source: doc/theses/mubeen_zulfiqar_MMath/performance.tex @ c9136d9

pthread-emulationqualifiedEnum
Last change on this file since c9136d9 was c9136d9, checked in by Peter A. Buhr <pabuhr@…>, 9 months ago

first proofread of chapter performance

  • Property mode set to 100644
File size: 25.3 KB
Line 
1\chapter{Performance}
2\label{c:Performance}
3
4This chapter uses the micro-benchmarks from \VRef[Chapter]{s:Benchmarks} to test a number of current memory allocators, including llheap.
5The goal is to see if llheap is competitive with the current best memory allocators.
6
7
8\section{Machine Specification}
9
10The performance experiments were run on two different multi-core architectures (x86 and ARM) to determine if there is consistency across platforms:
11\begin{itemize}
12\item
13\textbf{Nasus} AMD EPYC 7662, 64-core socket $\times$ 2, 2.0 GHz, GCC version 9.3.0
14\item
15\textbf{Algol} Huawei ARM TaiShan 2280 V2 Kunpeng 920, 24-core socket $\times$ 4, 2.6 GHz, GCC version 9.4.0
16\end{itemize}
17
18
19\section{Existing Memory Allocators}
20\label{sec:curAllocatorSec}
21
22With dynamic allocation being an important feature of C, there are many stand-alone memory allocators that have been designed for different purposes.
23For this thesis, 7 of the most popular and widely used memory allocators were selected for comparison.
24
25\subsection{glibc}
26glibc~\cite{glibc} is the default gcc thread-safe allocator.
27\\
28\textbf{Version:} Ubuntu GLIBC 2.31-0ubuntu9.7 2.31\\
29\textbf{Configuration:} Compiled by Ubuntu 20.04.\\
30\textbf{Compilation command:} N/A
31
32\subsection{dlmalloc}
33dlmalloc~\cite{dlmalloc} is a thread-safe allocator that is single threaded and single heap.
34It maintains free-lists of different sizes to store freed dynamic memory.
35\\
36\textbf{Version:} 2.8.6\\
37\textbf{Configuration:} Compiled with preprocessor @USE_LOCKS@.\\
38\textbf{Compilation command:} @gcc -g3 -O3 -Wall -Wextra -fno-builtin-malloc -fno-builtin-calloc@ @-fno-builtin-realloc -fno-builtin-free -fPIC -shared -DUSE_LOCKS -o libdlmalloc.so malloc-2.8.6.c@
39
40\subsection{hoard}
41Hoard~\cite{hoard} is a thread-safe allocator that is multi-threaded and using a heap layer framework. It has per-thread heaps that have thread-local free-lists, and a global shared heap.
42\\
43\textbf{Version:} 3.13\\
44\textbf{Configuration:} Compiled with hoard's default configurations and @Makefile@.\\
45\textbf{Compilation command:} @make all@
46
47\subsection{jemalloc}
48jemalloc~\cite{jemalloc} is a thread-safe allocator that uses multiple arenas. Each thread is assigned an arena. Each arena has chunks that contain contagious memory regions of same size. An arena has multiple chunks that contain regions of multiple sizes.
49\\
50\textbf{Version:} 5.2.1\\
51\textbf{Configuration:} Compiled with jemalloc's default configurations and @Makefile@.\\
52\textbf{Compilation command:} @autogen.sh; configure; make; make install@
53
54\subsection{pt3malloc}
55pt3malloc~\cite{pt3malloc} is a modification of dlmalloc.
56It is a thread-safe multi-threaded memory allocator that uses multiple heaps. pt3malloc heap has similar design to dlmalloc's heap.
57\\
58\textbf{Version:} 1.8\\
59\textbf{Configuration:} Compiled with pt3malloc's @Makefile@ using option ``linux-shared''.\\
60\textbf{Compilation command:} @make linux-shared@
61
62\subsection{rpmalloc}
63rpmalloc~\cite{rpmalloc} is a thread-safe allocator that is multi-threaded and uses per-thread heap. Each heap has multiple size-classes and each size-class contains memory regions of the relevant size.
64\\
65\textbf{Version:} 1.4.1\\
66\textbf{Configuration:} Compiled with rpmalloc's default configurations and ninja build system.\\
67\textbf{Compilation command:} @python3 configure.py; ninja@
68
69\subsection{tbb malloc}
70tbb malloc~\cite{tbbmallocmail
71} is a thread-safe allocator that is multi-threaded and uses private heap for each thread. Each private-heap has multiple bins of different sizes. Each bin contains free regions of the same size.
72\\
73\textbf{Version:} intel tbb 2020 update 2, tbb\_interface\_version == 11102\\
74\textbf{Configuration:} Compiled with tbbmalloc's default configurations and @Makefile@.\\
75\textbf{Compilation command:} @make@
76
77% \section{Experiment Environment}
78% We used our micro benchmark suite (FIX ME: cite mbench) to evaluate these memory allocators \ref{sec:curAllocatorSec} and our own memory allocator uHeap \ref{sec:allocatorSec}.
79
80\section{Experiments}
81% FIX ME: add experiment, knobs, graphs, description+analysis
82
83%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
84%% CHURN
85%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
86
87\subsection{Churn Micro-Benchmark}
88
89Churn tests allocators for speed under intensive dynamic memory usage (see \VRef{s:ChurnBenchmark}).
90This experiment was run with following configurations:
91\begin{description}[itemsep=0pt,parsep=0pt]
92\item[thread:]
931, 2, 4, 8, 16
94\item[spots:]
9516
96\item[obj:]
97100,000
98\item[max:]
99500
100\item[min:]
10150
102\item[step:]
10350
104\item[distro:]
105fisher
106\end{description}
107
108% -maxS          : 500
109% -minS          : 50
110% -stepS                 : 50
111% -distroS       : fisher
112% -objN          : 100000
113% -cSpots                : 16
114% -threadN       : 1, 2, 4, 8, 16
115
116\VRef[Figure]{fig:churn} shows the results for algol and nasus.
117The X-axis shows the number of threads.
118Each allocator's performance for each thread is shown in different colors.
119The Y-axis shows the total experiment time.
120
121\begin{figure}
122\centering
123    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/churn} }
124    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/churn} }
125\caption{Churn}
126\label{fig:churn}
127\end{figure}
128
129All allocators did well in this micro-benchmark, except for dmalloc on the ARM,
130
131
132%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
133%% THRASH
134%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
135
136\subsection{Cache Thrash}
137
138Thrash tests memory allocators for active false sharing (see \VRef{sec:benchThrashSec}).
139This experiment was run with following configurations:
140\begin{description}[itemsep=0pt,parsep=0pt]
141\item[thread:]
1421, 2, 4, 8, 16
143\item[iterations:]
1441,000
145\item[cacheRW:]
1461,000,000
147\item[size:]
1481
149\end{description}
150
151% * Each allocator was tested for its performance across different number of threads.
152% Experiment was repeated for each allocator for 1, 2, 4, 8, and 16 threads by setting the configuration -threadN.
153
154Results are shown in figure \ref{fig:cacheThrash} for both algol and nasus.
155X-axis shows number of threads. Each allocator's performance for each thread is shown in different colors.
156Y-axis shows the total time experiment took to finish.
157
158\begin{figure}
159\centering
160    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/cache-time-0-thrash} }
161    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/cache-time-0-thrash} }
162\caption{Cache Thrash}
163\label{fig:cacheThrash}
164\end{figure}
165
166%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
167%% SCRATCH
168%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
169
170\subsection{Cache Scratch}
171\label{s:CacheScratch}
172
173Scratch tests memory allocators for program-induced allocator-preserved passive false-sharing.
174This experiment was run with following configurations:
175\begin{description}[itemsep=0pt,parsep=0pt]
176\item[threads:]
1771, 2, 4, 8, 16
178\item[iterations:]
1791,000
180\item[cacheRW:]
1811,000,000
182\item[size:]
1831
184\end{description}
185
186% * Each allocator was tested for its performance across different number of threads.
187% Experiment was repeated for each allocator for 1, 2, 4, 8, and 16 threads by setting the configuration -threadN.
188
189Results are shown in figure \ref{fig:cacheScratch} for both algol and nasus.
190X-axis shows number of threads. Each allocator's performance for each thread is shown in different colors.
191Y-axis shows the total time experiment took to finish.
192
193\begin{figure}
194\centering
195    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/cache-time-0-scratch} }
196    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/cache-time-0-scratch} }
197\caption{Cache Scratch}
198\label{fig:cacheScratch}
199\end{figure}
200
201%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
202%% SPEED
203%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
204
205\subsection{Speed Micro-Benchmark}
206
207Speed testa memory allocators for runtime latency (see \VRef{s:SpeedMicroBenchmark}).
208This experiment was run with following configurations:
209\begin{description}[itemsep=0pt,parsep=0pt]
210\item[max:]
211500
212\item[min:]
21350
214\item[step:]
21550
216\item[distro:]
217fisher
218\item[objects:]
2191,000,000
220\item[workers:]
2211, 2, 4, 8, 16
222\end{description}
223
224% -maxS    :  500
225% -minS    :  50
226% -stepS   :  50
227% -distroS :  fisher
228% -objN    :  1000000
229% -threadN    : \{ 1, 2, 4, 8, 16 \} *
230
231%* Each allocator was tested for its performance across different number of threads.
232%Experiment was repeated for each allocator for 1, 2, 4, 8, and 16 threads by setting the configuration -threadN.
233
234Results for speed benchmark are shown in 12 figures, one figure for each chain of speed benchmark.
235X-axis shows number of threads. Each allocator's performance for each thread is shown in different colors.
236Y-axis shows the total time experiment took to finish. Less time an allocator takes to complete allocation, better it is so lower the better.
237
238\begin{itemize}
239\item Fig. \ref{fig:speed-3-malloc} shows results for chain: malloc
240\item Fig. \ref{fig:speed-4-realloc} shows results for chain: realloc
241\item Fig. \ref{fig:speed-5-free} shows results for chain: free
242\item Fig. \ref{fig:speed-6-calloc} shows results for chain: calloc
243\item Fig. \ref{fig:speed-7-malloc-free} shows results for chain: malloc-free
244\item Fig. \ref{fig:speed-8-realloc-free} shows results for chain: realloc-free
245\item Fig. \ref{fig:speed-9-calloc-free} shows results for chain: calloc-free
246\item Fig. \ref{fig:speed-10-malloc-realloc} shows results for chain: malloc-realloc
247\item Fig. \ref{fig:speed-11-calloc-realloc} shows results for chain: calloc-realloc
248\item Fig. \ref{fig:speed-12-malloc-realloc-free} shows results for chain: malloc-realloc-free
249\item Fig. \ref{fig:speed-13-calloc-realloc-free} shows results for chain: calloc-realloc-free
250\item Fig. \ref{fig:speed-14-malloc-calloc-realloc-free} shows results for chain: malloc-realloc-free-calloc
251\end{itemize}
252
253%speed-3-malloc.eps
254\begin{figure}
255\centering
256    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-3-malloc} }
257    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-3-malloc} }
258\caption{Speed benchmark chain: malloc}
259\label{fig:speed-3-malloc}
260\end{figure}
261
262%speed-4-realloc.eps
263\begin{figure}
264\centering
265    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-4-realloc} }
266    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-4-realloc} }
267\caption{Speed benchmark chain: realloc}
268\label{fig:speed-4-realloc}
269\end{figure}
270
271%speed-5-free.eps
272\begin{figure}
273\centering
274    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-5-free} }
275    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-5-free} }
276\caption{Speed benchmark chain: free}
277\label{fig:speed-5-free}
278\end{figure}
279
280%speed-6-calloc.eps
281\begin{figure}
282\centering
283    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-6-calloc} }
284    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-6-calloc} }
285\caption{Speed benchmark chain: calloc}
286\label{fig:speed-6-calloc}
287\end{figure}
288
289%speed-7-malloc-free.eps
290\begin{figure}
291\centering
292    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-7-malloc-free} }
293    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-7-malloc-free} }
294\caption{Speed benchmark chain: malloc-free}
295\label{fig:speed-7-malloc-free}
296\end{figure}
297
298%speed-8-realloc-free.eps
299\begin{figure}
300\centering
301    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-8-realloc-free} }
302    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-8-realloc-free} }
303\caption{Speed benchmark chain: realloc-free}
304\label{fig:speed-8-realloc-free}
305\end{figure}
306
307%speed-9-calloc-free.eps
308\begin{figure}
309\centering
310    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-9-calloc-free} }
311    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-9-calloc-free} }
312\caption{Speed benchmark chain: calloc-free}
313\label{fig:speed-9-calloc-free}
314\end{figure}
315
316%speed-10-malloc-realloc.eps
317\begin{figure}
318\centering
319    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-10-malloc-realloc} }
320    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-10-malloc-realloc} }
321\caption{Speed benchmark chain: malloc-realloc}
322\label{fig:speed-10-malloc-realloc}
323\end{figure}
324
325%speed-11-calloc-realloc.eps
326\begin{figure}
327\centering
328    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-11-calloc-realloc} }
329    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-11-calloc-realloc} }
330\caption{Speed benchmark chain: calloc-realloc}
331\label{fig:speed-11-calloc-realloc}
332\end{figure}
333
334%speed-12-malloc-realloc-free.eps
335\begin{figure}
336\centering
337    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-12-malloc-realloc-free} }
338    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-12-malloc-realloc-free} }
339\caption{Speed benchmark chain: malloc-realloc-free}
340\label{fig:speed-12-malloc-realloc-free}
341\end{figure}
342
343%speed-13-calloc-realloc-free.eps
344\begin{figure}
345\centering
346    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-13-calloc-realloc-free} }
347    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-13-calloc-realloc-free} }
348\caption{Speed benchmark chain: calloc-realloc-free}
349\label{fig:speed-13-calloc-realloc-free}
350\end{figure}
351
352%speed-14-{m,c,re}alloc-free.eps
353\begin{figure}
354\centering
355    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/speed-14-m-c-re-alloc-free} }
356    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/speed-14-m-c-re-alloc-free} }
357\caption{Speed benchmark chain: malloc-calloc-realloc-free}
358\label{fig:speed-14-malloc-calloc-realloc-free}
359\end{figure}
360
361%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
362%% MEMORY
363%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
364
365\subsection{Memory Micro-Benchmark}
366
367This experiment is run with following two configurations for each allocator.
368The difference between the two configurations is the number of producers and consumers.
369Configuration 1 has one producer and one consumer, and configuration 2 has 4 producers where each producer has 4 consumers.
370
371\noindent
372Configuration 1:
373\begin{description}[itemsep=0pt,parsep=0pt]
374\item[producer (K):]
3751
376\item[consumer (M):]
3771
378\item[round:]
379100,000
380\item[max:]
381500
382\item[min:]
38350
384\item[step:]
38550
386\item[distro:]
387fisher
388\item[objects (N):]
389100,000
390\end{description}
391
392% -threadA :  1
393% -threadF :  1
394% -maxS    :  500
395% -minS    :  50
396% -stepS   :  50
397% -distroS :  fisher
398% -objN    :  100000
399% -consumeS:  100000
400
401\noindent
402Configuration 2:
403\begin{description}[itemsep=0pt,parsep=0pt]
404\item[producer (K):]
4054
406\item[consumer (M):]
4074
408\item[round:]
409100,000
410\item[max:]
411500
412\item[min:]
41350
414\item[step:]
41550
416\item[distro:]
417fisher
418\item[objects (N):]
419100,000
420\end{description}
421
422% -threadA :  4
423% -threadF :  4
424% -maxS    :  500
425% -minS    :  50
426% -stepS   :  50
427% -distroS :  fisher
428% -objN    :  100000
429% -consumeS:  100000
430
431\begin{table}[b]
432\centering
433    \begin{tabular}{ |c|c|c| }
434     \hline
435    Memory Allocator & Configuration 1 Result & Configuration 2 Result\\
436     \hline
437    cfa & Fig. \ref{fig:mem-1-prod-1-cons-100-cfa} & Fig. \ref{fig:mem-4-prod-4-cons-100-cfa}\\
438     \hline
439    dl & Fig. \ref{fig:mem-1-prod-1-cons-100-dl} & Fig. \ref{fig:mem-4-prod-4-cons-100-dl}\\
440     \hline
441    glibc & Fig. \ref{fig:mem-1-prod-1-cons-100-glc} & Fig. \ref{fig:mem-4-prod-4-cons-100-glc}\\
442     \hline
443    hoard & Fig. \ref{fig:mem-1-prod-1-cons-100-hrd} & Fig. \ref{fig:mem-4-prod-4-cons-100-hrd}\\
444     \hline
445    je & Fig. \ref{fig:mem-1-prod-1-cons-100-je} & Fig. \ref{fig:mem-4-prod-4-cons-100-je}\\
446     \hline
447    pt3 & Fig. \ref{fig:mem-1-prod-1-cons-100-pt3} & Fig. \ref{fig:mem-4-prod-4-cons-100-pt3}\\
448     \hline
449    rp & Fig. \ref{fig:mem-1-prod-1-cons-100-rp} & Fig. \ref{fig:mem-4-prod-4-cons-100-rp}\\
450     \hline
451    tbb & Fig. \ref{fig:mem-1-prod-1-cons-100-tbb} & Fig. \ref{fig:mem-4-prod-4-cons-100-tbb}\\
452     \hline
453    \end{tabular}
454\caption{Memory benchmark results}
455\label{table:mem-benchmark-figs}
456\end{table}
457
458Results for memory benchmark are shown in 16 figures, two figures for each of the 8 allocators, one for each configuration.
459Table \ref{table:mem-benchmark-figs} shows the list of figures that contain memory benchmark results.
460
461Each figure has 2 graphs, one for each experiment environment.
462Each graph has following 5 subgraphs that show program's memory usage and statistics throughout the program lifetime.
463
464\begin{itemize}
465\item \textit{\textbf{current\_req\_mem(B)}} shows the amount of dynamic memory requested and currently in-use of the benchmark.
466\item \textit{\textbf{heap}}* shows the memory requested by the program (allocator) from the system that lies in the heap area.
467\item \textit{\textbf{mmap\_so}}* shows the memory requested by the program (allocator) from the system that lies in the mmap area.
468\item \textit{\textbf{mmap}}* shows the memory requested by the program (allocator or shared libraries) from the system that lies in the mmap area.
469\item \textit{\textbf{total\_dynamic}} shows the total usage of dynamic memory by the benchmark program which is a sum of heap, mmap, and mmap\_so.
470\end{itemize}
471
472* These statistics are gathered by monitoring the \textit{/proc/self/maps} file of the process in linux system.
473
474For each subgraph, x-axis shows the time during the program lifetime at which the data point was generated.
475Y-axis shows the memory usage in bytes.
476
477For the experiment, at a certain time in the program's life, the difference between the memory requested by the benchmark (\textit{current\_req\_mem(B)})
478and the memory that the process has received from system (\textit{heap}, \textit{mmap}) should be minimum.
479This difference is the memory overhead caused by the allocator and shows the level of fragmentation in the allocator.
480
481%mem-1-prod-1-cons-100-cfa.eps
482\begin{figure}
483\centering
484    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-cfa} }
485    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-cfa} }
486\caption{Memory benchmark results with 1 producer for cfa memory allocator}
487\label{fig:mem-1-prod-1-cons-100-cfa}
488\end{figure}
489
490%mem-1-prod-1-cons-100-dl.eps
491\begin{figure}
492\centering
493    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-dl} }
494    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-dl} }
495\caption{Memory benchmark results with 1 producer for dl memory allocator}
496\label{fig:mem-1-prod-1-cons-100-dl}
497\end{figure}
498
499%mem-1-prod-1-cons-100-glc.eps
500\begin{figure}
501\centering
502    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-glc} }
503    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-glc} }
504\caption{Memory benchmark results with 1 producer for glibc memory allocator}
505\label{fig:mem-1-prod-1-cons-100-glc}
506\end{figure}
507
508%mem-1-prod-1-cons-100-hrd.eps
509\begin{figure}
510\centering
511    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-hrd} }
512    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-hrd} }
513\caption{Memory benchmark results with 1 producer for hoard memory allocator}
514\label{fig:mem-1-prod-1-cons-100-hrd}
515\end{figure}
516
517%mem-1-prod-1-cons-100-je.eps
518\begin{figure}
519\centering
520    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-je} }
521    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-je} }
522\caption{Memory benchmark results with 1 producer for je memory allocator}
523\label{fig:mem-1-prod-1-cons-100-je}
524\end{figure}
525
526%mem-1-prod-1-cons-100-pt3.eps
527\begin{figure}
528\centering
529    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-pt3} }
530    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-pt3} }
531\caption{Memory benchmark results with 1 producer for pt3 memory allocator}
532\label{fig:mem-1-prod-1-cons-100-pt3}
533\end{figure}
534
535%mem-1-prod-1-cons-100-rp.eps
536\begin{figure}
537\centering
538    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-rp} }
539    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-rp} }
540\caption{Memory benchmark results with 1 producer for rp memory allocator}
541\label{fig:mem-1-prod-1-cons-100-rp}
542\end{figure}
543
544%mem-1-prod-1-cons-100-tbb.eps
545\begin{figure}
546\centering
547    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-1-prod-1-cons-100-tbb} }
548    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-1-prod-1-cons-100-tbb} }
549\caption{Memory benchmark results with 1 producer for tbb memory allocator}
550\label{fig:mem-1-prod-1-cons-100-tbb}
551\end{figure}
552
553%mem-4-prod-4-cons-100-cfa.eps
554\begin{figure}
555\centering
556    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-cfa} }
557    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-cfa} }
558\caption{Memory benchmark results with 4 producers for cfa memory allocator}
559\label{fig:mem-4-prod-4-cons-100-cfa}
560\end{figure}
561
562%mem-4-prod-4-cons-100-dl.eps
563\begin{figure}
564\centering
565    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-dl} }
566    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-dl} }
567\caption{Memory benchmark results with 4 producers for dl memory allocator}
568\label{fig:mem-4-prod-4-cons-100-dl}
569\end{figure}
570
571%mem-4-prod-4-cons-100-glc.eps
572\begin{figure}
573\centering
574    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-glc} }
575    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-glc} }
576\caption{Memory benchmark results with 4 producers for glibc memory allocator}
577\label{fig:mem-4-prod-4-cons-100-glc}
578\end{figure}
579
580%mem-4-prod-4-cons-100-hrd.eps
581\begin{figure}
582\centering
583    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-hrd} }
584    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-hrd} }
585\caption{Memory benchmark results with 4 producers for hoard memory allocator}
586\label{fig:mem-4-prod-4-cons-100-hrd}
587\end{figure}
588
589%mem-4-prod-4-cons-100-je.eps
590\begin{figure}
591\centering
592    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-je} }
593    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-je} }
594\caption{Memory benchmark results with 4 producers for je memory allocator}
595\label{fig:mem-4-prod-4-cons-100-je}
596\end{figure}
597
598%mem-4-prod-4-cons-100-pt3.eps
599\begin{figure}
600\centering
601    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-pt3} }
602    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-pt3} }
603\caption{Memory benchmark results with 4 producers for pt3 memory allocator}
604\label{fig:mem-4-prod-4-cons-100-pt3}
605\end{figure}
606
607%mem-4-prod-4-cons-100-rp.eps
608\begin{figure}
609\centering
610    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-rp} }
611    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-rp} }
612\caption{Memory benchmark results with 4 producers for rp memory allocator}
613\label{fig:mem-4-prod-4-cons-100-rp}
614\end{figure}
615
616%mem-4-prod-4-cons-100-tbb.eps
617\begin{figure}
618\centering
619    \subfigure[Algol]{ \includegraphics[width=0.9\textwidth]{evaluations/algol-perf-eps/mem-4-prod-4-cons-100-tbb} }
620    \subfigure[Nasus]{ \includegraphics[width=0.9\textwidth]{evaluations/nasus-perf-eps/mem-4-prod-4-cons-100-tbb} }
621\caption{Memory benchmark results with 4 producers for tbb memory allocator}
622\label{fig:mem-4-prod-4-cons-100-tbb}
623\end{figure}
624
625%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
626%% ANALYSIS
627%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Note: See TracBrowser for help on using the repository browser.