Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Operating Systems Software Linux

Linux Kernel 2.6.11 Released 312

Posted by CmdrTaco from the download-compile-reboot-repeat dept.
Xpilot writes "Linus Torvalds has just announced the availability of the newest Linux kernel release, 2.6.11. The newest addition to Linux that's stirring up some excitement is the inclusion of Infiniband support. You can get it from the usual mirrors at http://kernel.org/mirrors."
This discussion has been archived. No new comments can be posted.

Linux Kernel 2.6.11 Released More

Linux Kernel 2.6.11 Released

Comments Filter:

  • To Infiniband (Score:4, Funny)

    by Anonymous Coward on Wednesday March 02, 2005 @10:22AM (#11822794)
    And beyond!

  • infiniband..... (Score:4, Funny)

    by Anonymous Coward on Wednesday March 02, 2005 @10:22AM (#11822797)
    does that mean it goes PAST 11?

    • Re:infiniband..... (Score:5, Funny)

      by Anonymous Coward on Wednesday March 02, 2005 @10:28AM (#11822847)
      Nigel: "You're up to 11 on your server and up to 11 on your network and you need that extra little push to set it over the top and you got nowhere else to go. So this amp goes to infinity."

      Interviewer: "I see. But why not just make 11 louder?"

      Nigel: "But it goes to infinity."

      Interviewer: "Yeah, but what if you just made 11 as loud as infinity?"

      Nigel: "But...it goes to infinity."
  • Looks like Knoppix jumped the gun by including the 2.6 kernel in the new distro. If they had just waited a few hours...

    • Re:Knoppix jumped the gun... (Score:5, Funny)

      by Anonymous Coward on Wednesday March 02, 2005 @10:45AM (#11822979)
      Yeah, all those people waiting to boot their supercomputers using Knoppix to take advantage of Infiniband will just have to keep waiting.
    • The release of knoppix was announced few hours ago but things like the kernel are added much earlier. Like days or weeks earlier. They're not added at the last minute, because distro maintainence is a tough job. And many times ditros contain not the latest version of programs available at the time of the release. There are a lot of reasons. On a side note, people who want infiniband will be able to update the kernel on the knoppix cd.

  • infiniband? (Score:4, Insightful)

    by Quasar1999 ( 520073 ) on Wednesday March 02, 2005 @10:23AM (#11822806) Journal
    InfiniBand, which is derived from its underlying concept of "infinite bandwidth,"...

    Umm... I don't know about you... but that description didn't help me much... infinite bandwidth? What is this? How is this? How does linux get past physical hardware limitations that other os's can't?

    • Re:infiniband? (Score:5, Informative)

      by stecoop ( 759508 ) on Wednesday March 02, 2005 @10:29AM (#11822862) Journal
      Google knows all.

      Intell talks about Infiniband Architecture [intel.com]

      Initially InfiniBand Technology will be used to connect servers with remote storage and networking devices, and other servers. It will also be used inside servers for inter-processor communication (IPC) in parallel clusters. Customers requiring dense server deployments, such as ISPs, will also benefit from the small form factors being proposed. Other benefits include greater performance, lower latency, easier and faster sharing of data, built in security and quality of service, improved usability (the new form factor will be far easier to add/remove/upgrade than today's shared-bus I/O cards).

      Additionally, InfiniBand Architecture reduces total cost of ownership by focusing on data center reliability and scalability. The technology addresses reliability by creating multiple redundant paths between nodes (reducing hardware that needs to be purchased). It also moves from the load-and-store-based communications methods used by shared local bus I/O to a more reliable message passing approach.

      Scalability needs are addressed in two ways. First, the I/O fabric itself is designed to scale without encountering the latencies that some shared bus I/O architectures experience as workload increases. Second, the physical modularity of InfiniBand Technology will avoid the need for customers to buy excess capacity up-front in anticipation of future growth. Instead, they will be able to buy what they need at the outset and 'pay as they grow' to add capacity without impacting operations or installed systems.

      • Re:infiniband? (Score:5, Insightful)

        by micromoog ( 206608 ) on Wednesday March 02, 2005 @10:38AM (#11822930)
        So what is it in non-marketing terms?

        • Re:infiniband? (Score:5, Funny)

          by MrHanky ( 141717 ) on Wednesday March 02, 2005 @10:49AM (#11823028) Homepage Journal
          It's a rift in the time-space continuum.
          • I was thinking before I replied to this, because I know a reponse that is funny to me, goes like this: So what is it? Then when explained the response is: So what is it? Then when explained again the response is: So what is it? Only kidding. But jeese, does anyone watch Red Dwarf anymore? Most people would just think I am a jackass, not that they'd be wrong.

        • Re:infiniband? (Score:5, Informative)

          by wootest ( 694923 ) on Wednesday March 02, 2005 @10:51AM (#11823042)
          Virginia Tech used Infiniband to wire up their G5 cluster. It's basically very fast I/O with some good logic built-in - "The technology addresses reliability by creating multiple redundant paths between nodes (reducing hardware that needs to be purchased)." is basically the same as the change from linear, Token Ring-ish networks to big Ethernet meshes like the Internet. I don't claim to know much at all about this, but seemingly it's the superior alternative today, and it sounds like it should be as well.

          • Re:infiniband? (Score:2, Funny)

            by suso ( 153703 )
            is basically the same as the change from linear, Token Ring-ish networks to big Ethernet meshes ...

            Is that like Tolkien Elv-ish?

        • Re:infiniband? (Score:5, Informative)

          by Tomcat666 ( 210775 ) on Wednesday March 02, 2005 @11:03AM (#11823147) Homepage
          http://en.wikipedia.org/wiki/Infiniband [wikipedia.org]

          You're welcome.

        • what it is (Score:5, Interesting)

          by r00t ( 33219 ) on Wednesday March 02, 2005 @11:05AM (#11823172) Journal
          Infiniband is a "smart" fabric; it supports reliable data transmission. Note that, as with modem protocols, this causes ugly interactions with TCP retransmits. TCP is really designed to work over an Ethernet-like network, where congestion causes packet loss and not much else bad ever happens.

          You can use Infiniband as a LAN, for storage, or maybe for within a box. You could say that Infiniband starts where Hypertransport leaves off.

          For the short-haul usage, Infiniband is kind of big in terms of chip real estate. You can't cram it into a corner of a little FPGA like you can with RapidIO. For the long-haul usage, 1 gig or 10 gig Ethernet might be a better choice.

          Note that Intel, originally the primary sponsor behind Infiniband, no longer gives a damn. But hey, if you have money to burn...

        • It is mysterious and powerful. In fact, it's mystery is only exceeded by its power. ;)

      • ahem.. (Score:3, Funny)

        by carlmenezes ( 204187 )
        Google knows all.

        err...actually, google knows all who know all :)

    • Re:infiniband? (Score:2, Informative)

      by dhbiker ( 863466 )
      http://www.infinibandta.org/ibta/ [infinibandta.org] I think infinite bandwidth is more thatn a little misleading! but to take an excerpt from their marketing blurb "The first version of the specification for the technology was completed in October 2000 and the InfiniBand Trade Association is well on its way to establishing a new signaling rate specification beyond 100Gb/s"

      • Re:infiniband? (Score:5, Insightful)

        by pomakis ( 323200 ) <pomakis@pobox.com> on Wednesday March 02, 2005 @11:06AM (#11823191) Homepage
        I think infinite bandwidth is more thatn a little misleading! but to take an excerpt from their marketing blurb "The first version of the specification for the technology was completed in October 2000 and the InfiniBand Trade Association is well on its way to establishing a new signaling rate specification beyond 100Gb/s"

        100Gb/s? Then they're almost there! I'm sure infinity isn't much bigger than that.

    • Re:infiniband? (Score:5, Funny)

      by Kjella ( 173770 ) on Wednesday March 02, 2005 @10:34AM (#11822908) Homepage
      How does linux get past physical hardware limitations that other os's can't?

      Marketing. They transcend the physically possible on a regular basis. Though you missed the source. "How does [InfiniBand] get past physical hardware limitations that other [hardware] can't?" It is their marketing fluff, Linux merely supports the technology.

      Kjella

    • Re:infiniband? (Score:2, Informative)

      by TuringTest ( 533084 )
      Linked from the article:

      Infiniband (define)

      Both an I/O architecture and a specification for the transmission of data between processors and I/O devices that has been gradually replacing the PCI bus in high-end servers and PCs. Instead of sending data in parallel, which is what PCI does, InfiniBand sends data in serial and can carry multiple channels of data at the same time in a multiplexing signal. The principles of InfiniBand mirror those of mainframe computer systems that are inherently channel-based s
    • Umm... I don't know about you... but that description didn't help me much... infinite bandwidth? What is this? How is this?

      Agreed.

      They should have called it "SynerBand" as in, "Synergized Bandwidth". Alternatively, eSynerBand-Numa.iFlex2@@@ would have been a good choice.
    • It means that physical hardware limitations are the ONLY limitations. I think the "underlying concept" is that it should scale up as more connections are added. Other standards had inherent limitations so that adding more or better quality connections would help up to a certain point. Infiniband is not only for linux.

  • Article text (Score:5, Informative)

    by Anonymous Coward on Wednesday March 02, 2005 @10:26AM (#11822830)
    ...before the site is slashdotted...

    The Linux world is bracing for the final release of the new Linux 2.6.11 kernel, which will include a long list of driver updates and patches, with InfiniBand support perhaps being one of most interesting new additions.

    Late last night, Linux creator Linus Torvalds issued the fifth release candidate for the 2.6.11 kernel. The first 2.6.11 RC was issued on Jan. 12; the second on Jan 21; the third on Feb. 2; and the fourth on Feb. 12.

    In the RC5 posting, Torvalds indicated that it was likely the last RC before the final release.

    "Hey, I hoped -- rc4 was the last one, but we had some laptop resource conflicts, various ppc TLB flush issues, some possible stack overflows in networking and a number of other details warranting a quick -- rc5 before the final 2.6.11," Torvalds wrote.

    "This time it's really supposed to be a quickie, so people who can, please check it out, and we'll make the real 2.6.11 asap."

    The long list of updates in the 2.6.11 kernel includes architecture updates for x86-64, ia64, ppc, arm and mips, as well as updates to ACPI (define), DRI (Direct Rendering Infrastructure, which permits direct access to graphics hardware for X Window System users), ALSA (Advanced Linux Sound Architecture, which provides MIDI and audio functionality to the Linux), SCSI (define) and the XFS high-performance journaling filesystem.

    The 2.6.11 kernel will also be significant in that it includes driver support for the InfiniBand (define) interconnect architecture. InfiniBand, which is derived from its underlying concept of "infinite bandwidth," is a switched fabric interconnect technology for high-performance network devices that is common in a number of supercomputer clusters.

    The upcoming inclusion of InfiniBand support in the Linux kernel is a major step according to the InfiniBand Trade Association.

    "The inclusion of InfiniBand drivers in the upstream Linux kernel is a significant milestone," Ross Schibler, CTO of InfiniBand vendor Topspin Communications, told internetnews.com.

    InfiniBand support was available previously in various Linux distributions, but it wasn't part of the mainstream kernel.org Linux.

    "This now means that anyone that downloads a kernel will have automatic access to the software," explained Schibler. "It also means that any upcoming distributions (Red Hat, SUSE, etc.) will have the software included on their CDs. Previously SUSE had it on a distribution, but only in the 'unsupported' directory."

    Schibler sees the inclusion of InfiniBand as a testament to the maturation of the technology.

    "Now that the technology has matured to such a point that Linus has accepted it into the kernel, the way is paved for greater distribution of the code and accelerated deployment of the technology," Schibler said.

    The previous Linux kernel.org release, version 2.6.10 was issued on Dec. 24 after two release candidates. Linux distribution began including the 2.6.10 thereafter with Red Hat's Fedora Project being one of the first.

    Fedora Core 3 initially shipped with the 2.6.9 kernel and then upgraded to the 2.6.10 kernel on Jan 13. Mandrakelinux's 10.2 Beta 3 also includes the 2.6.10 release. SUSE Linux 9.2 currently includes the 2.6.8 kernel.

    Including the most recent kernel into a distribution is not a particularly easy task. The upcoming Debian, code-named Sarge, will only ship with the 2.6.8 kernel. In a release update e-mail, Debian Sarge release manager Andreas Barth related that a meeting was recently held to review the status of which kernel they would include.

    "The team leads involved eventually decided to stay with kernel 2.6.8 and 2.4.27, rather than bumping the 2.6 kernel to 2.6.10," Barth wrote. "This decision was made upon review of the known bugs in each of the 2.6 kernel versions; despite some significant bugs in the Debian 2.6.8 kernel tree, these bugs were weighed against the additional delays that a kernel version bump would introduce in t
    • Yay, there really are reasons NOT to include anything above 2.6.8 into a distribution. Most notable of them that somebody broke it on Sparc.

    • Re:Article text (Score:3, Insightful)

      by strider44 ( 650833 )
      Ross Schibler, CTO of InfiniBand vendor Topspin Communications, told internetnews.com. . . . "Now that the technology has matured to such a point that Linus has accepted it into the kernel, the way is paved for greater distribution of the code and accelerated deployment of the technology," Schibler said.

      That makes for an interesting comment, previously people have been ignoring linux and gunning for windows.
      • Obviously, it's server buyers that are the target market for this technology.

        Windows is merely one of the possible options on a server. Linux and other Unix-likes are others.
    • You mentioned an update to XFS. I remember one time I had a friend whose new Debian install wouldn't boot because the root partition wouldn't mount correctly for one reason or another. I asked him what filesystem he was using, and he said, "hang on, let me check," and he watched the kernel messages scroll by and saw "xfs" (which of course stands for x font server) and so determined that he was using XFS. It took quite a long time to convince him that Debian stable was not quite that cutting-edge.

  • Mac laptops (Score:5, Informative)

    by colinleroy ( 592025 ) on Wednesday March 02, 2005 @10:28AM (#11822853) Homepage
    And G4 laptops with an ATI finally get sleep support thanks to BenH's work!

    (I know, "why would you want to run Linux on a Mac". Don't bother asking).

  • someone tell nvidia! (Score:2, Informative)

    by Mondongo ( 43895 )
    I'm running 2.6.10 and the NVidia X Driver won't compile against it ... someone tell NVidia to keep up with this!
    • I upgraded to 2.6.10 yesterday (with typically bad timing) and had no problems with the Nvidia drivers. I compiled them, rebooted and X.org appeared in all its glory. This was on a gentoo system. Have you got the latest version of the drivers from Nvidia?

    • Re:someone tell nvidia! (Score:4, Interesting)

      by archen ( 447353 ) on Wednesday March 02, 2005 @10:46AM (#11823002)
      What distro do you use? It works for me using Gentoo, but I also use the unstable nvidia drivers (because I couldn't get the "stable" ones to work months back).

      Which reminds me, when is Linus going to leave the 2.6x tree alone so we don't have to worry about so much broken shit all the time? Imagine if Windows changed it's kernel a couple times a year and broke the video drivers each time. People would bitch endlessly, but I guess as Linux users, we just have to put up with it.
      • It works for me using Gentoo, but I also use the unstable nvidia drivers (because I couldn't get the "stable" ones to work months back).

        I had to do the same thing with Gentoo, but the stable drivers seem to be fixed now.
        • It is very rare that you have to upgrade the kernel, check if you need to, before upgrading.
        • Nvidia drivers are kernel modules, so you need to run their installation program every time you upgrade the kernel, since hookups in the kernel may change and a module compiled for a given version of kernel may not work flawlessly with another version. This does not mean upgrade breaks video drivers, if you had to rewrite the drivers, then I would call it breaking.
        • You are barking at the wrong tree. If Nvidia re
      • Which reminds me, when is Linus going to leave the 2.6x tree alone so we don't have to worry about so much broken shit all the time? Imagine if Windows changed it's kernel a couple times a year and broke the video drivers each time. People would bitch endlessly, but I guess as Linux users, we just have to put up with it.

        This is somewhat a Good Thing, and somewhat a Bad Thing. The latter is self explained, but the Good Thing about it is that the kernel developers are free to make not only good code, but gr
    • It should work fine. I ran 2.6.10 for a day or two and it compiled fine ("sudo nvidia-installer -f" to force a recompile). maybe you should run another make on your kernel sources.

  • Mirrors not caught up yet (Score:4, Informative)

    by philkerr ( 180450 ) on Wednesday March 02, 2005 @10:29AM (#11822867) Homepage
    The UK mirror isn't showing 2.6.11 yet, perhaps it might be best to wait a little bit so they catch up instead of hitting kernel.org

  • nVidia drivers don't quite work out of the box (Score:4, Informative)

    by Xpilot ( 117961 ) on Wednesday March 02, 2005 @10:29AM (#11822868) Homepage
    However, Con Kolivas maintains a patchset for desktop users which incorporates a fix that allows the nVidia drivers to work at his kernel patch page [optusnet.com.au]. If you don't want the other stuff and just the nVidia fix, you can find the patch split out [kolivas.org], and instructions on which patches to apply in his announcement of his patchset release [bhhdoa.org.au]. Check out the -ck patch though, it has a lot of cool stuff.

    (yay, I actually got a story submission in...hi mom!)
    • ck is a nice patchset. For something with slightly more, CKO [p.lodz.pl] offers everything ck does with Reiser4, Supermount, Alan Cox's -ac patchset, software suspend, updates to libata/ALSA/Bttv, and more.

      If anyone knows how to donate small amounts of money to the developers, please let me know: both ck and cko are on my list of projects to eventually donate to (linked to from my URL).

  • ACPI suspend? (Score:5, Insightful)

    by idlake ( 850372 ) on Wednesday March 02, 2005 @10:29AM (#11822869)
    Does ACPI suspend work on more laptops? Inability to suspend is a major problem with Linux on laptops right now, as there are more and more ACPI-only laptops. The situation is considerably worse compared to APM, in my experience.

    • Re:ACPI suspend? (Score:4, Informative)

      by lennarth ( 642915 ) on Wednesday March 02, 2005 @10:35AM (#11822913) Homepage Journal
      swsusp2 [sf.net] works like a charm on most modern 'tops.

    • Re:ACPI suspend? (Score:2, Informative)

      by gabbarbhai ( 719706 )
      Does ACPI suspend work on more laptops?

      Don't know about 'more laptops' but yes, as long as you compile your own kernel and put all USB, wireless card, agpgart, and related stuff in kernel modules. Unload these modules before suspend and reload them after. Of course, that also means that your USB stuff needs to be unplugged before you suspend. Works like a charm on Debian Sid and Ubuntu "Whory". No swsusp2 necessary for me.
      --Thinkpad R40 on Ubuntu Hoary (or Debian Sid depending on the day of the week)

      • Re:ACPI suspend? (Score:3, Insightful)

        by ckaminski ( 82854 )
        Why would I have to unplug USB devices if I want to just disable the software support?
      • I have always had the problem on R40, where it suspends OK, but refuses to wake up 'cause it doesn't recognize the LID button. There was a kernel patch that should take care of that, but I never succeeded to get through it. Did you experience this problem? And, is it gone in 2.6.11?

    • Re:ACPI suspend? (Score:3, Informative)

      by jamesshuang ( 598784 )
      If you can get your laptop to suspend to ram, but waking up results in a dead screen, perhaps this program would help: http://www.srcf.ucam.org/~mjg59/laptops/video-post _0.1.orig.tar.gz It does the video-post that most laptops don't do when coming back up, resulting in dead screen, but working CPU after an S3 suspend
    • Based on http://www.ubuntu.com/wiki/HoaryPMResults [ubuntu.com], a decently large number of machines will do ACPI suspend to RAM now. You probably want vbetool [ucam.org] to restore video state after resume.

  • Lotsa good stuff (Score:2, Informative)

    by ThoreauHD ( 213527 )
    I like the serial and usb2 fixes. Looks like it's tested as a pretty stable revision. If anyone gets this installed before I do, post some impressions if you would.

  • Not -so-natural high (Score:4, Funny)

    by necrodeep ( 96704 ) on Wednesday March 02, 2005 @10:30AM (#11822878)
    Kernel junkies of the world unite! Your next fix has arrived!
  • So it's now _officially_ all bug-free.

    Torvalds, you scoundrel you!
    Next you'll be telling us the kernel was made by the toothfairy for a lower TCO than windows...
    Oh wait..

  • NVidia Driver patches for 6629 with 2.6.11 (Score:5, Informative)

    by spankers ( 456500 ) on Wednesday March 02, 2005 @10:32AM (#11822891)
    zanders at nvForums has posted patches to improve performance with 2.6.x kernels. Here's the thread:
    http://www.nvnews.net/vbulletin/showthread.php?t=4 6676 [nvnews.net]

    This is the cumulative patch:
    http://www.nvnews.net/vbulletin/attachment.php?att achmentid=10558 [nvnews.net]
    • Isn't the NVIDIA driver closed? How does one patch a closed-source driver? I admit, I have very limited knowledge of how the NVidia drivers actually work. I had been using them up until I installed Mandrake 10.2B3 the other day. The nv driver that Mdk10.2 installed works like a dream, though. I assume its 3D performance is sub-par, but I don't use my computer to play games anyway.

      • Re:NVidia Driver patches for 6629 with 2.6.11 (Score:2, Informative)

        by Anonymous Coward
        The driver itself (all the interesting stuff) is closed. There's a glue layer thats open to bind the kernel and the binary nvidia driver together, that HAS to be GPL licensed.... so they are all patching the glue layer, not the actual driver. //fatal

  • A kernel patch for supercomputers ? (Score:5, Funny)

    by Gopal.V ( 532678 ) on Wednesday March 02, 2005 @10:33AM (#11822897) Homepage Journal
    TFA mentions that

    >InfiniBand, which is derived from its underlying concept of "infinite bandwidth," is a switched fabric interconnect technology for high-performance network devices that is common in a number of supercomputer clusters.

    So that works only for supercomputer clusters ?.

    Interestingly, the ChangeLog has some very small number of entries. The one I found most fun was:-

    Randy Dunlap:
    o [ide] make 1-bit fields unsigned
    I mean, other wise they would end up as "-1" or "0" (when you assume in code that "0" or "1" for 1 -bit fields). How did a sign-extension in the IDE (must be heavily used) be missed till version 2.6 ??. Typically, this looks like the average release - some bug fixes and a couple of big features which nobody (well almost nobody) would use on their boxes.
    • re:signs

      it'd be fine if they've been using "eq 0" and "ne 0" as tests.
    • I remember working with a guy who always added an extra bit to his bitfields 'because they always came out negative'. Somewhere in his (lack of) formal education someone forgot to tell him about unsigned ints...

      When I joined he'd been working with the company for 2 years and was their senior developer. You can imagine what a state the code was in... I basically deleted the lot and rewrote it (which pissed him off no end but pleased all the other developers).
    • It only works for supercomputer clusters only in that, if you connect your computers with infiniband stuff, you might as well start calling them a supercomputer cluster.

      Assuming you use the 1-bit fields with "if (field)" or "if (!field)", it doesn't matter if they're signed, except that you'll get a warning (but the desired effect) for "field = 1". Problems come up when you use the value as a larger type, but that's not a good idea for code clarity reasons anyway, when the field semantically stores either
  • ...for everyone who didn't know what it is as well:

    http://www.oreillynet.com/pub/a/network/2002/02/ 04 /windows.html

  • I'm new to Linux (only just tried Ubuntu and had a bad exprience)... can anyone explain what use the kernel is to the average user? I thought it was just used for a base of distros... It doesn't make sense to be able to upgrade an OS mid... use..?

    • Re:Help? Wikipedia to the rescue! (Score:5, Informative)

      by UnderScan ( 470605 ) <jjp6893&netscape,net> on Wednesday March 02, 2005 @10:57AM (#11823097)
      From http://en.wikipedia.org/wiki/Operating_system [wikipedia.org]:
      In general, the operating system is the first layer of software loaded into computer memory when it starts up. As the first software layer, all other software that gets loaded after it depends on this software to provide them with various common core services. These common core services include, but are not limited to: disk access, memory management, task scheduling, and user interfacing. Since these basic common services are assumed to be provided by the OS, there is no need to re-implement those same functions over and over again in every other piece of software that you may use. The portion of code that performs these core services is called the "kernel" of the operating system. Operating system kernels had been evolved from libraries that provided the core services into unending programs that control system resources because of the early needs of accounting for computer usage and then protecting those records.
      So that is the OS and the kernel. A new kernel version is new drivers and updated system services, which is a good thing. This is not the same as upgrading Win2000 to winXP or changing Linux distros as those involve many many more programs, libraries, & systems as compared (what is collectively known as an Operating System) to a kernel.

  • Score thus far (Score:5, Funny)

    by kmartshopper ( 836454 ) on Wednesday March 02, 2005 @10:55AM (#11823083)
    Linux 2.6.11 SCO 0 Better luck next time

  • Info for the masses (Score:5, Informative)

    by brsmith4 ( 567390 ) <brsmith4@gmail. c o m> on Wednesday March 02, 2005 @10:57AM (#11823104)
    For those of you that are unaware, since the poster doesn't explain at all what Infiniband is, I will explain it for you.

    Infiniband is a high-speed, low-latency interconnect used heavily with beowulf clusters (currently). Infiniband, like Myrinet, addressed many of the problems that are inherent with using interconnects like ethernet.

    The biggest problem with any TCP/IP based transport, in the world of supercomputing, is latency. The amount of error checking that is involved creates latencies that bring fine-grained (lots of memory reads/writes/swaps) calculations to their knees. As many clusters use MPI (Message Passing Interface) for sharing memory between nodes, a low-latency interconnect was needed to replace ethernet and TCP/IP. People have worked on reducing latencies over ethernet by designing raw transport stacks, relying on the switch and the quality/brevity of the ethernet connections (using short, shielded cables proved useful), to ensure accurate data transport, but none of these methods have proven viable.

    Infiniband has also been used as an interconnect for network storage devices as there are obvious advantages to this; eliminating much of that latency makes reads and writes to a device much simpler thus reducing overhead and improving overall throughput.

    More information on Infiniband can be found here [sourceforge.net] at the Infiniband sourceforge page. This should give a sufficiently technical overview of what it does without any of the marketing talk.
    • A lot of current development is going on at the OpenIB project [openib.org].. in fact, that's where the actual 2.6.11 IB kernel bits came from.

    • It's been a couple of years since I looked at IB, but I thougt it used IPv6 addressing (or some close variant thereof).

      Side note: when I worked at a Fibre Channel startup, some of our guys left to go work over at TopSpin.

      Tracy, Meher, if you're reading this -- Hi!

  • psmouse.c (Score:3, Informative)

    by MrNemesis ( 587188 ) on Wednesday March 02, 2005 @11:00AM (#11823124) Homepage Journal
    Does anyone know if any work is being done on the rather irritating problem I (and a fair few others) get with the 2.6 input system? Every time I use a mouse that goes through a KVM switch or a USB->PS2 adapter, the mouse would spazz around crazily and syslog would fill up with:

    lost synchronisation, throwing [1|2|3] bytes away

    Adding psmouse.proto=imps made the problem go away for most usage, but it still occurs under very heavy load, which makes mplaying UT impossible :(

  • Time to... (Score:3, Funny)

    by Cyn ( 50070 ) <cyn.cyn@org> on Wednesday March 02, 2005 @11:03AM (#11823159) Homepage
    ... update your SCO licenses! /rimshot

  • Show of hands... (Score:3, Funny)

    by Cyn ( 50070 ) <cyn.cyn@org> on Wednesday March 02, 2005 @11:07AM (#11823198) Homepage
    Who had been running 2.6.9 or earlier and just finished making a 2.6.10 kernel for the first time (e.g. because they suddenly needed new hardware support).

    *rhand* *grouse*
    • Even worse, I've been scratching my head, trying to get 2.6.10 to work.

      Until recently, I was running, compiling 2.4.x kernels with no problem; I then moved to FC3, and for the first time, was using 2.6.x kernels, and first time using grub.

      2.6.10 builds for me, and everything. I build the kernels, build and install the modules, and since I have SCSI, I run initrd; I add the kernel and initrd.img line to my grub.conf. However, I still get a kernel panic telling me that it can't find the system at boot..
      • I just recently went through similar. Is it complaining about pivot_root failing?

        If so, log in via rescue mode create a initrd directory in / . If not, well, dunno then.

  • ChangeLog (Score:3, Informative)

    by Sunspire ( 784352 ) on Wednesday March 02, 2005 @11:22AM (#11823393)
    The linked changelog is only the changes between 2.6.11-RC5 and the final version, that's why it's so short. Is there a complete changelog available somewhere?

    Also, does anyone know what the status of inotify support is? I think a lot of people would be glad to see it merged, as apps like Beagle require it and the new Gamin daemon (a FAM replacement) should work much better with it.
  • There doesn't appear to be a full changelog yet, but I have been following the release candidates, and it appears that Parallel ATA (AKA "Ultra ATA") Hard disk support is still not in the kernel.

    This is frustrating. I had purchased an Ultra ATA Hard disk drive (which came bundled with a Serial/Parallel ATA controller), and I had it working fine under SUSE Linux 9.0. What I didn't realize at the time was, Promise made proprietary drivers for SuSE Linux, and no other distro.

    I have wanted to switch over to

  • thinkpad? (Score:3, Interesting)

    by CAIMLAS ( 41445 ) on Wednesday March 02, 2005 @11:40AM (#11823605)
    anyone know if the X series of thinkpad IBM laptops will suspend properly now? mine doesn't return from sleep, and won't even begin to suspend to disk properly. (thinkpad x30)

    then again, X goofs up for me with the bios 'hybernation' feature too. though I think that's independent, as it still "works" - just with screen garbage on resume.

  • I just compiled the 2.6.10 kernel last night. I mean after several attempts over the past few years I finally got Gentoo installed (previous attempts had dependency clashes which I as a Gentoo newb didn't want to deal with). I must say that the documentation has improved though I still don't understand why the don't provide a few more scripts (hopefully cobining the partitioning, mounting, and chrooting phase --- I always seem to miss a step for forget to mount one of the partitions *GRR*).

    I guess I could

Slashdot Top Deals

Love may laugh at locksmiths, but he has a profound respect for money bags. -- Sidney Paternoster, "The Folly of the Wise"

Close