Improving performance of rav1d video decoder

Oh this stuff is what’s prompting the ffmpeg Twitter account to make a stand against Rust https://x.com/ffmpeg/status/1924137645988356437?s=46

I generally trust rbultje to benchmark correctly but the ravid tracking ticket has multithread numbers across multiple platforms that don't show that big a difference.

https://github.com/memorysafety/rav1d/issues/1294

Is that explained in replies? I only see the original tweet as I'm not logged in.