News for hackers. Stuff that matters.

A comment on Meta Is Using the Linux Scheduler Designed for Valve's Steam Deck on Its Servers. Parent comment

Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by Arainach on Tuesday December 23, 2025 @02:00PM (#46368195)
ACLs in Linux were tacked on later; not everything supports them properly. They were built into Windows NT from the start and are used consistently across kernel and userspace, making them far more useful in practice.

Also, as far as I know Linux doesn't support DENY ACLs, which Windows does.

[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by onraglanroad on Tuesday December 23, 2025 @02:25PM (#46368455)
Yes it does.
[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by 112233 on Tuesday December 23, 2025 @02:46PM (#46368664)
since when?
[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by onraglanroad on Tuesday December 23, 2025 @02:55PM (#46368787)
Since some of us could be bothered reading docs. Give it a try and see how it works out for you.
[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by 112233 on Tuesday December 23, 2025 @03:52PM (#46369340)
Some of us can! I certainly enjoy doing it, and according to "man 5 acl" what you assert is completely false. Unless you have a particular commit or document from kernel.org you had in mind?
[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by opello on Tuesday December 23, 2025 @05:53PM (#46370445)
> Each of these characters is replaced by the - character to denote that a permission is absent in the ACL entry.

Wouldn't the o::--- default ACL, like mode o-rwx, deny others access in the way you're describing?

[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by 112233 on Wednesday December 24, 2025 @03:03AM (#46373484)
See 6.2.1 of RFC8881, where NFSv4 ACLs are described. They are quite similar to Windows ACLs.

Here is kernel dev telling they are against adding NFSv4 ACL implementation. The relevant RichAcls patch never got merged: https://lkml.org/lkml/2016/3/15/52

[ Parent | on HN ]
Re:Meta Is Using the Linux Scheduler Designed for Valve's St... by opello on Wednesday December 24, 2025 @10:55AM (#46376624)
https://www.rfc-editor.org/rfc/rfc8881#section-6.2.1

I see what I misunderstood, even in the presence of an ALLOW entry, a DENY entry would prohibit access. I am familiar with that on the Windows side but haven't really dug into Linux ACLs. The ACCESS CHECK ALGORITHM[1] section of the acl(5) man page was pretty clear, I think.

[1] https://man7.org/linux/man-pages/man5/acl.5.html#ACCESS_CHEC...

[ Parent | on HN ]