Udaya Seshua
Udaya Seshua has a total of 2 co-authors. Here are the top co-authors by number of shared publications:
- Bart Vermeulen
NXP Semiconductors
1 shared publication - Nagaraju Bussa
Philips Research, Philips Innovation Campus, # 1, Murphy Road, Ulsoor 560008, Bangalore, India
1 shared publication
Latest Publications
- 1 January 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 498-503
A Run-Time Memory Protection Methodology
In this paper we present a novel methodology to help debug memory corruption errors during application debug. In this methodology an optimal balance between hardware and software instrumentation is chosen to check at run-time all memory accesses made by an application. To achieve this balance a set of benchmark applications is first analyzed to determine their memory access patterns. The analysis results are used to make our approach low-cost both from a software performance penalty and a hardware area point-of-view. Experimental results show that our innovative approach typically requires less than 2% of a CPU in silicon area for a less than 1% run-time performance overhead. Our method is both low-cost and applicable to high performance microprocessors as well as time-constrained embedded systems.
- 1 January 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 498-503
A Run-Time Memory Protection Methodology
In this paper we present a novel methodology to help debug memory corruption errors during application debug. In this methodology an optimal balance between hardware and software instrumentation is chosen to check at run-time all memory accesses made by an application. To achieve this balance a set of benchmark applications is first analyzed to determine their memory access patterns. The analysis results are used to make our approach low-cost both from a software performance penalty and a hardware area point-of-view. Experimental results show that our innovative approach typically requires less than 2% of a CPU in silicon area for a less than 1% run-time performance overhead. Our method is both low-cost and applicable to high performance microprocessors as well as time-constrained embedded systems.