From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH RFC 0/2] add nproc cgroup subsystem Date: Mon, 2 Mar 2015 08:49:26 -0500 Message-ID: <20150302134926.GA17694@htj.duckdns.org> References: <1424660891-12719-1-git-send-email-cyphar@cyphar.com> <20150227114940.GB3964@htj.duckdns.org> <54F09E62.8000007@gmail.com> <20150227170640.GK3964@htj.duckdns.org> <54F0BC51.4050506@gmail.com> <20150228115926.GA1005@htj.duckdns.org> <20150228164353.GR3964@htj.duckdns.org> <54F461F3.3030903@gmail.com> Mime-Version: 1.0 Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=sender:date:from:to:cc:subject:message-id:references:mime-version :content-type:content-disposition:in-reply-to:user-agent; bh=jqQ2/Gq2tlqy1QUcKqcaqi+0NmuJx57vnaL/w+aoWWQ=; b=W+g090H93snTdJOFFgEI2R3dHAxn1wKtRO4fEDeTri55UDqzdg6kaKEnBTJvmuvANt RU+69oUyBijN8iSMQhEWUC6rIOPveolXVxBvqk0kJfJgtkymmhmywg0iu2GX9og/I1e2 k9pTqxTuaD8+hhf6pUc/viZoixsjWkKahtP/z6yZyl7JynGKaCfxCZzdDnU+A8zd/zbn 0506/psYp1tNURtvL1NuqQnmQcSBSfJDcCb8UPjJ1kzWYhrA5kU8wbr4MMr1NAfJjf+J b5S+EgATxbbozYXP0TCSwPbYphN2pnUrmUvG7lowN/cpe1ZfuGfA7zqdrY03NUhdX95C oUgg== Content-Disposition: inline In-Reply-To: <54F461F3.3030903@gmail.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Austin S Hemmelgarn Cc: Tim Hockin , Frederic Weisbecker , lizefan@huawei.com, richard@nod.at, mingo@redhat.com, Aleksa Sarai , cgroups@vger.kernel.org, peterz@infradead.org, linux-kernel@vger.kernel.org On Mon, Mar 02, 2015 at 08:13:23AM -0500, Austin S Hemmelgarn wrote: > If 16-bit PID's aren't a concern anymore, then why do we still default to > treating it like a 16-bit signed int (the default for > /proc/sys/kernel/pid_max is 32768)? Inertia. It has to start there for backward compatibility. Now it's trivial to adjust dynamically and majority of the users don't need to worry about it, so there's no pressing reason to bump it up by default. 16bit pid_t was already a dying breed on 32bit config and it never was an option on 64bit. Any remotely modern distros in the past decade, whether 32 or 64bit, wouldn't have any problem with it. The only possibly problematic case would be legacy code which for some reason explicitly used 16bit integer types instead of pid_t, but at this point, we shouldn't be basing any design decisions on that. If anybody is still depending on that, there are different ways ton deal with the issue on their end including namespacing its pid space. Thanks. -- tejun