[PATCH 0/5] btrfs decoding fixes

[jeffm at suse.com]

From: Jeff Mahoney <jeffm at suse.com>

After submitting the last patch, I hunted down the last of the test case
failures.  These are all failures in the output side of a RW ioctl and
those test cases weren't easily runnable.  They are now, provided
a btrfs file system is made available to them via a BTRFS_MOUNTPOINT
environment variable during testing.

Three of the print_array conversions introduced bugs.  The search tree
buffer isn't an array at all -- it's a series of variable length
records, each with a header.  The btrfs_data_container element count
refers to the number of 64-bit items rather than the number of
3*64-bit-sized records.

There was a stray "..." instead of printing the search buffer output
in the BTRFS_IOC_TREE_SEARCH case and some broken whitespace in the _V2
case.

-Jeff

---
Jeff Mahoney (5):
  btrfs.c: fix print_array usage in btrfs_print_logical_ino_container
  btrfs.c: don't use print_array in btrfs_print_tree_search
  tests/btrfs.c: fix live btrfs_test_search_ioctls output
  tests: check decoding of btrfs RDWR ioctls returned data
  file_ioctl.c: fix FIDEDUPERANGE output

 btrfs.c             | 44 +++++++++++++++++++++++---------------------
 file_ioctl.c        |  2 +-
 tests/Makefile.am   |  2 ++
 tests/btrfs-vw.test | 29 +++++++++++++++++++++++++++++
 tests/btrfs-w.test  | 29 +++++++++++++++++++++++++++++
 tests/btrfs.c       |  7 +++++--
 6 files changed, 89 insertions(+), 24 deletions(-)
 create mode 100755 tests/btrfs-vw.test
 create mode 100755 tests/btrfs-w.test

-- 
2.7.1