Christoffer will be joining us once again, to continue the technical discussion on his document. Just to remind you, here is the link to his doc in github https://github.com/challfre/wg-lfscs/blob/mai

For this week's call, we are privileged to have Wenhui and Ruide from bytedance/TikTok who will help to clarify business requirements for Linux based safety critical applications. In particular, Linux

Hi, Given the amount of private requests, I am sharing the link to the database of kernel configurations for safety with the entire group https://docs.google.com/spreadsheets/d/1oiOmWTr94M7vP3sisFyBMf

Adding others, who may be interested in this invited talk.

Gab, now I am confused. Isn't it safety architecture? This also matches the WG name, as well as the definitions below. Regards Elana

Fully agreed. At this point – awareness is fundamentally important, but I doubt we can yet make any final decisions, where we draw the lines. Thanks, Gab, for sharing the makefile – this helps to high

Thanks, Robert, for your comment. This is all fine and true. And as noted, at the end of the day, the drivers end up calling the ioctl system interface. But then my question ends up being, for qualifi

Hi, To add to yesterday's discussion – see https://www.kernel.org/doc/html/latest/process/botching-up-ioctls.html This is basic common sense advice for any usage of ioctls. There is a nice summary of

Agreed. Lukas, the oversight could have been handled by a short private query directly to Eli and Gab.

Proposed Agenda: In next week's workshop we will have a session to wrap up the reference process and its application to analysis of kernel verification process. For today's call, I will present the er

Thanks, Gab. This code is extremely interesting. The memory_failure() code is indeed dependent on the underlying architecture for the actual implementation. And is extremely interesting, from safety p

That's fine. Except that as you see, there are legitimate hardware dependencies in this specific use case, which also limit the benefit of Linux. But as you say, it is a good first example. Regards El

Lukas, now I am even more confused. The subject line of this thread is "assumed system context for HW error reporting and reaction". In your mail below, you have reduced the scope to "A continued oper

In case there is complete control by C1 (from Chris' outline), our practical experience is that we can not rely on any software-based information on the system state leading to the error.

Hi Tony, Lukas, Tony – welcome, thanks for raising this, the topic is indeed interesting and relevant, as outlined by Lukas below. I'm not sure who is behind this paper, but the title ("A Decade of Wa