9fb074e7bb
Current approach doesn't work well in cases when multiple paths are predicted to be "cold". By "cold" paths I mean those containing "unreachable" instruction, call marked with 'cold' attribute and 'unwind' handler of 'invoke' instruction. The issue is that heuristics are applied one by one until the first match and essentially ignores relative hotness/coldness of other paths. New approach unifies processing of "cold" paths by assigning predefined absolute weight to each block estimated to be "cold". Then we propagate these weights up/down IR similarly to existing approach. And finally set up edge probabilities based on estimated block weights. One important difference is how we propagate weight up. Existing approach propagates the same weight to all blocks that are post-dominated by a block with some "known" weight. This is useless at least because it always gives 50\50 distribution which is assumed by default anyway. Worse, it causes the algorithm to skip further heuristics and can miss setting more accurate probability. New algorithm propagates the weight up only to the blocks that dominates and post-dominated by a block with some "known" weight. In other words, those blocks that are either always executed or not executed together. In addition new approach processes loops in an uniform way as well. Essentially loop exit edges are estimated as "cold" paths relative to back edges and should be considered uniformly with other coldness/hotness markers. Reviewed By: yrouban Differential Revision: https://reviews.llvm.org/D79485
54 lines
1.7 KiB
LLVM
54 lines
1.7 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-unknown < %s | \
|
|
; RUN: FileCheck %s --implicit-check-not=mtctr --implicit-check-not=bdnz
|
|
$test = comdat any
|
|
|
|
; No CTR loop due to frem (since it is always a call).
|
|
define void @test() nounwind comdat {
|
|
; CHECK-LABEL: test:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: mflr 0
|
|
; CHECK-NEXT: std 29, -24(1) # 8-byte Folded Spill
|
|
; CHECK-NEXT: std 30, -16(1) # 8-byte Folded Spill
|
|
; CHECK-NEXT: std 0, 16(1)
|
|
; CHECK-NEXT: stdu 1, -64(1)
|
|
; CHECK-NEXT: ld 29, 0(3)
|
|
; CHECK-NEXT: ld 30, 32(1)
|
|
; CHECK-NEXT: cmpld 30, 29
|
|
; CHECK-NEXT: bge- 0, .LBB0_2
|
|
; CHECK-NEXT: .p2align 5
|
|
; CHECK-NEXT: .LBB0_1: # %bounds.ok
|
|
; CHECK-NEXT: #
|
|
; CHECK-NEXT: lfsx 2, 0, 3
|
|
; CHECK-NEXT: xxlxor 1, 1, 1
|
|
; CHECK-NEXT: bl fmodf
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: addi 30, 30, 1
|
|
; CHECK-NEXT: stfsx 1, 0, 3
|
|
; CHECK-NEXT: cmpld 30, 29
|
|
; CHECK-NEXT: blt+ 0, .LBB0_1
|
|
; CHECK-NEXT: .LBB0_2: # %bounds.fail
|
|
; CHECK-NEXT: std 30, 32(1)
|
|
%pos = alloca i64, align 8
|
|
br label %forcond
|
|
|
|
forcond: ; preds = %bounds.ok, %0
|
|
%1 = load i64, i64* %pos
|
|
%.len1 = load i64, i64* undef
|
|
%bounds.cmp = icmp ult i64 %1, %.len1
|
|
br i1 %bounds.cmp, label %bounds.ok, label %bounds.fail
|
|
|
|
bounds.ok: ; preds = %forcond
|
|
%2 = load float, float* undef
|
|
%3 = frem float 0.000000e+00, %2
|
|
store float %3, float* undef
|
|
%4 = load i64, i64* %pos
|
|
%5 = add i64 %4, 1
|
|
store i64 %5, i64* %pos
|
|
br label %forcond
|
|
|
|
bounds.fail: ; preds = %forcond
|
|
unreachable
|
|
}
|
|
|