EPTLinux 3.13 is out!
So I'm surfing the web tonight and, out of random spike of interest, I decided to check what lkml.org has to offer for me today. I tend to visit it every now and then to gather what useful knowledge (or random rants) I can find there and rarely walk out empty-handed. It might be hard to imagine what one could find interesting in such technical gibberish (personally, I'm interested in kernel development and am working on a few minor projects related to kernel hacking lately), considering the issues tackled there are often so complex and specialized that for a complete stranger a hungry kitten on a street is a more important matter to take care of. Today though a surprise stroke me right in the face - turns out (2 < 2) is actually true. Who would've thought?
What was not a surprise (at least for the majority of people subscribed to the list or frequenting the site, myself excluded) was the release of 3.13 tarballs. It took Torvalds a while to finish the release (traveling, etc.), but it's here now, filled with some practical stuff, and "obviously", the 3.14 merge window is open.
+ Show Spoiler [Original announcement] +
The release got delayed by a week due to travels, but I suspect that's
just as well. We had a few fixes come in, and while it wasn't a lot, I
think we're better off for it. At least I hope so - I'll be very
disappointed if any of them cause more problems than they fix..
Anyway, the patch from rc8 is fairly small, with mainly some small
arch updates (arm, mips, powerpc, s390, sparc, x86 all had some minor
changes, some of them due to a networking fix for the bpf jit). And
drivers (mainly gpu and networking). And some generic networking
fixes. The appended shortlog gives more details.
Anyway, with this, the merge window for 3.14 is obviously open.
Linus
---
Aaro Koskinen (1):
MIPS: fix blast_icache32 on loongson2
Andreas Rohner (1):
nilfs2: fix segctor bug that causes file system corruption
Andrew Jones (1):
kvm: x86: fix apic_base enable check
Ben Skeggs (1):
drm/nouveau: fix null ptr dereferences on some boards
Benjamin Herrenschmidt (1):
powerpc: Check return value of instance-to-package OF call
Bjørn Mork (1):
net: usbnet: fix SG initialisation
Catalin Marinas (1):
Revert "arm64: Fix memory shareability attribute for ioremap_wc/cache"
Christian Engelmayer (1):
ieee802154: Fix memory leak in ieee802154_add_iface()
Christoph Paasch (1):
tcp: metrics: Avoid duplicate entries with the same destination-IP
Dan Carpenter (1):
cxgb4: silence shift wrapping static checker warning
Dave Airlie (1):
Revert "drm: copy mode type in drm_mode_connector_list_update()"
Eric Dumazet (2):
bpf: do not use reciprocal divide
parisc: fix SO_MAX_PACING_RATE typo
Eric W. Biederman (3):
vfs: In d_path don't call d_dname on a mount point
fork: Allow CLONE_PARENT after setns(CLONE_NEWPID)
vfs: Fix a regression in mounting proc
Gerald Schaefer (1):
net: rds: fix per-cpu helper usage
Hannes Frederic Sowa (2):
net: avoid reference counter overflows on fib_rules in multicast
forwarding
ipv6: simplify detection of first operational link-local address
on interface
Heiko Carstens (1):
s390/bpf,jit: fix 32 bit divisions, use unsigned divide instructions
Huacai Chen (1):
MIPS: fix case mismatch in local_r4k_flush_icache_range()
Hugh Dickins (1):
percpu_counter: unbreak __percpu_counter_add()
Ilia Mirkin (1):
drm/nouveau/mxm: fix null deref on load
Ivan Vecera (1):
be2net: add dma_mapping_error() check for dma_map_page()
Jan Kara (1):
writeback: Fix data corruption on NFS
Jean Delvare (1):
hwmon: (coretemp) Fix truncated name of alarm attributes
Jesse Barnes (1):
drm/i915/bdw: make sure south port interrupts are enabled properly v2
Jitendra Kalsaria (1):
qlge: Fix vlan netdev features.
John Stultz (2):
seqlock: Use raw_ prefix instead of _no_lockdep
sched_clock: Disable seqlock lockdep usage in sched_clock()
Linus Torvalds (2):
x86, fpu, amd: Clear exceptions in AMD FXSAVE workaround
Linux 3.13
Marek Lindner (1):
batman-adv: fix batman-adv header overhead calculation
Michael S. Tsirkin (1):
MAINTAINERS: add virtio-dev ML for virtio
Mika Westerberg (1):
e1000e: Fix compilation warning when !CONFIG_PM_SLEEP
Mikulas Patocka (1):
mm: fix crash when using XFS on loopback
Ming Lei (1):
lib/percpu_counter.c: fix __percpu_counter_add()
Neal Cardwell (1):
inet_diag: fix inet_diag_dump_icsk() to use correct state for
timewait sockets
NeilBrown (6):
md/raid5: Fix possible confusion when multiple write errors occur.
md/raid10: fix two bugs in handling of known-bad-blocks.
md/raid1: fix request counting bug in new 'barrier' code.
md/raid5: fix a recently broken BUG_ON().
md/raid10: fix bug when raid10 recovery fails to recover a block.
md: fix problem when adding device to read-only array with bitmap.
Paulo Zanoni (1):
drm/i915: fix DDI PLLs HW state readout code
Peter Korsgaard (1):
dm9601: add USB IDs for new dm96xx variants
Peter Zijlstra (1):
x86, mm, perf: Allow recursive faults from interrupts
Qais Yousef (1):
crash_dump: fix compilation error (on MIPS at least)
Rafael J. Wysocki (1):
Revert "ACPI: Add BayTrail SoC GPIO and LPSS ACPI IDs"
Richard Weinberger (1):
net,via-rhine: Fix tx_timeout handling
Rik van Riel (1):
sched: Calculate effective load even if local weight is 0
Robert Richter (1):
perf/x86/amd/ibs: Fix waking up from S3 for AMD family 10h
Russell King (1):
Revert "ARM: 7908/1: mm: Fix the arm_dma_limit calculation"
Soren Brinkmann (1):
clocksource: cadence_ttc: Fix mutex taken inside interrupt context
Stephen Boyd (1):
ARM: 7937/1: perf_event: Silence sparse warning
Stephen Warren (1):
i2c: Re-instate body of i2c_parent_is_i2c_adapter()
Steven Rostedt (1):
ftrace/x86: Load ftrace_ops in parameter not the variable holding it
Sudeep Holla (1):
ARM: 7934/1: DT/kernel: fix arch_match_cpu_phys_id to avoid
erroneous match
Taras Kondratiuk (2):
ARM: 7939/1: traps: fix opcode endianness when read from user memory
ARM: 7938/1: OMAP4/highbank: Flush L2 cache before disabling
Ville Syrjälä (1):
drm/i915: Don't grab crtc mutexes in intel_modeset_gem_init()
Yuval Mintz (1):
bnx2x: Don't release PCI bars on shutdown
Source - https://lkml.org/lkml/2014/1/19/148
Featuring!:
- Block layer revamp for high performance SSD storage!
- nftables!
- Improved support for Radeons: power management, GPU switching, R9 290X Hawaii support!
- Power capping framework!
- Intel MIC support!
- Improved performance on NUMA systems!
- Improved page table access scalability in hugepage workloads!
- Improvement in squashfs performance!
- New userspace socket rate capping mechanism!
- TCP Fast Open enabled by default!
- Support for the secure element!
- Support for the High-availability Seamless Redundancy protocol!
Notes
From a casual user's standpoint nothing really major has been added or modified in this version. Some performance tweaks over 3.12 are worth a mention though, especially the improved NUMA performance, as while they are nigh-invisible, things like that build up over time and make way for major boosts. Not 'featured' are also continuous tweaks to btrfs support and numerous (as usual) changes to the networking subsystem. Considering the latter has its own mailing list, the work there seems to be never done (even crossing out the never-ending need for driver writers). I've recently read an article on arstechnica about some cool features of ZFS and btrfs; if you are interested in such technical stuff, it's worth a read.
Why should I care? - 3.13 edition
If you are an everyday Ubuntu user, chances are you wouldn't even notice the kernel update's impact on your system (no hate on Ubuntu users, I swear). Some, if not most, of the features of this release are catered towards kernel developers and hackers, but there are some standing out as 'important' for a percentage of the userbase, namely the improved Radeon support. If you run Linux on a laptop with such a card, your battery will soon live a happier life, or you will enjoy a better framerate in Pacman.
As for me, the squashfs tweak is nice - probably won't get to make much use of it since I'm working with (really) outdated hardware, but in the future, who knows if I won't get to smile when my OpenWRT build launches a second quicker.
+ Show Spoiler [Disclaimer] +
This is most likely a one time thing, we'll see in the future. For a moment I wanted to throw out a random cinematic trailer for the release (sic!) but I figured it's one of these things that you start and then never end as your fuel tank goes empty an hour in. Hopefully this will be enough for some people. Huge thanks to the guys over at kernelnewbies.org for providing everyone with a concise list of changes, helped me a ton with compiling this short post.
Questions?
Feel free to chat about whatever related to the release, ask questions and answer them (even in random-links-zero-explanation-stack-overflow-style). I personally would love to know what this nftables business is all about - as I understand it, the bytecode compilation occurs in userspace, which immediately flashes the red light for me, but that might be due to my security ignorance. Will have to read more about it.
