Abstract:Indirect-branch prediction is becoming more important for modern processors as more programs are written in object-oriented languages. Previous hardware-based indirect-branch predictors generally require significant hardware storage or use aggressive algorithms which make the processor front-end more complex. In this paper, we propose a fast and cost-efficient indirect-branch prediction strategy, called Target Address Pointer (TAP) Prediction. TAP Prediction reuses the history-based branch direction predictor to detect occurrences of indirect branches, and then stores indirect-branch targets in the Branch Target Buffer (BTB). The key idea of TAP Prediction is to predict the Target Address Pointers, which generate virtual addresses to index the targets stored in the BTB, rather than to predict the indirect-branch targets directly. TAP Prediction also reuses the branch direction predictor to construct several small predictors. When fetching an indirect branch, these small predictors work in parallel to generate the target address pointer. Then TAP prediction accesses the BTB to fetch the predicted indirect-branch target using the generated virtual address. This mechanism could achieve time cost comparable to that of dedicated-storage-predictors, without requiring additional large amounts of storage. Our evaluation shows that for three representative direction predictors-Hybrid, Perceptrons, and O-GEHL-TAP schemes improve performance by 18.19%, 21.52%, and 20.59%, respectively, over the baseline processor with the most commonly-used BTB prediction. Compared with previous hardware-based indirect-branch predictors, the TAP-Perceptrons scheme achieves performance improvement equivalent to that provided by a 48KB TTC predictor, and it also outperforms the VPC predictor by 14.02%.

Indirect Branch Prediction in Dynamic Binary Translator

Improve Indirect Branch Prediction with Private Cache in Dynamic Binary Translation

SPC-Indexed Indirect Branch Hardware Cache Redirecting Technique in Binary Translation

SPTU: Improving Dynamic Binary Translation through Software Prediction with Target Updating

Return Instruction Analysis and Optimization in Dynamic Binary Translation

DTT: program structure-aware indirect branch optimization via direct-TPC-table in DBT system.

SWIP Prediction: Complexity-Effective Indirect-Branch Prediction Using Pointers

Compiler-Assisted Value Correlation for Indirect Branch Prediction

Vbiw: Optimizing Indirect Branch In Dynamic Binary Translation

A General Low-Cost Indirect Branch Prediction Using Target Address Pointers

CVP: an energy-efficient indirect branch prediction with compiler-guided value pattern.

TAP prediction: Reusing conditional branch predictor for indirect branches with Target Address Pointers

Spire: Improving Dynamic Binary Translation Through Spc-Indexed Indirect Branch Redirecting

Branch Prediction Based on Branch History Length Adaptation

Realization of Branch Prediction and Value Prediction in VLIW

A Hybrid Value Correlation Based Indirect Jump Prediction

An Energy-Efficient Branch Prediction Technique Via Global-History Noise Reduction

A Hybrid Branch Prediction Approach for High- Performance Processors

Branch Prediction Based on Dynamic Polarity Transformation

Transferring Optimize Technology in Dynamic Binary Translation

Improving branch prediction by combining perceptron with learning vector quantization neural network in embedded processor