Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.lang.c++ > #9295
| From | Andrey Tarasevich <andreytarasevich@hotmail.com> |
|---|---|
| Newsgroups | comp.lang.c++ |
| Subject | Re: Fun with member-function pointers |
| Date | 2011-08-24 09:30 -0700 |
| Organization | A noiseless patient Spider |
| Message-ID | <j338vq$krh$1@dont-email.me> (permalink) |
| References | (2 earlier) <9bhvopFfssU1@mid.individual.net> <j31l4o$iso$2@dont-email.me> <9bkgdaF95nU1@mid.individual.net> <j335g1$qs3$1@dont-email.me> <9bkmd6Fm6pU1@mid.individual.net> |
On 8/24/2011 9:15 AM, Default User wrote: > >> Again, beware of broken compilers that might not handle it correctly. But >> for your specific example I don't expect any surprises from the compiler. > > VS 2010 builds the example without complaint. Results are what I would want. Well, VS is actually the problematic one. It has a bunch of settings (see '/vms', '/vmm' and '/vmv') intended to control the trade-off between compiler compliance and member-function-pointer size/simplicity in this specific kind of situation (depending on the inheritance model). However, somehow they managed to let the Marcel's example slip through. When the VMT pointer is introduced into the class layout late in the hierarchy (i.e. the base class is not polymorphic, while the derived class is), the compiler fails to produce correct code. It is especially strange in light of the fact that all the infrastructure that is needed to handle this situation correctly is already built into the compiler (layout change is similar to that caused by ordinary multiple inheritance). Yet that managed to screw this specific case up. It is broken in VS 2005. It remains broken is VS 2010. Of course, one can work around this issue fairly easily by moving the VMT pointer all the way up the hierarchy, i.e. by forcefully making the base class polymorphic (not even mentioning that it will already be the case in [almost] all practical designs). Nevertheless, it is important to be aware of it. -- Best regards, Andrey Tarasevich
Back to comp.lang.c++ | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Fun with member-function pointers "Default User" <defaultuserbr@yahoo.com> - 2011-08-22 16:41 -0500
Re: Fun with member-function pointers Victor Bazarov <v.bazarov@comcast.invalid> - 2011-08-22 19:25 -0400
Re: Fun with member-function pointers "Default User" <defaultuserbr@yahoo.com> - 2011-08-22 19:06 -0500
Re: Fun with member-function pointers Juha Nieminen <nospam@thanks.invalid> - 2011-08-23 07:09 +0000
Re: Fun with member-function pointers Victor Bazarov <v.bazarov@comcast.invalid> - 2011-08-23 08:04 -0400
Re: Fun with member-function pointers Juha Nieminen <nospam@thanks.invalid> - 2011-08-23 13:34 +0000
Re: Fun with member-function pointers Victor Bazarov <v.bazarov@comcast.invalid> - 2011-08-23 13:02 -0400
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-23 19:12 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-23 18:51 -0700
Re: Fun with member-function pointers Marcel Müller <news.5.maazl@spamgourmet.com> - 2011-08-23 09:07 +0200
Re: Fun with member-function pointers Juha Nieminen <nospam@thanks.invalid> - 2011-08-23 10:09 +0000
Re: Fun with member-function pointers Marcel Müller <news.5.maazl@spamgourmet.com> - 2011-08-23 15:04 +0200
Re: Fun with member-function pointers "Default User" <defaultuserbr@yahoo.com> - 2011-08-23 10:36 -0500
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-23 18:46 -0700
Re: Fun with member-function pointers "Default User" <defaultuserbr@yahoo.com> - 2011-08-24 09:33 -0500
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-24 08:32 -0700
Re: Fun with member-function pointers "Default User" <defaultuserbr@yahoo.com> - 2011-08-24 11:15 -0500
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-24 09:30 -0700
Re: Fun with member-function pointers Marcel Müller <news.5.maazl@spamgourmet.com> - 2011-08-24 12:27 +0200
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-24 06:46 -0700
Re: Fun with member-function pointers Marcel Müller <news.5.maazl@spamgourmet.org> - 2011-08-24 18:57 +0200
Re: Fun with member-function pointers Paul <pchristor@yahoo.co.uk> - 2011-08-24 13:08 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-24 13:44 -0700
Re: Fun with member-function pointers Paul <pchristor@yahoo.co.uk> - 2011-08-25 04:23 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-25 07:31 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-25 08:21 -0700
Re: Fun with member-function pointers Paul <pchristor@yahoo.co.uk> - 2011-08-25 10:50 -0700
Re: Fun with member-function pointers Leigh Johnston <leigh@i42.co.uk> - 2011-08-25 13:57 +0100
Re: Fun with member-function pointers Leigh Johnston <leigh@i42.co.uk> - 2011-08-25 14:08 +0100
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-25 07:32 -0700
Re: Fun with member-function pointers Leigh Johnston <leigh@i42.co.uk> - 2011-08-25 15:45 +0100
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-25 07:48 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-23 18:43 -0700
Re: Fun with member-function pointers Paul <pchristor@yahoo.co.uk> - 2011-08-23 03:49 -0700
Re: Fun with member-function pointers Paul <pchristor@yahoo.co.uk> - 2011-08-23 06:44 -0700
Re: Fun with member-function pointers Andrey Tarasevich <andreytarasevich@hotmail.com> - 2011-08-23 18:28 -0700
Re: Fun with member-function pointers Jeff Flinn <TriumphSprint2000@hotmail.com> - 2011-08-25 11:29 -0400
csiph-web