RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-12-03 00:47:07 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Disable the vzeroupper insertion pass on PS4.

Browse Source 
Differential Revision: http://reviews.llvm.org/D16837

llvm-svn: 260764
This commit is contained in:
Yunzhong Gao 2016-02-12 23:37:57 +00:00
parent dec93a4fca
commit 9e56bc9706
5 changed files with 19 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
lib/Target/X86/X86.td
View File

@ -239,6 +239,11 @@ def FeatureSlowIncDec : SubtargetFeature<"slow-incdec", "SlowIncDec", "true",
def FeatureSoftFloat
: SubtargetFeature<"soft-float", "UseSoftFloat", "true",
"Use software floating point features.">;
// On at least some AMD processors, there is no performance hazard to writing
// only the lower parts of a YMM register without clearing the upper part.
def FeatureFastPartialYMMWrite
: SubtargetFeature<"fast-partial-ymm-write", "HasFastPartialYMMWrite",
"true", "Partial writes to YMM registers are fast">;
//===----------------------------------------------------------------------===//
// X86 processors supported.
@ -596,7 +601,8 @@ def : ProcessorModel<"btver2", BtVer2Model, [
FeatureXSAVE,
FeatureXSAVEOPT,
FeatureSlowSHLD,
FeatureLAHFSAHF
FeatureLAHFSAHF,
FeatureFastPartialYMMWrite
]>;
// Bulldozer

1
lib/Target/X86/X86Subtarget.cpp
View File

@ -285,6 +285,7 @@ void X86Subtarget::initializeEnvironment() {
HasSSEUnalignedMem = false;
HasCmpxchg16b = false;
UseLeaForSP = false;
HasFastPartialYMMWrite = false;
HasSlowDivide32 = false;
HasSlowDivide64 = false;
PadShortFunctions = false;

5
lib/Target/X86/X86Subtarget.h
View File

@ -189,6 +189,10 @@ protected:
/// the stack pointer. This is an optimization for Intel Atom processors.
bool UseLeaForSP;
/// True if there is no performance penalty to writing only the lower parts
/// of a YMM register without clearing the upper part.
bool HasFastPartialYMMWrite;
/// True if 8-bit divisions are significantly faster than
/// 32-bit divisions and should be used when possible.
bool HasSlowDivide32;
@ -421,6 +425,7 @@ public:
bool hasSSEUnalignedMem() const { return HasSSEUnalignedMem; }
bool hasCmpxchg16b() const { return HasCmpxchg16b; }
bool useLeaForSP() const { return UseLeaForSP; }
bool hasFastPartialYMMWrite() const { return HasFastPartialYMMWrite; }
bool hasSlowDivide32() const { return HasSlowDivide32; }
bool hasSlowDivide64() const { return HasSlowDivide64; }
bool padShortFunctions() const { return PadShortFunctions; }

2
lib/Target/X86/X86VZeroUpper.cpp
View File

@ -248,7 +248,7 @@ void VZeroUpperInserter::processBasicBlock(MachineBasicBlock &MBB) {
/// vzeroupper instructions before function calls.
bool VZeroUpperInserter::runOnMachineFunction(MachineFunction &MF) {
const X86Subtarget &ST = MF.getSubtarget<X86Subtarget>();
if (!ST.hasAVX() || ST.hasAVX512())
if (!ST.hasAVX() || ST.hasAVX512() || ST.hasFastPartialYMMWrite())
return false;
TII = ST.getInstrInfo();
MachineRegisterInfo &MRI = MF.getRegInfo();

5
test/CodeGen/X86/avx-vzeroupper.ll
View File

@ -1,4 +1,9 @@
; RUN: llc < %s -x86-use-vzeroupper -mtriple=x86_64-apple-darwin -mattr=+avx | FileCheck %s
; RUN: llc < %s -x86-use-vzeroupper -mtriple=x86_64-apple-darwin -mattr=+avx,+fast-partial-ymm-write | FileCheck --check-prefix=FASTYMM %s
; RUN: llc < %s -x86-use-vzeroupper -mtriple=x86_64-apple-darwin -mcpu=btver2 | FileCheck --check-prefix=BTVER2 %s
; FASTYMM-NOT: vzeroupper
; BTVER2-NOT: vzeroupper
declare i32 @foo()
declare <4 x float> @do_sse(<4 x float>)