Transparent Checkpoint-Restart for Hardware-Accelerated 3D Graphics
Northeastern University
arXiv:1312.6650 [cs.OS], (23 Dec 2013)
@article{2013arXiv1312.6650K,
author={Kazemi}, S. and {Garg}, R. and {Cooperman}, G.},
title={"{Transparent Checkpoint-Restart for Hardware-Accelerated 3D Graphics}"},
journal={ArXiv e-prints},
archivePrefix={"arXiv"},
eprint={1312.6650},
primaryClass={"cs.OS"},
keywords={Computer Science – Operating Systems},
year={2013},
month={dec},
adsurl={http://adsabs.harvard.edu/abs/2013arXiv1312.6650K},
adsnote={Provided by the SAO/NASA Astrophysics Data System}
}
A mechanism for transparent GPU-independent checkpoint-restart of 3D graphics is described. The approach is based on a record-prune-replay paradigm: all OpenGL calls relevant to the graphics driver state are recorded; calls not relevant to the internal driver state as of the last graphics frame prior to checkpoint are discarded; and and the remaining calls are replayed on restart. A previous approach, based on a shadow device driver, required more than 78,000 lines of OpenGL-specific code. In contrast, the new approach, based on record-prune-replay, is implemented in just 4,000 lines of code. The speed of this approach varies between 80 per cent and nearly 100 per cent of the speed of the native hardware acceleration for OpenGL 1.5, as measured when running ioquake3 under Linux.
December 24, 2013 by hgpu
