From patchwork Tue Apr 25 07:56:39 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222979 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 7E672C7EE24 for ; Tue, 25 Apr 2023 07:57:26 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525751.817148 (Exim 4.92) (envelope-from ) id 1prDY3-0005Iy-0R; Tue, 25 Apr 2023 07:57:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525751.817148; Tue, 25 Apr 2023 07:57:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDY2-0005Ir-Td; Tue, 25 Apr 2023 07:57:14 +0000 Received: by outflank-mailman (input) for mailman id 525751; Tue, 25 Apr 2023 07:57:13 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDY1-00053E-Eg for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:13 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id c943116a-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:12 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id BDFD1D75; Tue, 25 Apr 2023 00:57:55 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 6B5A63F587; Tue, 25 Apr 2023 00:57:08 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: c943116a-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Andrew Cooper , George Dunlap , Jan Beulich , Wei Liu , Henry Wang Subject: [PATCH v4 01/17] xen/arm: use NR_MEM_BANKS to override default NR_NODE_MEMBLKS Date: Tue, 25 Apr 2023 15:56:39 +0800 Message-Id: <20230425075655.4037980-2-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen As a memory range described in device tree cannot be split across multiple nodes. And it is very likely than if you have more than 64 nodes, you may need a lot more than 2 regions per node. So the default NR_NODE_MEMBLKS value (MAX_NUMNODES * 2) makes no sense on Arm. So, for Arm, we would just define NR_NODE_MEMBLKS as an alias to NR_MEM_BANKS. And in the future NR_MEM_BANKS will be user-configurable via kconfig, but for now leave NR_MEM_BANKS as 128 on Arm. This avoids having different way to define the value based NUMA vs non-NUMA. Further discussions can be found here[1]. [1] https://lists.xenproject.org/archives/html/xen-devel/2021-09/msg02322.html Signed-off-by: Wei Chen Signed-off-by: Henry Wang Acked-by: Jan Beulich --- v3 -> v4: 1. Add Acked-by tag from Jan. v2 -> v3: By checking the discussion in [1] and [2] [1] https://lists.xenproject.org/archives/html/xen-devel/2023-01/msg00595.html [2] https://lists.xenproject.org/archives/html/xen-devel/2021-09/msg02322.html 1. No change v1 -> v2: 1. Add code comments to explain using NR_MEM_BANKS for Arm 2. Refine commit messages. --- xen/arch/arm/include/asm/numa.h | 19 ++++++++++++++++++- xen/include/xen/numa.h | 9 +++++++++ 2 files changed, 27 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index e2bee2bd82..7d6ae36a19 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -3,9 +3,26 @@ #include +#include + typedef u8 nodeid_t; -#ifndef CONFIG_NUMA +#ifdef CONFIG_NUMA + +/* + * It is very likely that if you have more than 64 nodes, you may + * need a lot more than 2 regions per node. So, for Arm, we would + * just define NR_NODE_MEMBLKS as an alias to NR_MEM_BANKS. + * And in the future NR_MEM_BANKS will be bumped for new platforms, + * but for now leave NR_MEM_BANKS as it is on Arm. This avoid to + * have different way to define the value based NUMA vs non-NUMA. + * + * Further discussions can be found here: + * https://lists.xenproject.org/archives/html/xen-devel/2021-09/msg02322.html + */ +#define NR_NODE_MEMBLKS NR_MEM_BANKS + +#else /* Fake one node for now. See also node_online_map. */ #define cpu_to_node(cpu) 0 diff --git a/xen/include/xen/numa.h b/xen/include/xen/numa.h index 29b8c2df89..b86d0851fc 100644 --- a/xen/include/xen/numa.h +++ b/xen/include/xen/numa.h @@ -13,7 +13,16 @@ #define MAX_NUMNODES 1 #endif +/* + * Some architectures may have different considerations for + * number of node memory blocks. They can define their + * NR_NODE_MEMBLKS in asm/numa.h to reflect their architectural + * implementation. If the arch does not have specific implementation, + * the following default NR_NODE_MEMBLKS will be used. + */ +#ifndef NR_NODE_MEMBLKS #define NR_NODE_MEMBLKS (MAX_NUMNODES * 2) +#endif #define vcpu_to_node(v) (cpu_to_node((v)->processor)) From patchwork Tue Apr 25 07:56:40 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222980 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 9D2AEC7EE2A for ; Tue, 25 Apr 2023 07:57:29 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525752.817158 (Exim 4.92) (envelope-from ) id 1prDY6-0005ap-Br; Tue, 25 Apr 2023 07:57:18 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525752.817158; Tue, 25 Apr 2023 07:57:18 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDY6-0005ag-8s; Tue, 25 Apr 2023 07:57:18 +0000 Received: by outflank-mailman (input) for mailman id 525752; Tue, 25 Apr 2023 07:57:16 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDY4-00053E-Rv for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:16 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id cb5e765d-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:16 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 3D17C4B3; Tue, 25 Apr 2023 00:57:59 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id E70743F587; Tue, 25 Apr 2023 00:57:12 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: cb5e765d-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 02/17] xen/arm: implement helpers to get and update NUMA status Date: Tue, 25 Apr 2023 15:56:40 +0800 Message-Id: <20230425075655.4037980-3-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen NUMA has one global and one implementation specific switches. For ACPI NUMA implementation, Xen has acpi_numa, so we introduce device_tree_numa for device tree NUMA implementation. And use enumerations to indicate init, off and on status. arch_numa_disabled will get device_tree_numa status, but for arch_numa_setup we have not provided boot arguments to setup device_tree_numa. So we just return -EINVAL in this patch. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. Rename the first entry of enum dt_numa_status as DT_NUMA_DEFAULT. 2. Make enum dt_numa_status device_tree_numa as __ro_after_init and assign it explicitly to DT_NUMA_DEFAULT. 3. Update the year in copyright to 2023. 4. Don't move the x86 numa_disabled() and make Arm's numa_disabled() a static inline function for !CONFIG_NUMA. v1 -> v2: 1. Use arch_numa_disabled to replace numa_enable_with_firmware. 2. Introduce enumerations for device tree numa status. 3. Use common numa_disabled, drop Arm version numa_disabled. 4. Introduce arch_numa_setup for Arm. 5. Rename bad_srat to numa_bad. 6. Add numa_enable_with_firmware helper. 7. Add numa_disabled helper. 8. Refine commit message. --- xen/arch/arm/include/asm/numa.h | 17 +++++++++++ xen/arch/arm/numa.c | 50 +++++++++++++++++++++++++++++++++ 2 files changed, 67 insertions(+) create mode 100644 xen/arch/arm/numa.c diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 7d6ae36a19..83f60ad05b 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -22,6 +22,8 @@ typedef u8 nodeid_t; */ #define NR_NODE_MEMBLKS NR_MEM_BANKS +extern bool numa_disabled(void); + #else /* Fake one node for now. See also node_online_map. */ @@ -39,6 +41,21 @@ extern mfn_t first_valid_mfn; #define node_start_pfn(nid) (mfn_x(first_valid_mfn)) #define __node_distance(a, b) (20) +static inline bool numa_disabled(void) +{ + return true; +} + +static inline bool arch_numa_unavailable(void) +{ + return true; +} + +static inline bool arch_numa_broken(void) +{ + return true; +} + #endif #define arch_want_default_dmazone() (false) diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c new file mode 100644 index 0000000000..eb5d0632cb --- /dev/null +++ b/xen/arch/arm/numa.c @@ -0,0 +1,50 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Arm Architecture support layer for NUMA. + * + * Copyright (C) 2023 Arm Ltd + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + * + */ +#include +#include + +enum dt_numa_status { + DT_NUMA_DEFAULT, + DT_NUMA_ON, + DT_NUMA_OFF, +}; + +static enum dt_numa_status __ro_after_init device_tree_numa = DT_NUMA_DEFAULT; + +void __init numa_fw_bad(void) +{ + printk(KERN_ERR "NUMA: device tree numa info table not used.\n"); + device_tree_numa = DT_NUMA_OFF; +} + +bool __init arch_numa_unavailable(void) +{ + return device_tree_numa != DT_NUMA_ON; +} + +bool arch_numa_disabled(void) +{ + return device_tree_numa == DT_NUMA_OFF; +} + +int __init arch_numa_setup(const char *opt) +{ + return -EINVAL; +} From patchwork Tue Apr 25 07:56:41 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222981 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id B6260C77B71 for ; Tue, 25 Apr 2023 07:57:33 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525753.817168 (Exim 4.92) (envelope-from ) id 1prDYB-0005va-LF; Tue, 25 Apr 2023 07:57:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525753.817168; Tue, 25 Apr 2023 07:57:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYB-0005vT-Hr; Tue, 25 Apr 2023 07:57:23 +0000 Received: by outflank-mailman (input) for mailman id 525753; Tue, 25 Apr 2023 07:57:23 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYA-0005uB-Uu for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:22 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id ce1616ef-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:20 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id E0A634B3; Tue, 25 Apr 2023 00:58:03 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 4ECB33F587; Tue, 25 Apr 2023 00:57:16 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: ce1616ef-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Andrew Cooper , George Dunlap , Jan Beulich , Wei Liu , =?utf-8?q?Roger_Pau_Monn=C3=A9?= , Henry Wang Subject: [PATCH v4 03/17] xen/arm: implement node distance helpers for Arm Date: Tue, 25 Apr 2023 15:56:41 +0800 Message-Id: <20230425075655.4037980-4-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen We will parse NUMA nodes distances from device tree. So we need a matrix to record the distances between any two nodes we parsed. Accordingly, we provide this node_set_distance API for device tree NUMA to set the distance for any two nodes in this patch. When NUMA initialization failed, __node_distance will return NUMA_REMOTE_DISTANCE, this will help us avoid doing rollback for distance maxtrix when NUMA initialization failed. As both x86 and Arm have implemented __node_distance, so we move its declaration from asm/numa.h to xen/numa.h. At same time, the outdated u8 return value of x86 has been changed to unsigned char. Signed-off-by: Wei Chen Signed-off-by: Henry Wang Acked-by: Jan Beulich # non-Arm parts --- v3 -> v4: 1. s/definition/declaration/ in commit message. 2. Add Acked-by tag from Jan for non-Arm parts. 3. Drop unnecessary initializer for node_distance_map. Pre-set the distance map to NUMA_NO_DISTANCE. 4. Drop NUMA_DISTANCE_UDF_MIN and its usage. 5. Drop EXPORT_SYMBOL(__node_distance). 6. Rework __node_distance()'s return value logic. v2 -> v3: 1. Use __ro_after_init for node_distance_map. 2. Correct format of if condition identation in numa_set_distance(). 3. Drop the unnecessary change to the year of copyright. 4. Use ARRAY_SIZE() to determine node_distance_map's row, column size. v1 -> v2: 1. Use unsigned int/char instead of uint32_t/u8. 2. Re-org the commit message. --- xen/arch/arm/Makefile | 1 + xen/arch/arm/include/asm/numa.h | 12 +++++++ xen/arch/arm/numa.c | 55 +++++++++++++++++++++++++++++++++ xen/arch/x86/include/asm/numa.h | 1 - xen/arch/x86/srat.c | 2 +- xen/include/xen/numa.h | 1 + 6 files changed, 70 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/Makefile b/xen/arch/arm/Makefile index 4d076b278b..9073398d6e 100644 --- a/xen/arch/arm/Makefile +++ b/xen/arch/arm/Makefile @@ -38,6 +38,7 @@ obj-$(CONFIG_LIVEPATCH) += livepatch.o obj-y += mem_access.o obj-y += mm.o obj-y += monitor.o +obj-$(CONFIG_NUMA) += numa.o obj-y += p2m.o obj-y += percpu.o obj-y += platform.o diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 83f60ad05b..96c856a9f7 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -22,7 +22,19 @@ typedef u8 nodeid_t; */ #define NR_NODE_MEMBLKS NR_MEM_BANKS +/* + * In ACPI spec, 0-9 are the reserved values for node distance, + * 10 indicates local node distance, 20 indicates remote node + * distance. Set node distance map in device tree will follow + * the ACPI's definition. + */ +#define NUMA_DISTANCE_UDF_MAX 9 +#define NUMA_LOCAL_DISTANCE 10 +#define NUMA_REMOTE_DISTANCE 20 + extern bool numa_disabled(void); +extern void numa_set_distance(nodeid_t from, nodeid_t to, + unsigned int distance); #else diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index eb5d0632cb..e4f75f314b 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -28,6 +28,12 @@ enum dt_numa_status { static enum dt_numa_status __ro_after_init device_tree_numa = DT_NUMA_DEFAULT; +static unsigned char __ro_after_init +node_distance_map[MAX_NUMNODES][MAX_NUMNODES] = { + [0 ... MAX_NUMNODES - 1] = { [0 ... MAX_NUMNODES - 1] = NUMA_NO_DISTANCE } +}; + + void __init numa_fw_bad(void) { printk(KERN_ERR "NUMA: device tree numa info table not used.\n"); @@ -48,3 +54,52 @@ int __init arch_numa_setup(const char *opt) { return -EINVAL; } + +void __init numa_set_distance(nodeid_t from, nodeid_t to, + unsigned int distance) +{ + if ( from >= ARRAY_SIZE(node_distance_map) || + to >= ARRAY_SIZE(node_distance_map[0]) ) + { + printk(KERN_WARNING + "NUMA: invalid nodes: from=%"PRIu8" to=%"PRIu8" MAX=%"PRIu8"\n", + from, to, MAX_NUMNODES); + return; + } + + /* NUMA defines NUMA_NO_DISTANCE as unreachable and 0-9 are undefined */ + if ( distance >= NUMA_NO_DISTANCE || distance <= NUMA_DISTANCE_UDF_MAX || + (from == to && distance != NUMA_LOCAL_DISTANCE) ) + { + printk(KERN_WARNING + "NUMA: invalid distance: from=%"PRIu8" to=%"PRIu8" distance=%"PRIu32"\n", + from, to, distance); + return; + } + + node_distance_map[from][to] = distance; +} + +unsigned char __node_distance(nodeid_t from, nodeid_t to) +{ + if ( from == to ) + return NUMA_LOCAL_DISTANCE; + + /* + * When NUMA is off, any distance will be treated as unreachable, so + * directly return NUMA_NO_DISTANCE from here as an optimization. + */ + if ( numa_disabled() ) + return NUMA_NO_DISTANCE; + + /* + * Check whether the nodes are in the matrix range. + * When any node is out of range, except from and to nodes are the + * same, we treat them as unreachable. + */ + if ( from >= ARRAY_SIZE(node_distance_map) || + to >= ARRAY_SIZE(node_distance_map[0]) ) + return NUMA_NO_DISTANCE; + + return node_distance_map[from][to]; +} diff --git a/xen/arch/x86/include/asm/numa.h b/xen/arch/x86/include/asm/numa.h index 7866afa408..45456ac441 100644 --- a/xen/arch/x86/include/asm/numa.h +++ b/xen/arch/x86/include/asm/numa.h @@ -22,7 +22,6 @@ extern void init_cpu_to_node(void); #define arch_want_default_dmazone() (num_online_nodes() > 1) void srat_parse_regions(paddr_t addr); -extern u8 __node_distance(nodeid_t a, nodeid_t b); unsigned int arch_get_dma_bitsize(void); #endif diff --git a/xen/arch/x86/srat.c b/xen/arch/x86/srat.c index 56749ddca5..50faf5d352 100644 --- a/xen/arch/x86/srat.c +++ b/xen/arch/x86/srat.c @@ -328,7 +328,7 @@ unsigned int numa_node_to_arch_nid(nodeid_t n) return 0; } -u8 __node_distance(nodeid_t a, nodeid_t b) +unsigned char __node_distance(nodeid_t a, nodeid_t b) { unsigned index; u8 slit_val; diff --git a/xen/include/xen/numa.h b/xen/include/xen/numa.h index b86d0851fc..8356e47b61 100644 --- a/xen/include/xen/numa.h +++ b/xen/include/xen/numa.h @@ -114,6 +114,7 @@ extern bool numa_memblks_available(void); extern bool numa_update_node_memblks(nodeid_t node, unsigned int arch_nid, paddr_t start, paddr_t size, bool hotplug); extern void numa_set_processor_nodes_parsed(nodeid_t node); +extern unsigned char __node_distance(nodeid_t a, nodeid_t b); #else From patchwork Tue Apr 25 07:56:42 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222982 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id B137EC77B76 for ; Tue, 25 Apr 2023 07:57:35 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525754.817177 (Exim 4.92) (envelope-from ) id 1prDYE-0006G1-UQ; Tue, 25 Apr 2023 07:57:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525754.817177; Tue, 25 Apr 2023 07:57:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYE-0006Fs-RC; Tue, 25 Apr 2023 07:57:26 +0000 Received: by outflank-mailman (input) for mailman id 525754; Tue, 25 Apr 2023 07:57:25 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYD-0005uB-Ol for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:25 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id d0231bcb-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:24 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 5E193D75; Tue, 25 Apr 2023 00:58:07 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 0BA1B3F587; Tue, 25 Apr 2023 00:57:20 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d0231bcb-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 04/17] xen/arm: use arch_get_ram_range to get memory ranges from bootinfo Date: Tue, 25 Apr 2023 15:56:42 +0800 Message-Id: <20230425075655.4037980-5-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Implement the same helper "arch_get_ram_range" as x86 for NUMA code to get memory bank from Arm bootinfo. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Use arch_get_ram_range instead of arch_get_memory_map. --- xen/arch/arm/numa.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index e4f75f314b..3fee3789c7 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -103,3 +103,14 @@ unsigned char __node_distance(nodeid_t from, nodeid_t to) return node_distance_map[from][to]; } + +int __init arch_get_ram_range(unsigned int idx, paddr_t *start, paddr_t *end) +{ + if ( idx >= bootinfo.mem.nr_banks ) + return -ENOENT; + + *start = bootinfo.mem.bank[idx].start; + *end = *start + bootinfo.mem.bank[idx].size; + + return 0; +} From patchwork Tue Apr 25 07:56:43 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222983 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id B9D7EC6FD18 for ; Tue, 25 Apr 2023 07:57:40 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525759.817188 (Exim 4.92) (envelope-from ) id 1prDYI-0006ix-7K; Tue, 25 Apr 2023 07:57:30 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525759.817188; Tue, 25 Apr 2023 07:57:30 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYI-0006io-3O; Tue, 25 Apr 2023 07:57:30 +0000 Received: by outflank-mailman (input) for mailman id 525759; Tue, 25 Apr 2023 07:57:29 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYH-0005uB-AL for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:29 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id d22f07c2-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:27 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id B4B654B3; Tue, 25 Apr 2023 00:58:10 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 6A5033F882; Tue, 25 Apr 2023 00:57:24 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d22f07c2-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 05/17] xen/arm: build NUMA cpu_to_node map in dt_smp_init_cpus Date: Tue, 25 Apr 2023 15:56:43 +0800 Message-Id: <20230425075655.4037980-6-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen NUMA implementation has a cpu_to_node array to store CPU to NODE map. Xen is using CPU logical ID in runtime components, so we use CPU logical ID as CPU index in cpu_to_node. In device tree case, cpu_logical_map is created in dt_smp_init_cpus. So, when NUMA is enabled, dt_smp_init_cpus will fetch CPU NUMA id at the same time for cpu_to_node. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Use static inline to replace macros to perform function paramerters type check. 2. Add numa_disabled to gate the numa-node-id check for CONFIG_NUMA on but numa disabled user case. 3. Use macro instead of static inline function to stub numa_set_node. --- xen/arch/arm/include/asm/numa.h | 4 ++++ xen/arch/arm/smpboot.c | 36 +++++++++++++++++++++++++++++++++ 2 files changed, 40 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 96c856a9f7..97d4a67dea 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -68,6 +68,10 @@ static inline bool arch_numa_broken(void) return true; } +static inline void numa_set_node(unsigned int cpu, nodeid_t node) +{ +} + #endif #define arch_want_default_dmazone() (false) diff --git a/xen/arch/arm/smpboot.c b/xen/arch/arm/smpboot.c index 4a89b3a834..da7f2afd97 100644 --- a/xen/arch/arm/smpboot.c +++ b/xen/arch/arm/smpboot.c @@ -118,7 +118,12 @@ static void __init dt_smp_init_cpus(void) { [0 ... NR_CPUS - 1] = MPIDR_INVALID }; + static nodeid_t node_map[NR_CPUS] __initdata = + { + [0 ... NR_CPUS - 1] = NUMA_NO_NODE + }; bool bootcpu_valid = false; + unsigned int nid = 0; int rc; mpidr = system_cpuinfo.mpidr.bits & MPIDR_HWID_MASK; @@ -169,6 +174,28 @@ static void __init dt_smp_init_cpus(void) continue; } + if ( IS_ENABLED(CONFIG_NUMA) ) + { + /* + * When CONFIG_NUMA is set, try to fetch numa infomation + * from CPU dts node, otherwise the nid is always 0. + */ + if ( !dt_property_read_u32(cpu, "numa-node-id", &nid) ) + { + printk(XENLOG_WARNING + "cpu[%d] dts path: %s: doesn't have numa information!\n", + cpuidx, dt_node_full_name(cpu)); + /* + * During the early stage of NUMA initialization, when Xen + * found any CPU dts node doesn't have numa-node-id info, the + * NUMA will be treated as off, all CPU will be set to a FAKE + * node 0. So if we get numa-node-id failed here, we should + * set nid to 0. + */ + nid = 0; + } + } + /* * 8 MSBs must be set to 0 in the DT since the reg property * defines the MPIDR[23:0] @@ -228,9 +255,13 @@ static void __init dt_smp_init_cpus(void) { printk("cpu%d init failed (hwid %"PRIregister"): %d\n", i, hwid, rc); tmp_map[i] = MPIDR_INVALID; + node_map[i] = NUMA_NO_NODE; } else + { tmp_map[i] = hwid; + node_map[i] = nid; + } } if ( !bootcpu_valid ) @@ -246,6 +277,11 @@ static void __init dt_smp_init_cpus(void) continue; cpumask_set_cpu(i, &cpu_possible_map); cpu_logical_map(i) = tmp_map[i]; + + nid = node_map[i]; + if ( nid >= MAX_NUMNODES ) + nid = 0; + numa_set_node(i, nid); } } From patchwork Tue Apr 25 07:56:44 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222984 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 3C530C77B76 for ; Tue, 25 Apr 2023 07:57:43 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525760.817198 (Exim 4.92) (envelope-from ) id 1prDYK-00076a-Ho; Tue, 25 Apr 2023 07:57:32 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525760.817198; Tue, 25 Apr 2023 07:57:32 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYK-00076N-DO; Tue, 25 Apr 2023 07:57:32 +0000 Received: by outflank-mailman (input) for mailman id 525760; Tue, 25 Apr 2023 07:57:31 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYJ-00053E-8K for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:31 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id d41347b7-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:30 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 0E99E4B3; Tue, 25 Apr 2023 00:58:14 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id B7A583F587; Tue, 25 Apr 2023 00:57:27 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d41347b7-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 06/17] xen/arm: Add boot and secondary CPU to NUMA system Date: Tue, 25 Apr 2023 15:56:44 +0800 Message-Id: <20230425075655.4037980-7-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen In this patch, we make NUMA node online and add cpu to its NUMA node. This will make NUMA-aware components have NUMA affinity data to support their work. To keep the mostly the same behavior of x86, we use numa_detect_cpu_node to online node. The difference is that, we have prepared cpu_to_node in dt_smp_init_cpus, so we don't need to setup cpu_to_node in numa_detect_cpu_node. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Use unsigned int instead of int for cpu id. 2. Use static inline for stub to do type check. --- xen/arch/arm/include/asm/numa.h | 9 +++++++++ xen/arch/arm/numa.c | 10 ++++++++++ xen/arch/arm/setup.c | 5 +++++ 3 files changed, 24 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 97d4a67dea..b04ace26db 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -35,6 +35,7 @@ typedef u8 nodeid_t; extern bool numa_disabled(void); extern void numa_set_distance(nodeid_t from, nodeid_t to, unsigned int distance); +extern void numa_detect_cpu_node(unsigned int cpu); #else @@ -72,6 +73,14 @@ static inline void numa_set_node(unsigned int cpu, nodeid_t node) { } +static inline void numa_add_cpu(unsigned int cpu) +{ +} + +static inline void numa_detect_cpu_node(unsigned int cpu) +{ +} + #endif #define arch_want_default_dmazone() (false) diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index 3fee3789c7..7ae9ace15d 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -80,6 +80,16 @@ void __init numa_set_distance(nodeid_t from, nodeid_t to, node_distance_map[from][to] = distance; } +void numa_detect_cpu_node(unsigned int cpu) +{ + nodeid_t node = cpu_to_node[cpu]; + + if ( node == NUMA_NO_NODE ) + node = 0; + + node_set_online(node); +} + unsigned char __node_distance(nodeid_t from, nodeid_t to) { if ( from == to ) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 6f9f4d8c8a..09e18d32df 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -1205,6 +1205,11 @@ void __init start_xen(unsigned long boot_phys_offset, for_each_present_cpu ( i ) { + /* Detect and online node based on cpu_to_node[]. */ + numa_detect_cpu_node(i); + /* Set up node_to_cpumask based on cpu_to_node[]. */ + numa_add_cpu(i); + if ( (num_online_cpus() < nr_cpu_ids) && !cpu_online(i) ) { int ret = cpu_up(i); From patchwork Tue Apr 25 07:56:45 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222985 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 2B2C1C77B61 for ; Tue, 25 Apr 2023 07:57:45 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525764.817208 (Exim 4.92) (envelope-from ) id 1prDYN-0007dk-Ps; Tue, 25 Apr 2023 07:57:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525764.817208; Tue, 25 Apr 2023 07:57:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYN-0007dZ-Mr; Tue, 25 Apr 2023 07:57:35 +0000 Received: by outflank-mailman (input) for mailman id 525764; Tue, 25 Apr 2023 07:57:34 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYM-00053E-U1 for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:34 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id d62057d5-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:34 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 5C80D4B3; Tue, 25 Apr 2023 00:58:17 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 12B4A3F587; Tue, 25 Apr 2023 00:57:30 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d62057d5-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 07/17] xen/arm: introduce a helper to parse device tree processor node Date: Tue, 25 Apr 2023 15:56:45 +0800 Message-Id: <20230425075655.4037980-8-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Processor NUMA ID information is stored in device tree's processor node as "numa-node-id". We need a new helper to parse this ID from processor node. If we get this ID from processor node, this ID's validity still need to be checked. Once we got a invalid NUMA ID from any processor node, the device tree will be marked as NUMA information invalid. Since new helpers need to know the NUMA status, move the enum dt_numa_status to the Arm NUMA header. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. Move the enum dt_numa_status to the Arm NUMA header. 2. Update the year in copyright to 2023. v1 -> v2: 1. Move numa_disabled from fdt_numa_processor_affinity_init to fdt_parse_numa_cpu_node. 2. Move invalid NUMA id check to fdt_parse_numa_cpu_node. 3. Return ENODATA for normal dtb without NUMA info. 4. Use NUMA status helpers instead of SRAT functions. --- xen/arch/arm/Makefile | 1 + xen/arch/arm/include/asm/numa.h | 8 +++++ xen/arch/arm/numa.c | 8 +---- xen/arch/arm/numa_device_tree.c | 64 +++++++++++++++++++++++++++++++++ 4 files changed, 74 insertions(+), 7 deletions(-) create mode 100644 xen/arch/arm/numa_device_tree.c diff --git a/xen/arch/arm/Makefile b/xen/arch/arm/Makefile index 9073398d6e..bbc68e3735 100644 --- a/xen/arch/arm/Makefile +++ b/xen/arch/arm/Makefile @@ -39,6 +39,7 @@ obj-y += mem_access.o obj-y += mm.o obj-y += monitor.o obj-$(CONFIG_NUMA) += numa.o +obj-$(CONFIG_DEVICE_TREE_NUMA) += numa_device_tree.o obj-y += p2m.o obj-y += percpu.o obj-y += platform.o diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index b04ace26db..2987158d16 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -22,6 +22,14 @@ typedef u8 nodeid_t; */ #define NR_NODE_MEMBLKS NR_MEM_BANKS +enum dt_numa_status { + DT_NUMA_DEFAULT, + DT_NUMA_ON, + DT_NUMA_OFF, +}; + +extern enum dt_numa_status device_tree_numa; + /* * In ACPI spec, 0-9 are the reserved values for node distance, * 10 indicates local node distance, 20 indicates remote node diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index 7ae9ace15d..b953c2574a 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -20,13 +20,7 @@ #include #include -enum dt_numa_status { - DT_NUMA_DEFAULT, - DT_NUMA_ON, - DT_NUMA_OFF, -}; - -static enum dt_numa_status __ro_after_init device_tree_numa = DT_NUMA_DEFAULT; +enum dt_numa_status __ro_after_init device_tree_numa = DT_NUMA_DEFAULT; static unsigned char __ro_after_init node_distance_map[MAX_NUMNODES][MAX_NUMNODES] = { diff --git a/xen/arch/arm/numa_device_tree.c b/xen/arch/arm/numa_device_tree.c new file mode 100644 index 0000000000..83601c83e7 --- /dev/null +++ b/xen/arch/arm/numa_device_tree.c @@ -0,0 +1,64 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Arm Architecture support layer for device tree NUMA. + * + * Copyright (C) 2023 Arm Ltd + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + * + */ +#include +#include +#include +#include +#include + +/* Callback for device tree processor affinity */ +static int __init fdt_numa_processor_affinity_init(nodeid_t node) +{ + numa_set_processor_nodes_parsed(node); + device_tree_numa = DT_NUMA_ON; + + printk(KERN_INFO "DT: NUMA node %"PRIu8" processor parsed\n", node); + + return 0; +} + +/* Parse CPU NUMA node info */ +static int __init fdt_parse_numa_cpu_node(const void *fdt, int node) +{ + unsigned int nid; + + if ( numa_disabled() ) + return -EINVAL; + + /* + * device_tree_get_u32 will return NUMA_NO_NODE when this CPU + * DT node doesn't have numa-node-id. This can help us to + * distinguish a bad DTB and a normal DTB without NUMA info. + */ + nid = device_tree_get_u32(fdt, node, "numa-node-id", NUMA_NO_NODE); + if ( nid == NUMA_NO_NODE ) + { + numa_fw_bad(); + return -ENODATA; + } + else if ( nid >= MAX_NUMNODES ) + { + printk(XENLOG_ERR "DT: CPU NUMA node id %u is invalid\n", nid); + numa_fw_bad(); + return -EINVAL; + } + + return fdt_numa_processor_affinity_init(nid); +} From patchwork Tue Apr 25 07:56:46 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222986 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 82FC3C77B61 for ; Tue, 25 Apr 2023 07:57:50 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525768.817218 (Exim 4.92) (envelope-from ) id 1prDYT-0008EL-BE; Tue, 25 Apr 2023 07:57:41 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525768.817218; Tue, 25 Apr 2023 07:57:41 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYT-0008EC-6Y; Tue, 25 Apr 2023 07:57:41 +0000 Received: by outflank-mailman (input) for mailman id 525768; Tue, 25 Apr 2023 07:57:40 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYS-0005uB-4Y for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:40 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id d81a8885-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:37 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id B31DE4B3; Tue, 25 Apr 2023 00:58:20 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 697863F587; Tue, 25 Apr 2023 00:57:34 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: d81a8885-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 08/17] xen/arm: introduce a helper to parse device tree memory node Date: Tue, 25 Apr 2023 15:56:46 +0800 Message-Id: <20230425075655.4037980-9-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Memory blocks' NUMA ID information is stored in device tree's memory nodes as "numa-node-id". We need a new helper to parse and verify this ID from memory nodes. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Move numa_disabled check to fdt_parse_numa_memory_node. 2. Use numa_bad to replace bad_srat. 3. Replace tabs by spaces. 4. Align parameters. 5. return ENODATA for a normal dtb without numa info. 6. Un-addressed comment: "Why not parse numa-node-id and call fdt_numa_memory_affinity_init from xen/arch/arm/bootfdt.c:device_tree_get_meminfo. Is it because device_tree_get_meminfo is called too early?" I checked the device_tree_get_meminfo code and I think the answer is similar as I reply in RFC. I prefer a unify numa initialization entry. Don't want to make numa parse code in different places. 7. Use node id as dummy PXM for numa_update_node_memblks. --- xen/arch/arm/numa_device_tree.c | 89 +++++++++++++++++++++++++++++++++ 1 file changed, 89 insertions(+) diff --git a/xen/arch/arm/numa_device_tree.c b/xen/arch/arm/numa_device_tree.c index 83601c83e7..38f5e93d1d 100644 --- a/xen/arch/arm/numa_device_tree.c +++ b/xen/arch/arm/numa_device_tree.c @@ -34,6 +34,26 @@ static int __init fdt_numa_processor_affinity_init(nodeid_t node) return 0; } +/* Callback for parsing of the memory regions affinity */ +static int __init fdt_numa_memory_affinity_init(nodeid_t node, + paddr_t start, paddr_t size) +{ + if ( !numa_memblks_available() ) + { + dprintk(XENLOG_WARNING, + "Too many NUMA entries, try bigger NR_NODE_MEMBLKS\n"); + return -EINVAL; + } + + numa_fw_nid_name = "numa-node-id"; + if ( !numa_update_node_memblks(node, node, start, size, false) ) + return -EINVAL; + + device_tree_numa = DT_NUMA_ON; + + return 0; +} + /* Parse CPU NUMA node info */ static int __init fdt_parse_numa_cpu_node(const void *fdt, int node) { @@ -62,3 +82,72 @@ static int __init fdt_parse_numa_cpu_node(const void *fdt, int node) return fdt_numa_processor_affinity_init(nid); } + +/* Parse memory node NUMA info */ +static int __init fdt_parse_numa_memory_node(const void *fdt, int node, + const char *name, + unsigned int addr_cells, + unsigned int size_cells) +{ + unsigned int nid; + int ret = 0, len; + paddr_t addr, size; + const struct fdt_property *prop; + unsigned int idx, ranges; + const __be32 *addresses; + + if ( numa_disabled() ) + return -EINVAL; + + /* + * device_tree_get_u32 will return NUMA_NO_NODE when this memory + * DT node doesn't have numa-node-id. This can help us to + * distinguish a bad DTB and a normal DTB without NUMA info. + */ + nid = device_tree_get_u32(fdt, node, "numa-node-id", NUMA_NO_NODE); + if ( node == NUMA_NO_NODE ) + { + numa_fw_bad(); + return -ENODATA; + } + else if ( nid >= MAX_NUMNODES ) + { + printk(XENLOG_WARNING "Node id %u exceeds maximum value\n", nid); + goto invalid_data; + } + + prop = fdt_get_property(fdt, node, "reg", &len); + if ( !prop ) + { + printk(XENLOG_WARNING + "fdt: node `%s': missing `reg' property\n", name); + goto invalid_data; + } + + addresses = (const __be32 *)prop->data; + ranges = len / (sizeof(__be32)* (addr_cells + size_cells)); + for ( idx = 0; idx < ranges; idx++ ) + { + device_tree_get_reg(&addresses, addr_cells, size_cells, &addr, &size); + /* Skip zero size ranges */ + if ( !size ) + continue; + + ret = fdt_numa_memory_affinity_init(nid, addr, size); + if ( ret ) + goto invalid_data; + } + + if ( idx == 0 ) + { + printk(XENLOG_ERR + "bad property in memory node, idx=%d ret=%d\n", idx, ret); + goto invalid_data; + } + + return 0; + +invalid_data: + numa_fw_bad(); + return -EINVAL; +} From patchwork Tue Apr 25 07:56:47 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222988 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id B613FC6FD18 for ; Tue, 25 Apr 2023 07:59:13 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525783.817238 (Exim 4.92) (envelope-from ) id 1prDZo-0001pw-2B; Tue, 25 Apr 2023 07:59:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525783.817238; Tue, 25 Apr 2023 07:59:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZn-0001pl-U4; Tue, 25 Apr 2023 07:59:03 +0000 Received: by outflank-mailman (input) for mailman id 525783; Tue, 25 Apr 2023 07:59:02 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYU-00053E-1l for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:42 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id da1d13b8-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:40 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 18621D75; Tue, 25 Apr 2023 00:58:24 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id C25383F587; Tue, 25 Apr 2023 00:57:37 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: da1d13b8-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 09/17] xen/arm: introduce a helper to parse device tree NUMA distance map Date: Tue, 25 Apr 2023 15:56:47 +0800 Message-Id: <20230425075655.4037980-10-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen A NUMA aware device tree will provide a "distance-map" node to describe distance between any two nodes. This patch introduce a new helper to parse this distance map. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. The distance map default value is now NUMA_NO_DISTANCE, update the logic accordingly and add in-code comment as a note. v2 -> v3: 1. No change. v1 -> v2: 1. Get rid of useless braces. 2. Use new NUMA status helper. 3. Use PRIu32 to replace u in print messages. 4. Fix opposite = __node_distance(to, from). 5. disable dtb numa info table when we find an invalid data in dtb. --- xen/arch/arm/numa_device_tree.c | 108 ++++++++++++++++++++++++++++++++ 1 file changed, 108 insertions(+) diff --git a/xen/arch/arm/numa_device_tree.c b/xen/arch/arm/numa_device_tree.c index 38f5e93d1d..9d9d09273e 100644 --- a/xen/arch/arm/numa_device_tree.c +++ b/xen/arch/arm/numa_device_tree.c @@ -151,3 +151,111 @@ invalid_data: numa_fw_bad(); return -EINVAL; } + +/* Parse NUMA distance map v1 */ +static int __init fdt_parse_numa_distance_map_v1(const void *fdt, int node) +{ + const struct fdt_property *prop; + const __be32 *matrix; + unsigned int i, entry_count; + int len; + + printk(XENLOG_INFO "NUMA: parsing numa-distance-map\n"); + + prop = fdt_get_property(fdt, node, "distance-matrix", &len); + if ( !prop ) + { + printk(XENLOG_WARNING + "NUMA: No distance-matrix property in distance-map\n"); + goto invalid_data; + } + + if ( len % sizeof(__be32) != 0 ) + { + printk(XENLOG_WARNING + "distance-matrix in node is not a multiple of u32\n"); + goto invalid_data; + } + + entry_count = len / sizeof(__be32); + if ( entry_count == 0 ) + { + printk(XENLOG_WARNING "NUMA: Invalid distance-matrix\n"); + goto invalid_data; + } + + matrix = (const __be32 *)prop->data; + for ( i = 0; i + 2 < entry_count; i += 3 ) + { + unsigned int from, to, distance, opposite; + + from = dt_next_cell(1, &matrix); + to = dt_next_cell(1, &matrix); + distance = dt_next_cell(1, &matrix); + if ( (from == to && distance != NUMA_LOCAL_DISTANCE) || + (from != to && distance <= NUMA_LOCAL_DISTANCE) ) + { + printk(XENLOG_WARNING + "NUMA: Invalid distance: NODE#%"PRIu32"->NODE#%"PRIu32":%"PRIu32"\n", + from, to, distance); + goto invalid_data; + } + + printk(XENLOG_INFO "NUMA: distance: NODE#%"PRIu32"->NODE#%"PRIu32":%"PRIu32"\n", + from, to, distance); + + /* Get opposite way distance */ + opposite = __node_distance(to, from); + /* The default value in node_distance_map is NUMA_NO_DISTANCE */ + if ( opposite == NUMA_NO_DISTANCE ) + { + /* Bi-directions are not set, set both */ + numa_set_distance(from, to, distance); + numa_set_distance(to, from, distance); + } + else + { + /* + * Opposite way distance has been set to a different value. + * It may be a firmware device tree bug? + */ + if ( opposite != distance ) + { + /* + * In device tree NUMA distance-matrix binding: + * https://www.kernel.org/doc/Documentation/devicetree/bindings/numa.txt + * There is a notes mentions: + * "Each entry represents distance from first node to + * second node. The distances are equal in either + * direction." + * + * That means device tree doesn't permit this case. + * But in ACPI spec, it cares to specifically permit this + * case: + * "Except for the relative distance from a System Locality + * to itself, each relative distance is stored twice in the + * matrix. This provides the capability to describe the + * scenario where the relative distances for the two + * directions between System Localities is different." + * + * That means a real machine allows such NUMA configuration. + * So, place a WARNING here to notice system administrators, + * is it the specail case that they hijack the device tree + * to support their rare machines? + */ + printk(XENLOG_WARNING + "Un-matched bi-direction! NODE#%"PRIu32"->NODE#%"PRIu32":%"PRIu32", NODE#%"PRIu32"->NODE#%"PRIu32":%"PRIu32"\n", + from, to, distance, to, from, opposite); + } + + /* Opposite way distance has been set, just set this way */ + numa_set_distance(from, to, distance); + } + } + + return 0; + +invalid_data: + numa_fw_bad(); + return -EINVAL; +} From patchwork Tue Apr 25 07:56:48 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222993 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 83299C77B71 for ; Tue, 25 Apr 2023 07:59:19 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525799.817287 (Exim 4.92) (envelope-from ) id 1prDZu-00037K-Sc; Tue, 25 Apr 2023 07:59:10 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525799.817287; Tue, 25 Apr 2023 07:59:10 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZu-00035s-Hg; Tue, 25 Apr 2023 07:59:10 +0000 Received: by outflank-mailman (input) for mailman id 525799; Tue, 25 Apr 2023 07:59:09 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYX-0005uB-NO for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:45 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id dc0f7b28-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:44 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 779504B3; Tue, 25 Apr 2023 00:58:27 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 2D11D3F587; Tue, 25 Apr 2023 00:57:40 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: dc0f7b28-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 10/17] xen/arm: unified entry to parse all NUMA data from device tree Date: Tue, 25 Apr 2023 15:56:48 +0800 Message-Id: <20230425075655.4037980-11-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen In this function, we scan the whole device tree to parse CPU node id, memory node id and distance-map. Though early_scan_node will invoke a handler to process memory nodes. If we want to parse memory node id in that handler, we have to embed NUMA parse code in that handler. But we still need to scan whole device tree to find CPU NUMA id and distance-map. In this case, we include memory NUMA id parse in this function too. Another benefit is that we have a unique entry for device tree NUMA data parse. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Fix typos in commit message. 2. Fix code style and align parameters. 3. Use strncmp to replace memcmp. --- xen/arch/arm/include/asm/numa.h | 1 + xen/arch/arm/numa_device_tree.c | 30 ++++++++++++++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 2987158d16..15308f5a36 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -44,6 +44,7 @@ extern bool numa_disabled(void); extern void numa_set_distance(nodeid_t from, nodeid_t to, unsigned int distance); extern void numa_detect_cpu_node(unsigned int cpu); +extern int numa_device_tree_init(const void *fdt); #else diff --git a/xen/arch/arm/numa_device_tree.c b/xen/arch/arm/numa_device_tree.c index 9d9d09273e..bf5f112b92 100644 --- a/xen/arch/arm/numa_device_tree.c +++ b/xen/arch/arm/numa_device_tree.c @@ -259,3 +259,33 @@ invalid_data: numa_fw_bad(); return -EINVAL; } + +static int __init fdt_scan_numa_nodes(const void *fdt, int node, + const char *uname, int depth, + unsigned int address_cells, + unsigned int size_cells, void *data) +{ + int len, ret = 0; + const void *prop; + + prop = fdt_getprop(fdt, node, "device_type", &len); + if ( prop ) + { + if ( strncmp(prop, "cpu", len) == 0 ) + ret = fdt_parse_numa_cpu_node(fdt, node); + else if ( strncmp(prop, "memory", len) == 0 ) + ret = fdt_parse_numa_memory_node(fdt, node, uname, + address_cells, size_cells); + } + else if ( fdt_node_check_compatible(fdt, node, + "numa-distance-map-v1") == 0 ) + ret = fdt_parse_numa_distance_map_v1(fdt, node); + + return ret; +} + +/* Initialize NUMA from device tree */ +int __init numa_device_tree_init(const void *fdt) +{ + return device_tree_for_each_node(fdt, 0, fdt_scan_numa_nodes, NULL); +} From patchwork Tue Apr 25 07:56:49 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222990 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 13F5EC77B61 for ; Tue, 25 Apr 2023 07:59:17 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525779.817229 (Exim 4.92) (envelope-from ) id 1prDZm-0001ae-SE; Tue, 25 Apr 2023 07:59:02 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525779.817229; Tue, 25 Apr 2023 07:59:02 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZm-0001aS-LH; Tue, 25 Apr 2023 07:59:02 +0000 Received: by outflank-mailman (input) for mailman id 525779; Tue, 25 Apr 2023 07:59:01 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYb-0005uB-4u for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:49 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id de212a70-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:47 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id D010A4B3; Tue, 25 Apr 2023 00:58:30 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 85F5B3F587; Tue, 25 Apr 2023 00:57:44 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: de212a70-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 11/17] xen/arm: keep guest still be NUMA unware Date: Tue, 25 Apr 2023 15:56:49 +0800 Message-Id: <20230425075655.4037980-12-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen The NUMA information provided in the host Device-Tree are only for Xen. For dom0, we want to hide them as they may be different (for now, dom0 is still not aware of NUMA) The CPU and memory nodes are recreated from scratch for the domain. So we already skip the "numa-node-id" property for these two types of nodes. However, some devices like PCIe may have "numa-node-id" property too. We have to skip them as well. Signed-off-by: Wei Chen Signed-off-by: Henry Wang Reviewed-by: Stefano Stabellini --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Add Rb --- xen/arch/arm/domain_build.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index f80fdd1af2..2bf586fc45 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -1185,6 +1185,10 @@ static int __init write_properties(struct domain *d, struct kernel_info *kinfo, continue; } + /* Dom0 is currently NUMA unaware */ + if ( dt_property_name_is_equal(prop, "numa-node-id") ) + continue; + res = fdt_property(kinfo->fdt, prop->name, prop_data, prop_len); if ( res ) @@ -2567,6 +2571,8 @@ static int __init handle_node(struct domain *d, struct kernel_info *kinfo, DT_MATCH_TYPE("memory"), /* The memory mapped timer is not supported by Xen. */ DT_MATCH_COMPATIBLE("arm,armv7-timer-mem"), + /* Numa info doesn't need to be exposed to Domain-0 */ + DT_MATCH_COMPATIBLE("numa-distance-map-v1"), { /* sentinel */ }, }; static const struct dt_device_match timer_matches[] __initconst = From patchwork Tue Apr 25 07:56:50 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222994 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 37E88C6FD18 for ; Tue, 25 Apr 2023 07:59:20 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525798.817281 (Exim 4.92) (envelope-from ) id 1prDZu-0002zq-Bx; Tue, 25 Apr 2023 07:59:10 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525798.817281; Tue, 25 Apr 2023 07:59:10 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZu-0002xu-2V; Tue, 25 Apr 2023 07:59:10 +0000 Received: by outflank-mailman (input) for mailman id 525798; Tue, 25 Apr 2023 07:59:08 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYe-0005uB-Si for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:52 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id e0211fc5-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:50 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 23EC4FEC; Tue, 25 Apr 2023 00:58:34 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id CE20E3F587; Tue, 25 Apr 2023 00:57:47 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: e0211fc5-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 12/17] xen/arm: enable device tree based NUMA in system init Date: Tue, 25 Apr 2023 15:56:50 +0800 Message-Id: <20230425075655.4037980-13-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen In this patch, we can start to create NUMA system that is based on device tree. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. replace ~0 by INVALID_PADDR. 2. only print error messages for invalid dtb data. 3. remove unnecessary return. 4. remove the parameter of numa_init. --- xen/arch/arm/include/asm/numa.h | 5 +++ xen/arch/arm/numa.c | 57 +++++++++++++++++++++++++++++++++ xen/arch/arm/setup.c | 7 ++++ 3 files changed, 69 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 15308f5a36..55ac4665db 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -45,6 +45,7 @@ extern void numa_set_distance(nodeid_t from, nodeid_t to, unsigned int distance); extern void numa_detect_cpu_node(unsigned int cpu); extern int numa_device_tree_init(const void *fdt); +extern void numa_init(void); #else @@ -90,6 +91,10 @@ static inline void numa_detect_cpu_node(unsigned int cpu) { } +static inline void numa_init(void) +{ +} + #endif #define arch_want_default_dmazone() (false) diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index b953c2574a..f4635e0277 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -18,7 +18,11 @@ * */ #include +#include +#include #include +#include +#include enum dt_numa_status __ro_after_init device_tree_numa = DT_NUMA_DEFAULT; @@ -108,6 +112,59 @@ unsigned char __node_distance(nodeid_t from, nodeid_t to) return node_distance_map[from][to]; } +void __init numa_init(void) +{ + unsigned int idx; + paddr_t ram_start = INVALID_PADDR; + paddr_t ram_size = 0; + paddr_t ram_end = 0; + + /* NUMA has been turned off through Xen parameters */ + if ( numa_off ) + goto mem_init; + + /* Initialize NUMA from device tree when system is not ACPI booted */ + if ( acpi_disabled ) + { + int ret = numa_device_tree_init(device_tree_flattened); + if ( ret ) + { + numa_off = true; + if ( ret == -EINVAL ) + printk(XENLOG_WARNING + "Init NUMA from device tree failed, ret=%d\n", ret); + } + } + else + { + /* We don't support NUMA for ACPI boot currently */ + printk(XENLOG_WARNING + "ACPI NUMA has not been supported yet, NUMA off!\n"); + numa_off = true; + } + +mem_init: + /* + * Find the minimal and maximum address of RAM, NUMA will + * build a memory to node mapping table for the whole range. + */ + ram_start = bootinfo.mem.bank[0].start; + ram_size = bootinfo.mem.bank[0].size; + ram_end = ram_start + ram_size; + for ( idx = 1 ; idx < bootinfo.mem.nr_banks; idx++ ) + { + paddr_t bank_start = bootinfo.mem.bank[idx].start; + paddr_t bank_size = bootinfo.mem.bank[idx].size; + paddr_t bank_end = bank_start + bank_size; + + ram_size = ram_size + bank_size; + ram_start = min(ram_start, bank_start); + ram_end = max(ram_end, bank_end); + } + + numa_initmem_init(PFN_UP(ram_start), PFN_DOWN(ram_end)); +} + int __init arch_get_ram_range(unsigned int idx, paddr_t *start, paddr_t *end) { if ( idx >= bootinfo.mem.nr_banks ) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 09e18d32df..f05e233f3a 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -1121,6 +1121,13 @@ void __init start_xen(unsigned long boot_phys_offset, /* Parse the ACPI tables for possible boot-time configuration */ acpi_boot_table_init(); + /* + * Try to initialize NUMA system, if failed, the system will + * fallback to uniform system which means system has only 1 + * NUMA node. + */ + numa_init(); + end_boot_allocator(); /* From patchwork Tue Apr 25 07:56:51 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222989 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 46693C77B71 for ; Tue, 25 Apr 2023 07:59:15 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525789.817256 (Exim 4.92) (envelope-from ) id 1prDZp-0002Hm-MZ; Tue, 25 Apr 2023 07:59:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525789.817256; Tue, 25 Apr 2023 07:59:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZp-0002GG-Hc; Tue, 25 Apr 2023 07:59:05 +0000 Received: by outflank-mailman (input) for mailman id 525789; Tue, 25 Apr 2023 07:59:04 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYh-00053E-HI for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:55 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-sth1.inumbo.com (Halon) with ESMTP id e206d91b-e33e-11ed-b223-6b7b168915f2; Tue, 25 Apr 2023 09:57:54 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 79A8EFEC; Tue, 25 Apr 2023 00:58:37 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 272B63F587; Tue, 25 Apr 2023 00:57:50 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: e206d91b-e33e-11ed-b223-6b7b168915f2 From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 13/17] xen/arm: implement numa_node_to_arch_nid for device tree NUMA Date: Tue, 25 Apr 2023 15:56:51 +0800 Message-Id: <20230425075655.4037980-14-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Device tree based NUMA doesn't have the proximity domain like ACPI. So we can return node id directly as arch nid. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Use numa_node_to_arch_nid instead of dummy node_to_pxm. --- xen/arch/arm/include/asm/numa.h | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 55ac4665db..71b95a9a62 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -47,6 +47,15 @@ extern void numa_detect_cpu_node(unsigned int cpu); extern int numa_device_tree_init(const void *fdt); extern void numa_init(void); +/* + * Device tree NUMA doesn't have architecural node id. + * So we can just return node id as arch nid. + */ +static inline unsigned int numa_node_to_arch_nid(nodeid_t n) +{ + return n; +} + #else /* Fake one node for now. See also node_online_map. */ From patchwork Tue Apr 25 07:56:52 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222992 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 272C7C77B76 for ; Tue, 25 Apr 2023 07:59:18 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525792.817268 (Exim 4.92) (envelope-from ) id 1prDZs-0002iD-UQ; Tue, 25 Apr 2023 07:59:08 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525792.817268; Tue, 25 Apr 2023 07:59:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZs-0002i3-RB; Tue, 25 Apr 2023 07:59:08 +0000 Received: by outflank-mailman (input) for mailman id 525792; Tue, 25 Apr 2023 07:59:07 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYl-0005uB-O9 for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:57:59 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id e4132d8b-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:57:57 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id D17FB4B3; Tue, 25 Apr 2023 00:58:40 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 7EE673F587; Tue, 25 Apr 2023 00:57:54 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: e4132d8b-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 14/17] xen/arm: use CONFIG_NUMA to gate node_online_map in smpboot Date: Tue, 25 Apr 2023 15:56:52 +0800 Message-Id: <20230425075655.4037980-15-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen node_online_map in smpboot is still needed for Arm when NUMA is turned off by Kconfig. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. No change. --- xen/arch/arm/smpboot.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/arm/smpboot.c b/xen/arch/arm/smpboot.c index da7f2afd97..4f71cc974a 100644 --- a/xen/arch/arm/smpboot.c +++ b/xen/arch/arm/smpboot.c @@ -41,8 +41,10 @@ integer_param("maxcpus", max_cpus); /* CPU logical map: map xen cpuid to an MPIDR */ register_t __cpu_logical_map[NR_CPUS] = { [0 ... NR_CPUS-1] = MPIDR_INVALID }; +#ifndef CONFIG_NUMA /* Fake one node for now. See also asm/numa.h */ nodemask_t __read_mostly node_online_map = { { [0] = 1UL } }; +#endif /* Xen stack for bringing up the first CPU. */ static unsigned char __initdata cpu0_boot_stack[STACK_SIZE] From patchwork Tue Apr 25 07:56:53 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222991 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 40E46C6FD18 for ; Tue, 25 Apr 2023 07:59:17 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525796.817272 (Exim 4.92) (envelope-from ) id 1prDZt-0002mY-Dz; Tue, 25 Apr 2023 07:59:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525796.817272; Tue, 25 Apr 2023 07:59:09 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZt-0002lz-6o; Tue, 25 Apr 2023 07:59:09 +0000 Received: by outflank-mailman (input) for mailman id 525796; Tue, 25 Apr 2023 07:59:07 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYo-0005uB-CH for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:58:02 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id e5dcbc30-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:58:00 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id D8B0D4B3; Tue, 25 Apr 2023 00:58:43 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id CE0DB3F587; Tue, 25 Apr 2023 00:57:57 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: e5dcbc30-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Henry Wang , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk Subject: [PATCH v4 15/17] xen/arm: Set correct per-cpu cpu_core_mask Date: Tue, 25 Apr 2023 15:56:53 +0800 Message-Id: <20230425075655.4037980-16-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 In the common sysctl command XEN_SYSCTL_physinfo, the cores_per_socket is calculated based on the cpu_core_mask of CPU0. Currently on Arm this is a fixed value 1 (can be checked via xl info), which is not correct. This is because during the Arm cpu online process, set_cpu_sibling_map() only sets the per-cpu cpu_core_mask for itself. cores_per_socket refers to the number of cores that belong to the same socket (NUMA node). Therefore, this commit introduces a helper function numa_set_cpu_core_mask(cpu), which sets the per-cpu cpu_core_mask to the cpus in the same NUMA node as cpu. Calling this function at the boot time can ensure the correct cpu_core_mask, leading to the correct cores_per_socket to be returned by XEN_SYSCTL_physinfo. Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. New patch --- xen/arch/arm/include/asm/numa.h | 7 +++++++ xen/arch/arm/numa.c | 11 +++++++++++ xen/arch/arm/setup.c | 5 +++++ 3 files changed, 23 insertions(+) diff --git a/xen/arch/arm/include/asm/numa.h b/xen/arch/arm/include/asm/numa.h index 71b95a9a62..d4c89909d0 100644 --- a/xen/arch/arm/include/asm/numa.h +++ b/xen/arch/arm/include/asm/numa.h @@ -46,6 +46,7 @@ extern void numa_set_distance(nodeid_t from, nodeid_t to, extern void numa_detect_cpu_node(unsigned int cpu); extern int numa_device_tree_init(const void *fdt); extern void numa_init(void); +extern void numa_set_cpu_core_mask(int cpu); /* * Device tree NUMA doesn't have architecural node id. @@ -62,6 +63,12 @@ static inline unsigned int numa_node_to_arch_nid(nodeid_t n) #define cpu_to_node(cpu) 0 #define node_to_cpumask(node) (cpu_online_map) +static inline void numa_set_cpu_core_mask(int cpu) +{ + cpumask_or(per_cpu(cpu_core_mask, cpu), + per_cpu(cpu_core_mask, cpu), &cpu_possible_map); +} + /* * TODO: make first_valid_mfn static when NUMA is supported on Arm, this * is required because the dummy helpers are using it. diff --git a/xen/arch/arm/numa.c b/xen/arch/arm/numa.c index f4635e0277..74e4dc2c67 100644 --- a/xen/arch/arm/numa.c +++ b/xen/arch/arm/numa.c @@ -53,6 +53,17 @@ int __init arch_numa_setup(const char *opt) return -EINVAL; } +void numa_set_cpu_core_mask(int cpu) +{ + nodeid_t node = cpu_to_node[cpu]; + + if ( node == NUMA_NO_NODE ) + node = 0; + + cpumask_or(per_cpu(cpu_core_mask, cpu), + per_cpu(cpu_core_mask, cpu), &node_to_cpumask(node)); +} + void __init numa_set_distance(nodeid_t from, nodeid_t to, unsigned int distance) { diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index f05e233f3a..7cef913b7c 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -1226,6 +1226,11 @@ void __init start_xen(unsigned long boot_phys_offset, } printk("Brought up %ld CPUs\n", (long)num_online_cpus()); + + /* Set per-cpu cpu_core_mask to cpus that belongs to the same NUMA node. */ + for_each_online_cpu ( i ) + numa_set_cpu_core_mask(i); + /* TODO: smp_cpus_done(); */ /* This should be done in a vpmu driver but we do not have one yet. */ From patchwork Tue Apr 25 07:56:54 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222995 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 87469C77B7F for ; Tue, 25 Apr 2023 07:59:20 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525801.817290 (Exim 4.92) (envelope-from ) id 1prDZv-0003Cx-4H; Tue, 25 Apr 2023 07:59:11 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525801.817290; Tue, 25 Apr 2023 07:59:11 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZu-0003BT-W2; Tue, 25 Apr 2023 07:59:10 +0000 Received: by outflank-mailman (input) for mailman id 525801; Tue, 25 Apr 2023 07:59:09 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYr-0005uB-FT for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:58:05 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id e7d41e6c-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:58:03 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 31B9ED75; Tue, 25 Apr 2023 00:58:47 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id DBC503F587; Tue, 25 Apr 2023 00:58:00 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: e7d41e6c-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Stefano Stabellini , Julien Grall , Bertrand Marquis , Volodymyr Babchuk , Henry Wang Subject: [PATCH v4 16/17] xen/arm: Provide Kconfig options for Arm to enable NUMA Date: Tue, 25 Apr 2023 15:56:54 +0800 Message-Id: <20230425075655.4037980-17-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Arm platforms support both ACPI and device tree. We don't want users to select device tree NUMA or ACPI NUMA manually. We hope users can just enable NUMA for Arm, and device tree NUMA and ACPI NUMA can be selected depends on device tree feature and ACPI feature status automatically. In this case, these two kinds of NUMA support code can be co-exist in one Xen binary. Xen can check feature flags to decide using device tree or ACPI as NUMA based firmware. So in this patch, we introduce a generic option: CONFIG_ARM_NUMA for users to enable NUMA for Arm. And one CONFIG_DEVICE_TREE_NUMA option for ARM_NUMA to select when HAS_DEVICE_TREE option is enabled. Once when ACPI NUMA for Arm is supported, ACPI_NUMA can be selected here too. Signed-off-by: Wei Chen Signed-off-by: Henry Wang --- v3 -> v4: 1. No change. v2 -> v3: 1. No change. v1 -> v2: 1. Remove the condition of selecting DEVICE_TREE_NUMA. --- xen/arch/arm/Kconfig | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/xen/arch/arm/Kconfig b/xen/arch/arm/Kconfig index 239d3aed3c..e751ad50d1 100644 --- a/xen/arch/arm/Kconfig +++ b/xen/arch/arm/Kconfig @@ -39,6 +39,17 @@ config ACPI config ARM_EFI bool +config ARM_NUMA + bool "Arm NUMA (Non-Uniform Memory Access) Support (UNSUPPORTED)" if UNSUPPORTED + depends on HAS_DEVICE_TREE + select DEVICE_TREE_NUMA + help + Enable Non-Uniform Memory Access (NUMA) for Arm architecutres + +config DEVICE_TREE_NUMA + bool + select NUMA + config GICV3 bool "GICv3 driver" depends on !NEW_VGIC From patchwork Tue Apr 25 07:56:55 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Henry Wang X-Patchwork-Id: 13222987 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 116D2C77B61 for ; Tue, 25 Apr 2023 07:59:14 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.525785.817243 (Exim 4.92) (envelope-from ) id 1prDZo-0001sa-Cv; Tue, 25 Apr 2023 07:59:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 525785.817243; Tue, 25 Apr 2023 07:59:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDZo-0001rp-7L; Tue, 25 Apr 2023 07:59:04 +0000 Received: by outflank-mailman (input) for mailman id 525785; Tue, 25 Apr 2023 07:59:02 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1prDYw-0005uB-IE for xen-devel@lists.xenproject.org; Tue, 25 Apr 2023 07:58:10 +0000 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by se1-gles-flk1.inumbo.com (Halon) with ESMTP id eae4ecbe-e33e-11ed-8611-37d641c3527e; Tue, 25 Apr 2023 09:58:08 +0200 (CEST) Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 50CE84B3; Tue, 25 Apr 2023 00:58:52 -0700 (PDT) Received: from a015966.shanghai.arm.com (a015966.shanghai.arm.com [10.169.190.5]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id 7CF703F587; Tue, 25 Apr 2023 00:58:05 -0700 (PDT) X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: eae4ecbe-e33e-11ed-8611-37d641c3527e From: Henry Wang To: xen-devel@lists.xenproject.org Cc: Wei Chen , Andrew Cooper , George Dunlap , Jan Beulich , Julien Grall , Stefano Stabellini , Wei Liu , Henry Wang Subject: [PATCH v4 17/17] docs: update numa command line to support Arm Date: Tue, 25 Apr 2023 15:56:55 +0800 Message-Id: <20230425075655.4037980-18-Henry.Wang@arm.com> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20230425075655.4037980-1-Henry.Wang@arm.com> References: <20230425075655.4037980-1-Henry.Wang@arm.com> MIME-Version: 1.0 From: Wei Chen Current numa command in documentation is x86 only. Remove x86 from numa command's arch limitation in this patch. Signed-off-by: Wei Chen Signed-off-by: Henry Wang Acked-by: Jan Beulich --- v3 -> v4: 1. No change. v2 -> v3: 1. Add the Acked-by tag from Jan. v1 -> v2: 1. Update Arm NUMA status in SUPPORT.md to "Tech Preview". --- SUPPORT.md | 1 + docs/misc/xen-command-line.pandoc | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/SUPPORT.md b/SUPPORT.md index aa1940e55f..da4e3b9aa2 100644 --- a/SUPPORT.md +++ b/SUPPORT.md @@ -401,6 +401,7 @@ on embedded platforms and the x86 PV shim. Enables NUMA aware scheduling in Xen Status, x86: Supported + Status, Arm: Tech Preview ## Scalability diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index e0b89b7d33..2fea22dd70 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1890,7 +1890,7 @@ i.e. a limit on the number of guests it is possible to start each having assigned a device sharing a common interrupt line. Accepts values between 1 and 255. -### numa (x86) +### numa > `= on | off | fake= | noacpi` > Default: `on`