Marcin Ślusarz

I think we should merge this as is and work on other issues in a new bug report.

Please post cmake log, distribution version, libcapstone-dev package version, output of: ``` dpkg -L libcapstone-dev ``` and ``` cat `dpkg -L libcapstone-dev | grep pc$` ```

```ctest --output-on-failure``` will tell you why it failed.

I think we should replace $(XX) with (null) wherever we expect null address. Some time ago we used to print 0x0 instead of (null) for null pointers (see commit description...

Only operators that have "|" in the name are treated like that - "|" is an event separator in pmemcheck output and nobody has thought "|" could be part of...

*[`src/test/unittest/unittest.h` line 784 at r31](https://reviewable.io/reviews/pmem/pmdk/5475#-NAK75jd-ikBsCM1rBWb:-NAK75je8MF5xJrtrTbC:b-fzy9zz) ([raw file](https://github.com/pmem/pmdk/blob/4ebfd5acfe5d98762e61cf9bd49885218855e86e/src/test/unittest/unittest.h#L784)):* > ```objc > char *token; > > while ((token = strtok_r(tmp, delim, &tmp))) { > ``` Heh, I should have noticed this bug...

I agree with Kamil. I think pmempool_rm should verify that it knows what it's unlinking. So if file doesn't look like poolset or part with a header it should refuse...

See also: https://bugs.launchpad.net/ubuntu/+bug/1752378 (comment 15 and later)

Even simpler approach: Add new flag to pmemobj_tx_xadd_range[_direct] : POBJ_XADD_NO_DRAIN. Application would call pmemobj_tx_xadd_range multiple times with this flag and for the last region would call pmemobj_tx_xadd_range without this flag...

1. Can you explain how did you measure this? 2. and 3. look more like a request for dynamic/advanced space management in obj... 3. You can use allocations classes to...