[llvm-dev] Difficulty in Porting cloneLoop in LoopUnswitch

[Jingu Kang via llvm-dev]

Hi Raghesh,

I have also ported code for partially invariant condition from LoopUnswitch to SimpleLoopUnswitch. https://reviews.llvm.org/D99354 The SimpleLoopUnswitch clones unswitchable path of loop rather than cloning whole loop but the main logic is similar. It could be better to port your code to SimpleLoopUnswitch because the new pass manager is default.

Regards
JinGu Kang

From: llvm-dev <llvm-dev-bounces at lists.llvm.org> On Behalf Of raghesh via llvm-dev
Sent: Thursday, April 8, 2021 8:17 AM
To: Arthur Eubanks <aeubanks at google.com>
Cc: LLVM Developers Mailing List <llvm-dev at lists.llvm.org>
Subject: Re: [llvm-dev] Difficulty in Porting cloneLoop in LoopUnswitch

 Hi Arthur,

There are significant amounts of code that we added internally to LoopUnswitch. I presume we will be able to move that to SimpleUnswitch, but in the interest of time we thought of porting LoopUnswitch first.

Regards,
Raghesh
------------------------------
Raghesh Aloor
AMD India Pvt. Ltd.
Bengaluru.
------------------------------


On Thu, Apr 8, 2021 at 12:10 PM Arthur Eubanks <aeubanks at google.com<mailto:aeubanks at google.com>> wrote:
The new PM's equivalent of LoopUnswitch is the (not super well named) SimpleLoopUnswitch. Is there something specific that SimpleLoopUnswitch doesn't handle?

On Wed, Apr 7, 2021 at 11:36 PM raghesh via llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>> wrote:
Hi All,

Though the LoopUnswitch pass was not ported to the new pass manager by community,  we internally at AMD are in need to do that. While doing this we are facing issues to have a counterpart for the cloneLoop function, the body of which is given below.

Loop *llvm::cloneLoop(Loop *L, Loop *PL, ValueToValueMapTy &VM,
                      LoopInfo *LI, LPPassManager *LPM) {
  Loop &New = *LI->AllocateLoop();
  if (PL)
    PL->addChildLoop(&New);
  else
    LI->addTopLevelLoop(&New);

  if (LPM)
    LPM->addLoop(New);

  // Add all of the blocks in L to the new loop.
  for (Loop::block_iterator I = L->block_begin(), E = L->block_end();
       I != E; ++I)
    if (LI->getLoopFor(*I) == L)
      New.addBasicBlockToLoop(cast<BasicBlock>(VM[*I]), *LI);

  // Add all of the subloops to the new loop.
  for (Loop *I : *L)
    cloneLoop(I, &New, VM, LI, LPM);

  return &New;
}

In the new pass manager flow, we are trying to use LPMUpdater's addSiblingLoops in place of addLoop, which will be invoked only for the top level loop (loop to be unswitched). According to the comments given in AddSiblingLoops the loops nested inside will be traversed by the pass manager anyway. In that assumption we are not adding the subloops. With this we are facing assertion failures further.

We also tried capturing all the newly created loops and adding it towards the end of the LoopUnswitch pass, as done in the SimpleLoopUnswitch pass. That also did not help.

Another approach we tried is not to add the cloned loops at all to the pass manager.. This worked  but not sure about its implications. We might miss further optimization opportunities on these loops.

It would be a great help if someone could point out the right way to go about it.

Regards,
Raghesh
------------------------------
Raghesh Aloor
AMD India Pvt. Ltd.
Bengaluru.
------------------------------
_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20210408/43d0e6d2/attachment-0001.html>