From nobody Sun Nov 24 05:33:37 2024 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=gmail.com ARC-Seal: i=1; a=rsa-sha256; t=1726003587; cv=none; d=zohomail.com; s=zohoarc; b=gB24I17shjosRvWR2RlTDBSSs8PULONupk8tYayBgLXKSKgj5RLa9EFho2XKegb0fQ251VX1im+UV9wPfXckpE/eWRM/+F+RaraZl3i1N13UwiVkcb/JDmH+mKB8vcpkzAnpY0UoBMNPv6jwgf07QiaN7mGSwteZ7996HJ0b71Y= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1726003587; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=g/kA3YBpDLtsSb+1NwaLXjIWdJ2NRVqOLBBdSqFPbiveoSP43WxI65weZHJUaXdjq20ixBbMmUtp06t/M7TcOd7QsAM3H8l0XWxa24SIDd3n8Pp/UWVrxr+QjfYVfQYk6pblndkAUxrcaOomtE9biklwwkd7g81IrSaEqw9Uptw= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1726003587063734.7993528148451; Tue, 10 Sep 2024 14:26:27 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1so8Mr-0002Cf-IF; Tue, 10 Sep 2024 17:25:45 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1so8Mq-000268-0w; Tue, 10 Sep 2024 17:25:44 -0400 Received: from mail-yb1-xb2d.google.com ([2607:f8b0:4864:20::b2d]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1so8Mm-0002Ii-Vi; Tue, 10 Sep 2024 17:25:43 -0400 Received: by mail-yb1-xb2d.google.com with SMTP id 3f1490d57ef6-e1ce8a675f7so6167853276.3; Tue, 10 Sep 2024 14:25:40 -0700 (PDT) Received: from localhost ([2600:1700:830:3db0::14]) by smtp.gmail.com with UTF8SMTPSA id 3f1490d57ef6-e1d7ba0e668sm461104276.25.2024.09.10.14.25.38 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 10 Sep 2024 14:25:38 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1726003539; x=1726608339; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=SRPMElbSm9pf2vFooojIGQcmWBNBOpjvQQ3G0AHQuAX7At0k0u3MPEE4UJbw4ZpRVo CatotEsERVp6YxsdJ+S1hsDJifM4YE5mk6n5Y9CVnb2+m6WecNCnqtGFleKoBFUXK8Em qZVEokWdPuligm5u2KE4c+fNw1zD60fsOpXOwp5eF1OlUu99WaPe2CCmslpQu6tcHxuC 57ZY1hQsTt/2CmBj+7iUnBo4Mqj35vedJ3lyQ6hnqvtMxtwi8Yw8fNyldvHBXJ71BK+B CvH+i9h4JVLwb7sGpI5OWjtv3j1HI9V6nBcHo1vKcI78zAqGcrd0wN7VGZ+mgRO4T1Rk 7Tzw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1726003539; x=1726608339; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=h/xeL4BQwQO+XM6aFyhRD8QMfSr7p5N8VtxZa7Z1sPMdFEphxlFtlozy7TVjFi9Hfw nBPoLr5PCUEYSQvnYNx2mFzhxu2+J1mW3eaRfx8mHXRW241bRr9SWM/4tAsQiCXVV4I4 q0+SapQ+hStxVypWfDsNK4kuH8Wtd+48Z4k8s6nSTGGOiXZfV3COYcs0TN9SiuzkKbWg V/TPsDIWrRF1xJkbgnJNlXEyV8TMgJHpQ0x9o/4pDT8gT1WqDAw4bSnD5xEeF+W0Vg8Y T3NtpIoY+3Py6V15qH5PEr6zJpCxgF7bhegThiFYnIUortPJJQDDSb2K/9/OxSD3Q9EN BKXw== X-Gm-Message-State: AOJu0YxockQS1Ntio3A32jz6pWr/oWTW/k3yaqKRwbMpiQHSX/Jd2jXv mgsMnaFsh1CuK8BD/aRx/JuTQPReoiHM9K3/+JhUpzHqFl2sTfjVO+2DSp/H X-Google-Smtp-Source: AGHT+IG67i9Gy2Szse48B4leoupmQGLemvX+YA3ndQ0d7RhjJ7129f4+cmj8q9hLxOCadheqw6P9WA== X-Received: by 2002:a05:6902:905:b0:e12:97d9:4204 with SMTP id 3f1490d57ef6-e1d349f1235mr15244269276.41.1726003538972; Tue, 10 Sep 2024 14:25:38 -0700 (PDT) From: Gregor Haas To: qemu-devel@nongnu.org Cc: qemu-riscv@nongnu.org, atishp@rivosinc.com, dbarboza@ventanamicro.com, alistair.francis@wdc.com, Gregor Haas Subject: [PATCH v5 1/2] hw/riscv: Add support for generating OpenSBI domains in the device tree Date: Tue, 10 Sep 2024 14:25:31 -0700 Message-ID: <20240910212533.986734-3-gregorhaas1997@gmail.com> X-Mailer: git-send-email 2.46.0 In-Reply-To: <20240910212533.986734-1-gregorhaas1997@gmail.com> References: <20240910212533.986734-1-gregorhaas1997@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2607:f8b0:4864:20::b2d; envelope-from=gregorhaas1997@gmail.com; helo=mail-yb1-xb2d.google.com X-Spam_score_int: -17 X-Spam_score: -1.8 X-Spam_bar: - X-Spam_report: (-1.8 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @gmail.com) X-ZM-MESSAGEID: 1726003588712116600 Content-Type: text/plain; charset="utf-8" OpenSBI has support for domains, which are partitions of CPUs and memory in= to isolated compartments. Domains can be specified in the device tree accordin= g to a standardized format [1], which OpenSBI parses at boot time to initialize = all system domains. This patch enables simply specifying domains (and their associated memory regions) on the QEMU command line, from which these are t= hen rendered into the machine's device tree. At machine initialization time, a new create_fdt_opensbi_domains() function walks the peripherals/peripherals-anon containers, identifies all domains a= nd memregions, and parses them into the relevant device tree structures. [1] https://github.com/riscv-software-src/opensbi/blob/master/docs/domain_s= upport.md Signed-off-by: Gregor Haas Reviewed-by: Daniel Henrique Barboza --- MAINTAINERS | 7 + hw/riscv/Kconfig | 4 + hw/riscv/meson.build | 1 + hw/riscv/opensbi_domain.c | 558 ++++++++++++++++++++++++++++++ hw/riscv/virt.c | 3 + include/hw/riscv/opensbi_domain.h | 69 ++++ 6 files changed, 642 insertions(+) create mode 100644 hw/riscv/opensbi_domain.c create mode 100644 include/hw/riscv/opensbi_domain.h diff --git a/MAINTAINERS b/MAINTAINERS index 0c1bc69828..3b9f5b7432 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -357,6 +357,13 @@ F: target/riscv/XVentanaCondOps.decode F: target/riscv/insn_trans/trans_xventanacondops.c.inc F: disas/riscv-xventana* =20 +RISC-V OpenSBI domain support +M: Gregor Haas +L: qemu-riscv@nongnu.org +S: Maintained +F: hw/riscv/opensbi_domain.c +F: include/hw/riscv/opensbi_domain.h + RENESAS RX CPUs R: Yoshinori Sato S: Orphan diff --git a/hw/riscv/Kconfig b/hw/riscv/Kconfig index a2030e3a6f..db3a4d77ad 100644 --- a/hw/riscv/Kconfig +++ b/hw/riscv/Kconfig @@ -1,6 +1,9 @@ config RISCV_NUMA bool =20 +config RISCV_OPENSBI_DOMAIN + bool + config IBEX bool =20 @@ -40,6 +43,7 @@ config RISCV_VIRT imply TPM_TIS_SYSBUS select DEVICE_TREE select RISCV_NUMA + select RISCV_OPENSBI_DOMAIN select GOLDFISH_RTC select PCI select PCI_EXPRESS_GENERIC_BRIDGE diff --git a/hw/riscv/meson.build b/hw/riscv/meson.build index f872674093..f47626c164 100644 --- a/hw/riscv/meson.build +++ b/hw/riscv/meson.build @@ -1,6 +1,7 @@ riscv_ss =3D ss.source_set() riscv_ss.add(files('boot.c')) riscv_ss.add(when: 'CONFIG_RISCV_NUMA', if_true: files('numa.c')) +riscv_ss.add(when: 'CONFIG_RISCV_OPENSBI_DOMAIN', if_true: files('opensbi_= domain.c')) riscv_ss.add(files('riscv_hart.c')) riscv_ss.add(when: 'CONFIG_OPENTITAN', if_true: files('opentitan.c')) riscv_ss.add(when: 'CONFIG_RISCV_VIRT', if_true: files('virt.c')) diff --git a/hw/riscv/opensbi_domain.c b/hw/riscv/opensbi_domain.c new file mode 100644 index 0000000000..da315e59b2 --- /dev/null +++ b/hw/riscv/opensbi_domain.c @@ -0,0 +1,558 @@ +/* + * OpenSBI Domains + * + * Copyright (c) 2024 Gregor Haas + * + * Generates OpenSBI domain nodes in the machine's device tree + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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 f= or + * more details. + * + * You should have received a copy of the GNU General Public License along= with + * this program. If not, see . + */ + +#include "qemu/osdep.h" +#include "hw/riscv/opensbi_domain.h" +#include "hw/boards.h" +#include "hw/riscv/virt.h" +#include "qapi/error.h" +#include "qemu/cutils.h" +#include "qemu/error-report.h" +#include "sysemu/device_tree.h" + +#include + +static void create_fdt_domain_possible_harts(MachineState *ms, + OpenSBIDomainState *s, + char *path) { + unsigned long i, cpu; + unsigned long num_cpus; + + num_cpus =3D s->last_possible_hart - s->first_possible_hart + 1; + if (num_cpus) { + g_autofree uint32_t *phandles =3D g_malloc0_n(num_cpus, sizeof(uin= t32_t)); + + for (i =3D 0, cpu =3D s->first_possible_hart; i < num_cpus; i++, c= pu++) { + g_autofree char *cpu_name =3D g_strdup_printf("/cpus/cpu@%li",= cpu); + phandles[i] =3D cpu_to_fdt32(qemu_fdt_get_phandle( + ms->fdt, cpu_name)); + } + + qemu_fdt_setprop(ms->fdt, path, "possible-harts", + phandles, num_cpus * 4); + } +} + +static void create_fdt_domain_regions(MachineState *ms, + OpenSBIDomainState *s, + char *path) { + unsigned long i; + int num_regions =3D 0; + DeviceState *ds; + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (s->regions[i]) { + num_regions++; + } + } + + if (num_regions) { + g_autofree uint32_t *regions =3D + g_malloc0_n(num_regions, 2 * sizeof(uint32_t)); + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (s->regions[i]) { + ds =3D DEVICE(s->regions[i]); + g_autofree char *region_name =3D g_strdup_printf( + "/chosen/opensbi-domains/%s", ds->id); + regions[2 * i] =3D cpu_to_fdt32(qemu_fdt_get_phandle + (ms->fdt, region_name)); + regions[2 * i + 1] =3D cpu_to_fdt32(s->region_perms[i]); + } + } + + qemu_fdt_setprop(ms->fdt, path, "regions", + regions, num_regions * 8); + } +} + +struct DomainFDTState { + MachineState *ms; + bool regions; +}; + +static void create_fdt_one_domain(MachineState *ms, OpenSBIDomainState *s) +{ + DeviceState *ds =3D DEVICE(s); + g_autofree char *path =3D g_strdup_printf("/chosen/opensbi-domains/%s"= , ds->id); + + qemu_fdt_add_subnode(ms->fdt, path); + qemu_fdt_setprop_string(ms->fdt, path, "compatible", + "opensbi,domain,instance"); + qemu_fdt_setprop_cells(ms->fdt, path, "phandle", + qemu_fdt_alloc_phandle(ms->fdt)); + + create_fdt_domain_possible_harts(ms, s, path); + create_fdt_domain_regions(ms, s, path); + + if (s->boot_hart !=3D -1) { + g_autofree char *cpu_name =3D g_strdup_printf("/cpus/cpu@%i", s->b= oot_hart); + qemu_fdt_setprop_cell(ms->fdt, path, "boot-hart", + qemu_fdt_get_phandle(ms->fdt, cpu_name)); + if (s->assign) { + qemu_fdt_setprop_cell(ms->fdt, cpu_name, "opensbi-domain", + qemu_fdt_get_phandle(ms->fdt, path)); + } + } + + if (s->next_arg1 !=3D -1) { + qemu_fdt_setprop_cells(ms->fdt, path, "next-arg1", + (uint64_t) s->next_arg1 >> 32, s->next_arg1); + } + + if (s->next_addr !=3D -1) { + qemu_fdt_setprop_cells(ms->fdt, path, "next-addr", + (uint64_t) s->next_addr >> 32, s->next_addr); + } + + if (s->next_mode !=3D -1) { + qemu_fdt_setprop_cell(ms->fdt, path, "next-mode", + s->next_mode); + } + + if (s->system_reset_allowed) { + qemu_fdt_setprop(ms->fdt, path, "system-reset-allowed", NULL, 0); + } + + if (s->system_suspend_allowed) { + qemu_fdt_setprop(ms->fdt, path, "system-suspend-allowed", NULL, 0); + } +} + +static uint32_t create_fdt_one_device(MachineState *ms, char *device) +{ + uint32_t phandle; + int offs =3D fdt_path_offset(ms->fdt, device); + + if (offs < 0) { + error_report("%s: Could not find device %s: %s", __func__, + device, fdt_strerror(offs)); + exit(1); + } + + phandle =3D fdt_get_phandle(ms->fdt, offs); + if (!phandle) { + phandle =3D qemu_fdt_alloc_phandle(ms->fdt); + qemu_fdt_setprop_cell(ms->fdt, device, "phandle", phandle); + } + + return phandle; +} + +static void create_fdt_one_memregion(MachineState *ms, + OpenSBIMemregionState *s) +{ + g_autofree char *path; + int i, dev, num_devices; + DeviceState *ds =3D DEVICE(s); + + path =3D g_strdup_printf("/chosen/opensbi-domains/%s", ds->id); + qemu_fdt_add_subnode(ms->fdt, path); + qemu_fdt_setprop_string(ms->fdt, path, "compatible", + "opensbi,domain,memregion"); + qemu_fdt_setprop_cells(ms->fdt, path, "base", + (uint64_t) s->base >> 32, s->base); + + qemu_fdt_setprop_cell(ms->fdt, path, "order", + (uint32_t) s->order); + + if (s->mmio) { + qemu_fdt_setprop(ms->fdt, path, "mmio", NULL, 0); + + /* Get all phandles for related devices */ + num_devices =3D 0; + for (i =3D 0; i < OPENSBI_MEMREGION_DEVICES_MAX; i++) { + if (s->devices[i]) { + num_devices++; + } + } + + if (num_devices) { + g_autofree uint32_t *devices =3D + g_malloc0_n(num_devices, sizeof(uint32_t)); + for (i =3D 0, dev =3D 0; i < OPENSBI_MEMREGION_DEVICES_MAX && + dev < num_devices; i++) { + if (s->devices[i]) { + devices[dev++] =3D create_fdt_one_device(ms, + s->devices[i]); + } + } + + qemu_fdt_setprop(ms->fdt, path, "devices", devices, + num_devices * 4); + } + } + + qemu_fdt_setprop_cells(ms->fdt, path, "phandle", + qemu_fdt_alloc_phandle(ms->fdt)); +} + +static int create_fdt_domains(Object *obj, void *opaque) +{ + struct DomainFDTState *dfs =3D opaque; + OpenSBIDomainState *osds; + OpenSBIMemregionState *osms; + + osds =3D (OpenSBIDomainState *) + object_dynamic_cast(obj, TYPE_OPENSBI_DOMAIN); + osms =3D (OpenSBIMemregionState *) + object_dynamic_cast(obj, TYPE_OPENSBI_MEMREGION); + + if (dfs->regions) { + if (osms) { + create_fdt_one_memregion(dfs->ms, osms); + } + } else { + if (osds) { + create_fdt_one_domain(dfs->ms, osds); + } + } + + return 0; +} + +static const char *containers[] =3D { + "/peripheral", "/peripheral-anon" +}; + +void create_fdt_opensbi_domains(MachineState *s) +{ + int i; + MachineState *ms =3D MACHINE(s); + Object *container; + + struct DomainFDTState check =3D { + .ms =3D ms, + .regions =3D true + }; + + /* Make sure that top-level node exists */ + qemu_fdt_add_subnode(ms->fdt, "/chosen/opensbi-domains"); + qemu_fdt_setprop_string(ms->fdt, "/chosen/opensbi-domains", + "compatible", "opensbi,domain,config"); + + /* Do a scan through regions first */ + for (i =3D 0; i < ARRAY_SIZE(containers); i++) { + container =3D container_get(OBJECT(s), containers[i]); + object_child_foreach(container, create_fdt_domains, &check); + } + + /* Then scan through domains */ + check.regions =3D false; + for (i =3D 0; i < ARRAY_SIZE(containers); i++) { + container =3D container_get(OBJECT(s), containers[i]); + object_child_foreach(container, create_fdt_domains, &check); + } +} + +/* OpenSBI Memregions */ + +static void set_mmio(Object *obj, bool val, Error **err) +{ + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + s->mmio =3D val; +} + +static void set_device(Object *obj, const char *val, Error **err) +{ + int i; + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (!s->devices[i]) { + s->devices[i] =3D g_strdup(val); + break; + } + } +} + +static void opensbi_memregion_instance_init(Object *obj) +{ + int i; + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + + s->base =3D -1; + object_property_add_uint64_ptr(obj, "base", &s->base, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "base", + "The base address of the domain memory= region. If \"order\" is also specified, " + "this property should be a 2 ^ order a= ligned 64 bit address"); + + s->order =3D -1; + object_property_add_uint32_ptr(obj, "order", &s->order, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "order", + "The order of the domain memory region= . This property should have a 32 bit value " + "(i.e. one DT cell) in the range 3 <= =3D order <=3D __riscv_xlen."); + + s->mmio =3D false; + object_property_add_bool(obj, "mmio", NULL, set_mmio); + object_property_set_description(obj, "mmio", + "A boolean flag representing whether t= he domain memory region is a " + "memory-mapped I/O (MMIO) region."); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + g_autofree char *propname =3D g_strdup_printf("device%i", i); + object_property_add_str(obj, propname, NULL, set_device); + + g_autofree char *description =3D g_strdup_printf( + "Device %i (out of %i) for this memregion. This property s= hould be a device tree path to the device.", + i, OPENSBI_DOMAIN_MEMREGIONS_MAX); + object_property_set_description(obj, propname, description); + } +} + +static void opensbi_memregion_realize(DeviceState *ds, Error **errp) +{ + #if defined(TARGET_RISCV32) + int xlen =3D 32; + #elif defined(TARGET_RISCV64) + int xlen =3D 64; + #endif + + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(ds); + + if (s->base =3D=3D -1) { + error_setg(errp, "must specify base"); + return; + } + + if (s->order =3D=3D -1) { + error_setg(errp, "must specify order"); + return; + } + + /* Check order bounds */ + if (s->order < 3 || s->order > xlen) { + error_setg(errp, "order must be between 3 and %d", xlen); + return; + } + + /* Check base alignment */ + if (s->order < xlen && (s->base & (BIT(s->order) - 1))) { + error_setg(errp, "base not aligned to order"); + return; + } +} + +static void opensbi_memregion_class_init(ObjectClass *oc, void *opaque) +{ + DeviceClass *dc =3D DEVICE_CLASS(oc); + dc->realize =3D opensbi_memregion_realize; +} + +static const TypeInfo opensbi_memregion_info =3D { + .name =3D TYPE_OPENSBI_MEMREGION, + .parent =3D TYPE_DEVICE, + .instance_init =3D opensbi_memregion_instance_init, + .instance_size =3D sizeof(OpenSBIDomainState), + .class_init =3D opensbi_memregion_class_init +}; + +/* OpenSBI Domains */ + +static void set_sysreset_allowed(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->system_reset_allowed =3D val; +} + +static void set_suspend_allowed(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->system_suspend_allowed =3D val; +} + +static void set_assign(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->assign =3D val; +} + +static void set_possible_harts(Object *obj, const char *str, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + const char *firstcpu, *firstcpu_end, *lastcpu; + + firstcpu =3D str; + if (qemu_strtoul(firstcpu, &firstcpu_end, 0, + &s->first_possible_hart) < 0) { + error_setg(err, "could not convert firstcpu"); + return; + } + + lastcpu =3D qemu_strchrnul(str, '-'); + if (*lastcpu) { + if (lastcpu !=3D firstcpu_end) { + error_setg(err, "could not separate firstcpu and lastcpu"); + return; + } + + lastcpu++; + if (qemu_strtoul(lastcpu, NULL, 0, + &s->last_possible_hart) < 0) { + error_setg(err, "could not convert lastcpu"); + return; + } + } else { + s->last_possible_hart =3D s->first_possible_hart; + } +} + +static void opensbi_domain_instance_init(Object *obj) +{ + int i; + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + + s->boot_hart =3D VIRT_CPUS_MAX; + object_property_add_uint32_ptr(obj, "boot-hart", &s->boot_hart, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "boot-hart", + "The HART booting the domain instance.= "); + + s->first_possible_hart =3D -1; + s->last_possible_hart =3D -1; + object_property_add_str(obj, "possible-harts", NULL, set_possible_hart= s); + object_property_set_description(obj, "possible-harts", + "The contiguous list of CPUs for the d= omain instance, specified as firstcpu[-lastcpu]"); + + s->next_arg1 =3D -1; + object_property_add_uint64_ptr(obj, "next-arg1", &s->next_arg1, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-arg1", + "The 64 bit next booting stage arg1 fo= r the domain instance."); + + s->next_addr =3D -1; + object_property_add_uint64_ptr(obj, "next-addr", &s->next_addr, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-addr", + "The 64 bit next booting stage address= for the domain instance."); + + s->next_mode =3D -1; + object_property_add_uint32_ptr(obj, "next-mode", &s->next_mode, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-mode", + "The 32 bit next booting stage mode fo= r the domain instance."); + + s->system_reset_allowed =3D false; + object_property_add_bool(obj, "system-reset-allowed", NULL, + set_sysreset_allowed); + object_property_set_description(obj, "system-reset-allowed", + "Whether the domain instance is allowe= d to do system reset."); + + s->system_suspend_allowed =3D false; + object_property_add_bool(obj, "system-suspend-allowed", NULL, + set_suspend_allowed); + object_property_set_description(obj, "system-suspend-allowed", + "Whether the domain instance is allowe= d to do system suspend."); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + s->regions[i] =3D NULL; + g_autofree char *reg_propname =3D g_strdup_printf("region%i", i); + object_property_add_link(obj, reg_propname, TYPE_OPENSBI_MEMREGION, + (Object **) &s->regions[i], + qdev_prop_allow_set_link_before_realize, = 0); + + g_autofree char *reg_description =3D g_strdup_printf( + "Region %i (out of %i) for this domain.", + i, OPENSBI_DOMAIN_MEMREGIONS_MAX); + object_property_set_description(obj, reg_propname, reg_description= ); + + s->region_perms[i] =3D 0; + g_autofree char *perm_propname =3D g_strdup_printf("perms%i", i); + object_property_add_uint32_ptr(obj, perm_propname, &s->region_perm= s[i], + OBJ_PROP_FLAG_WRITE); + + g_autofree char *perm_description =3D g_strdup_printf( + "Permissions for region %i for this domain.", i); + object_property_set_description(obj, perm_propname, perm_descripti= on); + } + + object_property_add_bool(obj, "assign", NULL, set_assign); + object_property_set_description(obj, "assign", + "Whether to assign this domain to its = boot hart."); +} + +static void opensbi_domain_realize(DeviceState *ds, Error **errp) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(ds); + + if (!ds->id) { + error_setg(errp, "must specify an id"); + return; + } + + if (s->boot_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, "boot hart larger than maximum number of CPUs (%d= )", + VIRT_CPUS_MAX); + return; + } + + if (s->first_possible_hart =3D=3D -1) { + if (s->last_possible_hart !=3D -1) { + error_setg(errp, + "last possible hart set when first possible hart unse= t"); + return; + } + } else { + if (s->first_possible_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, + "first possible hart larger than maximum number of CP= Us (%d)", + VIRT_CPUS_MAX); + return; + } + + if (s->last_possible_hart !=3D -1) { + if (s->last_possible_hart < s->first_possible_hart) { + error_setg(errp, + "last possible hart larger than first possible ha= rt"); + return; + } + + if (s->last_possible_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, + "last possible hart larger than maximum number of= CPUS (%d)", + VIRT_CPUS_MAX); + return; + } + } + } +} + +static void opensbi_domain_class_init(ObjectClass *oc, void *opaque) +{ + DeviceClass *dc =3D DEVICE_CLASS(oc); + dc->realize =3D opensbi_domain_realize; +} + +static const TypeInfo opensbi_domain_info =3D { + .name =3D TYPE_OPENSBI_DOMAIN, + .parent =3D TYPE_DEVICE, + .instance_init =3D opensbi_domain_instance_init, + .instance_size =3D sizeof(OpenSBIDomainState), + .class_init =3D opensbi_domain_class_init +}; + +static void opensbi_register_types(void) +{ + type_register_static(&opensbi_domain_info); + type_register_static(&opensbi_memregion_info); +} + +type_init(opensbi_register_types) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index cef41c150a..760d8ce4d4 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -55,6 +55,7 @@ #include "hw/acpi/aml-build.h" #include "qapi/qapi-visit-common.h" #include "hw/virtio/virtio-iommu.h" +#include "hw/riscv/opensbi_domain.h" =20 /* KVM AIA only supports APLIC MSI. APLIC Wired is always emulated by QEMU= . */ static bool virt_use_kvm_aia(RISCVVirtState *s) @@ -1050,6 +1051,8 @@ static void finalize_fdt(RISCVVirtState *s) create_fdt_uart(s, virt_memmap, irq_mmio_phandle); =20 create_fdt_rtc(s, virt_memmap, irq_mmio_phandle); + + create_fdt_opensbi_domains(MACHINE(s)); } =20 static void create_fdt(RISCVVirtState *s, const MemMapEntry *memmap) diff --git a/include/hw/riscv/opensbi_domain.h b/include/hw/riscv/opensbi_d= omain.h new file mode 100644 index 0000000000..1ec8ed8288 --- /dev/null +++ b/include/hw/riscv/opensbi_domain.h @@ -0,0 +1,69 @@ +/* + * OpenSBI Domains + * + * Copyright (c) 2024 Gregor Haas + * + * Generates OpenSBI domain nodes in the machine's device tree + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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 f= or + * more details. + * + * You should have received a copy of the GNU General Public License along= with + * this program. If not, see . + */ + +#ifndef RISCV_DOMAIN_H +#define RISCV_DOMAIN_H + +#include "hw/sysbus.h" +#include "qom/object.h" +#include "cpu.h" + +#define TYPE_OPENSBI_MEMREGION "opensbi-memregion" +OBJECT_DECLARE_SIMPLE_TYPE(OpenSBIMemregionState, OPENSBI_MEMREGION) + +#define OPENSBI_MEMREGION_DEVICES_MAX 16 + +struct OpenSBIMemregionState { + /* public */ + DeviceState parent_obj; + + /* private */ + uint64_t base; + uint32_t order; + bool mmio; + char *devices[OPENSBI_MEMREGION_DEVICES_MAX]; +}; + +#define TYPE_OPENSBI_DOMAIN "opensbi-domain" +OBJECT_DECLARE_SIMPLE_TYPE(OpenSBIDomainState, OPENSBI_DOMAIN) + +#define OPENSBI_DOMAIN_MEMREGIONS_MAX 16 + +struct OpenSBIDomainState { + /* public */ + DeviceState parent_obj; + + /* private */ + OpenSBIMemregionState *regions[OPENSBI_DOMAIN_MEMREGIONS_MAX]; + unsigned int region_perms[OPENSBI_DOMAIN_MEMREGIONS_MAX]; + unsigned long first_possible_hart, last_possible_hart; + unsigned int boot_hart; + uint64_t next_arg1; + uint64_t next_addr; + uint32_t next_mode; + bool system_reset_allowed; + bool system_suspend_allowed; + + bool assign; +}; + +void create_fdt_opensbi_domains(MachineState *s); + +#endif /* RISCV_DOMAIN_H */ --=20 2.46.0 From nobody Sun Nov 24 05:33:37 2024 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=gmail.com ARC-Seal: i=1; a=rsa-sha256; t=1726003582; cv=none; d=zohomail.com; s=zohoarc; b=NXg9a8Tp6DFycoFUiypYkmSIkvzSqTzVl32Ai9Kq8wbAPCbLFfcI0X2u3BcGGxnFaIR5P6zxWtKPvxNaRfNx71SxUhwYe/ppLOpjqqvZ/lKYhD4im7kZqeMVZhHu2LhPTrGtMdSh26SgzX5pZ4hNWrqRu5MVw9wQ2+cLGzYgskw= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1726003582; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=Jrmu4NCoowqJaRgF++K+0kRHL0DZQBViuImzs7rf3E9Wy+LSNvt26q46WVX+YWzJ1jC+SL7M4iX58GQ0+Bk05tOheaPnLG/KMcOFH07QhX5lJbSypdE1hWy0amEQ14uatI7fXwcgG2zr+4J1ixG/NbJJWttThVIyD70AblJh5wM= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 172600358276531.10262063157029; Tue, 10 Sep 2024 14:26:22 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1so8Mq-00025x-1H; Tue, 10 Sep 2024 17:25:44 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1so8Mo-0001zY-8j; Tue, 10 Sep 2024 17:25:42 -0400 Received: from mail-yw1-x112b.google.com ([2607:f8b0:4864:20::112b]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1so8Ml-0002IT-2j; Tue, 10 Sep 2024 17:25:41 -0400 Received: by mail-yw1-x112b.google.com with SMTP id 00721157ae682-6cdae28014dso1947847b3.1; Tue, 10 Sep 2024 14:25:38 -0700 (PDT) Received: from localhost ([2600:1700:830:3db0::14]) by smtp.gmail.com with UTF8SMTPSA id 00721157ae682-6db9651d74fsm4410317b3.104.2024.09.10.14.25.36 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 10 Sep 2024 14:25:36 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1726003537; x=1726608337; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=YzAMzT/UJQ+snZKjTX/fc/B1ugRUZb4e1XapJOPKZQY6zMAtc4ttB4jQGlyI1TdE4q 9BnKmxsUbf+Ss2AUXI4jX0OBAarejW80yqUYz4lWnvjDCeISILZTDqxookVepsDgyaBo ghoWHuiQ+XumUrh3BwrHg3JI1HBvMMwRgTCb5CTWsPSlpb/awEOJHJOoYn1zs6QXQ10/ m9gZhxhkox53XqsML75KhzshnIG21H6JqJmJTmLoYFDwmXDlRoRRFOO9CK7RcyrpyUmt Vnw00KhoRkLwsxgrngexWzMxLxItEUH6Kly5Dn8iMKMir91Ot6IrxlsC1/IWQZsQ4EPB 1j4g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1726003537; x=1726608337; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=RtLCCYi/8lZclH6EGbiQfc3dGIl0M7+pchzl78qCIMM=; b=HtOS36ibv/quhBTW/sgvOaOr4jYAP+4dGFyZDScFZq3LSlGLLOwpF4J1kwXZ2wRqRS PRdWWZUgyitKisI4V3nuTJpCg1PEdzynoErrraKv5ByF+HDpYdi67MYhmzTTv6H0aqfB npzWRLrrV8ntBrCWoWG5hWDP9dK0bzJn2wdh0h0bmnFPZEfeAvjNFA/CByXycZ+H9KCD WvWDhGFLiUpg9GnndB5523EKLUpIhOQ/y6un12QQoiMI6lb3S6S2mXvPG0kRvDpc1AYh o3vvyFWekPMQbctyGA+ZYZWwqxuYMn08534fl3xzuNdWVm7JwPpz0WQ9nhcYAud9peS8 SnGw== X-Gm-Message-State: AOJu0Yz4c2YvtOFjpqCS8em1pWD/zlVt0ujaSai15PURMkVy+mTNtgQA aID+tydHCh5m5vYm5hESQWqx3FkWQsc2Bd1N2PkFdgemtt178+i+ar2ieNKQ X-Google-Smtp-Source: AGHT+IFwEg5lLB7Aa6pEZccj5P36ZV3I8VnoGAtDCAikpYRNcPXc65PepuX1pTYNwPrDrIyVlUDooA== X-Received: by 2002:a05:690c:83:b0:6d5:7617:950e with SMTP id 00721157ae682-6db953e8b87mr33961937b3.18.1726003537427; Tue, 10 Sep 2024 14:25:37 -0700 (PDT) From: Gregor Haas To: qemu-devel@nongnu.org Cc: qemu-riscv@nongnu.org, atishp@rivosinc.com, dbarboza@ventanamicro.com, alistair.francis@wdc.com, Gregor Haas Subject: [PATCH v5 1/2] Add support for generating OpenSBI domains in the device tree Date: Tue, 10 Sep 2024 14:25:30 -0700 Message-ID: <20240910212533.986734-2-gregorhaas1997@gmail.com> X-Mailer: git-send-email 2.46.0 In-Reply-To: <20240910212533.986734-1-gregorhaas1997@gmail.com> References: <20240910212533.986734-1-gregorhaas1997@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2607:f8b0:4864:20::112b; envelope-from=gregorhaas1997@gmail.com; helo=mail-yw1-x112b.google.com X-Spam_score_int: -17 X-Spam_score: -1.8 X-Spam_bar: - X-Spam_report: (-1.8 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @gmail.com) X-ZM-MESSAGEID: 1726003584612116600 Content-Type: text/plain; charset="utf-8" OpenSBI has support for domains, which are partitions of CPUs and memory in= to isolated compartments. Domains can be specified in the device tree accordin= g to a standardized format [1], which OpenSBI parses at boot time to initialize = all system domains. This patch enables simply specifying domains (and their associated memory regions) on the QEMU command line, from which these are t= hen rendered into the machine's device tree. At machine initialization time, a new create_fdt_opensbi_domains() function walks the peripherals/peripherals-anon containers, identifies all domains a= nd memregions, and parses them into the relevant device tree structures. [1] https://github.com/riscv-software-src/opensbi/blob/master/docs/domain_s= upport.md Signed-off-by: Gregor Haas Reviewed-by: Daniel Henrique Barboza --- MAINTAINERS | 7 + hw/riscv/Kconfig | 4 + hw/riscv/meson.build | 1 + hw/riscv/opensbi_domain.c | 558 ++++++++++++++++++++++++++++++ hw/riscv/virt.c | 3 + include/hw/riscv/opensbi_domain.h | 69 ++++ 6 files changed, 642 insertions(+) create mode 100644 hw/riscv/opensbi_domain.c create mode 100644 include/hw/riscv/opensbi_domain.h diff --git a/MAINTAINERS b/MAINTAINERS index 0c1bc69828..3b9f5b7432 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -357,6 +357,13 @@ F: target/riscv/XVentanaCondOps.decode F: target/riscv/insn_trans/trans_xventanacondops.c.inc F: disas/riscv-xventana* =20 +RISC-V OpenSBI domain support +M: Gregor Haas +L: qemu-riscv@nongnu.org +S: Maintained +F: hw/riscv/opensbi_domain.c +F: include/hw/riscv/opensbi_domain.h + RENESAS RX CPUs R: Yoshinori Sato S: Orphan diff --git a/hw/riscv/Kconfig b/hw/riscv/Kconfig index a2030e3a6f..db3a4d77ad 100644 --- a/hw/riscv/Kconfig +++ b/hw/riscv/Kconfig @@ -1,6 +1,9 @@ config RISCV_NUMA bool =20 +config RISCV_OPENSBI_DOMAIN + bool + config IBEX bool =20 @@ -40,6 +43,7 @@ config RISCV_VIRT imply TPM_TIS_SYSBUS select DEVICE_TREE select RISCV_NUMA + select RISCV_OPENSBI_DOMAIN select GOLDFISH_RTC select PCI select PCI_EXPRESS_GENERIC_BRIDGE diff --git a/hw/riscv/meson.build b/hw/riscv/meson.build index f872674093..f47626c164 100644 --- a/hw/riscv/meson.build +++ b/hw/riscv/meson.build @@ -1,6 +1,7 @@ riscv_ss =3D ss.source_set() riscv_ss.add(files('boot.c')) riscv_ss.add(when: 'CONFIG_RISCV_NUMA', if_true: files('numa.c')) +riscv_ss.add(when: 'CONFIG_RISCV_OPENSBI_DOMAIN', if_true: files('opensbi_= domain.c')) riscv_ss.add(files('riscv_hart.c')) riscv_ss.add(when: 'CONFIG_OPENTITAN', if_true: files('opentitan.c')) riscv_ss.add(when: 'CONFIG_RISCV_VIRT', if_true: files('virt.c')) diff --git a/hw/riscv/opensbi_domain.c b/hw/riscv/opensbi_domain.c new file mode 100644 index 0000000000..da315e59b2 --- /dev/null +++ b/hw/riscv/opensbi_domain.c @@ -0,0 +1,558 @@ +/* + * OpenSBI Domains + * + * Copyright (c) 2024 Gregor Haas + * + * Generates OpenSBI domain nodes in the machine's device tree + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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 f= or + * more details. + * + * You should have received a copy of the GNU General Public License along= with + * this program. If not, see . + */ + +#include "qemu/osdep.h" +#include "hw/riscv/opensbi_domain.h" +#include "hw/boards.h" +#include "hw/riscv/virt.h" +#include "qapi/error.h" +#include "qemu/cutils.h" +#include "qemu/error-report.h" +#include "sysemu/device_tree.h" + +#include + +static void create_fdt_domain_possible_harts(MachineState *ms, + OpenSBIDomainState *s, + char *path) { + unsigned long i, cpu; + unsigned long num_cpus; + + num_cpus =3D s->last_possible_hart - s->first_possible_hart + 1; + if (num_cpus) { + g_autofree uint32_t *phandles =3D g_malloc0_n(num_cpus, sizeof(uin= t32_t)); + + for (i =3D 0, cpu =3D s->first_possible_hart; i < num_cpus; i++, c= pu++) { + g_autofree char *cpu_name =3D g_strdup_printf("/cpus/cpu@%li",= cpu); + phandles[i] =3D cpu_to_fdt32(qemu_fdt_get_phandle( + ms->fdt, cpu_name)); + } + + qemu_fdt_setprop(ms->fdt, path, "possible-harts", + phandles, num_cpus * 4); + } +} + +static void create_fdt_domain_regions(MachineState *ms, + OpenSBIDomainState *s, + char *path) { + unsigned long i; + int num_regions =3D 0; + DeviceState *ds; + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (s->regions[i]) { + num_regions++; + } + } + + if (num_regions) { + g_autofree uint32_t *regions =3D + g_malloc0_n(num_regions, 2 * sizeof(uint32_t)); + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (s->regions[i]) { + ds =3D DEVICE(s->regions[i]); + g_autofree char *region_name =3D g_strdup_printf( + "/chosen/opensbi-domains/%s", ds->id); + regions[2 * i] =3D cpu_to_fdt32(qemu_fdt_get_phandle + (ms->fdt, region_name)); + regions[2 * i + 1] =3D cpu_to_fdt32(s->region_perms[i]); + } + } + + qemu_fdt_setprop(ms->fdt, path, "regions", + regions, num_regions * 8); + } +} + +struct DomainFDTState { + MachineState *ms; + bool regions; +}; + +static void create_fdt_one_domain(MachineState *ms, OpenSBIDomainState *s) +{ + DeviceState *ds =3D DEVICE(s); + g_autofree char *path =3D g_strdup_printf("/chosen/opensbi-domains/%s"= , ds->id); + + qemu_fdt_add_subnode(ms->fdt, path); + qemu_fdt_setprop_string(ms->fdt, path, "compatible", + "opensbi,domain,instance"); + qemu_fdt_setprop_cells(ms->fdt, path, "phandle", + qemu_fdt_alloc_phandle(ms->fdt)); + + create_fdt_domain_possible_harts(ms, s, path); + create_fdt_domain_regions(ms, s, path); + + if (s->boot_hart !=3D -1) { + g_autofree char *cpu_name =3D g_strdup_printf("/cpus/cpu@%i", s->b= oot_hart); + qemu_fdt_setprop_cell(ms->fdt, path, "boot-hart", + qemu_fdt_get_phandle(ms->fdt, cpu_name)); + if (s->assign) { + qemu_fdt_setprop_cell(ms->fdt, cpu_name, "opensbi-domain", + qemu_fdt_get_phandle(ms->fdt, path)); + } + } + + if (s->next_arg1 !=3D -1) { + qemu_fdt_setprop_cells(ms->fdt, path, "next-arg1", + (uint64_t) s->next_arg1 >> 32, s->next_arg1); + } + + if (s->next_addr !=3D -1) { + qemu_fdt_setprop_cells(ms->fdt, path, "next-addr", + (uint64_t) s->next_addr >> 32, s->next_addr); + } + + if (s->next_mode !=3D -1) { + qemu_fdt_setprop_cell(ms->fdt, path, "next-mode", + s->next_mode); + } + + if (s->system_reset_allowed) { + qemu_fdt_setprop(ms->fdt, path, "system-reset-allowed", NULL, 0); + } + + if (s->system_suspend_allowed) { + qemu_fdt_setprop(ms->fdt, path, "system-suspend-allowed", NULL, 0); + } +} + +static uint32_t create_fdt_one_device(MachineState *ms, char *device) +{ + uint32_t phandle; + int offs =3D fdt_path_offset(ms->fdt, device); + + if (offs < 0) { + error_report("%s: Could not find device %s: %s", __func__, + device, fdt_strerror(offs)); + exit(1); + } + + phandle =3D fdt_get_phandle(ms->fdt, offs); + if (!phandle) { + phandle =3D qemu_fdt_alloc_phandle(ms->fdt); + qemu_fdt_setprop_cell(ms->fdt, device, "phandle", phandle); + } + + return phandle; +} + +static void create_fdt_one_memregion(MachineState *ms, + OpenSBIMemregionState *s) +{ + g_autofree char *path; + int i, dev, num_devices; + DeviceState *ds =3D DEVICE(s); + + path =3D g_strdup_printf("/chosen/opensbi-domains/%s", ds->id); + qemu_fdt_add_subnode(ms->fdt, path); + qemu_fdt_setprop_string(ms->fdt, path, "compatible", + "opensbi,domain,memregion"); + qemu_fdt_setprop_cells(ms->fdt, path, "base", + (uint64_t) s->base >> 32, s->base); + + qemu_fdt_setprop_cell(ms->fdt, path, "order", + (uint32_t) s->order); + + if (s->mmio) { + qemu_fdt_setprop(ms->fdt, path, "mmio", NULL, 0); + + /* Get all phandles for related devices */ + num_devices =3D 0; + for (i =3D 0; i < OPENSBI_MEMREGION_DEVICES_MAX; i++) { + if (s->devices[i]) { + num_devices++; + } + } + + if (num_devices) { + g_autofree uint32_t *devices =3D + g_malloc0_n(num_devices, sizeof(uint32_t)); + for (i =3D 0, dev =3D 0; i < OPENSBI_MEMREGION_DEVICES_MAX && + dev < num_devices; i++) { + if (s->devices[i]) { + devices[dev++] =3D create_fdt_one_device(ms, + s->devices[i]); + } + } + + qemu_fdt_setprop(ms->fdt, path, "devices", devices, + num_devices * 4); + } + } + + qemu_fdt_setprop_cells(ms->fdt, path, "phandle", + qemu_fdt_alloc_phandle(ms->fdt)); +} + +static int create_fdt_domains(Object *obj, void *opaque) +{ + struct DomainFDTState *dfs =3D opaque; + OpenSBIDomainState *osds; + OpenSBIMemregionState *osms; + + osds =3D (OpenSBIDomainState *) + object_dynamic_cast(obj, TYPE_OPENSBI_DOMAIN); + osms =3D (OpenSBIMemregionState *) + object_dynamic_cast(obj, TYPE_OPENSBI_MEMREGION); + + if (dfs->regions) { + if (osms) { + create_fdt_one_memregion(dfs->ms, osms); + } + } else { + if (osds) { + create_fdt_one_domain(dfs->ms, osds); + } + } + + return 0; +} + +static const char *containers[] =3D { + "/peripheral", "/peripheral-anon" +}; + +void create_fdt_opensbi_domains(MachineState *s) +{ + int i; + MachineState *ms =3D MACHINE(s); + Object *container; + + struct DomainFDTState check =3D { + .ms =3D ms, + .regions =3D true + }; + + /* Make sure that top-level node exists */ + qemu_fdt_add_subnode(ms->fdt, "/chosen/opensbi-domains"); + qemu_fdt_setprop_string(ms->fdt, "/chosen/opensbi-domains", + "compatible", "opensbi,domain,config"); + + /* Do a scan through regions first */ + for (i =3D 0; i < ARRAY_SIZE(containers); i++) { + container =3D container_get(OBJECT(s), containers[i]); + object_child_foreach(container, create_fdt_domains, &check); + } + + /* Then scan through domains */ + check.regions =3D false; + for (i =3D 0; i < ARRAY_SIZE(containers); i++) { + container =3D container_get(OBJECT(s), containers[i]); + object_child_foreach(container, create_fdt_domains, &check); + } +} + +/* OpenSBI Memregions */ + +static void set_mmio(Object *obj, bool val, Error **err) +{ + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + s->mmio =3D val; +} + +static void set_device(Object *obj, const char *val, Error **err) +{ + int i; + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + if (!s->devices[i]) { + s->devices[i] =3D g_strdup(val); + break; + } + } +} + +static void opensbi_memregion_instance_init(Object *obj) +{ + int i; + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(obj); + + s->base =3D -1; + object_property_add_uint64_ptr(obj, "base", &s->base, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "base", + "The base address of the domain memory= region. If \"order\" is also specified, " + "this property should be a 2 ^ order a= ligned 64 bit address"); + + s->order =3D -1; + object_property_add_uint32_ptr(obj, "order", &s->order, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "order", + "The order of the domain memory region= . This property should have a 32 bit value " + "(i.e. one DT cell) in the range 3 <= =3D order <=3D __riscv_xlen."); + + s->mmio =3D false; + object_property_add_bool(obj, "mmio", NULL, set_mmio); + object_property_set_description(obj, "mmio", + "A boolean flag representing whether t= he domain memory region is a " + "memory-mapped I/O (MMIO) region."); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + g_autofree char *propname =3D g_strdup_printf("device%i", i); + object_property_add_str(obj, propname, NULL, set_device); + + g_autofree char *description =3D g_strdup_printf( + "Device %i (out of %i) for this memregion. This property s= hould be a device tree path to the device.", + i, OPENSBI_DOMAIN_MEMREGIONS_MAX); + object_property_set_description(obj, propname, description); + } +} + +static void opensbi_memregion_realize(DeviceState *ds, Error **errp) +{ + #if defined(TARGET_RISCV32) + int xlen =3D 32; + #elif defined(TARGET_RISCV64) + int xlen =3D 64; + #endif + + OpenSBIMemregionState *s =3D OPENSBI_MEMREGION(ds); + + if (s->base =3D=3D -1) { + error_setg(errp, "must specify base"); + return; + } + + if (s->order =3D=3D -1) { + error_setg(errp, "must specify order"); + return; + } + + /* Check order bounds */ + if (s->order < 3 || s->order > xlen) { + error_setg(errp, "order must be between 3 and %d", xlen); + return; + } + + /* Check base alignment */ + if (s->order < xlen && (s->base & (BIT(s->order) - 1))) { + error_setg(errp, "base not aligned to order"); + return; + } +} + +static void opensbi_memregion_class_init(ObjectClass *oc, void *opaque) +{ + DeviceClass *dc =3D DEVICE_CLASS(oc); + dc->realize =3D opensbi_memregion_realize; +} + +static const TypeInfo opensbi_memregion_info =3D { + .name =3D TYPE_OPENSBI_MEMREGION, + .parent =3D TYPE_DEVICE, + .instance_init =3D opensbi_memregion_instance_init, + .instance_size =3D sizeof(OpenSBIDomainState), + .class_init =3D opensbi_memregion_class_init +}; + +/* OpenSBI Domains */ + +static void set_sysreset_allowed(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->system_reset_allowed =3D val; +} + +static void set_suspend_allowed(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->system_suspend_allowed =3D val; +} + +static void set_assign(Object *obj, bool val, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + s->assign =3D val; +} + +static void set_possible_harts(Object *obj, const char *str, Error **err) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + const char *firstcpu, *firstcpu_end, *lastcpu; + + firstcpu =3D str; + if (qemu_strtoul(firstcpu, &firstcpu_end, 0, + &s->first_possible_hart) < 0) { + error_setg(err, "could not convert firstcpu"); + return; + } + + lastcpu =3D qemu_strchrnul(str, '-'); + if (*lastcpu) { + if (lastcpu !=3D firstcpu_end) { + error_setg(err, "could not separate firstcpu and lastcpu"); + return; + } + + lastcpu++; + if (qemu_strtoul(lastcpu, NULL, 0, + &s->last_possible_hart) < 0) { + error_setg(err, "could not convert lastcpu"); + return; + } + } else { + s->last_possible_hart =3D s->first_possible_hart; + } +} + +static void opensbi_domain_instance_init(Object *obj) +{ + int i; + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(obj); + + s->boot_hart =3D VIRT_CPUS_MAX; + object_property_add_uint32_ptr(obj, "boot-hart", &s->boot_hart, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "boot-hart", + "The HART booting the domain instance.= "); + + s->first_possible_hart =3D -1; + s->last_possible_hart =3D -1; + object_property_add_str(obj, "possible-harts", NULL, set_possible_hart= s); + object_property_set_description(obj, "possible-harts", + "The contiguous list of CPUs for the d= omain instance, specified as firstcpu[-lastcpu]"); + + s->next_arg1 =3D -1; + object_property_add_uint64_ptr(obj, "next-arg1", &s->next_arg1, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-arg1", + "The 64 bit next booting stage arg1 fo= r the domain instance."); + + s->next_addr =3D -1; + object_property_add_uint64_ptr(obj, "next-addr", &s->next_addr, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-addr", + "The 64 bit next booting stage address= for the domain instance."); + + s->next_mode =3D -1; + object_property_add_uint32_ptr(obj, "next-mode", &s->next_mode, + OBJ_PROP_FLAG_WRITE); + object_property_set_description(obj, "next-mode", + "The 32 bit next booting stage mode fo= r the domain instance."); + + s->system_reset_allowed =3D false; + object_property_add_bool(obj, "system-reset-allowed", NULL, + set_sysreset_allowed); + object_property_set_description(obj, "system-reset-allowed", + "Whether the domain instance is allowe= d to do system reset."); + + s->system_suspend_allowed =3D false; + object_property_add_bool(obj, "system-suspend-allowed", NULL, + set_suspend_allowed); + object_property_set_description(obj, "system-suspend-allowed", + "Whether the domain instance is allowe= d to do system suspend."); + + for (i =3D 0; i < OPENSBI_DOMAIN_MEMREGIONS_MAX; i++) { + s->regions[i] =3D NULL; + g_autofree char *reg_propname =3D g_strdup_printf("region%i", i); + object_property_add_link(obj, reg_propname, TYPE_OPENSBI_MEMREGION, + (Object **) &s->regions[i], + qdev_prop_allow_set_link_before_realize, = 0); + + g_autofree char *reg_description =3D g_strdup_printf( + "Region %i (out of %i) for this domain.", + i, OPENSBI_DOMAIN_MEMREGIONS_MAX); + object_property_set_description(obj, reg_propname, reg_description= ); + + s->region_perms[i] =3D 0; + g_autofree char *perm_propname =3D g_strdup_printf("perms%i", i); + object_property_add_uint32_ptr(obj, perm_propname, &s->region_perm= s[i], + OBJ_PROP_FLAG_WRITE); + + g_autofree char *perm_description =3D g_strdup_printf( + "Permissions for region %i for this domain.", i); + object_property_set_description(obj, perm_propname, perm_descripti= on); + } + + object_property_add_bool(obj, "assign", NULL, set_assign); + object_property_set_description(obj, "assign", + "Whether to assign this domain to its = boot hart."); +} + +static void opensbi_domain_realize(DeviceState *ds, Error **errp) +{ + OpenSBIDomainState *s =3D OPENSBI_DOMAIN(ds); + + if (!ds->id) { + error_setg(errp, "must specify an id"); + return; + } + + if (s->boot_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, "boot hart larger than maximum number of CPUs (%d= )", + VIRT_CPUS_MAX); + return; + } + + if (s->first_possible_hart =3D=3D -1) { + if (s->last_possible_hart !=3D -1) { + error_setg(errp, + "last possible hart set when first possible hart unse= t"); + return; + } + } else { + if (s->first_possible_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, + "first possible hart larger than maximum number of CP= Us (%d)", + VIRT_CPUS_MAX); + return; + } + + if (s->last_possible_hart !=3D -1) { + if (s->last_possible_hart < s->first_possible_hart) { + error_setg(errp, + "last possible hart larger than first possible ha= rt"); + return; + } + + if (s->last_possible_hart >=3D VIRT_CPUS_MAX) { + error_setg(errp, + "last possible hart larger than maximum number of= CPUS (%d)", + VIRT_CPUS_MAX); + return; + } + } + } +} + +static void opensbi_domain_class_init(ObjectClass *oc, void *opaque) +{ + DeviceClass *dc =3D DEVICE_CLASS(oc); + dc->realize =3D opensbi_domain_realize; +} + +static const TypeInfo opensbi_domain_info =3D { + .name =3D TYPE_OPENSBI_DOMAIN, + .parent =3D TYPE_DEVICE, + .instance_init =3D opensbi_domain_instance_init, + .instance_size =3D sizeof(OpenSBIDomainState), + .class_init =3D opensbi_domain_class_init +}; + +static void opensbi_register_types(void) +{ + type_register_static(&opensbi_domain_info); + type_register_static(&opensbi_memregion_info); +} + +type_init(opensbi_register_types) diff --git a/hw/riscv/virt.c b/hw/riscv/virt.c index cef41c150a..760d8ce4d4 100644 --- a/hw/riscv/virt.c +++ b/hw/riscv/virt.c @@ -55,6 +55,7 @@ #include "hw/acpi/aml-build.h" #include "qapi/qapi-visit-common.h" #include "hw/virtio/virtio-iommu.h" +#include "hw/riscv/opensbi_domain.h" =20 /* KVM AIA only supports APLIC MSI. APLIC Wired is always emulated by QEMU= . */ static bool virt_use_kvm_aia(RISCVVirtState *s) @@ -1050,6 +1051,8 @@ static void finalize_fdt(RISCVVirtState *s) create_fdt_uart(s, virt_memmap, irq_mmio_phandle); =20 create_fdt_rtc(s, virt_memmap, irq_mmio_phandle); + + create_fdt_opensbi_domains(MACHINE(s)); } =20 static void create_fdt(RISCVVirtState *s, const MemMapEntry *memmap) diff --git a/include/hw/riscv/opensbi_domain.h b/include/hw/riscv/opensbi_d= omain.h new file mode 100644 index 0000000000..1ec8ed8288 --- /dev/null +++ b/include/hw/riscv/opensbi_domain.h @@ -0,0 +1,69 @@ +/* + * OpenSBI Domains + * + * Copyright (c) 2024 Gregor Haas + * + * Generates OpenSBI domain nodes in the machine's device tree + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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 f= or + * more details. + * + * You should have received a copy of the GNU General Public License along= with + * this program. If not, see . + */ + +#ifndef RISCV_DOMAIN_H +#define RISCV_DOMAIN_H + +#include "hw/sysbus.h" +#include "qom/object.h" +#include "cpu.h" + +#define TYPE_OPENSBI_MEMREGION "opensbi-memregion" +OBJECT_DECLARE_SIMPLE_TYPE(OpenSBIMemregionState, OPENSBI_MEMREGION) + +#define OPENSBI_MEMREGION_DEVICES_MAX 16 + +struct OpenSBIMemregionState { + /* public */ + DeviceState parent_obj; + + /* private */ + uint64_t base; + uint32_t order; + bool mmio; + char *devices[OPENSBI_MEMREGION_DEVICES_MAX]; +}; + +#define TYPE_OPENSBI_DOMAIN "opensbi-domain" +OBJECT_DECLARE_SIMPLE_TYPE(OpenSBIDomainState, OPENSBI_DOMAIN) + +#define OPENSBI_DOMAIN_MEMREGIONS_MAX 16 + +struct OpenSBIDomainState { + /* public */ + DeviceState parent_obj; + + /* private */ + OpenSBIMemregionState *regions[OPENSBI_DOMAIN_MEMREGIONS_MAX]; + unsigned int region_perms[OPENSBI_DOMAIN_MEMREGIONS_MAX]; + unsigned long first_possible_hart, last_possible_hart; + unsigned int boot_hart; + uint64_t next_arg1; + uint64_t next_addr; + uint32_t next_mode; + bool system_reset_allowed; + bool system_suspend_allowed; + + bool assign; +}; + +void create_fdt_opensbi_domains(MachineState *s); + +#endif /* RISCV_DOMAIN_H */ --=20 2.46.0 From nobody Sun Nov 24 05:33:37 2024 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=gmail.com ARC-Seal: i=1; a=rsa-sha256; t=1726003599; cv=none; d=zohomail.com; s=zohoarc; b=LxXvDMBsf8u5zjDKcHFvJEhgAoXcXKDpy5Im2nGgyDIsreRUgySU/Hglrn96DwHTXb6B7z4D8vpsKIvlq9hIsEyPh55TERcea6Pq97ry6n4T7hl4Nj6X8EMg8gsfY1s1IClzLHFAc2ebI3U3rGL9ggfzKy6sQ7ArIBWCqQFh04Y= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1726003599; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=HXSM6HNJszpKQyq7tPjFTacYfm2lIL1O8ndJeYautCGfdxVMJ/HBMiv6JEbkTnEzIE4oGBkqENCHpd2/Vk+PqJBFcsV0UStNfWEYOXBMZdxBzdSc56Q0sJSnpRKpr8gC5fsgbLGIsZZVm6qaVfqIqR2572bb4e2css0r2stGU9U= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 172600359967878.54026450145386; Tue, 10 Sep 2024 14:26:39 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1so8Ms-0002GG-9k; Tue, 10 Sep 2024 17:25:46 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1so8Mr-0002B0-4h; Tue, 10 Sep 2024 17:25:45 -0400 Received: from mail-yw1-x1130.google.com ([2607:f8b0:4864:20::1130]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1so8Mo-0002Iz-U0; Tue, 10 Sep 2024 17:25:44 -0400 Received: by mail-yw1-x1130.google.com with SMTP id 00721157ae682-6d3f4218081so59068327b3.1; Tue, 10 Sep 2024 14:25:42 -0700 (PDT) Received: from localhost ([2600:1700:830:3db0::14]) by smtp.gmail.com with UTF8SMTPSA id 00721157ae682-6db9650df6dsm4428357b3.103.2024.09.10.14.25.40 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 10 Sep 2024 14:25:40 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1726003541; x=1726608341; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=NNw/AwJWudejZh8znGmBJrK+PnlYsdROYO8YiMIOkwa9ILdkfeZHr9I+R7rWw661Aq zznGhVbvTnO/Q/0fUbIAMngKu6g8Q9tl+Vgd1zgHQRatchdy8JognfSYQa3YgihaEyNL jp4RttNWIcZHZdM/+meFaG+YCyqeW9XGzFCQ9o0Xk6tQfQ1a02wC1P28f4cMccTpbB9/ y+MAusBtm5nlWc3m1PQmD6pPI4EwGIud8lmz3JbLDoGjY86jfhxZOJCcuSWinQTiFNxr 67xbLuBfkZQvvKdiWRoBUzAI52/dpJ8rr5d7x9mk7ozzrIs49Unk83OQwVf/nuWKvCCS D6Ag== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1726003541; x=1726608341; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=Ici0+Jq1O2UdHCYqayx8tu3n7/R8R0FX9KP9tjMmUrFKmQue0JPfj1PymRORz8d/5M 2/Xu/4/5JNcjaN2VHpl02kDAYYiMZagk5vmFaNGvjBMaPEijLtH0rUGsrtRkYRCwvRSq nGIlZH/raS2mmSLJ3BWbWnflonWzUTIz7YOPVmLzHz9VDCeQSHICUmqO/S4VpDlilpus v3VG9ioEuacvX+XP0kud9Dou4NOZOwXMRhv9bBTS8oRyHY1xcUYvhZV6LP15SqL0t8ZG YwRw5Osowkd0k7N+NzvtzbdgLoNm4XkF+S5lBF4lzfd578uK9I/YJ4a+5EfHmc1+2N9j ux6g== X-Gm-Message-State: AOJu0YyhKUdZHxxPM/Qa78hqRkYKQBkgq153rug0Y+BK+QzrpAZNYYxF sIVOaV06TZ3tMTzgYRpdm2HJT0fwwGRFgz0Yd9w2eOB4W2JVGV2CoQjavOew X-Google-Smtp-Source: AGHT+IF9uGnRMRk4MebQSBNtEVoh/GtzWzYdf0FBa4If4bNikE0RBCGI9Do70Bi5nVkaqO8pB/iRuQ== X-Received: by 2002:a05:690c:350e:b0:6af:b0cc:abda with SMTP id 00721157ae682-6db45164af1mr168531427b3.40.1726003541368; Tue, 10 Sep 2024 14:25:41 -0700 (PDT) From: Gregor Haas To: qemu-devel@nongnu.org Cc: qemu-riscv@nongnu.org, atishp@rivosinc.com, dbarboza@ventanamicro.com, alistair.francis@wdc.com, Gregor Haas Subject: [PATCH v5 2/2] docs/system/riscv: Add documentation for command-line OpenSBI domains Date: Tue, 10 Sep 2024 14:25:33 -0700 Message-ID: <20240910212533.986734-5-gregorhaas1997@gmail.com> X-Mailer: git-send-email 2.46.0 In-Reply-To: <20240910212533.986734-1-gregorhaas1997@gmail.com> References: <20240910212533.986734-1-gregorhaas1997@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2607:f8b0:4864:20::1130; envelope-from=gregorhaas1997@gmail.com; helo=mail-yw1-x1130.google.com X-Spam_score_int: -17 X-Spam_score: -1.8 X-Spam_bar: - X-Spam_report: (-1.8 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @gmail.com) X-ZM-MESSAGEID: 1726003600990116600 Content-Type: text/plain; charset="utf-8" Signed-off-by: Gregor Haas --- MAINTAINERS | 1 + docs/system/riscv/opensbi_domains.rst | 156 ++++++++++++++++++++++++++ docs/system/target-riscv.rst | 10 ++ 3 files changed, 167 insertions(+) create mode 100644 docs/system/riscv/opensbi_domains.rst diff --git a/MAINTAINERS b/MAINTAINERS index 3b9f5b7432..5eb65cc499 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -361,6 +361,7 @@ RISC-V OpenSBI domain support M: Gregor Haas L: qemu-riscv@nongnu.org S: Maintained +F: docs/system/riscv/opensbi_domains.rst F: hw/riscv/opensbi_domain.c F: include/hw/riscv/opensbi_domain.h =20 diff --git a/docs/system/riscv/opensbi_domains.rst b/docs/system/riscv/open= sbi_domains.rst new file mode 100644 index 0000000000..b8bbf52738 --- /dev/null +++ b/docs/system/riscv/opensbi_domains.rst @@ -0,0 +1,156 @@ +OpenSBI Domains +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + +OpenSBI has support for domains, which are partitions of CPUs and memory i= nto +isolated compartments. Domains can be specified in the device tree accordi= ng to +a standardized format [1_], which OpenSBI parses at boot time to initializ= e all +system domains. Depending on the specific QEMU machine being used, these d= omain +configurations can be specified on the QEMU command line. Currently, this = is +only possible for the ``virt`` machine. To enable this functionality for a= new +machine, the initialization code must call ``create_fdt_opensbi_domains`` = after +all device tree nodes for peripherals have been initialized. This is to en= sure +that references to devices work for MMIO regions. + +There are two "devices" that are used to configure OpenSBI domains with th= is +mechanism: ``opensbi-memregion`` and ``opensbi-domain``. + +Memregions +---------- + +OpenSBI memregions can be added to the machine's device tree with the foll= owing +flag: + +.. code-block:: bash + + -device opensbi-memregion + +For this device flag, the following options are implemented: + +- ``id``: The name of the memregion. This name is later used to link this + region to a domain if desired, in which case this argument is required. +- ``base`` (required): The base address of this memregion. This address mu= st + be aligned to ``2 ^ order``. +- ``order`` (required): The ``log2`` of the memregion's size, which must be + between 3 and ``__riscv_xlen`` inclusive. +- ``mmio`` (optional): A boolean indicating whether the specified physical + address range belongs to an MMIO-mapped peripheral device. +- ``deviceX`` (optional): If ``mmio`` is indicated, this is the device tree + path to the ``X``-th device corresponding to this physical address range, + where ``0 <=3D X < OPENSBI_MEMREGION_DEVICES_MAX`` (default 16). + +Domains +------- + +OpenSBI domains can be added to the machine's device tree with the followi= ng +flag: + +.. code-block:: bash + + - device opensbi-domain + +For this device flag, the following options are implemented: + +- ``id`` (required): The name of the domain, which becomes its identifier = in + the device tree +- ``boot-hart`` (optional): The HART booting the domain instance. +- ``possible-harts`` (optional): The contiguous list of CPUs for the domain + instance, specified as ``firstcpu[-lastcpu]`` (e.g. ``0-3``). +- ``next-arg1`` (optional): The 64 bit next booting stage arg1 for the dom= ain + instance. +- ``next-addr`` (optional): The 64 bit next booting stage address for the + domain instance. +- ``next-mode`` (optional): The 32 bit next booting stage mode for the dom= ain + instance. +- ``system-reset-allowed`` (optional): Whether the domain instance is allo= wed + to do system reset. +- ``system-suspend-allowed`` (optional): Whether the domain instance is al= lowed + to do system suspend. + +Furthermore, memregions can be linked to domains using the following optio= ns: + +- ``regionX`` (optional): The ``id`` of the ``X``-th region for this domai= n, + where ``0 <=3D X < OPENSBI_DOMAIN_MEMREGIONS_MAX`` (default 16). +- ``permsX`` (optional): Access permissions for the ``X``-th region for th= is + domain, ``0 <=3D X < OPENSBI_DOMAIN_MEMREGIONS_MAX`` (default 16). This = must be + encoded using OpenSBI's permission encoding scheme in ``sbi_domain.h``, = and + copied below at the time of writing for convenience + +.. code-block:: c + + /** Flags representing memory region attributes */ + #define SBI_MEMREGION_M_READABLE (1UL << 0) + #define SBI_MEMREGION_M_WRITABLE (1UL << 1) + #define SBI_MEMREGION_M_EXECUTABLE (1UL << 2) + #define SBI_MEMREGION_SU_READABLE (1UL << 3) + #define SBI_MEMREGION_SU_WRITABLE (1UL << 4) + #define SBI_MEMREGION_SU_EXECUTABLE (1UL << 5) + +Example +------- + +A complete example command line is shown below: + +.. code-block:: bash + + $ qemu-system-riscv64 -machine virt -bios fw_jump.bin -cpu max -smp 2 = -m 4G -nographic \ + -device opensbi-memregion,id=3Dmem,base=3D0xBC000000,order=3D2= 6,mmio=3Dfalse \ + -device opensbi-memregion,id=3Duart,base=3D0x10000000,order=3D= 12,mmio=3Dtrue,device0=3D"/soc/serial@10000000" \ + -device opensbi-domain,id=3Ddomain,possible-harts=3D0-1,boot-h= art=3D0x0,next-addr=3D0xBC000000,next-mode=3D1,region0=3Dmem,perms0=3D0x3f,= region1=3Duart,perms1=3D0x3f + +As a result of the above configuration, QEMU will add the following subnod= es to +the device tree: + +.. code-block:: dts + + chosen { + opensbi-domains { + compatible =3D "opensbi,domain,config"; + + domain { + next-mode =3D <0x01>; + next-addr =3D <0x00 0xbc000000>; + boot-hart =3D <0x03>; + regions =3D <0x8000 0x3f 0x8002 0x3f>; + possible-harts =3D <0x03 0x01>; + phandle =3D <0x8003>; + compatible =3D "opensbi,domain,instance"; + }; + + uart { + phandle =3D <0x8002>; + devices =3D <0x1800000>; + mmio; + order =3D <0x0c>; + base =3D <0x00 0x10000000>; + compatible =3D "opensbi,domain,memregion"; + }; + + mem { + phandle =3D <0x8000>; + order =3D <0x1a>; + base =3D <0x00 0xbc000000>; + compatible =3D "opensbi,domain,memregion"; + }; + }; + }; + +This results in OpenSBI output as below, where regions 01-03 are inherited= from +the root domain and regions 00 and 04 correspond to the user specified one= s: + +.. code-block:: console + + Domain1 Name : domain + Domain1 Boot HART : 0 + Domain1 HARTs : 0,1 + Domain1 Region00 : 0x0000000010000000-0x0000000010000fff M: (= I,R,W,X) S/U: (R,W,X) + Domain1 Region01 : 0x0000000002000000-0x000000000200ffff M: (= I,R,W) S/U: () + Domain1 Region02 : 0x0000000080080000-0x000000008009ffff M: (= R,W) S/U: () + Domain1 Region03 : 0x0000000080000000-0x000000008007ffff M: (= R,X) S/U: () + Domain1 Region04 : 0x00000000bc000000-0x00000000bfffffff M: (= R,W,X) S/U: (R,W,X) + Domain1 Next Address : 0x00000000bc000000 + Domain1 Next Arg1 : 0x0000000000000000 + Domain1 Next Mode : S-mode + Domain1 SysReset : no + Domain1 SysSuspend : no + +.. _1: https://github.com/riscv-software-src/opensbi/blob/master/docs/doma= in_support.md diff --git a/docs/system/target-riscv.rst b/docs/system/target-riscv.rst index ba195f1518..bb776ea5f3 100644 --- a/docs/system/target-riscv.rst +++ b/docs/system/target-riscv.rst @@ -92,3 +92,13 @@ the images they need. * ``-bios `` =20 Tells QEMU to load the specified file as the firmware. + +RISC-V Devices +-------------- + +There are some RISC-V specific devices that can be specified: + +.. toctree:: + :maxdepth: 1 + + riscv/opensbi_domains --=20 2.46.0 From nobody Sun Nov 24 05:33:37 2024 Delivered-To: importer@patchew.org Authentication-Results: mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass(p=none dis=none) header.from=gmail.com ARC-Seal: i=1; a=rsa-sha256; t=1726003590; cv=none; d=zohomail.com; s=zohoarc; b=VpfGf6Oz9YW9eWcM10xegec1PiE3Vt2BhXj7FYjKGygDRSbr0iN12C6wI9c4Tx0+kOb05w34HuAQ+BqzM+SP97P7B4u+ddCB0TXXYswASYu8ux4SECf0rHlGufkcOKGitUzJ6jLrMw1I+9qx17yd+Wxzihcoloyew5Hzz859OHo= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=zohomail.com; s=zohoarc; t=1726003590; h=Content-Transfer-Encoding:Cc:Cc:Date:Date:From:From:In-Reply-To:List-Subscribe:List-Post:List-Id:List-Archive:List-Help:List-Unsubscribe:MIME-Version:Message-ID:References:Sender:Subject:Subject:To:To:Message-Id:Reply-To; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=iOwpJqTMOz3srYeeDyvWPSopVR3VRUVHFGxWpdY/22OSEOrZqxdmezf2W8QFiH8DL1Iuxscsc+gzMhBUWaRGWAzHxG6KzZtFGtxmONmtmldnungflnt7zm2GxPmnZ8rk49Uhm9P8hDOemTKy5OvHS+2ta2Zup6Dcc5SRNEHpths= ARC-Authentication-Results: i=1; mx.zohomail.com; dkim=pass; spf=pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) smtp.mailfrom=qemu-devel-bounces+importer=patchew.org@nongnu.org; dmarc=pass header.from= (p=none dis=none) Return-Path: Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) by mx.zohomail.com with SMTPS id 1726003590729128.34416034497804; Tue, 10 Sep 2024 14:26:30 -0700 (PDT) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1so8Mr-0002C3-D4; Tue, 10 Sep 2024 17:25:45 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1so8Mq-00026t-56; Tue, 10 Sep 2024 17:25:44 -0400 Received: from mail-yw1-x1134.google.com ([2607:f8b0:4864:20::1134]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.90_1) (envelope-from ) id 1so8Mn-0002Ir-Po; Tue, 10 Sep 2024 17:25:43 -0400 Received: by mail-yw1-x1134.google.com with SMTP id 00721157ae682-6d9f0cf5ae3so13530797b3.0; Tue, 10 Sep 2024 14:25:41 -0700 (PDT) Received: from localhost ([2600:1700:830:3db0::14]) by smtp.gmail.com with UTF8SMTPSA id 00721157ae682-6db965216e3sm4454407b3.130.2024.09.10.14.25.39 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Tue, 10 Sep 2024 14:25:39 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1726003540; x=1726608340; darn=nongnu.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=nioAumAqgu4yGw1IaxfbGkquatPraTKskkaEuRlxIB/+9+TZFKUk37dMPTIqMUcLc6 LLS26u/1Y8EZ/xZVqjVcDl5YD8r8tZhJOmnEmWc0Zm3Lk+MDDnlk95ee16NAgCuwxgZQ bzn21z0O3C4pFTo5THDsEuPXzfR6nuyz9gqHsXqG9IWlVC1w4aAUd5n36JQk070jlwJa trCSJ3ZqH3QuZIiy8wCENCYvZis0O+AA+hFG4WB68y87W71/Wswb7K5M6t1rovm0PSYE kMRMfSbElnXhc6GsnlBBEeBq3ptZBqky1xC1cT+D6590zL2SWs5Kkiy98Q4jTtUR2KoC eZuQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1726003540; x=1726608340; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=Sul7/BXCsjPRlqt3G2+RLEijKUxX/bYzUBFroC3XHX8=; b=ixiRzkLerLmQ4cdwNUoYhn2ENArjVAAkvE55OT8+Q7AMSjxH1Qpx9OmIZEM/34z0eA wiZL3FOagkU+lc9pzAbeKvX92fzHbalwVtp9/kZz9DCr1Od8KLUpubQpr2nFFzPWN+Lu GdzKI7brSGaP0kahOxyKmcIw/xojgx2jFS9fCANkYYGcJx/356k7AnMPrE5oFI//LWwe tpxS1MF0Nhe8OYdIM7cDgm0BeEZ3rWPFxvSpza7FGGmicFqsZSbHc5L+G+tOXMxGx92r Gm8eORj0Z9J6EdqelXv/9N4VB7pwQhk4Yc79O63DGavZ7Q9DRY+gt3ylLiAcE7+tChEQ ABJg== X-Gm-Message-State: AOJu0YxwbhE2Q2e7FNPpkKrMfsBogY5R6l+RTzyhLGEFdfWBYOKgWT1w RgxQzjx4mnzXWD7tDwHDIrhdvlEONLmleNcqH/9TqA/QapUESrpZt2ndzI1K X-Google-Smtp-Source: AGHT+IFfDruHfq0pQWyOfyEhNzR4eaXJs3gqUFwTM1bVD3qjzAAZrpjLjEX48asu+9UbFnrBrohKBw== X-Received: by 2002:a05:690c:60c3:b0:6db:4809:ed4e with SMTP id 00721157ae682-6dba6dc047bmr13909407b3.22.1726003540215; Tue, 10 Sep 2024 14:25:40 -0700 (PDT) From: Gregor Haas To: qemu-devel@nongnu.org Cc: qemu-riscv@nongnu.org, atishp@rivosinc.com, dbarboza@ventanamicro.com, alistair.francis@wdc.com, Gregor Haas Subject: [PATCH v5 2/2] Add documentation for command-line OpenSBI domains Date: Tue, 10 Sep 2024 14:25:32 -0700 Message-ID: <20240910212533.986734-4-gregorhaas1997@gmail.com> X-Mailer: git-send-email 2.46.0 In-Reply-To: <20240910212533.986734-1-gregorhaas1997@gmail.com> References: <20240910212533.986734-1-gregorhaas1997@gmail.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Received-SPF: pass (zohomail.com: domain of gnu.org designates 209.51.188.17 as permitted sender) client-ip=209.51.188.17; envelope-from=qemu-devel-bounces+importer=patchew.org@nongnu.org; helo=lists.gnu.org; Received-SPF: pass client-ip=2607:f8b0:4864:20::1134; envelope-from=gregorhaas1997@gmail.com; helo=mail-yw1-x1134.google.com X-Spam_score_int: -17 X-Spam_score: -1.8 X-Spam_bar: - X-Spam_report: (-1.8 / 5.0 requ) BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_NONE=-0.0001, SPF_HELO_NONE=0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+importer=patchew.org@nongnu.org Sender: qemu-devel-bounces+importer=patchew.org@nongnu.org X-ZohoMail-DKIM: pass (identity @gmail.com) X-ZM-MESSAGEID: 1726003592769116600 Content-Type: text/plain; charset="utf-8" Signed-off-by: Gregor Haas --- MAINTAINERS | 1 + docs/system/riscv/opensbi_domains.rst | 156 ++++++++++++++++++++++++++ docs/system/target-riscv.rst | 10 ++ 3 files changed, 167 insertions(+) create mode 100644 docs/system/riscv/opensbi_domains.rst diff --git a/MAINTAINERS b/MAINTAINERS index 3b9f5b7432..5eb65cc499 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -361,6 +361,7 @@ RISC-V OpenSBI domain support M: Gregor Haas L: qemu-riscv@nongnu.org S: Maintained +F: docs/system/riscv/opensbi_domains.rst F: hw/riscv/opensbi_domain.c F: include/hw/riscv/opensbi_domain.h =20 diff --git a/docs/system/riscv/opensbi_domains.rst b/docs/system/riscv/open= sbi_domains.rst new file mode 100644 index 0000000000..b8bbf52738 --- /dev/null +++ b/docs/system/riscv/opensbi_domains.rst @@ -0,0 +1,156 @@ +OpenSBI Domains +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + +OpenSBI has support for domains, which are partitions of CPUs and memory i= nto +isolated compartments. Domains can be specified in the device tree accordi= ng to +a standardized format [1_], which OpenSBI parses at boot time to initializ= e all +system domains. Depending on the specific QEMU machine being used, these d= omain +configurations can be specified on the QEMU command line. Currently, this = is +only possible for the ``virt`` machine. To enable this functionality for a= new +machine, the initialization code must call ``create_fdt_opensbi_domains`` = after +all device tree nodes for peripherals have been initialized. This is to en= sure +that references to devices work for MMIO regions. + +There are two "devices" that are used to configure OpenSBI domains with th= is +mechanism: ``opensbi-memregion`` and ``opensbi-domain``. + +Memregions +---------- + +OpenSBI memregions can be added to the machine's device tree with the foll= owing +flag: + +.. code-block:: bash + + -device opensbi-memregion + +For this device flag, the following options are implemented: + +- ``id``: The name of the memregion. This name is later used to link this + region to a domain if desired, in which case this argument is required. +- ``base`` (required): The base address of this memregion. This address mu= st + be aligned to ``2 ^ order``. +- ``order`` (required): The ``log2`` of the memregion's size, which must be + between 3 and ``__riscv_xlen`` inclusive. +- ``mmio`` (optional): A boolean indicating whether the specified physical + address range belongs to an MMIO-mapped peripheral device. +- ``deviceX`` (optional): If ``mmio`` is indicated, this is the device tree + path to the ``X``-th device corresponding to this physical address range, + where ``0 <=3D X < OPENSBI_MEMREGION_DEVICES_MAX`` (default 16). + +Domains +------- + +OpenSBI domains can be added to the machine's device tree with the followi= ng +flag: + +.. code-block:: bash + + - device opensbi-domain + +For this device flag, the following options are implemented: + +- ``id`` (required): The name of the domain, which becomes its identifier = in + the device tree +- ``boot-hart`` (optional): The HART booting the domain instance. +- ``possible-harts`` (optional): The contiguous list of CPUs for the domain + instance, specified as ``firstcpu[-lastcpu]`` (e.g. ``0-3``). +- ``next-arg1`` (optional): The 64 bit next booting stage arg1 for the dom= ain + instance. +- ``next-addr`` (optional): The 64 bit next booting stage address for the + domain instance. +- ``next-mode`` (optional): The 32 bit next booting stage mode for the dom= ain + instance. +- ``system-reset-allowed`` (optional): Whether the domain instance is allo= wed + to do system reset. +- ``system-suspend-allowed`` (optional): Whether the domain instance is al= lowed + to do system suspend. + +Furthermore, memregions can be linked to domains using the following optio= ns: + +- ``regionX`` (optional): The ``id`` of the ``X``-th region for this domai= n, + where ``0 <=3D X < OPENSBI_DOMAIN_MEMREGIONS_MAX`` (default 16). +- ``permsX`` (optional): Access permissions for the ``X``-th region for th= is + domain, ``0 <=3D X < OPENSBI_DOMAIN_MEMREGIONS_MAX`` (default 16). This = must be + encoded using OpenSBI's permission encoding scheme in ``sbi_domain.h``, = and + copied below at the time of writing for convenience + +.. code-block:: c + + /** Flags representing memory region attributes */ + #define SBI_MEMREGION_M_READABLE (1UL << 0) + #define SBI_MEMREGION_M_WRITABLE (1UL << 1) + #define SBI_MEMREGION_M_EXECUTABLE (1UL << 2) + #define SBI_MEMREGION_SU_READABLE (1UL << 3) + #define SBI_MEMREGION_SU_WRITABLE (1UL << 4) + #define SBI_MEMREGION_SU_EXECUTABLE (1UL << 5) + +Example +------- + +A complete example command line is shown below: + +.. code-block:: bash + + $ qemu-system-riscv64 -machine virt -bios fw_jump.bin -cpu max -smp 2 = -m 4G -nographic \ + -device opensbi-memregion,id=3Dmem,base=3D0xBC000000,order=3D2= 6,mmio=3Dfalse \ + -device opensbi-memregion,id=3Duart,base=3D0x10000000,order=3D= 12,mmio=3Dtrue,device0=3D"/soc/serial@10000000" \ + -device opensbi-domain,id=3Ddomain,possible-harts=3D0-1,boot-h= art=3D0x0,next-addr=3D0xBC000000,next-mode=3D1,region0=3Dmem,perms0=3D0x3f,= region1=3Duart,perms1=3D0x3f + +As a result of the above configuration, QEMU will add the following subnod= es to +the device tree: + +.. code-block:: dts + + chosen { + opensbi-domains { + compatible =3D "opensbi,domain,config"; + + domain { + next-mode =3D <0x01>; + next-addr =3D <0x00 0xbc000000>; + boot-hart =3D <0x03>; + regions =3D <0x8000 0x3f 0x8002 0x3f>; + possible-harts =3D <0x03 0x01>; + phandle =3D <0x8003>; + compatible =3D "opensbi,domain,instance"; + }; + + uart { + phandle =3D <0x8002>; + devices =3D <0x1800000>; + mmio; + order =3D <0x0c>; + base =3D <0x00 0x10000000>; + compatible =3D "opensbi,domain,memregion"; + }; + + mem { + phandle =3D <0x8000>; + order =3D <0x1a>; + base =3D <0x00 0xbc000000>; + compatible =3D "opensbi,domain,memregion"; + }; + }; + }; + +This results in OpenSBI output as below, where regions 01-03 are inherited= from +the root domain and regions 00 and 04 correspond to the user specified one= s: + +.. code-block:: console + + Domain1 Name : domain + Domain1 Boot HART : 0 + Domain1 HARTs : 0,1 + Domain1 Region00 : 0x0000000010000000-0x0000000010000fff M: (= I,R,W,X) S/U: (R,W,X) + Domain1 Region01 : 0x0000000002000000-0x000000000200ffff M: (= I,R,W) S/U: () + Domain1 Region02 : 0x0000000080080000-0x000000008009ffff M: (= R,W) S/U: () + Domain1 Region03 : 0x0000000080000000-0x000000008007ffff M: (= R,X) S/U: () + Domain1 Region04 : 0x00000000bc000000-0x00000000bfffffff M: (= R,W,X) S/U: (R,W,X) + Domain1 Next Address : 0x00000000bc000000 + Domain1 Next Arg1 : 0x0000000000000000 + Domain1 Next Mode : S-mode + Domain1 SysReset : no + Domain1 SysSuspend : no + +.. _1: https://github.com/riscv-software-src/opensbi/blob/master/docs/doma= in_support.md diff --git a/docs/system/target-riscv.rst b/docs/system/target-riscv.rst index ba195f1518..bb776ea5f3 100644 --- a/docs/system/target-riscv.rst +++ b/docs/system/target-riscv.rst @@ -92,3 +92,13 @@ the images they need. * ``-bios `` =20 Tells QEMU to load the specified file as the firmware. + +RISC-V Devices +-------------- + +There are some RISC-V specific devices that can be specified: + +.. toctree:: + :maxdepth: 1 + + riscv/opensbi_domains --=20 2.46.0