llvm-mirror/test/CodeGen/AArch64/eliminate-trunc.ll
Wei Mi 4f28dc3b0c [LSR] Recommit: Allow formula containing Reg for SCEVAddRecExpr related with outerloop.
The recommit includes some changes of testcases. No functional change to the patch.

In RateRegister of existing LSR, if a formula contains a Reg which is a SCEVAddRecExpr,
and this SCEVAddRecExpr's loop is an outerloop, the formula will be marked as Loser
and dropped.

Suppose we have an IR that %for.body is outerloop and %for.body2 is innerloop. LSR only
handle inner loop now so only %for.body2 will be handled.

Using the logic above, formula like
reg(%array) + reg({1,+, %size}<%for.body>) + 1*reg({0,+,1}<%for.body2>) will be dropped
no matter what because reg({1,+, %size}<%for.body>) is a SCEVAddRecExpr type reg related
with outerloop. Only formula like
reg(%array) + 1*reg({{1,+, %size}<%for.body>,+,1}<nuw><nsw><%for.body2>) will be kept
because the SCEVAddRecExpr related with outerloop is folded into the initial value of the
SCEVAddRecExpr related with current loop.

But in some cases, we do need to share the basic induction variable
reg{0 ,+, 1}<%for.body2> among LSR Uses to reduce the final total number of induction
variables used by LSR, so we don't want to drop the formula like
reg(%array) + reg({1,+, %size}<%for.body>) + 1*reg({0,+,1}<%for.body2>) unconditionally.

From the existing comment, it tries to avoid considering multiple level loops at the same time.
However, existing LSR only handles innermost loop, so for any SCEVAddRecExpr with a loop other
than current loop, it is an invariant and will be simple to handle, and the formula doesn't have
to be dropped.

Differential Revision: https://reviews.llvm.org/D26429

llvm-svn: 294814
2017-02-11 00:50:23 +00:00

40 lines
1.6 KiB
LLVM

; RUN: llc -verify-machineinstrs -o - %s -mtriple=aarch64-none-apple-ios7.0 -mcpu=cyclone | FileCheck %s
; Check trunc i64 operation is translated as a subregister access
; eliminating an i32 induction varible.
; CHECK-NOT: add {{x[0-9]+}}, {{x[0-9]+}}, #1
; CHECK: add {{w[0-9]+}}, {{w[0-9]+}}, #1
; CHECK-NEXT: cmp {{w[0-9]+}}, {{w[0-9]+}}
define void @test1_signed([8 x i8]* nocapture %a, i8* nocapture readonly %box, i8 %limit, i64 %inv) minsize {
entry:
%conv = zext i8 %limit to i32
%cmp223 = icmp eq i8 %limit, 0
br i1 %cmp223, label %for.end15, label %for.body4.lr.ph.us
for.body4.us:
%indvars.iv = phi i64 [ 0, %for.body4.lr.ph.us ], [ %indvars.iv.next, %for.body4.us ]
%arrayidx6.us = getelementptr inbounds [8 x i8], [8 x i8]* %a, i64 %indvars.iv, i64 %inv
%0 = load i8, i8* %arrayidx6.us, align 1
%idxprom7.us = zext i8 %0 to i64
%arrayidx8.us = getelementptr inbounds i8, i8* %box, i64 %idxprom7.us
%1 = load i8, i8* %arrayidx8.us, align 1
store i8 %1, i8* %arrayidx6.us, align 1
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
%2 = trunc i64 %indvars.iv.next to i32
%cmp2.us = icmp slt i32 %2, %conv
br i1 %cmp2.us, label %for.body4.us, label %for.cond1.for.inc13_crit_edge.us
for.body4.lr.ph.us:
%indvars.iv26 = phi i64 [ %indvars.iv.next27, %for.cond1.for.inc13_crit_edge.us ], [ 0, %entry ]
br label %for.body4.us
for.cond1.for.inc13_crit_edge.us:
%indvars.iv.next27 = add nuw nsw i64 %indvars.iv26, 1
%exitcond28 = icmp eq i64 %indvars.iv26, 3
br i1 %exitcond28, label %for.end15, label %for.body4.lr.ph.us
for.end15:
ret void
}