Slashdot Mirror
Linus Says Pre-2.6 is Coming
gomoX writes "As seen on C|Net , Linus has announced that the pre-2.6 series will be starting in early July. Despite not having been able to meet the release goal for 2.6 in June 2003, the next stable version is not as far away as you may think. You can take your guess based on the fact there was a 9 month period between first test version of 2.4 and the official release of 2.4.0 on January 2001."
In fact, 2.5 isn't that bad right now... certainly, it would be crazy to use it on a production system unless you really know what you're doing[1], but it's quite usable, and the scheduling has really improved.
;)
[1] in which case you probably wouldn't use it on said production system...
So what's the easiest way to use new versions, use something like Gentoo or Linux From Scratch? Jonah Hex
Horror & SciFi Erotic Nudes
(of course, a beta version of Linux is probably more stable than a release version of Windows, but I can actually do something with Linux if/when it blows up in my face.)
Free Software: Like love, it grows best when given away.
So what hot new features can we expect from the new 2.6 Kernel
Even the regular Gentoo kernel has a lot of extra patches in it, including the O(1) Scheduler, and Low-latency scheduling; works great for me.
pb Reply or e-mail; don't vaguely moderate.
I truly believe the time for linux has come. It has always been stable and powerful (as all you slashdotters already know), but now it is really as easy to use as Windows....I tried Mandrake 7.2 but gave up after 1 month for various reasons. Now I have Mandrake 9.1, and I was very pleasantly surprised how polished it was... and easy to use! It is now my primary OS at home. MS, eat your heart out!
I am using a rock solid 2.5.70 since its released and its performs just great! And having Morton and Torvalds at OSDL is a good thing (tm) :)
--
One by one the penguins steal my sanity...
Your "Overt Sauce" model is clearly breaked ! Windows has been way beyond 2.6 (3.1 to be exact) since what ? 11 years ?
KernelTrap is running a story on an interview Alan Cox gave at LinuxUser & Developer Expo 2003 in Birmingham, U.K. A summary of Alan's talk is also available.
Will this be simply a kernel upgrade and I'm running 2.6? Or... will I have to wait for a distro to release their 2.6 version?
"To deny our own impulses, is to deny the very thing that makes us human." - Mouse
not to be nitpicky or anything, but technically, shouldn't future versions of linux be referred to as GNU/SCO?
SP4 does nothing else but fix bugs. Kernel 2.6 adds quite alot of new features so its not really that much of a valid comparsion.
Please, Please, Please don't compare the two. Being an Windows administrator (a minor devil in some pantheons?), I wouldn't wish SP4 equivalent instability generating changes/autoupdate problems upon anyone.
Ok, I give up, why you?
Microsoft deserves to be raked over the coals, and the government IS out to get you.
That doesn't mean that the editors aren't lazy and slower than sloths, though.
-- Darl McBride
What they don't get in timeliness they make up for in volume (dupes).
My question is this:
There was some hesitancy, upon the official release of kernel 2.4, based upon some bugs etc...
I'm wondering, does the kernel - generally speaking - get more and more stable. For example, will the first release of 2.6 be more stable than the first release of 2.4. I realise that there are new additions to the kernel, and with that new problems will probably emerge. However, comparatively speaking, does it make sense that the kernel's evolution will lean towards stability with each release in the cycle, or will it generally be unnoticable?
Just curious.
This wasn't just plain terrible, this was fancy terrible. This was terrible with raisins in it. - Dorothy Parker
I have absolutly nothing to add of a technical nature to this story, so I will delude you with a rambling trip down memory lane (comprised completly of anecdotes from 2.2).
;)
My first taste of linux was phatLinux on my brand new p2-400 (128MB of pc 100 ram I liked). 3 months later I had built a sub 400 dollar computer to play around with and bought (yes paid money for) Linux Mandrake 6.5 from Wal-Mart. From there I began learning about this kernel thing (and my joys when I found make menuconfig and make xconfig, have you ever tried make config? ewww...) Well that went fine and fun, I added options, made modules all the fun stuff you do, but it was still in the same 2.2 vein that came with mandrake. Then 2.4.0 final was relased and I compiled and installed my first new kernel. Everything was new and faster. DevFS was a godsend, the ppp and bsd compression routines made my modem fast (or somthing I went from 2.5 kb/s downloads to 5-6 kb/s after recompiling). Since then I have also come to love dri, premptive and low latency patches, and all these other backported goodies. I am waiting on 2.6 final before I play with any of the new features (I didn't play with 2.3 or 2.5). Ok I am done. And I didn't even mention Gentoo... oh wait... damn.
There is nothing wrong with being gay. It's getting caught where the trouble lies.
Me neither, but the new scheduler is nice, POSIX ACLs look sweet and ALSA included in the main kernel release, and the cryptoapi. Plus all the incremental upgrades to drivers etc.
Well do you know what statics and approxomations are?
Without knowing the detail of a black box system, you can predict future behaviour of the system due to past observations. Of course not certainly, but likely.
IMHO also: thats not insightful!
--
Karma 50, and all I got was this lousy T-Shirt.
The Windows kernel hasn't changed significantly since the NT4 -> Win2K change. The biggest improvement in the XP kernel was pre-faulting the pages of large processes. Meanwhile, in 2.6, the block I/O layer, VM layer, scheduler, and sound system are brand new. And the whole kernel was made preemptible! Shortly after 2.6, ReiserFS 4 (which looks very promising from initial benchmarks) will be released. In all, 2.4 -> 2.6 will be like NT4 -> XP!
PS> Before anyone bitches about rewrites being a bad thing, look at things this way. Such extensive changes are necessary for the continually growing range of systems Linux is expected to run on. 2.0 and 2.2 were greatfor single CPU servers, or SMP machines with only a few processors. 2.4 is very usable for heavy-duty machines with many more processors. 2.6 (along with the changes that help interactivity) will make an excellent kernel for desktop machines and workstations. In 2.8, the focus will be on optimizing the core algorithms to run on large-scale NUMA machines.
A deep unwavering belief is a sure sign you're missing something...
Kernel is as stable as the rocky mountains. Been compiling Mozilla non-stop in gcc for the last 3 hours with no problems.
It'll be nice when it finally comes out, because I'll be able to point people doing audio work to Linux. Right now I have to say "well, Linux is better than Windows for this, but only if you apply the low-latency, pre-emptible kernel, and variable HZ (with HZ set to 1000) patches," which is a bit more involved than most people who are just doing audio work want to deal with. Once 2.6.x comes out I can just point them to the stock kernel.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Jesus. The guy offers some historical perspective and you gotta be all surly... Deep breaths, man.
"You can take your guess based on ....
No you can't. Linus has always maintained that a kernel will be released "when it's done". Why would he change now?
Dude - do you what the word guess means?
guess - a. To predict (a result or an event) without sufficient information. b. To assume, presume, or assert (a fact) without sufficient information.
I also agree - NOT insightful.
Well, the nanosecond patch is critical for make on fast computers, since it uses filesystem timestamps. If you're running gentoo on a brand new desktop it might be a good idea.
The fbdev patch reduces the size of the framebuffer, so if you like framebuffered consoles, it will reduce your kernel size.
If you have multiple processors, the Shared page table patch will help reduce page table sizes, and thereby improve performance, marginally. More RAM = more file cache / less disk paging; shared data -> higher cache coherency = faster kernel performance in memory mapping.
Additionally there seems to have been some mucking around with tweaking the adaptive scheduler so X gets more time when it needs it. The performance metrics have been kind of squishy, but the general consensus is that X and related 'interactive' processes are more responsive.
I Browse at +4 Flamebait
Open Source Sysadmin
I played around a bit with .71-.73, but the big thing that got me was that my mouse speed in gnome was sped up by about 10x over 2.4. I had to set the accel down to the lowest setting in gnome to make it usable, compared to about the 25-50% setting with 2.4. Of course, there is no similar setting for GDM.
I'm guessing this is due to the new keyboard/mouse modules, but who knows. Hopefully this is one of the things that will get shaken out when 2.5 and 2.6 become more mainstream and the KDE/GNOME folks set things up to work nicer with the devel kernels.
Don't forget merger of larger projects into the kernel, namely:
CryptoAPI
IPsec
ALSA
XFS
No more patching the kernel/building module packages for those features!
Also, improved latency with (O)1 Scheduler and other I/O performance tweaks will be native to the kernel.
Finally, 2.6 - and at last support for my Zaurus should make it into the systems at work where 'recompiling the kernel' is a dirty word.
As long as Red Hat build it in to their stock kernel that is.
Beep beep.
There's a lot of complaining about code-freezes for the kernel not being code-freezes. People gripe about major changes being introduced in the last days of the development version.
I think the problem is the standard explanation of 'even kernels are production, odd kernels are development.' Whether he says so or not, it's clear that branching to an even version does not mean that it's a production kernel...branching to an even version begins the code freeze. Up until they call it 2.6, there's going to be large changes to the codebase. Once Linus calls it 2.6, everyone knows they can't put in major changes, but basic bug-fixes only. Therefore, it's never until a few months (or a year) after the even series starts that it's really a production kernel.
Software development managers would hate this...lots of kernel developers hate this...but love him or leave him, that's how Linus works.
I'm sure I'm not the only one who's wondering if Reiser 4 will go into the stock 2.6! So: does anyone know?
I know this is kind of off the topic but could someone tell me if Slashdot has message boards, or just these comments. I'm trying to find some answer for some random question I have regarding a carreer in technology and some question on my network here at home. Maybe I'm blind but I can't seem to find any message boards on the site to ask my questions. If Slashdot doesn't have any message boards then could someone recommend some message boards where true techs hang out? Thanks
Hey moron they are named after the years where they were released. Any self respecting computer user knows that, unless they are too busy with their head up Linus's ass.
I bet you have never heard of an "estimated guess" then?
What about that? Will we be finaly able to switch kernels without a reboot?
:D
I could google for it, but hearing peoples' comments about these things is much more interesting...
Looking for people to chat about multicopters, coding, music. skype: gtsiros
(of course, a beta version of Linux is probably more stable than a release version of Windows
I'm posting this anonymously because I'd be ashamed to have my name tied to defending Windows. Your comment is definitely true for versions of Windows up through ME. But I've been running XP on a Sony Vaio for the past fourteen months, and I have never had to reboot. In fact, I don't even know if XP has a blue screen of death, cuz if it does I've sure never seen it. Despite its thousand liabilities and the fact that Windows helps fund a scummy corporation, I don't think that XP can be criticized for instability.
What's the name of the new filtering tool that will "obsolete" iptables?
I think you mean "educated" guess.
preeminent or pre-eminent ( P ) Pronunciation Key (pr-m-nnt) adj. Superior to or notable above all others; outstanding.
Not gonna comment on parent poster's spelling skills, but you, sir, are in dire need of a dictionary.
You must be the only guy who's never heard of Windows 98.
In Soviet America the banks rob you!
t my mouse speed in gnome was sped up by about 10x over 2.4
great news! 900% speedup from Linux kernel 2.4 to 2.5.75
On my own systems, I noticed a huge improvement in both support and responsiveness when transitioning from 2.2.x to 2.4.0-2.4.10...after 2.4.13, I noticed even further improvement in responsiveness. However, I can deadlock any and all 2.2.x and 2.4.x kernels very easily--run VMWare and try to access a CD-ROM drive. This happens on two wildly different systems and both when using ide-scsi emulation for my cdrecorder and when not using it.
I hope that 2.6.x will yield further improvements, and I'm keeping optimistic on that one. BTW, my prediction is May 22, 2004 for a definite 2.6.0.
No, sorry. But usually statistics show that most approximations are way off, and even more so in the software business.
I'll second that, I did a test install on a workstation at work, and it now suffers much more random crashes.
;)
Guess my work box needed a copy of Linux on it anyway though
Well I was talking about statistics not approximations.
So your are talking about statistics regarding statistics.
Now is there something biting in it's ass?
--
Karma 50, and all I got was this lousy T-Shirt.
I've downloaded Knoppix and I've not been able to get the networking working on my Asus A7N8X deluxe MB. If anyone has any ideas please let me know.
I'm hoping when they upgrade the kernal in the knoppix distro I can finally use mozilla etc. on the net and can see how things work.
Really? Where did you get that kernel? It is not on kernel.org yet.
Do you care about the security of your wireless mouse?
Sorry,
I just had to say, that's funny!
I'm amazed how MS added network support in only the 3.61 days between the release of 3.1 and 3.11.
"For every expert, there is an equal and opposite expert"
Does nobody even know how their system works anymore? xset people! xset!!
Win9x = 9x kernel
WinXP = NT kernel
Almost completely different beast.
Plus, what's to say this bug isn't simply a problem with vmware?
When all freedom is outlawed only the outlaws have freedom
These vaporware announcements must stop. Already since Linus announced this I have noticed a drop-off in downloads of the 2.4 kernel.
:)
These anti-competitive practices need to be stopped. A kernel-lead developer such as Linus should have to compete on the same level as everyone else for his kernel to get used.
With the windows kernel change, there was also a significant change from Win9x to XP
True, but you can't include 9x in your argument.
The Windows XP kernel is based off of the Windows NT4 kernel.
The Win98 kernel (also the basis of the ME kernel) is based off of the Win95 kernel.
Think of it this way, 2.6 is based (very loosely, albeit) on the 2.4 kernel, (2.2 kernel, etc.. to Linux-0.x).
Your comparison of the W2K kernel to the W98 kernel is like comparing the Linux kernel to the SCO kernel.
No matter how much FUD SCO publishes toward the contrary, the Linux kernel does not use the same codebase of their Unix(tm) kernel.
I merely use SCO as a reference because that's all everyone talks about, and almost everyone wanes incessantly about how Linux is not SCO, so I figured you could understand it this way.
To fend off the flames, s/"SCO"/"AT&T"|"Solaris"/
nor has it meen mentioned on lkml
pat
Humans are slow, innaccurate, and brilliant; computers are fast, acurrate, and dumb; together they are unbeatable
do you mean in ten more years Linux might DIE ??!
Is the framebuffer stuff fixed yet? Last time I liked (around 2.5.6x or so) it was very badly broken.
See my journal, I write things there
This may sound a bit wacked but here it goes. For a lot of uses I'd like to not mount a file system but instead access the block device from a single user level process via custom syscalls any sugestions on how to do this in 2.6 plus what is the best file system to use. I tend towards ext3 but a simple BSD filesystem may be a better answer. If you don't mount the root file system then a embedded linux is hack proof.
Heck, its open source I can put the kernel version number at what ever level I want to.
Well.. maybe. Or Maybe not. But Definitely not sort of.
Kernel is as stable as the rocky mountains. Been compiling Mozilla non-stop in gcc for the last 3 hours with no problems.
A whole 3 hours? Wow, must be ready for release!
Reminds me of Microsoft's crazy uptimes of "three, six months."
From what I heard, Linus Torvalds himself didn't initialy enjoy the patch submission of another developer which caused prioritization improvments to X server and Input Devices (Keyboard, Mouse). You are talking about that patch. To my remembrance, Linus said, somewhere along the lines of, "Windows NT did this...bad...verry bad...We are supposed to entreat all applications/devices equaly, not priority in such of others...bend over, Ralph#~Jim#~Miguel, I'm spanking you today...ok, you can have the patch included. No hard feelings?"
This improvment you speak of is some sort of modification to allow automatic renice(1). In the past, using a Dual Pentium Pro UltraWide SCSI graphics workstation, I had a problem with keyboard and mouse keys becoming stuck in the software state as well as audio playback skipping. Perhaps this patch is what makes the lame determination upon what peice of software needs priority over other software; that still doesn't help the system dig itself out of a whole when every peice of software wants highest priority or whatnot. This code, honestly, is a Bad Thing(TM) != Good Thing(TM).
(1)renice, give priority to applications. System level application "nice" with numerical parameter and string parameter; -20 being lowest priority, 0 being normal priority, +20 being highest priority; followed by the name of the daemon or thread or application that will have its priority changed.
SP4 does nothing else but fix bugs.
Yup. We fixed a bug by applying it on our SQL server at work. It rebooted randomly 4 times the next day.
Really.
Yeah, slashdot really should have tags so people can spot the joke ... :)
Been compiling Mozilla non-stop in gcc for the last 3 hours with no problems..... ./configure has just about finished! And who said you needed the latest hardware?
on my 486. Oh look -
You are in a twisty maze of processor lines, all alike.
There is a lot of hype here.
Is that the 'locate' database updating itself? On Debian I think that happens by default every 24 hours, at some time in the middle of the night. Basically builds a list of all the files on your hard drive, so you can do "locate blah" and quickly find where files are without searching the hard drive each time.
/dev/hda" (or whatever your hard drive is). If it says off, do "hdparm -d1 /dev/hda" to turn it on.
Another possibility might be to check if DMA is on; sometimes the kernel won't turn it on, depending on your config and hardware. Try "hdparm -d
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Umm, no the nanosecond patch is not critical for make on fast computers. Sometimes a file might get rebuild when it shouldn't. Thats all.
Shared page tables are not in 2.5, nor are they likely to be in 2.6. It has nothing to do with multiple CPU machines.
No true bigot runs an rpm-based distribution. Real Bigots run Debian, Slackware, Gentoo, or Linux From Scratch.
I thought SCO told us that Linux was the Commie OS?
Don't tell me they're lying about that too?
Cheers
Stor
"Yeah well there's a lot of stuff that should be, but isn't"
Afterall Linux is open source, meaning anyone can hack the Linux kernal & create their own Linux 2.6.
So why doesn't that happen? & what gives Linus & co the authority to determine the next Linux revisions? I'm not saying he shouldn't, I just wondering why it happens this way when AFICS open source should means its a free for all.
Well crap...just this morning I updated to 2.4. Now I can't feel all cool anymore.
I have a girlfriend whose name doesn't end in
What happened to devfsd and lvm? I know they were talking about replacing lvm, and I was wondering if the new code is in place?
Also, I read somewhere that the developers were unhappy about devfsd, since 'nobody was using it'. I'm using it, so I'm hoping they don't remove it.
Je ne parle pas francais.
If updatedb (the program updating the locate database) only makes your hard drive for 5-10 seconds as the grandparent post said then you must have a bloody fast machine or a bloody small hard drive.
"The obvious mathematical breakthrough would be development of an easy way to factor large prime numbers." Bill Gates,
slashdot really should have <humour> tags
You know... at first I read that as a <rumour> tag.
Do you care about the security of your wireless mouse?
I've been hearing though other channels that the IDE layer rewrite improves the IDE subsystem to the point where SCSI emulation won't be needed to drive an IDE CD burner. Can anyone confirm or deny this? If so, this will probably become my main reason to switch to 2.6 (although there are quite a few secondary ones too). Thanks linux team (and IDE rewrite folks)!
"Can't you see that everyone is buying station wagons?"
Win2K and XP do in fact have their own special BSOD's, they're just hidden behind what I've come to call pseudostability. Right click "my computer",properties, Advanced tab, click "startup and recovery..." and note that the default setting for 2K and XP is to reboot on bluescreening, a behavior slightly better than doing nothing, but not nearly as nice as preventing it to begin with.
You are correct in assuming you don't need ide-scsi to emulate a SCSI host for burning cdroms in 2.6, but it has nothing at all to do with the IDE rewrite.
2.6 has support for queueing "generic scsi" commands through the block layer, using the same mechanism and transport as the regular read/write file system requests. So we can overload the sg (scsi generic) SG_IO and provide the same functionality for non-scsi attached devices (such as atapi burners). With a recent cdrecord, you can give the device with -dev=/dev/hdc for instance.
Additionally, cd burning is now zero copy. The user space data buffer is mapped directly into the kernel for the dma operations. DMA is supported on a 4-byte boundary, where 2.4 and previous has required sector alignment (512 bytes) for any atapi dma operations.
>Hotplug CPU Removal Support.
I know Linux is wonderful but running without a cpu?
SP4 went onto our 2k workstations fine, but wrecked our metaframe box.
Off Linus?
Wow, you can deadlock the kernel by inserting random code into the kernel ( vmware binary kernel modules)? Thats amazing!!
As soon as you mess with binary kernel modules you can not expect Linux to guarantee stability, the modules could do anything they like. Really.
Is that the 'locate' database updating itself?
Normally, cron updatedb is heavily niced (19 on my box). So performance shouldn't suffer too much, except perhaps for I/O-heavy work. But locking up? No.
When DMA is off, hard drive accesses can block, especially with a stock 2.4.x (non-preemptible, non-low-latency-patched) kernel. In some brief measurements I did, running updatedb could cause blocks of up to 500ms, which leads to some pretty crappy interactive performance until it finishes.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
It's like to say:
Your watch is 10 times faster now - enjoy!
BTW, if you're willing to break the terms of the VMWare license, you can see the source code for the VMWare kernel modules. This is how VMWare 2.x kernel module patches were made available on the net for those users using kernels newer than 2.4.6.
There are very, very, very few kernels with which the precompiled "stock" modules that ship in the VMWare RPM work properly. When one of these kernels is not found, vmware-config.pl runs through compiling the modules, and if you care to catch it quick enough, the module source appears in /tmp. The perl script also quite frequently forgets to delete the source. Have at it.