Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Linux Software

Linux 2.2.2 Released 116

Posted by justin++ from the die-bugs-die dept.
Dave Miller, the ultracool hacker who does (definitely) work at Red Hat, wrote in to let us know Linux 2.2.2 has been released. A summary of changes should appear on Cutting Edge Linux shortly. This new kernel should fix the recent autofs bug (fixed in ac7). Update: 02/22 11:35 by J :Dave Whitinger, the guy behind the fabulous Linux Today site, wrote in with the linux-kernel-announce(ment). Update: 02/24 03:13 by CT : Dave Miller does work at Red Hat.
This discussion has been archived. No new comments can be posted.

Linux 2.2.2 Released More

Linux 2.2.2 Released

Comments Filter:
  • Well, it's 2/22 and kernel 2.2.2 is out. It makes me wonder if anyone planned this. Interesting, eh?

    (Maybe we'll get 2.2.3 tomorrow. *grin*)

  • :( (Score:1)

    by Matrix ( 290 )
    Too bad that I still have problems with 2.2.2 and sound. The Ultrasound project won't compile on 2.2 and OSS sounds VERY staticky on one channel .. Would anyone by any chance know why OSS would be staticky in one channel but the Ultrasound project sounds fine in both?
  • Wow... it's like a Little Caesar's commercial... "twotwotwo!"
  • a way to ensure integrity of a filesystem in case of powerdown or other calamity.

    When a file is written, an entry is first made to the journal indicating what was to be written, then an attempt is made at writing to the filesystem, if the write was successful then the journal entry is removed or marked done else the write can be attempted again or rolled back.

    In the case of an error in the filesystem, the journal can be used to bring it back into sane state.

    Think of journaling as a form of transaction processing for a filesystem. This is my understanding of the process.

    NTFS is a journaling filesystem with advanced Access Control Lists. It's the best and only real thing that NT has going for it; too bad the rest of the OS has so many problems. When NTFS read/write is available for Linux then we can say that Linux is a better NT than NT? --- Note play on words... :)
  • Posted by a VALUED MEMBER of the space m:

    ftp://pulsar.midsouth.rr.com/pu b/linux-2.2.2.tar.gz [rr.com]
  • Posted by chiem:

    Actually, I won:

    chiem@blowfish:~[1]$ uptime
    1:23am up 2:49, 3 users, load average: 0.11, 0.14, 0.07

    Which means I had it done at ~10:33 pm on the 22nd.

    --k
  • Is that the same Dave Miller that used to run a BBS on an Apple //gs, or am I suffering from Small World Syndrome?

    --Threed

  • 2.2.2 (Score:1)

    by robin ( 1321 )
    [...]
    Some time has to be fixed as the standard, after all.

    Typically, UTC (Universal Coordinated Time, a very close approximation to GMT) is taken as the standard time. I gather astronomers prefer sidereal time, possibly even in America. HTH.

    Oh yeah, this conversion table [bcca.org] might come in handy some time.


    --
    W.A.S.T.E.
  • Actually Linux users should change their underwear everytime they recompile the kernel.
    Windows users should change them everytime they reboot.

    heh heh
    Kashani
  • Yes, it does. Not only that but it'll make you richer and take away that sick personality streak.
  • when loading the module aha152x:

    bug: kernel timer added twice at c021a72d (of course, that last "thing" changes any time I try to load the module)
  • Unlike 2.2.2-pre5, it builds on Alpha(164LX, at least).
    I'm rebooting now to see if it works...
  • For the following reasons (to start)
    1. Long-time users of linux
      1. Many who follow the "bleeding-edge" have multiple machines.
      2. The majority don't follow the "bleeding edge" anyway.
    2. Newbies (apply rule 1.2 above)
    3. No-one is going to accept your patches.
  • Anyone know when they are starting the 2.3 tree? I'm quite content with my shiny new 2.2 kernel (my /dev/sound finally works!), but I'd also like to see what they've got in the works for the future. Have they begun a tentative feature list for 2.3?

    --Lenny
  • A journalling filesystem would be nice, although that will likely be saved until Linux 3.0 . ext2fs seems pretty decent, though I know little about the topic. Hmmm...wonder where I could find some good literature on filesystems (theory and comparisons)?

    --Lenny
  • I wouldn't go out of my way to to do it, but besides that, why the hell not? I'm not about to bring down my router box to upgrade the kernel (it still runs 2.0.36) until I'm sure the upgrade would be without serious complications. (I've heard about troubles w/ ipchains and ipamasq) But for my personal box, I figure any update to the stable kernel, especially in the early stages is a pretty good idea
  • Alan Cox mentions on www.linux.org.uk that realplayer is not cool with the new sound stuff or something. In any case it doesn't work quite right. He says some people already have a workaround. I couldn't find where this was discussed -- does anybody know what the workaround is? Or do I have to go browse the kernel mailing list archives...

  • Turn off "Socket Filtering" in your configuration.
    Or CONFIG_FILTER in your .config file.

    This will #ifdef out the socket filtering code,
    which is the culprit. I'm gonna see if my feeble
    C/kernel skills can do anything about it
    tommorrow. But I am sure it'll be fixed before
    I wake up.

  • I have been recently. (for 2.2)
  • Thats _really_ cool.
  • Provided of course that you have enough memory to avoid swapping... NT is very swapaholic with less than 64MB.

    Oink! Oink!
  • "Uncanny" would be more appropriate.
  • Yes, but be sure to include penicillin.h.

    Brian
  • I believe Dominic Giampalo (the guy behind BFS, the file system underlying the BeOS) has written a book by the name of "Practical File System Design with the Be File System". Despite the name, my understanding (note: I haven't read the book) is that it's a bit more generic than it sounds...

    - Darchmare
    - Axis Mutatis, http://www.axismutatis.net
  • Quake 2 works intermittently for me. It seems to work better if my sound modules aren't loaded yet. If the sound modules are loaded, and I try to start up Quake 2, I get nothing. It starts to load, and switches video modes, but then freezes. If I kill quake2 from a telnet session, things go back to normal, but attempting to kill it from the machine itself is nearly impossible because you can't do anything but SysRq commands. And those don't always bring the machine back into a usable state. Check here [nols.com] for information on my setup.
  • I just booted up my new 2.2.2 kernel and tried Quake II and it froze up the whole system, just like 2.2.1. what's up with this? I remember reading /. comments about this when 2.2.0 was released, but I guess either the issue hasn't been addressed, or it's just considered a serious bug in quake....
  • It's me again. I just realized that I should have mentioned what kind of system I have.

    Hardware: generic ET6000 / 2.25 MB; SB AWE64 PnP.

    Software: linux 2.2.2; glibc 2.0.7; svgalib 1.3.1; Quake 2 v. 3.19 (glibc).

    Quake 2 works fine with Linux 2.0.33, but freezes in 2.2.[12], leaving the system unusable so I have to reboot.

    Based on the comments so far, it sounds like this has to do with sound or glibc or both.

    If anyone else wants to share his experience, I'd appreciate it.
  • hmm... what is a journaling FS?
  • if some people recompile a kernel everytime an update is issued. =]
  • I'm pretty sure Hurd can do this, the platform-independent MkLinux part can be restarted without taking down the whole Mach system. But, is the uptime stored by MkLinux or Mach?

    --
    Stephen Pitts
  • Dunno about the sound...But I had the same problem with APM (running redhat 5.2). Actually its not a kernel problem, but its an inittab thing. Edit /etc/rc.d/init.d/halt. Go to the last line and change it from:
    "eval $command -d"
    to
    "eval $command -d -p"

    This fixed my APM problem.

    Hope this helps...

    -Bubba

  • I'm running a SB16 PnP which worked fine under 2.0.36, but which doesn't work under 2.2.1. I get the fabled "device or resource busy" when initializing the sb module (it's choking on the DMA address). Has anyone had any luck with a similar setup? Is this fixed in 2.2?

    wog

  • sndconfig is broken in 2.2.x for SB cards because the module expects different arguments than it did in 2.0.x

    I have already run isapnp and modprobe -- isapnp works, but modprobe fails with "device or resource busy".


    wog
  • That's a nice idea, but will be difficult without an extremely modular kernel. What would be interesting is the ability to swap in and out modules that are being used; i.e. upgrade to a new (for example) scsi.o or e2fs.o. Remember, a modular kernel (like Mach) is slower than a monolithic one (compare OS/2 Version 4 to Version 3; compare Windows NT to 9[58]; compare Linux to AIX). I have nothing against Mach (or Hurd), but they do compromise performance.

    I wouldn't be surprised if someone found a way to create memmgr.o and then change that while the system was running. OK, so I would be impressed.

    Cheers,
    Joshua. (Geez, it's late.)

  • I can't get any mp3s to play correctly through 2.2.2.

    cd-audio works fine, quake audio plays fine, but the mp3s loop the first 0.5s of the mp3.
    both x11amp and mpg123 fail.
    reboot into 2.2.1, mp3s work fine.
    What's up with this?
  • Enough said... I would recommend ALSA... I had similar problems... Under OSS/Free, I could not get it to play sound on my GUS PnP without the thing spewing out tons and tons of DMA/IRQ errors, even though my DMA/IRQs were okay. IMHO, OSS/Free is what I would call crippleware, masquerading as open source... ALSA rules.
  • For me, the 2.2.2 kernel was up for maybe 5 hours before it hung: I used mpg123 to play some mp3's (mpg123 has worked with 2.0.x, 2.1.x, and 2.2.1). When the last mp3 was finnished, there was some debug information, but no shell prompt. The Alt-SysRq-stuff didn't quite work: I usually do S-U-B (Sync-Umount-Boot) in situations like this, but this time the 'S' only printed out more debug-information, the 'U' didn't do anything (I tried it three times). The 'B' rebooted as expected though (and it spent 25 minutes fixing the disks).

    I have a Cyrix CPU. I've had problems with the 2.2.1 kernel too (random hangs and reboots). Does this have to do with Cyrix maybe? I've read articles about some undocumented 'hang-cyrix-opcodes', but why would any program include anything like that?

    Gavare
    (Let's just hope I can post this before the system hangs again ;)
  • I will beat you to it and be the first to compile.


    "Nothing great was ever achieved without enthusiasm." - Emerson

  • What? (Score:1)

    by X-Type ( 15655 )
    I do not understand.
  • I'm still using a libc5 system with kernel 2.0.36 (on an i486). Is it really worth the effort to try the move to 2.2.x? I'm pretty happy with the state of the system now, and unless there's some really super-duper reason to do the upgrade boogie (like security problems) I generally leave things be.

    --
  • If you are on 2.0 and are leary about using ipchains, try: http://www.animx.eu.org/linux/ipfw [eu.org] It's a simple perl script I wrote. I didn't like the ipfwadm command too well, so I wrote this. It's pretty basic (it'll only do what ipfwadm -l or whatever it is can list. At the time I didn't know about the more verbose listing of rules that ipfwadm had). This script can build a rules file from your current rules (just run the file that it outputs. it sets it as executable) to setup the rules with the new kernel. It does rely on the files in /proc to work. It can also auto detect which program it needs to run by looking at what files are found in /proc.

    I would love to hear if it helps anyone, it's been pretty private (Never found a good place to post it), but I release it under GNU GPL! If anyone makes a change, I'd love to hear about it as well.

    mail me [mailto].

  • I think I figured out what it was. I noticed quake2 not working as well, strace told me it was stopping on a read.. TURNS OUT it's the friggin mouse.. Just move the mouse and it works. (for me atleast)
  • > "640k ought to be enough for anybody." - Bill Gates

    AFAIK Gates never said this. This sentence was uttered by some chairman of IBM who was responsible for hardware limit of 640 KB of directly adressable memory. Again, AFAIK these words were said when he was asked if he thinks 640k limit will have the same fate as 64k limit did. As we all know 640k limit got to the same place but much faster due to accelerated speed of technical progress.
  • Compiles just fine on my Celeron-A... haven't rebooted yet to see if it actually works, but it compiles!

    later,
    Robbie
  • Nope, I got the patch and it compiled fine...

    Later,
    Robbie
  • Hmmm... I am, at this point, running 2.2.2 on an Intel system. Compiled with no problems, and I got the patch. I even have loopback enabled... I don't know why it isn't working for you, sorry.

    Later,
    Robbie
  • It compile for anyone else on an intel box? The object file 'loopback.o' isn't being very cooperative. Complains about "dereferencing pointer to incomplete type" in the header file sock.h.
  • I prior to 2.2.2 I had 2.2.1-ac6 so I unpatched the ac6 stuff and then applied the 2.2.2 patch. Tried to compile and got the loopback.o problem. Then got the full sources and tried to compile again. Same problem. *scratchin' my head* Any ideas?
  • i'm running 2.2.2 right now and x11amp sounds just fine over my SB16PnP. It took a little work to get it that way, but once i got it set up in 2.2.0 it didn't give any more trouble. Here's how i did it, maybe this'll help.

    On advice from a HOWTO, i downloaded and compiled isapnptools-1.17. pnpdump --config gave the expected output, so i just redirected to isapnp.conf to save myself some work.

    Compiled the kernel with sound as a module. Ended up with adlib_card.o, opl3.o, sb.o, soundcore.o, sound.o, and uart401.o in /lib/modules.

    Finally, i put in a script to initilize the thing on boot.

    #! /bin/bash
    case "$1" in
    start)
    echo "Setting up sound card..."
    # Plug n Play config
    isapnp /etc/isapnp.conf
    # Install basic sound modules
    insmod soundcore
    insmod sound
    # init'ize uart thing, because sb needs it
    insmod uart401
    # Set up the actual sound card! Pulled numbers out of isapnp.conf
    insmod sb io=0x220 irq=9 dma=1 dma16=5 mpu_io=0x300 type=6
    # Midi support. Note that this automatically loads opl3.o as well
    modprobe adlib_card io=0x388
    echo "Done."
    esac

    exit 0

    i'm not sure how many of those insmods could be done with modprobe, don't feel much like changing it because it works fine as-is for me.
  • I rebooted my Linux box 2 times since i have upgraded to the 2.2.x series (2.2.0 to 2.2.1 and 2.2.1 to 2.2.2) but my root partition can't be unmounted. The log:
    Shutting down system loggers:
    syslogd klogd /etc/rc.d/rc0.d/K70syslog: kill : (119) - no such pid
    This is very strange. I'm running syslogd and i didn't killed it. But it's not the major problem:
    Unmounting file systems
    umount: / : device is busy
    mount: / is busy
    Remounting remaining filesystems (if any) readonly
    mount: / is busy
    The system is halted - System halted... I don't want this device busy every time i have to reboot.
  • I had the same problem using the 2.2.1 (uptime of 25 days). I ran a ps and +-15mb of my ram was eated by something. Today i rebooted and actually i'm using 20mb running X with the same apps when it used +- 38mb. I think that with an uptime of 25 days 27mb is normal, but not 38mb. It's a RedHat 5.1 with all the updates and the Changes required compiled all from scratch with gcc 2.7.2.3. I have 128mb of ram + 128 of swap.

Slashdot Top Deals

Software production is assumed to be a line function, but it is run like a staff function. -- Paul Licker

Close