[PATCH] Teach DeadArgElimination not to eliminate return values of functions with 'returned' arguments

[Bill Wendling]

On Jun 27, 2013, at 4:07 PM, Stephen Lin <swlin at post.harvard.edu> wrote:

>> I don't object anymore. I still think you are weakening the canonical form,
>> but I see the practical reasons for it.
>> 
> 
> Thanks!
> 
> And we can still change it if you, Nick, or anyone else comes up with
> an alternative proposal and the community comes to a consensus that it
> is better, no problem.
> 
Not commenting on the thread, but just saying that the patch LGTM. :-)

-bw