[thirdparty/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 pagetype_names[] = {
	__def_pagetype_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}
};

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 = 0;
	char *type = "";

	/*
	 * page->_mapcount space in struct page is used by slab pages to
	 * encode own info, and we must avoid calling page_folio() again.
	 */
	if (!folio_test_slab(folio)) {
		mapcount = atomic_read(&page->_mapcount) + 1;
		if (folio_test_large(folio))
			mapcount += folio_entire_mapcount(folio);
	}

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