mar 2019 07:33:57 -0400 Received: from mail-qt1-f193.google.com ([209.85.160.193]:36980 "EHLO mail-qt1-f193.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726006AbfCTLd4 (ORCPT ); Wed, 20 Mar 2019 07:33:56 -0400 Received: by mail-qt1-f193.google.com with SMTP id z16so572843qtn.4 for ; Wed, 20 Mar 2019 04:33:55 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=joelfernandes.org; s=google; h=date:user-agent:in-reply-to:references:mime-version :content-transfer-encoding:subject:to:cc:from:message-id; bh=d4Z5Mqmi4LJpbgIxofUDgn3+z5lt3XVSz/v98zM6dGA=; b=mRAEamYdksjPsuTT3gCoilF0nl1WVeaMZxrp1TGoeVeQ2DV/jRTfE84LyIcdoznqET 1JMcsEKSvlvTNUOo4WathfgwTJMeKT2I7E8FU4K5gKtP7VYRExdXFSSyLq4DyQo4J9+0 AWCC6EAjjQKAtdovCCY6NFWpZHC98Dq7Nlu3o= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:user-agent:in-reply-to:references :mime-version:content-transfer-encoding:subject:to:cc:from :message-id; bh=d4Z5Mqmi4LJpbgIxofUDgn3+z5lt3XVSz/v98zM6dGA=; b=RqcL4s/ENn0FYbUBRDVxIORTOwLHSaexeAeWDYrJnOrCERRruanes/r4TWd2yGmeFE PySHQ6bPMxZ5TJ4KTGc2CpNGtw+uAT+pAohRmu4ghlaoyC/b4mAyhCPEMZvgAs2HcNDC 66pVRN67pOugr/RI1VUq6dj39V0PHuC9ORwLqOUWriDwfJUe3Gto2MdOQRMqlEjc7jIg iQLzvo/65nRmwHVDmd/NANh/mNetaj79HnsU/G87s5Jj8/ozG9f1tdQXDpgj/tw9UJRX SBw5UZ8Eiz94xSgCgWHuNn71nhEcBcA9k0MCs3Qak0xt6d9WdcWOZ6uYn+fl2uYAIX8h DWIw== X-Gm-Message-State: APjAAAUsT8Wxg7BrIU0s3nZX19dTloOtDZAq3JZMH7KhbB4CWDKrRinT gxfOJPWvhuydqNYAQFx3bxqarw== X-Google-Smtp-Source: APXvYqzkMfTtf1uGPNkRoUzX4rFJebsuzIDcrB0bSeB6XFbGSXfe+v9DxOGTZ1whbs3Rlygi+TBr6w== X-Received: by 2002:a0c:9e9a:: with SMTP id r26mr6005557qvd.57.1553081635138; Wed, 20 Mar 2019 04:33:55 -0700 (PDT) Received: from [192.168.0.109] (c-73-216-90-110.hsd1.va.comcast.net. [73.216.90.110]) by smtp.gmail.com with ESMTPSA id 91sm912671qtf.62.2019.03.20.04.33.53 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Wed, 20 Mar 2019 04:33:54 -0700 (PDT) Date: Wed, 20 Mar 2019 07:33:51 -0400 User-Agent: K-9 Mail for Android In-Reply-To: References: <20190317015306.GA167393@google.com> <20190317114238.ab6tvvovpkpozld5@brauner.io> <20190318002949.mqknisgt7cmjmt7n@brauner.io> <20190318235052.GA65315@google.com> <20190319221415.baov7x6zoz7hvsno@brauner.io> <20190319231020.tdcttojlbmx57gke@brauner.io> <20190320015249.GC129907@google.com> <20190320035953.mnhax3vd47ya4zzm@brauner.io> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Subject: Re: pidfd design To: Daniel Colascione , Christian Brauner CC: Suren Baghdasaryan , Steven Rostedt , Sultan Alsawaf , Tim Murray , Michal Hocko , Greg Kroah-Hartman , =?ISO-8859-1?Q?Arve_Hj=F8nnev=E5g?= , Todd Kjos , Martijn Coenen , Ingo Molnar , Peter Zijlstra , LKML , "open list:ANDROID DRIVERS" , linux-mm , kernel-team , Oleg Nesterov , Andy Lutomirski , "Serge E. Hallyn" , Kees Cook From: Joel Fernandes Message-ID: <4A06C5BB-9171-4E70-BE31-9574B4083A9F@joelfernandes.org> Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On March 20, 2019 3:02:32 AM EDT, Daniel Colascione = wrote: >On Tue, Mar 19, 2019 at 8:59 PM Christian Brauner > wrote: >> >> On Tue, Mar 19, 2019 at 07:42:52PM -0700, Daniel Colascione wrote: >> > On Tue, Mar 19, 2019 at 6:52 PM Joel Fernandes > wrote: >> > > >> > > On Wed, Mar 20, 2019 at 12:10:23AM +0100, Christian Brauner >wrote: >> > > > On Tue, Mar 19, 2019 at 03:48:32PM -0700, Daniel Colascione >wrote: >> > > > > On Tue, Mar 19, 2019 at 3:14 PM Christian Brauner > wrote: >> > > > > > So I dislike the idea of allocating new inodes from the >procfs super >> > > > > > block=2E I would like to avoid pinning the whole pidfd >concept exclusively >> > > > > > to proc=2E The idea is that the pidfd API will be useable >through procfs >> > > > > > via open("/proc/") because that is what users expect >and really >> > > > > > wanted to have for a long time=2E So it makes sense to have >this working=2E >> > > > > > But it should really be useable without it=2E That's why >translate_pid() >> > > > > > and pidfd_clone() are on the table=2E What I'm saying is, >once the pidfd >> > > > > > api is "complete" you should be able to set CONFIG_PROCFS=3DN >- even >> > > > > > though that's crazy - and still be able to use pidfds=2E This >is also a >> > > > > > point akpm asked about when I did the pidfd_send_signal >work=2E >> > > > > >> > > > > I agree that you shouldn't need CONFIG_PROCFS=3DY to use >pidfds=2E One >> > > > > crazy idea that I was discussing with Joel the other day is >to just >> > > > > make CONFIG_PROCFS=3DY mandatory and provide a new >get_procfs_root() >> > > > > system call that returned, out of thin air and independent of >the >> > > > > mount table, a procfs root directory file descriptor for the >caller's >> > > > > PID namspace and suitable for use with openat(2)=2E >> > > > >> > > > Even if this works I'm pretty sure that Al and a lot of others >will not >> > > > be happy about this=2E A syscall to get an fd to /proc? >> > >> > Why not? procfs provides access to a lot of core kernel >functionality=2E >> > Why should you need a mountpoint to get to it? >> > >> > > That's not going >> > > > to happen and I don't see the need for a separate syscall just >for that=2E >> > >> > We need a system call for the same reason we need a getrandom(2): >you >> > have to bootstrap somehow when you're in a minimal environment=2E >> > >> > > > (I do see the point of making CONFIG_PROCFS=3Dy the default btw= =2E) >> > >> > I'm not proposing that we make CONFIG_PROCFS=3Dy the default=2E I'm >> > proposing that we *hardwire* it as the default and just declare >that >> > it's not possible to build a Linux kernel that doesn't include >procfs=2E >> > Why do we even have that button? >> > >> > > I think his point here was that he wanted a handle to procfs no >matter where >> > > it was mounted and then can later use openat on that=2E Agreed that >it may be >> > > unnecessary unless there is a usecase for it, and especially if >the /proc >> > > directory being the defacto mountpoint for procfs is a universal >convention=2E >> > >> > If it's a universal convention and, in practice, everyone needs >proc >> > mounted anyway, so what's the harm in hardwiring CONFIG_PROCFS=3Dy? >If >> > we advertise /proc as not merely some kind of optional debug >interface >> > but *the* way certain kernel features are exposed --- and there's >> > nothing wrong with that --- then we should give programs access to >> > these core kernel features in a way that doesn't depend on >userspace >> > kernel configuration, and you do that by either providing a >> > procfs-root-getting system call or just hardwiring the "/proc/" >prefix >> > into VFS=2E >> > >> > > > Inode allocation from the procfs mount for the file descriptors >Joel >> > > > wants is not correct=2E Their not really procfs file descriptors >so this >> > > > is a nack=2E We can't just hook into proc that way=2E >> > > >> > > I was not particular about using procfs mount for the FDs but >that's the only >> > > way I knew how to do it until you pointed out anon_inode (my grep >skills >> > > missed that), so thank you! >> > > >> > > > > C'mon: /proc is used by everyone today and almost every >program breaks >> > > > > if it's not around=2E The string "/proc" is already de facto >kernel ABI=2E >> > > > > Let's just drop the pretense of /proc being optional and bake >it into >> > > > > the kernel proper, then give programs a way to get to /proc >that isn't >> > > > > tied to any particular mount configuration=2E This way, we >don't need a >> > > > > translate_pid(), since callers can just use procfs to do the >same >> > > > > thing=2E (That is, if I understand correctly what translate_pid >does=2E) >> > > > >> > > > I'm not sure what you think translate_pid() is doing since >you're not >> > > > saying what you think it does=2E >> > > > Examples from the old patchset: >> > > > translate_pid(pid, ns, -1) - get pid in our pid namespace >> > >> > Ah, it's a bit different from what I had in mind=2E It's fair to want >to >> > translate PIDs between namespaces, but the only way to make the >> > translate_pid under discussion robust is to have it accept and >produce >> > pidfds=2E (At that point, you might as well call it translate_pidfd= =2E) >We >> > should not be adding new APIs to the kernel that accept numeric >PIDs: >> >> The traditional pid-based api is not going away=2E There are users that >> have the requirement to translate pids between namespaces and also >doing >> introspection on these namespaces independent of pidfds=2E We will not >> restrict the usefulness of this syscall by making it only work with >> pidfds=2E >> >> > it's not possible to use these APIs correctly except under very >> > limited circumstances --- mostly, talking about init or a parent >> >> The pid-based api is one of the most widely used apis of the kernel >and >> people have been using it quite successfully for a long time=2E Yes, >it's >> rac, but it's here to stay=2E >> >> > talking about its child=2E >> > >> > Really, we need a few related operations, and we shouldn't >necessarily >> > mingle them=2E >> >> Yes, we've established that previously=2E >> >> > >> > 1) Given a numeric PID, give me a pidfd: that works today: you just >> > open /proc/ >> >> Agreed=2E >> >> > >> > 2) Given a pidfd, give me a numeric PID: that works today: you just >> > openat(pidfd, "stat", O_RDONLY) and read the first token (which is >> > always the numeric PID)=2E >> >> Agreed=2E >> >> > >> > 3) Given a pidfd, send a signal: that's what pidfd_send_signal >does, >> > and it's a good start on the rest of these operations=2E >> >> Agreed=2E >> >> > 5) Given a pidfd in NS1, get a pidfd in NS2=2E That's what >translate_pid >> > is for=2E My preferred signature for this routine is >translate_pid(int >> > pidfd, int nsfd) -> pidfd=2E We don't need two namespace arguments=2E >Why >> > not? Because the pidfd *already* names a single process, uniquely! >> >> Given that people are interested in pids we can't just always return >a >> pidfd=2E That would mean a user would need to do get the pidfd read >from >> /stat and then close the pidfd=2E If you do that for a 100 pids >or >> more you end up allocating and closing file descriptors constantly >for >> no reason=2E We can't just debate pids away=2E So it will also need to = be >> able to yield pids e=2Eg=2E through a flag argument=2E > >Sure, but that's still not a reason that we should care about pidfds >working separately from procfs=2E=2E Agreed=2E I can't imagine pidfd being anything but a proc pid directory ha= ndle=2E So I am confused what Christian meant=2E Pidfd *is* a procfs direct= ory fid always=2E That's what I gathered from his pidfd_send_signal patch = but let me know if I'm way off in the woods=2E For my next revision, I am thinking of adding the flag argument Christian = mentioned to make translate_pid return an anon_inode FD which can be used f= or death status, given a =2E Since it is thought that translate_pid ca= n be made to return a pid FD, I think it is ok to have it return a pid stat= us FD for the purposes of the death status as well=2E Joel Fernandes, Android kernel team Sent from k9-mail on Android