On 09/09/2015 13:41, David M. Lloyd wrote:
If you have access to a Windows development environment, it seems to
me that you could help directly with the testing required to determine
a solution.
I think my proposed remapping approach will work, but it is contingent
on testing the following on Windows:
* Determine if it is possible to remap the pages previously mapped to
a file view without first unmapping them. Preferably, the pages would
be remapped as inaccessible (equivalent to POSIX PROT_NONE).
* Determine if such remapping releases the file to be deleted, or, if
not, determine if the previously mapped file view can be unmapped
without affecting the remapped pages.
If the answer to both of these can be shown to be affirmative, then I
think there is a real viable solution which allows immediate release
of backing resources, with the address space being reclaimed by GC.
There are a few contributions from Microsoft on this mailing list and
they might be able to jump in on this discussion. When we last explored
the remapping approach (about 5 years ago) then we didn't find a way to
do it atomically so we gave up. It's good to explore this again with
fresh eyes.
-Alan
