1// SPDX-License-Identifier: GPL-2.0-only
2/*
3 * coretemp.c - Linux kernel module for hardware monitoring
4 *
5 * Copyright (C) 2007 Rudolf Marek <r.marek@assembler.cz>
6 *
7 * Inspired from many hwmon drivers
8 */
9
10#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
11
12#include <linux/module.h>
13#include <linux/init.h>
14#include <linux/slab.h>
15#include <linux/jiffies.h>
16#include <linux/hwmon.h>
17#include <linux/sysfs.h>
18#include <linux/hwmon-sysfs.h>
19#include <linux/err.h>
20#include <linux/mutex.h>
21#include <linux/list.h>
22#include <linux/platform_device.h>
23#include <linux/cpu.h>
24#include <linux/smp.h>
25#include <linux/moduleparam.h>
26#include <linux/pci.h>
27#include <asm/msr.h>
28#include <asm/processor.h>
29#include <asm/cpu_device_id.h>
30
31#define DRVNAME	"coretemp"
32
33/*
34 * force_tjmax only matters when TjMax can't be read from the CPU itself.
35 * When set, it replaces the driver's suboptimal heuristic.
36 */
37static int force_tjmax;
38module_param_named(tjmax, force_tjmax, int, 0444);
39MODULE_PARM_DESC(tjmax, "TjMax value in degrees Celsius");
40
41#define PKG_SYSFS_ATTR_NO	1	/* Sysfs attribute for package temp */
42#define BASE_SYSFS_ATTR_NO	2	/* Sysfs Base attr no for coretemp */
43#define NUM_REAL_CORES		128	/* Number of Real cores per cpu */
44#define CORETEMP_NAME_LENGTH	28	/* String Length of attrs */
45#define MAX_CORE_ATTRS		4	/* Maximum no of basic attrs */
46#define TOTAL_ATTRS		(MAX_CORE_ATTRS + 1)
47#define MAX_CORE_DATA		(NUM_REAL_CORES + BASE_SYSFS_ATTR_NO)
48
49#ifdef CONFIG_SMP
50#define for_each_sibling(i, cpu) \
51	for_each_cpu(i, topology_sibling_cpumask(cpu))
52#else
53#define for_each_sibling(i, cpu)	for (i = 0; false; )
54#endif
55
56/*
57 * Per-Core Temperature Data
58 * @last_updated: The time when the current temperature value was updated
59 *		earlier (in jiffies).
60 * @cpu_core_id: The CPU Core from which temperature values should be read
61 *		This value is passed as "id" field to rdmsr/wrmsr functions.
62 * @status_reg: One of IA32_THERM_STATUS or IA32_PACKAGE_THERM_STATUS,
63 *		from where the temperature values should be read.
64 * @attr_size:  Total number of pre-core attrs displayed in the sysfs.
65 * @is_pkg_data: If this is 1, the temp_data holds pkgtemp data.
66 *		Otherwise, temp_data holds coretemp data.
67 * @valid: If this is 1, the current temperature is valid.
68 */
69struct temp_data {
70	int temp;
71	int ttarget;
72	int tjmax;
73	unsigned long last_updated;
74	unsigned int cpu;
75	u32 cpu_core_id;
76	u32 status_reg;
77	int attr_size;
78	bool is_pkg_data;
79	bool valid;
80	struct sensor_device_attribute sd_attrs[TOTAL_ATTRS];
81	char attr_name[TOTAL_ATTRS][CORETEMP_NAME_LENGTH];
82	struct attribute *attrs[TOTAL_ATTRS + 1];
83	struct attribute_group attr_group;
84	struct mutex update_lock;
85};
86
87/* Platform Data per Physical CPU */
88struct platform_data {
89	struct device		*hwmon_dev;
90	u16			pkg_id;
91	u16			cpu_map[NUM_REAL_CORES];
92	struct ida		ida;
93	struct cpumask		cpumask;
94	struct temp_data	*core_data[MAX_CORE_DATA];
95	struct device_attribute name_attr;
96};
97
98/* Keep track of how many zone pointers we allocated in init() */
99static int max_zones __read_mostly;
100/* Array of zone pointers. Serialized by cpu hotplug lock */
101static struct platform_device **zone_devices;
102
103static ssize_t show_label(struct device *dev,
104				struct device_attribute *devattr, char *buf)
105{
106	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
107	struct platform_data *pdata = dev_get_drvdata(dev);
108	struct temp_data *tdata = pdata->core_data[attr->index];
109
110	if (tdata->is_pkg_data)
111		return sprintf(buf, "Package id %u\n", pdata->pkg_id);
112
113	return sprintf(buf, "Core %u\n", tdata->cpu_core_id);
114}
115
116static ssize_t show_crit_alarm(struct device *dev,
117				struct device_attribute *devattr, char *buf)
118{
119	u32 eax, edx;
120	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
121	struct platform_data *pdata = dev_get_drvdata(dev);
122	struct temp_data *tdata = pdata->core_data[attr->index];
123
124	mutex_lock(&tdata->update_lock);
125	rdmsr_on_cpu(tdata->cpu, tdata->status_reg, &eax, &edx);
126	mutex_unlock(&tdata->update_lock);
127
128	return sprintf(buf, "%d\n", (eax >> 5) & 1);
129}
130
131static ssize_t show_tjmax(struct device *dev,
132			struct device_attribute *devattr, char *buf)
133{
134	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
135	struct platform_data *pdata = dev_get_drvdata(dev);
136
137	return sprintf(buf, "%d\n", pdata->core_data[attr->index]->tjmax);
138}
139
140static ssize_t show_ttarget(struct device *dev,
141				struct device_attribute *devattr, char *buf)
142{
143	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
144	struct platform_data *pdata = dev_get_drvdata(dev);
145
146	return sprintf(buf, "%d\n", pdata->core_data[attr->index]->ttarget);
147}
148
149static ssize_t show_temp(struct device *dev,
150			struct device_attribute *devattr, char *buf)
151{
152	u32 eax, edx;
153	struct sensor_device_attribute *attr = to_sensor_dev_attr(devattr);
154	struct platform_data *pdata = dev_get_drvdata(dev);
155	struct temp_data *tdata = pdata->core_data[attr->index];
156
157	mutex_lock(&tdata->update_lock);
158
159	/* Check whether the time interval has elapsed */
160	if (!tdata->valid || time_after(jiffies, tdata->last_updated + HZ)) {
161		rdmsr_on_cpu(tdata->cpu, tdata->status_reg, &eax, &edx);
162		/*
163		 * Ignore the valid bit. In all observed cases the register
164		 * value is either low or zero if the valid bit is 0.
165		 * Return it instead of reporting an error which doesn't
166		 * really help at all.
167		 */
168		tdata->temp = tdata->tjmax - ((eax >> 16) & 0x7f) * 1000;
169		tdata->valid = 1;
170		tdata->last_updated = jiffies;
171	}
172
173	mutex_unlock(&tdata->update_lock);
174	return sprintf(buf, "%d\n", tdata->temp);
175}
176
177struct tjmax_pci {
178	unsigned int device;
179	int tjmax;
180};
181
182static const struct tjmax_pci tjmax_pci_table[] = {
183	{ 0x0708, 110000 },	/* CE41x0 (Sodaville ) */
184	{ 0x0c72, 102000 },	/* Atom S1240 (Centerton) */
185	{ 0x0c73, 95000 },	/* Atom S1220 (Centerton) */
186	{ 0x0c75, 95000 },	/* Atom S1260 (Centerton) */
187};
188
189struct tjmax {
190	char const *id;
191	int tjmax;
192};
193
194static const struct tjmax tjmax_table[] = {
195	{ "CPU  230", 100000 },		/* Model 0x1c, stepping 2	*/
196	{ "CPU  330", 125000 },		/* Model 0x1c, stepping 2	*/
197};
198
199struct tjmax_model {
200	u8 model;
201	u8 mask;
202	int tjmax;
203};
204
205#define ANY 0xff
206
207static const struct tjmax_model tjmax_model_table[] = {
208	{ 0x1c, 10, 100000 },	/* D4xx, K4xx, N4xx, D5xx, K5xx, N5xx */
209	{ 0x1c, ANY, 90000 },	/* Z5xx, N2xx, possibly others
210				 * Note: Also matches 230 and 330,
211				 * which are covered by tjmax_table
212				 */
213	{ 0x26, ANY, 90000 },	/* Atom Tunnel Creek (Exx), Lincroft (Z6xx)
214				 * Note: TjMax for E6xxT is 110C, but CPU type
215				 * is undetectable by software
216				 */
217	{ 0x27, ANY, 90000 },	/* Atom Medfield (Z2460) */
218	{ 0x35, ANY, 90000 },	/* Atom Clover Trail/Cloverview (Z27x0) */
219	{ 0x36, ANY, 100000 },	/* Atom Cedar Trail/Cedarview (N2xxx, D2xxx)
220				 * Also matches S12x0 (stepping 9), covered by
221				 * PCI table
222				 */
223};
224
225static int adjust_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev)
226{
227	/* The 100C is default for both mobile and non mobile CPUs */
228
229	int tjmax = 100000;
230	int tjmax_ee = 85000;
231	int usemsr_ee = 1;
232	int err;
233	u32 eax, edx;
234	int i;
235	u16 devfn = PCI_DEVFN(0, 0);
236	struct pci_dev *host_bridge = pci_get_domain_bus_and_slot(0, 0, devfn);
237
238	/*
239	 * Explicit tjmax table entries override heuristics.
240	 * First try PCI host bridge IDs, followed by model ID strings
241	 * and model/stepping information.
242	 */
243	if (host_bridge && host_bridge->vendor == PCI_VENDOR_ID_INTEL) {
244		for (i = 0; i < ARRAY_SIZE(tjmax_pci_table); i++) {
245			if (host_bridge->device == tjmax_pci_table[i].device) {
246				pci_dev_put(host_bridge);
247				return tjmax_pci_table[i].tjmax;
248			}
249		}
250	}
251	pci_dev_put(host_bridge);
252
253	for (i = 0; i < ARRAY_SIZE(tjmax_table); i++) {
254		if (strstr(c->x86_model_id, tjmax_table[i].id))
255			return tjmax_table[i].tjmax;
256	}
257
258	for (i = 0; i < ARRAY_SIZE(tjmax_model_table); i++) {
259		const struct tjmax_model *tm = &tjmax_model_table[i];
260		if (c->x86_model == tm->model &&
261		    (tm->mask == ANY || c->x86_stepping == tm->mask))
262			return tm->tjmax;
263	}
264
265	/* Early chips have no MSR for TjMax */
266
267	if (c->x86_model == 0xf && c->x86_stepping < 4)
268		usemsr_ee = 0;
269
270	if (c->x86_model > 0xe && usemsr_ee) {
271		u8 platform_id;
272
273		/*
274		 * Now we can detect the mobile CPU using Intel provided table
275		 * http://softwarecommunity.intel.com/Wiki/Mobility/720.htm
276		 * For Core2 cores, check MSR 0x17, bit 28 1 = Mobile CPU
277		 */
278		err = rdmsr_safe_on_cpu(id, 0x17, &eax, &edx);
279		if (err) {
280			dev_warn(dev,
281				 "Unable to access MSR 0x17, assuming desktop"
282				 " CPU\n");
283			usemsr_ee = 0;
284		} else if (c->x86_model < 0x17 && !(eax & 0x10000000)) {
285			/*
286			 * Trust bit 28 up to Penryn, I could not find any
287			 * documentation on that; if you happen to know
288			 * someone at Intel please ask
289			 */
290			usemsr_ee = 0;
291		} else {
292			/* Platform ID bits 52:50 (EDX starts at bit 32) */
293			platform_id = (edx >> 18) & 0x7;
294
295			/*
296			 * Mobile Penryn CPU seems to be platform ID 7 or 5
297			 * (guesswork)
298			 */
299			if (c->x86_model == 0x17 &&
300			    (platform_id == 5 || platform_id == 7)) {
301				/*
302				 * If MSR EE bit is set, set it to 90 degrees C,
303				 * otherwise 105 degrees C
304				 */
305				tjmax_ee = 90000;
306				tjmax = 105000;
307			}
308		}
309	}
310
311	if (usemsr_ee) {
312		err = rdmsr_safe_on_cpu(id, 0xee, &eax, &edx);
313		if (err) {
314			dev_warn(dev,
315				 "Unable to access MSR 0xEE, for Tjmax, left"
316				 " at default\n");
317		} else if (eax & 0x40000000) {
318			tjmax = tjmax_ee;
319		}
320	} else if (tjmax == 100000) {
321		/*
322		 * If we don't use msr EE it means we are desktop CPU
323		 * (with exeception of Atom)
324		 */
325		dev_warn(dev, "Using relative temperature scale!\n");
326	}
327
328	return tjmax;
329}
330
331static bool cpu_has_tjmax(struct cpuinfo_x86 *c)
332{
333	u8 model = c->x86_model;
334
335	return model > 0xe &&
336	       model != 0x1c &&
337	       model != 0x26 &&
338	       model != 0x27 &&
339	       model != 0x35 &&
340	       model != 0x36;
341}
342
343static int get_tjmax(struct cpuinfo_x86 *c, u32 id, struct device *dev)
344{
345	int err;
346	u32 eax, edx;
347	u32 val;
348
349	/*
350	 * A new feature of current Intel(R) processors, the
351	 * IA32_TEMPERATURE_TARGET contains the TjMax value
352	 */
353	err = rdmsr_safe_on_cpu(id, MSR_IA32_TEMPERATURE_TARGET, &eax, &edx);
354	if (err) {
355		if (cpu_has_tjmax(c))
356			dev_warn(dev, "Unable to read TjMax from CPU %u\n", id);
357	} else {
358		val = (eax >> 16) & 0xff;
359		/*
360		 * If the TjMax is not plausible, an assumption
361		 * will be used
362		 */
363		if (val) {
364			dev_dbg(dev, "TjMax is %d degrees C\n", val);
365			return val * 1000;
366		}
367	}
368
369	if (force_tjmax) {
370		dev_notice(dev, "TjMax forced to %d degrees C by user\n",
371			   force_tjmax);
372		return force_tjmax * 1000;
373	}
374
375	/*
376	 * An assumption is made for early CPUs and unreadable MSR.
377	 * NOTE: the calculated value may not be correct.
378	 */
379	return adjust_tjmax(c, id, dev);
380}
381
382static int create_core_attrs(struct temp_data *tdata, struct device *dev,
383			     int index)
384{
385	int i;
386	static ssize_t (*const rd_ptr[TOTAL_ATTRS]) (struct device *dev,
387			struct device_attribute *devattr, char *buf) = {
388			show_label, show_crit_alarm, show_temp, show_tjmax,
389			show_ttarget };
390	static const char *const suffixes[TOTAL_ATTRS] = {
391		"label", "crit_alarm", "input", "crit", "max"
392	};
393
394	for (i = 0; i < tdata->attr_size; i++) {
395		/*
396		 * We map the attr number to core id of the CPU
397		 * The attr number is always core id + 2
398		 * The Pkgtemp will always show up as temp1_*, if available
399		 */
400		int attr_no = tdata->is_pkg_data ? 1 : tdata->cpu_core_id + 2;
401
402		snprintf(tdata->attr_name[i], CORETEMP_NAME_LENGTH,
403			 "temp%d_%s", attr_no, suffixes[i]);
404		sysfs_attr_init(&tdata->sd_attrs[i].dev_attr.attr);
405		tdata->sd_attrs[i].dev_attr.attr.name = tdata->attr_name[i];
406		tdata->sd_attrs[i].dev_attr.attr.mode = 0444;
407		tdata->sd_attrs[i].dev_attr.show = rd_ptr[i];
408		tdata->sd_attrs[i].index = index;
409		tdata->attrs[i] = &tdata->sd_attrs[i].dev_attr.attr;
410	}
411	tdata->attr_group.attrs = tdata->attrs;
412	return sysfs_create_group(&dev->kobj, &tdata->attr_group);
413}
414
415
416static int chk_ucode_version(unsigned int cpu)
417{
418	struct cpuinfo_x86 *c = &cpu_data(cpu);
419
420	/*
421	 * Check if we have problem with errata AE18 of Core processors:
422	 * Readings might stop update when processor visited too deep sleep,
423	 * fixed for stepping D0 (6EC).
424	 */
425	if (c->x86_model == 0xe && c->x86_stepping < 0xc && c->microcode < 0x39) {
426		pr_err("Errata AE18 not fixed, update BIOS or microcode of the CPU!\n");
427		return -ENODEV;
428	}
429	return 0;
430}
431
432static struct platform_device *coretemp_get_pdev(unsigned int cpu)
433{
434	int id = topology_logical_die_id(cpu);
435
436	if (id >= 0 && id < max_zones)
437		return zone_devices[id];
438	return NULL;
439}
440
441static struct temp_data *init_temp_data(unsigned int cpu, int pkg_flag)
442{
443	struct temp_data *tdata;
444
445	tdata = kzalloc(sizeof(struct temp_data), GFP_KERNEL);
446	if (!tdata)
447		return NULL;
448
449	tdata->status_reg = pkg_flag ? MSR_IA32_PACKAGE_THERM_STATUS :
450							MSR_IA32_THERM_STATUS;
451	tdata->is_pkg_data = pkg_flag;
452	tdata->cpu = cpu;
453	tdata->cpu_core_id = topology_core_id(cpu);
454	tdata->attr_size = MAX_CORE_ATTRS;
455	mutex_init(&tdata->update_lock);
456	return tdata;
457}
458
459static int create_core_data(struct platform_device *pdev, unsigned int cpu,
460			    int pkg_flag)
461{
462	struct temp_data *tdata;
463	struct platform_data *pdata = platform_get_drvdata(pdev);
464	struct cpuinfo_x86 *c = &cpu_data(cpu);
465	u32 eax, edx;
466	int err, index;
467
468	/*
469	 * Get the index of tdata in pdata->core_data[]
470	 * tdata for package: pdata->core_data[1]
471	 * tdata for core: pdata->core_data[2] .. pdata->core_data[NUM_REAL_CORES + 1]
472	 */
473	if (pkg_flag) {
474		index = PKG_SYSFS_ATTR_NO;
475	} else {
476		index = ida_alloc_max(&pdata->ida, NUM_REAL_CORES - 1, GFP_KERNEL);
477		if (index < 0)
478			return index;
479
480		pdata->cpu_map[index] = topology_core_id(cpu);
481		index += BASE_SYSFS_ATTR_NO;
482	}
483
484	tdata = init_temp_data(cpu, pkg_flag);
485	if (!tdata) {
486		err = -ENOMEM;
487		goto ida_free;
488	}
489
490	/* Test if we can access the status register */
491	err = rdmsr_safe_on_cpu(cpu, tdata->status_reg, &eax, &edx);
492	if (err)
493		goto exit_free;
494
495	/* We can access status register. Get Critical Temperature */
496	tdata->tjmax = get_tjmax(c, cpu, &pdev->dev);
497
498	/*
499	 * Read the still undocumented bits 8:15 of IA32_TEMPERATURE_TARGET.
500	 * The target temperature is available on older CPUs but not in this
501	 * register. Atoms don't have the register at all.
502	 */
503	if (c->x86_model > 0xe && c->x86_model != 0x1c) {
504		err = rdmsr_safe_on_cpu(cpu, MSR_IA32_TEMPERATURE_TARGET,
505					&eax, &edx);
506		if (!err) {
507			tdata->ttarget
508			  = tdata->tjmax - ((eax >> 8) & 0xff) * 1000;
509			tdata->attr_size++;
510		}
511	}
512
513	pdata->core_data[index] = tdata;
514
515	/* Create sysfs interfaces */
516	err = create_core_attrs(tdata, pdata->hwmon_dev, index);
517	if (err)
518		goto exit_free;
519
520	return 0;
521exit_free:
522	pdata->core_data[index] = NULL;
523	kfree(tdata);
524ida_free:
525	if (!pkg_flag)
526		ida_free(&pdata->ida, index - BASE_SYSFS_ATTR_NO);
527	return err;
528}
529
530static void
531coretemp_add_core(struct platform_device *pdev, unsigned int cpu, int pkg_flag)
532{
533	if (create_core_data(pdev, cpu, pkg_flag))
534		dev_err(&pdev->dev, "Adding Core %u failed\n", cpu);
535}
536
537static void coretemp_remove_core(struct platform_data *pdata, int indx)
538{
539	struct temp_data *tdata = pdata->core_data[indx];
540
541	/* if we errored on add then this is already gone */
542	if (!tdata)
543		return;
544
545	/* Remove the sysfs attributes */
546	sysfs_remove_group(&pdata->hwmon_dev->kobj, &tdata->attr_group);
547
548	kfree(pdata->core_data[indx]);
549	pdata->core_data[indx] = NULL;
550
551	if (indx >= BASE_SYSFS_ATTR_NO)
552		ida_free(&pdata->ida, indx - BASE_SYSFS_ATTR_NO);
553}
554
555static int coretemp_device_add(int zoneid)
556{
557	struct platform_device *pdev;
558	struct platform_data *pdata;
559	int err;
560
561	/* Initialize the per-zone data structures */
562	pdata = kzalloc(sizeof(*pdata), GFP_KERNEL);
563	if (!pdata)
564		return -ENOMEM;
565
566	pdata->pkg_id = zoneid;
567	ida_init(&pdata->ida);
568
569	pdev = platform_device_alloc(DRVNAME, zoneid);
570	if (!pdev) {
571		err = -ENOMEM;
572		goto err_free_pdata;
573	}
574
575	err = platform_device_add(pdev);
576	if (err)
577		goto err_put_dev;
578
579	platform_set_drvdata(pdev, pdata);
580	zone_devices[zoneid] = pdev;
581	return 0;
582
583err_put_dev:
584	platform_device_put(pdev);
585err_free_pdata:
586	kfree(pdata);
587	return err;
588}
589
590static void coretemp_device_remove(int zoneid)
591{
592	struct platform_device *pdev = zone_devices[zoneid];
593	struct platform_data *pdata = platform_get_drvdata(pdev);
594
595	ida_destroy(&pdata->ida);
596	kfree(pdata);
597	platform_device_unregister(pdev);
598}
599
600static int coretemp_cpu_online(unsigned int cpu)
601{
602	struct platform_device *pdev = coretemp_get_pdev(cpu);
603	struct cpuinfo_x86 *c = &cpu_data(cpu);
604	struct platform_data *pdata;
605
606	/*
607	 * Don't execute this on resume as the offline callback did
608	 * not get executed on suspend.
609	 */
610	if (cpuhp_tasks_frozen)
611		return 0;
612
613	/*
614	 * CPUID.06H.EAX[0] indicates whether the CPU has thermal
615	 * sensors. We check this bit only, all the early CPUs
616	 * without thermal sensors will be filtered out.
617	 */
618	if (!cpu_has(c, X86_FEATURE_DTHERM))
619		return -ENODEV;
620
621	pdata = platform_get_drvdata(pdev);
622	if (!pdata->hwmon_dev) {
623		struct device *hwmon;
624
625		/* Check the microcode version of the CPU */
626		if (chk_ucode_version(cpu))
627			return -EINVAL;
628
629		/*
630		 * Alright, we have DTS support.
631		 * We are bringing the _first_ core in this pkg
632		 * online. So, initialize per-pkg data structures and
633		 * then bring this core online.
634		 */
635		hwmon = hwmon_device_register_with_groups(&pdev->dev, DRVNAME,
636							  pdata, NULL);
637		if (IS_ERR(hwmon))
638			return PTR_ERR(hwmon);
639		pdata->hwmon_dev = hwmon;
640
641		/*
642		 * Check whether pkgtemp support is available.
643		 * If so, add interfaces for pkgtemp.
644		 */
645		if (cpu_has(c, X86_FEATURE_PTS))
646			coretemp_add_core(pdev, cpu, 1);
647	}
648
649	/*
650	 * Check whether a thread sibling is already online. If not add the
651	 * interface for this CPU core.
652	 */
653	if (!cpumask_intersects(&pdata->cpumask, topology_sibling_cpumask(cpu)))
654		coretemp_add_core(pdev, cpu, 0);
655
656	cpumask_set_cpu(cpu, &pdata->cpumask);
657	return 0;
658}
659
660static int coretemp_cpu_offline(unsigned int cpu)
661{
662	struct platform_device *pdev = coretemp_get_pdev(cpu);
663	struct platform_data *pd;
664	struct temp_data *tdata;
665	int i, indx = -1, target;
666
667	/* No need to tear down any interfaces for suspend */
668	if (cpuhp_tasks_frozen)
669		return 0;
670
671	/* If the physical CPU device does not exist, just return */
672	pd = platform_get_drvdata(pdev);
673	if (!pd->hwmon_dev)
674		return 0;
675
676	for (i = 0; i < NUM_REAL_CORES; i++) {
677		if (pd->cpu_map[i] == topology_core_id(cpu)) {
678			indx = i + BASE_SYSFS_ATTR_NO;
679			break;
680		}
681	}
682
683	/* Too many cores and this core is not populated, just return */
684	if (indx < 0)
685		return 0;
686
687	tdata = pd->core_data[indx];
688
689	cpumask_clear_cpu(cpu, &pd->cpumask);
690
691	/*
692	 * If this is the last thread sibling, remove the CPU core
693	 * interface, If there is still a sibling online, transfer the
694	 * target cpu of that core interface to it.
695	 */
696	target = cpumask_any_and(&pd->cpumask, topology_sibling_cpumask(cpu));
697	if (target >= nr_cpu_ids) {
698		coretemp_remove_core(pd, indx);
699	} else if (tdata && tdata->cpu == cpu) {
700		mutex_lock(&tdata->update_lock);
701		tdata->cpu = target;
702		mutex_unlock(&tdata->update_lock);
703	}
704
705	/*
706	 * If all cores in this pkg are offline, remove the interface.
707	 */
708	tdata = pd->core_data[PKG_SYSFS_ATTR_NO];
709	if (cpumask_empty(&pd->cpumask)) {
710		if (tdata)
711			coretemp_remove_core(pd, PKG_SYSFS_ATTR_NO);
712		hwmon_device_unregister(pd->hwmon_dev);
713		pd->hwmon_dev = NULL;
714		return 0;
715	}
716
717	/*
718	 * Check whether this core is the target for the package
719	 * interface. We need to assign it to some other cpu.
720	 */
721	if (tdata && tdata->cpu == cpu) {
722		target = cpumask_first(&pd->cpumask);
723		mutex_lock(&tdata->update_lock);
724		tdata->cpu = target;
725		mutex_unlock(&tdata->update_lock);
726	}
727	return 0;
728}
729static const struct x86_cpu_id __initconst coretemp_ids[] = {
730	X86_MATCH_VENDOR_FEATURE(INTEL, X86_FEATURE_DTHERM, NULL),
731	{}
732};
733MODULE_DEVICE_TABLE(x86cpu, coretemp_ids);
734
735static enum cpuhp_state coretemp_hp_online;
736
737static int __init coretemp_init(void)
738{
739	int i, err;
740
741	/*
742	 * CPUID.06H.EAX[0] indicates whether the CPU has thermal
743	 * sensors. We check this bit only, all the early CPUs
744	 * without thermal sensors will be filtered out.
745	 */
746	if (!x86_match_cpu(coretemp_ids))
747		return -ENODEV;
748
749	max_zones = topology_max_packages() * topology_max_die_per_package();
750	zone_devices = kcalloc(max_zones, sizeof(struct platform_device *),
751			      GFP_KERNEL);
752	if (!zone_devices)
753		return -ENOMEM;
754
755	for (i = 0; i < max_zones; i++) {
756		err = coretemp_device_add(i);
757		if (err)
758			goto outzone;
759	}
760
761	err = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "hwmon/coretemp:online",
762				coretemp_cpu_online, coretemp_cpu_offline);
763	if (err < 0)
764		goto outzone;
765	coretemp_hp_online = err;
766	return 0;
767
768outzone:
769	while (i--)
770		coretemp_device_remove(i);
771	kfree(zone_devices);
772	return err;
773}
774module_init(coretemp_init)
775
776static void __exit coretemp_exit(void)
777{
778	int i;
779
780	cpuhp_remove_state(coretemp_hp_online);
781	for (i = 0; i < max_zones; i++)
782		coretemp_device_remove(i);
783	kfree(zone_devices);
784}
785module_exit(coretemp_exit)
786
787MODULE_AUTHOR("Rudolf Marek <r.marek@assembler.cz>");
788MODULE_DESCRIPTION("Intel Core temperature monitor");
789MODULE_LICENSE("GPL");
790