1. 28 Oct, 2010 10 commits
    • Theodore Ts'o's avatar
      jbd2: Add sanity check for attempts to start handle during umount · 5c2178e7
      Theodore Ts'o authored
      An attempt to modify the file system during the call to
      jbd2_destroy_journal() can lead to a system lockup.  So add some
      checking to make it much more obvious when this happens to and to
      determine where the offending code is located.
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      5c2178e7
    • Sergey Senozhatsky's avatar
      ext4: fix NULL pointer dereference in print_daily_error_info() · a1c6c569
      Sergey Senozhatsky authored
      Fix NULL pointer dereference in print_daily_error_info, when   
      called on unmounted fs (EXT4_SB(sb) returns NULL), by removing error 
      reporting timer in ext4_put_super.
      
      Google-Bug-Id: 3017663
      Signed-off-by: default avatarSergey Senozhatsky <sergey.senozhatsky@gmail.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      a1c6c569
    • Lukas Czerner's avatar
      ext4: don't hold spinlock while calling ext4_issue_discard() · 53fdcf99
      Lukas Czerner authored
      We can't hold the block group spinlock because we ext4_issue_discard()
      calls wait and hence can get rescheduled.
      
      Google-Bug-Id: 3017678
      Signed-off-by: default avatarLukas Czerner <lczerner@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      53fdcf99
    • Lukas Czerner's avatar
      ext4: check for negative error code from sb_issue_discard · 58298709
      Lukas Czerner authored
      sb_issue_discard() is returning negative error code, so check for
      -EOPNOTSUPP.
      Signed-off-by: default avatarLukas Czerner <lczerner@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      58298709
    • Eric Sandeen's avatar
      ext4: don't bump up LONG_MAX nr_to_write by a factor of 8 · b443e733
      Eric Sandeen authored
      I'm uneasy with lots of stuff going on in ext4_da_writepages(),
      but bumping nr_to_write from LLONG_MAX to -8 clearly isn't
      making anything better, so avoid the multiplier in that case.
      Signed-off-by: default avatarEric Sandeen <sandeen@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      b443e733
    • Eric Sandeen's avatar
      ext4: stop looping in ext4_num_dirty_pages when max_pages reached · 659c6009
      Eric Sandeen authored
      Today we simply break out of the inner loop when we have accumulated
      max_pages; this keeps scanning forwad and doing pagevec_lookup_tag()
      in the while (!done) loop, this does potentially a lot of work
      with no net effect.
      
      When we have accumulated max_pages, just clean up and return.
      Signed-off-by: default avatarEric Sandeen <sandeen@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      659c6009
    • Curt Wohlgemuth's avatar
      ext4: use dedicated slab caches for group_info structures · fb1813f4
      Curt Wohlgemuth authored
      ext4_group_info structures are currently allocated with kmalloc().
      With a typical 4K block size, these are 136 bytes each -- meaning
      they'll each consume a 256-byte slab object.  On a system with many
      ext4 large partitions, that's a lot of wasted kernel slab space.
      (E.g., a single 1TB partition will have about 8000 block groups, using
      about 2MB of slab, of which nearly 1MB is wasted.)
      
      This patch creates an array of slab pointers created as needed --
      depending on the superblock block size -- and uses these slabs to
      allocate the group info objects.
      
      Google-Bug-Id: 2980809
      Signed-off-by: default avatarCurt Wohlgemuth <curtw@google.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      fb1813f4
    • Wen Congyang's avatar
      ext4: avoid null dereference in trace_ext4_mballoc_discard · b853fd36
      Wen Congyang authored
      ac->inode is set to null in function ext4_mb_release_group_pa(),
      and then trace_ext4_mballoc_discard(ac) is called, the kernel
      will panic.
      
      BUG: unable to handle kernel NULL pointer dereference at 000000a4
      IP: [<f87e1714>] ftrace_raw_event_ext4__mballoc+0x54/0xc0 [ext4]
      *pdpt = 0000000000abd001 *pde = 0000000000000000
      Oops: 0000 [#1] SMP
      
      Pid: 550, comm: flush-8:16 Not tainted 2.6.36-rc1 #1 SE7320EP2/Altos G530
      EIP: 0060:[<f87e1714>] EFLAGS: 00010206 CPU: 1
      EIP is at ftrace_raw_event_ext4__mballoc+0x54/0xc0 [ext4]
      EAX: f32ac840 EBX: f3f1cf88 ECX: f32ac840 EDX: 00000000
      ESI: f32ac83c EDI: f880b9d8 EBP: 00000000 ESP: f4b77ae4
       DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
      Process flush-8:16 (pid: 550, ti=f4b76000 task=f613e540 task.ti=f4b76000)
      Call Trace:
       [<f87f5ac1>] ? ext4_mb_release_group_pa+0x121/0x150 [ext4]
       [<f87f8356>] ? ext4_mb_discard_group_preallocations+0x336/0x400 [ext4]
       [<f87fb7f1>] ? ext4_mb_new_blocks+0x3d1/0x4f0 [ext4]
       [<c05a6c5b>] ? __make_request+0x10b/0x440
       [<f87f1fb4>] ? ext4_ext_map_blocks+0x1334/0x1980 [ext4]
       [<c04ac78a>] ? rb_reserve_next_event+0xaa/0x3b0
       [<f87d18d6>] ? ext4_map_blocks+0xd6/0x1d0 [ext4]
       [<f87d2da7>] ? mpage_da_map_blocks+0xc7/0x8a0 [ext4]
       [<c04c8a68>] ? find_get_pages_tag+0x38/0x110
       [<c04d23a5>] ? __pagevec_release+0x15/0x20
       [<f87d3ca5>] ? ext4_da_writepages+0x2b5/0x5d0 [ext4]
       [<c04cfbe0>] ? __writepage+0x0/0x30
       [<c04d0e34>] ? do_writepages+0x14/0x30
       [<c0526600>] ? writeback_single_inode+0xa0/0x240
       [<c0526971>] ? writeback_sb_inodes+0xc1/0x180
       [<c0526ab8>] ? writeback_inodes_wb+0x88/0x140
       [<c0526d7b>] ? wb_writeback+0x20b/0x320
       [<c045aca7>] ? lock_timer_base+0x27/0x50
       [<c0526fe0>] ? wb_do_writeback+0x150/0x190
       [<c05270a8>] ? bdi_writeback_thread+0x88/0x1f0
       [<c043b680>] ? complete+0x40/0x60
       [<c0527020>] ? bdi_writeback_thread+0x0/0x1f0
       [<c0469474>] ? kthread+0x74/0x80
       [<c0469400>] ? kthread+0x0/0x80
       [<c040a23e>] ? kernel_thread_helper+0x6/0x10
      Signed-off-by: default avatarWen Congyang <wency@cn.fujitsu.com>
      Acked-by: default avatarSteven Rostedt <rostedt@goodmis.org>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      b853fd36
    • Brian King's avatar
      jbd2: Fix I/O hang in jbd2_journal_release_jbd_inode · 39e3ac25
      Brian King authored
      This fixes a hang seen in jbd2_journal_release_jbd_inode
      on a lot of Power 6 systems running with ext4. When we get
      in the hung state, all I/O to the disk in question gets blocked
      where we stay indefinitely. Looking at the task list, I can see
      we are stuck in jbd2_journal_release_jbd_inode waiting on a
      wake up. I added some debug code to detect this scenario and
      dump additional data if we were stuck in jbd2_journal_release_jbd_inode
      for longer than 30 minutes. When it hit, I was able to see that
      i_flags was 0, suggesting we missed the wake up.
      
      This patch changes i_flags to be an unsigned long, uses bit operators
      to access it, and adds barriers around the accesses. Prior to applying
      this patch, we were regularly hitting this hang on numerous systems
      in our test environment. After applying the patch, the hangs no longer
      occur.
      Signed-off-by: default avatarBrian King <brking@linux.vnet.ibm.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      39e3ac25
    • Theodore Ts'o's avatar
      ext4: fix EOFBLOCKS_FL handling · 58590b06
      Theodore Ts'o authored
      It turns out we have several problems with how EOFBLOCKS_FL is
      handled.  First of all, there was a fencepost error where we were not
      clearing the EOFBLOCKS_FL when fill in the last uninitialized block,
      but rather when we allocate the next block _after_ the uninitalized
      block.  Secondly we were not testing to see if we needed to clear the
      EOFBLOCKS_FL when writing to the file O_DIRECT or when were converting
      an uninitialized block (which is the most common case).
      
      Google-Bug-Id: 2928259
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      58590b06
  2. 29 Sep, 2010 2 commits
  3. 28 Sep, 2010 9 commits
  4. 27 Sep, 2010 19 commits