Hello.

I would like to give a talk about KVA allocator in the kernel and about
improvements i have done.

See below the presentation:

ftp://vps418301.ovh.net/incoming/Reworking_of_KVA_allocator_in_Linux_kernel.pdf

Thank you in advance!

--
Vlad Rezki

Execute only memory can protect from attacks that involve reading executable code. This feature already exists on some CPUs and is enabled for userspace.

This talk will explain how we are working on creating a virtualized “not-readable” permission bit for guest page tables for x86 and the impact to the kernel. This bit can be used to create execute-only memory for userspace programs as done...

The Red-Black tree and Radix tree are used in many places in the kernel to store ranges. Both of these trees have drawbacks when used for ranges. The Red-Black tree requires writing your own insertion & search code. It is also designed with the assumption that memory accesses are cheap, which is no longer true. The Radix tree performs acceptably well when ranges are aligned to a power of 2,...

Implementing safety-critical systems usually requires adhering to meticulously defined development processes that specify how code is supposed to be developed, integrated and reviewed, driven by the assumption that a disciplined approach leads to reliably high quality. While known to produce code that can satisfy the highest quality standards, Linux kernel development does not follow such...

DRM is merging new drivers at a brisk pace, and with lima and panfrost to support ARM Mali GPUs the last obvious gap in not yet reverse-engineered hardware is getting closed. Plus new features, more contributors, more patches - in general upstream graphics is as healthy as it's never been before.

Time for some celebratory drinks, except this talk will be none of that. Now that we've achieved...

Storage hardware with built-in “inline” encryption support is becoming increasingly common, especially on mobile SoCs running Android; it's also now part of the UFS and eMMC standards. These devices en/decrypt data between the application processor and disk without generating disk latency or cpu overhead. Inline encryption hardware can be programmed to hold multiple encryption keys...

There is a lot of similar and duplicated code in architecture specific
bits of memory management.

For instance, most architectures have

#define PGALLOC_GFP (GFP_KERNEL | __GFP_ZERO)

for allocating page table pages and many of them use similar, if not
identical, implementation of pte_alloc_one*().

But that's only the tip of the iceberg.

There are several early_alloc() or similarily...

In the linux kernel, most operations affecting a process's address space are protected by by mmap_sem (a per-process read-write semaphore).

This simple design is increasingly a problem for multi-threaded applications, and often causes threads that operate on separate parts of their address space to end up blocking on each other due to false sharing issues - mmap_sem only supports locking the...

Big systems are becoming more common these days. Having thousands of CPUs is
no more a dream and some applications are attempting to spread over all
these CPUs by creating threads.
This leads to contention on the mm->mmap_sem which is protecting the memory
layout shared by these threads.
There were multiple attempts to get rid of the mmap_sem's contention or the
mmap_sem itself, Speculative...

Background

Memory pressure is inevitable in many environments. A decade size survey[1] of DRAM to CPU ratio in virtual machines and physical machines for data centers implies that the pressure will be even more common and severe. As an answer to this problem, heterogeneous memory systems utilizing recently evolved memory devices such as non-volatile memory along with the DRAM are...

With virtualization being the key to the success of cloud computing, Intel's
introduction of the Scalable IO Virtualization (SIOV) aims to further the cause by shifting the creation of assignable virtual devices from hardware to a more software assisted approach. Using SIOV, a device resource can be mapped directly to guest or other user space drivers for near native DMA (Direct Memory...

What could be more fun than talking about kernel documentation? Things we
could get into:

• The state of the RST transition, what remains to be done, whether it's
  all just useless churn that makes the documentation worse, etc.

• Things we'd like to improve in the documentation toolchain.

• Overall organization of Documentation/ and moving docs when the need
  arises. It seems I end...

The ABI between Linux and user software mostly sits at the user/privileged boundary, although many architectures extend this with a small amount of special-case code that sits in userspace, such as in special pages or shared libraries (vDSOs) mapped into each user process [1] that user code can call into.

The reasons for this are a bit arbitrary: system interface libraries such as glibc and...

In this talk Dmitry will highlight some of the areas for improvement related to release quality, security, and developer experience and productivity. Then try to show that the existing processes, approaches and tools poorly cope with the current scale and rate of change and don't provide adequate quality and developer experience. Lastly Dmitry will advocate that only pervasive changes to the...

ZRAM is a compressed RAM based block device implementation which has gotten a lot of use recently primarily in the Android world. ZRAM consists of the block device front-end, compressor back-end and memory allocator back-end. Compressor back-end is accessed via a common API, and therefore it is easy with ZRAM to select the particular compression algorithm that fits your special purpose. As...