Re: misc/50754: src/sys/arch/epoc32/stand/e32boot/ldd/epoc32.cpp:62: shallow copy or deep copy ?

[David Holland <dholland-bugs%netbsd.org@localhost>]

The following reply was made to PR misc/50754; it has been noted by GNATS.

From: David Holland <dholland-bugs%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: misc/50754: src/sys/arch/epoc32/stand/e32boot/ldd/epoc32.cpp:62:
 shallow copy or deep copy ?
Date: Tue, 31 May 2016 02:23:55 +0000

 On Wed, Feb 03, 2016 at 09:50:00AM +0000, dcb314%hotmail.com@localhost wrote:
  > [src/sys/arch/epoc32/stand/e32boot/ldd/epoc32.cpp:62]: (style) Value of pointer 'cpu', which points to allocated memory, is copied in copy constructor instead of allocating new memory.
  > 
  > For objects with pointers to other objects, it is usual to prefer
  > deep copy to shallow copy.
 
 Perhaps in CS101 style-checker land. In real programs, structure
 members (er, pardon me, fields of objects) that point to things may be
 part of the same object (in which case cloning them along with
 everything else is usually the right thing) or references to external
 objects (in which case cloning them willy-nilly is almost always
 wrong) or, as appears to be the case this time, global references to
 singletons, in which case copying them is certainly wrong.
 
 This isn't up to the usual standards of your checker and your
 reporting...
 
 (also, 5 seconds of grepping shows that the offending copy constructor
 is not used)
 
 -- 
 David A. Holland
 dholland%netbsd.org@localhost