Google Git
Sign in
android / platform / external / tensorflow / 133f11f1ced / . / test / Transforms / loop-fusion.mlir
blob: 2458049f22c207f83c996a62eadf01564df595aa [file] [log] [blame]
// RUN: mlir-opt %s -loop-fusion -split-input-file -verify | FileCheck %s
// TODO(andydavis) Add more tests:
// *) Add nested fusion test cases when non-constant loop bound support is
// added to iteration domain in dependence check.
// *) Add a test w/ floordiv/ceildiv/mod when supported in dependence check.
// *) Add tests which check fused computation slice indexing and loop bounds.
// TODO(andydavis) Test clean up: move memref allocs to func args.
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_raw_dep_for_locality() {
func @should_fuse_raw_dep_for_locality() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %m[%i1] : memref<10xf32>
}
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %0[%1] : memref<1xf32>
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %3 = load %0[%2] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_reduction_to_pointwise() {
func @should_fuse_reduction_to_pointwise() {
%a = alloc() : memref<10x10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
for %i1 = 0 to 10 {
%v0 = load %b[%i0] : memref<10xf32>
%v1 = load %a[%i0, %i1] : memref<10x10xf32>
%v3 = addf %v0, %v1 : f32
store %v3, %b[%i0] : memref<10xf32>
}
}
for %i2 = 0 to 10 {
%v4 = load %b[%i2] : memref<10xf32>
store %v4, %c[%i2] : memref<10xf32>
}
// Should fuse in entire inner loop on %i1 from source loop nest, as %i1
// is not used in the access function of the store/load on %b.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %4 = load %0[%3] : memref<1xf32>
// CHECK-NEXT: %5 = load %1[%i0, %i1] : memref<10x10xf32>
// CHECK-NEXT: %6 = addf %4, %5 : f32
// CHECK-NEXT: %7 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %6, %0[%7] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: %8 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %9 = load %0[%8] : memref<1xf32>
// CHECK-NEXT: store %9, %2[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-DAG: [[MAP_SHIFT_MINUS_ONE_R1:#map[0-9]+]] = (d0) -> (d0 - 1)
// CHECK-DAG: [[MAP_SHIFT_BY_ONE:#map[0-9]+]] = (d0) -> (d0 + 1)
// CHECK-DAG: [[MAP_SHIFT_MINUS_IV_R2_EVEN:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d0 + d2)
// CHECK-DAG: [[MAP_SHIFT_MINUS_IV_R2_ODD:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d1 + d3)
// CHECK-LABEL: func @should_fuse_loop_nests_with_shifts() {
func @should_fuse_loop_nests_with_shifts() {
%a = alloc() : memref<10x10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 9 {
for %i1 = 0 to 9 {
%idx = affine.apply (d0) -> (d0 + 1) (%i0)
%idy = affine.apply (d0) -> (d0 + 1) (%i1)
store %cf7, %a[%idx, %idy] : memref<10x10xf32>
}
}
for %i2 = 1 to 10 {
for %i3 = 1 to 10 {
%v0 = load %a[%i2, %i3] : memref<10x10xf32>
}
}
// Source slice affine apply sequence:
// *) First two affine apply's map from the dst to src iteration space.
// *) Third affine apply is access function around src store.
// *) Fourth affine apply shifts the stores access function by '-1', because
// of the offset induced by reducing the memref shape from 10x10 to 9x9.
// *) Fifth affine apply shifts the loads access function by '-1', because
// of the offset induced by reducing the memref shape from 10x10 to 9x9.
// NOTE: Should create a private memref with reduced shape 9x9xf32.
// CHECK: for %i0 = 1 to 10 {
// CHECK-NEXT: for %i1 = 1 to 10 {
// CHECK-NEXT: %1 = affine.apply [[MAP_SHIFT_MINUS_ONE_R1]](%i0)
// CHECK-NEXT: %2 = affine.apply [[MAP_SHIFT_MINUS_ONE_R1]](%i1)
// CHECK-NEXT: %3 = affine.apply [[MAP_SHIFT_BY_ONE]](%1)
// CHECK-NEXT: %4 = affine.apply [[MAP_SHIFT_BY_ONE]](%2)
// CHECK-NEXT: %5 = affine.apply [[MAP_SHIFT_MINUS_IV_R2_EVEN]](%i0, %i1, %3, %4)
// CHECK-NEXT: %6 = affine.apply [[MAP_SHIFT_MINUS_IV_R2_ODD]](%i0, %i1, %3, %4)
// CHECK-NEXT: store %cst, %0[%5, %6] : memref<1x1xf32>
// CHECK-NEXT: %7 = affine.apply [[MAP_SHIFT_MINUS_IV_R2_EVEN]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %8 = affine.apply [[MAP_SHIFT_MINUS_IV_R2_ODD]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %9 = load %0[%7, %8] : memref<1x1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-DAG: [[MAP_D2_D0_DIFF:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d0 + d2)
// CHECK-DAG: [[MAP_D3_D1_DIFF:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d1 + d3)
// CHECK-LABEL: func @should_fuse_loop_nest() {
func @should_fuse_loop_nest() {
%a = alloc() : memref<10x10xf32>
%b = alloc() : memref<10x10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
for %i1 = 0 to 10 {
store %cf7, %a[%i0, %i1] : memref<10x10xf32>
}
}
for %i2 = 0 to 10 {
for %i3 = 0 to 10 {
%v0 = load %a[%i3, %i2] : memref<10x10xf32>
store %v0, %b[%i2, %i3] : memref<10x10xf32>
}
}
for %i4 = 0 to 10 {
for %i5 = 0 to 10 {
%v1 = load %b[%i4, %i5] : memref<10x10xf32>
}
}
// Expecting private memref for '%a' first, then private memref for '%b'.
// CHECK-DAG: [[NEWA:%[0-9]+]] = alloc() : memref<1x1xf32>
// CHECK-DAG: [[NEWB:%[0-9]+]] = alloc() : memref<1x1xf32>
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %2 = affine.apply [[MAP_D2_D0_DIFF]](%i1, %i0, %i1, %i0)
// CHECK-NEXT: %3 = affine.apply [[MAP_D3_D1_DIFF]](%i1, %i0, %i1, %i0)
// CHECK-NEXT: store %cst, [[NEWA]][%2, %3] : memref<1x1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP_D2_D0_DIFF]](%i1, %i0, %i1, %i0)
// CHECK-NEXT: %5 = affine.apply [[MAP_D3_D1_DIFF]](%i1, %i0, %i1, %i0)
// CHECK-NEXT: %6 = load [[NEWA]][%4, %5] : memref<1x1xf32>
// CHECK-NEXT: %7 = affine.apply [[MAP_D2_D0_DIFF]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %8 = affine.apply [[MAP_D3_D1_DIFF]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: store %6, [[NEWB]][%7, %8] : memref<1x1xf32>
// CHECK-NEXT: %9 = affine.apply [[MAP_D2_D0_DIFF]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %10 = affine.apply [[MAP_D3_D1_DIFF]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %11 = load [[NEWB]][%9, %10] : memref<1x1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_across_intermediate_loop_with_no_deps() {
func @should_fuse_across_intermediate_loop_with_no_deps() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
%v0 = load %a[%i0] : memref<10xf32>
store %v0, %b[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %c[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v1 = load %b[%i2] : memref<10xf32>
}
// Should fuse first loop (past second loop with no dependences) into third.
// Note that fusion creates a private memref '%2' for the fused loop nest.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %2[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: %3 = load %1[%i1] : memref<10xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: store %3, %0[%4] : memref<1xf32>
// CHECK-NEXT: %5 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: %6 = load %0[%5] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_all_loops() {
func @should_fuse_all_loops() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
// Set up flow dependences from first and second loops to third.
for %i0 = 0 to 10 {
store %cf7, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %b[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v0 = load %a[%i2] : memref<10xf32>
%v1 = load %b[%i2] : memref<10xf32>
}
// Should fuse first and second loops into third.
// Expecting private memref for '%a' first, then private memref for '%b'.
// CHECK-DAG: [[NEWA:%[0-9]+]] = alloc() : memref<1xf32>
// CHECK-DAG: [[NEWB:%[0-9]+]] = alloc() : memref<1xf32>
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, [[NEWA]][%2] : memref<1xf32>
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, [[NEWB]][%3] : memref<1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %5 = load [[NEWA]][%4] : memref<1xf32>
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %7 = load [[NEWB]][%6] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_first_and_second_loops() {
func @should_fuse_first_and_second_loops() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %a[%i1] : memref<10xf32>
store %cf7, %b[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v1 = load %c[%i2] : memref<10xf32>
}
// Should fuse first loop into the second (last loop should not be fused).
// Should create private memref '%2' for fused loop.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %0[%3] : memref<1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %5 = load %0[%4] : memref<1xf32>
// CHECK-NEXT: store %cst, %1[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: %6 = load %2[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_not_fuse_would_create_cycle() {
func @should_not_fuse_would_create_cycle() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
// Set up the following dependences:
// 1) loop0 -> loop1 on memref '%a'
// 2) loop0 -> loop2 on memref '%b'
// 3) loop1 -> loop2 on memref '%c'
for %i0 = 0 to 10 {
%v0 = load %a[%i0] : memref<10xf32>
store %cf7, %b[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %a[%i1] : memref<10xf32>
%v1 = load %c[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v2 = load %b[%i2] : memref<10xf32>
store %cf7, %c[%i2] : memref<10xf32>
}
// Should not fuse: fusing loop first loop into last would create a cycle.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = load %0[%i0] : memref<10xf32>
// CHECK-NEXT: store %cst, %1[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i1] : memref<10xf32>
// CHECK-NEXT: %4 = load %2[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i2 = 0 to 10 {
// CHECK-NEXT: %5 = load %1[%i2] : memref<10xf32>
// CHECK-NEXT: store %cst, %2[%i2] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_not_fuse_across_waw_dep() {
func @should_not_fuse_across_waw_dep() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %m[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v1 = load %m[%i2] : memref<10xf32>
}
// Fusing loop %i0 to %i2 would violate the WAW dependence between %i0 and %i1
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i2 = 0 to 10 {
// CHECK-NEXT: %1 = load %0[%i2] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_and_move_to_preserve_war_dep() {
func @should_fuse_and_move_to_preserve_war_dep() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
%v0 = load %a[%i0] : memref<10xf32>
store %v0, %b[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %a[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v1 = load %b[%i2] : memref<10xf32>
}
// Loops '%i1' and '%i2' have no dependences. We can fuse a slice of '%i0'
// into '%i2' if we move the fused loop nest before '%i1', which preserves
// the WAR dependence from load '%a' in '%i0' to the store '%a' in loop '%i1'.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %2 = load %1[%i0] : memref<10xf32>
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %2, %0[%3] : memref<1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %5 = load %0[%4] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %1[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_with_private_memref_if_top_level_access() {
func @should_fuse_with_private_memref_if_top_level_access() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %m[%i1] : memref<10xf32>
}
%c0 = constant 4 : index
%v1 = load %m[%c0] : memref<10xf32>
// Top-level load to '%m' should prevent fusion.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %1[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: store %cst, %0[%2] : memref<1xf32>
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: %4 = load %0[%3] : memref<1xf32>
// CHECK-NEXT: }
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_no_top_level_access() {
func @should_fuse_no_top_level_access() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %m[%i1] : memref<10xf32>
}
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %0[%1] : memref<1xf32>
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %3 = load %0[%2] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
#set0 = (d0) : (1 == 0)
// CHECK-LABEL: func @should_not_fuse_if_inst_at_top_level() {
func @should_not_fuse_if_inst_at_top_level() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %m[%i1] : memref<10xf32>
}
%c0 = constant 4 : index
if #set0(%c0) {
}
// Top-level IfOp should prevent fusion.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: %1 = load %0[%i1] : memref<10xf32>
// CHECK-NEXT: }
return
}
// -----
#set0 = (d0) : (1 == 0)
// CHECK-LABEL: func @should_not_fuse_if_inst_in_loop_nest() {
func @should_not_fuse_if_inst_in_loop_nest() {
%m = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%c4 = constant 4 : index
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
if #set0(%c4) {
}
%v0 = load %m[%i1] : memref<10xf32>
}
// IfOp in ForInst should prevent fusion.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK: for %i1 = 0 to 10 {
// CHECK-NEXT: if #set0(%c4) {
// CHECK-NEXT: }
// CHECK-NEXT: %1 = load %0[%i1] : memref<10xf32>
// CHECK-NEXT: }
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5) -> (-d0 + d3)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5) -> (-d1 + d4)
// CHECK: [[MAP2:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5) -> (-d2 + d5)
// CHECK-LABEL: func @permute_and_fuse() {
func @permute_and_fuse() {
%m = alloc() : memref<10x20x30xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
for %i1 = 0 to 20 {
for %i2 = 0 to 30 {
store %cf7, %m[%i0, %i1, %i2] : memref<10x20x30xf32>
}
}
}
for %i3 = 0 to 30 {
for %i4 = 0 to 10 {
for %i5 = 0 to 20 {
%v0 = load %m[%i4, %i5, %i3] : memref<10x20x30xf32>
"foo"(%v0) : (f32) -> ()
}
}
}
// CHECK: for %i0 = 0 to 30 {
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: for %i2 = 0 to 20 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: %2 = affine.apply [[MAP1]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: %3 = affine.apply [[MAP2]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: store %cst, %0[%1, %2, %3] : memref<1x1x1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: %5 = affine.apply [[MAP1]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: %6 = affine.apply [[MAP2]](%i1, %i2, %i0, %i1, %i2, %i0)
// CHECK-NEXT: %7 = load %0[%4, %5, %6] : memref<1x1x1xf32>
// CHECK-NEXT: "foo"(%7) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0, d1) -> (d0 * 4 + d1)
// CHECK-DAG: [[MAP1:#map[0-9]+]] = (d0) -> (d0 floordiv 4)
// CHECK-DAG: [[MAP2:#map[0-9]+]] = (d0) -> (d0 mod 4)
// Reshape from a 64 x f32 to 16 x 4 x f32.
// CHECK-LABEL: func @fuse_reshape_64_16_4
func @fuse_reshape_64_16_4(%in : memref<64xf32>) {
%out = alloc() : memref<16x4xf32>
for %i0 = 0 to 64 {
%v = load %in[%i0] : memref<64xf32>
%idx = affine.apply (d0) -> (d0 floordiv 4) (%i0)
%idy = affine.apply (d0) -> (d0 mod 4) (%i0)
store %v, %out[%idx, %idy] : memref<16x4xf32>
}
for %i1 = 0 to 16 {
for %i2 = 0 to 4 {
%w = load %out[%i1, %i2] : memref<16x4xf32>
"foo"(%w) : (f32) -> ()
}
}
return
// CHECK: for %i0 =
// CHECK-NEXT: for %i1 =
// CHECK-NOT: for
// CHECK: }
// CHECK-NEXT: }
// CHECK-NEXT: return
}
// -----
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0) -> (d0 floordiv 4)
// CHECK-DAG: [[MAP1:#map[0-9]+]] = (d0) -> (d0 mod 4)
// CHECK-DAG: [[MAP2:#map[0-9]+]] = (d0, d1) -> (d0 * 4 + d1)
// CHECK-DAG: [[MAP3:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// Reshape a 16x4xf32 to 64xf32.
// CHECK-LABEL: func @fuse_reshape_16_4_64
func @fuse_reshape_16_4_64() {
%in = alloc() : memref<16x4xf32>
%out = alloc() : memref<64xf32>
for %i0 = 0 to 16 {
for %i1 = 0 to 4 {
%v = load %in[%i0, %i1] : memref<16x4xf32>
%idx = affine.apply (d0, d1) -> (4*d0 + d1) (%i0, %i1)
store %v, %out[%idx] : memref<64xf32>
}
}
for %i2 = 0 to 64 {
%w = load %out[%i2] : memref<64xf32>
"foo"(%w) : (f32) -> ()
}
// CHECK: for %i0 = 0 to 64 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0)
// CHECK-NEXT: %3 = affine.apply [[MAP1]](%i0)
// CHECK-NEXT: %4 = load %1[%2, %3] : memref<16x4xf32>
// CHECK-NEXT: %5 = affine.apply [[MAP2]](%2, %3)
// CHECK-NEXT: %6 = affine.apply [[MAP3]](%i0, %5)
// CHECK-NEXT: store %4, %0[%6] : memref<1xf32>
// CHECK-NEXT: %7 = affine.apply [[MAP3]](%i0, %i0)
// CHECK-NEXT: %8 = load %0[%7] : memref<1xf32>
// CHECK-NEXT: "foo"(%8) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// All three loop nests below (6-d one, 2-d one, 2-d one is fused into a single
// 2-d loop nest).
func @R6_to_R2_reshape_square() -> memref<64x9xi32> {
%in = alloc() : memref<2x2x3x3x16x1xi32>
%out = alloc() : memref<64x9xi32>
%live_out = alloc() : memref<64x9xi32>
// Initialize input.
for %i0 = 0 to 2 {
for %i1 = 0 to 2 {
for %i2 = 0 to 3 {
for %i3 = 0 to 3 {
for %i4 = 0 to 16 {
for %i5 = 0 to 1 {
%val = "foo"(%i0, %i1, %i2, %i3, %i4, %i5) : (index, index, index, index, index, index) -> i32
store %val, %in[%i0, %i1, %i2, %i3, %i4, %i5] : memref<2x2x3x3x16x1xi32>
}
}
}
}
}
}
for %ii = 0 to 64 {
for %jj = 0 to 9 {
// Convert output coordinates to linear index.
%a0 = affine.apply (d0, d1) -> (d0 * 9 + d1) (%ii, %jj)
%0 = affine.apply (d0) -> (d0 floordiv (2 * 3 * 3 * 16 * 1))(%a0)
%1 = affine.apply (d0) -> ((d0 mod 288) floordiv (3 * 3 * 16 * 1))(%a0)
%2 = affine.apply (d0) -> (((d0 mod 288) mod 144) floordiv (3 * 16 * 1))(%a0)
%3 = affine.apply (d0) -> ((((d0 mod 288) mod 144) mod 48) floordiv (16 * 1))(%a0)
%4 = affine.apply (d0) -> ((((d0 mod 288) mod 144) mod 48) mod 16)(%a0)
%5 = affine.apply (d0) -> (((((d0 mod 144) mod 144) mod 48) mod 16) mod 1)(%a0)
%v = load %in[%0, %1, %2, %3, %4, %5] : memref<2x2x3x3x16x1xi32>
store %v, %out[%ii, %jj] : memref<64x9xi32>
}
}
for %i = 0 to 64 {
for %j = 0 to 9 {
%a = load %out[%i, %j] : memref<64x9xi32>
%b = muli %a, %a : i32
store %b, %live_out[%i, %j] : memref<64x9xi32>
}
}
return %live_out : memref<64x9xi32>
}
// Everything above is fused to a single 2-d loop nest, and the 6-d tensor %in
// is eliminated if -memref-dataflow-opt is also supplied.
//
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0, d1) -> ((d0 * 9 + d1) floordiv 288)
// CHECK-DAG: [[MAP1:#map[0-9]+]] = (d0, d1) -> (((d0 * 9 + d1) mod 288) floordiv 144)
// CHECK-DAG: [[MAP2:#map[0-9]+]] = (d0, d1) -> ((((d0 * 9 + d1) mod 288) mod 144) floordiv 48)
// CHECK-DAG: [[MAP3:#map[0-9]+]] = (d0, d1) -> (((((d0 * 9 + d1) mod 288) mod 144) mod 48) floordiv 16)
// CHECK-DAG: [[MAP4:#map[0-9]+]] = (d0, d1) -> (((((d0 * 9 + d1) mod 288) mod 144) mod 48) mod 16)
// CHECK-DAG: [[MAP5:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d2 - (d0 * 37 + d1 * 36) floordiv 36)
// CHECK-DAG: [[MAP6:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d3)
// CHECK-DAG: [[MAP7:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d4)
// CHECK-DAG: [[MAP8:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d5)
// CHECK-DAG: [[MAP9:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d6)
// CHECK-DAG: [[MAP10:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7) -> (d7)
// CHECK-DAG: [[MAP11:#map[0-9]+]] = (d0, d1) -> (d0 * 9 + d1)
// CHECK-DAG: [[MAP12:#map[0-9]+]] = (d0) -> (d0 floordiv 288)
// CHECK-DAG: [[MAP13:#map[0-9]+]] = (d0) -> ((d0 mod 288) floordiv 144)
// CHECK-DAG: [[MAP14:#map[0-9]+]] = (d0) -> (((d0 mod 288) mod 144) floordiv 48)
// CHECK-DAG: [[MAP15:#map[0-9]+]] = (d0) -> ((((d0 mod 288) mod 144) mod 48) floordiv 16)
// CHECK-DAG: [[MAP16:#map[0-9]+]] = (d0) -> ((((d0 mod 288) mod 144) mod 48) mod 16)
// CHECK-DAG: [[MAP17:#map[0-9]+]] = (d0) -> (0)
// CHECK-DAG: [[MAP18:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d0 + d2)
// CHECK-DAG: [[MAP19:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d1 + d3)
//
// CHECK-LABEL: func @R6_to_R2_reshape
// CHECK: %0 = alloc() : memref<1x2x3x3x16x1xi32>
// CHECK: %1 = alloc() : memref<1x1xi32>
// CHECK: %2 = alloc() : memref<64x9xi32>
// CHECK-NEXT: for %i0 = 0 to 64 {
// CHECK-NEXT: for %i1 = 0 to 9 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i1)
// CHECK-NEXT: %4 = affine.apply [[MAP1]](%i0, %i1)
// CHECK-NEXT: %5 = affine.apply [[MAP2]](%i0, %i1)
// CHECK-NEXT: %6 = affine.apply [[MAP3]](%i0, %i1)
// CHECK-NEXT: %7 = affine.apply [[MAP4]](%i0, %i1)
// CHECK-NEXT: %8 = "foo"(%3, %4, %5, %6, %7, %c0) : (index, index, index, index, index, index) -> i32
// CHECK-NEXT: %9 = affine.apply [[MAP5]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: %10 = affine.apply [[MAP6]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: %11 = affine.apply [[MAP7]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: %12 = affine.apply [[MAP8]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: %13 = affine.apply [[MAP9]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: %14 = affine.apply [[MAP10]](%i0, %i1, %3, %4, %5, %6, %7, %c0)
// CHECK-NEXT: store %8, %0[%9, %10, %11, %12, %13, %14] : memref<1x2x3x3x16x1xi32>
// CHECK-NEXT: %15 = affine.apply [[MAP11]](%i0, %i1)
// CHECK-NEXT: %16 = affine.apply [[MAP12]](%15)
// CHECK-NEXT: %17 = affine.apply [[MAP13]](%15)
// CHECK-NEXT: %18 = affine.apply [[MAP14]](%15)
// CHECK-NEXT: %19 = affine.apply [[MAP15]](%15)
// CHECK-NEXT: %20 = affine.apply [[MAP16]](%15)
// CHECK-NEXT: %21 = affine.apply [[MAP17]](%15)
// CHECK-NEXT: %22 = affine.apply [[MAP5]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %23 = affine.apply [[MAP6]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %24 = affine.apply [[MAP7]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %25 = affine.apply [[MAP8]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %26 = affine.apply [[MAP9]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %27 = affine.apply [[MAP10]](%i0, %i1, %16, %17, %18, %19, %20, %21)
// CHECK-NEXT: %28 = load %0[%22, %23, %24, %25, %26, %27] : memref<1x2x3x3x16x1xi32>
// CHECK-NEXT: %29 = affine.apply [[MAP18]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %30 = affine.apply [[MAP19]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: store %28, %1[%29, %30] : memref<1x1xi32>
// CHECK-NEXT: %31 = affine.apply [[MAP18]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %32 = affine.apply [[MAP19]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %33 = load %1[%31, %32] : memref<1x1xi32>
// CHECK-NEXT: %34 = muli %33, %33 : i32
// CHECK-NEXT: store %34, %2[%i0, %i1] : memref<64x9xi32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return %2 : memref<64x9xi32>
// -----
// CHECK-LABEL: func @fuse_symbolic_bounds
func @fuse_symbolic_bounds(%M : index, %N : index) {
%N_plus_5 = affine.apply (d0) -> (d0 + 5)(%N)
%m = alloc(%M, %N_plus_5) : memref<? x ? x f32>
%c0 = constant 0.0 : f32
%s = constant 5 : index
for %i0 = 0 to %M {
for %i1 = 0 to (d0) -> (d0 + 5) (%N) {
store %c0, %m[%i0, %i1] : memref<? x ? x f32>
}
}
for %i2 = 0 to %M {
for %i3 = 0 to %N {
%idy = affine.apply (d0)[s0] -> (d0 + s0) (%i3)[%s]
%v = load %m[%i2, %idy] : memref<? x ? x f32>
}
}
return
}
// -----
// CHECK-DAG: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_reduction_at_depth1
func @should_fuse_reduction_at_depth1() {
%a = alloc() : memref<10x100xf32>
%b = alloc() : memref<10xf32>
for %i0 = 0 to 10 {
for %i1 = 0 to 100 {
%v0 = load %b[%i0] : memref<10xf32>
%v1 = load %a[%i0, %i1] : memref<10x100xf32>
%v2 = "maxf"(%v0, %v1) : (f32, f32) -> f32
store %v2, %b[%i0] : memref<10xf32>
}
}
for %i2 = 0 to 10 {
for %i3 = 0 to 100 {
%v3 = load %b[%i2] : memref<10xf32>
%v4 = load %a[%i2, %i3] : memref<10x100xf32>
%v5 = subf %v4, %v3 : f32
store %v5, %b[%i2] : memref<10xf32>
}
}
// This test should fuse the src reduction loop at depth 1 in the destination
// loop nest, which improves locality and enables subsequence passes to
// decrease the reduction memref size and possibly place it in a faster
// memory space.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: for %i1 = 0 to 100 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %3 = load %0[%2] : memref<1xf32>
// CHECK-NEXT: %4 = load %1[%i0, %i1] : memref<10x100xf32>
// CHECK-NEXT: %5 = "maxf"(%3, %4) : (f32, f32) -> f32
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %5, %0[%6] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 100 {
// CHECK-NEXT: %7 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %8 = load %0[%7] : memref<1xf32>
// CHECK-NEXT: %9 = load %1[%i0, %i2] : memref<10x100xf32>
// CHECK-NEXT: %10 = subf %9, %8 : f32
// CHECK-NEXT: %11 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %10, %0[%11] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2) -> (-d0 + d1)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2) -> (d2)
// CHECK-LABEL: func @should_fuse_at_src_depth1_and_dst_depth1
func @should_fuse_at_src_depth1_and_dst_depth1() {
%a = alloc() : memref<100x16xf32>
%b = alloc() : memref<100x16xf32>
for %i0 = 0 to 100 {
for %i1 = 0 to 16 {
%v0 = load %a[%i0, %i1] : memref<100x16xf32>
"op0"(%v0) : (f32) -> ()
}
for %i2 = 0 to 16 {
%v1 = "op1"() : () -> (f32)
store %v1, %b[%i0, %i2] : memref<100x16xf32>
}
}
for %i3 = 0 to 100 {
for %i4 = 0 to 16 {
%v2 = load %b[%i3, %i4] : memref<100x16xf32>
"op2"(%v2) : (f32) -> ()
}
}
// We can slice iterations of the '%i0' and '%i1' loops in the the source
// loop nest, but slicing at depth 2 and inserting the slice in the
// destination loop nest at depth2 causes extra computation. Instead,
// the fusion algorithm should detect that the source loop should be sliced
// at depth 1 and the slice should be inserted at depth 1.
// CHECK: for %i0 = 0 to 100 {
// CHECK-NEXT: for %i1 = 0 to 16 {
// CHECK-NEXT: %2 = load %1[%i0, %i1] : memref<100x16xf32>
// CHECK-NEXT: "op0"(%2) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 16 {
// CHECK-NEXT: %3 = "op1"() : () -> f32
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0, %i2)
// CHECK-NEXT: %5 = affine.apply [[MAP1]](%i0, %i0, %i2)
// CHECK-NEXT: store %3, %0[%4, %5] : memref<1x16xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i3 = 0 to 16 {
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i0, %i0, %i3)
// CHECK-NEXT: %7 = affine.apply [[MAP1]](%i0, %i0, %i3)
// CHECK-NEXT: %8 = load %0[%6, %7] : memref<1x16xf32>
// CHECK-NEXT: "op2"(%8) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (d0 * 10 + d1)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2) -> (d0 * -10 - d1 + d2)
// CHECK-LABEL: func @should_fuse_src_depth1_at_dst_depth2
func @should_fuse_src_depth1_at_dst_depth2() {
%a = alloc() : memref<100xf32>
%c0 = constant 0.0 : f32
for %i0 = 0 to 100 {
store %c0, %a[%i0] : memref<100xf32>
}
for %i1 = 0 to 10 {
for %i2 = 0 to 10 {
%a0 = affine.apply (d0, d1) -> (d0 * 10 + d1) (%i1, %i2)
%v0 = load %a[%a0] : memref<100xf32>
}
}
// The source loop nest slice loop bound is a function of both destination
// loop IVs, so we should slice at depth 1 and insert the slice at depth 2.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i1)
// CHECK-NEXT: %2 = affine.apply [[MAP1]](%i0, %i1, %1)
// CHECK-NEXT: store %cst, %0[%2] : memref<1xf32>
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i1)
// CHECK-NEXT: %4 = affine.apply [[MAP1]](%i0, %i1, %3)
// CHECK-NEXT: %5 = load %0[%4] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @fusion_at_depth0_not_currently_supported
func @fusion_at_depth0_not_currently_supported() {
%0 = alloc() : memref<10xf32>
%c0 = constant 0 : index
%cst = constant 0.000000e+00 : f32
for %i0 = 0 to 10 {
store %cst, %0[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%1 = load %0[%c0] : memref<10xf32>
}
// NOTE: Should shrink memref size to 1 element access by load in dst loop
// nest, and make the store in the slice store to the same element.
// CHECK-DAG: %0 = alloc() : memref<1xf32>
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%c0] : memref<1xf32>
// CHECK-NEXT: %1 = load %0[%c0_0] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (-d0 + d4)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (-d1 + d5)
// CHECK: [[MAP2:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (-d2 + d6)
// CHECK: [[MAP3:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (-d3 + d7)
// CHECK: [[MAP4:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (d8)
// CHECK: [[MAP5:#map[0-9]+]] = (d0, d1, d2, d3, d4, d5, d6, d7, d8, d9) -> (d9)
// CHECK-LABEL: func @should_fuse_deep_loop_nests
func @should_fuse_deep_loop_nests() {
%0 = alloc() : memref<2x2x3x3x16x10xf32, 2>
%1 = alloc() : memref<2x2x3x3x16x10xf32, 2>
%2 = alloc() : memref<3x3x3x3x16x10xf32, 2>
%c0 = constant 0 : index
%c1 = constant 1 : index
%c1_0 = constant 1 : index
%cst = constant 0.000000e+00 : f32
for %i0 = 0 to 2 {
for %i1 = 0 to 2 {
for %i2 = 0 to 3 {
for %i3 = 0 to 3 {
for %i4 = 0 to 16 {
for %i5 = 0 to 10 {
%3 = load %0[%i0, %i1, %i2, %i3, %i4, %i5]
: memref<2x2x3x3x16x10xf32, 2>
}
}
for %i6 = 0 to 16 {
for %i7 = 0 to 10 {
store %cst, %1[%i0, %i1, %i2, %i3, %i6, %i7]
: memref<2x2x3x3x16x10xf32, 2>
}
}
}
}
}
}
for %i8 = 0 to 3 {
for %i9 = 0 to 3 {
for %i10 = 0 to 2 {
for %i11 = 0 to 2 {
for %i12 = 0 to 3 {
for %i13 = 0 to 3 {
for %i14 = 0 to 2 {
for %i15 = 0 to 2 {
for %i16 = 0 to 16 {
for %i17 = 0 to 10 {
%5 = load %0[%i14, %i15, %i12, %i13, %i16, %i17]
: memref<2x2x3x3x16x10xf32, 2>
}
}
for %i18 = 0 to 16 {
for %i19 = 0 to 10 {
%6 = load %1[%i10, %i11, %i8, %i9, %i18, %i19]
: memref<2x2x3x3x16x10xf32, 2>
}
}
}
}
}
}
}
}
}
}
// The first four loops of the source loop nest can be sliced with iteration
// bounds which are a function of the first four loops of destination loop nest,
// where the destination loops nests have been interchanged.
// CHECK-DAG: %0 = alloc() : memref<1x1x1x1x16x10xf32, 2>
// CHECK: for %i0 = 0 to 3 {
// CHECK-NEXT: for %i1 = 0 to 3 {
// CHECK-NEXT: for %i2 = 0 to 2 {
// CHECK-NEXT: for %i3 = 0 to 2 {
// CHECK-NEXT: for %i4 = 0 to 3 {
// CHECK-NEXT: for %i5 = 0 to 3 {
// CHECK-NEXT: for %i6 = 0 to 16 {
// CHECK-NEXT: for %i7 = 0 to 10 {
// CHECK-NEXT: %3 = load %1[%i2, %i3, %i0, %i1, %i6, %i7] : memref<2x2x3x3x16x10xf32, 2>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: for %i8 = 0 to 16 {
// CHECK-NEXT: for %i9 = 0 to 10 {
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: %5 = affine.apply [[MAP1]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: %6 = affine.apply [[MAP2]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: %7 = affine.apply [[MAP3]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: %8 = affine.apply [[MAP4]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: %9 = affine.apply [[MAP5]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i8, %i9)
// CHECK-NEXT: store %cst, %0[%4, %5, %6, %7, %8, %9] : memref<1x1x1x1x16x10xf32, 2>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: for %i10 = 0 to 2 {
// CHECK-NEXT: for %i11 = 0 to 2 {
// CHECK-NEXT: for %i12 = 0 to 16 {
// CHECK-NEXT: for %i13 = 0 to 10 {
// CHECK-NEXT: %10 = load %1[%i10, %i11, %i4, %i5, %i12, %i13] : memref<2x2x3x3x16x10xf32, 2>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: for %i14 = 0 to 16 {
// CHECK-NEXT: for %i15 = 0 to 10 {
// CHECK-NEXT: %11 = affine.apply [[MAP0]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %12 = affine.apply [[MAP1]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %13 = affine.apply [[MAP2]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %14 = affine.apply [[MAP3]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %15 = affine.apply [[MAP4]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %16 = affine.apply [[MAP5]](%i2, %i3, %i0, %i1, %i2, %i3, %i0, %i1, %i14, %i15)
// CHECK-NEXT: %17 = load %0[%11, %12, %13, %14, %15, %16] : memref<1x1x1x1x16x10xf32, 2>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2) -> (-d0 + d1)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2) -> (d2)
// CHECK-LABEL: func @should_fuse_at_depth1_and_reduce_slice_trip_count
func @should_fuse_at_depth1_and_reduce_slice_trip_count() {
%a = alloc() : memref<4x256xf32>
%b = alloc() : memref<4x256xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
for %i0 = 0 to 4 {
for %i1 = 0 to 256 {
%v0 = load %b[%i0, %i1] : memref<4x256xf32>
}
for %i2 = 0 to 256 {
store %cf0, %a[%i0, %i2] : memref<4x256xf32>
}
}
for %d0 = 0 to 4 {
for %d1 = 0 to 16 {
%v1 = load %a[%d0, %d1] : memref<4x256xf32>
}
}
// The cost of fusing at depth 2 is greater than the cost of fusing at depth 1
// for two reasons:
// 1) Inserting the unsliceable src loop %i1 to a higher depth removes
// redundant computation and reduces costs.
// 2) Inserting the sliceable src loop %i2 at depth 1, we can still reduce
// its trip count to 16 (from 256) reducing costs.
// NOTE: the size of the private memref created for the fused loop nest
// is reduced from the original shape from 4x256 to 4x16 because of the
// data accessed by the load.
// CHECK-DAG: %0 = alloc() : memref<1x16xf32>
// CHECK: for %i0 = 0 to 4 {
// CHECK-NEXT: for %i1 = 0 to 256 {
// CHECK-NEXT: %2 = load %1[%i0, %i1] : memref<4x256xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 16 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0, %i2)
// CHECK-NEXT: %4 = affine.apply [[MAP1]](%i0, %i0, %i2)
// CHECK-NEXT: store %cst, %0[%3, %4] : memref<1x16xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i3 = 0 to 16 {
// CHECK-NEXT: %5 = affine.apply [[MAP0]](%i0, %i0, %i3)
// CHECK-NEXT: %6 = affine.apply [[MAP1]](%i0, %i0, %i3)
// CHECK-NEXT: %7 = load %0[%5, %6] : memref<1x16xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_fuse_at_depth1_with_trip_count_20
func @should_fuse_at_depth1_with_trip_count_20() {
%a = alloc() : memref<100xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
for %i0 = 0 to 100 {
store %cf0, %a[%i0]: memref<100xf32>
}
for %i1 = 0 to 5 {
for %i2 = 0 to 10 {
%v0 = load %a[%i2]: memref<100xf32>
}
for %i3 = 0 to 10 {
for %i4 = 0 to 20 {
%v1 = load %a[%i4]: memref<100xf32>
}
}
}
// NOTE: The size of the private memref created for fusion is shrunk to 20xf32
// CHECK-DAG: %0 = alloc() : memref<20xf32>
// CHECK: for %i0 = 0 to 5 {
// CHECK-NEXT: for %i1 = 0 to 20 {
// CHECK-NEXT: store %cst, %0[%i1] : memref<20xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 10 {
// CHECK-NEXT: %1 = load %0[%i2] : memref<20xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i3 = 0 to 10 {
// CHECK-NEXT: for %i4 = 0 to 20 {
// CHECK-NEXT: %2 = load %0[%i4] : memref<20xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_fuse_at_depth1_with_trip_count_19
func @should_fuse_at_depth1_with_trip_count_19() {
%a = alloc() : memref<100xf32>
%c0 = constant 0 : index
%cf0 = constant 0.0 : f32
for %i0 = 0 to 100 {
store %cf0, %a[%i0]: memref<100xf32>
}
for %i1 = 0 to 5 {
for %i2 = 0 to 19 {
%v0 = load %a[%i2]: memref<100xf32>
}
for %i3 = 0 to 10 {
for %i4 = 0 to 10 {
%v1 = load %a[%i4]: memref<100xf32>
}
}
}
// NOTE: The size of the private memref created for fusion is shrunk to 19xf32
// CHECK-DAG: %0 = alloc() : memref<19xf32>
// CHECK: for %i0 = 0 to 5 {
// CHECK-NEXT: for %i1 = 0 to 19 {
// CHECK-NEXT: store %cst, %0[%i1] : memref<19xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 19 {
// CHECK-NEXT: %1 = load %0[%i2] : memref<19xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i3 = 0 to 10 {
// CHECK-NEXT: for %i4 = 0 to 10 {
// CHECK-NEXT: %2 = load %0[%i4] : memref<19xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_with_private_memrefs_with_diff_shapes() {
func @should_fuse_with_private_memrefs_with_diff_shapes() {
%m = alloc() : memref<100xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 100 {
store %cf7, %m[%i0] : memref<100xf32>
}
for %i1 = 0 to 17 {
%v0 = load %m[%i1] : memref<100xf32>
}
for %i2 = 0 to 82 {
%v1 = load %m[%i2] : memref<100xf32>
}
// Should create two new private memrefs customized to the shapes accessed
// by loops %i1 and %i2.
// CHECK-DAG: %0 = alloc() : memref<1xf32>
// CHECK-DAG: %1 = alloc() : memref<1xf32>
// CHECK: for %i0 = 0 to 82 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %1[%2] : memref<1xf32>
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %4 = load %1[%3] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 17 {
// CHECK-NEXT: %5 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: store %cst, %0[%5] : memref<1xf32>
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: %7 = load %0[%6] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_not_fuse_live_out_arg(%arg0: memref<10xf32>) {
func @should_not_fuse_live_out_arg(%arg0: memref<10xf32>) {
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %arg0[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %arg0[%i1] : memref<10xf32>
}
// This tests that the loop nest '%i0' should not be removed after fusion
// because it writes to memref argument '%arg0'.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %arg0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %0 = load %arg0[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @should_not_fuse_escaping_memref() -> memref<10xf32>
func @should_not_fuse_escaping_memref() -> memref<10xf32> {
%cf7 = constant 7.0 : f32
%m = alloc() : memref<10xf32>
for %i0 = 0 to 10 {
store %cf7, %m[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %m[%i1] : memref<10xf32>
}
// This tests that the loop nest '%i0' should not be removed after fusion
// because it writes to memref '%m' which is returned by the function.
// CHECK-DAG: %0 = alloc() : memref<10xf32>
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %1 = load %0[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return %0 : memref<10xf32>
return %m : memref<10xf32>
}
// -----
// This should fuse with the %in becoming a 1x1x1.
func @R3_to_R2_reshape() {
%in = alloc() : memref<2x3x16xi32>
%c0 = constant 0 : index
for %i0 = 0 to 2 {
for %i1 = 0 to 3 {
for %i2 = 0 to 16 {
%val = "foo"(%i0, %i1, %i2) : (index, index, index) -> i32
store %val, %in[%i0, %i1, %i2] : memref<2x3x16xi32>
}
}
}
for %ii = 0 to 32 {
for %jj = 0 to 3 {
%a0 = affine.apply (d0, d1) -> (d0 * 3 + d1) (%ii, %jj)
%idx = affine.apply (d0) -> (d0 floordiv (3 * 16)) (%a0)
%v = load %in[%idx, %jj, %c0]
: memref<2x3x16xi32>
}
}
return
}
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> ((d0 * 3 + d1) floordiv 48)
// CHECK-NEXT: [[MAP2:#map[0-9]+]] = (d0, d1, d2, d3, d4) -> (d2 - (d0 * 25 + d1 * 24) floordiv 24)
// CHECK-NEXT: [[MAP3:#map[0-9]+]] = (d0, d1, d2, d3, d4) -> (-d1 + d3)
// CHECK-NEXT: [[MAP4:#map[0-9]+]] = (d0, d1, d2, d3, d4) -> (d4)
// CHECK-NEXT: [[MAP5:#map[0-9]+]] = (d0, d1) -> (d0 * 3 + d1)
// CHECK-NEXT: [[MAP6:#map[0-9]+]] = (d0) -> (d0 floordiv 48)
// CHECK-LABEL: func @R3_to_R2_reshape()
// CHECK-DAG: %0 = alloc() : memref<1x1x1xi32>
// CHECK: for %i0 = 0 to 32 {
// CHECK-NEXT: for %i1 = 0 to 3 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i1)
// CHECK-NEXT: %2 = "foo"(%1, %i1, %c0) : (index, index, index) -> i32
// CHECK-NEXT: %3 = affine.apply [[MAP2]](%i0, %i1, %1, %i1, %c0)
// CHECK-NEXT: %4 = affine.apply [[MAP3]](%i0, %i1, %1, %i1, %c0)
// CHECK-NEXT: %5 = affine.apply [[MAP4]](%i0, %i1, %1, %i1, %c0)
// CHECK-NEXT: store %2, %0[%3, %4, %5] : memref<1x1x1xi32>
// CHECK-NEXT: %6 = affine.apply [[MAP5]](%i0, %i1)
// CHECK-NEXT: %7 = affine.apply [[MAP6]](%6)
// CHECK-NEXT: %8 = affine.apply [[MAP2]](%i0, %i1, %7, %i1, %c0_0)
// CHECK-NEXT: %9 = affine.apply [[MAP3]](%i0, %i1, %7, %i1, %c0_0)
// CHECK-NEXT: %10 = affine.apply [[MAP4]](%i0, %i1, %7, %i1, %c0_0)
// CHECK-NEXT: %11 = load %0[%8, %9, %10] : memref<1x1x1xi32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
// -----
// CHECK-LABEL: func @should_not_fuse_multi_output_producer() {
func @should_not_fuse_multi_output_producer() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %a[%i0] : memref<10xf32>
store %cf7, %b[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %a[%i1] : memref<10xf32>
}
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: store %cst, %1[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %2 = load %0[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @fusion_preventing_deps_on_middle_loop() {
func @fusion_preventing_deps_on_middle_loop() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
%v0 = load %a[%i0] : memref<10xf32>
store %v0, %b[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %a[%i1] : memref<10xf32>
%v1 = load %c[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v2 = load %b[%i2] : memref<10xf32>
store %v2, %c[%i2] : memref<10xf32>
}
// Loops '%i0' and '%i2' cannot fuse along producer/consumer edge on memref
// '%b', because of the WAR dep from '%i0' to '%i1' on memref '%a' and
// because of the WAR dep from '%i1' to '%i2' on memref '%c'.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = load %0[%i0] : memref<10xf32>
// CHECK-NEXT: store %3, %1[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %0[%i1] : memref<10xf32>
// CHECK-NEXT: %4 = load %2[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 10 {
// CHECK-NEXT: %5 = load %1[%i2] : memref<10xf32>
// CHECK-NEXT: store %5, %2[%i2] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_and_move_to_preserve_war_dep() {
func @should_fuse_and_move_to_preserve_war_dep() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
%v0 = load %b[%i0] : memref<10xf32>
store %v0, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 3 {
%v2 = load %c[%i1] : memref<10xf32>
}
for %i2 = 0 to 5 {
store %cf7, %b[%i2] : memref<10xf32>
}
for %i3 = 0 to 10 {
%v1 = load %a[%i3] : memref<10xf32>
store %cf7, %c[%i3] : memref<10xf32>
}
// Dependence graph:
//
// %i0 ---------
// | |
// --- %i1 | %b | %a
// | | |
// %c | %i2 <-- |
// | |
// --> %i3 <--------
//
// It is possible to fuse loop '%i0' into '%i3' and preserve dependences
// if the fused loop nest is inserted between loops '%i1' and '%i2'.
// CHECK-DAG: %0 = alloc() : memref<1xf32>
// CHECK: for %i0 = 0 to 3 {
// CHECK-NEXT: %3 = load %2[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %4 = load %1[%i1] : memref<10xf32>
// CHECK-NEXT: %5 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: store %4, %0[%5] : memref<1xf32>
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: %7 = load %0[%6] : memref<1xf32>
// CHECK-NEXT: store %cst, %2[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 5 {
// CHECK-NEXT: store %cst, %1[%i2] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK-LABEL: func @fusion_preventing_dep_on_constant() {
func @fusion_preventing_dep_on_constant() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
%v0 = load %b[%i0] : memref<10xf32>
store %cf7, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %b[%i1] : memref<10xf32>
}
%cf11 = constant 11.0 : f32
for %i2 = 0 to 10 {
%v2 = load %a[%i2] : memref<10xf32>
store %cf11, %c[%i2] : memref<10xf32>
}
// Loops '%i0' and '%i2' cannot fuse along producer/consumer edge on memref
// '%a', because of the WAR dep from '%i0' to '%i1' on memref '%b' and
// because of the SSA value dep from '%cf11' def to use in '%i2'.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = load %1[%i0] : memref<10xf32>
// CHECK-NEXT: store %cst, %0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %1[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: %cst_0 = constant 1.100000e+01 : f32
// CHECK-NEXT: for %i2 = 0 to 10 {
// CHECK-NEXT: %4 = load %0[%i2] : memref<10xf32>
// CHECK-NEXT: store %cst_0, %2[%i2] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_and_preserve_dep_on_constant() {
func @should_fuse_and_preserve_dep_on_constant() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%c = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
%cf11 = constant 11.0 : f32
for %i0 = 0 to 10 {
%v0 = load %b[%i0] : memref<10xf32>
store %cf7, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
store %cf7, %b[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v2 = load %a[%i2] : memref<10xf32>
store %cf11, %c[%i2] : memref<10xf32>
}
// Loops '%i0' and '%i2' can fuse along producer/consumer edge on memref
// '%a', and preserve the WAR dep from '%i0' to '%i1' on memref '%b', and
// the SSA value dep from '%cf11' def to use in '%i2'.
// CHECK: %cst_0 = constant 1.100000e+01 : f32
// CHECK-NEXT: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = load %1[%i0] : memref<10xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %0[%4] : memref<1xf32>
// CHECK-NEXT: %5 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %6 = load %0[%5] : memref<1xf32>
// CHECK-NEXT: store %cst_0, %2[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: store %cst, %1[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2) -> (d1)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2) -> (-d0 + d2)
// CHECK: [[MAP2:#map[0-9]+]] = (d0, d1) -> (d0 * 16 - d1 + 15)
// CHECK: [[MAP3:#map[0-9]+]] = (d0, d1) -> (d0 * 16 + d1)
// CHECK-LABEL: func @should_fuse_at_depth_above_loop_carried_dependence(%arg0: memref<64x4xf32>, %arg1: memref<64x4xf32>) {
func @should_fuse_at_depth_above_loop_carried_dependence(%arg0: memref<64x4xf32>, %arg1: memref<64x4xf32>) {
%out = alloc() : memref<64x4xf32>
%0 = constant 0.0 : f32
for %i0 = 0 to 64 {
for %i1 = 0 to 4 {
store %0, %out[%i0, %i1] : memref<64x4xf32>
}
}
for %i2 = 0 to 4 {
for %i3 = 0 to 4 {
for %i4 = 0 to 16 {
%1 = affine.apply (d0, d1) -> (d0 * 16 - d1 + 15)(%i3, %i4)
%2 = load %arg1[%1, %i2] : memref<64x4xf32>
"op0"(%2) : (f32) -> ()
}
for %i5 = 0 to 4 {
for %i6 = 0 to 16 {
%3 = affine.apply (d0, d1) -> (d0 * 16 - d1 + 15)(%i5, %i6)
%4 = load %arg0[%3, %i3] : memref<64x4xf32>
"op1"(%4) : (f32) -> ()
}
for %i7 = 0 to 16 {
%5 = "op2"() : () -> (f32)
%6 = affine.apply (d0, d1) -> (d0 * 16 + d1)(%i5, %i7)
%7 = load %out[%6, %i2] : memref<64x4xf32>
%8 = addf %7, %5 : f32
store %8, %out[%6, %i2] : memref<64x4xf32>
}
}
}
}
// We can fuse source loop nest '%i0' into dst loop nest '%i2', but the
// depth at which we can insert the src loop nest slice into the dst loop
// lest must be decreased because of a loop carried dependence on loop '%i3'.
// As a result, the source loop nest is inserted at dst loop nest depth 1,
// just above the loop with the carried depenence. In addition, the source
// loop nest iteration bounds on its loop '%i1' are reduced to 1, so the
// memref size can be reduced to 128x1xf32.
// CHECK: %0 = alloc() : memref<64x1xf32>
// CHECK: for %i0 = 0 to 4 {
// CHECK-NEXT: for %i1 = 0 to 64 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i1, %i0)
// CHECK-NEXT: %2 = affine.apply [[MAP1]](%i0, %i1, %i0)
// CHECK-NEXT: store %cst, %0[%1, %2] : memref<64x1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i2 = 0 to 4 {
// CHECK-NEXT: for %i3 = 0 to 16 {
// CHECK-NEXT: %3 = affine.apply [[MAP2]](%i2, %i3)
// CHECK-NEXT: %4 = load %arg1[%3, %i0] : memref<64x4xf32>
// CHECK-NEXT: "op0"(%4) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: for %i4 = 0 to 4 {
// CHECK-NEXT: for %i5 = 0 to 16 {
// CHECK-NEXT: %5 = affine.apply [[MAP2]](%i4, %i5)
// CHECK-NEXT: %6 = load %arg0[%5, %i2] : memref<64x4xf32>
// CHECK-NEXT: "op1"(%6) : (f32) -> ()
// CHECK-NEXT: }
// CHECK-NEXT: for %i6 = 0 to 16 {
// CHECK-NEXT: %7 = "op2"() : () -> f32
// CHECK-NEXT: %8 = affine.apply [[MAP3]](%i4, %i6)
// CHECK-NEXT: %9 = affine.apply [[MAP0]](%i0, %8, %i0)
// CHECK-NEXT: %10 = affine.apply [[MAP1]](%i0, %8, %i0)
// CHECK-NEXT: %11 = load %0[%9, %10] : memref<64x1xf32>
// CHECK-NEXT: %12 = addf %11, %7 : f32
// CHECK-NEXT: %13 = affine.apply [[MAP0]](%i0, %8, %i0)
// CHECK-NEXT: %14 = affine.apply [[MAP1]](%i0, %8, %i0)
// CHECK-NEXT: store %12, %0[%13, %14] : memref<64x1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_after_private_memref_creation() {
func @should_fuse_after_private_memref_creation() {
%a = alloc() : memref<10xf32>
%b = alloc() : memref<10xf32>
%cf7 = constant 7.0 : f32
for %i0 = 0 to 10 {
store %cf7, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%v0 = load %a[%i1] : memref<10xf32>
store %v0, %b[%i1] : memref<10xf32>
}
for %i2 = 0 to 10 {
%v1 = load %a[%i2] : memref<10xf32>
store %v1, %b[%i2] : memref<10xf32>
}
// On the first visit to '%i2', the fusion algorithm can not fuse loop nest
// '%i0' into '%i2' because of the dependences '%i0' and '%i2' each have on
// '%i1'. However, once the loop nest '%i0' is fused into '%i1' with a
// private memref, the dependence between '%i0' and '%i1' on memref '%a' no
// longer exists, so '%i0' can now be fused into '%i2'.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %1[%3] : memref<1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %5 = load %1[%4] : memref<1xf32>
// CHECK-NEXT: store %5, %2[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i1 = 0 to 10 {
// CHECK-NEXT: %6 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: store %cst, %0[%6] : memref<1xf32>
// CHECK-NEXT: %7 = affine.apply [[MAP0]](%i1, %i1)
// CHECK-NEXT: %8 = load %0[%7] : memref<1xf32>
// CHECK-NEXT: store %8, %2[%i1] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1) -> (-d0 + d1)
// CHECK-LABEL: func @should_fuse_after_one_loop_interchange() {
func @should_fuse_after_one_loop_interchange() {
%a = alloc() : memref<10xf32>
%cf0 = constant 0.0 : f32
for %i0 = 0 to 10 {
store %cf0, %a[%i0] : memref<10xf32>
}
for %i1 = 0 to 5 {
for %i2 = 0 to 10 {
%v0 = load %a[%i2] : memref<10xf32>
store %v0, %a[%i2] : memref<10xf32>
}
}
// The dependence between the load and store is carried on loop '%i1', and
// cannot be fused with loop '%i0' without violating this dependence.
// Once loops '%i1' and %i2' are interchanged, loop '%i0' can be fused
// at loop depth 1, because the loop carrying the dependence has been
// interchanged and is now at depth 2.
// CHECK: for %i0 = 0 to 10 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %cst, %0[%1] : memref<1xf32>
// CHECK-NEXT: for %i1 = 0 to 5 {
// CHECK-NEXT: %2 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: %3 = load %0[%2] : memref<1xf32>
// CHECK-NEXT: %4 = affine.apply [[MAP0]](%i0, %i0)
// CHECK-NEXT: store %3, %0[%4] : memref<1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
// CHECK: [[MAP0:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d0 + d2)
// CHECK: [[MAP1:#map[0-9]+]] = (d0, d1, d2, d3) -> (-d1 + d3)
// CHECK-LABEL: func @should_fuse_after_two_loop_interchanges() {
func @should_fuse_after_two_loop_interchanges() {
%a = alloc() : memref<6x8xf32>
%cf0 = constant 0.0 : f32
for %i0 = 0 to 6 {
for %i1 = 0 to 8 {
store %cf0, %a[%i0, %i1] : memref<6x8xf32>
}
}
for %i2 = 0 to 4 {
for %i3 = 0 to 6 {
for %i4 = 0 to 2 {
for %i5 = 0 to 8 {
%v0 = load %a[%i3, %i5] : memref<6x8xf32>
%v1 = addf %v0, %v0 : f32
store %v1, %a[%i3, %i5] : memref<6x8xf32>
}
}
}
}
// The dependence between the load and store is carried on loops '%i2' and
// '%i4', and cannot be fused with loop '%i0' without violating this
// dependence.
// Once loop '%i2' is interchanged with loop '%i3', and again with loop
// '%i5', then loop '%i0' can be fused at loop depth 2, because the loop
// carring the dependences have been interchanged with loops at depth > 2.
// CHECK: for %i0 = 0 to 6 {
// CHECK-NEXT: for %i1 = 0 to 8 {
// CHECK-NEXT: %1 = affine.apply [[MAP0]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %2 = affine.apply [[MAP1]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: store %cst, %0[%1, %2] : memref<1x1xf32>
// CHECK-NEXT: for %i2 = 0 to 4 {
// CHECK-NEXT: for %i3 = 0 to 2 {
// CHECK-NEXT: %3 = affine.apply [[MAP0]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %4 = affine.apply [[MAP1]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %5 = load %0[%3, %4] : memref<1x1xf32>
// CHECK-NEXT: %6 = addf %5, %5 : f32
// CHECK-NEXT: %7 = affine.apply [[MAP0]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: %8 = affine.apply [[MAP1]](%i0, %i1, %i0, %i1)
// CHECK-NEXT: store %6, %0[%7, %8] : memref<1x1xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return
return
}
// -----
func @should_fuse_live_out_writer(%arg0 : memref<10xf32>) -> memref<10xf32> {
%cst = constant 0.000000e+00 : f32
for %i0 = 0 to 10 {
store %cst, %arg0[%i0] : memref<10xf32>
}
for %i1 = 0 to 10 {
%1 = load %arg0[%i1] : memref<10xf32>
store %1, %arg0[%i1] : memref<10xf32>
}
return %arg0 : memref<10xf32>
// CHECK: %cst = constant 0.000000e+00 : f32
// CHECK-NEXT: for %i0 = 0 to 10 {
// CHECK-NEXT: store %cst, %arg0[%i0] : memref<10xf32>
// CHECK-NEXT: %0 = load %arg0[%i0] : memref<10xf32>
// CHECK-NEXT: store %0, %arg0[%i0] : memref<10xf32>
// CHECK-NEXT: }
// CHECK-NEXT: return %arg0 : memref<10xf32>
}
// -----
// The fused slice has 16 iterations from along %i0.
// CHECK-DAG: [[MAP_LB:#map[0-9]+]] = (d0) -> (d0 * 16)
// CHECK-DAG: [[MAP_UB:#map[0-9]+]] = (d0) -> (d0 * 16 + 16)
#map = (d0, d1) -> (d0 * 16 + d1)
// CHECK-LABEL: slice_tile
func @slice_tile(%arg1: memref<32x8xf32>, %arg2: memref<32x8xf32>, %0 : f32) -> memref<32x8xf32> {
for %i0 = 0 to 32 {
for %i1 = 0 to 8 {
store %0, %arg2[%i0, %i1] : memref<32x8xf32>
}
}
for %i = 0 to 2 {
for %j = 0 to 8 {
for %k = 0 to 8 {
for %kk = 0 to 16 {
%1 = affine.apply #map(%k, %kk)
%2 = load %arg1[%1, %j] : memref<32x8xf32>
%3 = "foo"(%2) : (f32) -> f32
}
for %ii = 0 to 16 {
%6 = affine.apply #map(%i, %ii)
%7 = load %arg2[%6, %j] : memref<32x8xf32>
%8 = addf %7, %7 : f32
store %8, %arg2[%6, %j] : memref<32x8xf32>
}
}
}
}
return %arg2 : memref<32x8xf32>
}
// CHECK: for %i0 = 0 to 2 {
// CHECK-NEXT: for %i1 = 0 to 8 {
// CHECK-NEXT: for %i2 = [[MAP_LB]](%i0) to [[MAP_UB]](%i0) {
// CHECK-NEXT: store %arg2, %arg1[%i2, %i1] : memref<32x8xf32>
// CHECK-NEXT: }
// CHECK-NEXT: for %i3 = 0 to 8 {
// CHECK-NEXT: for %i4 = 0 to 16 {
// CHECK-NEXT: %0 = affine.apply #map{{[0-9]+}}(%i3, %i4)
// CHECK-NEXT: %1 = load %arg0[%0, %i1] : memref<32x8xf32>
// CHECK-NEXT: %2 = "foo"(%1) : (f32) -> f32
// CHECK-NEXT: }
// CHECK-NEXT: for %i5 = 0 to 16 {
// CHECK-NEXT: %3 = affine.apply #map{{[0-9]+}}(%i0, %i5)
// CHECK-NEXT: %4 = load %arg1[%3, %i1] : memref<32x8xf32>
// CHECK-NEXT: %5 = addf %4, %4 : f32
// CHECK-NEXT: store %5, %arg1[%3, %i1] : memref<32x8xf32>
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: }
// CHECK-NEXT: return %arg1 : memref<32x8xf32>
// CHECK-NEXT:}