After two months of development, Linus Torvalds unveiled a few days ago the launch ofto new Linux kernel version 5.17.
Among the most notable changes highlights a new performance management system for the processors AMD, support for BPF programs compiled laptops, the transition from the pseudo-random number generator to the BLAKE2s algorithm, the new fscache backend to cache network file systems, among other things.
The new version received 14203 fixes from the developers of 1995, the size of the patch is 37 MB (changes affected 11366 files, 506043 lines of code were added, 250954 lines were removed).
Main novelties of the Linux Kernel 5.17
In this new version the possibility of nested mapping of user IDs of mounted file systems is implemented, which is used to map files of a certain user on an external partition mounted with another user on the current system. The added feature allows you to use the mapping recursively on file systems for which the mapping has already been applied.
The subsystem fscache has been completely rewritten. The new implementation is distinguished by a significant simplification of the code and the replacement of complicated programming and object state tracking operations with simpler mechanisms. Support for the new fscache is implemented in the CIFS file system.
Btrfs is optimized for register and fsync operations for large directories, implemented by copying only index keys and reducing the amount of recorded metadata, plus indexing and searching by free space record size support has been provided, which reduced latency and seek time by about 30%, which allowed the defragmentation operations to be interrupted.
Ext4 migrated to a new mount API which separates the stages of parsing the mount options and configuring the superblock, plus support for the lazytime and nolazytime mount options was removed, which were added as a temporary change to ease the transition from util-linux to using the flag MS_LAZYTIME and added support for setting and reading tags in the FS (ioctl FS_IOC_GETFSLABEL and FS_IOC_SETFSLABEL).
The controller amd-pstate has been added to provide dynamic frequency control for optimal performance. The driver supports new AMD CPUs and APUs, including some Zen 2 and Zen 3 generation chips, and was developed in collaboration with Valve to improve power management efficiency. For adaptive frequency switching, the CPPC (Collaborative Processor Performance Control) mechanism is used, which allows you to change indicators more accurately (not limited to three performance levels) and respond more quickly to changes in state than Previously used ACPI-based P-state drivers. (CPU frequency).
On the other hand, it is highlighted that an updated implementation is proposed from the pseudorandom number generator RDRAND, which is responsible for the operation of the /dev/random and /dev/urandom devices, notable for the transition to using the BLAKE2s hash function instead of SHA1 for entropy mixing operations. The change made it possible to increase the security of the pseudo-random number generator by getting rid of the troublesome SHA1 algorithm and removing the overwriting of the RNG initialization vector. Since the BLAKE2s algorithm is ahead of SHA1 in terms of performance, its use also had a positive effect on performance.
Added protection against processor vulnerabilities caused by speculative execution of instructions after unconditional jump-forward operations. The problem arises from the preemptive processing of instructions immediately following the jump instruction in memory (SLS, Straight Line Speculation). Enabling security requires a build with GCC version 12, which is currently in testing.
The subsystem drm (Direct Rendering Manager) and the i915 driver have added support for screens to display sensitive information, for example, some laptops are equipped with screens with a built-in Confidential View mode, making it difficult to view from the outside. Added changes allow you to plug in specialized drivers for such screens and control private browsing modes by setting properties in regular KMS drivers.
The controller amdgpu includes support for STB debugging technology (Smart Trace Buffer) for all AMD GPUs that support it. STB facilitates failure analysis and identifies the source of problems by storing in a special buffer information about the functions performed before the last failure.
Of the other changes that stand out:
- The i915 driver adds support for Intel Raptor Lake S chips and enables support for Intel Alder Lake P graphics by default.
- The fbcon/fbdev drivers returned support for hardware-accelerated scrolling in the console.
- Continuous integration of changes to support Apple M1 chips.
- Implemented the ability to use the simpledrm driver on systems with an Apple M1 chip to generate output via the firmware-provided frame buffer.
- bpf_loop() handler in the eBPF subsystem, which provides an alternative way of organizing loops in eBPF programs, faster and easier for the verifier to verify.
If you are interested in knowing more about it, you can check the details In the following link.