The 2014 Kernel Summit

Day 1

Discussions from the invitation-only day include:

• Error handling for I/O memory management units: what do we do when DMA goes bad?

• Stable tree maintenance: what is the condition of the stable tree and how can things be done better?

• The state of linux-next: how is the kernel's integration tree doing and what could be improved?

• Kernel tinification: how can we make the kernel work on the smallest of systems?

• What makes Linus happy (or not)?: this year's guide to the care and feeding of the kernel's benevolent dictator.

• Performance regressions: how is the development community doing in the eternal battle against the introduction of performance problems?

• Kernel self tests: creating a simple set of self tests to quickly determine if a given kernel build runs properly.

• Two sessions on review: how do we get better review of new kernel ABIs? And, for that matter, how can we get more reviewers in general?

• One year of Coverity work: using defect information from the Coverity scanner to improve the kernel.

And, of course, there is the obligatory group photo:

Day 2

Unfortunately, LWN was unable to attend the workshop sessions. Attendees to various sessions have posted reports; we are collecting them here.

• Device model/driver properties/ACPI_DSD session, posted by Darren Hart

• Testing by Paul McKenney. See also a followup by Shuah Khan.

• printk() (Paul McKenney)

• mount (Paul McKenney)

• Power management dependencies (Laurent Pinchart).

• Power-aware scheduling (Morten Rasmussen).

Day 3

Many of the third-day sessions took the form of tutorials for tools like Coccinelle — valuable sessions, but difficult to write up in this format. Beyond that, though, this day also featured:

• The kernel.org update with an emphasis on the new two-factor authentication system.

• The power-aware scheduling miniconference; a report from Morten Rasmussen on the latest developments in power-aware scheduling.

• A report from the networking miniconference; a whirlwind summary of what is happening in the networking subsystem.