Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[LV] Drop incorrect inbounds for reverse vector pointer when folding tail #120730

Merged
merged 21 commits into from
Jan 7, 2025
Merged

[LV] Drop incorrect inbounds for reverse vector pointer when folding tail #120730

Changes from 2 commits
Commits
Show all changes
21 commits
Select commit Hold shift + click to select a range
466b722
[𝘀𝗽𝗿] initial version
fmayer Dec 20, 2024
50784b9
[𝘀𝗽𝗿] changes to main this commit is based on
fmayer Dec 20, 2024
5987219
simplify
fmayer Dec 20, 2024
e4e48cf
address comment
fmayer Dec 20, 2024
7631b15
[𝘀𝗽𝗿] changes introduced through rebase
fmayer Dec 20, 2024
2575f7f
address commentsrebase
fmayer Dec 20, 2024
8897101
address
fmayer Dec 20, 2024
e16f349
fmt
fmayer Dec 20, 2024
24ec5a2
tests
fmayer Dec 20, 2024
cbcb7db
test
fmayer Dec 20, 2024
b2f212d
[𝘀𝗽𝗿] changes introduced through rebase
fmayer Dec 20, 2024
c2e2812
fix tests i accidentally didn't run because they depend on assertions…
fmayer Dec 20, 2024
338cea0
[𝘀𝗽𝗿] changes introduced through rebase
fmayer Jan 3, 2025
4f8aec1
rebase
fmayer Jan 3, 2025
1d6843d
[𝘀𝗽𝗿] changes introduced through rebase
fmayer Jan 3, 2025
8b6a170
rebase
fmayer Jan 3, 2025
32eade6
[𝘀𝗽𝗿] changes introduced through rebase
fmayer Jan 3, 2025
c841374
rebase
fmayer Jan 3, 2025
2993e44
comment
fmayer Jan 6, 2025
f55fdb1
fmt
fmayer Jan 6, 2025
410dad6
cmt
fmayer Jan 6, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
61 changes: 29 additions & 32 deletions ...rms/LoopVectorize/vplan-noinbounds-gep.ll → ...LoopVectorize/X86/vplan-noinbounds-gep.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,16 +1,14 @@
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 3
; RUN: opt -passes=loop-vectorize -force-vector-width=4 -enable-vplan-native-path -S %s | FileCheck %s
; RUN: opt -passes=loop-vectorize -force-vector-width=4 -S %s | FileCheck %s

source_filename = "<stdin>"
target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"

; Function Attrs: nofree norecurse nosync nounwind memory(none)
define i1 @fn() local_unnamed_addr #0 {
define i1 @fn(ptr %nno) local_unnamed_addr #0 {
; CHECK-LABEL: define i1 @fn(
; CHECK-SAME: ) local_unnamed_addr #[[ATTR0:[0-9]+]] {
; CHECK-SAME: ptr [[NNO:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: entry:
; CHECK-NEXT: [[NNO:%.*]] = alloca [12 x i32], align 16
; CHECK-NEXT: br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK: vector.ph:
; CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
Expand All @@ -19,15 +17,15 @@ define i1 @fn() local_unnamed_addr #0 {
; CHECK-NEXT: [[VEC_IND:%.*]] = phi <4 x i64> [ <i64 10, i64 9, i64 8, i64 7>, [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP11:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT: [[OFFSET_IDX:%.*]] = sub i64 10, [[INDEX]]
; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[TMP22:%.*]] = add i64 [[OFFSET_IDX]], 0
; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i64> poison, i64 [[INDEX]], i64 0
; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i64> [[BROADCAST_SPLATINSERT]], <4 x i64> poison, <4 x i32> zeroinitializer
; CHECK-NEXT: [[VEC_IV:%.*]] = add <4 x i64> [[BROADCAST_SPLAT]], <i64 0, i64 1, i64 2, i64 3>
; CHECK-NEXT: [[TMP1:%.*]] = icmp ule <4 x i64> [[VEC_IV]], splat (i64 10)
; CHECK-NEXT: [[TMP2:%.*]] = and <4 x i64> [[VEC_IND]], splat (i64 1)
; CHECK-NEXT: [[TMP3:%.*]] = icmp eq <4 x i64> [[TMP2]], zeroinitializer
; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds nuw [12 x i32], ptr [[NNO]], i64 0, i64 [[TMP0]]
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr i32, ptr [[TMP4]], i32 0
; CHECK-NEXT: [[TMP23:%.*]] = getelementptr inbounds nuw [12 x i32], ptr [[NNO]], i64 0, i64 [[TMP22]]
; CHECK-NEXT: [[TMP5:%.*]] = getelementptr i32, ptr [[TMP23]], i32 0
; CHECK-NEXT: [[TMP6:%.*]] = getelementptr i32, ptr [[TMP5]], i32 -3
; CHECK-NEXT: [[REVERSE:%.*]] = shufflevector <4 x i1> [[TMP1]], <4 x i1> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP6]], i32 4, <4 x i1> [[REVERSE]], <4 x i32> poison)
Expand All @@ -50,56 +48,55 @@ define i1 @fn() local_unnamed_addr #0 {
; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ -2, [[MIDDLE_BLOCK]] ], [ 10, [[ENTRY:%.*]] ]
; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP14]], [[MIDDLE_BLOCK]] ], [ 0, [[ENTRY]] ]
; CHECK-NEXT: br label [[FOR_BODY20:%.*]]
; CHECK: for.body20:
; CHECK: loop.header:
; CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_INC35:%.*]] ]
; CHECK-NEXT: [[SUM_01:%.*]] = phi i32 [ [[BC_MERGE_RDX]], [[SCALAR_PH]] ], [ [[SUM_1:%.*]], [[FOR_INC35]] ]
; CHECK-NEXT: [[REM4:%.*]] = and i64 [[INDVARS_IV]], 1
; CHECK-NEXT: [[CMP21:%.*]] = icmp eq i64 [[REM4]], 0
; CHECK-NEXT: [[ARRAYIDX24:%.*]] = getelementptr inbounds nuw [12 x i32], ptr [[NNO]], i64 0, i64 [[INDVARS_IV]]
; CHECK-NEXT: [[TMP15:%.*]] = load i32, ptr [[ARRAYIDX24]], align 4
; CHECK-NEXT: [[GEP:%.*]] = getelementptr inbounds nuw [12 x i32], ptr [[NNO]], i64 0, i64 [[INDVARS_IV]]
; CHECK-NEXT: [[TMP15:%.*]] = load i32, ptr [[GEP]], align 4
; CHECK-NEXT: br i1 [[CMP21]], label [[IF_THEN22:%.*]], label [[FOR_INC35]]
; CHECK: if.then22:
; CHECK: if.then:
; CHECK-NEXT: [[MUL:%.*]] = shl i32 [[TMP15]], 1
; CHECK-NEXT: [[REM27:%.*]] = urem i32 [[MUL]], 10
; CHECK-NEXT: br label [[FOR_INC35]]
; CHECK: for.inc35:
; CHECK: loop.latch:
; CHECK-NEXT: [[REM27_PN:%.*]] = phi i32 [ [[REM27]], [[IF_THEN22]] ], [ [[TMP15]], [[FOR_BODY20]] ]
; CHECK-NEXT: [[SUM_1]] = or i32 [[REM27_PN]], [[SUM_01]]
; CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nsw i64 [[INDVARS_IV]], -1
; CHECK-NEXT: [[CMP19_NOT:%.*]] = icmp eq i64 [[INDVARS_IV]], 0
; CHECK-NEXT: br i1 [[CMP19_NOT]], label [[FOR_END36]], label [[FOR_BODY20]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK: for.end36:
; CHECK: exit:
; CHECK-NEXT: [[SUM_1_LCSSA:%.*]] = phi i32 [ [[SUM_1]], [[FOR_INC35]] ], [ [[TMP14]], [[MIDDLE_BLOCK]] ]
; CHECK-NEXT: [[CMP41:%.*]] = icmp eq i32 [[SUM_1_LCSSA]], 0
; CHECK-NEXT: ret i1 [[CMP41]]
;
entry:
%nno = alloca [12 x i32], align 16
br label %for.body20
br label %loop.header

for.body20: ; preds = %entry, %for.inc35
%indvars.iv = phi i64 [ 10, %entry ], [ %indvars.iv.next, %for.inc35 ]
%sum.01 = phi i32 [ 0, %entry ], [ %sum.1, %for.inc35 ]
%rem4 = and i64 %indvars.iv, 1
loop.header: ; preds = %entry, %loop.latch
%iv = phi i64 [ 10, %entry ], [ %iv.next, %loop.latch ]
%sum.01 = phi i32 [ 0, %entry ], [ %sum.1, %loop.latch ]
%rem4 = and i64 %iv, 1
%cmp21 = icmp eq i64 %rem4, 0
%arrayidx24 = getelementptr inbounds nuw [12 x i32], ptr %nno, i64 0, i64 %indvars.iv
%0 = load i32, ptr %arrayidx24, align 4
br i1 %cmp21, label %if.then22, label %for.inc35
%gep = getelementptr inbounds nuw [12 x i32], ptr %nno, i64 0, i64 %iv
%0 = load i32, ptr %gep, align 4
br i1 %cmp21, label %if.then, label %loop.latch

if.then22: ; preds = %for.body20
if.then: ; preds = %loop.header
%mul = shl i32 %0, 1
%rem27 = urem i32 %mul, 10
br label %for.inc35
br label %loop.latch

for.inc35: ; preds = %for.body20, %if.then22
%rem27.pn = phi i32 [ %rem27, %if.then22 ], [ %0, %for.body20 ]
loop.latch: ; preds = %loop.header, %if.then
%rem27.pn = phi i32 [ %rem27, %if.then ], [ %0, %loop.header ]
%sum.1 = or i32 %rem27.pn, %sum.01
%indvars.iv.next = add nsw i64 %indvars.iv, -1
%cmp19.not = icmp eq i64 %indvars.iv, 0
br i1 %cmp19.not, label %for.end36, label %for.body20
%iv.next = add nsw i64 %iv, -1
%cmp19.not = icmp eq i64 %iv, 0
br i1 %cmp19.not, label %exit, label %loop.header

for.end36: ; preds = %for.inc35
%sum.1.lcssa = phi i32 [ %sum.1, %for.inc35 ]
exit: ; preds = %loop.latch
%sum.1.lcssa = phi i32 [ %sum.1, %loop.latch ]
%cmp41 = icmp eq i32 %sum.1.lcssa, 0
ret i1 %cmp41
}
Expand Down
Loading