Should the browser use all available memory?
jmmv.dev
jmmv.dev
I've always thought this should be the standard way of doing caching. It's also explained at length in this article by the author of Varnish: https://varnish-cache.org/docs/trunk/phk/notes.html. Contrary to what TFA is saying, performance under "normal" use would be the same: the data is in RAM. It doesn't change that it is backed by a file or not.
And when I said performance would suffer, I was not implying it'd be terrible. I was trying to say that the extra open/close system calls are overhead that you don't directly pay with direct memory accesses. But... the overall impact of those on system performance would likely be invisible. Also clarified this.
No suitable to the OP's purpose: have reserved "optional" memory you could use for caching.
"The MADV_FREE operation can be applied only to private anonymous pages (see mmap(2)). In Linux before version 4.12, when freeing pages on a swapless system, the pages in the given range are freed instantly, regardless of memory pressure."
You mean Skype running on top of chrome?
If I were making the rules, I would use some conservative browser with a lot of non-standard settings, like Palemoon, and limit it to a gigabyte of real memory. If the website survives, then we can talk.