mm: __set_page_dirty uses spin_lock_irqsave instead of spin_lock_irq
authorKOSAKI Motohiro <[email protected]>
Thu, 6 Feb 2014 20:04:28 +0000 (12:04 -0800)
committerLinus Torvalds <[email protected]>
Thu, 6 Feb 2014 21:48:51 +0000 (13:48 -0800)
commit227d53b397a32a7614667b3ecaf1d89902fb6c12
treecfdccad47eb1404da7a657cd260af3ba1b3c3120
parent7bc35fdde6724549a0239b71e08b9f33d8bf2bfb
mm: __set_page_dirty uses spin_lock_irqsave instead of spin_lock_irq

To use spin_{un}lock_irq is dangerous if caller disabled interrupt.
During aio buffer migration, we have a possibility to see the following
call stack.

aio_migratepage  [disable interrupt]
  migrate_page_copy
    clear_page_dirty_for_io
      set_page_dirty
        __set_page_dirty_buffers
          __set_page_dirty
            spin_lock_irq

This mean, current aio migration is a deadlockable.  spin_lock_irqsave
is a safer alternative and we should use it.

Signed-off-by: KOSAKI Motohiro <[email protected]>
Reported-by: David Rientjes [email protected]>
Cc: <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
fs/buffer.c