Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753656Ab3ERX2W (ORCPT ); Sat, 18 May 2013 19:28:22 -0400 Received: from hydra.sisk.pl ([212.160.235.94]:34449 "EHLO hydra.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753562Ab3ERX1L (ORCPT ); Sat, 18 May 2013 19:27:11 -0400 From: "Rafael J. Wysocki" To: ACPI Devel Maling List Cc: LKML , Greg Kroah-Hartman , Toshi Kani , Wen Congyang , Tang Chen , Yasuaki Ishimatsu , Andrew Morton , Jiang Liu , Vasilis Liaskovitis , linux-mm@kvack.org Subject: [PATCH 4/5] ACPI / scan: Add second pass of companion offlining to hot-remove code Date: Sun, 19 May 2013 01:34:14 +0200 Message-ID: <3662688.5fMZaG7XgD@vostro.rjw.lan> User-Agent: KMail/4.9.5 (Linux/3.9.0+; KDE/4.9.5; x86_64; ; ) In-Reply-To: <2250271.rGYN6WlBxf@vostro.rjw.lan> References: <2250271.rGYN6WlBxf@vostro.rjw.lan> MIME-Version: 1.0 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="utf-8" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4308 Lines: 131 From: Rafael J. Wysocki As indicated by comments in mm/memory_hotplug.c:remove_memory(), if CONFIG_MEMCG is set, it may not be possible to offline all of the memory blocks held by one module (FRU) in one pass (because one of them may be used by the others to store page cgroup in that case and that block has to be offlined before the other ones). To handle that arguably corner case, add a second pass of companion device offlining to acpi_scan_hot_remove() and make it ignore errors returned in the first pass (and make it skip the second pass if the first one is successful). Signed-off-by: Rafael J. Wysocki --- drivers/acpi/scan.c | 67 ++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 50 insertions(+), 17 deletions(-) Index: linux-pm/drivers/acpi/scan.c =================================================================== --- linux-pm.orig/drivers/acpi/scan.c +++ linux-pm/drivers/acpi/scan.c @@ -131,6 +131,7 @@ static acpi_status acpi_bus_offline_comp { struct acpi_device *device = NULL; struct acpi_device_physical_node *pn; + bool second_pass = (bool)data; acpi_status status = AE_OK; if (acpi_bus_get_device(handle, &device)) @@ -141,15 +142,26 @@ static acpi_status acpi_bus_offline_comp list_for_each_entry(pn, &device->physical_node_list, node) { int ret; + if (second_pass) { + /* Skip devices offlined by the first pass. */ + if (pn->put_online) + continue; + } else { + pn->put_online = false; + } ret = device_offline(pn->dev); if (acpi_force_hot_remove) continue; - if (ret < 0) { - status = AE_ERROR; - break; + if (ret >= 0) { + pn->put_online = !ret; + } else { + *ret_p = pn->dev; + if (second_pass) { + status = AE_ERROR; + break; + } } - pn->put_online = !ret; } mutex_unlock(&device->physical_node_lock); @@ -185,6 +197,7 @@ static int acpi_scan_hot_remove(struct a acpi_handle not_used; struct acpi_object_list arg_list; union acpi_object arg; + struct device *errdev; acpi_status status; unsigned long long sta; @@ -197,22 +210,42 @@ static int acpi_scan_hot_remove(struct a lock_device_hotplug(); - status = acpi_walk_namespace(ACPI_TYPE_ANY, handle, ACPI_UINT32_MAX, - NULL, acpi_bus_offline_companions, NULL, - NULL); - if (ACPI_SUCCESS(status) || acpi_force_hot_remove) - status = acpi_bus_offline_companions(handle, 0, NULL, NULL); - - if (ACPI_FAILURE(status) && !acpi_force_hot_remove) { - acpi_bus_online_companions(handle, 0, NULL, NULL); + /* + * Carry out two passes here and ignore errors in the first pass, + * because if the devices in question are memory blocks and + * CONFIG_MEMCG is set, one of the blocks may hold data structures + * that the other blocks depend on, but it is not known in advance which + * block holds them. + * + * If the first pass is successful, the second one isn't needed, though. + */ + errdev = NULL; + acpi_walk_namespace(ACPI_TYPE_ANY, handle, ACPI_UINT32_MAX, + NULL, acpi_bus_offline_companions, + (void *)false, (void **)&errdev); + acpi_bus_offline_companions(handle, 0, (void *)false, (void **)&errdev); + if (errdev) { + errdev = NULL; acpi_walk_namespace(ACPI_TYPE_ANY, handle, ACPI_UINT32_MAX, - acpi_bus_online_companions, NULL, NULL, - NULL); + NULL, acpi_bus_offline_companions, + (void *)true , (void **)&errdev); + if (!errdev || acpi_force_hot_remove) + acpi_bus_offline_companions(handle, 0, (void *)true, + (void **)&errdev); + + if (errdev && !acpi_force_hot_remove) { + dev_warn(errdev, "Offline failed.\n"); + acpi_bus_online_companions(handle, 0, NULL, NULL); + acpi_walk_namespace(ACPI_TYPE_ANY, handle, + ACPI_UINT32_MAX, + acpi_bus_online_companions, NULL, + NULL, NULL); - unlock_device_hotplug(); + unlock_device_hotplug(); - put_device(&device->dev); - return -EBUSY; + put_device(&device->dev); + return -EBUSY; + } } ACPI_DEBUG_PRINT((ACPI_DB_INFO, -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/