282 lines
10 KiB
Python
282 lines
10 KiB
Python
# SPDX-FileCopyrightText: 2021 Guillaume DIDIER
|
|
#
|
|
# SPDX-License-Identifier: Apache-2.0
|
|
# SPDX-License-Identifier: MIT
|
|
|
|
import pandas as pd
|
|
import matplotlib.pyplot as plt
|
|
import seaborn as sns
|
|
from sys import exit
|
|
import numpy as np
|
|
from scipy import optimize
|
|
import sys
|
|
|
|
# TODO
|
|
# sys.argv[1] should be the root
|
|
# with root-result_lite.csv.bz2 the result
|
|
# and .stats.csv
|
|
# root.slices a slice mapping - done
|
|
# root.cores a core + socket mapping - done -> move to analyse csv ?
|
|
#
|
|
# Facet plot with actual dot cloud + plot the linear regression
|
|
# each row is a slice
|
|
# each row is an origin core
|
|
# each column a helper core if applicable
|
|
|
|
|
|
stats = pd.read_csv(sys.argv[1] + ".stats.csv",
|
|
dtype={
|
|
"main_core": np.int8,
|
|
"helper_core": np.int8,
|
|
# "address": int,
|
|
"hash": np.int8,
|
|
# "time": np.int16,
|
|
"clflush_remote_hit": np.float64,
|
|
"clflush_shared_hit": np.float64,
|
|
# "clflush_miss_f": np.int32,
|
|
# "clflush_local_hit_f": np.int32,
|
|
"clflush_miss_n": np.float64,
|
|
"clflush_local_hit_n": np.float64,
|
|
# "reload_miss": np.int32,
|
|
# "reload_remote_hit": np.int32,
|
|
# "reload_shared_hit": np.int32,
|
|
# "reload_local_hit": np.int32
|
|
}
|
|
)
|
|
|
|
slice_mapping = pd.read_csv(sys.argv[1] + ".slices.csv")
|
|
core_mapping = pd.read_csv(sys.argv[1] + ".cores.csv")
|
|
|
|
print(core_mapping.to_string())
|
|
print(slice_mapping.to_string())
|
|
|
|
print("core {} is mapped to '{}'".format(4, repr(core_mapping.iloc[4])))
|
|
|
|
min_time_miss = stats["clflush_miss_n"].min()
|
|
max_time_miss = stats["clflush_miss_n"].max()
|
|
|
|
|
|
def remap_core(key):
|
|
def remap(core):
|
|
remapped = core_mapping.iloc[core]
|
|
return remapped[key]
|
|
|
|
return remap
|
|
|
|
|
|
stats["main_socket"] = stats["main_core"].apply(remap_core("socket"))
|
|
stats["main_core_fixed"] = stats["main_core"].apply(remap_core("core"))
|
|
stats["main_ht"] = stats["main_core"].apply(remap_core("hthread"))
|
|
stats["helper_socket"] = stats["helper_core"].apply(remap_core("socket"))
|
|
stats["helper_core_fixed"] = stats["helper_core"].apply(remap_core("core"))
|
|
stats["helper_ht"] = stats["helper_core"].apply(remap_core("hthread"))
|
|
|
|
# slice_mapping = {3: 0, 1: 1, 2: 2, 0: 3}
|
|
|
|
stats["slice_group"] = stats["hash"].apply(lambda h: slice_mapping["slice_group"].iloc[h])
|
|
|
|
graph_lower_miss = int((min_time_miss // 10) * 10)
|
|
graph_upper_miss = int(((max_time_miss + 9) // 10) * 10)
|
|
|
|
print("Graphing from {} to {}".format(graph_lower_miss, graph_upper_miss))
|
|
|
|
g_ = sns.FacetGrid(stats, col="main_core_fixed", row="slice_group")
|
|
|
|
g_.map(sns.distplot, 'clflush_miss_n', bins=range(graph_lower_miss, graph_upper_miss), color="b")
|
|
#g.map(sns.scatterplot, 'slice_group', 'clflush_local_hit_n', color="g")
|
|
plt.show()
|
|
|
|
|
|
|
|
# also explains remote
|
|
# shared needs some thinking as there is something weird happening there.
|
|
|
|
#
|
|
# M 0 1 2 3 4 5 6 7
|
|
#
|
|
|
|
|
|
print(stats.head())
|
|
|
|
num_core = len(stats["main_core_fixed"].unique())
|
|
print("Found {}".format(num_core))
|
|
|
|
|
|
def miss_topology(main_core_fixed, slice_group, C, h):
|
|
return C + h * abs(main_core_fixed - slice_group) + h * abs(slice_group + 1)
|
|
|
|
def miss_topology_df(x, C, h):
|
|
return x.apply(lambda x, C, h: miss_topology(x["main_core_fixed"], x["slice_group"], C, h), args=(C, h), axis=1)
|
|
|
|
|
|
|
|
res_miss = optimize.curve_fit(miss_topology_df, stats[["main_core_fixed", "slice_group"]], stats["clflush_miss_n"])
|
|
print("Miss topology:")
|
|
print(res_miss)
|
|
|
|
|
|
memory = -1
|
|
gpu_if_any = num_core
|
|
|
|
|
|
def exclusive_hit_topology_gpu(main_core, slice_group, helper_core, C, h1, h2):
|
|
round_trip = gpu_if_any - memory
|
|
|
|
if slice_group <= num_core/2:
|
|
# send message towards higher cores first
|
|
if helper_core < slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(round_trip - (helper_core - memory))
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
else:
|
|
# send message toward lower cores first
|
|
if helper_core > slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - memory)
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
return r
|
|
|
|
|
|
def exclusive_hit_topology_gpu_df(x, C, h1, h2):
|
|
return x.apply(lambda x, C, h1, h2: exclusive_hit_topology_gpu(x["main_core_fixed"], x["slice_group"], x["helper_core_fixed"], C, h1, h2), args=(C, h1, h2), axis=1)
|
|
|
|
|
|
def exclusive_hit_topology_gpu2(main_core, slice_group, helper_core, C, h1, h2):
|
|
round_trip = gpu_if_any + 1 - memory
|
|
|
|
if slice_group <= num_core/2:
|
|
# send message towards higher cores first
|
|
if helper_core < slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(round_trip - (helper_core - memory))
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
else:
|
|
# send message toward lower cores first
|
|
if helper_core > slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - memory)
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
return r
|
|
|
|
|
|
def exclusive_hit_topology_gpu2_df(x, C, h1, h2):
|
|
return x.apply(lambda x, C, h1, h2: exclusive_hit_topology_gpu2(x["main_core_fixed"], x["slice_group"], x["helper_core_fixed"], C, h1, h2), args=(C, h1, h2), axis=1)
|
|
|
|
|
|
# unlikely
|
|
def exclusive_hit_topology_nogpu(main_core, slice_group, helper_core, C, h1, h2):
|
|
round_trip = (num_core-1) - memory
|
|
|
|
if slice_group <= num_core/2:
|
|
# send message towards higher cores first
|
|
if helper_core < slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(round_trip - (helper_core - memory))
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
else:
|
|
# send message toward lower cores first
|
|
if helper_core > slice_group:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - memory)
|
|
else:
|
|
r = C + h1 * abs(main_core - slice_group) + h2 * abs(helper_core - slice_group)
|
|
return r
|
|
|
|
|
|
def exclusive_hit_topology_nogpu_df(x, C, h1, h2):
|
|
return x.apply(lambda x, C, h1, h2: exclusive_hit_topology_nogpu(x["main_core_fixed"], x["slice_group"], x["helper_core_fixed"], C, h1, h2), args=(C, h1, h2), axis=1)
|
|
|
|
|
|
#res_no_gpu = optimize.curve_fit(exclusive_hit_topology_nogpu_df, stats[["main_core_fixed", "slice_group", "helper_core_fixed"]], stats["clflush_remote_hit"])
|
|
#print("Exclusive hit topology (No GPU):")
|
|
#print(res_no_gpu)
|
|
|
|
res_gpu = optimize.curve_fit(exclusive_hit_topology_gpu_df, stats[["main_core_fixed", "slice_group", "helper_core_fixed"]], stats["clflush_remote_hit"])
|
|
print("Exclusive hit topology (GPU):")
|
|
print(res_gpu)
|
|
|
|
#res_gpu2 = optimize.curve_fit(exclusive_hit_topology_gpu2_df, stats[["main_core_fixed", "slice_group", "helper_core_fixed"]], stats["clflush_remote_hit"])
|
|
#print("Exclusive hit topology (GPU2):")
|
|
#print(res_gpu2)
|
|
|
|
|
|
|
|
def remote_hit_topology_2(x, C, h):
|
|
main_core = x["main_core_fixed"]
|
|
slice_group = x["slice_group"]
|
|
helper_core = x["helper_core_fixed"]
|
|
return C + h * abs(main_core - slice_group) + h * abs(slice_group - helper_core) + h * abs(helper_core - main_core)
|
|
|
|
|
|
def shared_hit_topology_1(x, C, h):
|
|
main_core = x["main_core_fixed"]
|
|
slice_group = x["slice_group"]
|
|
helper_core = x["helper_core_fixed"]
|
|
return C + h * abs(main_core - slice_group) + h * max(abs(slice_group - main_core), abs(slice_group - helper_core))
|
|
|
|
|
|
def plot_func(function, *params):
|
|
def plot_it(x, **kwargs):
|
|
# plot_x = []
|
|
# plot_y = []
|
|
# for x in set(x):
|
|
# plot_y.append(function(x, *params))
|
|
# plot_x = x
|
|
print(x)
|
|
plot_y = function(x, *params)
|
|
sns.lineplot(x, plot_y, **kwargs)
|
|
return plot_it
|
|
|
|
stats["predicted_miss"] = miss_topology_df(stats, *(res_miss[0]))
|
|
|
|
figure_median_I = sns.FacetGrid(stats, col="main_core_fixed")
|
|
figure_median_I.map(sns.scatterplot, 'slice_group', 'clflush_miss_n', color="b")
|
|
figure_median_I.map(sns.lineplot, 'slice_group', 'predicted_miss', color="b")
|
|
figure_median_I.set_titles(col_template="$A$ = {col_name}")
|
|
figure_median_I.tight_layout()
|
|
|
|
import tikzplotlib
|
|
|
|
tikzplotlib.save("fig-median-I.tex", axis_width=r'0.175\textwidth', axis_height=r'0.25\textwidth')
|
|
plt.show()
|
|
|
|
#stats["predicted_remote_hit_no_gpu"] = exclusive_hit_topology_nogpu_df(stats, *(res_no_gpu[0]))
|
|
stats["predicted_remote_hit_gpu"] = exclusive_hit_topology_gpu_df(stats, *(res_gpu[0]))
|
|
#stats["predicted_remote_hit_gpu2"] = exclusive_hit_topology_gpu_df(stats, *(res_gpu2[0]))
|
|
|
|
|
|
stats_A0 = stats[stats["main_core_fixed"] == 0]
|
|
figure_median_E_A0 = sns.FacetGrid(stats_A0, col="slice_group")
|
|
figure_median_E_A0.map(sns.scatterplot, 'helper_core_fixed', 'clflush_remote_hit', color="r")
|
|
figure_median_E_A0.map(sns.lineplot, 'helper_core_fixed', 'predicted_remote_hit_gpu', color="r")
|
|
figure_median_E_A0.set_titles(col_template="$S$ = {col_name}")
|
|
|
|
tikzplotlib.save("fig-median-E-A0.tex", axis_width=r'0.175\textwidth', axis_height=r'0.25\textwidth')
|
|
plt.show()
|
|
|
|
g = sns.FacetGrid(stats, row="main_core_fixed")
|
|
|
|
g.map(sns.scatterplot, 'slice_group', 'clflush_miss_n', color="b")
|
|
g.map(sns.scatterplot, 'slice_group', 'clflush_local_hit_n', color="g")
|
|
|
|
g0 = sns.FacetGrid(stats, row="slice_group")
|
|
|
|
g0.map(sns.scatterplot, 'main_core_fixed', 'clflush_miss_n', color="b")
|
|
g0.map(sns.scatterplot, 'main_core_fixed', 'clflush_local_hit_n', color="g") # this gives away the trick I think !
|
|
# possibility of sending a general please discard this everyone around one of the ring + wait for ACK - direction depends on the core.
|
|
|
|
|
|
|
|
g2 = sns.FacetGrid(stats, row="main_core_fixed", col="slice_group")
|
|
g2.map(sns.scatterplot, 'helper_core_fixed', 'clflush_remote_hit', color="r")
|
|
g2.map(sns.lineplot, 'helper_core_fixed', 'predicted_remote_hit_gpu', color="r")
|
|
#g2.map(sns.lineplot, 'helper_core_fixed', 'predicted_remote_hit_gpu2', color="g")
|
|
#g2.map(sns.lineplot, 'helper_core_fixed', 'predicted_remote_hit_no_gpu', color="g")
|
|
#g2.map(plot_func(exclusive_hit_topology_nogpu_df, *(res_no_gpu[0])), 'helper_core_fixed', color="g")
|
|
|
|
g3 = sns.FacetGrid(stats, row="main_core_fixed", col="slice_group")
|
|
g3.map(sns.scatterplot, 'helper_core_fixed', 'clflush_shared_hit', color="y")
|
|
|
|
# more ideas needed
|
|
|
|
plt.show()
|