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: [Patch] Fix c++/PR42844


> --- gcc/testsuite/ChangeLog ? ? (revision 158206)
> +++ gcc/testsuite/ChangeLog ? ? (working copy)
> @@ -1,7 +1,8 @@
> -2010-04-11 ?Eric Botcazou ?<ebotcazou@adacore.com>
> -
> - ? ? ? * gnat.dg/pack9.adb: Remove -cargs option.
> - ? ? ? * gnat.dg/aggr12.ad[sb]: New test.
> +2010-04-11 ?Fabien Chêne ?<fabien.chene@gmail.com>
> + ? ? ? * g++.dg/init/pr42844.C: New.
> + ? ? ? * g++.dg/cpp0x/defaulted2.C: Adjust.
> + ? ? ? * g++.dg/tree-ssa/pr27549.C: Likewise.
> + ? ? ? * g++.old-deja/g++.mike/dyncast8.C: Likewise.
>
>
> It is better to not send the changelog as a patch because:
>
> 1. It will not apply cleanly in most cases.
> 2. It may overwrite something else like in the above case.

Well, then what is the right way to post patches ?
Should I write the Changelog out-of-tree, and do a manual diff;
Then post 'svn diff plus manual changelog diff' ?

> The wording should be "user-provided" because it is an adjective.
>
> I am not sure the message is clear enough for the user to identify and
> fix the problem.

FYI, instead of user-provided default constructor, Commeau, gives the
following error message: no explicitly declared default constructor.
Is it better ? Any better idea ?

-- 
Fab


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