From 1878563e87db3bc578fa1a31513ae5abf9f92304 Mon Sep 17 00:00:00 2001
From: Dominic Kempf <dominic.kempf@iwr.uni-heidelberg.de>
Date: Thu, 8 Feb 2018 16:50:39 +0100
Subject: [PATCH] Fix minimization over number of quadrature points
So far, it did not set the qp tuple correctly and
thus sometimes resulted in too big numbers being
chosen.
---
python/dune/perftool/sumfact/vectorization.py | 22 ++++++++++++-------
1 file changed, 14 insertions(+), 8 deletions(-)
diff --git a/python/dune/perftool/sumfact/vectorization.py b/python/dune/perftool/sumfact/vectorization.py
index 38ec9b16..d7fb6776 100644
--- a/python/dune/perftool/sumfact/vectorization.py
+++ b/python/dune/perftool/sumfact/vectorization.py
@@ -95,10 +95,12 @@ def explicit_costfunction(sf):
return 1000000000000
-def strategy_cost(strategy):
+def strategy_cost(strat_tuple):
+ qp, strategy = strat_tuple
func = get_backend(interface="vectorization_strategy",
selector=lambda: get_form_option("vectorization_strategy"))
keys = set(sf.cache_key for sf in strategy.values())
+ set_quadrature_points(qp)
# Sum over all the sum factorization kernels in the realization
score = 0.0
@@ -110,6 +112,13 @@ def strategy_cost(strategy):
return score
+def fixedqp_strategy_costfunction(qp):
+ def _cost(strategy):
+ return strategy_cost((qp, strategy))
+
+ return _cost
+
+
def stringify_vectorization_strategy(strategy):
result = []
qp, strategy = strategy
@@ -189,11 +198,10 @@ def decide_vectorization_strategy():
for qp in quad_points:
for strat in fixed_quad_vectorization_opportunity_generator(frozenset(stage1_sumfacts), width, qp):
if strategy == int(get_form_option("vectorization_list_index")):
- set_quadrature_points(qp)
# Output the strategy and its cost into a separate file
if get_global_context_value("form_type") == "jacobian_apply":
with open("strategycosts.csv", "a") as f:
- f.write("{} {}\n".format(strategy, strategy_cost(strat)))
+ f.write("{} {}\n".format(strategy, strategy_cost((qp, strat))))
return qp, strat
strategy = strategy + 1
@@ -210,12 +218,12 @@ def decide_vectorization_strategy():
for key in keys:
key_sumfacts = frozenset(sf for sf in active_sumfacts if sf.input_key == key)
minimum = min(fixed_quad_vectorization_opportunity_generator(key_sumfacts, width, qp),
- key=strategy_cost)
+ key=fixedqp_strategy_costfunction(qp))
sfdict = add_to_frozendict(sfdict, minimum)
else:
# Find the minimum cost strategy between all the quadrature point tuples
optimal_strategies = {qp: fixed_quadrature_optimal_vectorization(active_sumfacts, width, qp) for qp in quad_points}
- qp = min(optimal_strategies, key=lambda qp: strategy_cost(optimal_strategies[qp]))
+ qp = min(optimal_strategies, key=lambda qp: strategy_cost((qp, optimal_strategies[qp])))
sfdict = optimal_strategies[qp]
set_quadrature_points(qp)
@@ -238,8 +246,6 @@ def fixed_quadrature_optimal_vectorization(sumfacts, width, qp):
opportunities and score them individually, but we need to do a divide and conquer
approach.
"""
- set_quadrature_points(qp)
-
# Find the sets of simultaneously realizable kernels (thats an equivalence relation)
keys = frozenset(sf.input_key for sf in sumfacts)
@@ -249,7 +255,7 @@ def fixed_quadrature_optimal_vectorization(sumfacts, width, qp):
for key in keys:
key_sumfacts = frozenset(sf for sf in sumfacts if sf.input_key == key)
minimum = min(fixed_quad_vectorization_opportunity_generator(key_sumfacts, width, qp),
- key=strategy_cost)
+ key=fixedqp_strategy_costfunction(qp))
sfdict = add_to_frozendict(sfdict, minimum)
return sfdict
--
GitLab