<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 6/27/14, 3:05 PM, Hal Finkel wrote:<br>
</div>
<blockquote
cite="mid:780323.148.1403895933357.JavaMail.javamailuser@localhost"
type="cite">
<pre wrap="">----- Original Message -----
</pre>
<blockquote type="cite">
<pre wrap="">From: "Justin Holewinski" <a class="moz-txt-link-rfc2396E" href="mailto:jholewinski@nvidia.com"><jholewinski@nvidia.com></a>
To: <a class="moz-txt-link-abbreviated" href="mailto:llvm-commits@cs.uiuc.edu">llvm-commits@cs.uiuc.edu</a>
Sent: Friday, June 27, 2014 1:35:45 PM
Subject: [llvm] r211938 - [NVPTX] Clean up argument lowering code and properly handle alignment for structs and
vectors
Author: jholewinski
Date: Fri Jun 27 13:35:44 2014
New Revision: 211938
URL: <a class="moz-txt-link-freetext" href="http://llvm.org/viewvc/llvm-project?rev=211938&view=rev">http://llvm.org/viewvc/llvm-project?rev=211938&view=rev</a>
Log:
[NVPTX] Clean up argument lowering code and properly handle alignment
for structs and vectors
Added:
llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll
Modified:
llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp
Modified: llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp
URL:
<a class="moz-txt-link-freetext" href="http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp?rev=211938&r1=211937&r2=211938&view=diff">http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp?rev=211938&r1=211937&r2=211938&view=diff</a>
==============================================================================
--- llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/NVPTX/NVPTXISelLowering.cpp Fri Jun 27
13:35:44 2014
@@ -67,6 +67,17 @@ static bool IsPTXVectorType(MVT VT) {
}
}
+static uint64_t GCD( int a, int b)
+{
+ if (a < b) std::swap(a,b);
+ while (b > 0) {
+ uint64_t c = b;
+ b = a % b;
+ a = c;
+ }
+ return a;
+}
</pre>
</blockquote>
<pre wrap="">
There is a GreatestCommonDivisor64 in include/llvm/Support/MathExtras.h, can you use that instead?</pre>
</blockquote>
<br>
Very likely; thanks for the pointer!<br>
<br>
<blockquote
cite="mid:780323.148.1403895933357.JavaMail.javamailuser@localhost"
type="cite">
<pre wrap="">
-Hal
</pre>
<blockquote type="cite">
<pre wrap="">+
/// ComputePTXValueVTs - For the given Type \p Ty, returns the set
of primitive
/// EVTs that compose it. Unlike ComputeValueVTs, this will break
apart vectors
/// into their primitive components.
@@ -518,26 +529,12 @@ NVPTXTargetLowering::getPrototype(Type *
} else if (isa<PointerType>(retTy)) {
O << ".param .b" << getPointerTy().getSizeInBits() << " _";
} else {
- if ((retTy->getTypeID() == Type::StructTyID) ||
isa<VectorType>(retTy)) {
- SmallVector<EVT, 16> vtparts;
- ComputeValueVTs(*this, retTy, vtparts);
- unsigned totalsz = 0;
- for (unsigned i = 0, e = vtparts.size(); i != e; ++i) {
- unsigned elems = 1;
- EVT elemtype = vtparts[i];
- if (vtparts[i].isVector()) {
- elems = vtparts[i].getVectorNumElements();
- elemtype = vtparts[i].getVectorElementType();
- }
- // TODO: no need to loop
- for (unsigned j = 0, je = elems; j != je; ++j) {
- unsigned sz = elemtype.getSizeInBits();
- if (elemtype.isInteger() && (sz < 8))
- sz = 8;
- totalsz += sz / 8;
- }
- }
- O << ".param .align " << retAlignment << " .b8 _[" <<
totalsz << "]";
+ if((retTy->getTypeID() == Type::StructTyID) ||
+ isa<VectorType>(retTy)) {
+ O << ".param .align "
+ << retAlignment
+ << " .b8 _["
+ << getDataLayout()->getTypeAllocSize(retTy) << "]";
} else {
assert(false && "Unknown return type");
}
@@ -706,7 +703,8 @@ SDValue NVPTXTargetLowering::LowerCall(T
if (Ty->isAggregateType()) {
// aggregate
SmallVector<EVT, 16> vtparts;
- ComputeValueVTs(*this, Ty, vtparts);
+ SmallVector<uint64_t, 16> Offsets;
+ ComputePTXValueVTs(*this, Ty, vtparts, &Offsets, 0);
unsigned align = getArgumentAlignment(Callee, CS, Ty,
paramCount + 1);
// declare .param .align <align> .b8 .param<n>[<size>];
@@ -718,34 +716,26 @@ SDValue NVPTXTargetLowering::LowerCall(T
Chain = DAG.getNode(NVPTXISD::DeclareParam, dl,
DeclareParamVTs,
DeclareParamOps);
InFlag = Chain.getValue(1);
- unsigned curOffset = 0;
for (unsigned j = 0, je = vtparts.size(); j != je; ++j) {
- unsigned elems = 1;
EVT elemtype = vtparts[j];
- if (vtparts[j].isVector()) {
- elems = vtparts[j].getVectorNumElements();
- elemtype = vtparts[j].getVectorElementType();
- }
- for (unsigned k = 0, ke = elems; k != ke; ++k) {
- unsigned sz = elemtype.getSizeInBits();
- if (elemtype.isInteger() && (sz < 8))
- sz = 8;
- SDValue StVal = OutVals[OIdx];
- if (elemtype.getSizeInBits() < 16) {
- StVal = DAG.getNode(ISD::ANY_EXTEND, dl, MVT::i16,
StVal);
- }
- SDVTList CopyParamVTs = DAG.getVTList(MVT::Other,
MVT::Glue);
- SDValue CopyParamOps[] = { Chain,
- DAG.getConstant(paramCount,
MVT::i32),
- DAG.getConstant(curOffset,
MVT::i32),
- StVal, InFlag };
- Chain = DAG.getMemIntrinsicNode(NVPTXISD::StoreParam,
dl,
- CopyParamVTs,
CopyParamOps,
- elemtype,
MachinePointerInfo());
- InFlag = Chain.getValue(1);
- curOffset += sz / 8;
- ++OIdx;
+ unsigned ArgAlign = GCD(align, Offsets[j]);
+ if (elemtype.isInteger() && (sz < 8))
+ sz = 8;
+ SDValue StVal = OutVals[OIdx];
+ if (elemtype.getSizeInBits() < 16) {
+ StVal = DAG.getNode(ISD::ANY_EXTEND, dl, MVT::i16,
StVal);
}
+ SDVTList CopyParamVTs = DAG.getVTList(MVT::Other,
MVT::Glue);
+ SDValue CopyParamOps[] = { Chain,
+ DAG.getConstant(paramCount,
MVT::i32),
+ DAG.getConstant(Offsets[j],
MVT::i32),
+ StVal, InFlag };
+ Chain = DAG.getMemIntrinsicNode(NVPTXISD::StoreParam, dl,
+ CopyParamVTs,
CopyParamOps,
+ elemtype,
MachinePointerInfo(),
+ ArgAlign);
+ InFlag = Chain.getValue(1);
+ ++OIdx;
}
if (vtparts.size() > 0)
--OIdx;
@@ -930,13 +920,15 @@ SDValue NVPTXTargetLowering::LowerCall(T
}
// struct or vector
SmallVector<EVT, 16> vtparts;
+ SmallVector<uint64_t, 16> Offsets;
const PointerType *PTy = dyn_cast<PointerType>(Args[i].Ty);
assert(PTy && "Type of a byval parameter should be pointer");
- ComputeValueVTs(*this, PTy->getElementType(), vtparts);
+ ComputePTXValueVTs(*this, PTy->getElementType(), vtparts,
&Offsets, 0);
// declare .param .align <align> .b8 .param<n>[<size>];
unsigned sz = Outs[OIdx].Flags.getByValSize();
SDVTList DeclareParamVTs = DAG.getVTList(MVT::Other, MVT::Glue);
+ unsigned ArgAlign = Outs[OIdx].Flags.getByValAlign();
// The ByValAlign in the Outs[OIdx].Flags is alway set at this
point,
// so we don't need to worry about natural alignment or not.
// See TargetLowering::LowerCallTo().
@@ -948,38 +940,28 @@ SDValue NVPTXTargetLowering::LowerCall(T
Chain = DAG.getNode(NVPTXISD::DeclareParam, dl, DeclareParamVTs,
DeclareParamOps);
InFlag = Chain.getValue(1);
- unsigned curOffset = 0;
for (unsigned j = 0, je = vtparts.size(); j != je; ++j) {
- unsigned elems = 1;
EVT elemtype = vtparts[j];
- if (vtparts[j].isVector()) {
- elems = vtparts[j].getVectorNumElements();
- elemtype = vtparts[j].getVectorElementType();
+ int curOffset = Offsets[j];
+ unsigned PartAlign = GCD(ArgAlign, curOffset);
+ SDValue srcAddr =
+ DAG.getNode(ISD::ADD, dl, getPointerTy(), OutVals[OIdx],
+ DAG.getConstant(curOffset, getPointerTy()));
+ SDValue theVal = DAG.getLoad(elemtype, dl, tempChain, srcAddr,
+ MachinePointerInfo(), false,
false, false,
+ PartAlign);
+ if (elemtype.getSizeInBits() < 16) {
+ theVal = DAG.getNode(ISD::ANY_EXTEND, dl, MVT::i16, theVal);
}
- for (unsigned k = 0, ke = elems; k != ke; ++k) {
- unsigned sz = elemtype.getSizeInBits();
- if (elemtype.isInteger() && (sz < 8))
- sz = 8;
- SDValue srcAddr =
- DAG.getNode(ISD::ADD, dl, getPointerTy(), OutVals[OIdx],
- DAG.getConstant(curOffset, getPointerTy()));
- SDValue theVal = DAG.getLoad(elemtype, dl, tempChain,
srcAddr,
- MachinePointerInfo(), false,
false, false,
- 0);
- if (elemtype.getSizeInBits() < 16) {
- theVal = DAG.getNode(ISD::ANY_EXTEND, dl, MVT::i16,
theVal);
- }
- SDVTList CopyParamVTs = DAG.getVTList(MVT::Other,
MVT::Glue);
- SDValue CopyParamOps[] = { Chain,
DAG.getConstant(paramCount, MVT::i32),
- DAG.getConstant(curOffset,
MVT::i32), theVal,
- InFlag };
- Chain = DAG.getMemIntrinsicNode(NVPTXISD::StoreParam, dl,
CopyParamVTs,
- CopyParamOps, elemtype,
- MachinePointerInfo());
+ SDVTList CopyParamVTs = DAG.getVTList(MVT::Other, MVT::Glue);
+ SDValue CopyParamOps[] = { Chain, DAG.getConstant(paramCount,
MVT::i32),
+ DAG.getConstant(curOffset,
MVT::i32), theVal,
+ InFlag };
+ Chain = DAG.getMemIntrinsicNode(NVPTXISD::StoreParam, dl,
CopyParamVTs,
+ CopyParamOps, elemtype,
+ MachinePointerInfo());
- InFlag = Chain.getValue(1);
- curOffset += sz / 8;
- }
+ InFlag = Chain.getValue(1);
}
++paramCount;
}
@@ -1088,7 +1070,6 @@ SDValue NVPTXTargetLowering::LowerCall(T
// Generate loads from param memory/moves from registers for
result
if (Ins.size() > 0) {
- unsigned resoffset = 0;
if (retTy && retTy->isVectorTy()) {
EVT ObjectVT = getValueType(retTy);
unsigned NumElts = ObjectVT.getVectorNumElements();
@@ -1097,14 +1078,15 @@ SDValue NVPTXTargetLowering::LowerCall(T
ObjectVT) ==
NumElts &&
"Vector was not scalarized");
unsigned sz = EltVT.getSizeInBits();
- bool needTruncate = sz < 16 ? true : false;
+ bool needTruncate = sz < 8 ? true : false;
if (NumElts == 1) {
// Just a simple load
SmallVector<EVT, 4> LoadRetVTs;
- if (needTruncate) {
- // If loading i1 result, generate
- // load i16
+ if (EltVT == MVT::i1 || EltVT == MVT::i8) {
+ // If loading i1/i8 result, generate
+ // load.b8 i16
+ // if i1
// trunc i16 to i1
LoadRetVTs.push_back(MVT::i16);
} else
@@ -1128,9 +1110,10 @@ SDValue NVPTXTargetLowering::LowerCall(T
} else if (NumElts == 2) {
// LoadV2
SmallVector<EVT, 4> LoadRetVTs;
- if (needTruncate) {
- // If loading i1 result, generate
- // load i16
+ if (EltVT == MVT::i1 || EltVT == MVT::i8) {
+ // If loading i1/i8 result, generate
+ // load.b8 i16
+ // if i1
// trunc i16 to i1
LoadRetVTs.push_back(MVT::i16);
LoadRetVTs.push_back(MVT::i16);
@@ -1173,9 +1156,10 @@ SDValue NVPTXTargetLowering::LowerCall(T
EVT VecVT = EVT::getVectorVT(F->getContext(), EltVT,
VecSize);
for (unsigned i = 0; i < NumElts; i += VecSize) {
SmallVector<EVT, 8> LoadRetVTs;
- if (needTruncate) {
- // If loading i1 result, generate
- // load i16
+ if (EltVT == MVT::i1 || EltVT == MVT::i8) {
+ // If loading i1/i8 result, generate
+ // load.b8 i16
+ // if i1
// trunc i16 to i1
for (unsigned j = 0; j < VecSize; ++j)
LoadRetVTs.push_back(MVT::i16);
@@ -1214,10 +1198,13 @@ SDValue NVPTXTargetLowering::LowerCall(T
}
} else {
SmallVector<EVT, 16> VTs;
- ComputePTXValueVTs(*this, retTy, VTs);
+ SmallVector<uint64_t, 16> Offsets;
+ ComputePTXValueVTs(*this, retTy, VTs, &Offsets, 0);
assert(VTs.size() == Ins.size() && "Bad value decomposition");
+ unsigned RetAlign = getArgumentAlignment(Callee, CS, retTy,
0);
for (unsigned i = 0, e = Ins.size(); i != e; ++i) {
unsigned sz = VTs[i].getSizeInBits();
+ unsigned AlignI = GCD(RetAlign, Offsets[i]);
bool needTruncate = sz < 8 ? true : false;
if (VTs[i].isInteger() && (sz < 8))
sz = 8;
@@ -1243,19 +1230,18 @@ SDValue NVPTXTargetLowering::LowerCall(T
SmallVector<SDValue, 4> LoadRetOps;
LoadRetOps.push_back(Chain);
LoadRetOps.push_back(DAG.getConstant(1, MVT::i32));
- LoadRetOps.push_back(DAG.getConstant(resoffset, MVT::i32));
+ LoadRetOps.push_back(DAG.getConstant(Offsets[i], MVT::i32));
LoadRetOps.push_back(InFlag);
SDValue retval = DAG.getMemIntrinsicNode(
NVPTXISD::LoadParam, dl,
DAG.getVTList(LoadRetVTs), LoadRetOps,
- TheLoadType, MachinePointerInfo());
+ TheLoadType, MachinePointerInfo(), AlignI);
Chain = retval.getValue(1);
InFlag = retval.getValue(2);
SDValue Ret0 = retval.getValue(0);
if (needTruncate)
Ret0 = DAG.getNode(ISD::TRUNCATE, dl, Ins[i].VT, Ret0);
InVals.push_back(Ret0);
- resoffset += sz / 8;
}
}
}
Added: llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll
URL:
<a class="moz-txt-link-freetext" href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll?rev=211938&view=auto">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll?rev=211938&view=auto</a>
==============================================================================
--- llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll (added)
+++ llvm/trunk/test/CodeGen/NVPTX/arg-lowering.ll Fri Jun 27 13:35:44
2014
@@ -0,0 +1,13 @@
+; RUN: llc < %s -march=nvptx -mcpu=sm_20 | FileCheck %s
+
+; CHECK: .visible .func (.param .align 16 .b8 func_retval0[16])
foo0(
+; CHECK: .param .align 4 .b8 foo0_param_0[8]
+define <4 x float> @foo0({float, float} %arg0) {
+ ret <4 x float> <float 1.0, float 1.0, float 1.0, float 1.0>
+}
+
+; CHECK: .visible .func (.param .align 8 .b8 func_retval0[8]) foo1(
+; CHECK: .param .align 8 .b8 foo1_param_0[16]
+define <2 x float> @foo1({float, float, i64} %arg0) {
+ ret <2 x float> <float 1.0, float 1.0>
+}
_______________________________________________
llvm-commits mailing list
<a class="moz-txt-link-abbreviated" href="mailto:llvm-commits@cs.uiuc.edu">llvm-commits@cs.uiuc.edu</a>
<a class="moz-txt-link-freetext" href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a>
</pre>
</blockquote>
<pre wrap="">
--
Hal Finkel
Assistant Computational Scientist
Leadership Computing Facility
Argonne National Laboratory
</pre>
</blockquote>
<br>
<br>
<pre class="moz-signature" cols="72">--
Thanks,
Justin Holewinski</pre>
<DIV>
<HR>
</DIV>
<DIV>This email message is for the sole use of the intended recipient(s) and may
contain confidential information. Any unauthorized review, use, disclosure
or distribution is prohibited. If you are not the intended recipient,
please contact the sender by reply email and destroy all copies of the original
message. </DIV>
<DIV>
<HR>
</DIV>
</body>
</html>