Merge stable-1.6 into master

fix make warnings about duplicated targets

obj: undo/redo user buffers

Contributors:
    Piotr Balcer (@pbalcer)
    Rafal Borawski (@rborawski)
    Jan Michalski (@janekmi)
    Weronika Lewandowska (@wlemkows)
    Marcin Slusarz (@marcinslusarz)

Contributors:
    Piotr Balcer (@pbalcer)
    Rafal Borawski (@rborawski)
    Jan Michalski (@janekmi)
    Weronika Lewandowska (@wlemkows)
    Marcin Slusarz (@marcinslusarz)

test: enable running bad block tests on real pmem

test: fix printing length of bad blocks in case of DAX devices

Merge stable-1.5 into stable-1.6

test (py): handle tests inheritance

Tx add range bug

test (py): generalize obj_basic_integration

Removes make warnings, caused by e84ddb91

common: detect libunwind on Travis

"1" is not a valid value for USE_LIBUNWIND. Instead of providing
the correct value, let pmdk build system detect whether it is available
or not. All distros we test on Travis ship with libunwind that
supports pkg-config.

 obj: prevent ulog_foreach from ever reading uninitialized data 

test: fix memcheck.log.match in obj_tx_add_range(_direct) tests

This patch changes snapshotting so that memory that's directly
after the snapshot that is being created will be zeroed. This
ensures that it's safe to read later on (during abort).

On some systems (for example OpenSUSE) memcheck generates errors
which we are suppressing. This patch adds optional lines to
memcheck.log.match to match those suppressed errors.

Fixes test introduced in: https://github.com/pmem/pmdk/pull/3921

test: force-disable pmemcheck in a few pmempool_transform tests

Merge stable-1.6 into master

Merge stable-1.5 into stable-1.6

test: print length of bad blocks in blocks

test: output/error stream redirection compatibility fix 

[1.7] test: obj_pmalloc_mt persistency fix

test: obj_ulog_size/ fix stack problem

Update PACKAGE_MAINTAINER

added a new method tx_merge_flags to merge flags for two ranges which
are being merged into one

... so that the tests were correct also in case of real hardware.

Ref: pmem/issues#1047

Ref: pmem/issues#1121

Ref: pmem/pmdk#3911
Ref: pmem/issues#1118