maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

Make our data-locality schedule tree transforms externally accessible 

Other passes which perform different optimizations might be interested in
also applying data-locality transformations as part of their overall
transformation.

llvm-svn: 245824
This commit is contained in:
Tobias Grosser 2015-08-24 06:01:47 +00:00
parent ae8c17ef58
commit fa57e9b7e6
2 changed files with 172 additions and 141 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

124
polly/include/polly/ScheduleOptimizer.h
View File

@ -12,8 +12,132 @@
#ifndef POLLY_SCHEDULE_OPTIMIZER_H #ifndef POLLY_SCHEDULE_OPTIMIZER_H
#define POLLY_SCHEDULE_OPTIMIZER_H #define POLLY_SCHEDULE_OPTIMIZER_H
#include "isl/ctx.h"
#include "llvm/ADT/ArrayRef.h"
struct isl_schedule;
struct isl_schedule_node;
struct isl_union_map;
namespace polly { namespace polly {
extern bool DisablePollyTiling; extern bool DisablePollyTiling;
class Scop;
} }
class ScheduleTreeOptimizer {
public:
/// @brief Apply schedule tree transformations.
///
/// This function takes an (possibly already optimized) schedule tree and
/// applies a set of additional optimizations on the schedule tree. The
/// transformations applied include:
///
/// - Tiling
/// - Prevectorization
///
/// @param Schedule The schedule object the transformations will be applied
/// to.
/// @returns The transformed schedule.
static __isl_give isl_schedule *
optimizeSchedule(__isl_take isl_schedule *Schedule);
/// @brief Apply schedule tree transformations.
///
/// This function takes a node in an (possibly already optimized) schedule
/// tree and applies a set of additional optimizations on this schedule tree
/// node and its descendents. The transformations applied include:
///
/// - Tiling
/// - Prevectorization
///
/// @param Node The schedule object post-transformations will be applied to.
/// @returns The transformed schedule.
static __isl_give isl_schedule_node *
optimizeScheduleNode(__isl_take isl_schedule_node *Node);
/// @brief Decide if the @p NewSchedule is profitable for @p S.
///
/// @param S The SCoP we optimize.
/// @param NewSchedule The new schedule we computed.
///
/// @return True, if we believe @p NewSchedule is an improvement for @p S.
static bool isProfitableSchedule(polly::Scop &S,
__isl_keep isl_union_map *NewSchedule);
private:
/// @brief Tile a schedule node.
///
/// @param Node The node to tile.
/// @param Identifier An name that identifies this kind of tiling and
/// that is used to mark the tiled loops in the
/// generated AST.
/// @param TileSizes A vector of tile sizes that should be used for
/// tiling.
/// @param DefaultTileSize A default tile size that is used for dimensions
/// that are not covered by the TileSizes vector.
static __isl_give isl_schedule_node *
tileNode(__isl_take isl_schedule_node *Node, const char *Identifier,
llvm::ArrayRef<int> TileSizes, int DefaultTileSize);
/// @brief Check if this node is a band node we want to tile.
///
/// We look for innermost band nodes where individual dimensions are marked as
/// permutable.
///
/// @param Node The node to check.
static bool isTileableBandNode(__isl_keep isl_schedule_node *Node);
/// @brief Pre-vectorizes one scheduling dimension of a schedule band.
///
/// prevectSchedBand splits out the dimension DimToVectorize, tiles it and
/// sinks the resulting point loop.
///
/// Example (DimToVectorize=0, VectorWidth=4):
///
/// | Before transformation:
/// |
/// | A[i,j] -> [i,j]
/// |
/// | for (i = 0; i < 128; i++)
/// | for (j = 0; j < 128; j++)
/// | A(i,j);
///
/// | After transformation:
/// |
/// | for (it = 0; it < 32; it+=1)
/// | for (j = 0; j < 128; j++)
/// | for (ip = 0; ip <= 3; ip++)
/// | A(4 * it + ip,j);
///
/// The goal of this transformation is to create a trivially vectorizable
/// loop. This means a parallel loop at the innermost level that has a
/// constant number of iterations corresponding to the target vector width.
///
/// This transformation creates a loop at the innermost level. The loop has
/// a constant number of iterations, if the number of loop iterations at
/// DimToVectorize can be divided by VectorWidth. The default VectorWidth is
/// currently constant and not yet target specific. This function does not
/// reason about parallelism.
static __isl_give isl_schedule_node *
prevectSchedBand(__isl_take isl_schedule_node *Node, unsigned DimToVectorize,
int VectorWidth);
/// @brief Apply additional optimizations on the bands in the schedule tree.
///
/// We are looking for an innermost band node and apply the following
/// transformations:
///
/// - Tile the band
/// - if the band is tileable
/// - if the band has more than one loop dimension
///
/// - Prevectorize the schedule of the band (or the point loop in case of
/// tiling).
/// - if vectorization is enabled
///
/// @param Node The schedule node to (possibly) optimize.
/// @param User A pointer to forward some use information (currently unused).
static isl_schedule_node *optimizeBand(isl_schedule_node *Node, void *User);
};
#endif #endif

189
polly/lib/Transform/ScheduleOptimizer.cpp
View File

@ -160,135 +160,10 @@ static cl::list<int>
cl::Hidden, cl::ZeroOrMore, cl::CommaSeparated, cl::Hidden, cl::ZeroOrMore, cl::CommaSeparated,
cl::cat(PollyCategory)); cl::cat(PollyCategory));
namespace {
class IslScheduleOptimizer : public ScopPass {
public:
static char ID;
explicit IslScheduleOptimizer() : ScopPass(ID) { LastSchedule = nullptr; }
~IslScheduleOptimizer() { isl_schedule_free(LastSchedule); }
bool runOnScop(Scop &S) override;
void printScop(raw_ostream &OS, Scop &S) const override;
void getAnalysisUsage(AnalysisUsage &AU) const override;
private:
isl_schedule *LastSchedule;
/// @brief Decide if the @p NewSchedule is profitable for @p S.
///
/// @param S The SCoP we optimize.
/// @param NewSchedule The new schedule we computed.
///
/// @return True, if we believe @p NewSchedule is an improvement for @p S.
bool isProfitableSchedule(Scop &S, __isl_keep isl_union_map *NewSchedule);
/// @brief Tile a schedule node.
///
/// @param Node The node to tile.
/// @param Identifier An name that identifies this kind of tiling and
/// that is used to mark the tiled loops in the
/// generated AST.
/// @param TileSizes A vector of tile sizes that should be used for
/// tiling.
/// @param DefaultTileSize A default tile size that is used for dimensions
/// that are not covered by the TileSizes vector.
static __isl_give isl_schedule_node *
tileNode(__isl_take isl_schedule_node *Node, const char *Identifier,
ArrayRef<int> TileSizes, int DefaultTileSize);
/// @brief Check if this node is a band node we want to tile.
///
/// We look for innermost band nodes where individual dimensions are marked as
/// permutable.
///
/// @param Node The node to check.
static bool isTileableBandNode(__isl_keep isl_schedule_node *Node);
/// @brief Pre-vectorizes one scheduling dimension of a schedule band.
///
/// prevectSchedBand splits out the dimension DimToVectorize, tiles it and
/// sinks the resulting point loop.
///
/// Example (DimToVectorize=0, VectorWidth=4):
///
/// | Before transformation:
/// |
/// | A[i,j] -> [i,j]
/// |
/// | for (i = 0; i < 128; i++)
/// | for (j = 0; j < 128; j++)
/// | A(i,j);
///
/// | After transformation:
/// |
/// | for (it = 0; it < 32; it+=1)
/// | for (j = 0; j < 128; j++)
/// | for (ip = 0; ip <= 3; ip++)
/// | A(4 * it + ip,j);
///
/// The goal of this transformation is to create a trivially vectorizable
/// loop. This means a parallel loop at the innermost level that has a
/// constant number of iterations corresponding to the target vector width.
///
/// This transformation creates a loop at the innermost level. The loop has
/// a constant number of iterations, if the number of loop iterations at
/// DimToVectorize can be divided by VectorWidth. The default VectorWidth is
/// currently constant and not yet target specific. This function does not
/// reason about parallelism.
static __isl_give isl_schedule_node *
prevectSchedBand(__isl_take isl_schedule_node *Node, unsigned DimToVectorize,
int VectorWidth);
/// @brief Apply additional optimizations on the bands in the schedule tree.
///
/// We are looking for an innermost band node and apply the following
/// transformations:
///
/// - Tile the band
/// - if the band is tileable
/// - if the band has more than one loop dimension
///
/// - Prevectorize the schedule of the band (or the point loop in case of
/// tiling).
/// - if vectorization is enabled
///
/// @param Node The schedule node to (possibly) optimize.
/// @param User A pointer to forward some use information (currently unused).
static isl_schedule_node *optimizeBand(isl_schedule_node *Node, void *User);
/// @brief Apply post-scheduling transformations.
///
/// This function applies a set of additional local transformations on the
/// schedule tree as it computed by the isl scheduler. Local transformations
/// applied include:
///
/// - Tiling
/// - Prevectorization
///
/// @param Schedule The schedule object post-transformations will be applied
/// on.
/// @returns The transformed schedule.
static __isl_give isl_schedule *
addPostTransforms(__isl_take isl_schedule *Schedule);
using llvm::Pass::doFinalization;
virtual bool doFinalization() override {
isl_schedule_free(LastSchedule);
LastSchedule = nullptr;
return true;
}
};
}
char IslScheduleOptimizer::ID = 0;
__isl_give isl_schedule_node * __isl_give isl_schedule_node *
IslScheduleOptimizer::prevectSchedBand(__isl_take isl_schedule_node *Node, ScheduleTreeOptimizer::prevectSchedBand(__isl_take isl_schedule_node *Node,
unsigned DimToVectorize, unsigned DimToVectorize,
int VectorWidth) { int VectorWidth) {
assert(isl_schedule_node_get_type(Node) == isl_schedule_node_band); assert(isl_schedule_node_get_type(Node) == isl_schedule_node_band);
auto Space = isl_schedule_node_band_get_space(Node); auto Space = isl_schedule_node_band_get_space(Node);
@ -319,9 +194,9 @@ IslScheduleOptimizer::prevectSchedBand(__isl_take isl_schedule_node *Node,
} }
__isl_give isl_schedule_node * __isl_give isl_schedule_node *
IslScheduleOptimizer::tileNode(__isl_take isl_schedule_node *Node, ScheduleTreeOptimizer::tileNode(__isl_take isl_schedule_node *Node,
const char *Identifier, ArrayRef<int> TileSizes, const char *Identifier, ArrayRef<int> TileSizes,
int DefaultTileSize) { int DefaultTileSize) {
auto Ctx = isl_schedule_node_get_ctx(Node); auto Ctx = isl_schedule_node_get_ctx(Node);
auto Space = isl_schedule_node_band_get_space(Node); auto Space = isl_schedule_node_band_get_space(Node);
auto Dims = isl_space_dim(Space, isl_dim_set); auto Dims = isl_space_dim(Space, isl_dim_set);
@ -346,7 +221,7 @@ IslScheduleOptimizer::tileNode(__isl_take isl_schedule_node *Node,
return Node; return Node;
} }
bool IslScheduleOptimizer::isTileableBandNode( bool ScheduleTreeOptimizer::isTileableBandNode(
__isl_keep isl_schedule_node *Node) { __isl_keep isl_schedule_node *Node) {
if (isl_schedule_node_get_type(Node) != isl_schedule_node_band) if (isl_schedule_node_get_type(Node) != isl_schedule_node_band)
return false; return false;
@ -375,8 +250,8 @@ bool IslScheduleOptimizer::isTileableBandNode(
} }
__isl_give isl_schedule_node * __isl_give isl_schedule_node *
IslScheduleOptimizer::optimizeBand(__isl_take isl_schedule_node *Node, ScheduleTreeOptimizer::optimizeBand(__isl_take isl_schedule_node *Node,
void *User) { void *User) {
if (!isTileableBandNode(Node)) if (!isTileableBandNode(Node))
return Node; return Node;
@ -405,7 +280,7 @@ IslScheduleOptimizer::optimizeBand(__isl_take isl_schedule_node *Node,
for (int i = Dims - 1; i >= 0; i--) for (int i = Dims - 1; i >= 0; i--)
if (isl_schedule_node_band_member_get_coincident(Node, i)) { if (isl_schedule_node_band_member_get_coincident(Node, i)) {
Node = IslScheduleOptimizer::prevectSchedBand(Node, i, PrevectorWidth); Node = prevectSchedBand(Node, i, PrevectorWidth);
break; break;
} }
@ -413,17 +288,22 @@ IslScheduleOptimizer::optimizeBand(__isl_take isl_schedule_node *Node,
} }
__isl_give isl_schedule * __isl_give isl_schedule *
IslScheduleOptimizer::addPostTransforms(__isl_take isl_schedule *Schedule) { ScheduleTreeOptimizer::optimizeSchedule(__isl_take isl_schedule *Schedule) {
isl_schedule_node *Root = isl_schedule_get_root(Schedule); isl_schedule_node *Root = isl_schedule_get_root(Schedule);
Root = optimizeScheduleNode(Root);
isl_schedule_free(Schedule); isl_schedule_free(Schedule);
Root = isl_schedule_node_map_descendant_bottom_up(
Root, IslScheduleOptimizer::optimizeBand, NULL);
auto S = isl_schedule_node_get_schedule(Root); auto S = isl_schedule_node_get_schedule(Root);
isl_schedule_node_free(Root); isl_schedule_node_free(Root);
return S; return S;
} }
bool IslScheduleOptimizer::isProfitableSchedule( __isl_give isl_schedule_node *ScheduleTreeOptimizer::optimizeScheduleNode(
__isl_take isl_schedule_node *Node) {
Node = isl_schedule_node_map_descendant_bottom_up(Node, optimizeBand, NULL);
return Node;
}
bool ScheduleTreeOptimizer::isProfitableSchedule(
Scop &S, __isl_keep isl_union_map *NewSchedule) { Scop &S, __isl_keep isl_union_map *NewSchedule) {
// To understand if the schedule has been optimized we check if the schedule // To understand if the schedule has been optimized we check if the schedule
// has changed at all. // has changed at all.
@ -440,6 +320,33 @@ bool IslScheduleOptimizer::isProfitableSchedule(
return changed; return changed;
} }
namespace {
class IslScheduleOptimizer : public ScopPass {
public:
static char ID;
explicit IslScheduleOptimizer() : ScopPass(ID) { LastSchedule = nullptr; }
~IslScheduleOptimizer() { isl_schedule_free(LastSchedule); }
bool runOnScop(Scop &S) override;
void printScop(raw_ostream &OS, Scop &S) const override;
void getAnalysisUsage(AnalysisUsage &AU) const override;
private:
isl_schedule *LastSchedule;
using llvm::Pass::doFinalization;
virtual bool doFinalization() override {
isl_schedule_free(LastSchedule);
LastSchedule = nullptr;
return true;
}
};
}
char IslScheduleOptimizer::ID = 0;
bool IslScheduleOptimizer::runOnScop(Scop &S) { bool IslScheduleOptimizer::runOnScop(Scop &S) {
// Skip empty SCoPs but still allow code generation as it will delete the // Skip empty SCoPs but still allow code generation as it will delete the
@ -562,10 +469,10 @@ bool IslScheduleOptimizer::runOnScop(Scop &S) {
isl_printer_free(P); isl_printer_free(P);
}); });
isl_schedule *NewSchedule = addPostTransforms(Schedule); isl_schedule *NewSchedule = ScheduleTreeOptimizer::optimizeSchedule(Schedule);
isl_union_map *NewScheduleMap = isl_schedule_get_map(NewSchedule); isl_union_map *NewScheduleMap = isl_schedule_get_map(NewSchedule);
if (!isProfitableSchedule(S, NewScheduleMap)) { if (!ScheduleTreeOptimizer::isProfitableSchedule(S, NewScheduleMap)) {
isl_union_map_free(NewScheduleMap); isl_union_map_free(NewScheduleMap);
isl_schedule_free(NewSchedule); isl_schedule_free(NewSchedule);
return false; return false;