From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932837AbcAMJTq (ORCPT ); Wed, 13 Jan 2016 04:19:46 -0500 Received: from mx2.suse.de ([195.135.220.15]:37165 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932077AbcAMJTm (ORCPT ); Wed, 13 Jan 2016 04:19:42 -0500 Date: Wed, 13 Jan 2016 10:19:30 +0100 (CET) From: Miroslav Benes To: Jessica Yu cc: Rusty Russell , Josh Poimboeuf , Seth Jennings , Jiri Kosina , Vojtech Pavlik , Jonathan Corbet , linux-api@vger.kernel.org, live-patching@vger.kernel.org, x86@kernel.org, linux-kernel@vger.kernel.org, linux-s390@vger.kernel.org, linux-doc@vger.kernel.org Subject: Re: [RFC PATCH v3 4/6] livepatch: reuse module loader code to write relocations In-Reply-To: <1452281304-28618-5-git-send-email-jeyu@redhat.com> Message-ID: References: <1452281304-28618-1-git-send-email-jeyu@redhat.com> <1452281304-28618-5-git-send-email-jeyu@redhat.com> User-Agent: Alpine 2.00 (LNX 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 8 Jan 2016, Jessica Yu wrote: > static int klp_write_object_relocations(struct module *pmod, > struct klp_object *obj) > { > - int ret = 0; > - unsigned long val; > - struct klp_reloc *reloc; > + int i, len, ret = 0; > + char *secname; > + const char *objname; > > if (WARN_ON(!klp_is_object_loaded(obj))) > return -EINVAL; > > - if (WARN_ON(!obj->relocs)) > - return -EINVAL; > + objname = klp_is_module(obj) ? obj->name : "vmlinux"; > > module_disable_ro(pmod); > + /* For each klp rela section for this object */ > + for (i = 1; i < pmod->info->hdr->e_shnum; i++) { > + if (!(pmod->info->sechdrs[i].sh_flags & SHF_RELA_LIVEPATCH)) > + continue; One more thing. If the module does not specify it is a live patch module in modinfo (with MODULE_INFO(livepatch, "Y")), but it is a perfect live patch module otherwise (it calls klp_register_patch in its init function), the kernel crashes here. pmod->info is not initialized at all. This should be fixed. Perhaps the easiest would be to call klp_write_object_relocations() in klp_init_object_loaded() only if is_livepatch_module() returns true. Similar to a check for obj->relocs before. Miroslav