From 0c2f06d34510c5b8ecb0890950ce06892a4423a3 Mon Sep 17 00:00:00 2001
From: Dominic Kempf <dominic.kempf@iwr.uni-heidelberg.de>
Date: Thu, 17 Nov 2016 14:03:53 +0100
Subject: [PATCH] Saving some work
---
python/dune/perftool/loopy/__init__.py | 1 +
.../loopy/transformations/collect_rotate.py | 184 ++++++++++++++++++
.../loopy/transformations/vectorview.py | 9 +
3 files changed, 194 insertions(+)
create mode 100644 python/dune/perftool/loopy/transformations/collect_rotate.py
diff --git a/python/dune/perftool/loopy/__init__.py b/python/dune/perftool/loopy/__init__.py
index 2ef32a24..e645b4c8 100644
--- a/python/dune/perftool/loopy/__init__.py
+++ b/python/dune/perftool/loopy/__init__.py
@@ -1,4 +1,5 @@
""" Export the interface interesting to the rest of the project """
from dune.perftool.loopy.transformations.collect_precompute import collect_vector_data_precompute
+from dune.perftool.loopy.transformations.collect_rotate import collect_vector_data_rotate
from dune.perftool.loopy.transformations.duplicate import heuristic_duplication
diff --git a/python/dune/perftool/loopy/transformations/collect_rotate.py b/python/dune/perftool/loopy/transformations/collect_rotate.py
new file mode 100644
index 00000000..d702ab8c
--- /dev/null
+++ b/python/dune/perftool/loopy/transformations/collect_rotate.py
@@ -0,0 +1,184 @@
+""" A kernel transformation that precomputes quantities until a vector register
+is filled and then does vector computations """
+
+from dune.perftool.loopy.vcl import get_vcl_type_size
+from dune.perftool.loopy.transformations.vectorview import (add_temporary_with_vector_view,
+ add_vector_view,
+ get_vector_view_name,
+ )
+from dune.perftool.tools import get_pymbolic_basename
+
+from loopy.kernel.creation import parse_domains
+from loopy.symbolic import pw_aff_to_expr
+
+from pymbolic.mapper.dependency import DependencyMapper
+from pymbolic.mapper.substitutor import substitute
+
+import pymbolic.primitives as prim
+import loopy as lp
+import numpy as np
+
+
+def collect_vector_data_rotate(knl, insns, inames):
+ #
+ # Process/Assert/Standardize the input
+ #
+
+ # inames input -> tuple
+ if isinstance(inames, str):
+ inames = inames.split(",")
+ inames = tuple(i.strip() for i in inames)
+
+ # insns -> list of Instruction instances
+ if isinstance(insns, lp.match.MatchExpressionBase):
+ insns = lp.find_instructions(knl, insns)
+ else:
+ if isinstance(insns, str):
+ insns = [i.strip() for i in insns.split(",")]
+ insns = [knl.id_to_insn[i] for i in insns]
+
+ # Analyse the inames of the given instructions and identify inames
+ # that they all have in common. Those inames will also be iname dependencies
+ # of inserted instructions.
+ common_inames = frozenset([]).union(*(insn.within_inames for insn in insns)) - frozenset(inames)
+
+ # Determine the vector lane width
+ # TODO infer the numpy type here
+ vec_size = get_vcl_type_size(np.float64)
+
+ #
+ # Inspect the given instructions for dependent quantities
+ #
+
+ quantities = {}
+ for insn in insns:
+ for expr in DependencyMapper()(insn.expression):
+ basename = get_pymbolic_basename(expr)
+ quantities.setdefault(basename, frozenset())
+ quantities[basename] = quantities[basename].union(frozenset([expr]))
+ assert all(len(q) == 1 for q in quantities.values())
+
+ # Add vector size buffers for all these quantities
+ replacemap_arr = {}
+ replacemap_vec = {}
+ for quantity in quantities:
+ expr, = quantities[quantity]
+ arrname = quantity + '_buffered_arr'
+ knl = add_temporary_with_vector_view(knl,
+ arrname,
+ dtype=np.float64,
+ shape=(vec_size,),
+ dim_tags="c",
+ base_storage=quantity + '_base_storage',
+ )
+
+ replacemap_arr[quantity] = prim.Subscript(prim.Variable(arrname), (prim.Variable('rotate_index'),))
+ replacemap_vec[expr] = prim.Variable(get_vector_view_name(arrname))
+
+ write_match = lp.match.Or(tuple(lp.match.Writes(q) for q in quantities))
+ iname_match = lp.match.And(tuple(lp.match.Iname(i) for i in inames))
+ match = lp.match.And((write_match, iname_match))
+ write_insns = lp.find_instructions(knl, match)
+
+ other_insns = [i for i in knl.instructions if i.id not in [j.id for j in insns + write_insns]]
+ new_insns = []
+ temporaries = knl.temporary_variables
+
+ for insn in write_insns:
+ if isinstance(insn, lp.Assignment):
+ new_insns.append(insn.copy(assignee=replacemap_arr[get_pymbolic_basename(insn.assignee)],
+ )
+ )
+ elif isinstance(insn, lp.CInstruction):
+ pass
+ else:
+ raise NotImplementedError
+
+ #
+ # Add two counter variables to the kernel
+ #
+
+ # Insert a flat consecutive counter 'total_index'
+ temporaries['total_index'] = lp.TemporaryVariable('total_index', # name
+ dtype=np.int32,
+ )
+ new_insns.append(lp.Assignment(prim.Variable("total_index"), # assignee
+ 0, # expression
+ within_inames=common_inames,
+ within_inames_is_final=True,
+ id="assign_total_index",
+ ))
+ new_insns.append(lp.Assignment(prim.Variable("total_index"), # assignee
+ prim.Sum((prim.Variable("total_index"), 1)), # expression
+ within_inames=common_inames.union(inames),
+ within_inames_is_final=True,
+ depends_on=frozenset([i.id for i in write_insns]).union(frozenset({"assign_total_index"})),
+ depends_on_is_final=True,
+ id="update_total_index",
+ ))
+
+ # Insert a rotating index, that counts 0 , .. , vecsize - 1
+ temporaries['rotate_index'] = lp.TemporaryVariable('rotate_index', # name
+ dtype=np.int32,
+ )
+ new_insns.append(lp.Assignment(prim.Variable("rotate_index"), # assignee
+ 0, # expression
+ within_inames=common_inames,
+ within_inames_is_final=True,
+ id="assign_rotate_index",
+ ))
+ new_insns.append(lp.Assignment(prim.Variable("rotate_index"), # assignee
+ prim.Remainder(prim.Sum((prim.Variable("rotate_index"), 1)), vec_size), # expression
+ within_inames=common_inames.union(inames),
+ within_inames_is_final=True,
+ depends_on=frozenset([i.id for i in write_insns]).union(frozenset({"assign_rotate_index"})),
+ depends_on_is_final=True,
+ id="update_rotate_index",
+ ))
+
+ #
+ # Construct a flat loop for the given instructions
+ #
+
+# new_insns = []
+# other_insns = [i for i in knl.instructions if i.id not in [j.id for j in insns]]
+#
+# size = prim.Product(tuple(pw_aff_to_expr(knl.get_iname_bounds(i).size) for i in inames))
+# size = prim.FloorDiv(size, vec_size)
+#
+# temporaries = knl.temporary_variables
+# temporaries["flatsize"] = lp.TemporaryVariable("flatsize",
+# dtype=np.int32,
+# shape=(),
+# )
+# new_insns.append(lp.Assignment(prim.Variable("flatsize"),
+# size,
+# )
+# )
+#
+# # Add an additional domain to the kernel
+# new_iname = "flat_{}".format("_".join(inames))
+# domain = "{{ [{0}] : 0<={0}<flatsize }}".format(new_iname, str(size))
+# domain = parse_domains(domain, {})
+# knl = knl.copy(domains=knl.domains + domain,
+# temporary_variables=temporaries)
+#
+# # Split and tag the flat iname
+# knl = lp.split_iname(knl, new_iname, vec_size, inner_tag="vec")
+# new_inames = ("{}_outer".format(new_iname), "{}_inner".format(new_iname))
+# knl = lp.assume(knl, "flatsize mod {} = 0".format(vec_size))
+#
+# for insn in insns:
+# # Get a vector view of the lhs expression
+# lhsname = get_pymbolic_basename(insn.assignee)
+# knl = add_vector_view(knl, lhsname)
+# lhsname = get_vector_view_name(lhsname)
+#
+# new_insns.append(lp.Assignment(prim.Subscript(prim.Variable(lhsname), tuple(prim.Variable(i) for i in new_inames)),
+# prim.Subscript(prim.Variable(get_vector_view_name("wk_precomputed")), tuple(prim.Variable(i) for i in new_inames)),
+# within_inames=frozenset(new_inames),
+# within_inames_is_final=True,
+# )
+# )
+
+ return knl.copy(instructions=new_insns + other_insns)
diff --git a/python/dune/perftool/loopy/transformations/vectorview.py b/python/dune/perftool/loopy/transformations/vectorview.py
index 86aefa81..dd36969b 100644
--- a/python/dune/perftool/loopy/transformations/vectorview.py
+++ b/python/dune/perftool/loopy/transformations/vectorview.py
@@ -46,3 +46,12 @@ def add_vector_view(knl, tmpname):
)
return knl.copy(temporary_variables=temporaries)
+
+
+def add_temporary_with_vector_view(knl, name, *args, **kwargs):
+ temps = knl.temporary_variables
+ assert name not in temps
+ temps[name] = lp.TemporaryVariable(name, *args, **kwargs)
+ knl = knl.copy(temporary_variables=temps)
+ knl = add_vector_view(knl, name)
+ return knl
\ No newline at end of file
--
GitLab