This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [patchv2] compile: Fix crash on cv-qualified self-reference
- From: Jan Kratochvil <jan dot kratochvil at redhat dot com>
- To: Pedro Alves <palves at redhat dot com>
- Cc: Yao Qi <qiyaoltc at gmail dot com>, gdb-patches at sourceware dot org, Phil Muldoon <pmuldoon at redhat dot com>
- Date: Wed, 1 Jul 2015 17:12:02 +0200
- Subject: Re: [patchv2] compile: Fix crash on cv-qualified self-reference
- Authentication-results: sourceware.org; auth=none
- References: <20150418172843 dot GA17777 at host1 dot jankratochvil dot net> <20150516132555 dot GB17266 at host1 dot jankratochvil dot net> <86lhf0p1hf dot fsf at gmail dot com> <20150701132406 dot GA13975 at host1 dot jankratochvil dot net> <5593F10D dot 4020903 at redhat dot com> <20150701141003 dot GA19545 at host1 dot jankratochvil dot net> <55940059 dot 30603 at redhat dot com>
On Wed, 01 Jul 2015 16:59:37 +0200, Pedro Alves wrote:
> Well, I'd say that having the core GDB Type be aware of GccType
> directly would be a misdesign, not a feature.
You are right, it would be reversed, anyway not the insance crap of
TYPE_CODE_*.
> So a trivial 1-1 conversion or your code would be:
I find it complicates the code even more without C++.
> I'm very much looking forward to C++ as well,
So why that does not happen? You have added a disadvantage of more
requirements for the code compliance with the new exception macros but there
is still no advantage from that.
> but switch/case vs
> virtual methods here is really more about syntax sugar than design.
C++ may look so to someone but is not just a syntax sugar.
Jan