Apparent memory leak on linux platforms

[Martyn Gigg]

Run the script below with the memory manager open, noting the memory at the start. At then end of the script, even though all workspaces have been deleted, the memory is not back to where it started.

_range = range(0, 3)
Load('WISH00016748.raw',OutputWorkspace='WISH00016748')

for i in _range:
        CloneWorkspace(InputWorkspace='WISH00016748', OutputWorkspace=str(i))
        
for j in _range:
        DeleteWorkspace(Workspace=str(j))
        
DeleteWorkspace('WISH00016748')

[Martyn Gigg]

It looks as if this is yet again down to more complex memory management going on NOT an actual leak. If you run another algorithm immediately afterwards then the memory is reported correctly back to the OS and that algorithm doesn't use additional memory.

It is confusing for users though so add a releaseFreeMemory call at the end of script execution.

[Martyn Gigg]

Add releaseFreeMemory call when script stops. Refs #6081

Changeset: 6e5a581132a8c9fab905adce6c68e08d37de0512

[Owen Arnold]

This works all memory appears to be recovered. Compared to the last nightly build that is not returning memory to the system, this is a big improvement.

[Martyn Gigg]

Add releaseFreeMemory call when script stops. Refs #6081

Changeset: 6e5a581132a8c9fab905adce6c68e08d37de0512

[Martyn Gigg]

Add releaseFreeMemory call when script stops. Refs #6081

Changeset: 6e5a581132a8c9fab905adce6c68e08d37de0512

[Stuart Campbell]

This ticket has been transferred to github issue ​6927