Merge branch 'coredump-vma-snapshot-fix-for-v5.18' of https://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace into for-next/execve

- Fix missing mmap_lock in file_files_note (Eric W. Biederman)

6 files changed, 97 insertions, 84 deletions

diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index eaf39b1bdbbb..ca5296cae979 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1641,17 +1641,16 @@ static void fill_siginfo_note(struct memelfnote *note, user_siginfo_t *csigdata,
  *   long file_ofs
  * followed by COUNT filenames in ASCII: "FILE1" NUL "FILE2" NUL...
  */
-static int fill_files_note(struct memelfnote *note)
+static int fill_files_note(struct memelfnote *note, struct coredump_params *cprm)
 {
-	struct mm_struct *mm = current->mm;
-	struct vm_area_struct *vma;
 	unsigned count, size, names_ofs, remaining, n;
 	user_long_t *data;
 	user_long_t *start_end_ofs;
 	char *name_base, *name_curpos;
+	int i;
 
 	/* *Estimated* file count and total data size needed */
-	count = mm->map_count;
+	count = cprm->vma_count;
 	if (count > UINT_MAX / 64)
 		return -EINVAL;
 	size = count * 64;
@@ -1673,11 +1672,12 @@ static int fill_files_note(struct memelfnote *note)
 	name_base = name_curpos = ((char *)data) + names_ofs;
 	remaining = size - names_ofs;
 	count = 0;
-	for (vma = mm->mmap; vma != NULL; vma = vma->vm_next) {
+	for (i = 0; i < cprm->vma_count; i++) {
+		struct core_vma_metadata *m = &cprm->vma_meta[i];
 		struct file *file;
 		const char *filename;
 
-		file = vma->vm_file;
+		file = m->file;
 		if (!file)
 			continue;
 		filename = file_path(file, name_curpos, remaining);
@@ -1697,9 +1697,9 @@ static int fill_files_note(struct memelfnote *note)
 		memmove(name_curpos, filename, n);
 		name_curpos += n;
 
-		*start_end_ofs++ = vma->vm_start;
-		*start_end_ofs++ = vma->vm_end;
-		*start_end_ofs++ = vma->vm_pgoff;
+		*start_end_ofs++ = m->start;
+		*start_end_ofs++ = m->end;
+		*start_end_ofs++ = m->pgoff;
 		count++;
 	}
 
@@ -1710,7 +1710,7 @@ static int fill_files_note(struct memelfnote *note)
 	 * Count usually is less than mm->map_count,
 	 * we need to move filenames down.
 	 */
-	n = mm->map_count - count;
+	n = cprm->vma_count - count;
 	if (n != 0) {
 		unsigned shift_bytes = n * 3 * sizeof(data[0]);
 		memmove(name_base - shift_bytes, name_base,
@@ -1822,7 +1822,7 @@ static int fill_thread_core_info(struct elf_thread_core_info *t,
 
 static int fill_note_info(struct elfhdr *elf, int phdrs,
 			  struct elf_note_info *info,
-			  const kernel_siginfo_t *siginfo, struct pt_regs *regs)
+			  struct coredump_params *cprm)
 {
 	struct task_struct *dump_task = current;
 	const struct user_regset_view *view = task_user_regset_view(dump_task);
@@ -1894,7 +1894,7 @@ static int fill_note_info(struct elfhdr *elf, int phdrs,
 	 * Now fill in each thread's information.
 	 */
 	for (t = info->thread; t != NULL; t = t->next)
-		if (!fill_thread_core_info(t, view, siginfo->si_signo, &info->size))
+		if (!fill_thread_core_info(t, view, cprm->siginfo->si_signo, &info->size))
 			return 0;
 
 	/*
@@ -1903,13 +1903,13 @@ static int fill_note_info(struct elfhdr *elf, int phdrs,
 	fill_psinfo(psinfo, dump_task->group_leader, dump_task->mm);
 	info->size += notesize(&info->psinfo);
 
-	fill_siginfo_note(&info->signote, &info->csigdata, siginfo);
+	fill_siginfo_note(&info->signote, &info->csigdata, cprm->siginfo);
 	info->size += notesize(&info->signote);
 
 	fill_auxv_note(&info->auxv, current->mm);
 	info->size += notesize(&info->auxv);
 
-	if (fill_files_note(&info->files) == 0)
+	if (fill_files_note(&info->files, cprm) == 0)
 		info->size += notesize(&info->files);
 
 	return 1;
@@ -2051,7 +2051,7 @@ static int elf_note_info_init(struct elf_note_info *info)
 
 static int fill_note_info(struct elfhdr *elf, int phdrs,
 			  struct elf_note_info *info,
-			  const kernel_siginfo_t *siginfo, struct pt_regs *regs)
+			  struct coredump_params *cprm)
 {
 	struct core_thread *ct;
 	struct elf_thread_status *ets;
@@ -2072,13 +2072,13 @@ static int fill_note_info(struct elfhdr *elf, int phdrs,
 	list_for_each_entry(ets, &info->thread_list, list) {
 		int sz;
 
-		sz = elf_dump_thread_status(siginfo->si_signo, ets);
+		sz = elf_dump_thread_status(cprm->siginfo->si_signo, ets);
 		info->thread_status_size += sz;
 	}
 	/* now collect the dump for the current */
 	memset(info->prstatus, 0, sizeof(*info->prstatus));
-	fill_prstatus(&info->prstatus->common, current, siginfo->si_signo);
-	elf_core_copy_regs(&info->prstatus->pr_reg, regs);
+	fill_prstatus(&info->prstatus->common, current, cprm->siginfo->si_signo);
+	elf_core_copy_regs(&info->prstatus->pr_reg, cprm->regs);
 
 	/* Set up header */
 	fill_elf_header(elf, phdrs, ELF_ARCH, ELF_CORE_EFLAGS);
@@ -2094,18 +2094,18 @@ static int fill_note_info(struct elfhdr *elf, int phdrs,
 	fill_note(info->notes + 1, "CORE", NT_PRPSINFO,
 		  sizeof(*info->psinfo), info->psinfo);
 
-	fill_siginfo_note(info->notes + 2, &info->csigdata, siginfo);
+	fill_siginfo_note(info->notes + 2, &info->csigdata, cprm->siginfo);
 	fill_auxv_note(info->notes + 3, current->mm);
 	info->numnote = 4;
 
-	if (fill_files_note(info->notes + info->numnote) == 0) {
+	if (fill_files_note(info->notes + info->numnote, cprm) == 0) {
 		info->notes_files = info->notes + info->numnote;
 		info->numnote++;
 	}
 
 	/* Try to dump the FPU. */
-	info->prstatus->pr_fpvalid = elf_core_copy_task_fpregs(current, regs,
-							       info->fpu);
+	info->prstatus->pr_fpvalid =
+		elf_core_copy_task_fpregs(current, cprm->regs, info->fpu);
 	if (info->prstatus->pr_fpvalid)
 		fill_note(info->notes + info->numnote++,
 			  "CORE", NT_PRFPREG, sizeof(*info->fpu), info->fpu);
@@ -2191,8 +2191,7 @@ static void fill_extnum_info(struct elfhdr *elf, struct elf_shdr *shdr4extnum,
 static int elf_core_dump(struct coredump_params *cprm)
 {
 	int has_dumped = 0;
-	int vma_count, segs, i;
-	size_t vma_data_size;
+	int segs, i;
 	struct elfhdr elf;
 	loff_t offset = 0, dataoff;
 	struct elf_note_info info = { };
@@ -2200,16 +2199,12 @@ static int elf_core_dump(struct coredump_params *cprm)
 	struct elf_shdr *shdr4extnum = NULL;
 	Elf_Half e_phnum;
 	elf_addr_t e_shoff;
-	struct core_vma_metadata *vma_meta;
-
-	if (dump_vma_snapshot(cprm, &vma_count, &vma_meta, &vma_data_size))
-		return 0;
 
 	/*
 	 * The number of segs are recored into ELF header as 16bit value.
 	 * Please check DEFAULT_MAX_MAP_COUNT definition when you modify here.
 	 */
-	segs = vma_count + elf_core_extra_phdrs();
+	segs = cprm->vma_count + elf_core_extra_phdrs();
 
 	/* for notes section */
 	segs++;
@@ -2223,7 +2218,7 @@ static int elf_core_dump(struct coredump_params *cprm)
 	 * Collect all the non-memory information about the process for the
 	 * notes.  This also sets up the file header.
 	 */
-	if (!fill_note_info(&elf, e_phnum, &info, cprm->siginfo, cprm->regs))
+	if (!fill_note_info(&elf, e_phnum, &info, cprm))
 		goto end_coredump;
 
 	has_dumped = 1;
@@ -2248,7 +2243,7 @@ static int elf_core_dump(struct coredump_params *cprm)
 
 	dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);
 
-	offset += vma_data_size;
+	offset += cprm->vma_data_size;
 	offset += elf_core_extra_data_size();
 	e_shoff = offset;
 
@@ -2268,8 +2263,8 @@ static int elf_core_dump(struct coredump_params *cprm)
 		goto end_coredump;
 
 	/* Write program headers for segments dump */
-	for (i = 0; i < vma_count; i++) {
-		struct core_vma_metadata *meta = vma_meta + i;
+	for (i = 0; i < cprm->vma_count; i++) {
+		struct core_vma_metadata *meta = cprm->vma_meta + i;
 		struct elf_phdr phdr;
 
 		phdr.p_type = PT_LOAD;
@@ -2306,8 +2301,8 @@ static int elf_core_dump(struct coredump_params *cprm)
 	/* Align to page */
 	dump_skip_to(cprm, dataoff);
 
-	for (i = 0; i < vma_count; i++) {
-		struct core_vma_metadata *meta = vma_meta + i;
+	for (i = 0; i < cprm->vma_count; i++) {
+		struct core_vma_metadata *meta = cprm->vma_meta + i;
 
 		if (!dump_user_range(cprm, meta->start, meta->dump_size))
 			goto end_coredump;
@@ -2324,7 +2319,6 @@ static int elf_core_dump(struct coredump_params *cprm)
 end_coredump:
 	free_note_info(&info);
 	kfree(shdr4extnum);
-	kvfree(vma_meta);
 	kfree(phdr4note);
 	return has_dumped;
 }
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 7fa6e6632d9d..08d0c8797828 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -1465,7 +1465,7 @@ static bool elf_fdpic_dump_segments(struct coredump_params *cprm,
 static int elf_fdpic_core_dump(struct coredump_params *cprm)
 {
 	int has_dumped = 0;
-	int vma_count, segs;
+	int segs;
 	int i;
 	struct elfhdr *elf = NULL;
 	loff_t offset = 0, dataoff;
@@ -1480,8 +1480,6 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	elf_addr_t e_shoff;
 	struct core_thread *ct;
 	struct elf_thread_status *tmp;
-	struct core_vma_metadata *vma_meta = NULL;
-	size_t vma_data_size;
 
 	/* alloc memory for large data structures: too large to be on stack */
 	elf = kmalloc(sizeof(*elf), GFP_KERNEL);
@@ -1491,9 +1489,6 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	if (!psinfo)
 		goto end_coredump;
 
-	if (dump_vma_snapshot(cprm, &vma_count, &vma_meta, &vma_data_size))
-		goto end_coredump;
-
 	for (ct = current->signal->core_state->dumper.next;
 					ct; ct = ct->next) {
 		tmp = elf_dump_thread_status(cprm->siginfo->si_signo,
@@ -1513,7 +1508,7 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	tmp->next = thread_list;
 	thread_list = tmp;
 
-	segs = vma_count + elf_core_extra_phdrs();
+	segs = cprm->vma_count + elf_core_extra_phdrs();
 
 	/* for notes section */
 	segs++;
@@ -1558,7 +1553,7 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 	/* Page-align dumped data */
 	dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);
 
-	offset += vma_data_size;
+	offset += cprm->vma_data_size;
 	offset += elf_core_extra_data_size();
 	e_shoff = offset;
 
@@ -1578,8 +1573,8 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 		goto end_coredump;
 
 	/* write program headers for segments dump */
-	for (i = 0; i < vma_count; i++) {
-		struct core_vma_metadata *meta = vma_meta + i;
+	for (i = 0; i < cprm->vma_count; i++) {
+		struct core_vma_metadata *meta = cprm->vma_meta + i;
 		struct elf_phdr phdr;
 		size_t sz;
 
@@ -1628,7 +1623,7 @@ static int elf_fdpic_core_dump(struct coredump_params *cprm)
 
 	dump_skip_to(cprm, dataoff);
 
-	if (!elf_fdpic_dump_segments(cprm, vma_meta, vma_count))
+	if (!elf_fdpic_dump_segments(cprm, cprm->vma_meta, cprm->vma_count))
 		goto end_coredump;
 
 	if (!elf_core_write_extra_data(cprm))
@@ -1652,7 +1647,6 @@ end_coredump:
 		thread_list = thread_list->next;
 		kfree(tmp);
 	}
-	kvfree(vma_meta);
 	kfree(phdr4note);
 	kfree(elf);
 	kfree(psinfo);
diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c
index 5f0bf24bb3b8..626898150011 100644
--- a/fs/binfmt_flat.c
+++ b/fs/binfmt_flat.c
@@ -37,6 +37,7 @@
 #include <linux/flat.h>
 #include <linux/uaccess.h>
 #include <linux/vmalloc.h>
+#include <linux/coredump.h>
 
 #include <asm/byteorder.h>
 #include <asm/unaligned.h>
@@ -97,7 +98,9 @@ static int load_flat_shared_library(int id, struct lib_info *p);
 #endif
 
 static int load_flat_binary(struct linux_binprm *);
+#ifdef CONFIG_COREDUMP
 static int flat_core_dump(struct coredump_params *cprm);
+#endif
 
 static struct linux_binfmt flat_format = {
 	.module		= THIS_MODULE,
@@ -114,12 +117,14 @@ static struct linux_binfmt flat_format = {
  * Currently only a stub-function.
  */
 
+#ifdef CONFIG_COREDUMP
 static int flat_core_dump(struct coredump_params *cprm)
 {
 	pr_warn("Process %s:%d received signr %d and should have core dumped\n",
 		current->comm, current->pid, cprm->siginfo->si_signo);
 	return 1;
 }
+#endif
 
 /****************************************************************************/
 /*
diff --git a/fs/coredump.c b/fs/coredump.c
index b73817712dd2..7ed7d601e5e0 100644
--- a/fs/coredump.c
+++ b/fs/coredump.c
@@ -54,6 +54,9 @@
 
 #include <trace/events/sched.h>
 
+static bool dump_vma_snapshot(struct coredump_params *cprm);
+static void free_vma_snapshot(struct coredump_params *cprm);
+
 static int core_uses_pid;
 static unsigned int core_pipe_limit;
 static char core_pattern[CORENAME_MAX_SIZE] = "core";
@@ -532,6 +535,7 @@ void do_coredump(const kernel_siginfo_t *siginfo)
 		 * by any locks.
 		 */
 		.mm_flags = mm->flags,
+		.vma_meta = NULL,
 	};
 
 	audit_core_dumps(siginfo->si_signo);
@@ -746,6 +750,9 @@ void do_coredump(const kernel_siginfo_t *siginfo)
 			pr_info("Core dump to |%s disabled\n", cn.corename);
 			goto close_fail;
 		}
+		if (!dump_vma_snapshot(&cprm))
+			goto close_fail;
+
 		file_start_write(cprm.file);
 		core_dumped = binfmt->core_dump(&cprm);
 		/*
@@ -759,6 +766,7 @@ void do_coredump(const kernel_siginfo_t *siginfo)
 			dump_emit(&cprm, "", 1);
 		}
 		file_end_write(cprm.file);
+		free_vma_snapshot(&cprm);
 	}
 	if (ispipe && core_pipe_limit)
 		wait_for_dump_helpers(cprm.file);
@@ -1092,18 +1100,29 @@ static struct vm_area_struct *next_vma(struct vm_area_struct *this_vma,
 	return gate_vma;
 }
 
+static void free_vma_snapshot(struct coredump_params *cprm)
+{
+	if (cprm->vma_meta) {
+		int i;
+		for (i = 0; i < cprm->vma_count; i++) {
+			struct file *file = cprm->vma_meta[i].file;
+			if (file)
+				fput(file);
+		}
+		kvfree(cprm->vma_meta);
+		cprm->vma_meta = NULL;
+	}
+}
+
 /*
  * Under the mmap_lock, take a snapshot of relevant information about the task's
  * VMAs.
  */
-int dump_vma_snapshot(struct coredump_params *cprm, int *vma_count,
-		      struct core_vma_metadata **vma_meta,
-		      size_t *vma_data_size_ptr)
+static bool dump_vma_snapshot(struct coredump_params *cprm)
 {
 	struct vm_area_struct *vma, *gate_vma;
 	struct mm_struct *mm = current->mm;
 	int i;
-	size_t vma_data_size = 0;
 
 	/*
 	 * Once the stack expansion code is fixed to not change VMA bounds
@@ -1111,36 +1130,37 @@ int dump_vma_snapshot(struct coredump_params *cprm, int *vma_count,
 	 * mmap_lock in read mode.
 	 */
 	if (mmap_write_lock_killable(mm))
-		return -EINTR;
+		return false;
 
+	cprm->vma_data_size = 0;
 	gate_vma = get_gate_vma(mm);
-	*vma_count = mm->map_count + (gate_vma ? 1 : 0);
+	cprm->vma_count = mm->map_count + (gate_vma ? 1 : 0);
 
-	*vma_meta = kvmalloc_array(*vma_count, sizeof(**vma_meta), GFP_KERNEL);
-	if (!*vma_meta) {
+	cprm->vma_meta = kvmalloc_array(cprm->vma_count, sizeof(*cprm->vma_meta), GFP_KERNEL);
+	if (!cprm->vma_meta) {
 		mmap_write_unlock(mm);
-		return -ENOMEM;
+		return false;
 	}
 
 	for (i = 0, vma = first_vma(current, gate_vma); vma != NULL;
 			vma = next_vma(vma, gate_vma), i++) {
-		struct core_vma_metadata *m = (*vma_meta) + i;
+		struct core_vma_metadata *m = cprm->vma_meta + i;
 
 		m->start = vma->vm_start;
 		m->end = vma->vm_end;
 		m->flags = vma->vm_flags;
 		m->dump_size = vma_dump_size(vma, cprm->mm_flags);
+		m->pgoff = vma->vm_pgoff;
+
+		m->file = vma->vm_file;
+		if (m->file)
+			get_file(m->file);
 	}
 
 	mmap_write_unlock(mm);
 
-	if (WARN_ON(i != *vma_count)) {
-		kvfree(*vma_meta);
-		return -EFAULT;
-	}
-
-	for (i = 0; i < *vma_count; i++) {
-		struct core_vma_metadata *m = (*vma_meta) + i;
+	for (i = 0; i < cprm->vma_count; i++) {
+		struct core_vma_metadata *m = cprm->vma_meta + i;
 
 		if (m->dump_size == DUMP_SIZE_MAYBE_ELFHDR_PLACEHOLDER) {
 			char elfmag[SELFMAG];
@@ -1153,9 +1173,8 @@ int dump_vma_snapshot(struct coredump_params *cprm, int *vma_count,
 			}
 		}
 
-		vma_data_size += m->dump_size;
+		cprm->vma_data_size += m->dump_size;
 	}
 
-	*vma_data_size_ptr = vma_data_size;
-	return 0;
+	return true;
 }
diff --git a/include/linux/binfmts.h b/include/linux/binfmts.h
index 5d651c219c99..3dc20c4f394c 100644
--- a/include/linux/binfmts.h
+++ b/include/linux/binfmts.h
@@ -8,6 +8,7 @@
 #include <uapi/linux/binfmts.h>
 
 struct filename;
+struct coredump_params;
 
 #define CORENAME_MAX_SIZE 128
 
@@ -77,18 +78,6 @@ struct linux_binprm {
 #define BINPRM_FLAGS_PRESERVE_ARGV0_BIT 3
 #define BINPRM_FLAGS_PRESERVE_ARGV0 (1 << BINPRM_FLAGS_PRESERVE_ARGV0_BIT)
 
-/* Function parameter for binfmt->coredump */
-struct coredump_params {
-	const kernel_siginfo_t *siginfo;
-	struct pt_regs *regs;
-	struct file *file;
-	unsigned long limit;
-	unsigned long mm_flags;
-	loff_t written;
-	loff_t pos;
-	loff_t to_skip;
-};
-
 /*
  * This structure defines the functions that are used to load the binary formats that
  * linux accepts.
diff --git a/include/linux/coredump.h b/include/linux/coredump.h
index 248a68c668b4..08a1d3e7e46d 100644
--- a/include/linux/coredump.h
+++ b/include/linux/coredump.h
@@ -12,22 +12,34 @@ struct core_vma_metadata {
 	unsigned long start, end;
 	unsigned long flags;
 	unsigned long dump_size;
+	unsigned long pgoff;
+	struct file   *file;
+};
+
+struct coredump_params {
+	const kernel_siginfo_t *siginfo;
+	struct pt_regs *regs;
+	struct file *file;
+	unsigned long limit;
+	unsigned long mm_flags;
+	loff_t written;
+	loff_t pos;
+	loff_t to_skip;
+	int vma_count;
+	size_t vma_data_size;
+	struct core_vma_metadata *vma_meta;
 };
 
 /*
  * These are the only things you should do on a core-file: use only these
  * functions to write out all the necessary info.
  */
-struct coredump_params;
 extern void dump_skip_to(struct coredump_params *cprm, unsigned long to);
 extern void dump_skip(struct coredump_params *cprm, size_t nr);
 extern int dump_emit(struct coredump_params *cprm, const void *addr, int nr);
 extern int dump_align(struct coredump_params *cprm, int align);
 int dump_user_range(struct coredump_params *cprm, unsigned long start,
 		    unsigned long len);
-int dump_vma_snapshot(struct coredump_params *cprm, int *vma_count,
-		      struct core_vma_metadata **vma_meta,
-		      size_t *vma_data_size_ptr);
 extern void do_coredump(const kernel_siginfo_t *siginfo);
 #else
 static inline void do_coredump(const kernel_siginfo_t *siginfo) {}