From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964817AbbFKUKY (ORCPT ); Thu, 11 Jun 2015 16:10:24 -0400 Received: from smtp-outbound-2.vmware.com ([208.91.2.13]:36585 "EHLO smtp-outbound-2.vmware.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751663AbbFKUKV (ORCPT ); Thu, 11 Jun 2015 16:10:21 -0400 From: Philip Moltmann To: "dmitry.torokhov@gmail.com" CC: "linux-kernel@vger.kernel.org" , "pv-drivers@vmware.com" , Xavier Deguillard , "gregkh@linuxfoundation.org" , "akpm@linux-foundation.org" Subject: Re: [PATCH 6/9] VMware balloon: Do not limit the amount of frees and allocations in non-sleep mode. Thread-Topic: [PATCH 6/9] VMware balloon: Do not limit the amount of frees and allocations in non-sleep mode. Thread-Index: AQHQdtqwxW0ZMpOiZkq53vEjDzl4pZ1QlxmAgFf1xAA= Date: Thu, 11 Jun 2015 20:10:07 +0000 Message-ID: <1434053408.25956.31.camel@vmware.com> References: <1429032576-2656-1-git-send-email-moltmann@vmware.com> <1429032576-2656-8-git-send-email-moltmann@vmware.com> <20150416205555.GA21618@dtor-ws> In-Reply-To: <20150416205555.GA21618@dtor-ws> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-mailer: Evolution 3.16.2.1 (3.16.2.1-2.fc22) x-originating-ip: [10.113.160.246] Content-Type: text/plain; charset="utf-8" Content-ID: <9FEFFB2B0EFE7F4A965A517BDBC3BC94@pa-exch1.vmware.com> MIME-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from base64 to 8bit by nfs id t5BKARfS013434 Hi, sorry for taking so long to address your concerns. > What happens if you run this new driver on an older hypervisor that > does not support batched operations? When the driver starts or when it gets reset the driver checks for the capabilities of the hypervisor in vmballoon_send_start. Then it resets it state and only uses the available functionality. A reset happens any time the VM get hot migrated, snapshotted, resumed, etc. I tested this driver on various versions of ESXi to have a full set of possible capabilities. > > - if (allocations >= rate) { > > + if (rate != -1 && allocations >= rate) { > > /* We allocated enough pages, let's take a > > break. */ > > Why don't you make the rate UINT_MAX when doing fast allocations, > then > you would not need to treat -1 as a special case. Good suggestion. > > /* free pages to reach target */ > > list_for_each_entry_safe(page, next, &b->pages, lru) { > > list_del(&page->lru); > > b->ops->add_page(b, num_pages++, page); > > > > + > > Seems line unintended whitespace addition. I could not find this in the code. I will send out a new patch series addressing your concerns. Thanks Philip{.n++%ݶw{.n+{G{ayʇڙ,jfhz_(階ݢj"mG?&~iOzv^m ?I