From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751330AbdALPaC (ORCPT ); Thu, 12 Jan 2017 10:30:02 -0500 Received: from mx1.redhat.com ([209.132.183.28]:59454 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751267AbdALP35 (ORCPT ); Thu, 12 Jan 2017 10:29:57 -0500 From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Glisse?= To: akpm@linux-foundation.org, , linux-mm@kvack.org Cc: John Hubbard , =?UTF-8?q?J=C3=A9r=C3=B4me=20Glisse?= , Dan Williams , Ross Zwisler Subject: [HMM v16 02/15] mm/ZONE_DEVICE/devmem_pages_remove: allow early removal of device memory v2 Date: Thu, 12 Jan 2017 11:30:29 -0500 Message-Id: <1484238642-10674-3-git-send-email-jglisse@redhat.com> In-Reply-To: <1484238642-10674-1-git-send-email-jglisse@redhat.com> References: <1484238642-10674-1-git-send-email-jglisse@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.29]); Thu, 12 Jan 2017 15:29:34 +0000 (UTC) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Some device driver manage multiple physical devices memory from a single fake device driver. In that case the fake device might outlive the real device and ZONE_DEVICE and its resource allocated for a real device would waste resources in the meantime. This patch allow early removal of ZONE_DEVICE and associated resource, before device driver is tear down. Changed since v1: - s/devm_memremap_pages_remove/devm_memunmap_pages - use ret value of devm_memremap_pages as arg for devm_memunmap_pages Signed-off-by: Jérôme Glisse Cc: Dan Williams Cc: Ross Zwisler --- include/linux/memremap.h | 6 ++++++ kernel/memremap.c | 14 ++++++++++++++ 2 files changed, 20 insertions(+) diff --git a/include/linux/memremap.h b/include/linux/memremap.h index f7e0609..aee8477 100644 --- a/include/linux/memremap.h +++ b/include/linux/memremap.h @@ -53,6 +53,7 @@ struct dev_pagemap { void *devm_memremap_pages(struct device *dev, struct resource *res, struct percpu_ref *ref, struct vmem_altmap *altmap); struct dev_pagemap *find_dev_pagemap(resource_size_t phys); +int devm_memunmap_pages(struct device *dev, void *start); static inline bool dev_page_allow_migrate(const struct page *page) { @@ -78,6 +79,11 @@ static inline struct dev_pagemap *find_dev_pagemap(resource_size_t phys) return NULL; } +static inline int devm_memunmap_pages(struct device *dev, void *start) +{ + return -EINVAL; +} + static inline bool dev_page_allow_migrate(const struct page *page) { return false; diff --git a/kernel/memremap.c b/kernel/memremap.c index 07665eb..f0d4ea2 100644 --- a/kernel/memremap.c +++ b/kernel/memremap.c @@ -387,6 +387,20 @@ void *devm_memremap_pages(struct device *dev, struct resource *res, } EXPORT_SYMBOL(devm_memremap_pages); +static int devm_page_map_match(struct device *dev, void *data, void *match_data) +{ + struct page_map *page_map = data; + + return __va(page_map->res.start) == match_data; +} + +int devm_memunmap_pages(struct device *dev, void *start) +{ + return devres_release(dev, &devm_memremap_pages_release, + &devm_page_map_match, start); +} +EXPORT_SYMBOL(devm_memunmap_pages); + unsigned long vmem_altmap_offset(struct vmem_altmap *altmap) { /* number of pfns from base where pfn_to_page() is valid */ -- 2.4.3