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=-8.2 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS, URIBL_BLOCKED,USER_AGENT_SANE_1 autolearn=ham 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 BB67AC3A5A2 for ; Tue, 3 Sep 2019 17:03:28 +0000 (UTC) Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 8744C20870 for ; Tue, 3 Sep 2019 17:03:28 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 8744C20870 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([::1]:49578 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1i5CDH-0004Zq-BB for qemu-devel@archiver.kernel.org; Tue, 03 Sep 2019 13:03:27 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:43162) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1i5BxP-0002QF-J5 for qemu-devel@nongnu.org; Tue, 03 Sep 2019 12:47:05 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1i5BxN-00087F-UZ for qemu-devel@nongnu.org; Tue, 03 Sep 2019 12:47:03 -0400 Received: from mx1.redhat.com ([209.132.183.28]:6020) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1i5BxN-00086n-Ll for qemu-devel@nongnu.org; Tue, 03 Sep 2019 12:47:01 -0400 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id D255718C4269; Tue, 3 Sep 2019 16:47:00 +0000 (UTC) Received: from work-vm (unknown [10.36.118.29]) by smtp.corp.redhat.com (Postfix) with ESMTPS id C568B6061E; Tue, 3 Sep 2019 16:46:57 +0000 (UTC) Date: Tue, 3 Sep 2019 17:46:55 +0100 From: "Dr. David Alan Gilbert" To: Ying Fang Message-ID: <20190903164655.GP2744@work-vm> References: <20190827080512.2417-1-fangying1@huawei.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20190827080512.2417-1-fangying1@huawei.com> User-Agent: Mutt/1.12.1 (2019-06-15) X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.6.2 (mx1.redhat.com [10.5.110.62]); Tue, 03 Sep 2019 16:47:00 +0000 (UTC) X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 209.132.183.28 Subject: Re: [Qemu-devel] [PATCH] qmp: Fix memory leak in migrate_params_test_apply X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: lcf.lichaofeng@huawei.com, zhanghailiang , qemu-devel@nongnu.org, zhouyibo3@huawei.com, quintela@redhat.com Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" * Ying Fang (fangying1@huawei.com) wrote: > Address Sanitizer shows memory leak in migrate_params_test_apply > migration/migration.c:1253 and the stack is as bellow: > > Direct leak of 45 byte(s) in 9 object(s) allocated from: > #0 0xffffbd7fc1db in __interceptor_malloc (/lib64/libasan.so.4+0xd31db) > #1 0xffffbd514163 in g_malloc (/lib64/libglib-2.0.so.0+0x57163) > #2 0xffffbd52f43b in g_strdup (/lib64/libglib-2.0.so.0+0x7243b) > #3 0xaaaadfa4d623 in migrate_params_test_apply migration/migration.c:1253 > #4 0xaaaadfa4d623 in qmp_migrate_set_parameters migration/migration.c:1422 > #5 0xaaaadfa963f3 in hmp_migrate_set_parameter monitor/hmp-cmds.c:1867 > #6 0xaaaadfa8afe3 in handle_hmp_command monitor/hmp.c:1082 > #7 0xaaaadf479c57 in qmp_human_monitor_command monitor/misc.c:140 > #8 0xaaaadfadf87b in qmp_marshal_human_monitor_command qapi/qapi-commands-misc.c:1024 > #9 0xaaaadfc7797b in do_qmp_dispatch qapi/qmp-dispatch.c:131 > #10 0xaaaadfc7797b in qmp_dispatch qapi/qmp-dispatch.c:174 > #11 0xaaaadfa84fff in monitor_qmp_dispatch monitor/qmp.c:120 > #12 0xaaaadfa85bbf in monitor_qmp_bh_dispatcher monitor/qmp.c:209 > #13 0xaaaadfd2228f in aio_bh_call util/async.c:89 > #14 0xaaaadfd2228f in aio_bh_poll util/async.c:117 > #15 0xaaaadfd29bc3 in aio_dispatch util/aio-posix.c:459 > #16 0xaaaadfd21ff7 in aio_ctx_dispatch util/async.c:260 > #17 0xffffbd50e2f7 in g_main_context_dispatch (/lib64/libglib-2.0.so.0+0x512f7) > #18 0xaaaadfd278d7 in glib_pollfds_poll util/main-loop.c:218 > #19 0xaaaadfd278d7 in os_host_main_loop_wait util/main-loop.c:241 > #20 0xaaaadfd278d7 in main_loop_wait util/main-loop.c:517 > #21 0xaaaadf67b5e7 in main_loop vl.c:1806 > #22 0xaaaadf15d453 in main vl.c:4488 > > Cc: zhanghailiang > Signed-off-by: Ying Fang > --- > migration/migration.c | 6 ++++++ > 1 file changed, 6 insertions(+) > > diff --git a/migration/migration.c b/migration/migration.c > index 8b9f2fe30a..05e44ff7cc 100644 > --- a/migration/migration.c > +++ b/migration/migration.c > @@ -1250,11 +1250,17 @@ static void migrate_params_test_apply(MigrateSetParameters *params, > > if (params->has_tls_creds) { > assert(params->tls_creds->type == QTYPE_QSTRING); > + if (dest->tls_creds) { > + g_free(dest->tls_creds); > + } > dest->tls_creds = g_strdup(params->tls_creds->u.s); > } > > if (params->has_tls_hostname) { > assert(params->tls_hostname->type == QTYPE_QSTRING); > + if (dest->tls_hostname) { > + g_free(dest->tls_hostname); > + } > dest->tls_hostname = g_strdup(params->tls_hostname->u.s); > } Thanks for reporting the leak, but I don't think this is the right fix: In the call chain we have, qmp_migrate_set_parameters calls: migrate_params_test_apply(params, &tmp); tmp is a stack allocated variable that becomes the 'dest' we see here. Then at the top of migrate_params_test_apply we have: *dest = migrate_get_current()->parameters; so that's probably bad; that's a shallow copy, so dest->tls_authz points to the same storage as the real current migration parameters. whne the code does: if (params->has_tls_creds) { assert(params->tls_creds->type == QTYPE_QSTRING); dest->tls_creds = g_strdup(params->tls_creds->u.s); } it's only changing the pointer in the 'tmp' not the main copy because of migrate_params_check fails then the parameters get entirely unchanged. So if you do a free on dest->tls_hostname you end up freeing the real parameter that's still getting used, not the tmp. So I think we need to: a) change migrate_params_test_apply so that it returns a MigrationParameters * rather than taking &tmp b) Make migrate_params_test use QAPI_CLONE to clone instead of doing: *dest = migrate_get_current()->parameters; c) Then do a qapi_free_MigrateParameters in qmp_migrate_set_parameters on both the true and false paths. Does that make sense? Dave > > -- > 2.19.1 > > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK