Slashdot Mirror


Linux 3.19 Kernel To Start 2015 With Many New Features

An anonymous reader writes Linux 3.18 was recently released, thus making Linux 3.19 the version under development as the year comes to a close. Linux 3.19 as the first big kernel update of 2015 is bringing in the new year with many new features: among them are AMDKFD HSA kernel driver, Intel "Skylake" graphics support, Radeon and NVIDIA driver improvements, RAID5/6 improvements for Btrfs, LZ4 compression for SquashFS, better multi-touch support, new input drivers, x86 laptop improvements, etc.

66 comments

  1. Re:DirectX? by Anonymous Coward · · Score: 0

    To make Linux useful for playing games.

    A true gamer games within his mind.

  2. Re:DirectX? by Anonymous Coward · · Score: 0

    Run Windows applications on Linux, BSD, Solaris and Mac OS X.

    Not possible, sorry. :-)

  3. A Device Manager? by Anonymous Coward · · Score: 0

    It's the simple things that make a difference.

    1. Re:A Device Manager? by __aaclcg7560 · · Score: 1

      ls -al /dev/

    2. Re:A Device Manager? by Anonymous Coward · · Score: 0

      It's the simple things that make a difference.

      Coming to Linux in 2050.

    3. Re: A Device Manager? by Anonymous Coward · · Score: 0

      Linux is just a Kernel...

    4. Re:A Device Manager? by Sam36 · · Score: 0

      Dumbest response ever.. Almost as a bad as complaining that your Ferrari doesn't have a cup holder and automatic transmission....

  4. Re: DirectX? by Anonymous Coward · · Score: 0

    Works fairly well.... Definitely some bugs, but overall 8/10.

  5. Re:systemd? by Anonymous Coward · · Score: 0, Troll

    The Linux kernel is depricated - systemd will take over its functionality.

  6. Will RTL8192 still be broken? by Anonymous Coward · · Score: 1

    It worked always great, then I think after 3.12 it doesn't work anymore. The problem is that most distributions I tried installed with 3.08 or something below, and they all want to upgrade to a newer kernel and all probably want to upgrade utilities - that also could be broken for that chipset. So long story short, due to linux being broken one piece at time, I had to buy a mac :(

    1. Re:Will RTL8192 still be broken? by jones_supa · · Score: 1

      Sounds somewhat worrying. That's a very popular chip.

      There's always a few things you can do:

      - Talk to LKML.
      - Post a bug report in bugzilla.kernel.org.
      - Find the specific patch which caused the regression with git bisect. Canonical has a good guide on the topic (use "man git-bisect" for more info).

    2. Re: Will RTL8192 still be broken? by Anonymous Coward · · Score: 1

      Could you be more specific on a device that isn't working. I'm pretty sure I have run a device with a version of this chip not long ago without issue. Knowing a specific device might help track down any issues.

    3. Re:Will RTL8192 still be broken? by Anonymous Coward · · Score: 0

      I just don't care anymore, I wasted too much time on that laptop, it's just there to be a voip testing server.

    4. Re:Will RTL8192 still be broken? by Anonymous Coward · · Score: 0

      Congrats on not being a virgin anymore!

    5. Re:Will RTL8192 still be broken? by Eunuchswear · · Score: 1

      It worked always great, then I think after 3.12 it doesn't work anymore.

      Seems to be working for me with Debian's 3.16.

      Bug report numbers?

      --
      Watch this Heartland Institute video
  7. Re:DirectX? by Anonymous Coward · · Score: 0

    Idiot

  8. Re:systemd? by Anonymous Coward · · Score: 0

    Ah! But it could and should happen!

    Due to its nature, updates to the Linux kernel tend to render machines unusable until the rest of the OS catches up.

    This could be managed, maybe even have two kernels running at the same time, the old and the new!

  9. "New Features" by coop247 · · Score: 1

    driver improvements, RAID5/6 improvements for Btrfs, LZ4 compression for SquashFS, better multi-touch support, new input drivers, x86 laptop improvements

    Not sure "new features" is the right summary of changes.

    --
    //TODO: Insert catchy phrase
    1. Re:"New Features" by fph+il+quozientatore · · Score: 1

      By the way, wtf is a "x86 laptop improvement"?

      --
      My first program:

      Hell Segmentation fault

    2. Re:"New Features" by r1348 · · Score: 1
    3. Re:"New Features" by Anonymous Coward · · Score: 0

      Anyone running Linux on a laptop doesn't require real productivity.

    4. Re:"New Features" by SuricouRaven · · Score: 1

      x86/64, but with a ton of stuff crammed into the ACPI interface, some of which is actually there for power management, none of which is properly documented because every manufacturer likes to put their hotkeys and screen close sensor on a different ACPI event.

    5. Re: "New Features" by Anonymous Coward · · Score: 0

      I've been using Linux on laptops for serious software development for years. Blow it out your ass.

  10. System Hardware. Or yum install hardinfo by raymorris · · Score: 3, Informative

    The kernel and friends manage hotplug devices quite nicely.
    I take that to mean you want a clickity-click GUI, so you can see what the system has already detected and handled properly for you, and do things without needing to understand what you're doing. If that's what you're looking for, hardinfo is a well-known option. Your choice of graphical desktop environment probably has one it provides by default as well. Look under "System" or similar.

  11. Re:systemd? by Anonymous Coward · · Score: 0

    Redhat and the spikey haired arch script kiddies will never let this happen. Therefore 2015 will be the year of the BSD uprising.

  12. Re:systemd? by Anonymous Coward · · Score: 1

    Red Hat is why shitstemd exists...

  13. Re:systemd? by phantomfive · · Score: 4, Funny

    The Linux kernel is depricated[sic] - systemd will take over its functionality.

    +1 hilarious.

    --
    "First they came for the slanderers and i said nothing."
  14. Re:systemd? by Anonymous Coward · · Score: 0

    Absolute bollocks. I'm running kernel 3.18 right now on a Ubuntu 9.04 machine, and everything works fine.

  15. Re:systemd? by twistedcubic · · Score: 1

    This is not at all true. The opposite is true. The latest kernel works perfectly fine on any 15 year old x86 machine.

  16. Re:systemd? by Barsteward · · Score: 0

    I wish we could score posts (score: -2 Immature Twat)

    --
    "The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
  17. Re:systemd? by Anonymous Coward · · Score: 0

    Oh but FreeBSD are looking to move to a Systemd like init too, and laughing at the linux users complaining about this direction

  18. Re:systemd? by Anonymous Coward · · Score: 1, Interesting

    The basic idea is good, but SystemD has implemented it poorly. FreeBSD is sitting back and taking notes of what not to do. SystemD is pioneering the idea, could have been done better, but FreeBSD will get the benefit of hindsight. FreeBSD will still manage to mess some stuff up and OpenBSD will refine it 10 years from now.

  19. Re:systemd? by Barsteward · · Score: 2

    "but SystemD has implemented it poorly." - thats a matter of opinion

    "SystemD is pioneering the idea" - surely Sun, Apple and Ubuntu have similar in place before systemd came about.

    --
    "The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
  20. Re:systemd? by Anonymous Coward · · Score: 1

    I'm glad to hear that systemd is being reimplemented in Emacs Lisp so it can run seamlessly, as all functionality should, as part of Emacs.

  21. Re: DirectX? by Anonymous Coward · · Score: 0

    Tan Oracle Database under Windows for months...

  22. Re:System Hardware. Or yum install hardinfo by Blaskowicz · · Score: 1

    Not really : "hardinfo" itself is not known or not known under that name, and a report about the installed hardware is a bit worthless (lspci and lsusb do about that).
    The Device Manager is not only a unique GUI (stable during two decades of Windows versions), it allows to choose or install drivers and even to configure the drivers. You can do things that are seemingly impossible in linux like limiting a wireless card to a maximum speed (to get a connection "slower" but more stable), or other things. It would be not only having the simple GUI (from times when Windows was easier to use) do lsmod and modprobe kind of work, but also configuring the kernel modules (or kernel), which is something an advanced user is likely to not know about (do I need to set up a build environment and recompile kernel modules?, compile kernel?)

    There does exist a useful GUI under linux, the "Proprietary Driver Manager" which allows to switch between nvidia and nouveau (for instance) by clicking a radio button.

  23. Re:System Hardware. Or yum install hardinfo by SuricouRaven · · Score: 1

    Manual override of connection speed is via the 'iwconfig' command. I've done it before. I don't know if you can set a maximum speed, but you can set a fixed speed.

  24. Re:systemd? by Anonymous Coward · · Score: 0

    "but SystemD has implemented it poorly." - thats a matter of opinion

    The point is that that's the opinion of people who dislike systemd. That doesn't automatically mean they'll also dislike "a Systemd like init".

  25. Re:systemd? by Anonymous Coward · · Score: 0

    While those other companies did create similar things well before SystemD, it is the first one making it to the masses, and not limited to research or niche.

  26. Lockup issue by xluap · · Score: 3, Interesting

    Has the lockup issue been solved?

    http://www.phoronix.com/scan.p...

    1. Re:Lockup issue by jones_supa · · Score: 2

      That's something different.

      See this LKML page. Search with CTRL+F for "frequent lockups in 3.18rc4".

      The thread is clearly still going on. As far as I can tell, the bug has not been fixed.

    2. Re:Lockup issue by Anonymous Coward · · Score: 2, Informative

      As far as I know, they found the cause (it seems the kernel doesn't handle well some clock issues and clock overflow), and Linus send a RFC-like patch that kinda "fixes" the thing but it is so hackish that it needs to be discussed, generalized and ironed out. DaveJ doesn't have access to the hardware triggering the issue anymore, so it may be a while before they can actually fix it in a proper manner.

      On the way to figuring out the cause for that bug, they found other lockup issues with similar symptoms but completely different causes and fixed them.

  27. Re:systemd? by unixisc · · Score: 1

    But is systemd GPLv3? It would have to be in order to be a part of emacs

  28. Re:systemd? by Grishnakh · · Score: 1

    The point is that that's the opinion of people who dislike systemd. That doesn't automatically mean they'll also dislike "a Systemd like init".

    People like that might not (those who think it's implemented poorly). The people who will are the ones who insist on an init system that follows the "Unix philosophy". They're not going to be satisfied with anything other than sysvinit.

  29. Re:systemd? by Anonymous Coward · · Score: 0

    Stop trolling. Technical objections != dogmatic hate. A replacement for sysvinit does not have to be exactly like it. Do you remember the same pushback against upstart? Why not?

    To quote the future: systemd can bite my shiny metal ass.

  30. Re:systemd? by Grishnakh · · Score: 1

    Great, we look forward to seeing your new upstart-based distro, since everyone else has abandoned it.

  31. Re:DirectX? by loonycyborg · · Score: 1

    Not essential anymore. All major game engines tend to have opengl codepaths and even full linux ports.

  32. Re:systemd? by Barsteward · · Score: 1

    "Technical objections != dogmatic hate" - after reading a lot of posts by those that don't like systemd in these forums, that statement does not stand up to scrutiny. Its been "dogmatic hate" under the guise of very flimsy "technical objections" that were really personal likes/dislikes. I'd love to see a list of "technical objections" against systemd that have not already been (or cannot be) refuted or have different ways of doing the same thing.

    It would be nice to have an adult discussion of the pro/cons of systemd without the emotive claptrap especially against LP.

    --
    "The hands that help are better far than lips that pray." - Robert Ingersoll (1833-1899)
  33. Re:systemd? by Anonymous Coward · · Score: 0

    Not any of binary distributions, I bet. Even distros that use -i386- as architecture designation are usually compiled with optimizations for P4 or even later now.

  34. Re:systemd? by Eunuchswear · · Score: 1

    Great, we look forward to seeing your new upstart-based distro.

    That'd be Debian. Jessie will support systemd, upstart or sysvinit.

    --
    Watch this Heartland Institute video
  35. Re:systemd? by drinkypoo · · Score: 1

    are they using -march or -mtune?

    I do understand that many won't run out of the box with less than a 586-class processor now, but I don't think P4 is common yet

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  36. Re:systemd? by Anonymous Coward · · Score: 0

    Red Hat is why shitstemd exists...

    Then let RedHat have systemd and everyone with a clue can stay with the superior SysVInit. I like text log files instead of the binary garbage coming to a Linux system near you and that has plagued Microsoft Windows for decades.

  37. old hardware left behind by bzipitidoo · · Score: 1

    One of my PCs is a Gateway GT5628 PC with an Intel Q6600 chipset. Shutdown used to work every time on this PC, with kernels around the 2.6.32 version. By 2.6.38, shutdown was unreliable. About half the time shutdown works, and the other half the computer goes through the shutdown process successfully and at the very end, fails to turn itself off, sitting on the text screen with "power down" displayed on the monitor. I have to hold the power button for 4 seconds to complete the shutdown.

    I haven't submitted any bug report. It would be nice if shutdown worked every time like it used to, but it's a minor problem with an easy workaround, so minor I figured no one would care to hunt it down and fix it. I haven't. I could try a bunch of old kernels out to narrow down when this feature was broken, but haven't felt it was worth my time.

    Linux is very good about supporting old hardware, but inevitably some does get left behind. They deliberately dropped support for the 386 somewhere around kernel version 3.5. Other old hardware simply isn't checked. When was the last time anyone tried a mouse that plugs into the serial port? Not USB, not PS/2, but ye olde 9 pin (or 25 pin!) serial port? Last time I fooled around with one about 5 years ago, I couldn't get XWindows to recognize it. The fastest "fix" is to just get a USB or PS/2 mouse. Or, at the price of systems these days, a whole new computer.

    --
    Intellectual Property is a monopolistic, selfish, and defective concept. It is "tyranny over the mind of man"
    1. Re:old hardware left behind by Anonymous Coward · · Score: 0

      When was the last time anyone tried a mouse that plugs into the serial port?

      Last week. Needs a custom xorg.conf entry, but otherwise works fine on debian wheezy.

  38. Re:System Hardware. Or yum install hardinfo by hairyfeet · · Score: 2, Insightful

    Which leads us to the sadly all too true obligatory XKCD which is why Linux on the desktop is so low its getting its ass handed to it by "other" and has gotten so low its literally below the margin for error.

    Considering that every time Linux starts to get stable the devs take a big steaming shit on it, Pulse, KDE 4, Gnome 3, Systemd, not to mention Torvalds constant kernel fiddling, is anybody really surprised by the plummeting numbers? Its a damned shame but as long as devs would rather put out yet another release instead of fixing the bugs in the previous release Linux will always remain in alpha quality.

    --
    ACs don't waste your time replying, your posts are never seen by me.
  39. Re:System Hardware. Or yum install hardinfo by Anonymous Coward · · Score: 0

    You can. iwconfig wl0 rate 11M auto = anything up to 11Mbit

  40. Re:DirectX? by Anonymous Coward · · Score: 0

    Does anyone still use DirectX?

  41. ftape by Anonymous Coward · · Score: 0

    Can I use my QIC-80 again yet?

  42. SYSTEMD by Anonymous Coward · · Score: 0

    SystemD Systemd systemd
    Fuck :-(

  43. Re:DirectX? by Anonymous Coward · · Score: 0

    Really? At least on FPS side I have not seen any AAA-games with Linux support in years and I am writing this as a Debian user who boots Windows at home only to play games. Ref: http://store.steampowered.com/search/?term=linux#sort_by=Released_DESC&sort_order=DESC&term=linux&category1=998&tags=19&page=1

  44. Re:DirectX? by Ultra64 · · Score: 1

    Really curious to know how this is "Flamebait".
     

  45. Re:systemd? by Pi1grim · · Score: 1

    Have you actually taken a look at the log format that journald uses? Text is stored verbatim in them, so you can even dig through them with grep. Binary meta-data being added to it makes wonderful things possible - getting logs by unit, time and other parameters without whipping out a mile-long regexp. So please read up on the topic you attempt to bash or you will end up looking pretty stupid to anyone with a clue. Just like you did now.