From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pa0-x22c.google.com (mail-pa0-x22c.google.com [IPv6:2607:f8b0:400e:c03::22c]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 2FF531A0356 for ; Wed, 24 Feb 2016 22:22:16 +1100 (AEDT) Received: by mail-pa0-x22c.google.com with SMTP id fl4so11333997pad.0 for ; Wed, 24 Feb 2016 03:22:16 -0800 (PST) Subject: Re: [PATCH v8 8/8] livepatch: Detect offset for the ftrace location during build To: Torsten Duwe References: <20160212164517.GO12548@pathway.suse.cz> <20160216054701.GA20570@linux.vnet.ibm.com> <20160216082302.GA20522@lst.de> <20160216103028.GA10730@linux.vnet.ibm.com> <20160216103907.GB25103@lst.de> <20160216135742.GT12548@pathway.suse.cz> <1455678521.2956.4.camel@ellerman.id.au> <20160223170017.GB21932@lst.de> <56CD4FA0.4090109@gmail.com> <56CD53E7.3010108@gmail.com> <20160224092318.GA24232@lst.de> Cc: Michael Ellerman , Petr Mladek , Jessica Yu , Jiri Kosina , linux-kernel@vger.kernel.org, Steven Rostedt , Kamalesh Babulal , live-patching@vger.kernel.org, Miroslav Benes , linuxppc-dev@lists.ozlabs.org From: Balbir Singh Message-ID: <56CD925E.4060000@gmail.com> Date: Wed, 24 Feb 2016 22:22:06 +1100 MIME-Version: 1.0 In-Reply-To: <20160224092318.GA24232@lst.de> Content-Type: text/plain; charset=windows-1252 List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On 24/02/16 20:23, Torsten Duwe wrote: > On Wed, Feb 24, 2016 at 05:55:35PM +1100, Balbir Singh wrote: >> >> >> We need to remove the SQUASH_TOC_SAVE_INSNS bits as well, now that the ppc64_profile_stub_insns does not save r2 > Sure -- this was meant to _replace_ the changes from patch 2/8, not on top. > And yes, it exposes duplicate definitions, but does not cause them AFAICS. > The two unasked questions about it were: Is Michael's solution on a similar > basis? Is this worth any further effort e.g. put into v9? > > My bad you did mention _replace_, but I think 2/8 and 6/8 of tightly bound together, so the replacement is not straight forward. Yes, it is heading in a similar direction, but it focuses mostly on ftrace. I think v9 makes sense, but I'll let Michael comment on this as well] Personally, I think your v8 or v9 + Michael's changes - RECORD_C_MCOUNT + some changes (yet to code them based on v8/v9/ftrace stability) should get the full live patching working. Balbir Singh.