Home > Linux > Linux: 2.6.1-Rc3 Released Linux: 2.6.1-Rc3 Released Eugenia Loli 2004-01-08 Linux 26 Comments Linus Torvalds has released Linux Kernel 2.6.1-rc3. This release includes mostly small bug fixes. About The Author Eugenia Loli Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker. Follow me on Twitter @EugeniaLoli 26 Comments 2004-01-08 8:01 pm correction, it was a local root exploit not a remote one 2004-01-08 8:29 pm A friend of mine had a spare PII 500 mhz with 128 megs of RAM running RH 9 and didn’t ever expect it to be responsive or fast. He’s a newbie and asked me to come over and compile 2.6 for him. The system after the install felt like it received a hardware upgrade. It was /much/ more responsive in every area, and apps launched faster too. Wow, I love my 2.6. 🙂 2004-01-08 8:40 pm 2.6 is an order of magnitude better than 2.4 in many, many areas. 2004-01-08 8:48 pm Is this related at all to the recent remote root Linux exploit? Because I was under the impression that that particular problem was fixed in RC2. It was a memory management error, and could lead to a local user being able to do whatever at kernel level. You won’t see “remote exploits” targeted at the Linux kernel. 2004-01-08 9:07 pm Looks like mister 12.242.164.— still doesn’t understand the difference between local and remote exploit… Anyway, kernel 2.6 is a lot faster, but it seems to interact really badly with the new kwin III in KDE. Any KDE 3.2-beta2 users also running 2.6? Do you notice resize/expose being a lot worse than in 3.1? Anyway, 2.6.1 is going to be a very exciting release. It includes a lot of the -mm patchset work, which should make the VM even more robust. 2004-01-08 9:27 pm The mremap bug was fixed in rc2. In parenthesis, Linus explains that he included rc2 changelog. 2004-01-08 10:02 pm This was already argued to death. Can we just ignore the troll? I actually have a problem. I tried applying the RC2 patch to RC1, and it wasnt applying. This is my first test of an RC kernel. Are you suppose to apply the RC patches over the other RC patches, or to the vanilla source? 2004-01-08 10:11 pm over vanilla. those are not incremental patches. 2004-01-08 10:30 pm Yep, on paper, 2.6.0 IS really faster than 2.4.x, thanks to preemption and scheduler. But with some hardware, I’m faster to a kernel panic too: ISDN-Card, USB-Bluetooth, USB-Chipcard-Reader… really no exotic things and they work beautiful on 2.4.x Oddly enough, my Asus (P4P800VM) doesn’t boot with PNP support compiled in which is required by my ISDN-Card. But don’t work at my office with PNP enabled too 🙁 Again, with 2.4.x I’ve no problems with PNP compiled in… Linus is right. He ask, why I use and what I expect from a .0 Software… I think, now it’s the time only to read the changelog of 2.6.0. Maybe on my 2nd PC – where I only need IDE, Sound, Graphics and Network – I will use 2.6.0 2004-01-08 11:21 pm I run the combo. I can’t say I notice very much speed-improvements over what I did with 2.4 / KDE-3.2. Don’t know if that’s because I have several Gigahertz and a GB of RAM, cause a friend of mine on a 1.2GHz (or so) system says the speed-improvement is awsome! What have bothered me after the 2.6 upgrade is that if I get network-traffic at 100KB/s+ over a timespan of some few minutes, the _WHOLE_ system is halted. Don’t know if that’s a problem in the kernel itself, or in the Orinoco-drivers for my wlan-card. I used to use the hostap-drivers, but couldn’t get them to compile with 2.6, so now I’m using the orinoco-driver instead. Whatever is causing it, I hope it’ll be fixed soon =) 2004-01-09 12:31 am I have 1 ghz and I see can see the difference. I think the slower the PC the more of a speed boost you’ll notice when compared to 2.4 kernel. Oh yeah I am running MDK 9.2 with 2.6 kernel. 2004-01-09 12:51 am Installing 2.6 made it seem like I trippled my processor speed, doubled my memory and quadripled my I/O devices performance. Very impressive. Who says you need to upgrade your hardware to get more out of your system? 2004-01-09 2:20 am You sound like an infomercial. And already, Linux Kernel 2.6 is off to a rocky start, with a vulnerability just weeks after its release. 2004-01-09 3:04 am A local vulnerability, which was patched quickly – more quickly than what you’ll ever get out from Microsoft. Quit trolling. 2004-01-09 4:55 am ===== I like the facts, L$ or Lin-sux is breached more than Windows. ===== Where do you get this from? Do you have some kind of URL where I get the picture? ===== Check it out, Apache is attacked and hacked, but little is said about it. ===== Oh … and Apache only runs on Linux, right? There isn’t even a small chance that Apache would be running or used by some ISPs on, say … Windows, Solaris, Free/Net/OpenBSD, etc? Right? Besides, with 67,4% market share Apache is by far more often used for website hosting than Microsoft IIS which has only 20,8% market share. If you don’t trust the numbers, check netcraft. 2004-01-09 5:44 am Look, the little troll got pissed. aww, how cute. 2004-01-09 5:50 am Of course, everyone will have recognized the famous name-stealing troll. Again, let me make things clear: the Real Archie Steel is pro-Linux. The Fake Archie Steel (or should I say “Steal”) is an anti-Linux troll. These two posts are therefore fake: http://www.osnews.com/comment.php?news_id=5599&offset=15&rows=21#18… http://www.osnews.com/comment.php?news_id=5599&offset=15&rows=21#18… To annoy him, I’ve adopted a new nickname that he can’t use without being contradictory. I’m not going to respond to anymore of his comments – anyway, as they’re typically flame-bait, they usually get reported and modded down. @Anonymous (IP: 12.242.164.—) There is no need to be so hostile towards the facts. The fact is that the vulnerability you’re referring to was a local one, and this is quite different from a remote exploit, something which you’ve yet to recognize. The fact is also that it was patched quickly. If someone’s afraid of the facts here, it’s you, not me. I know they sometimes fly in the face of your blind worship of said kernel, I don’t worship Kernel 2.6, I don’t use it yet. Anyway it’s just an OS. You’re the zealot here, trolling Kernel threads to harp on its alleged “insecurity.” but sometimes the truth hurts. Is that why you seek to distort it? No one mentioned Microsoft besides you, but since you did bring question the superiority of Windows, I’ll assume you were trying to form a real sentence here, but in fact you are quite mistaken. I did not address Window’s technical merits here, but rather the speed at which Microsoft has usually fixed vulnerabilities in the past. As I’ve often said, I’ve got nothing against Windows. I prefer Linux, but I work with both. Daily. I probably know Windows better than you do. I do have a strong biais against Microsoft, due to their abysmal corporate behavior. Again, you’re trolling: trying to pick a fight by making provocative statements. And the fact that there are Linux trolls on Windows thread is not an appropriate reason for your behaviour. “Others do it too” has never been a good excuse. here are some more facts for you to munch on: Yes, I’ve read this article already. I follow Linux security news regularly. See, if you read the articles instead of just the headlines, you realize that while Linux machines are more often the target of active attacks (i.e. actively conducted by a cracker) by a factor of about 3:1, Windows are more often the target of passive security threats such as viruses, worms and trojan horses. For your information, a Trojan Horse is a program that gives remote access to a hacker, but that propagates like a Virus. One could argue that this is proof enough that Linux is more sophisticated than Windows, since a cracker can have a program do the job for him on a Windows box, while a Linux box requires his active participation, but that’s not really the point (even though it’s probably true). According to the article you cited, damage due to overt attacks (the one in which Linux lead) are estimated at 707 million US$. When one includes damages from viruses and worms (which are essentially a Windows problem, since there are proportionately 50 times more viruses for Windows than Linux), the price tag goes up to 28.2 billion US$. In other words, the economic damages due to Windows’ vulnerabilities is about 60 times those due to Linux’s. Since Windows servers outnumber Linux servers 3 to 1, one can say that, from a security standpoint, Windows servers cost 20 times as much as Linux servers. It’s no wonder “thrustworthy computing” is the mot du jour at Microsoft – and why Linux is growing at an accelerated pace in Server space, despite MS’s FUD campaigns (the best money can buy!). Those are the facts, which clearly you aren’t basing your opinions on. Any more links for me? 🙂 2004-01-09 5:53 am The name-stealing troll posted an additional comment: http://www.osnews.com/comment.php?news_id=5599&offset=15&rows=24#18… Moderators, please send this troll down below (and this comment as well). 2004-01-09 6:07 am the supposed increase in speed is upposed to come from tyhe fact (afaik) that the kerenl now checks for user input more frequently, thats means that if you have a fast computer its already checks more often than is necessary so no noticeable difference (probably a small boost if you lowered it a bit in the kernel or something) but on a slower computer this should be more noticeable… an example the difference between earning $20,000 and $40,000 a year is a larger real world difference than $1bln and $2bln. just an idea… my linux kernel book is still in the mail ps- is it perhaps time for a passworded login for this forum. these trolls are really ruining things. sadly it appears that no one has anything ever good to say about micrsoft/m$ and from lacking that perspective we are missing out (trolls don’t say anything good, they just make it look worse by association) 2004-01-09 7:34 am Your message says one thing, while your name says another: you’re not making any sense. Your post is a contradiction in itself. You took the bait – hook, line and sinker…as long as you’ll keep stealing my name (because you fear what I have to say), you’ll identify yourself as a fake. Checkmate! Poor troll. No one takes you seriously, and no one believes that you are the real Archie Steel (that would be me). Proof: I don’t even need to report you, others do it. Mods, please mod this post down so it can join those of the name-stealing troll. 2004-01-09 8:42 am Funny how ChocolateCheeseCake (IP: 168.143.123.—) and Archie Steel says: Linux Rules (IP: 209.234.157.—) use the same e-mail address: Lin_sux@evil.com http://www.osnews.com/comment.php?news_id=5601#185661 http://www.osnews.com/comment.php?news_id=5599&offset=15&rows=22#18… http://www.osnews.com/moderation.php?news_id=5599#185654 IIRC he posted here too with an acronym for NetBSD http://www.osnews.com/comment.php?news_id=5598 I’m wondering if it’s the same as the Anonymous (IP: 12.242.164.—) troll. 209.234.157.* and 168.143.123.* also troll on http://www.deadly.org Here are is an example: http://www.deadly.org/article.php3?sid=20031121083351 Both ranges are owned by large ISP’s in the USA. I’m wondering, is it a proxy IP? Is it dynamic dialup? 2004-01-09 10:10 am who cares:) 2004-01-09 1:11 pm In fact linux kernel 2.6.1 final is released. 🙂 2004-01-09 3:08 pm At last, we have a happy conclusion to the security nightmare that was 2.6.0. I’m glad I didn’t upgrade my production servers, despite the claims that the parity of the version number indicates stability. 2004-01-09 6:52 pm I thought the kde/gnome applications and mozilla load slowly because the dynamic linker (ld) is slow in looking up unresolved symbols? Which kernel improvements in 2.6 make applications start faster? It would be nice to see how much faster mozilla loads (first time load) under 2.6 as compared to 2.4. 2004-01-09 7:05 pm At last, we have a happy conclusion to the security nightmare that was 2.6.0. I’m glad I didn’t upgrade my production servers, despite the claims that the parity of the version number indicates stability. Somehow I doubt that you really have Linux “production servers”, seeing as how you keep trolling the Linux threads. In any case, you’d be a really poor admin if you were to switch all of your servers to a .0 kernel. A good admin would never take that risk, and would rather wait until the Kernel has matured enough. Also, if it works well, why upgrade? This isn’t Microsoft products we’re talking about – there are still servers running the 2.2.X kernels out there. In any case, this is another irrelevant criticism as the bug you’re referring to is a local exploit, something which you seem determined to ignore (therefore proving to all that you know very little about computer security). Even if you had upgraded your imaginary production servers, this specific bug couldn’t have been exploited by a hacker over the Internet. So quit spreading your FUD. Incidentally, for a desktop, one can use 2.6.X without too much problems, although you need to do some tweaking as not all apps are optimized to run with it (and a few don’t run, such as lm_sensors). I’ve tried it and it works well, though I am going to wait a few version numbers more before using it.