The world's worst Linux kernel module

https://github.com/davidmalcolm/antipatterns.ko

by pabs3 • 5 years ago

178 12 5 years ago

github.com

pabs3• 5 years ago

The static analysis tools being developed to check this:

https://news.ycombinator.com/item?id=28638885

anonydsfsfs• 5 years ago

Caveat: this is my first ever kernel module, so I may have committed more mistakes than I intended.

lol. Reminds me of the "hypocrite commits" fiasco where researchers at the University of Minnesota submitted kernel patches with what they thought were security vulnerabilities, but because of their inexperience, the first patch they sent was actually valid and had no vulnerabilities whatsoever[1].

[1] https://www-users.cse.umn.edu/~kjlu/papers/full-disclosure.p...

IshKebab• 5 years ago

It doesn't sound like it was "because of their inexperience" but rather because determining whether Linux C code is buggy is really hard. That's kind of the whole point.

forumninja• 5 years ago

GitHub Copilot will learn this code and suggest it. Bravo!

recov• 5 years ago

Potential idea to somehow have 51% of the programs on github be malicious to taint copilot?

riboflavin123• 5 years ago

I've never worked with the C/make pipeline. Can anyone explain why this commit is getting so many votes?

cratermoon2• 5 years ago

Is it normal kernel style to use if blocks without braces if it's only one line? See https://github.com/davidmalcolm/antipatterns.ko/blob/main/bu...

zamadatix• 5 years ago

From https://www.kernel.org/doc/html/v4.10/process/coding-style.h...

"Do not unnecessarily use braces where a single statement will do."

cratermoon2• 5 years ago

I wonder if anyone has done an analysis of kernel code to find all the cases where brace-less statements introduced bugs.

ToFab123• 5 years ago

Apples famous "gotofail" SSL Security Bug was due to this

https://embeddedgurus.com/barr-code/2014/03/apples-gotofail-...

clarge1120• 5 years ago

Mistakenly read this as “World’s FIRST…”

The world's worst Linux kernel module

Related Stories