Microsoft Security Researcher Proposes Unprivileged Chroot For Linux

[phoronix]

Phoronix: Microsoft Security Researcher Proposes Unprivileged Chroot For Linux

Security researcher and Microsoft engineer Mickaël Salaün is proposing unprivileged chroot support for the Linux kernel...

https://www.phoronix.com/scan.php?page=news_item&px=2021-Unprivileged-Chroot

[rmnscnce]

No, Jim. This is not a sign of Windows shifting to use the Linux kernel.

At least not yet

[macemoneta]

How would this be different from proot?

https://proot-me.github.io/

[tildearrow]

Every thread with "Microsoft" on it:

*looks at comments*

*EEE-related comment*

[kpedersen]

How would this be different from proot?

Microsoft's prootEEE has a much nicer ring to it.

[CTTY]

How would this be different from proot?
https://proot-me.github.io/

Isnt proot using the mount namespace? If so, the difference is explained in the quote in the article.

[sindr]

How would this be different from proot?
https://proot-me.github.io/

PRoot relies on ptrace which has negative performance implications to say the least.

[latalante]

I have been using chroot since 2013 as an unprivileged user.
https://github.com/sabotage-linux/sa...42ace4c02244a6 bb005e069d037b

[skeevy420]

I apologize ahead of time for this

I have been using chroot since 2013 as an unprivileged user.
https://github.com/sabotage-linux/sa...42ace4c02244a6 bb005e069d037b

I know times have been tough lately, COVID is rampant, money is hard to come by, the good people on both sides, but I've had the best people, the best, working on a new unprivileged chroot solution. It chroots 17x faster than any other existing chroot implementation and it takes care of all the White Privilege in the process. I call it The Super Duper Chroot. It's the BEST unprivileged chroot method and nothing comes close. The Russians won't know what happened when they try to mess with our systems and get trapped inside the Super Duper Chroot full of Fake News.

[macemoneta]

PRoot relies on ptrace which has negative performance implications to say the least.

Sure, but only when root operations are intercepted. The rest of the time, performance isn't impacted.