PROJECT TITLE :

Hardware Support for Safe Execution of Native Client Applications

ABSTRACT:

Over the past few years, there was vast growth in the world of the net browser as an applications platform. One example of this trend is Google's Native Shopper (NaCl) platform, that is a software-fault isolation mechanism that permits the running of native x86 or ARM code on the browser. One in every of the protection mechanisms utilized by NaCl is that every one branches must jump to the start of a legitimate instruction. So as to realize this criteria though, all return instructions are replaced by a particular branch instruction sequence, which we tend to decision NaCl returns, that are absolute to come back to a legitimate instruction. But, these NaCl returns lose the advantage of the highly correct come back-address stack (RAS) in exchange for the less accurate indirect branch predictor. In this paper, we propose a NaCl-RAS mechanism that can establish and accurately predict seventy six.9 on average compared to the 39.5 of a ancient BTB predictor.