Intel Shadow Stack For Linux Hits Last-Minute Snag With Issue Raised By Torvalds

[phoronix]

Phoronix: Intel Shadow Stack For Linux Hits Last-Minute Snag With Issue Raised By Torvalds

Intel Shaodw Stack support was submitted for Linux 6.4 at the start of the merge window but now with this two-week merge window drawing to a close, it hasn't been pulled yet and Linus Torvalds raised technical issues with the proposed patches that now jeopardize its arrival this cycle...

https://www.phoronix.com/news/Intel-Shadow-Stack-Issue-6.4

[euduvda]

Yeah, bummer!

Code Owners get mad when other reviewers don't get some glaring issue!
But they are Code Owners for a reason...

Glad the issue was caught, now wait for fixes and we can have another nice feature.

There's time for that or the merge window is already closed?

[AlDunsmuir]

Phoronix: Intel Shadow Stack For Linux Hits Last-Minute Snag With Issue Raised By Torvalds

Intel Shaodw Stack support

https://www.phoronix.com/news/Intel-...tack-Issue-6.4

Should be “Intel Shadow Stack support”

[NotMine999]

Intel programmers get schooled by Linus.

Seriously, Linus has a solid point on "code quality".

Before submitting a pull to Linus, thoughtfully-carefully-slowly walk through your own code...

(yeah I know that process is not taught in schools for like decades now)

...and consider how it will behave on architectures that do not care about what your code does.

Don't take the Microsoft approach "If it compiles...SHIP IT...so we can make the deadline date...and we might fix it later".

The Microsoft approach pre-dates the Facebook approach of "move fast & break stuff".

[schmidtbag]

Intel programmers get schooled by Linus.

Seriously, Linus has a solid point on "code quality".

Before submitting a pull to Linus, thoughtfully-carefully-slowly walk through your own code...

(yeah I know that process is not taught in schools for like decades now)

...and consider how it will behave on architectures that do not care about what your code does.

Don't take the Microsoft approach "If it compiles...SHIP IT...so we can make the deadline date...and we might fix it later".

The Microsoft approach pre-dates the Facebook approach of "move fast & break stuff".

Agreed. On top of everything you said, it's only further emphasized when we're talking about security - something you can't just brush off like its no big deal. You'd think Intel would learn their lesson about that by now...