mm: use down_read_killable for locking mmap_sem in access_remote_vm
authorKonstantin Khlebnikov <khlebnikov@yandex-team.ru>
Fri, 12 Jul 2019 04:00:07 +0000 (21:00 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 31 Jul 2019 05:27:09 +0000 (07:27 +0200)
[ Upstream commit 1e426fe28261b03f297992e89da3320b42816f4e ]

This function is used by ptrace and proc files like /proc/pid/cmdline and
/proc/pid/environ.

Access_remote_vm never returns error codes, all errors are ignored and
only size of successfully read data is returned.  So, if current task was
killed we'll simply return 0 (bytes read).

Mmap_sem could be locked for a long time or forever if something goes
wrong.  Using a killable lock permits cleanup of stuck tasks and
simplifies investigation.

Link: http://lkml.kernel.org/r/156007494202.3335.16782303099589302087.stgit@buzz
Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Reviewed-by: Michal Koutný <mkoutny@suse.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Cyrill Gorcunov <gorcunov@gmail.com>
Cc: Kirill Tkhai <ktkhai@virtuozzo.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Roman Gushchin <guro@fb.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
mm/memory.c
mm/nommu.c

index e0010cb..fb5655b 100644 (file)
@@ -4491,7 +4491,9 @@ int __access_remote_vm(struct task_struct *tsk, struct mm_struct *mm,
        void *old_buf = buf;
        int write = gup_flags & FOLL_WRITE;
 
-       down_read(&mm->mmap_sem);
+       if (down_read_killable(&mm->mmap_sem))
+               return 0;
+
        /* ignore errors, just check how much was successfully transferred */
        while (len) {
                int bytes, ret, offset;
index e4aac33..1d63ecf 100644 (file)
@@ -1779,7 +1779,8 @@ int __access_remote_vm(struct task_struct *tsk, struct mm_struct *mm,
        struct vm_area_struct *vma;
        int write = gup_flags & FOLL_WRITE;
 
-       down_read(&mm->mmap_sem);
+       if (down_read_killable(&mm->mmap_sem))
+               return 0;
 
        /* the access must start within one of the target process's mappings */
        vma = find_vma(mm, addr);