[PATCHES] A module inliner pass with a greedy call site queue

[Hal Finkel]

----- Original Message -----
> From: "Yin Ma" <yinma at codeaurora.org>
> To: "Chandler Carruth" <chandlerc at google.com>
> Cc: "Jiangning Liu" <liujiangning1 at gmail.com>, "Gerolf Hoflehner" <ghoflehner at apple.com>, "Nick Lewycky"
> <nicholas at mxc.ca>, "James Molloy" <james.molloy at arm.com>, "Hal Finkel" <hfinkel at anl.gov>, "Commit Messages and
> Patches for LLVM" <llvm-commits at cs.uiuc.edu>
> Sent: Wednesday, August 13, 2014 5:26:52 PM
> Subject: RE: [PATCHES] A module inliner pass with a greedy call site queue
>
> Hi Chandler,
> 
> 
> 
> I am not familiar how to do consensus thing for merging. What is the
> good
> 
> way to do it?

First, have you submitted a proposal for a BoF at the developers' meeting? We need to make sure that is done.  Second, I think that we need to collect pertinent test cases (where we understand the behavior of current implementations and the desired behavior) and lay out semantic requirements (like insensitivity to unrelated module changes). I think it will be hard to make progress here without these things documented.

Thanks again,
Hal

> 
> 
> 
> Thanks,
> 
> 
> 
> Yin
> 
> 
> 
> From: Chandler Carruth [mailto:chandlerc at google.com]
> Sent: Tuesday, August 12, 2014 1:19 PM
> To: Yin Ma
> Cc: Jiangning Liu; Gerolf Hoflehner; Nick Lewycky; James Molloy; Hal
> Finkel; Commit Messages and Patches for LLVM
> Subject: Re: [PATCHES] A module inliner pass with a greedy call site
> queue
> 
> 
> 
> 
> 
> 
> 
> 
> On Tue, Aug 12, 2014 at 1:16 PM, Yin Ma < yinma at codeaurora.org >
> wrote:
> 
> 
> 
> Hi Everybody,
> 
> 
> 
> It has been several weeks after I submitted this greedy inliner
> patch. Some people like
> 
> and some hesitate. This inliner has shown promising potentials in
> various situations and
> 
> default off. I would like to know whether we can merge this into LLVM
> at first so the people
> 
> who like it can continue tuning it up? For people who still has
> question about what is the
> 
> right inliner, I am about to submit a BOF proposal to discuss this
> issue in the coming LLVM
> 
> dev conference. How do you guys think?
> 
> 
> I don't think we should merge this until there is some consensus. I
> think the BOF is already moving in the right direction to try to
> find it.

-- 
Hal Finkel
Assistant Computational Scientist
Leadership Computing Facility
Argonne National Laboratory