[all-commits] [llvm/llvm-project] d84b32: [MacroFusion] Limit the max fused number as 2 to r...

[QingShan Zhang via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: d84b320dfd0a7dbedacc287ede5e5bc4c0f113ba
      https://github.com/llvm/llvm-project/commit/d84b320dfd0a7dbedacc287ede5e5bc4c0f113ba
  Author: QingShan Zhang <qshanz at cn.ibm.com>
  Date:   2019-12-04 (Wed, 04 Dec 2019)

  Changed paths:
    M llvm/lib/CodeGen/MacroFusion.cpp
    M llvm/test/CodeGen/AArch64/macro-fusion.ll

  Log Message:
  -----------
  [MacroFusion] Limit the max fused number as 2 to reduce the dependency

This is the example:

int foo(int a, int b, int c, int d) {
  return a + b + c + d;
}

And this is the Dependency Graph:
+------+       +------+       +------+       +------+
|  A   |       |  B   |       |  C   |       |  D   |
+--+--++       +---+--+       +--+---+       +--+---+
   ^  ^            ^  ^          ^              ^
   |  |            |  |          |              |
   |  |            |  |New1      +--------------+
   |  |            |  |          |
   |  |            |  |       +--+---+
   |  |New2        |  +-------+ ADD1 |
   |  |            |          +--+---+
   |  |            |    Fuse     ^
   |  |            +-------------+
   |  +------------+
   |               |
   |   Fuse     +--+---+
   +----------->+ ADD2 |
   |            +------+
+--+---+
| ADD3 |
+------+

We need also create an artificial edge from ADD1 to A if
https://reviews.llvm.org/D69998 is landed. That will force the Node A scheduled
before the ADD1 and ADD2. But in fact, it is ok to schedule the Node A
in-between ADD3 and ADD2, as ADD3 and ADD2 are NOT a fusion pair because
ADD2 has been matched to ADD1. We are creating these unnecessary dependency
edges that override the heuristics.

Differential Revision: https://reviews.llvm.org/D70066