[LLVMdev] [RFC] LCSSA vs. SSAUpdater ... FIGHT!

[Andrew Trick]

On Feb 11, 2014, at 5:03 AM, Chandler Carruth <chandlerc at google.com> wrote:

> > I still don’t particularly like that we force all LLVM clients to perform LCSSA when all they end up doing is rotating and simplifying loops (no LICM/unroll). So it is a tradeoff.
> 
> Again, LoopSimplify does not require LCSSA today. Nothing to do there. If we folded rotate into simplify (which we should probably do) then we would be done.

That’s a good point, and I’m totally on board with your massive simplification of the loop passes. Just for the record, the disadvantage that I’m pointing out is that LCSSA will be computed for all loops even is the code has no LICM opportunities. As opposed to the SSAUpdater approach which simply does nothing if there’s nothing to do.
-Andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140212/bd4b3929/attachment.html>