[cfe-commits] PATCH: Differentiate between definitely and possibly uninitialized warnings

Chris Lattner clattner at apple.com
Thu Apr 7 13:37:39 PDT 2011

Yes, the diagnostics you changed sound good.  I'm just asking for the next step :)


On Apr 6, 2011, at 11:50 PM, Chandler Carruth <chandlerc at google.com> wrote:

> On Wed, Apr 6, 2011 at 11:13 PM, Chris Lattner <clattner at apple.com> wrote:
> Looks good to me.  Please also change "is possibly uninitialized" to "may be uninitialized" while you're at it, thanks!
> On some further discussion, I think I understand what these messages were aiming at: the 'possibly' isn't about whether we've proven there to be a path leading to uninitialized use, but the fact that we haven't proven *all* paths to cause an uninitialized use....
> I still want some different wording here. Do you think the wording I proposed in the patch is actually OK for this? Is there a better way to phrase it?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20110407/1ea8d547/attachment.html>

More information about the cfe-commits mailing list