UNIX systems for modern architectures: symmetric multiprocessing and caching for kernel programmers
UNIX systems for modern architectures: symmetric multiprocessing and caching for kernel programmers
Surpassing the TLB performance of superpages with less operating system support
ASPLOS VI Proceedings of the sixth international conference on Architectural support for programming languages and operating systems
A new page table for 64-bit address spaces
SOSP '95 Proceedings of the fifteenth ACM symposium on Operating systems principles
Reducing TLB and memory overhead using online superpage promotion
ISCA '95 Proceedings of the 22nd annual international symposium on Computer architecture
ACM Computing Surveys (CSUR)
Performance analysis using the MIPS R10000 performance counters
Supercomputing '96 Proceedings of the 1996 ACM/IEEE conference on Supercomputing
ACM SIGARCH Computer Architecture News
Practical, transparent operating system support for superpages
ACM SIGOPS Operating Systems Review - OSDI '02: Proceedings of the 5th symposium on Operating systems design and implementation
ACM SIGOPS Operating Systems Review
Practical, transparent operating system support for superpages
OSDI '02 Proceedings of the 5th symposium on Operating systems design and implementationCopyright restrictions prevent ACM from being able to make the PDFs for this conference available for downloading
Advanced non-distributed operating systems course
ACM SIGCSE Bulletin
Proceedings of the 2006 workshop on Memory system performance and correctness
ATEC '99 Proceedings of the annual conference on USENIX Annual Technical Conference
Using 4KB page size for virtual memory is obsolete
IRI'09 Proceedings of the 10th IEEE international conference on Information Reuse & Integration
SpecTLB: a mechanism for speculative address translation
Proceedings of the 38th annual international symposium on Computer architecture
Performance characteristics of explicit superpage support
ISCA'10 Proceedings of the 2010 international conference on Computer Architecture
Exploiting semantics of virtual memory to improve the efficiency of the on-chip memory system
Euro-Par'12 Proceedings of the 18th international conference on Parallel Processing
CoLT: Coalesced Large-Reach TLBs
MICRO-45 Proceedings of the 2012 45th Annual IEEE/ACM International Symposium on Microarchitecture
Efficient virtual memory for big memory servers
Proceedings of the 40th Annual International Symposium on Computer Architecture
Hi-index | 0.00 |
Many commercial microprocessor architectures support translation lookaside buffer (TLB) entries with multiple page sizes. This support can be used to substantially reduce the overhead introduced by TLB misses incurred when the processor runs an application with a large working set. Applications are currently not able to take advantage of this hardware feature because most commercial operating systems support only one page size. In this paper we present a design that provides general purpose operating system support that allows applications to use multiple page sizes. The paper describes why providing this support is not as simple as it first seems. If not designed carefully, adding this support will require significant modifications and add a lot of overhead to the operating system. The paper shows how our approach simplifies the design without sacrificing functionality and performance. In addition, applications need not be modified to make use of this feature. The design has been implemented on IRIX 6.4 operating system running on the SGI Origin platform. We include some performance results at the end to show how our design allows applications to take advantage of large pages to gain performance improvements.