Linux Plumbers Conf 2013

Summarising the current status of IVI on Wayland

Automotive diagnostics under Linux — what is missing, what are the next steps?

Making BlueZ fit for the automotive industry.

Update where we are with the ‘New Challenges for Linux Network Support’ work.

Efficient device management for in vehicle infotainment with consideration for smart devices connectivity

Overview of netmap and derived tools and discussion on how they can help improving network virtualization.

This talk shares our experience with Open vSwitch from the past year, both as users and active developers.

Discussion around the use of hardware assisted filtering capabilities found in modern networking hardware to offload the filtering of packets to the hardware.

Explore ways to extend and merge existing networking mechanisms to address the needs of virtual network infrastructure

A look at the Contrail vRouter kernel module and steps taken to enhance forwarding performance

5s boots delivered, today.

Outline plans for future Upstart development.

Discussing the new cgroups userspace.

The Gummiboot Boot Loader

A quick introduction of systemd-nspawn for developers

Update on the Boot Loader Specification

Automatically detecting Linux partitions via GPT type IDs

A brief introduction to what TCP repair is.

A look at performance issues found and addressed in VXLAN and NVGRE tunnel protocols, and steps we are exploring to further improve performance.

This talk will give an overview of what the Intel DPDK is and what it is not.

The implementation of the new Network Service Headers IETF draft in both Linux and Open vSwitch.

Locator/ID Separation Protocol implementation in Linux and Open vSwitch.

How can we best combine the benefits of Open vSwitch and other kernel components like Netfilter in a single system?

Discuss current and future switch features that are or will be embedded into the NIC as well as how these features could be managed in a Linux environment.

A look on the common dilemma for using Linux as the foundation of a networking OS

Implementing Per CPU atomics for userspace

Discuss the current status of lock elision in Linux

Present state of full dynticks, current limitations and what remains to be done

New Scheduler support allowing cooperative scheduling on top of a a 1:1 threading model.

Scalability Issues in Linux Kernel

Discussion of userspace RCU new APIs and use-cases

Allow physical memory to change which NUMA node it belongs to at runtime to improve
long running virtual machine performance.

Scaling the userspace sysfs interface for physical memory configuration for large memory systems

vmsplice() could move (rather can copy) pages between processes, but performance would be greatly improved if this supported THP.

We need mechanisms that do for update-mostly workloads that what RCU does for read-mostly workloads.

Discussion of issues around enabling, disabling, and removing Virtual Functions

Discussion of synchronization issues with device addition and removal

Discussion on how to integrate the native PCIe hotplug (pciehp) with ACPI-based PCI hotplug (acpiphp).

Discussion on the status of support for ACPI-based hotplug of devices in the Linux kernel

Discussion on current status, issues and challeges faced, and observations made in converting legacy pm_ops to dev_pm_ops

Discussion on interfaces to use for communicating the needs of user space related to power management to the kernel.

Update on the current progress and solving/discussing current problems for x86_64/i586.

Using the Clang Integrated Assembler to build the Linux Kernel on x86/ARM.

VLAIS is a gcc extension which isn’t supported by clang. The unnecessary use of VLAIS in the Linux Kernel prevents the kernel from being built with clang or any other C standard compliant compiler.

Problems observed while making Android userland build with clang – and their solutions

Linux on the Hexagon DSP with LLVM

Current state of unified secure boot code and plans for getting the architecture override into UEFI

What changes do we need to make to the kernel to support the Secure Boot security model?

Discuss how to make kexec/kdump work with secureboot

ACPI vs Device Tree – Moving Forward: Discussion on device enumeration methods

Discuss plans to ensure systemd-based distros continue to work on lxc.

Present the device namespace implementation by cellrox and discuss upstreaming.

Overview of what features to expect for LXC 1.0 and hints on how to help us get there.

Ever wanted to run Android in a container or run a Linux distro in a container on Android?

Current state and expected timeline for completion of unprivileged containers

A tour over checkpoint-restore project features and plans followed by a discussion on integrating this with LXC.

Redesign of container management and its use at Google.

Now that the kernel supports both LXC and OpenVZ do we really need two different tool sets? If not, how should we unify the user visible container control plane.

Use Cases for Containers in OpenStack

State of the Common Display Framework and roadmap to mainline

A look at the issues preventing better usage of media decode and 2D composition hardware

Splitting DRM and KMS nodes to allow advanced access-control and sharing between subsystems.

Developers creating high-end graphics applications on open platforms have stone age tools, but some have learned to use them well.

The progress of open source graphics on ARM in the last year.

Rebuilding Debian with Clang

The Clang Static Analyzer can be used to look for various specific issues in the Linux kernel. What other checkers can be added specifically for the Linux kernel?

Discussion around issues with a KMS based HWComposer, the Atomic Display Framework, and other KMS extentions

Discussion on dma-buf fences and Android sync driver

Discussion around how phasing out the fbdev api might affect Android

Discussions on upstreaming ION-like functionality around dma-buf allocation and device constraint solving.

Persistent memory devices are coming, so we need a plan on how to take advantage of them through the file system.

An interface to offload bulk data copying

Setting up a file system over a dm-thin device, although easy, can have poor performance if not done properly and can even cause data loss. This talk will walk through the dm-thin device configuration to be used by xfs filesystem to achieve good performance and avoid data loss.

Making the idle framework and the scheduler to collaborate

Energy-efficiency benefits of extended idle and RCU callback offloading.

When packing tasks can improve power efficiency and scheduling latency

Integrating and extending the existing power management frameworks with the scheduler.

Discussions around KDBUS and if it is sufficient to replace the in-kernel binder driver

Discussions around what the ConfigFS gadget driver needs to replace the Android Gadget driver

Discussions on getting Android netfilter changes merged upstream

SCSI midlayer scalability work done by Christoph Hellwig

Shingled drives need changes to file and storage to get the best efficiency.

A dedicated metadata cache would help distributes file systems and in general should reduce per file system efforts to cache metadata.

Reserved time to discuss new ideas in an open conversation.

The Design and Implementation of a virtual Trusted Platform Module in qemu

Overview of the current work to fix the filesystem freeze API

Passing non-PCI devices into guests and allow for user space device drivers

Discussing VirtIO on Xen

EFI support in Xen – the current state of development