mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-10-22 12:33:33 +02:00
8e0ed2340c
Summary: This change enables frame pointer elimination in non-leaf functions. The -fomit-frame-pointer option still needs to be used when compiling via clang (or an equivalent method of not setting the 'no-frame-pointer-elim*' function attributes if generating llvm IR via some other method) to take advantage of this optimization. This change should be NFC when compiling via clang without -fomit-frame-pointer. Reviewers: t.p.northover Subscribers: aemerson, rengolin, tberghammer, qcolombet, llvm-commits, danalbert, mcrosier, srhines Differential Revision: http://reviews.llvm.org/D17730 llvm-svn: 262495
38 lines
1.1 KiB
LLVM
38 lines
1.1 KiB
LLVM
; RUN: llc -verify-machineinstrs -mtriple=arm64-apple-ios7.0 -disable-fp-elim -o - %s | FileCheck %s
|
|
|
|
; When generating DAG selection tables, TableGen used to only flag an
|
|
; instruction as needing a chain on its own account if it had a built-in pattern
|
|
; which used the chain. This meant that the AArch64 load/stores weren't
|
|
; recognised and so both loads from %locvar below were coalesced into a single
|
|
; LS8_LDR instruction (same operands other than the non-existent chain) and the
|
|
; increment was lost at return.
|
|
|
|
; This was obviously a Bad Thing.
|
|
|
|
declare void @bar(i8*)
|
|
|
|
define i64 @test_chains() {
|
|
; CHECK-LABEL: test_chains:
|
|
|
|
%locvar = alloca i8
|
|
|
|
call void @bar(i8* %locvar)
|
|
; CHECK: bl {{_?bar}}
|
|
|
|
%inc.1 = load i8, i8* %locvar
|
|
%inc.2 = zext i8 %inc.1 to i64
|
|
%inc.3 = add i64 %inc.2, 1
|
|
%inc.4 = trunc i64 %inc.3 to i8
|
|
store i8 %inc.4, i8* %locvar
|
|
|
|
; CHECK: ldurb {{w[0-9]+}}, [x29, [[LOCADDR:#-?[0-9]+]]]
|
|
; CHECK: add {{w[0-9]+}}, {{w[0-9]+}}, #1
|
|
; CHECK: sturb w[[STRVAL:[0-9]+]], [x29, [[LOCADDR]]]
|
|
; CHECK; and w0, w[[STRVAL]], #0xff
|
|
|
|
%ret.1 = load i8, i8* %locvar
|
|
%ret.2 = zext i8 %ret.1 to i64
|
|
ret i64 %ret.2
|
|
; CHECK: ret
|
|
}
|