From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from aserp1040.oracle.com ([141.146.126.69]:21568 "EHLO aserp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932595AbcGOLyh (ORCPT ); Fri, 15 Jul 2016 07:54:37 -0400 Subject: Re: Linux 4.1.28 To: Thomas Voegtle References: <5786F7BC.9090406@oracle.com> Cc: LKML , stable , lwn@lwn.net, lukasz.odzioba@intel.com From: Sasha Levin Message-ID: <5788CEF1.8010900@oracle.com> Date: Fri, 15 Jul 2016 07:54:25 -0400 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: stable-owner@vger.kernel.org List-ID: On 07/15/2016 07:38 AM, Thomas Voegtle wrote: > On Wed, 13 Jul 2016, Sasha Levin wrote: > >> I'm announcing the release of the 4.1.28 kernel. > > I have a serious memleak with 4.1.28 (like 20mb/s) > I stripped down my kernel config and started a bisect, which came to: > > # first bad commit: [c5ad33184354260be6d05de57e46a5498692f6d6] mm/swap.c: > flush lru pvecs on compound page arrival > => > commit c5ad33184354260be6d05de57e46a5498692f6d6 > Author: Lukasz Odzioba > Date: Fri Jun 24 14:50:01 2016 -0700 > > mm/swap.c: flush lru pvecs on compound page arrival > > > Reverting this on top 4.1.28 helps. Config attached. Yup, this was reported and a fix is already queued for 4.1.29. Thanks, Sasha