Re:Performance patches in Go 1.11
by kodablah on Friday June 22, 2018 @10:17AM
(#17374106)
The ModInverse and GCD improvements will be very welcome in a demo I just wrote for trustless card playing[0].
I'm beginning to wonder if there will become more value in "lifting" binaries to LLVM IR, applying all opts to them, and re-compiling them. McSema does this[1] albeit you have to setup IDA and what not. Has anyone applied this to Go programs and measured its benefits for their specific use cases?
I'm beginning to wonder if there will become more value in "lifting" binaries to LLVM IR, applying all opts to them, and re-compiling them. McSema does this[1] albeit you have to setup IDA and what not. Has anyone applied this to Go programs and measured its benefits for their specific use cases?
0 - https://github.com/cretz/go-mental-poker 1 - https://github.com/trailofbits/mcsema