llvm-6502/lib/CodeGen/SelectionDAG/LegalizeTypes.cpp

679 lines
24 KiB
C++
Raw Normal View History

//===-- LegalizeTypes.cpp - Common code for DAG type legalizer ------------===//
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
//
//===----------------------------------------------------------------------===//
//
// This file implements the SelectionDAG::LegalizeTypes method. It transforms
// an arbitrary well-formed SelectionDAG to only consist of legal types. This
// is common code shared among the LegalizeTypes*.cpp files.
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
//
//===----------------------------------------------------------------------===//
#include "LegalizeTypes.h"
#include "llvm/CallingConv.h"
#include "llvm/Constants.h"
#include "llvm/DerivedTypes.h"
#include "llvm/Support/CommandLine.h"
#include "llvm/Support/MathExtras.h"
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
using namespace llvm;
#ifndef NDEBUG
static cl::opt<bool>
ViewLegalizeTypesDAGs("view-legalize-types-dags", cl::Hidden,
cl::desc("Pop up a window to show dags before legalize types"));
#else
static const bool ViewLegalizeTypesDAGs = 0;
#endif
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
/// run - This is the main entry point for the type legalizer. This does a
/// top-down traversal of the dag, legalizing types as it goes.
void DAGTypeLegalizer::run() {
// Create a dummy node (which is not added to allnodes), that adds a reference
// to the root node, preventing it from being deleted, and tracking any
// changes of the root.
HandleSDNode Dummy(DAG.getRoot());
// The root of the dag may dangle to deleted nodes until the type legalizer is
// done. Set it to null to avoid confusion.
DAG.setRoot(SDOperand());
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Walk all nodes in the graph, assigning them a NodeID of 'ReadyToProcess'
// (and remembering them) if they are leaves and assigning 'NewNode' if
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// non-leaves.
for (SelectionDAG::allnodes_iterator I = DAG.allnodes_begin(),
E = DAG.allnodes_end(); I != E; ++I) {
if (I->getNumOperands() == 0) {
I->setNodeId(ReadyToProcess);
Worklist.push_back(I);
} else {
I->setNodeId(NewNode);
}
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Now that we have a set of nodes to process, handle them all.
while (!Worklist.empty()) {
SDNode *N = Worklist.back();
Worklist.pop_back();
assert(N->getNodeId() == ReadyToProcess &&
"Node should be ready if on worklist!");
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Scan the values produced by the node, checking to see if any result
// types are illegal.
unsigned i = 0;
unsigned NumResults = N->getNumValues();
do {
MVT ResultVT = N->getValueType(i);
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
switch (getTypeAction(ResultVT)) {
default:
assert(false && "Unknown action!");
case Legal:
break;
case PromoteInteger:
PromoteIntegerResult(N, i);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
goto NodeDone;
case ExpandInteger:
ExpandIntegerResult(N, i);
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
goto NodeDone;
case SoftenFloat:
SoftenFloatResult(N, i);
goto NodeDone;
case ExpandFloat:
ExpandFloatResult(N, i);
goto NodeDone;
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
case Scalarize:
ScalarizeResult(N, i);
goto NodeDone;
case Split:
SplitResult(N, i);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
goto NodeDone;
}
} while (++i < NumResults);
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Scan the operand list for the node, handling any nodes with operands that
// are illegal.
{
unsigned NumOperands = N->getNumOperands();
bool NeedsRevisit = false;
for (i = 0; i != NumOperands; ++i) {
MVT OpVT = N->getOperand(i).getValueType();
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
switch (getTypeAction(OpVT)) {
default:
assert(false && "Unknown action!");
case Legal:
continue;
case PromoteInteger:
NeedsRevisit = PromoteIntegerOperand(N, i);
break;
case ExpandInteger:
NeedsRevisit = ExpandIntegerOperand(N, i);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
break;
case SoftenFloat:
NeedsRevisit = SoftenFloatOperand(N, i);
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
break;
case ExpandFloat:
NeedsRevisit = ExpandFloatOperand(N, i);
break;
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
case Scalarize:
NeedsRevisit = ScalarizeOperand(N, i);
break;
case Split:
NeedsRevisit = SplitOperand(N, i);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
break;
}
LegalizeTypes support for EXTRACT_VECTOR_ELT. The approach taken is different to that in LegalizeDAG when it is a question of expanding or promoting the result type: for example, if extracting an i64 from a <2 x i64>, when i64 needs expanding, it bitcasts the vector to <4 x i32>, extracts the appropriate two i32's, and uses those for the Lo and Hi parts. Likewise, when extracting an i16 from a <4 x i16>, and i16 needs promoting, it bitcasts the vector to <2 x i32>, extracts the appropriate i32, twiddles the bits if necessary, and uses that as the promoted value. This puts more pressure on bitcast legalization, and I've added the appropriate cases. They needed to be added anyway since users can generate such bitcasts too if they want to. Also, when considering various cases (Legal, Promote, Expand, Scalarize, Split) it is a pain that expand can correspond to Expand, Scalarize or Split, so I've changed the LegalizeTypes enum so it lists those different cases - now Expand only means splitting a scalar in two. The code produced is the same as by LegalizeDAG for all relevant testcases, except for 2007-10-31-extractelement-i64.ll, where the code seems to have improved (see below; can an expert please tell me if it is better or not). Before < vs after >. < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 28(%esp) < movl (%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 28(%esp) < movl 8(%esp), %eax < movl %eax, 24(%esp) < movq 24(%esp), %mm0 < movq %mm0, 56(%esp) --- > subl $44, %esp > movaps %xmm0, 16(%esp) > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movq (%esp), %mm0 > movq %mm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 16(%esp), %eax < movl %eax, 48(%esp) < movl 20(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 4(%esp), %eax < movl %eax, 60(%esp) < movl (%esp), %eax < movl %eax, 56(%esp) --- > pshufd $1, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) < subl $92, %esp < movaps %xmm0, 64(%esp) --- > subl $44, %esp < movl 24(%esp), %eax < movl %eax, 48(%esp) < movl 28(%esp), %eax < movl %eax, 52(%esp) < movaps %xmm0, (%esp) < movl 12(%esp), %eax < movl %eax, 60(%esp) < movl 8(%esp), %eax < movl %eax, 56(%esp) --- > pshufd $3, %xmm0, %xmm1 > movd %xmm1, 4(%esp) > movhlps %xmm0, %xmm0 > movd %xmm0, (%esp) > movd %xmm1, 12(%esp) > movd %xmm0, 8(%esp) git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47672 91177308-0d34-0410-b5e6-96231b3b80d8
2008-02-27 13:34:40 +00:00
break;
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
}
// If the node needs revisiting, don't add all users to the worklist etc.
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
if (NeedsRevisit)
continue;
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
if (i == NumOperands)
DEBUG(cerr << "Legally typed node: "; N->dump(&DAG); cerr << "\n");
}
NodeDone:
// If we reach here, the node was processed, potentially creating new nodes.
// Mark it as processed and add its users to the worklist as appropriate.
N->setNodeId(Processed);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
for (SDNode::use_iterator UI = N->use_begin(), E = N->use_end();
UI != E; ++UI) {
SDNode *User = UI->getUser();
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
int NodeID = User->getNodeId();
assert(NodeID != ReadyToProcess && NodeID != Processed &&
"Invalid node id for user of unprocessed node!");
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// This node has two options: it can either be a new node or its Node ID
// may be a count of the number of operands it has that are not ready.
if (NodeID > 0) {
User->setNodeId(NodeID-1);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// If this was the last use it was waiting on, add it to the ready list.
if (NodeID-1 == ReadyToProcess)
Worklist.push_back(User);
continue;
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Otherwise, this node is new: this is the first operand of it that
// became ready. Its new NodeID is the number of operands it has minus 1
// (as this node is now processed).
assert(NodeID == NewNode && "Unknown node ID!");
User->setNodeId(User->getNumOperands()-1);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// If the node only has a single operand, it is now ready.
if (User->getNumOperands() == 1)
Worklist.push_back(User);
}
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// If the root changed (e.g. it was a dead load, update the root).
DAG.setRoot(Dummy.getValue());
//DAG.viewGraph();
// Remove dead nodes. This is important to do for cleanliness but also before
// the checking loop below. Implicit folding by the DAG.getNode operators can
// cause unreachable nodes to be around with their flags set to new.
DAG.RemoveDeadNodes();
// In a debug build, scan all the nodes to make sure we found them all. This
// ensures that there are no cycles and that everything got processed.
#ifndef NDEBUG
for (SelectionDAG::allnodes_iterator I = DAG.allnodes_begin(),
E = DAG.allnodes_end(); I != E; ++I) {
bool Failed = false;
// Check that all result types are legal.
for (unsigned i = 0, NumVals = I->getNumValues(); i < NumVals; ++i)
if (!isTypeLegal(I->getValueType(i))) {
cerr << "Result type " << i << " illegal!\n";
Failed = true;
}
// Check that all operand types are legal.
for (unsigned i = 0, NumOps = I->getNumOperands(); i < NumOps; ++i)
if (!isTypeLegal(I->getOperand(i).getValueType())) {
cerr << "Operand type " << i << " illegal!\n";
Failed = true;
}
if (I->getNodeId() != Processed) {
if (I->getNodeId() == NewNode)
cerr << "New node not 'noticed'?\n";
else if (I->getNodeId() > 0)
cerr << "Operand not processed?\n";
else if (I->getNodeId() == ReadyToProcess)
cerr << "Not added to worklist?\n";
Failed = true;
}
if (Failed) {
I->dump(&DAG); cerr << "\n";
abort();
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
}
#endif
}
/// AnalyzeNewNode - The specified node is the root of a subtree of potentially
/// new nodes. Correct any processed operands (this may change the node) and
/// calculate the NodeId.
void DAGTypeLegalizer::AnalyzeNewNode(SDNode *&N) {
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// If this was an existing node that is already done, we're done.
if (N->getNodeId() != NewNode)
return;
// Okay, we know that this node is new. Recursively walk all of its operands
// to see if they are new also. The depth of this walk is bounded by the size
// of the new tree that was constructed (usually 2-3 nodes), so we don't worry
// about revisiting of nodes.
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
//
// As we walk the operands, keep track of the number of nodes that are
// processed. If non-zero, this will become the new nodeid of this node.
// Already processed operands may need to be remapped to the node that
// replaced them, which can result in our node changing. Since remapping
// is rare, the code tries to minimize overhead in the non-remapping case.
SmallVector<SDOperand, 8> NewOps;
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
unsigned NumProcessed = 0;
for (unsigned i = 0, e = N->getNumOperands(); i != e; ++i) {
SDOperand OrigOp = N->getOperand(i);
SDOperand Op = OrigOp;
if (Op.Val->getNodeId() == Processed)
RemapNode(Op);
if (Op.Val->getNodeId() == NewNode)
AnalyzeNewNode(Op.Val);
else if (Op.Val->getNodeId() == Processed)
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
++NumProcessed;
if (!NewOps.empty()) {
// Some previous operand changed. Add this one to the list.
NewOps.push_back(Op);
} else if (Op != OrigOp) {
// This is the first operand to change - add all operands so far.
for (unsigned j = 0; j < i; ++j)
NewOps.push_back(N->getOperand(j));
NewOps.push_back(Op);
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
}
// Some operands changed - update the node.
if (!NewOps.empty())
N = DAG.UpdateNodeOperands(SDOperand(N, 0), &NewOps[0], NewOps.size()).Val;
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
N->setNodeId(N->getNumOperands()-NumProcessed);
if (N->getNodeId() == ReadyToProcess)
Worklist.push_back(N);
}
namespace {
/// NodeUpdateListener - This class is a DAGUpdateListener that listens for
/// updates to nodes and recomputes their ready state.
class VISIBILITY_HIDDEN NodeUpdateListener :
public SelectionDAG::DAGUpdateListener {
DAGTypeLegalizer &DTL;
public:
NodeUpdateListener(DAGTypeLegalizer &dtl) : DTL(dtl) {}
virtual void NodeDeleted(SDNode *N, SDNode *E) {
assert(N->getNodeId() != DAGTypeLegalizer::Processed &&
N->getNodeId() != DAGTypeLegalizer::ReadyToProcess &&
"RAUW deleted processed node!");
// It is possible, though rare, for the deleted node N to occur as a
// target in a map, so note the replacement N -> E in ReplacedNodes.
assert(E && "Node not replaced?");
for (unsigned i = 0, e = E->getNumValues(); i != e; ++i)
DTL.NoteReplacement(SDOperand(N, i), SDOperand(E, i));
}
virtual void NodeUpdated(SDNode *N) {
// Node updates can mean pretty much anything. It is possible that an
// operand was set to something already processed (f.e.) in which case
// this node could become ready. Recompute its flags.
assert(N->getNodeId() != DAGTypeLegalizer::Processed &&
N->getNodeId() != DAGTypeLegalizer::ReadyToProcess &&
"RAUW updated processed node!");
DTL.ReanalyzeNode(N);
}
};
}
/// ReplaceValueWith - The specified value was legalized to the specified other
/// value. If they are different, update the DAG and NodeIDs replacing any uses
/// of From to use To instead.
void DAGTypeLegalizer::ReplaceValueWith(SDOperand From, SDOperand To) {
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
if (From == To) return;
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// If expansion produced new nodes, make sure they are properly marked.
AnalyzeNewNode(To.Val);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Anything that used the old node should now use the new one. Note that this
// can potentially cause recursive merging.
NodeUpdateListener NUL(*this);
DAG.ReplaceAllUsesOfValueWith(From, To, &NUL);
// The old node may still be present in a map like ExpandedIntegers or
// PromotedIntegers. Inform maps about the replacement.
NoteReplacement(From, To);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
}
/// ReplaceNodeWith - Replace uses of the 'from' node's results with the 'to'
/// node's results. The from and to node must define identical result types.
void DAGTypeLegalizer::ReplaceNodeWith(SDNode *From, SDNode *To) {
if (From == To) return;
// If expansion produced new nodes, make sure they are properly marked.
AnalyzeNewNode(To);
assert(From->getNumValues() == To->getNumValues() &&
"Node results don't match");
// Anything that used the old node should now use the new one. Note that this
// can potentially cause recursive merging.
NodeUpdateListener NUL(*this);
DAG.ReplaceAllUsesWith(From, To, &NUL);
// The old node may still be present in a map like ExpandedIntegers or
// PromotedIntegers. Inform maps about the replacement.
for (unsigned i = 0, e = From->getNumValues(); i != e; ++i) {
assert(From->getValueType(i) == To->getValueType(i) &&
"Node results don't match");
NoteReplacement(SDOperand(From, i), SDOperand(To, i));
}
}
/// RemapNode - If the specified value was already legalized to another value,
/// replace it by that value.
void DAGTypeLegalizer::RemapNode(SDOperand &N) {
DenseMap<SDOperand, SDOperand>::iterator I = ReplacedNodes.find(N);
if (I != ReplacedNodes.end()) {
// Use path compression to speed up future lookups if values get multiply
// replaced with other values.
RemapNode(I->second);
N = I->second;
}
}
/// ExpungeNode - If this is a deleted value that was kept around to speed up
/// remapping, remove it globally now. The only map that can have a deleted
/// node as a source is ReplacedNodes. Other maps can have deleted nodes as
/// targets, but since their looked-up values are always immediately remapped
/// using RemapNode, resulting in a not-deleted node, this is harmless as long
/// as ReplacedNodes/RemapNode always performs correct mappings. The mapping
/// will always be correct as long as ExpungeNode is called on the source when
/// adding a new node to ReplacedNodes, and called on the target when adding
/// a new node to any map.
void DAGTypeLegalizer::ExpungeNode(SDOperand N) {
SDOperand Replacement = N;
RemapNode(Replacement);
if (Replacement != N) {
// Remove N from all maps - this is expensive but extremely rare.
ReplacedNodes.erase(N);
for (DenseMap<SDOperand, SDOperand>::iterator I = ReplacedNodes.begin(),
E = ReplacedNodes.end(); I != E; ++I) {
if (I->second == N)
I->second = Replacement;
}
for (DenseMap<SDOperand, SDOperand>::iterator I = PromotedIntegers.begin(),
E = PromotedIntegers.end(); I != E; ++I) {
assert(I->first != N);
if (I->second == N)
I->second = Replacement;
}
for (DenseMap<SDOperand, SDOperand>::iterator I = SoftenedFloats.begin(),
E = SoftenedFloats.end(); I != E; ++I) {
assert(I->first != N);
if (I->second == N)
I->second = Replacement;
}
for (DenseMap<SDOperand, SDOperand>::iterator I = ScalarizedVectors.begin(),
E = ScalarizedVectors.end(); I != E; ++I) {
assert(I->first != N);
if (I->second == N)
I->second = Replacement;
}
for (DenseMap<SDOperand, std::pair<SDOperand, SDOperand> >::iterator
I = ExpandedIntegers.begin(), E = ExpandedIntegers.end(); I != E; ++I){
assert(I->first != N);
if (I->second.first == N)
I->second.first = Replacement;
if (I->second.second == N)
I->second.second = Replacement;
}
for (DenseMap<SDOperand, std::pair<SDOperand, SDOperand> >::iterator
I = ExpandedFloats.begin(), E = ExpandedFloats.end(); I != E; ++I) {
assert(I->first != N);
if (I->second.first == N)
I->second.first = Replacement;
if (I->second.second == N)
I->second.second = Replacement;
}
for (DenseMap<SDOperand, std::pair<SDOperand, SDOperand> >::iterator
I = SplitVectors.begin(), E = SplitVectors.end(); I != E; ++I) {
assert(I->first != N);
if (I->second.first == N)
I->second.first = Replacement;
if (I->second.second == N)
I->second.second = Replacement;
}
}
}
void DAGTypeLegalizer::SetPromotedInteger(SDOperand Op, SDOperand Result) {
ExpungeNode(Result);
AnalyzeNewNode(Result.Val);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
SDOperand &OpEntry = PromotedIntegers[Op];
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
assert(OpEntry.Val == 0 && "Node is already promoted!");
OpEntry = Result;
}
void DAGTypeLegalizer::SetSoftenedFloat(SDOperand Op, SDOperand Result) {
ExpungeNode(Result);
AnalyzeNewNode(Result.Val);
SDOperand &OpEntry = SoftenedFloats[Op];
assert(OpEntry.Val == 0 && "Node is already converted to integer!");
OpEntry = Result;
}
void DAGTypeLegalizer::SetScalarizedVector(SDOperand Op, SDOperand Result) {
ExpungeNode(Result);
AnalyzeNewNode(Result.Val);
SDOperand &OpEntry = ScalarizedVectors[Op];
assert(OpEntry.Val == 0 && "Node is already scalarized!");
OpEntry = Result;
}
void DAGTypeLegalizer::GetExpandedInteger(SDOperand Op, SDOperand &Lo,
SDOperand &Hi) {
std::pair<SDOperand, SDOperand> &Entry = ExpandedIntegers[Op];
RemapNode(Entry.first);
RemapNode(Entry.second);
assert(Entry.first.Val && "Operand isn't expanded");
Lo = Entry.first;
Hi = Entry.second;
}
void DAGTypeLegalizer::SetExpandedInteger(SDOperand Op, SDOperand Lo,
SDOperand Hi) {
ExpungeNode(Lo);
ExpungeNode(Hi);
// Lo/Hi may have been newly allocated, if so, add nodeid's as relevant.
AnalyzeNewNode(Lo.Val);
AnalyzeNewNode(Hi.Val);
// Remember that this is the result of the node.
std::pair<SDOperand, SDOperand> &Entry = ExpandedIntegers[Op];
assert(Entry.first.Val == 0 && "Node already expanded");
Entry.first = Lo;
Entry.second = Hi;
}
void DAGTypeLegalizer::GetExpandedFloat(SDOperand Op, SDOperand &Lo,
SDOperand &Hi) {
std::pair<SDOperand, SDOperand> &Entry = ExpandedFloats[Op];
RemapNode(Entry.first);
RemapNode(Entry.second);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
assert(Entry.first.Val && "Operand isn't expanded");
Lo = Entry.first;
Hi = Entry.second;
}
void DAGTypeLegalizer::SetExpandedFloat(SDOperand Op, SDOperand Lo,
SDOperand Hi) {
ExpungeNode(Lo);
ExpungeNode(Hi);
// Lo/Hi may have been newly allocated, if so, add nodeid's as relevant.
AnalyzeNewNode(Lo.Val);
AnalyzeNewNode(Hi.Val);
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
// Remember that this is the result of the node.
std::pair<SDOperand, SDOperand> &Entry = ExpandedFloats[Op];
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
assert(Entry.first.Val == 0 && "Node already expanded");
Entry.first = Lo;
Entry.second = Hi;
}
void DAGTypeLegalizer::GetSplitVector(SDOperand Op, SDOperand &Lo,
SDOperand &Hi) {
std::pair<SDOperand, SDOperand> &Entry = SplitVectors[Op];
RemapNode(Entry.first);
RemapNode(Entry.second);
assert(Entry.first.Val && "Operand isn't split");
Lo = Entry.first;
Hi = Entry.second;
}
void DAGTypeLegalizer::SetSplitVector(SDOperand Op, SDOperand Lo,
SDOperand Hi) {
ExpungeNode(Lo);
ExpungeNode(Hi);
// Lo/Hi may have been newly allocated, if so, add nodeid's as relevant.
AnalyzeNewNode(Lo.Val);
AnalyzeNewNode(Hi.Val);
// Remember that this is the result of the node.
std::pair<SDOperand, SDOperand> &Entry = SplitVectors[Op];
assert(Entry.first.Val == 0 && "Node already split");
Entry.first = Lo;
Entry.second = Hi;
}
//===----------------------------------------------------------------------===//
// Utilities.
//===----------------------------------------------------------------------===//
/// BitConvertToInteger - Convert to an integer of the same size.
SDOperand DAGTypeLegalizer::BitConvertToInteger(SDOperand Op) {
unsigned BitWidth = Op.getValueType().getSizeInBits();
return DAG.getNode(ISD::BIT_CONVERT, MVT::getIntegerVT(BitWidth), Op);
}
SDOperand DAGTypeLegalizer::CreateStackStoreLoad(SDOperand Op,
MVT DestVT) {
// Create the stack frame object.
SDOperand FIPtr = DAG.CreateStackTemporary(DestVT);
// Emit a store to the stack slot.
SDOperand Store = DAG.getStore(DAG.getEntryNode(), Op, FIPtr, NULL, 0);
// Result is a load from the stack slot.
return DAG.getLoad(DestVT, Store, FIPtr, NULL, 0);
}
/// JoinIntegers - Build an integer with low bits Lo and high bits Hi.
SDOperand DAGTypeLegalizer::JoinIntegers(SDOperand Lo, SDOperand Hi) {
MVT LVT = Lo.getValueType();
MVT HVT = Hi.getValueType();
MVT NVT = MVT::getIntegerVT(LVT.getSizeInBits() + HVT.getSizeInBits());
Lo = DAG.getNode(ISD::ZERO_EXTEND, NVT, Lo);
Hi = DAG.getNode(ISD::ANY_EXTEND, NVT, Hi);
Hi = DAG.getNode(ISD::SHL, NVT, Hi, DAG.getConstant(LVT.getSizeInBits(),
TLI.getShiftAmountTy()));
return DAG.getNode(ISD::OR, NVT, Lo, Hi);
}
/// SplitInteger - Return the lower LoVT bits of Op in Lo and the upper HiVT
/// bits in Hi.
void DAGTypeLegalizer::SplitInteger(SDOperand Op,
MVT LoVT, MVT HiVT,
SDOperand &Lo, SDOperand &Hi) {
assert(LoVT.getSizeInBits() + HiVT.getSizeInBits() ==
Op.getValueType().getSizeInBits() && "Invalid integer splitting!");
Lo = DAG.getNode(ISD::TRUNCATE, LoVT, Op);
Hi = DAG.getNode(ISD::SRL, Op.getValueType(), Op,
DAG.getConstant(LoVT.getSizeInBits(),
TLI.getShiftAmountTy()));
Hi = DAG.getNode(ISD::TRUNCATE, HiVT, Hi);
}
/// SplitInteger - Return the lower and upper halves of Op's bits in a value type
/// half the size of Op's.
void DAGTypeLegalizer::SplitInteger(SDOperand Op,
SDOperand &Lo, SDOperand &Hi) {
MVT HalfVT = MVT::getIntegerVT(Op.getValueType().getSizeInBits()/2);
SplitInteger(Op, HalfVT, HalfVT, Lo, Hi);
}
/// MakeLibCall - Generate a libcall taking the given operands as arguments and
/// returning a result of type RetVT.
SDOperand DAGTypeLegalizer::MakeLibCall(RTLIB::Libcall LC, MVT RetVT,
const SDOperand *Ops, unsigned NumOps,
bool isSigned) {
TargetLowering::ArgListTy Args;
Args.reserve(NumOps);
TargetLowering::ArgListEntry Entry;
for (unsigned i = 0; i != NumOps; ++i) {
Entry.Node = Ops[i];
Entry.Ty = Entry.Node.getValueType().getTypeForMVT();
Entry.isSExt = isSigned;
Entry.isZExt = !isSigned;
Args.push_back(Entry);
}
SDOperand Callee = DAG.getExternalSymbol(TLI.getLibcallName(LC),
TLI.getPointerTy());
const Type *RetTy = RetVT.getTypeForMVT();
std::pair<SDOperand,SDOperand> CallInfo =
TLI.LowerCallTo(DAG.getEntryNode(), RetTy, isSigned, !isSigned, false,
CallingConv::C, false, Callee, Args, DAG);
return CallInfo.first;
}
SDOperand DAGTypeLegalizer::GetVectorElementPointer(SDOperand VecPtr, MVT EltVT,
SDOperand Index) {
// Make sure the index type is big enough to compute in.
if (Index.getValueType().bitsGT(TLI.getPointerTy()))
Index = DAG.getNode(ISD::TRUNCATE, TLI.getPointerTy(), Index);
else
Index = DAG.getNode(ISD::ZERO_EXTEND, TLI.getPointerTy(), Index);
// Calculate the element offset and add it to the pointer.
unsigned EltSize = EltVT.getSizeInBits() / 8; // FIXME: should be ABI size.
Index = DAG.getNode(ISD::MUL, Index.getValueType(), Index,
DAG.getConstant(EltSize, Index.getValueType()));
return DAG.getNode(ISD::ADD, Index.getValueType(), Index, VecPtr);
}
/// GetSplitDestVTs - Compute the VTs needed for the low/hi parts of a type
/// which is split into two not necessarily identical pieces.
void DAGTypeLegalizer::GetSplitDestVTs(MVT InVT, MVT &LoVT, MVT &HiVT) {
if (!InVT.isVector()) {
LoVT = HiVT = TLI.getTypeToTransformTo(InVT);
} else {
MVT NewEltVT = InVT.getVectorElementType();
unsigned NumElements = InVT.getVectorNumElements();
if ((NumElements & (NumElements-1)) == 0) { // Simple power of two vector.
NumElements >>= 1;
LoVT = HiVT = MVT::getVectorVT(NewEltVT, NumElements);
} else { // Non-power-of-two vectors.
unsigned NewNumElts_Lo = 1 << Log2_32(NumElements);
unsigned NewNumElts_Hi = NumElements - NewNumElts_Lo;
LoVT = MVT::getVectorVT(NewEltVT, NewNumElts_Lo);
HiVT = MVT::getVectorVT(NewEltVT, NewNumElts_Hi);
}
}
}
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
//===----------------------------------------------------------------------===//
// Entry Point
//===----------------------------------------------------------------------===//
/// LegalizeTypes - This transforms the SelectionDAG into a SelectionDAG that
/// only uses types natively supported by the target.
///
/// Note that this is an involved process that may invalidate pointers into
/// the graph.
void SelectionDAG::LegalizeTypes() {
if (ViewLegalizeTypesDAGs) viewGraph();
One mundane change: Change ReplaceAllUsesOfValueWith to *optionally* take a deleted nodes vector, instead of requiring it. One more significant change: Implement the start of a legalizer that just works on types. This legalizer is designed to run before the operation legalizer and ensure just that the input dag is transformed into an output dag whose operand and result types are all legal, even if the operations on those types are not. This design/impl has the following advantages: 1. When finished, this will *significantly* reduce the amount of code in LegalizeDAG.cpp. It will remove all the code related to promotion and expansion as well as splitting and scalarizing vectors. 2. The new code is very simple, idiomatic, and modular: unlike LegalizeDAG.cpp, it has no 3000 line long functions. :) 3. The implementation is completely iterative instead of recursive, good for hacking on large dags without blowing out your stack. 4. The implementation updates nodes in place when possible instead of deallocating and reallocating the entire graph that points to some mutated node. 5. The code nicely separates out handling of operations with invalid results from operations with invalid operands, making some cases simpler and easier to understand. 6. The new -debug-only=legalize-types option is very very handy :), allowing you to easily understand what legalize types is doing. This is not yet done. Until the ifdef added to SelectionDAGISel.cpp is enabled, this does nothing. However, this code is sufficient to legalize all of the code in 186.crafty, olden and freebench on an x86 machine. The biggest issues are: 1. Vectors aren't implemented at all yet 2. SoftFP is a mess, I need to talk to Evan about it. 3. No lowering to libcalls is implemented yet. 4. Various operations are missing etc. 5. There are FIXME's for stuff I hax0r'd out, like softfp. Hey, at least it is a step in the right direction :). If you'd like to help, just enable the #ifdef in SelectionDAGISel.cpp and compile code with it. If this explodes it will tell you what needs to be implemented. Help is certainly appreciated. Once this goes in, we can do three things: 1. Add a new pass of dag combine between the "type legalizer" and "operation legalizer" passes. This will let us catch some long-standing isel issues that we miss because operation legalization often obfuscates the dag with target-specific nodes. 2. We can rip out all of the type legalization code from LegalizeDAG.cpp, making it much smaller and simpler. When that happens we can then reimplement the core functionality left in it in a much more efficient and non-recursive way. 3. Once the whole legalizer is non-recursive, we can implement whole-function selectiondags maybe... git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@42981 91177308-0d34-0410-b5e6-96231b3b80d8
2007-10-15 06:10:22 +00:00
DAGTypeLegalizer(*this).run();
}