This is the mail archive of the
mailing list for the GDB project.
Re: [patch] fix exp/12117
On Wed, Oct 13, 2010 at 11:44 AM, Tom Tromey <email@example.com> wrote:
>>>>>> "Doug" == Doug Evans <firstname.lastname@example.org> writes:
> Doug> This patch fixes pr exp/12117.
> Doug> http://sourceware.org/bugzilla/show_bug.cgi?id=12117
> Doug> I will check this in in two days if there are no objections.
> It looks reasonable to me.
> Not an objection, but it seems odd that check_typedef can strip
> qualifiers from a type. ?It looks like it tries to do the right thing in
> some cases, but not all. ?Couldn't it just call make_qualified_type at
> the right point? ?Or unconditionally use that instead of checking
No disagreement that it's odd, and there may yet be more bugs in this area.
I thought of calling make_qualified_type, but I wasn't entirely
comfortable with it.
a) it seems like it's not just c/v, e.g., it's also the address space
[and perhaps here's a case where there are more bugs in this area :-)]
b) it seems odd to have to build such types on the fly
[I can imagine a proliferation of such objects that aren't attached to
anything concrete in the debug info, and just cause confusion]