Hi,
>> Or you mean something else?
>
>I was thinking that the recursive part could be a simple loop to set
>access, but your way is fine too.
Ok, great. Note, before committing I mean to also simplify it a bit, the
TREE_STATIC check of the recursive part has no reason to exist, doesn't exist
On 09/04/2013 10:42 AM, Paolo Carlini wrote:
Indeed. I think the recursive part already does that, because only the
first time is called complain == true (thus does exactly what the
current code does), then when the recursion proper starts, complain ==
false.
Ah yes, I see.
Or you mean someth
Hi,
On 09/04/2013 03:11 PM, Jason Merrill wrote:
It looks to me like this will result in duplicate diagnostics for
invalid members in a nested anonymous union. Maybe make the recursive
part only handle access setting?
Indeed. I think the recursive part already does that, because only the
firs
It looks to me like this will result in duplicate diagnostics for
invalid members in a nested anonymous union. Maybe make the recursive
part only handle access setting?
Jason
... assuming the general idea makes sense, this version may be better
because, at the cost of 3 lines of code duplication, it cuts the
unnecessary function calls, eg exactly zero if the struct doesn't have
anonymous aggregates at all. The patch is also easier to read ;)
Booted and tested x86_6
