Re: [Merge] lp:~johill-lanl/epics-base/epics-base-enclosure-of into lp:epics-base

[Jeff Hill <johill@lanl.gov>]

> Using reinterpret_cast<> to bypass these checks seems questionable to me. I'm worried 
> that this might involve Undefined Behavior.

This is certainly a reasonable point. My restricted use of enclosureOf (i.e. only with single inheritance c++ classes) would probably never actually cause troubles with compiler portability or undefined behavior, but now I am considering this question; why would we accept any risk at all in this area.

At the root of this functionality is a desire to reduce the size of the data structures; we might avoid storing a pointer to the enclosing class within interface adapter classes. However, given that only large memory chips are manufactured today then perhaps its best to just store the additional pointer.

Thanks for your review.

PS: The latest C++ standard is actually relaxing its restrictions related to what is and isn't considered to be POD, BTW.



-- 
https://code.launchpad.net/~johill-lanl/epics-base/epics-base-enclosure-of/+merge/196010
Your team EPICS Core Developers is requested to review the proposed merge of lp:~johill-lanl/epics-base/epics-base-enclosure-of into lp:epics-base.