 Ian Goldberg
|
13c2afb818
Make compressed DPFs the default
|
2 years ago |
|
Ian Goldberg
|
63ef055b2e
A mode that replicates the tests from the Duoram paper
|
2 years ago |
|
Ian Goldberg
|
f3cb64ea1b
Add the ability to reset the Lamport clock
|
2 years ago |
|
Ian Goldberg
|
a09fb25516
Some more missing yield()s
|
2 years ago |
|
Ian Goldberg
|
2b29022d8c
Writes (= reads + updates)
|
2 years ago |
|
Ian Goldberg
|
e9fa1a0289
Simultaneous independent reads and updates
|
2 years ago |
|
Ian Goldberg
|
3565cedf03
Don't separately account for halftriples that are created as a byproduct of RDPF creation
|
2 years ago |
|
Ian Goldberg
|
c39726c869
Add a third simultaneous read
|
2 years ago |
|
Ian Goldberg
|
121f757edb
The server wasn't appending in -a mode
|
2 years ago |
|
Ian Goldberg
|
212a0d420e
Demonstrate independent simultaneous reads
|
2 years ago |
|
Ian Goldberg
|
c6682b9553
Make it compile on Ubuntu 20.04 (g++ 9.4.0)
|
2 years ago |
|
Ian Goldberg
|
0fa58457c5
Multithread RDPF creation when it makes sense
|
2 years ago |
|
Ian Goldberg
|
4d3fc01b66
Be able to set the number of CPU threads for each processing thread (independently) during preprocessing
|
2 years ago |
|
Ian Goldberg
|
33a85d89ca
Simplify ParallelEval API
|
2 years ago |
|
Ian Goldberg
|
b6abfb1b85
Duoram reads and updates are now multithreaded
|
2 years ago |
|
Ian Goldberg
|
92b29fc996
Templatize additive and XOR shared indices into Duoram Shapes
|
2 years ago |
|
Ian Goldberg
|
45357b1778
Add -a mode for preprocessing with appending
|
2 years ago |
|
Ian Goldberg
|
3999846d2c
Only print precomputed value stats if they're more than 0
|
2 years ago |
|
Ian Goldberg
|
2c32812a93
A little more templatization
|
2 years ago |
|
Ian Goldberg
|
94fcf8f8b0
Parallel RDPF evaluator
|
2 years ago |
|
Ian Goldberg
|
0a53ee11af
Start preparing the tupletime online mode for parallelism
|
2 years ago |
|
Ian Goldberg
|
e09f4e3f3b
Keep separate track of the number of threads we can use for computation and for communication
|
2 years ago |
|
Ian Goldberg
|
0e1836d604
Preventing the threads from writing into memory in the inner loop yields the expected speedup
|
2 years ago |
|
Ian Goldberg
|
b899b24b6d
Parallel StreamEval
|
2 years ago |
|
Ian Goldberg
|
af5c08e306
Do something sensible in online mode when asked for multiple threads
|
2 years ago |
|
Ian Goldberg
|
e127dd94f0
Be clear about MPCTIO::{rdpftriple,rdpfpair}
|
2 years ago |
|
Ian Goldberg
|
1216a94599
Avoid default captures in lambda expressions
|
2 years ago |
|
Ian Goldberg
|
0ef40dbb69
Capture the loop counter by copy, not reference
|
2 years ago |
|
Ian Goldberg
|
5ea8872081
Some more missing yields
|
2 years ago |
|
Ian Goldberg
|
dfb63bdb5d
Add yield()s in the right places
|
2 years ago |
