このエントリーをはてなブックマークに追加
ID 62358
FullText URL
Author
Ban, Yuya
Yamauchi, Toshihiro Graduate School of Natural Science and Technology, Okayama University ORCID Kaken ID publons researchmap
Abstract
Use-after-free (UAF) vulnerabilities, which are abused by exploiting a dangling pointer that refers to a freed memory, execute an arbitrary code. The vulnerability is caused by bug in a program. In particular, it is contained in a large scale program such as browser. HeapRevolver [1] [2], which prohibits freed memory area from being reused for a certain period, has been proposed. HeapRevolver in Windows uses the number of the freed memory areas for prohibiting as a trigger to release the freed memory area. In other words, HeapRevolver uses the number of the freed memory areas as a threshold for releasing. However, when the size of individual freed memory areas is large, the HeapRevolver on Windows increases the memory overhead. In this paper, we propose improved HeapRevolver for Windows considering the size and number of the freed memory areas. Improved HeapRevolver enables to prohibit the reuse of the certain number of the freed memory areas at any time via the size and number of the freed memory areas as a threshold. The evaluation results show that the improved HeapRevolver can prevent attacks that exploiting UAF vulnerabilities. In particular, when the size of individual freed memory areas is small in the programs, it is effective to decrease the attack success rate.
Keywords
Security
Use-After-Free
dangling pointer
memory allocation
Note
© 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
6th International Symposium on Computing and Networking (CANDAR) - Across Practical Development and Theoretical Research, Takayama, JAPAN, Nov. 27-30, 2018
Published Date
2018-11
Publication Title
2018 Sixth International Symposium on Computing and Networking Workshops (CANDARW)
Publisher
IEEE
Start Page
398
End Page
404
ISBN
978-1-5386-9184-7
Content Type
Conference Paper
Related Url
isVersionOf https://doi.org/10.1109/candarw.2018.00080
language
English
Copyright Holders
© 2018 IEEE.
File Version
author
DOI