497a6d6722
This reverts commit ed004cf42bf57ca79b57bc3076ef83a8477426ea. The original commit exposed an independent cost issue, triggering an assertion. That issue has been fixed in 3457e7efc3. Reland the patch now that the assertion has been fixed.
215 lines
7.2 KiB
LLVM
215 lines
7.2 KiB
LLVM
; REQUIRES: asserts
|
|
; RUN: opt < %s -passes=loop-vectorize -debug-only=loop-vectorize -disable-output -S 2>&1 | FileCheck %s
|
|
|
|
target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
|
|
target triple = "arm64-apple-ios5.0.0"
|
|
|
|
define void @selects_1(ptr nocapture %dst, i32 %A, i32 %B, i32 %C, i32 %N) {
|
|
; CHECK: LV: Checking a loop in 'selects_1'
|
|
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%cond> = select ir<%cmp1>, ir<10>, ir<%and>
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%cond6> = select ir<%cmp2>, ir<30>, ir<%and>
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%cond11> = select ir<%cmp7>, ir<%cond>, ir<%cond6>
|
|
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%cond> = select ir<%cmp1>, ir<10>, ir<%and>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%cond6> = select ir<%cmp2>, ir<30>, ir<%and>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%cond11> = select ir<%cmp7>, ir<%cond>, ir<%cond6>
|
|
|
|
; CHECK: LV: Selecting VF: 4
|
|
|
|
entry:
|
|
%n = zext i32 %N to i64
|
|
br label %loop
|
|
|
|
loop:
|
|
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
|
|
%arrayidx = getelementptr inbounds i32, ptr %dst, i64 %iv
|
|
%0 = load i32, ptr %arrayidx, align 4
|
|
%and = and i32 %0, 2047
|
|
%cmp1 = icmp eq i32 %and, %A
|
|
%cond = select i1 %cmp1, i32 10, i32 %and
|
|
%cmp2 = icmp eq i32 %and, %B
|
|
%cond6 = select i1 %cmp2, i32 30, i32 %and
|
|
%cmp7 = icmp ugt i32 %cond, %C
|
|
%cond11 = select i1 %cmp7, i32 %cond, i32 %cond6
|
|
store i32 %cond11, ptr %arrayidx, align 4
|
|
%iv.next = add nuw nsw i64 %iv, 1
|
|
%ec = icmp eq i64 %iv.next, %n
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
ret void
|
|
}
|
|
|
|
define i32 @multi_user_cmp(ptr readonly %a, i64 noundef %n) {
|
|
; CHECK: LV: Checking a loop in 'multi_user_cmp'
|
|
; CHECK: Cost of 4 for VF 16: WIDEN ir<%cmp1> = fcmp olt ir<%load1>, ir<0.000000e+00>
|
|
; CHECK: LV: Selecting VF: 16.
|
|
entry:
|
|
br label %loop
|
|
|
|
loop:
|
|
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
|
|
%all.off.next = phi i1 [ true, %entry ], [ %all.off, %loop ]
|
|
%any.0.off09 = phi i1 [ false, %entry ], [ %.any.0.off0, %loop ]
|
|
%arrayidx = getelementptr inbounds float, ptr %a, i64 %iv
|
|
%load1 = load float, ptr %arrayidx, align 4
|
|
%cmp1 = fcmp olt float %load1, 0.000000e+00
|
|
%.any.0.off0 = select i1 %cmp1, i1 true, i1 %any.0.off09
|
|
%all.off = select i1 %cmp1, i1 %all.off.next, i1 false
|
|
%iv.next = add nuw nsw i64 %iv, 1
|
|
%ec = icmp eq i64 %iv.next, %n
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
%0 = select i1 %.any.0.off0, i32 2, i32 3
|
|
%1 = select i1 %all.off, i32 1, i32 %0
|
|
ret i32 %1
|
|
}
|
|
|
|
define i32 @select_vpinst_for_tail_folding(i8 %n) {
|
|
; CHECK: LV: Checking a loop in 'select_vpinst_for_tail_folding'
|
|
; CHECK: Cost of 1 for VF 2: EMIT vp<{{.+}}> = select vp<{{.+}}>, ir<%red.next>, ir<%red>
|
|
; CHECK: Cost of 1 for VF 4: EMIT vp<{{.+}}> = select vp<{{.+}}>, ir<%red.next>, ir<%red>
|
|
; CHECK: LV: Selecting VF: 4
|
|
|
|
entry:
|
|
%c = icmp ne i8 %n, 0
|
|
%ext = zext i1 %c to i32
|
|
br label %loop
|
|
|
|
loop:
|
|
%iv = phi i32 [ %ext, %entry ], [ %iv.next, %loop ]
|
|
%red = phi i32 [ 0, %entry ], [ %red.next, %loop ]
|
|
%iv.next = add i32 %iv, 1
|
|
%red.next = mul i32 %red, %iv
|
|
%ec = icmp eq i32 %iv, 12
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
ret i32 %red.next
|
|
}
|
|
|
|
define i32 @select_xor_cond(ptr %src, i1 %c.0) {
|
|
; CHECK: LV: Checking a loop in 'select_xor_cond'
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%sel> = select ir<%c>, ir<false>, ir<%c.0>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%sel> = select ir<%c>, ir<false>, ir<%c.0>
|
|
; CHECK: Cost of 1 for VF 8: WIDEN ir<%sel> = select ir<%c>, ir<false>, ir<%c.0>
|
|
; CHECK: Cost of 1 for VF 16: WIDEN ir<%sel> = select ir<%c>, ir<false>, ir<%c.0>
|
|
; CHECK: LV: Selecting VF: 4.
|
|
|
|
entry:
|
|
br label %loop
|
|
|
|
loop:
|
|
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
|
|
%p = phi i32 [ 0, %entry ], [ %ext.sel, %loop ]
|
|
%gep.src = getelementptr i8, ptr %src, i64 %iv
|
|
%0 = load i8, ptr %gep.src, align 1
|
|
%c = icmp eq i8 %0, 0
|
|
%not.c = xor i1 %c, true
|
|
%sel = select i1 %not.c, i1 %c.0, i1 false
|
|
%ext.sel = zext i1 %sel to i32
|
|
%iv.next = add i64 %iv, 1
|
|
%ec = icmp eq i64 %iv, 19
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
%ext.c = zext i1 %c to i32
|
|
%res = add i32 %ext.c, %p
|
|
ret i32 %res
|
|
}
|
|
|
|
define void @select_invariant_cmp_cond(ptr %dst, ptr %src, i32 %a, i32 %b, i64 %n) "target-cpu"="neoverse-v2" {
|
|
; CHECK: LV: Checking a loop in 'select_invariant_cmp_cond'
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%sel> = select ir<%cmp>, ir<%trunc>, ir<0>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%sel> = select ir<%cmp>, ir<%trunc>, ir<0>
|
|
; CHECK: LV: Selecting VF: 4.
|
|
entry:
|
|
%cmp = icmp ugt i32 %a, %b
|
|
br label %loop
|
|
|
|
loop:
|
|
%iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ]
|
|
%gep.src = getelementptr inbounds i32, ptr %src, i64 %iv
|
|
%load = load i32, ptr %gep.src, align 4
|
|
%conv = sext i32 %load to i64
|
|
%or = or i64 %conv, 1
|
|
%trunc = trunc i64 %or to i32
|
|
%sel = select i1 %cmp, i32 %trunc, i32 0
|
|
%gep.dst = getelementptr inbounds i32, ptr %dst, i64 %iv
|
|
store i32 %sel, ptr %gep.dst, align 4
|
|
%iv.next = add nuw nsw i64 %iv, 1
|
|
%ec = icmp eq i64 %iv.next, %n
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
ret void
|
|
}
|
|
|
|
declare void @use(i64, i32)
|
|
|
|
define void @logical_and_select_inverted(i1 %cmp, ptr %start, ptr %end) {
|
|
; CHECK: LV: Checking a loop in 'logical_and_select_inverted'
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%narrow> = select ir<%trunc>, ir<false>, ir<%cmp>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%narrow> = select ir<%trunc>, ir<false>, ir<%cmp>
|
|
; CHECK: Cost of 1 for VF 8: WIDEN ir<%narrow> = select ir<%trunc>, ir<false>, ir<%cmp>
|
|
; CHECK: Cost of 1 for VF 16: WIDEN ir<%narrow> = select ir<%trunc>, ir<false>, ir<%cmp>
|
|
; CHECK: LV: Selecting VF: 16.
|
|
|
|
entry:
|
|
br label %loop
|
|
|
|
loop:
|
|
%for = phi i64 [ %zext, %loop ], [ 0, %entry ]
|
|
%p = phi i32 [ %spec.select, %loop ], [ 0, %entry ]
|
|
%ptr = phi ptr [ %ptr.next, %loop ], [ %start, %entry ]
|
|
%ld = load i8, ptr %ptr, align 8
|
|
%trunc = trunc i8 %ld to i1
|
|
%not = xor i1 %trunc, true
|
|
%zext = zext i1 %not to i64
|
|
%or.1 = or i64 %for, 1
|
|
%narrow = select i1 %not, i1 %cmp, i1 false
|
|
%spec.select = zext i1 %narrow to i32
|
|
%or.2 = or i32 %p, 1
|
|
%ptr.next = getelementptr i8, ptr %ptr, i64 40
|
|
%ec = icmp eq ptr %ptr.next, %end
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
call void @use(i64 %or.1, i32 %or.2)
|
|
ret void
|
|
}
|
|
|
|
define void @logical_or_select_inverted(i1 %cmp, ptr %start, ptr %end) {
|
|
; CHECK: LV: Checking a loop in 'logical_or_select_inverted'
|
|
; CHECK: Cost of 1 for VF 2: WIDEN ir<%narrow> = select ir<%trunc>, ir<%cmp>, ir<true>
|
|
; CHECK: Cost of 1 for VF 4: WIDEN ir<%narrow> = select ir<%trunc>, ir<%cmp>, ir<true>
|
|
; CHECK: Cost of 1 for VF 8: WIDEN ir<%narrow> = select ir<%trunc>, ir<%cmp>, ir<true>
|
|
; CHECK: Cost of 1 for VF 16: WIDEN ir<%narrow> = select ir<%trunc>, ir<%cmp>, ir<true>
|
|
; CHECK: LV: Selecting VF: 16.
|
|
|
|
entry:
|
|
br label %loop
|
|
|
|
loop:
|
|
%for = phi i64 [ %zext, %loop ], [ 0, %entry ]
|
|
%p = phi i32 [ %spec.select, %loop ], [ 0, %entry ]
|
|
%ptr = phi ptr [ %ptr.next, %loop ], [ %start, %entry ]
|
|
%ld = load i8, ptr %ptr, align 8
|
|
%trunc = trunc i8 %ld to i1
|
|
%not = xor i1 %trunc, true
|
|
%zext = zext i1 %not to i64
|
|
%or.1 = or i64 %for, 1
|
|
%narrow = select i1 %not, i1 true, i1 %cmp
|
|
%spec.select = zext i1 %narrow to i32
|
|
%or.2 = or i32 %p, 1
|
|
%ptr.next = getelementptr i8, ptr %ptr, i64 40
|
|
%ec = icmp eq ptr %ptr.next, %end
|
|
br i1 %ec, label %exit, label %loop
|
|
|
|
exit:
|
|
call void @use(i64 %or.1, i32 %or.2)
|
|
ret void
|
|
}
|