This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [C++ PATCH]: Rationalize build_compound_expr


On Fri, 2003-07-25 at 08:21, Nathan Sidwell wrote:
> Hi,
> I started looking at PR 11512 and decided that it was time
> build_compound_expr took sensible parameters. It is silly to build
> TREE_LISTs merely to pull them appart again. In doing so, I came across
> the following
> 1) Some unreachable code in build_compound_expr to do with NOP stripping
> -- that is looking at the type of the *list*, not its TREE_VALUE
> 2) Code like 'int (1, 2)' pedwarns about treating it as a compound
> expression, but does not perform overload resolution on it - this patch
> does so consistently.
> 3) One of those cases with initializing a RESULT_DECL was a mere warning
> it is now a pedwarn
> 4) The case for a reference init was completely silent about this. Now
> it is not.
> 
> booted & tested on i686-pc-linux-gnu, ok?

Yes, this is a long-standing wart, nice to see it go!

-- 
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]