[linux.git] / mm / debug.c

// SPDX-License-Identifier: GPL-2.0
/*
 * mm/debug.c
 *
 * mm/ specific debug routines.
 *
 */

#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/trace_events.h>
#include <linux/memcontrol.h>
#include <trace/events/mmflags.h>
#include <linux/migrate.h>
#include <linux/page_owner.h>
#include <linux/ctype.h>

#include "internal.h"
#include <trace/events/migrate.h>

/*
 * Define EM() and EMe() so that MIGRATE_REASON from trace/events/migrate.h can
 * be used to populate migrate_reason_names[].
 */
#undef EM
#undef EMe
#define EM(a, b)	b,
#define EMe(a, b)	b

const char *migrate_reason_names[MR_TYPES] = {
	MIGRATE_REASON
};

const struct trace_print_flags pageflag_names[] = {
	__def_pageflag_names,
	{0, NULL}
};

const struct trace_print_flags gfpflag_names[] = {
	__def_gfpflag_names,
	{0, NULL}
};

const struct trace_print_flags vmaflag_names[] = {
	__def_vmaflag_names,
	{0, NULL}
};

#define DEF_PAGETYPE_NAME(_name) [PGTY_##_name - 0xf0] =  __stringify(_name)

static const char *page_type_names[] = {
	DEF_PAGETYPE_NAME(slab),
	DEF_PAGETYPE_NAME(hugetlb),
	DEF_PAGETYPE_NAME(offline),
	DEF_PAGETYPE_NAME(guard),
	DEF_PAGETYPE_NAME(table),
	DEF_PAGETYPE_NAME(buddy),
	DEF_PAGETYPE_NAME(unaccepted),
};

static const char *page_type_name(unsigned int page_type)
{
	unsigned i = (page_type >> 24) - 0xf0;

	if (i >= ARRAY_SIZE(page_type_names))
		return "unknown";
	return page_type_names[i];
}

static void __dump_folio(struct folio *folio, struct page *page,
		unsigned long pfn, unsigned long idx)
{
	struct address_space *mapping = folio_mapping(folio);
	int mapcount = atomic_read(&page->_mapcount);
	char *type = "";

	mapcount = page_mapcount_is_type(mapcount) ? 0 : mapcount + 1;
	pr_warn("page: refcount:%d mapcount:%d mapping:%p index:%#lx pfn:%#lx\n",
			folio_ref_count(folio), mapcount, mapping,
			folio->index + idx, pfn);
	if (folio_test_large(folio)) {
		pr_warn("head: order:%u mapcount:%d entire_mapcount:%d nr_pages_mapped:%d pincount:%d\n",
				folio_order(folio),
				folio_mapcount(folio),
				folio_entire_mapcount(folio),
				folio_nr_pages_mapped(folio),
				atomic_read(&folio->_pincount));
	}

#ifdef CONFIG_MEMCG
	if (folio->memcg_data)
		pr_warn("memcg:%lx\n", folio->memcg_data);
#endif
	if (folio_test_ksm(folio))
		type = "ksm ";
	else if (folio_test_anon(folio))
		type = "anon ";
	else if (mapping)
		dump_mapping(mapping);
	BUILD_BUG_ON(ARRAY_SIZE(pageflag_names) != __NR_PAGEFLAGS + 1);

	/*
	 * Accessing the pageblock without the zone lock. It could change to
	 * "isolate" again in the meantime, but since we are just dumping the
	 * state for debugging, it should be fine to accept a bit of
	 * inaccuracy here due to racing.
	 */
	pr_warn("%sflags: %pGp%s\n", type, &folio->flags,
		is_migrate_cma_folio(folio, pfn) ? " CMA" : "");
	if (page_has_type(&folio->page))
		pr_warn("page_type: %x(%s)\n", folio->page.page_type >> 24,
				page_type_name(folio->page.page_type));

	print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32,
			sizeof(unsigned long), page,
			sizeof(struct page), false);
	if (folio_test_large(folio))
		print_hex_dump(KERN_WARNING, "head: ", DUMP_PREFIX_NONE, 32,
			sizeof(unsigned long), folio,
			2 * sizeof(struct page), false);
}

static void __dump_page(const struct page *page)
{
	struct folio *foliop, folio;
	struct page precise;
	unsigned long pfn = page_to_pfn(page);
	unsigned long idx, nr_pages = 1;
	int loops = 5;

again:
	memcpy(&precise, page, sizeof(*page));
	foliop = page_folio(&precise);
	if (foliop == (struct folio *)&precise) {
		idx = 0;
		if (!folio_test_large(foliop))
			goto dump;
		foliop = (struct folio *)page;
	} else {
		idx = folio_page_idx(foliop, page);
	}

	if (idx < MAX_FOLIO_NR_PAGES) {
		memcpy(&folio, foliop, 2 * sizeof(struct page));
		nr_pages = folio_nr_pages(&folio);
		foliop = &folio;
	}

	if (idx > nr_pages) {
		if (loops-- > 0)
			goto again;
		pr_warn("page does not match folio\n");
		precise.compound_head &= ~1UL;
		foliop = (struct folio *)&precise;
		idx = 0;
	}

dump:
	__dump_folio(foliop, &precise, pfn, idx);
}

void dump_page(const struct page *page, const char *reason)
{
	if (PagePoisoned(page))
		pr_warn("page:%p is uninitialized and poisoned", page);
	else
		__dump_page(page);
	if (reason)
		pr_warn("page dumped because: %s\n", reason);
	dump_page_owner(page);
}
EXPORT_SYMBOL(dump_page);

#ifdef CONFIG_DEBUG_VM

void dump_vma(const struct vm_area_struct *vma)
{
	pr_emerg("vma %px start %px end %px mm %px\n"
		"prot %lx anon_vma %px vm_ops %px\n"
		"pgoff %lx file %px private_data %px\n"
		"flags: %#lx(%pGv)\n",
		vma, (void *)vma->vm_start, (void *)vma->vm_end, vma->vm_mm,
		(unsigned long)pgprot_val(vma->vm_page_prot),
		vma->anon_vma, vma->vm_ops, vma->vm_pgoff,
		vma->vm_file, vma->vm_private_data,
		vma->vm_flags, &vma->vm_flags);
}
EXPORT_SYMBOL(dump_vma);

void dump_mm(const struct mm_struct *mm)
{
	pr_emerg("mm %px task_size %lu\n"
		"mmap_base %lu mmap_legacy_base %lu\n"
		"pgd %px mm_users %d mm_count %d pgtables_bytes %lu map_count %d\n"
		"hiwater_rss %lx hiwater_vm %lx total_vm %lx locked_vm %lx\n"
		"pinned_vm %llx data_vm %lx exec_vm %lx stack_vm %lx\n"
		"start_code %lx end_code %lx start_data %lx end_data %lx\n"
		"start_brk %lx brk %lx start_stack %lx\n"
		"arg_start %lx arg_end %lx env_start %lx env_end %lx\n"
		"binfmt %px flags %lx\n"
#ifdef CONFIG_AIO
		"ioctx_table %px\n"
#endif
#ifdef CONFIG_MEMCG
		"owner %px "
#endif
		"exe_file %px\n"
#ifdef CONFIG_MMU_NOTIFIER
		"notifier_subscriptions %px\n"
#endif
#ifdef CONFIG_NUMA_BALANCING
		"numa_next_scan %lu numa_scan_offset %lu numa_scan_seq %d\n"
#endif
		"tlb_flush_pending %d\n"
		"def_flags: %#lx(%pGv)\n",

		mm, mm->task_size,
		mm->mmap_base, mm->mmap_legacy_base,
		mm->pgd, atomic_read(&mm->mm_users),
		atomic_read(&mm->mm_count),
		mm_pgtables_bytes(mm),
		mm->map_count,
		mm->hiwater_rss, mm->hiwater_vm, mm->total_vm, mm->locked_vm,
		(u64)atomic64_read(&mm->pinned_vm),
		mm->data_vm, mm->exec_vm, mm->stack_vm,
		mm->start_code, mm->end_code, mm->start_data, mm->end_data,
		mm->start_brk, mm->brk, mm->start_stack,
		mm->arg_start, mm->arg_end, mm->env_start, mm->env_end,
		mm->binfmt, mm->flags,
#ifdef CONFIG_AIO
		mm->ioctx_table,
#endif
#ifdef CONFIG_MEMCG
		mm->owner,
#endif
		mm->exe_file,
#ifdef CONFIG_MMU_NOTIFIER
		mm->notifier_subscriptions,
#endif
#ifdef CONFIG_NUMA_BALANCING
		mm->numa_next_scan, mm->numa_scan_offset, mm->numa_scan_seq,
#endif
		atomic_read(&mm->tlb_flush_pending),
		mm->def_flags, &mm->def_flags
	);
}
EXPORT_SYMBOL(dump_mm);

static bool page_init_poisoning __read_mostly = true;

static int __init setup_vm_debug(char *str)
{
	bool __page_init_poisoning = true;

	/*
	 * Calling vm_debug with no arguments is equivalent to requesting
	 * to enable all debugging options we can control.
	 */
	if (*str++ != '=' || !*str)
		goto out;

	__page_init_poisoning = false;
	if (*str == '-')
		goto out;

	while (*str) {
		switch (tolower(*str)) {
		case'p':
			__page_init_poisoning = true;
			break;
		default:
			pr_err("vm_debug option '%c' unknown. skipped\n",
			       *str);
		}

		str++;
	}
out:
	if (page_init_poisoning && !__page_init_poisoning)
		pr_warn("Page struct poisoning disabled by kernel command line option 'vm_debug'\n");

	page_init_poisoning = __page_init_poisoning;

	return 1;
}
__setup("vm_debug", setup_vm_debug);

void page_init_poison(struct page *page, size_t size)
{
	if (page_init_poisoning)
		memset(page, PAGE_POISON_PATTERN, size);
}

void vma_iter_dump_tree(const struct vma_iterator *vmi)
{
#if defined(CONFIG_DEBUG_VM_MAPLE_TREE)
	mas_dump(&vmi->mas);
	mt_dump(vmi->mas.tree, mt_dump_hex);
#endif	/* CONFIG_DEBUG_VM_MAPLE_TREE */
}

#endif		/* CONFIG_DEBUG_VM */
Commit	Line	Data
b2441318	1	// SPDX-License-Identifier: GPL-2.0
31c9afa6 SL	2	/*
	3	* mm/debug.c
	4	*
	5	* mm/ specific debug routines.
	6	*
	7	*/
	8
82742a3a SL	9	#include <linux/kernel.h>
82742a3a SL	10	#include <linux/mm.h>
af658dca	11	#include <linux/trace_events.h>
82742a3a	12	#include <linux/memcontrol.h>
420adbe9	13	#include <trace/events/mmflags.h>
7cd12b4a	14	#include <linux/migrate.h>
4e462112	15	#include <linux/page_owner.h>
f682a97a	16	#include <linux/ctype.h>
82742a3a	17
edf14cdb	18	#include "internal.h"
8eb42bea JH	19	#include <trace/events/migrate.h>
	20
	21	/*
	22	* Define EM() and EMe() so that MIGRATE_REASON from trace/events/migrate.h can
	23	* be used to populate migrate_reason_names[].
	24	*/
	25	#undef EM
	26	#undef EMe
	27	#define EM(a, b) b,
	28	#define EMe(a, b) b
edf14cdb	29
9a2f45ff	30	const char *migrate_reason_names[MR_TYPES] = {
8eb42bea	31	MIGRATE_REASON
7cd12b4a VB	32	};
7cd12b4a VB	33
edf14cdb VB	34	const struct trace_print_flags pageflag_names[] = {
	35	__def_pageflag_names,
	36	{0, NULL}
	37	};
	38
	39	const struct trace_print_flags gfpflag_names[] = {
	40	__def_gfpflag_names,
	41	{0, NULL}
420adbe9 VB	42	};
420adbe9 VB	43
edf14cdb VB	44	const struct trace_print_flags vmaflag_names[] = {
	45	__def_vmaflag_names,
	46	{0, NULL}
82742a3a SL	47	};
82742a3a SL	48
4ffca5a9 MWO	49	#define DEF_PAGETYPE_NAME(_name) [PGTY_##_name - 0xf0] = __stringify(_name)
	50
	51	static const char *page_type_names[] = {
	52	DEF_PAGETYPE_NAME(slab),
	53	DEF_PAGETYPE_NAME(hugetlb),
	54	DEF_PAGETYPE_NAME(offline),
	55	DEF_PAGETYPE_NAME(guard),
	56	DEF_PAGETYPE_NAME(table),
	57	DEF_PAGETYPE_NAME(buddy),
	58	DEF_PAGETYPE_NAME(unaccepted),
	59	};
	60
	61	static const char *page_type_name(unsigned int page_type)
	62	{
	63	unsigned i = (page_type >> 24) - 0xf0;
	64
	65	if (i >= ARRAY_SIZE(page_type_names))
	66	return "unknown";
	67	return page_type_names[i];
	68	}
	69
fae7d834 MWO	70	static void __dump_folio(struct folio folio, struct page page,
fae7d834 MWO	71	unsigned long pfn, unsigned long idx)
82742a3a	72	{
fae7d834	73	struct address_space *mapping = folio_mapping(folio);
7441d349	74	int mapcount = atomic_read(&page->_mapcount);
5b57b8f2	75	char *type = "";
fc36def9	76
4ffca5a9	77	mapcount = page_mapcount_is_type(mapcount) ? 0 : mapcount + 1;
fae7d834 MWO	78	pr_warn("page: refcount:%d mapcount:%d mapping:%p index:%#lx pfn:%#lx\n",
	79	folio_ref_count(folio), mapcount, mapping,
	80	folio->index + idx, pfn);
	81	if (folio_test_large(folio)) {
05c5323b	82	pr_warn("head: order:%u mapcount:%d entire_mapcount:%d nr_pages_mapped:%d pincount:%d\n",
fae7d834	83	folio_order(folio),
05c5323b	84	folio_mapcount(folio),
91ec7f28	85	folio_entire_mapcount(folio),
eec20426	86	folio_nr_pages_mapped(folio),
94688e8e	87	atomic_read(&folio->_pincount));
452b557c	88	}
91f5345a MWO	89
91f5345a MWO	90	#ifdef CONFIG_MEMCG
fae7d834 MWO	91	if (folio->memcg_data)
fae7d834 MWO	92	pr_warn("memcg:%lx\n", folio->memcg_data);
91f5345a	93	#endif
fae7d834	94	if (folio_test_ksm(folio))
5b57b8f2	95	type = "ksm ";
fae7d834	96	else if (folio_test_anon(folio))
5b57b8f2	97	type = "anon ";
3e9d80a8 MWO	98	else if (mapping)
3e9d80a8 MWO	99	dump_mapping(mapping);
edf14cdb	100	BUILD_BUG_ON(ARRAY_SIZE(pageflag_names) != __NR_PAGEFLAGS + 1);
ff8e8116	101
fae7d834 MWO	102	/*
	103	* Accessing the pageblock without the zone lock. It could change to
	104	* "isolate" again in the meantime, but since we are just dumping the
	105	* state for debugging, it should be fine to accept a bit of
	106	* inaccuracy here due to racing.
	107	*/
	108	pr_warn("%sflags: %pGp%s\n", type, &folio->flags,
	109	is_migrate_cma_folio(folio, pfn) ? " CMA" : "");
8f790d0c	110	if (page_has_type(&folio->page))
4ffca5a9 MWO	111	pr_warn("page_type: %x(%s)\n", folio->page.page_type >> 24,
4ffca5a9 MWO	112	page_type_name(folio->page.page_type));
f2421a16	113
e0392cf7	114	print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32,
46e8a3a0 VB	115	sizeof(unsigned long), page,
46e8a3a0 VB	116	sizeof(struct page), false);
fae7d834	117	if (folio_test_large(folio))
6197ab98	118	print_hex_dump(KERN_WARNING, "head: ", DUMP_PREFIX_NONE, 32,
fae7d834 MWO	119	sizeof(unsigned long), folio,
	120	2 * sizeof(struct page), false);
	121	}
	122
	123	static void __dump_page(const struct page *page)
	124	{
	125	struct folio *foliop, folio;
	126	struct page precise;
	127	unsigned long pfn = page_to_pfn(page);
	128	unsigned long idx, nr_pages = 1;
	129	int loops = 5;
	130
	131	again:
	132	memcpy(&precise, page, sizeof(*page));
	133	foliop = page_folio(&precise);
	134	if (foliop == (struct folio *)&precise) {
	135	idx = 0;
	136	if (!folio_test_large(foliop))
	137	goto dump;
	138	foliop = (struct folio *)page;
	139	} else {
	140	idx = folio_page_idx(foliop, page);
	141	}
	142
	143	if (idx < MAX_FOLIO_NR_PAGES) {
	144	memcpy(&folio, foliop, 2 * sizeof(struct page));
	145	nr_pages = folio_nr_pages(&folio);
	146	foliop = &folio;
	147	}
	148
	149	if (idx > nr_pages) {
	150	if (loops-- > 0)
	151	goto again;
	152	pr_warn("page does not match folio\n");
	153	precise.compound_head &= ~1UL;
	154	foliop = (struct folio *)&precise;
	155	idx = 0;
	156	}
	157
	158	dump:
	159	__dump_folio(foliop, &precise, pfn, idx);
82742a3a SL	160	}
82742a3a SL	161
b3a32033	162	void dump_page(const struct page page, const char reason)
82742a3a	163	{
be7c701f MWO	164	if (PagePoisoned(page))
	165	pr_warn("page:%p is uninitialized and poisoned", page);
	166	else
	167	__dump_page(page);
	168	if (reason)
	169	pr_warn("page dumped because: %s\n", reason);
4e462112	170	dump_page_owner(page);
82742a3a SL	171	}
	172	EXPORT_SYMBOL(dump_page);
	173
	174	#ifdef CONFIG_DEBUG_VM
	175
82742a3a SL	176	void dump_vma(const struct vm_area_struct *vma)
82742a3a SL	177	{
763ecb03	178	pr_emerg("vma %px start %px end %px mm %px\n"
152a2d19 MW	179	"prot %lx anon_vma %px vm_ops %px\n"
152a2d19 MW	180	"pgoff %lx file %px private_data %px\n"
b8eceeb9	181	"flags: %#lx(%pGv)\n",
763ecb03	182	vma, (void )vma->vm_start, (void )vma->vm_end, vma->vm_mm,
82742a3a SL	183	(unsigned long)pgprot_val(vma->vm_page_prot),
82742a3a SL	184	vma->anon_vma, vma->vm_ops, vma->vm_pgoff,
b8eceeb9 VB	185	vma->vm_file, vma->vm_private_data,
b8eceeb9 VB	186	vma->vm_flags, &vma->vm_flags);
82742a3a SL	187	}
	188	EXPORT_SYMBOL(dump_vma);
	189
31c9afa6 SL	190	void dump_mm(const struct mm_struct *mm)
31c9afa6 SL	191	{
763ecb03	192	pr_emerg("mm %px task_size %lu\n"
763ecb03	193	"mmap_base %lu mmap_legacy_base %lu\n"
152a2d19	194	"pgd %px mm_users %d mm_count %d pgtables_bytes %lu map_count %d\n"
31c9afa6	195	"hiwater_rss %lx hiwater_vm %lx total_vm %lx locked_vm %lx\n"
70f8a3ca	196	"pinned_vm %llx data_vm %lx exec_vm %lx stack_vm %lx\n"
31c9afa6 SL	197	"start_code %lx end_code %lx start_data %lx end_data %lx\n"
	198	"start_brk %lx brk %lx start_stack %lx\n"
	199	"arg_start %lx arg_end %lx env_start %lx env_end %lx\n"
0258b5fd	200	"binfmt %px flags %lx\n"
31c9afa6	201	#ifdef CONFIG_AIO
152a2d19	202	"ioctx_table %px\n"
31c9afa6 SL	203	#endif
31c9afa6 SL	204	#ifdef CONFIG_MEMCG
152a2d19	205	"owner %px "
31c9afa6	206	#endif
152a2d19	207	"exe_file %px\n"
31c9afa6	208	#ifdef CONFIG_MMU_NOTIFIER
984cfe4e	209	"notifier_subscriptions %px\n"
31c9afa6 SL	210	#endif
	211	#ifdef CONFIG_NUMA_BALANCING
	212	"numa_next_scan %lu numa_scan_offset %lu numa_scan_seq %d\n"
	213	#endif
31c9afa6	214	"tlb_flush_pending %d\n"
b8eceeb9	215	"def_flags: %#lx(%pGv)\n",
31c9afa6	216
763ecb03	217	mm, mm->task_size,
763ecb03	218	mm->mmap_base, mm->mmap_legacy_base,
31c9afa6 SL	219	mm->pgd, atomic_read(&mm->mm_users),
31c9afa6 SL	220	atomic_read(&mm->mm_count),
af5b0f6a	221	mm_pgtables_bytes(mm),
31c9afa6 SL	222	mm->map_count,
31c9afa6 SL	223	mm->hiwater_rss, mm->hiwater_vm, mm->total_vm, mm->locked_vm,
44dc1b1f	224	(u64)atomic64_read(&mm->pinned_vm),
70f8a3ca	225	mm->data_vm, mm->exec_vm, mm->stack_vm,
31c9afa6 SL	226	mm->start_code, mm->end_code, mm->start_data, mm->end_data,
	227	mm->start_brk, mm->brk, mm->start_stack,
	228	mm->arg_start, mm->arg_end, mm->env_start, mm->env_end,
0258b5fd	229	mm->binfmt, mm->flags,
31c9afa6 SL	230	#ifdef CONFIG_AIO
	231	mm->ioctx_table,
	232	#endif
	233	#ifdef CONFIG_MEMCG
	234	mm->owner,
	235	#endif
	236	mm->exe_file,
	237	#ifdef CONFIG_MMU_NOTIFIER
984cfe4e	238	mm->notifier_subscriptions,
31c9afa6 SL	239	#endif
	240	#ifdef CONFIG_NUMA_BALANCING
	241	mm->numa_next_scan, mm->numa_scan_offset, mm->numa_scan_seq,
	242	#endif
16af97dc	243	atomic_read(&mm->tlb_flush_pending),
b8eceeb9 VB	244	mm->def_flags, &mm->def_flags
b8eceeb9 VB	245	);
31c9afa6	246	}
c2fdc235	247	EXPORT_SYMBOL(dump_mm);
31c9afa6	248
f682a97a AD	249	static bool page_init_poisoning __read_mostly = true;
	250
	251	static int __init setup_vm_debug(char *str)
	252	{
	253	bool __page_init_poisoning = true;
	254
	255	/*
	256	* Calling vm_debug with no arguments is equivalent to requesting
	257	* to enable all debugging options we can control.
	258	*/
	259	if (str++ != '=' \|\| !str)
	260	goto out;
	261
	262	__page_init_poisoning = false;
	263	if (*str == '-')
	264	goto out;
	265
	266	while (*str) {
	267	switch (tolower(*str)) {
	268	case'p':
	269	__page_init_poisoning = true;
	270	break;
	271	default:
	272	pr_err("vm_debug option '%c' unknown. skipped\n",
	273	*str);
	274	}
	275
	276	str++;
	277	}
	278	out:
	279	if (page_init_poisoning && !__page_init_poisoning)
	280	pr_warn("Page struct poisoning disabled by kernel command line option 'vm_debug'\n");
	281
	282	page_init_poisoning = __page_init_poisoning;
	283
	284	return 1;
	285	}
	286	__setup("vm_debug", setup_vm_debug);
	287
	288	void page_init_poison(struct page *page, size_t size)
	289	{
	290	if (page_init_poisoning)
	291	memset(page, PAGE_POISON_PATTERN, size);
	292	}
b50e195f LH	293
	294	void vma_iter_dump_tree(const struct vma_iterator *vmi)
	295	{
	296	#if defined(CONFIG_DEBUG_VM_MAPLE_TREE)
	297	mas_dump(&vmi->mas);
	298	mt_dump(vmi->mas.tree, mt_dump_hex);
	299	#endif /* CONFIG_DEBUG_VM_MAPLE_TREE */
	300	}
	301
82742a3a	302	#endif /* CONFIG_DEBUG_VM */