derek.holloway
7502bf64a4
CodeQL / Analyze (actions) (push) Canceled after 0s
CodeQL / Analyze (python) (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Multi-Arch CI / setup (push) Canceled after 0s
Publish build_manylinux_x86_64 images / publish_build_manylinux_x86_64 (push) Canceled after 0s
Setup (Multi-Arch) / setup (push) Canceled after 0s
Multi-Arch CI / Linux::${{ fromJSON(needs.setup.outputs.linux_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubi10 image / publish_no_rocm_image_ubi10 (push) Canceled after 0s
Multi-Arch CI / Windows::${{ fromJSON(needs.setup.outputs.windows_build_config || '{}').build_variant_label || 'skip' }} (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Multi-Arch CI / Manifest Diff (push) Canceled after 0s
TheRock Manifest Diff Report / Generate Manifest Diff Report (push) Canceled after 0s
Multi-Arch CI / CI Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_ocl_rt images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04_rocgdb images / publish_no_rocm_image_ubuntu24_04_rocgdb (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_ocl_rt (push) Canceled after 0s
Publish TheRock Docker image / build-and-push-image (push) Canceled after 0s
Unit Tests / Unit Tests :: ubuntu-24.04 (push) Canceled after 0s
Unit Tests / Unit Tests :: windows-2022 (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_media (push) Canceled after 0s
Unit Tests / Unit Tests Summary (push) Canceled after 0s
Publish no_rocm_image_ubuntu24_04 images / publish_no_rocm_image_ubuntu24_04_openmpi (push) Canceled after 0s
118 lines
4.8 KiB
LLVM
118 lines
4.8 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --version 6
|
|
; RUN: opt < %s "-passes=print<scalar-evolution>" -disable-output 2>&1 | FileCheck %s
|
|
|
|
; Check that, when constructing a SCEV for rounding a value up to the nearest
|
|
; multiple of a constant, we always use that largest possible value for the
|
|
; added constant, even if instcombine has cleared some bits in the IR because
|
|
; the input is already known to be a multiple of some smaller value.
|
|
|
|
define i32 @mul_2_round_to_16_a(i32 %val) {
|
|
; CHECK-LABEL: 'mul_2_round_to_16_a'
|
|
; CHECK-NEXT: Classifying expressions for: @mul_2_round_to_16_a
|
|
; CHECK-NEXT: %mul = mul i32 %val, 2
|
|
; CHECK-NEXT: --> (2 * %val) U: [0,-1) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: %add = add i32 %mul, 15
|
|
; CHECK-NEXT: --> (15 + (2 * %val)) U: [15,14) S: [-2147483633,-2147483634)
|
|
; CHECK-NEXT: %round = and i32 %add, -16
|
|
; CHECK-NEXT: --> (16 * ((15 + (2 * %val)) /u 16))<nuw> U: [0,-15) S: [-2147483648,2147483633)
|
|
; CHECK-NEXT: Determining loop execution counts for: @mul_2_round_to_16_a
|
|
;
|
|
%mul = mul i32 %val, 2
|
|
%add = add i32 %mul, 15
|
|
%round = and i32 %add, -16
|
|
ret i32 %round
|
|
}
|
|
|
|
define i32 @mul_2_round_to_16_b(i32 %val) {
|
|
; CHECK-LABEL: 'mul_2_round_to_16_b'
|
|
; CHECK-NEXT: Classifying expressions for: @mul_2_round_to_16_b
|
|
; CHECK-NEXT: %mul = mul i32 %val, 2
|
|
; CHECK-NEXT: --> (2 * %val) U: [0,-1) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: %add = add i32 %mul, 14
|
|
; CHECK-NEXT: --> (14 + (2 * %val)) U: [0,-1) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: %round = and i32 %add, -16
|
|
; CHECK-NEXT: --> (16 * ((15 + (2 * %val)) /u 16))<nuw> U: [0,-15) S: [-2147483648,2147483633)
|
|
; CHECK-NEXT: Determining loop execution counts for: @mul_2_round_to_16_b
|
|
;
|
|
%mul = mul i32 %val, 2
|
|
%add = add i32 %mul, 14
|
|
%round = and i32 %add, -16
|
|
ret i32 %round
|
|
}
|
|
|
|
define i32 @mul_2_round_to_16_c(i32 %val) {
|
|
; CHECK-LABEL: 'mul_2_round_to_16_c'
|
|
; CHECK-NEXT: Classifying expressions for: @mul_2_round_to_16_c
|
|
; CHECK-NEXT: %mul = mul i32 %val, 2
|
|
; CHECK-NEXT: --> (2 * %val) U: [0,-1) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: %add = add i32 %mul, 14
|
|
; CHECK-NEXT: --> (14 + (2 * %val)) U: [0,-1) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: %div = udiv i32 %add, 16
|
|
; CHECK-NEXT: --> ((15 + (2 * %val)) /u 16) U: [0,268435456) S: [0,268435456)
|
|
; CHECK-NEXT: %round = mul i32 %div, 16
|
|
; CHECK-NEXT: --> (16 * ((15 + (2 * %val)) /u 16))<nuw> U: [0,-15) S: [-2147483648,2147483633)
|
|
; CHECK-NEXT: Determining loop execution counts for: @mul_2_round_to_16_c
|
|
;
|
|
%mul = mul i32 %val, 2
|
|
%add = add i32 %mul, 14
|
|
%div = udiv i32 %add, 16
|
|
%round = mul i32 %div, 16
|
|
ret i32 %round
|
|
}
|
|
|
|
define i32 @mul_4_round_to_16(i32 %val) {
|
|
; CHECK-LABEL: 'mul_4_round_to_16'
|
|
; CHECK-NEXT: Classifying expressions for: @mul_4_round_to_16
|
|
; CHECK-NEXT: %mul = mul i32 %val, 4
|
|
; CHECK-NEXT: --> (4 * %val) U: [0,-3) S: [-2147483648,2147483645)
|
|
; CHECK-NEXT: %add = add i32 %mul, 12
|
|
; CHECK-NEXT: --> (12 + (4 * %val)) U: [0,-3) S: [-2147483648,2147483645)
|
|
; CHECK-NEXT: %round = and i32 %add, -16
|
|
; CHECK-NEXT: --> (16 * ((15 + (4 * %val)) /u 16))<nuw> U: [0,-15) S: [-2147483648,2147483633)
|
|
; CHECK-NEXT: Determining loop execution counts for: @mul_4_round_to_16
|
|
;
|
|
%mul = mul i32 %val, 4
|
|
%add = add i32 %mul, 12
|
|
%round = and i32 %add, -16
|
|
ret i32 %round
|
|
}
|
|
|
|
; N is not a power of 2, cannot do transformation
|
|
define i32 @invalid1(i32 %val) {
|
|
; CHECK-LABEL: 'invalid1'
|
|
; CHECK-NEXT: Classifying expressions for: @invalid1
|
|
; CHECK-NEXT: %mul = mul i32 %val, 3
|
|
; CHECK-NEXT: --> (3 * %val) U: full-set S: full-set
|
|
; CHECK-NEXT: %add = add i32 %mul, 7
|
|
; CHECK-NEXT: --> (7 + (3 * %val)) U: full-set S: full-set
|
|
; CHECK-NEXT: %div = udiv i32 %add, 9
|
|
; CHECK-NEXT: --> ((7 + (3 * %val)) /u 9) U: [0,477218589) S: [0,477218589)
|
|
; CHECK-NEXT: %round = mul i32 %div, 9
|
|
; CHECK-NEXT: --> (9 * ((7 + (3 * %val)) /u 9))<nuw> U: [0,-3) S: [0,-3)
|
|
; CHECK-NEXT: Determining loop execution counts for: @invalid1
|
|
;
|
|
%mul = mul i32 %val, 3
|
|
%add = add i32 %mul, 7
|
|
%div = udiv i32 %add, 9
|
|
%round = mul i32 %div, 9
|
|
ret i32 %round
|
|
}
|
|
|
|
; M is greater then N, cannot do transformation
|
|
define i32 @invalid2(i32 %val) {
|
|
; CHECK-LABEL: 'invalid2'
|
|
; CHECK-NEXT: Classifying expressions for: @invalid2
|
|
; CHECK-NEXT: %mul = mul i32 %val, 4
|
|
; CHECK-NEXT: --> (4 * %val) U: [0,-3) S: [-2147483648,2147483645)
|
|
; CHECK-NEXT: %add = add i32 %mul, 1
|
|
; CHECK-NEXT: --> (1 + (4 * %val))<nuw><nsw> U: [1,-2) S: [-2147483647,2147483646)
|
|
; CHECK-NEXT: %round = and i32 %add, -2
|
|
; CHECK-NEXT: --> (2 * ((1 + (4 * %val))<nuw><nsw> /u 2))<nuw> U: [0,-3) S: [-2147483648,2147483647)
|
|
; CHECK-NEXT: Determining loop execution counts for: @invalid2
|
|
;
|
|
%mul = mul i32 %val, 4
|
|
%add = add i32 %mul, 1
|
|
%round = and i32 %add, -2
|
|
ret i32 %round
|
|
}
|