pcpu.c 9.21 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
/******************************************************************************
 * pcpu.c
 * Management physical cpu in dom0, get pcpu info and provide sys interface
 *
 * Copyright (c) 2012 Intel Corporation
 * Author: Liu, Jinsong <jinsong.liu@intel.com>
 * Author: Jiang, Yunhong <yunhong.jiang@intel.com>
 *
 * 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; or, when distributed
 * separately from the Linux kernel or incorporated into other
 * software packages, subject to the following license:
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this source file (the "Software"), to deal in the Software without
 * restriction, including without limitation the rights to use, copy, modify,
 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
 * and to permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

34 35
#define pr_fmt(fmt) "xen_cpu: " fmt

36 37 38 39 40 41 42
#include <linux/interrupt.h>
#include <linux/spinlock.h>
#include <linux/cpu.h>
#include <linux/stat.h>
#include <linux/capability.h>

#include <xen/xen.h>
43
#include <xen/acpi.h>
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80
#include <xen/xenbus.h>
#include <xen/events.h>
#include <xen/interface/platform.h>
#include <asm/xen/hypervisor.h>
#include <asm/xen/hypercall.h>


/*
 * @cpu_id: Xen physical cpu logic number
 * @flags: Xen physical cpu status flag
 * - XEN_PCPU_FLAGS_ONLINE: cpu is online
 * - XEN_PCPU_FLAGS_INVALID: cpu is not present
 */
struct pcpu {
	struct list_head list;
	struct device dev;
	uint32_t cpu_id;
	uint32_t flags;
};

static struct bus_type xen_pcpu_subsys = {
	.name = "xen_cpu",
	.dev_name = "xen_cpu",
};

static DEFINE_MUTEX(xen_pcpu_lock);

static LIST_HEAD(xen_pcpus);

static int xen_pcpu_down(uint32_t cpu_id)
{
	struct xen_platform_op op = {
		.cmd			= XENPF_cpu_offline,
		.interface_version	= XENPF_INTERFACE_VERSION,
		.u.cpu_ol.cpuid		= cpu_id,
	};

81
	return HYPERVISOR_platform_op(&op);
82 83 84 85 86 87 88 89 90 91
}

static int xen_pcpu_up(uint32_t cpu_id)
{
	struct xen_platform_op op = {
		.cmd			= XENPF_cpu_online,
		.interface_version	= XENPF_INTERFACE_VERSION,
		.u.cpu_ol.cpuid		= cpu_id,
	};

92
	return HYPERVISOR_platform_op(&op);
93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
}

static ssize_t show_online(struct device *dev,
			   struct device_attribute *attr,
			   char *buf)
{
	struct pcpu *cpu = container_of(dev, struct pcpu, dev);

	return sprintf(buf, "%u\n", !!(cpu->flags & XEN_PCPU_FLAGS_ONLINE));
}

static ssize_t __ref store_online(struct device *dev,
				  struct device_attribute *attr,
				  const char *buf, size_t count)
{
	struct pcpu *pcpu = container_of(dev, struct pcpu, dev);
	unsigned long long val;
	ssize_t ret;

	if (!capable(CAP_SYS_ADMIN))
		return -EPERM;

	if (kstrtoull(buf, 0, &val) < 0)
		return -EINVAL;

	switch (val) {
	case 0:
		ret = xen_pcpu_down(pcpu->cpu_id);
		break;
	case 1:
		ret = xen_pcpu_up(pcpu->cpu_id);
		break;
	default:
		ret = -EINVAL;
	}

	if (ret >= 0)
		ret = count;
	return ret;
}
static DEVICE_ATTR(online, S_IRUGO | S_IWUSR, show_online, store_online);

135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161
static struct attribute *pcpu_dev_attrs[] = {
	&dev_attr_online.attr,
	NULL
};

static umode_t pcpu_dev_is_visible(struct kobject *kobj,
				   struct attribute *attr, int idx)
{
	struct device *dev = kobj_to_dev(kobj);
	/*
	 * Xen never offline cpu0 due to several restrictions
	 * and assumptions. This basically doesn't add a sys control
	 * to user, one cannot attempt to offline BSP.
	 */
	return dev->id ? attr->mode : 0;
}

static const struct attribute_group pcpu_dev_group = {
	.attrs = pcpu_dev_attrs,
	.is_visible = pcpu_dev_is_visible,
};

static const struct attribute_group *pcpu_dev_groups[] = {
	&pcpu_dev_group,
	NULL
};

162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
static bool xen_pcpu_online(uint32_t flags)
{
	return !!(flags & XEN_PCPU_FLAGS_ONLINE);
}

static void pcpu_online_status(struct xenpf_pcpuinfo *info,
			       struct pcpu *pcpu)
{
	if (xen_pcpu_online(info->flags) &&
	   !xen_pcpu_online(pcpu->flags)) {
		/* the pcpu is onlined */
		pcpu->flags |= XEN_PCPU_FLAGS_ONLINE;
		kobject_uevent(&pcpu->dev.kobj, KOBJ_ONLINE);
	} else if (!xen_pcpu_online(info->flags) &&
		    xen_pcpu_online(pcpu->flags)) {
		/* The pcpu is offlined */
		pcpu->flags &= ~XEN_PCPU_FLAGS_ONLINE;
		kobject_uevent(&pcpu->dev.kobj, KOBJ_OFFLINE);
	}
}

static struct pcpu *get_pcpu(uint32_t cpu_id)
{
	struct pcpu *pcpu;

	list_for_each_entry(pcpu, &xen_pcpus, list) {
		if (pcpu->cpu_id == cpu_id)
			return pcpu;
	}

	return NULL;
}

static void pcpu_release(struct device *dev)
{
	struct pcpu *pcpu = container_of(dev, struct pcpu, dev);

	list_del(&pcpu->list);
	kfree(pcpu);
}

static void unregister_and_remove_pcpu(struct pcpu *pcpu)
{
	struct device *dev;

	if (!pcpu)
		return;

	dev = &pcpu->dev;
	/* pcpu remove would be implicitly done */
	device_unregister(dev);
}

static int register_pcpu(struct pcpu *pcpu)
{
	struct device *dev;
	int err = -EINVAL;

	if (!pcpu)
		return err;

	dev = &pcpu->dev;
	dev->bus = &xen_pcpu_subsys;
	dev->id = pcpu->cpu_id;
	dev->release = pcpu_release;
227
	dev->groups = pcpu_dev_groups;
228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258

	err = device_register(dev);
	if (err) {
		pcpu_release(dev);
		return err;
	}

	return 0;
}

static struct pcpu *create_and_register_pcpu(struct xenpf_pcpuinfo *info)
{
	struct pcpu *pcpu;
	int err;

	if (info->flags & XEN_PCPU_FLAGS_INVALID)
		return ERR_PTR(-ENODEV);

	pcpu = kzalloc(sizeof(struct pcpu), GFP_KERNEL);
	if (!pcpu)
		return ERR_PTR(-ENOMEM);

	INIT_LIST_HEAD(&pcpu->list);
	pcpu->cpu_id = info->xen_cpuid;
	pcpu->flags = info->flags;

	/* Need hold on xen_pcpu_lock before pcpu list manipulations */
	list_add_tail(&pcpu->list, &xen_pcpus);

	err = register_pcpu(pcpu);
	if (err) {
259
		pr_warn("Failed to register pcpu%u\n", info->xen_cpuid);
260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279
		return ERR_PTR(-ENOENT);
	}

	return pcpu;
}

/*
 * Caller should hold the xen_pcpu_lock
 */
static int sync_pcpu(uint32_t cpu, uint32_t *max_cpu)
{
	int ret;
	struct pcpu *pcpu = NULL;
	struct xenpf_pcpuinfo *info;
	struct xen_platform_op op = {
		.cmd                   = XENPF_get_cpuinfo,
		.interface_version     = XENPF_INTERFACE_VERSION,
		.u.pcpu_info.xen_cpuid = cpu,
	};

280
	ret = HYPERVISOR_platform_op(&op);
281 282 283 284 285 286 287 288 289 290 291 292 293
	if (ret)
		return ret;

	info = &op.u.pcpu_info;
	if (max_cpu)
		*max_cpu = info->max_present;

	pcpu = get_pcpu(cpu);

	/*
	 * Only those at cpu present map has its sys interface.
	 */
	if (info->flags & XEN_PCPU_FLAGS_INVALID) {
294
		unregister_and_remove_pcpu(pcpu);
295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347
		return 0;
	}

	if (!pcpu) {
		pcpu = create_and_register_pcpu(info);
		if (IS_ERR_OR_NULL(pcpu))
			return -ENODEV;
	} else
		pcpu_online_status(info, pcpu);

	return 0;
}

/*
 * Sync dom0's pcpu information with xen hypervisor's
 */
static int xen_sync_pcpus(void)
{
	/*
	 * Boot cpu always have cpu_id 0 in xen
	 */
	uint32_t cpu = 0, max_cpu = 0;
	int err = 0;
	struct pcpu *pcpu, *tmp;

	mutex_lock(&xen_pcpu_lock);

	while (!err && (cpu <= max_cpu)) {
		err = sync_pcpu(cpu, &max_cpu);
		cpu++;
	}

	if (err)
		list_for_each_entry_safe(pcpu, tmp, &xen_pcpus, list)
			unregister_and_remove_pcpu(pcpu);

	mutex_unlock(&xen_pcpu_lock);

	return err;
}

static void xen_pcpu_work_fn(struct work_struct *work)
{
	xen_sync_pcpus();
}
static DECLARE_WORK(xen_pcpu_work, xen_pcpu_work_fn);

static irqreturn_t xen_pcpu_interrupt(int irq, void *dev_id)
{
	schedule_work(&xen_pcpu_work);
	return IRQ_HANDLED;
}

Liu Jinsong's avatar
Liu Jinsong committed
348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366
/* Sync with Xen hypervisor after cpu hotadded */
void xen_pcpu_hotplug_sync(void)
{
	schedule_work(&xen_pcpu_work);
}
EXPORT_SYMBOL_GPL(xen_pcpu_hotplug_sync);

/*
 * For hypervisor presented cpu, return logic cpu id;
 * For hypervisor non-presented cpu, return -ENODEV.
 */
int xen_pcpu_id(uint32_t acpi_id)
{
	int cpu_id = 0, max_id = 0;
	struct xen_platform_op op;

	op.cmd = XENPF_get_cpuinfo;
	while (cpu_id <= max_id) {
		op.u.pcpu_info.xen_cpuid = cpu_id;
367
		if (HYPERVISOR_platform_op(&op)) {
Liu Jinsong's avatar
Liu Jinsong committed
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382
			cpu_id++;
			continue;
		}

		if (acpi_id == op.u.pcpu_info.acpi_id)
			return cpu_id;
		if (op.u.pcpu_info.max_present > max_id)
			max_id = op.u.pcpu_info.max_present;
		cpu_id++;
	}

	return -ENODEV;
}
EXPORT_SYMBOL_GPL(xen_pcpu_id);

383 384 385 386 387 388 389 390 391 392 393
static int __init xen_pcpu_init(void)
{
	int irq, ret;

	if (!xen_initial_domain())
		return -ENODEV;

	irq = bind_virq_to_irqhandler(VIRQ_PCPU_STATE, 0,
				      xen_pcpu_interrupt, 0,
				      "xen-pcpu", NULL);
	if (irq < 0) {
394
		pr_warn("Failed to bind pcpu virq\n");
395 396 397 398 399
		return irq;
	}

	ret = subsys_system_register(&xen_pcpu_subsys, NULL);
	if (ret) {
400
		pr_warn("Failed to register pcpu subsys\n");
401 402 403 404 405
		goto err1;
	}

	ret = xen_sync_pcpus();
	if (ret) {
406
		pr_warn("Failed to sync pcpu info\n");
407 408 409 410 411 412 413 414 415 416 417 418
		goto err2;
	}

	return 0;

err2:
	bus_unregister(&xen_pcpu_subsys);
err1:
	unbind_from_irqhandler(irq, NULL);
	return ret;
}
arch_initcall(xen_pcpu_init);