The Linux Kernel Mailing List
 help / color / mirror / Atom feed
From: haoran.jiang@linux.dev
To: "Huacai Chen" <chenhuacai@kernel.org>
Cc: loongarch@lists.linux.dev, linux-kernel@vger.kernel.org,
	kernel@xen0n.name, yangtiezhu@loongson.cn,
	"Haoran Jiang" <jianghaoran@kylinos.cn>,
	"Xiujie Jiang" <jiangxiujie@kylinos.cn>
Subject: Re: [PATCH] LoongArch: Add support to dump the kernel page tables
Date: Tue, 30 Jun 2026 12:09:59 +0000	[thread overview]
Message-ID: <c26214ada1e531dc7fabb51c65a19b3fbf60a997@linux.dev> (raw)
In-Reply-To: <CAAhV-H6MTP_eU-8CeHzChvBGmjxP3KFcZDMypiFH4EsF4F3u_A@mail.gmail.com>

2026年6月29日 22:00, "Huacai Chen" <chenhuacai@kernel.org mailto:chenhuacai@kernel.org?to=%22Huacai%20Chen%22%20%3Cchenhuacai%40kernel.org%3E > 写到:


> 
> Hi, Haoran,
> 
> On Mon, Jun 29, 2026 at 3:44 PM <haoran.jiang@linux.dev> wrote:
> 
> > 
> > From: Haoran Jiang <jianghaoran@kylinos.cn>
> > 
> >  In a similar manner to riscv,arm64,x86 etc.,this patch allows
> >  dumping the page tables of the LoongArch page-mapped virtual memory
> >  region via a debugfs file, which is useful for kernel developers to
> >  inspect page table layouts and verify permissions and type settings.
> > 
> I like the RISC-V version that put everything in ptdump.c.
> 
> Huacai
> 
1,The ptdump functionality can be enabled via either CONFIG_PTDUMP_DEBUGFS
or CONFIG_DEBUG_WX. For example, RISC-V enables it through DEBUG_WX,
while other architectures support both options.

CONFIG_DEBUG_WX performs a kernel page table permission check at boot
time to enforce the W^X principle. It depends on CONFIG_STRICT_KERNEL_RWX;
debug_checkwx() is only invoked when CONFIG_STRICT_KERNEL_RWX is enabled.

However, on LoongArch, the kernel text segment resides in the direct
mapping region, where permission control cannot be enforced via page
tables. Therefore, CONFIG_STRICT_KERNEL_RWX is not necessary to
implement, and enabling DEBUG_WX is also unnecessary. Instead, this
patch adopts CONFIG_PTDUMP_DEBUGFS to enable the ptdump feature for
LoongArch.


2,However, for page-mapped regions, the page table permissions should conform to the W^X principle. The MODULES region can meet this requirement via `CONFIG_STRICT_MODULE_RWX`.But other regions, such as vmalloc, currently still do not comply with the W^X principle (as can be verified by checking `cat /sys/kernel/debug/check_wx_pages`). I intend to add `PAGE_NO_EXEC` to the `PAGE_KERNEL` definition。

 #define PAGE_KERNEL    __pgprot(_PAGE_PRESENT | __READABLE | __WRITEABLE | \___

__ PAGE_GLOBAL | _PAGE_KERN | _CACHE_CC | _PAGE_NO_EXEC)


 #define PAGE_KERNEL_SUC __pgprot(_PAGE_PRESENT | __READABLE | __WRITEABLE | \

_PAGE_GLOBAL | _PAGE_KERN |  _CACHE_SUC | _PAGE_NO_EXEC)


 #define PAGE_KERNEL_WUC __pgprot(_PAGE_PRESENT | __READABLE | __WRITEABLE | \

PAGE_GLOBAL | _PAGE_KERN |  _CACHE_WUC | _PAGE_NO_EXEC)


eBPF memory allocated from the MODULES region is not affected by this change, because it already gets executable permissions via `set_memory_rox`.

The vmalloc, PCI I/O, vmemmap, KFENCE, KASAN, and Fixmap regions should all be non-executable by default.

Would this modification have any other side effects?


thanks!

> > 
> > Co-developed-by: Xiujie Jiang <jiangxiujie@kylinos.cn>
> >  Signed-off-by: Xiujie Jiang <jiangxiujie@kylinos.cn>
> >  Signed-off-by: Haoran Jiang <jianghaoran@kylinos.cn>
> >  ---
> >  arch/loongarch/Kconfig | 1 +
> >  arch/loongarch/include/asm/ptdump.h | 61 ++++
> >  arch/loongarch/mm/Makefile | 2 +
> >  arch/loongarch/mm/ptdump.c | 417 ++++++++++++++++++++++++++++
> >  arch/loongarch/mm/ptdump_debugfs.c | 22 ++
> >  5 files changed, 503 insertions(+)
> >  create mode 100644 arch/loongarch/include/asm/ptdump.h
> >  create mode 100644 arch/loongarch/mm/ptdump.c
> >  create mode 100644 arch/loongarch/mm/ptdump_debugfs.c
> > 
> >  diff --git a/arch/loongarch/Kconfig b/arch/loongarch/Kconfig
> >  index bec7cc1d72ee..e97fe97e1955 100644
> >  --- a/arch/loongarch/Kconfig
> >  +++ b/arch/loongarch/Kconfig
> >  @@ -25,6 +25,7 @@ config LOONGARCH
> >  select ARCH_HAS_NON_OVERLAPPING_ADDRESS_SPACE
> >  select ARCH_HAS_PREEMPT_LAZY
> >  select ARCH_HAS_PTE_SPECIAL if 64BIT
> >  + select ARCH_HAS_PTDUMP
> >  select ARCH_HAS_SET_MEMORY
> >  select ARCH_HAS_SET_DIRECT_MAP
> >  select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
> >  diff --git a/arch/loongarch/include/asm/ptdump.h b/arch/loongarch/include/asm/ptdump.h
> >  new file mode 100644
> >  index 000000000000..4590ea56abe6
> >  --- /dev/null
> >  +++ b/arch/loongarch/include/asm/ptdump.h
> >  @@ -0,0 +1,61 @@
> >  +/* SPDX-License-Identifier: GPL-2.0-only */
> >  +
> >  +#ifndef __ASM_PTDUMP_H
> >  +#define __ASM_PTDUMP_H
> >  +
> >  +#include <linux/ptdump.h>
> >  +#include <linux/mm_types.h>
> >  +#include <linux/seq_file.h>
> >  +
> >  +/*
> >  + * The page dumper groups page table entries of the same type into a single
> >  + * description. It uses pg_state to track the range information while
> >  + * iterating over the pte entries. When the continuity is broken it then
> >  + * dumps out a description of the range.
> >  + */
> >  +struct pg_state {
> >  + struct ptdump_state ptdump;
> >  + struct seq_file *seq;
> >  + const struct addr_marker *marker;
> >  + unsigned long start_address;
> >  + unsigned long start_pa;
> >  + unsigned long last_pa;
> >  + int level;
> >  + u64 current_prot;
> >  + bool check_wx;
> >  + unsigned long wx_pages;
> >  +};
> >  +
> >  +/* Address marker */
> >  +struct addr_marker {
> >  + unsigned long start_address;
> >  + const char *name;
> >  +};
> >  +
> >  +/* Private information for debugfs */
> >  +struct ptd_mm_info {
> >  + struct mm_struct *mm;
> >  + const struct addr_marker *markers;
> >  + unsigned long base_addr;
> >  + unsigned long end;
> >  +};
> >  +
> >  +/* Page Table Entry */
> >  +struct prot_bits {
> >  + u64 mask;
> >  + u64 val;
> >  + const char *set;
> >  + const char *clear;
> >  +};
> >  +
> >  +/* Page Level */
> >  +struct pg_level {
> >  + const char *name;
> >  + u64 mask;
> >  +};
> >  +
> >  +void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo);
> >  +void __init ptdump_debugfs_register(struct ptd_mm_info *info, const char *name);
> >  +
> >  +#endif /* __ASM_PTDUMP_H */
> >  +
> >  diff --git a/arch/loongarch/mm/Makefile b/arch/loongarch/mm/Makefile
> >  index 2aae3773de77..da9e6fee0439 100644
> >  --- a/arch/loongarch/mm/Makefile
> >  +++ b/arch/loongarch/mm/Makefile
> >  @@ -10,5 +10,7 @@ obj-y += init.o cache.o tlb.o tlbex.o extable.o \
> >  obj-$(CONFIG_HIGHMEM) += highmem.o
> >  obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
> >  obj-$(CONFIG_KASAN) += kasan_init.o
> >  +obj-$(CONFIG_PTDUMP) += ptdump.o
> >  +obj-$(CONFIG_PTDUMP_DEBUGFS) += ptdump_debugfs.o
> > 
> >  KASAN_SANITIZE_kasan_init.o := n
> >  diff --git a/arch/loongarch/mm/ptdump.c b/arch/loongarch/mm/ptdump.c
> >  new file mode 100644
> >  index 000000000000..cab6dad18150
> >  --- /dev/null
> >  +++ b/arch/loongarch/mm/ptdump.c
> >  @@ -0,0 +1,417 @@
> >  +// SPDX-License-Identifier: GPL-2.0-only
> >  +/*
> >  + * Derived from riscv implementation
> >  + */
> >  +
> >  +#include <linux/init.h>
> >  +#include <linux/debugfs.h>
> >  +#include <linux/seq_file.h>
> >  +#include <linux/ptdump.h>
> >  +#include <linux/pgtable.h>
> >  +#include <asm/kasan.h>
> >  +#include <asm/ptdump.h>
> >  +
> >  +#define _PAGE_PLV_VAL 0x0
> >  +
> >  +#define pt_dump_seq_printf(m, fmt, args...) \
> >  +({ \
> >  + if (m) \
> >  + seq_printf(m, fmt, ##args); \
> >  +})
> >  +
> >  +#define pt_dump_seq_puts(m, fmt) \
> >  +({ \
> >  + if (m) \
> >  + seq_puts(m, fmt); \
> >  +})
> >  +
> >  +enum address_markers_idx {
> >  + PCI_IO_START_NR,
> >  + PCI_IO_END_NR,
> >  + MODULES_START_NR,
> >  + MODULES_END_NR,
> >  + VMALLOC_START_NR,
> >  + VMALLOC_END_NR,
> >  +#ifdef CONFIG_SPARSEMEM_VMEMMAP
> >  + VMEMMAP_START_NR,
> >  + VMEMMAP_END_NR,
> >  +#endif
> >  +#ifdef CONFIG_KFENCE
> >  + KFENCE_AREA_START_NR,
> >  + KFENCE_AREA_END_NR,
> >  +#endif
> >  +#ifdef CONFIG_KASAN
> >  + KASAN_SHADOW_START_NR,
> >  + KASAN_SHADOW_END_NR,
> >  +#endif
> >  + FIXMAP_START_NR,
> >  + FIXMAP_END_NR,
> >  + END_OF_SPACE_NR
> >  +};
> >  +
> >  +
> >  +
> >  +static struct addr_marker address_markers[] = {
> >  + {0, "PCI I/O start"},
> >  + {0, "PCI I/O end"},
> >  + {0, "modules start"},
> >  + {0, "modules end"},
> >  + {0, "vmalloc() area"},
> >  + {0, "vmalloc() end"},
> >  +#ifdef CONFIG_SPARSEMEM_VMEMMAP
> >  + {0, "vmemmap start"},
> >  + {0, "vmemmap end"},
> >  +#endif
> >  +#ifdef CONFIG_KFENCE
> >  + {0, "kfence area start"},
> >  + {0, "kfence area end"},
> >  +#endif
> >  +#ifdef CONFIG_KASAN
> >  + {0, "Kasan shadow start"},
> >  + {0, "Kasan shadow end"},
> >  +#endif
> >  + {0, "Fixmap start"},
> >  + {0, "Fixmap end"},
> >  + {-1, NULL},
> >  +};
> >  +
> >  +
> >  +static struct ptd_mm_info kernel_ptd_info = {
> >  + .mm = &init_mm,
> >  + .markers = address_markers,
> >  + .base_addr = 0,
> >  + .end = ULONG_MAX,
> >  +};
> >  +
> >  +static const struct prot_bits pte_bits[] = {
> >  + {
> >  + .mask = _PAGE_VALID,
> >  + .val = _PAGE_VALID,
> >  + .set = "V",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_DIRTY,
> >  + .val = _PAGE_DIRTY,
> >  + .set = "D",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_PLV,
> >  + .val = _PAGE_PLV_VAL,
> >  + .set = "KERN",
> >  + .clear = "USR ",
> >  + }, {
> >  + .mask = _CACHE_MASK,
> >  + .val = _CACHE_MASK,
> >  + .set = " ",
> >  + .clear = "SUC",
> >  + }, {
> >  + .mask = _PAGE_GLOBAL,
> >  + .val = _PAGE_GLOBAL,
> >  + .set = "G",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_PRESENT,
> >  + .val = _PAGE_PRESENT,
> >  + .set = "P",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_WRITE,
> >  + .val = _PAGE_WRITE,
> >  + .set = "W",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_HGLOBAL,
> >  + .val = _PAGE_HGLOBAL,
> >  + .set = "HG",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_NO_READ,
> >  + .val = _PAGE_NO_READ,
> >  + .set = "NR",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_NO_EXEC,
> >  + .val = _PAGE_NO_EXEC,
> >  + .set = "NX",
> >  + .clear = " ",
> >  + }, {
> >  + .mask = _PAGE_RPLV,
> >  + .val = _PAGE_RPLV,
> >  + .set = "RPLV",
> >  + .clear = " ",
> >  + }
> >  +};
> >  +
> >  +static struct pg_level pg_level[] = {
> >  + { /* pgd */
> >  + .name = "PGD",
> >  + }, { /* p4d */
> >  + .name = (CONFIG_PGTABLE_LEVELS > 4) ? "P4D" : "PGD",
> >  + }, { /* pud */
> >  + .name = (CONFIG_PGTABLE_LEVELS > 3) ? "PUD" : "PGD",
> >  + }, { /* pmd */
> >  + .name = (CONFIG_PGTABLE_LEVELS > 2) ? "PMD" : "PGD",
> >  + }, { /* pte */
> >  + .name = "PTE",
> >  + },
> >  +};
> >  +
> >  +static void dump_prot(struct pg_state *st)
> >  +{
> >  +
> >  + unsigned int i;
> >  +
> >  + for (i = 0; i < ARRAY_SIZE(pte_bits); i++) {
> >  + const char *s;
> >  +
> >  + if (pte_bits[i].mask == _CACHE_MASK) {
> >  + if ((st->current_prot & pte_bits[i].mask) == _CACHE_CC)
> >  + s = "CC ";
> >  + else if ((st->current_prot & pte_bits[i].mask) == _CACHE_WUC)
> >  + s = "WUC";
> >  + else
> >  + s = pte_bits[i].clear;
> >  + } else if (pte_bits[i].mask == _PAGE_GLOBAL) {
> >  + if ((st->current_prot & pte_bits[i].mask) == pte_bits[i].val) {
> >  + if (st->level != 4)
> >  + s = "PSE";
> >  + else
> >  + s = pte_bits[i].set;
> >  + } else {
> >  + s = pte_bits[i].clear;
> >  + }
> >  +
> >  + } else {
> >  + if ((st->current_prot & pte_bits[i].mask) == pte_bits[i].val)
> >  + s = pte_bits[i].set;
> >  + else
> >  + s = pte_bits[i].clear;
> >  + }
> >  +
> >  +
> >  + if (s)
> >  + pt_dump_seq_printf(st->seq, " %s", s);
> >  + }
> >  +
> >  +}
> >  +
> >  +
> >  +#ifdef CONFIG_64BIT
> >  +#define ADDR_FORMAT "0x%016lx"
> >  +#else
> >  +#define ADDR_FORMAT "0x%08lx"
> >  +#endif
> >  +static void dump_addr(struct pg_state *st, unsigned long addr)
> >  +{
> >  + static const char units[] = "KMGTPE";
> >  + const char *unit = units;
> >  + unsigned long delta;
> >  +
> >  + pt_dump_seq_printf(st->seq, ADDR_FORMAT "-" ADDR_FORMAT " ",
> >  + st->start_address, addr);
> >  +
> >  + pt_dump_seq_printf(st->seq, " " ADDR_FORMAT " ", st->start_pa);
> >  + delta = (addr - st->start_address) >> 10;
> >  +
> >  + while (!(delta & 1023) && unit[1]) {
> >  + delta >>= 10;
> >  + unit++;
> >  + }
> >  +
> >  + pt_dump_seq_printf(st->seq, "%9lu%c %s", delta, *unit,
> >  + pg_level[st->level].name);
> >  +}
> >  +
> >  +static void note_prot_wx(struct pg_state *st, unsigned long addr)
> >  +{
> >  + if (!st->check_wx)
> >  + return;
> >  + if ((st->current_prot & _PAGE_WRITE) == 0)
> >  + return;
> >  + if ((st->current_prot & _PAGE_NO_EXEC) == _PAGE_NO_EXEC)
> >  + return;
> >  +
> >  + st->wx_pages += (addr - st->start_address) / PAGE_SIZE;
> >  +
> >  +}
> >  +
> >  +static void note_page(struct ptdump_state *pt_st, unsigned long addr,
> >  + int level, u64 val)
> >  +{
> >  + struct pg_state *st = container_of(pt_st, struct pg_state, ptdump);
> >  + u64 pa = PFN_PHYS(pte_pfn(__pte(val)));
> >  + u64 prot = 0;
> >  +
> >  + if (level >= 0)
> >  + prot = val & pg_level[level].mask;
> >  +
> >  + if (st->level == -1) {
> >  + st->level = level;
> >  + st->current_prot = prot;
> >  + st->start_address = addr;
> >  + st->start_pa = pa;
> >  + st->last_pa = pa;
> >  + pt_dump_seq_printf(st->seq, "---[ %s ]---\n", st->marker->name);
> >  + } else if (prot != st->current_prot ||
> >  + level != st->level || addr >= st->marker[1].start_address) {
> >  + if (st->current_prot) {
> >  + note_prot_wx(st, addr);
> >  + dump_addr(st, addr);
> >  + dump_prot(st);
> >  + pt_dump_seq_puts(st->seq, "\n");
> >  + }
> >  +
> >  + while (addr >= st->marker[1].start_address) {
> >  + st->marker++;
> >  + pt_dump_seq_printf(st->seq, "---[ %s ]---\n",
> >  + st->marker->name);
> >  + }
> >  +
> >  + st->start_address = addr;
> >  + st->start_pa = pa;
> >  + st->last_pa = pa;
> >  + st->current_prot = prot;
> >  + st->level = level;
> >  + } else {
> >  + st->last_pa = pa;
> >  + }
> >  +}
> >  +
> >  +static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> >  +{
> >  + note_page(pt_st, addr, 4, pte_val(pte));
> >  +}
> >  +
> >  +static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> >  +{
> >  + note_page(pt_st, addr, 3, pmd_val(pmd));
> >  +}
> >  +
> >  +static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> >  +{
> >  + note_page(pt_st, addr, 2, pud_val(pud));
> >  +}
> >  +
> >  +static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> >  +{
> >  + note_page(pt_st, addr, 1, p4d_val(p4d));
> >  +}
> >  +
> >  +static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> >  +{
> >  + note_page(pt_st, addr, 0, pgd_val(pgd));
> >  +}
> >  +
> >  +static void note_page_flush(struct ptdump_state *pt_st)
> >  +{
> >  + pte_t pte_zero = {0};
> >  +
> >  + note_page(pt_st, 0, -1, pte_val(pte_zero));
> >  +}
> >  +
> >  +void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo)
> >  +{
> >  + struct pg_state st = {
> >  + .seq = s,
> >  + .marker = pinfo->markers,
> >  + .level = -1,
> >  + .ptdump = {
> >  + .note_page_pte = note_page_pte,
> >  + .note_page_pmd = note_page_pmd,
> >  + .note_page_pud = note_page_pud,
> >  + .note_page_p4d = note_page_p4d,
> >  + .note_page_pgd = note_page_pgd,
> >  + .note_page_flush = note_page_flush,
> >  + .range = (struct ptdump_range[]) {
> >  + {pinfo->base_addr, pinfo->end},
> >  + {0, 0}
> >  + }
> >  + }
> >  + };
> >  +
> >  + ptdump_walk_pgd(&st.ptdump, pinfo->mm, NULL);
> >  +}
> >  +
> >  +bool ptdump_check_wx(void)
> >  +{
> >  + struct pg_state st = {
> >  + .seq = NULL,
> >  + .marker = (struct addr_marker[]) {
> >  + {0, NULL},
> >  + {-1, NULL},
> >  + },
> >  + .level = -1,
> >  + .check_wx = true,
> >  + .ptdump = {
> >  + .note_page_pte = note_page_pte,
> >  + .note_page_pmd = note_page_pmd,
> >  + .note_page_pud = note_page_pud,
> >  + .note_page_p4d = note_page_p4d,
> >  + .note_page_pgd = note_page_pgd,
> >  + .note_page_flush = note_page_flush,
> >  + .range = (struct ptdump_range[]) {
> >  + {vm_map_base, ULONG_MAX},
> >  + {0, 0}
> >  + }
> >  + }
> >  + };
> >  +
> >  + ptdump_walk_pgd(&st.ptdump, &init_mm, NULL);
> >  +
> >  + if (st.wx_pages) {
> >  + pr_warn("Checked W+X mappings: failed, %lu W+X pages found\n",
> >  + st.wx_pages);
> >  +
> >  + return false;
> >  + }
> >  +
> >  + pr_info("Checked W+X mappings: passed, no W+X pages found\n");
> >  +
> >  + return true;
> >  +}
> >  +
> >  +
> >  +static int __init ptdump_init(void)
> >  +{
> >  + unsigned int i, j;
> >  +
> >  + address_markers[PCI_IO_START_NR].start_address = (unsigned long)PCI_IOBASE;
> >  + address_markers[PCI_IO_END_NR].start_address = (unsigned long)PCI_IOBASE + IO_SPACE_LIMIT;
> >  +
> >  + address_markers[MODULES_START_NR].start_address = MODULES_VADDR;
> >  + address_markers[MODULES_END_NR].start_address = MODULES_END;
> >  +
> >  + address_markers[VMALLOC_START_NR].start_address = VMALLOC_START;
> >  + address_markers[VMALLOC_END_NR].start_address = VMALLOC_END;
> >  +
> >  +#ifdef CONFIG_SPARSEMEM_VMEMMAP
> >  + address_markers[VMEMMAP_START_NR].start_address = (unsigned long)vmemmap;
> >  + address_markers[VMEMMAP_END_NR].start_address = VMEMMAP_END;
> >  +#endif
> >  +
> >  +#ifdef CONFIG_KFENCE
> >  + address_markers[KFENCE_AREA_START_NR].start_address = KFENCE_AREA_START;
> >  + address_markers[KFENCE_AREA_END_NR].start_address = KFENCE_AREA_END;
> >  +#endif
> >  +
> >  +#ifdef CONFIG_KASAN
> >  + address_markers[KASAN_SHADOW_START_NR].start_address = KASAN_SHADOW_START;
> >  + address_markers[KASAN_SHADOW_END_NR].start_address = KASAN_SHADOW_END;
> >  +#endif
> >  +
> >  + address_markers[FIXMAP_START_NR].start_address = FIXADDR_START;
> >  + address_markers[FIXMAP_END_NR].start_address = FIXADDR_TOP;
> >  +
> >  + kernel_ptd_info.base_addr = vm_map_base;
> >  +
> >  + for (i = 0; i < ARRAY_SIZE(pg_level); i++) {
> >  + for (j = 0; j < ARRAY_SIZE(pte_bits); j++)
> >  + pg_level[i].mask |= pte_bits[j].mask;
> >  + }
> >  +
> >  + ptdump_debugfs_register(&kernel_ptd_info, "kernel_page_tables");
> >  +
> >  + return 0;
> >  +}
> >  +
> >  +device_initcall(ptdump_init);
> >  diff --git a/arch/loongarch/mm/ptdump_debugfs.c b/arch/loongarch/mm/ptdump_debugfs.c
> >  new file mode 100644
> >  index 000000000000..aa6169f6aa3a
> >  --- /dev/null
> >  +++ b/arch/loongarch/mm/ptdump_debugfs.c
> >  @@ -0,0 +1,22 @@
> >  +// SPDX-License-Identifier: GPL-2.0
> >  +#include <linux/debugfs.h>
> >  +#include <linux/seq_file.h>
> >  +
> >  +#include <asm/ptdump.h>
> >  +#include <asm/pgtable.h>
> >  +#include <asm/kasan.h>
> >  +
> >  +static int ptdump_show(struct seq_file *m, void *v)
> >  +{
> >  + struct ptd_mm_info *info = m->private;
> >  +
> >  + ptdump_walk(m, info);
> >  + return 0;
> >  +}
> >  +DEFINE_SHOW_ATTRIBUTE(ptdump);
> >  +
> >  +void __init ptdump_debugfs_register(struct ptd_mm_info *info, const char *name)
> >  +{
> >  + debugfs_create_file(name, 0400, NULL, info, &ptdump_fops);
> >  +}
> >  +
> >  --
> >  2.43.0
> >
>

      reply	other threads:[~2026-06-30 12:10 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-29  7:43 [PATCH] LoongArch: Add support to dump the kernel page tables haoran.jiang
2026-06-29 14:00 ` Huacai Chen
2026-06-30 12:09   ` haoran.jiang [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=c26214ada1e531dc7fabb51c65a19b3fbf60a997@linux.dev \
    --to=haoran.jiang@linux.dev \
    --cc=chenhuacai@kernel.org \
    --cc=jianghaoran@kylinos.cn \
    --cc=jiangxiujie@kylinos.cn \
    --cc=kernel@xen0n.name \
    --cc=linux-kernel@vger.kernel.org \
    --cc=loongarch@lists.linux.dev \
    --cc=yangtiezhu@loongson.cn \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox