1. 14 2月, 2017 3 次提交
  2. 27 1月, 2017 3 次提交
  3. 20 1月, 2017 3 次提交
    • L
      Btrfs: fix truncate down when no_holes feature is enabled · 91298eec
      Liu Bo 提交于
      For such a file mapping,
      
      [0-4k][hole][8k-12k]
      
      In NO_HOLES mode, we don't have the [hole] extent any more.
      Commit c1aa4575 ("Btrfs: fix shrinking truncate when the no_holes feature is enabled")
       fixed disk isize not being updated in NO_HOLES mode when data is not flushed.
      
      However, even if data has been flushed, we can still have trouble
      in updating disk isize since we updated disk isize to 'start' of
      the last evicted extent.
      Reviewed-by: NChris Mason <clm@fb.com>
      Signed-off-by: NLiu Bo <bo.li.liu@oracle.com>
      Signed-off-by: NDavid Sterba <dsterba@suse.com>
      91298eec
    • C
      Btrfs: Fix deadlock between direct IO and fast fsync · 97dcdea0
      Chandan Rajendra 提交于
      The following deadlock is seen when executing generic/113 test,
      
       ---------------------------------------------------------+----------------------------------------------------
        Direct I/O task                                           Fast fsync task
       ---------------------------------------------------------+----------------------------------------------------
        btrfs_direct_IO
          __blockdev_direct_IO
           do_blockdev_direct_IO
            do_direct_IO
             btrfs_get_blocks_direct
              while (blocks needs to written)
               get_more_blocks (first iteration)
                btrfs_get_blocks_direct
                 btrfs_create_dio_extent
                   down_read(&BTRFS_I(inode) >dio_sem)
                   Create and add extent map and ordered extent
                   up_read(&BTRFS_I(inode) >dio_sem)
                                                                  btrfs_sync_file
                                                                    btrfs_log_dentry_safe
                                                                     btrfs_log_inode_parent
                                                                      btrfs_log_inode
                                                                       btrfs_log_changed_extents
                                                                        down_write(&BTRFS_I(inode) >dio_sem)
                                                                         Collect new extent maps and ordered extents
                                                                          wait for ordered extent completion
               get_more_blocks (second iteration)
                btrfs_get_blocks_direct
                 btrfs_create_dio_extent
                   down_read(&BTRFS_I(inode) >dio_sem)
       --------------------------------------------------------------------------------------------------------------
      
      In the above description, Btrfs direct I/O code path has not yet started
      submitting bios for file range covered by the initial ordered
      extent. Meanwhile, The fast fsync task obtains the write semaphore and
      waits for I/O on the ordered extent to get completed. However, the
      Direct I/O task is now blocked on obtaining the read semaphore.
      
      To resolve the deadlock, this commit modifies the Direct I/O code path
      to obtain the read semaphore before invoking
      __blockdev_direct_IO(). The semaphore is then given up after
      __blockdev_direct_IO() returns. This allows the Direct I/O code to
      complete I/O on all the ordered extents it creates.
      Signed-off-by: NChandan Rajendra <chandan@linux.vnet.ibm.com>
      Reviewed-by: NFilipe Manana <fdmanana@suse.com>
      Signed-off-by: NDavid Sterba <dsterba@suse.com>
      97dcdea0
    • W
      btrfs: fix false enospc error when truncating heavily reflinked file · 47b5d646
      Wang Xiaoguang 提交于
      Below test script can reveal this bug:
          dd if=/dev/zero of=fs.img bs=$((1024*1024)) count=100
          dev=$(losetup --show -f fs.img)
          mkdir -p /mnt/mntpoint
          mkfs.btrfs  -f $dev
          mount $dev /mnt/mntpoint
          cd /mnt/mntpoint
      
          echo "workdir is: /mnt/mntpoint"
          blocksize=$((128 * 1024))
          dd if=/dev/zero of=testfile bs=$blocksize count=1
          sync
          count=$((17*1024*1024*1024/blocksize))
          echo "file size is:" $((count*blocksize))
          for ((i = 1; i <= $count; i++)); do
              dst_offset=$((blocksize * i))
              xfs_io -f -c "reflink testfile 0 $dst_offset $blocksize"\
                      testfile > /dev/null
          done
          sync
          truncate --size 0 testfile
      
      The last truncate operation will fail for ENOSPC reason, but indeed
      it should not fail.
      
      In btrfs_truncate(), we use a temporary block_rsv to do truncate
      operation. With every btrfs_truncate_inode_items() call, we migrate space
      to this block_rsv, but forget to cleanup previous reservation, which
      will make this block_rsv's reserved bytes keep growing, and this reserved
      space will only be released in the end of btrfs_truncate(), this metadata
      leak will impact other's metadata reservation. In this case, it's
      "btrfs_start_transaction(root, 2);" fails for enospc error, which make
      this truncate operation fail.
      
      Call btrfs_block_rsv_release() to fix this bug.
      Signed-off-by: NWang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
      Reviewed-by: NDavid Sterba <dsterba@suse.com>
      Signed-off-by: NDavid Sterba <dsterba@suse.com>
      47b5d646
  4. 09 1月, 2017 1 次提交
  5. 04 1月, 2017 1 次提交
  6. 12 12月, 2016 1 次提交
  7. 09 12月, 2016 1 次提交
  8. 06 12月, 2016 6 次提交
  9. 30 11月, 2016 8 次提交
  10. 29 11月, 2016 1 次提交
  11. 01 11月, 2016 1 次提交
  12. 28 10月, 2016 1 次提交
    • C
      block: better op and flags encoding · ef295ecf
      Christoph Hellwig 提交于
      Now that we don't need the common flags to overflow outside the range
      of a 32-bit type we can encode them the same way for both the bio and
      request fields.  This in addition allows us to place the operation
      first (and make some room for more ops while we're at it) and to
      stop having to shift around the operation values.
      
      In addition this allows passing around only one value in the block layer
      instead of two (and eventuall also in the file systems, but we can do
      that later) and thus clean up a lot of code.
      
      Last but not least this allows decreasing the size of the cmd_flags
      field in struct request to 32-bits.  Various functions passing this
      value could also be updated, but I'd like to avoid the churn for now.
      Signed-off-by: NChristoph Hellwig <hch@lst.de>
      Signed-off-by: NJens Axboe <axboe@fb.com>
      ef295ecf
  13. 25 10月, 2016 2 次提交
    • W
      btrfs: pass correct args to btrfs_async_run_delayed_refs() · dd4b857a
      Wang Xiaoguang 提交于
      In btrfs_truncate_inode_items()->btrfs_async_run_delayed_refs(), we
      swap the arg2 and arg3 wrongly, fix this.
      
      This bug just impacts asynchronous delayed refs handle when we truncate inodes.
      In delayed_ref_async_start(), there is such codes:
      
          trans = btrfs_join_transaction(async->root);
          if (trans->transid > async->transid)
              goto end;
          ret = btrfs_run_delayed_refs(trans, async->root, async->count);
      
      From this codes, we can see that this just influence whether can we handle
      delayed refs or the number of delayed refs to handle, this may impact
      performance, but will not result in missing delayed refs, all delayed refs will
      be handled in btrfs_commit_transaction().
      Signed-off-by: NWang Xiaoguang <wangxg.fnst@cn.fujitsu.com>
      Reviewed-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
      Signed-off-by: NDavid Sterba <dsterba@suse.com>
      dd4b857a
    • G
      btrfs: qgroup: Prevent qgroup->reserved from going subzero · 0b34c261
      Goldwyn Rodrigues 提交于
      While free'ing qgroup->reserved resources, we much check if
      the page has not been invalidated by a truncate operation
      by checking if the page is still dirty before reducing the
      qgroup resources. Resources in such a case are free'd when
      the entire extent is released by delayed_ref.
      
      This fixes a double accounting while releasing resources
      in case of truncating a file, reproduced by the following testcase.
      
      SCRATCH_DEV=/dev/vdb
      SCRATCH_MNT=/mnt
      mkfs.btrfs -f $SCRATCH_DEV
      mount -t btrfs $SCRATCH_DEV $SCRATCH_MNT
      cd $SCRATCH_MNT
      btrfs quota enable $SCRATCH_MNT
      btrfs subvolume create a
      btrfs qgroup limit 500m a $SCRATCH_MNT
      sync
      for c in {1..15}; do
      dd if=/dev/zero  bs=1M count=40 of=$SCRATCH_MNT/a/file;
      done
      
      sleep 10
      sync
      sleep 5
      
      touch $SCRATCH_MNT/a/newfile
      
      echo "Removing file"
      rm $SCRATCH_MNT/a/file
      
      Fixes: b9d0b389 ("btrfs: Add handler for invalidate page")
      Signed-off-by: NGoldwyn Rodrigues <rgoldwyn@suse.com>
      Reviewed-by: NQu Wenruo <quwenruo@cn.fujitsu.com>
      Signed-off-by: NDavid Sterba <dsterba@suse.com>
      0b34c261
  14. 11 10月, 2016 1 次提交
    • A
      [btrfs] fix check_direct_IO() for non-iovec iterators · cd27e455
      Al Viro 提交于
      looking for duplicate ->iov_base makes sense only for
      iovec-backed iterators; for kvec-backed ones it's pointless,
      for bvec-backed ones it's pointless and broken on 32bit (we
      walk through an array of struct bio_vec accessing them as if
      they were struct iovec; works by accident on 64bit, but on
      32bit it'll blow up) and for pipe-backed ones it's pointless
      and ends up oopsing.
      Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
      cd27e455
  15. 08 10月, 2016 1 次提交
  16. 28 9月, 2016 1 次提交
  17. 27 9月, 2016 3 次提交
  18. 26 9月, 2016 2 次提交