[cfe-commits] [patch] Warn on self assignment for member variables

Ted Kremenek kremenek at apple.com
Mon Jun 25 13:14:13 PDT 2012


On Jun 25, 2012, at 12:42 PM, Nico Weber <thakis at chromium.org> wrote:

> Hi Ted,
> 
> On Mon, Jun 25, 2012 at 12:40 PM, Ted Kremenek <kremenek at apple.com> wrote:
>> Hi Nico,
>> 
>> How expensive is this check?  This code makes me nervous:
> 
> I didn't measure, but note that this is only executed if the lhs and
> rhs are both MemberExprs, and they both refer to the same Decl. This
> isn't true for most assignments.

Makes sense.

> 
> Is it enough to do measure the impact on a full build of chromium, or
> would you like to see more data?

Measuring against full build of Chromium seems reasonable to me.  It would be nice for us to establish a common set of representative codebases to do this kind of performance measurement on.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20120625/83d13190/attachment.html>


More information about the cfe-commits mailing list