Merge branch 'for-next/coredump' into for-next/core

* for-next/coredump:
  arm64: Change elfcore for_each_mte_vma() to use VMA iterator
  arm64: mte: Document the core dump file format
  arm64: mte: Dump the MTE tags in the core file
  arm64: mte: Define the number of bytes for storing the tags in a page
  elf: Introduce the ARM MTE ELF segment type
  elfcore: Replace CONFIG_{IA64, UML} checks with a new option

Documentation/arm64/memory-tagging-extension.rst

@@ -213,6 +213,29 @@ address ABI control and MTE configuration of a process as per the
 Documentation/arm64/tagged-address-abi.rst and above. The corresponding
 ``regset`` is 1 element of 8 bytes (``sizeof(long))``).
 
+Core dump support
+-----------------
+
+The allocation tags for user memory mapped with ``PROT_MTE`` are dumped
+in the core file as additional ``PT_ARM_MEMTAG_MTE`` segments. The
+program header for such segment is defined as:
+
+:``p_type``: ``PT_ARM_MEMTAG_MTE``
+:``p_flags``: 0
+:``p_offset``: segment file offset
+:``p_vaddr``: segment virtual address, same as the corresponding
+  ``PT_LOAD`` segment
+:``p_paddr``: 0
+:``p_filesz``: segment size in file, calculated as ``p_mem_sz / 32``
+  (two 4-bit tags cover 32 bytes of memory)
+:``p_memsz``: segment size in memory, same as the corresponding
+  ``PT_LOAD`` segment
+:``p_align``: 0
+
+The tags are stored in the core file at ``p_offset`` as two 4-bit tags
+in a byte. With the tag granule of 16 bytes, a 4K page requires 128
+bytes in the core file.
+
 Example of correct usage
 ========================
 

arch/arm64/Kconfig

@@ -10,6 +10,7 @@ config ARM64
 	select ACPI_SPCR_TABLE if ACPI
 	select ACPI_PPTT if ACPI
 	select ARCH_HAS_DEBUG_WX
+	select ARCH_BINFMT_ELF_EXTRA_PHDRS
 	select ARCH_BINFMT_ELF_STATE
 	select ARCH_CORRECT_STACKTRACE_ON_KRETPROBE
 	select ARCH_ENABLE_HUGEPAGE_MIGRATION if HUGETLB_PAGE && MIGRATION

arch/arm64/include/asm/mte-def.h

@@ -11,6 +11,7 @@
 #define MTE_TAG_SHIFT		56
 #define MTE_TAG_SIZE		4
 #define MTE_TAG_MASK		GENMASK((MTE_TAG_SHIFT + (MTE_TAG_SIZE - 1)), MTE_TAG_SHIFT)
+#define MTE_PAGE_TAG_STORAGE	(MTE_GRANULES_PER_PAGE * MTE_TAG_SIZE / 8)
 
 #define __MTE_PREAMBLE		ARM64_ASM_PREAMBLE ".arch_extension memtag\n"
 

arch/arm64/kernel/Makefile

@@ -61,6 +61,7 @@ obj-$(CONFIG_ARM64_ACPI_PARKING_PROTOCOL)	+= acpi_parking_protocol.o
 obj-$(CONFIG_PARAVIRT)			+= paravirt.o
 obj-$(CONFIG_RANDOMIZE_BASE)		+= kaslr.o
 obj-$(CONFIG_HIBERNATION)		+= hibernate.o hibernate-asm.o
+obj-$(CONFIG_ELF_CORE)			+= elfcore.o
 obj-$(CONFIG_KEXEC_CORE)		+= machine_kexec.o relocate_kernel.o	\
 					   cpu-reset.o
 obj-$(CONFIG_KEXEC_FILE)		+= machine_kexec_file.o kexec_image.o

arch/arm64/kernel/elfcore.c

+// SPDX-License-Identifier: GPL-2.0-only
+
+#include <linux/coredump.h>
+#include <linux/elfcore.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+
+#include <asm/cpufeature.h>
+#include <asm/mte.h>
+
+#ifndef VMA_ITERATOR
+#define VMA_ITERATOR(name, mm, addr)	\
+	struct mm_struct *name = mm
+#define for_each_vma(vmi, vma)		\
+	for (vma = vmi->mmap; vma; vma = vma->vm_next)
+#endif
+
+#define for_each_mte_vma(vmi, vma)					\
+	if (system_supports_mte())					\
+		for_each_vma(vmi, vma)					\
+			if (vma->vm_flags & VM_MTE)
+
+static unsigned long mte_vma_tag_dump_size(struct vm_area_struct *vma)
+{
+	if (vma->vm_flags & VM_DONTDUMP)
+		return 0;
+
+	return vma_pages(vma) * MTE_PAGE_TAG_STORAGE;
+}
+
+/* Derived from dump_user_range(); start/end must be page-aligned */
+static int mte_dump_tag_range(struct coredump_params *cprm,
+			      unsigned long start, unsigned long end)
+{
+	unsigned long addr;
+
+	for (addr = start; addr < end; addr += PAGE_SIZE) {
+		char tags[MTE_PAGE_TAG_STORAGE];
+		struct page *page = get_dump_page(addr);
+
+		/*
+		 * get_dump_page() returns NULL when encountering an empty
+		 * page table entry that would otherwise have been filled with
+		 * the zero page. Skip the equivalent tag dump which would
+		 * have been all zeros.
+		 */
+		if (!page) {
+			dump_skip(cprm, MTE_PAGE_TAG_STORAGE);
+			continue;
+		}
+
+		/*
+		 * Pages mapped in user space as !pte_access_permitted() (e.g.
+		 * PROT_EXEC only) may not have the PG_mte_tagged flag set.
+		 */
+		if (!test_bit(PG_mte_tagged, &page->flags)) {
+			put_page(page);
+			dump_skip(cprm, MTE_PAGE_TAG_STORAGE);
+			continue;
+		}
+
+		mte_save_page_tags(page_address(page), tags);
+		put_page(page);
+		if (!dump_emit(cprm, tags, MTE_PAGE_TAG_STORAGE))
+			return 0;
+	}
+
+	return 1;
+}
+
+Elf_Half elf_core_extra_phdrs(void)
+{
+	struct vm_area_struct *vma;
+	int vma_count = 0;
+	VMA_ITERATOR(vmi, current->mm, 0);
+
+	for_each_mte_vma(vmi, vma)
+		vma_count++;
+
+	return vma_count;
+}
+
+int elf_core_write_extra_phdrs(struct coredump_params *cprm, loff_t offset)
+{
+	struct vm_area_struct *vma;
+	VMA_ITERATOR(vmi, current->mm, 0);
+
+	for_each_mte_vma(vmi, vma) {
+		struct elf_phdr phdr;
+
+		phdr.p_type = PT_ARM_MEMTAG_MTE;
+		phdr.p_offset = offset;
+		phdr.p_vaddr = vma->vm_start;
+		phdr.p_paddr = 0;
+		phdr.p_filesz = mte_vma_tag_dump_size(vma);
+		phdr.p_memsz = vma->vm_end - vma->vm_start;
+		offset += phdr.p_filesz;
+		phdr.p_flags = 0;
+		phdr.p_align = 0;
+
+		if (!dump_emit(cprm, &phdr, sizeof(phdr)))
+			return 0;
+	}
+
+	return 1;
+}
+
+size_t elf_core_extra_data_size(void)
+{
+	struct vm_area_struct *vma;
+	size_t data_size = 0;
+	VMA_ITERATOR(vmi, current->mm, 0);
+
+	for_each_mte_vma(vmi, vma)
+		data_size += mte_vma_tag_dump_size(vma);
+
+	return data_size;
+}
+
+int elf_core_write_extra_data(struct coredump_params *cprm)
+{
+	struct vm_area_struct *vma;
+	VMA_ITERATOR(vmi, current->mm, 0);
+
+	for_each_mte_vma(vmi, vma) {
+		if (vma->vm_flags & VM_DONTDUMP)
+			continue;
+
+		if (!mte_dump_tag_range(cprm, vma->vm_start, vma->vm_end))
+			return 0;
+	}
+
+	return 1;
+}

arch/arm64/lib/mte.S

@@ -134,7 +134,7 @@ SYM_FUNC_END(mte_copy_tags_to_user)
 /*
  * Save the tags in a page
  *   x0 - page address
- *   x1 - tag storage
+ *   x1 - tag storage, MTE_PAGE_TAG_STORAGE bytes
  */
 SYM_FUNC_START(mte_save_page_tags)
 	multitag_transfer_size x7, x5
@@ -158,7 +158,7 @@ SYM_FUNC_END(mte_save_page_tags)
 /*
  * Restore the tags in a page
  *   x0 - page address
- *   x1 - tag storage
+ *   x1 - tag storage, MTE_PAGE_TAG_STORAGE bytes
  */
 SYM_FUNC_START(mte_restore_page_tags)
 	multitag_transfer_size x7, x5

arch/arm64/mm/mteswap.c

@@ -12,7 +12,7 @@ static DEFINE_XARRAY(mte_pages);
 void *mte_allocate_tag_storage(void)
 {
 	/* tags granule is 16 bytes, 2 tags stored per byte */
-	return kmalloc(PAGE_SIZE / 16 / 2, GFP_KERNEL);
+	return kmalloc(MTE_PAGE_TAG_STORAGE, GFP_KERNEL);
 }
 
 void mte_free_tag_storage(char *storage)

arch/ia64/Kconfig

@@ -8,6 +8,7 @@ menu "Processor type and features"
 
 config IA64
 	bool
+	select ARCH_BINFMT_ELF_EXTRA_PHDRS
 	select ARCH_HAS_DMA_MARK_CLEAN
 	select ARCH_HAS_STRNCPY_FROM_USER
 	select ARCH_HAS_STRNLEN_USER

arch/x86/um/Kconfig

@@ -8,6 +8,7 @@ endmenu
 
 config UML_X86
 	def_bool y
+	select ARCH_BINFMT_ELF_EXTRA_PHDRS if X86_32
 
 config 64BIT
 	bool "64-bit kernel" if "$(SUBARCH)" = "x86"

fs/Kconfig.binfmt

@@ -36,6 +36,9 @@ config COMPAT_BINFMT_ELF
 config ARCH_BINFMT_ELF_STATE
 	bool
 
+config ARCH_BINFMT_ELF_EXTRA_PHDRS
+	bool
+
 config ARCH_HAVE_ELF_PROT
 	bool
 

include/linux/elfcore.h

@@ -114,7 +114,7 @@ static inline int elf_core_copy_task_fpregs(struct task_struct *t, struct pt_reg
 #endif
 }
 
-#if (defined(CONFIG_UML) && defined(CONFIG_X86_32)) || defined(CONFIG_IA64)
+#ifdef CONFIG_ARCH_BINFMT_ELF_EXTRA_PHDRS
 /*
  * These functions parameterize elf_core_dump in fs/binfmt_elf.c to write out
  * extra segments containing the gate DSO contents.  Dumping its
@@ -149,6 +149,6 @@ static inline size_t elf_core_extra_data_size(void)
 {
 	return 0;
 }
-#endif
+#endif /* CONFIG_ARCH_BINFMT_ELF_EXTRA_PHDRS */
 
 #endif /* _LINUX_ELFCORE_H */

include/uapi/linux/elf.h

@@ -40,6 +40,9 @@ typedef __s64	Elf64_Sxword;
 
 #define PT_GNU_STACK	(PT_LOOS + 0x474e551)
 
+/* ARM MTE memory tag segment type */
+#define PT_ARM_MEMTAG_MTE	(PT_LOPROC + 0x1)
+
 /*
  * Extended Numbering
  *