From mboxrd@z Thu Jan 1 00:00:00 1970 From: Edward Cree Subject: Re: [RFC PATCH v2 bpf-next 0/2] verifier liveness simplification Date: Thu, 4 Oct 2018 18:35:45 +0100 Message-ID: References: <0252cca7-82e4-24d3-8682-e1a613d6cd78@netronome.com> <858638d7-dad4-cb70-478c-29a54a131e68@netronome.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit Cc: To: Jiong Wang , , Return-path: Received: from dispatch1-us1.ppe-hosted.com ([148.163.129.52]:33858 "EHLO dispatch1-us1.ppe-hosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727406AbeJEAaI (ORCPT ); Thu, 4 Oct 2018 20:30:08 -0400 In-Reply-To: <858638d7-dad4-cb70-478c-29a54a131e68@netronome.com> Content-Language: en-GB Sender: netdev-owner@vger.kernel.org List-ID: On 03/10/18 16:36, Jiong Wang wrote: > On 28/09/2018 14:36, Edward Cree wrote: > > But what you've described sounds interesting; perhaps it would also > >  help later with loop-variable handling? > > Haven't considered how to use this for loop-variable handling, guess you mean > applying what I have described to your previous loop detection RFC? I will look > into your RFC later. Tbh I was thinking more of John Fastabend's version (I'm not sure if he ever  got round to posting patches, but he discussed the design towards the end of  https://www.mail-archive.com/netdev@vger.kernel.org/msg216285.html ) which  is building 'proper compiler data structures' and thus might be interested  in proper use-def chains.  (Or it might not; I'm not really a compiler-guru  so it's not immediately obvious to me.) My approach was much less interested in the 'provenance' of the induction  variable, just that it was increasing appropriately, so use-def chains are  not really relevant to it. -Ed