Reducing Minor Page Fault Overheads through Enhanced Page Walker

نویسندگان

چکیده

Application virtual memory footprints are growing rapidly in all systems from servers down to smartphones. To address this demand, system integrators incorporating ever larger amounts of main memory, warranting rethinking management. In current systems, applications produce page fault exceptions whenever they access regions that not backed by a physical page. As application grow, induce more and minor faults. Handling each can take few thousands CPU cycles blocks the till OS kernel finds free frame. These faults be detrimental performance when their frequency occurrence is high spread across runtime. Specifically, lazy allocation-induced increasingly impacting performance. Our evaluation several workloads indicates an overhead due as 29% execution time. article, we propose mitigate problem through hardware, software co-design approach. first parallelize portions allocation run ahead time separate thread. Then Minor Fault Offload Engine (MFOE), per-core hardware accelerator for handling. MFOE equipped with pre-allocated frame table it uses service fault. On fault, quickly picks table, makes entry TLB, updates satisfy The pre-allocation tables periodically refreshed background thread, which also data structures account handled We evaluate gem5 architectural simulator modified Linux running on top simulated containing accelerator. results show improves average critical path handling latency 33× tail 51×. Among evaluated applications, observed improvement runtime 6.6%.

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

On Reducing Dynamic Web Page Construction Times

Many web sites incorporate dynamic web pages to deliver customized contents to their users. However, dynamic pages result in increased user response times due to their construction overheads. In this paper, we consider mechanisms for reducing these overheads by utilizing the excess capacity with which web servers are typically provisioned. Specifically, we present a caching technique that integ...

متن کامل

Reducing Parallel Overheads Through Dynamic Serialization

If parallelism can be successfully exploited in a program, signi cant reductions in execution time can be achieved. However, if sections of the code are dominated by parallel overheads, the overall program performance can degrade. We propose a framework, based on an inspector-executor model, for identifying loops that are dominated by parallel overheads and dynamically serializing these loops. ...

متن کامل

URL-Enhanced Adaptive Page-Refresh Models

We study the refresh model required to keep an up to date copy of a web page. This has applications for more efficient and frequent crawling, in the case of search engines, and for higher hit rates for proxy servers with pre-fetching. Two main models have been proposed namely the uniform refresh model (Cho and Garcia-Molina, 2000) and the adaptive page refresh model (Edwards et al., 2001), with...

متن کامل

Web Page Prediction Enhanced with Confidence Mechanism

In this work we comparatively present and evaluate different prediction techniques used to anticipate and prefetch web pages and files accessed via browsers. The goal is to reduce the delays necessary to load the web pages and files visited by the users. We have included into our analysis Markov chains, Hidden Markov Models and graph algorithms. We have enhanced all these predictors with confid...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: ACM Transactions on Architecture and Code Optimization

سال: 2022

ISSN: ['1544-3973', '1544-3566']

DOI: https://doi.org/10.1145/3547142