mm/mmap.c: remove verify_mm_writelocked()
authorYangtao Li <[email protected]>
Fri, 28 Dec 2018 08:34:09 +0000 (00:34 -0800)
committerLinus Torvalds <[email protected]>
Fri, 28 Dec 2018 20:11:47 +0000 (12:11 -0800)
We should get rid of this function.  It no longer serves its purpose.
This is a historical artifact from 2005 where do_brk was called outside of
the core mm.  We do have a proper abstraction in vm_brk_flags and that one
does the locking properly so there is no need to use this function.

Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Yangtao Li <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Acked-by: David Rientjes <[email protected]>
Cc: Dan Williams <[email protected]>
Cc: Dominik Brodowski <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: David Woodhouse <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
mm/mmap.c

index 7bb64381e77c09e2af0a3b693a93d5d48659ae7b..f901065c4c64cf63b455497fb993e7c4eacfc184 100644 (file)
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -2973,16 +2973,6 @@ out:
        return ret;
 }
 
-static inline void verify_mm_writelocked(struct mm_struct *mm)
-{
-#ifdef CONFIG_DEBUG_VM
-       if (unlikely(down_read_trylock(&mm->mmap_sem))) {
-               WARN_ON(1);
-               up_read(&mm->mmap_sem);
-       }
-#endif
-}
-
 /*
  *  this is really a simplified "do_mmap".  it only handles
  *  anonymous maps.  eventually we may be able to do some
@@ -3009,12 +2999,6 @@ static int do_brk_flags(unsigned long addr, unsigned long len, unsigned long fla
        if (error)
                return error;
 
-       /*
-        * mm->mmap_sem is required to protect against another thread
-        * changing the mappings in case we sleep.
-        */
-       verify_mm_writelocked(mm);
-
        /*
         * Clear old maps.  this also does some error checking for us
         */