Weekly Edition
Return to the Kernel page
| |||||||||||||
Oracle's "Unbreakable Enterprise Kernel"
Oracle has sent out a
press release proclaiming the virtues of its new kernel. "Based
on the combined efforts of Oracle's Linux, Database, Middleware, and
Hardware engineering teams, the Unbreakable Enterprise Kernel is: Fast:
More than 75 percent performance gain demonstrated in OLTP performance
tests over a Red Hat Compatible Kernel; 200 percent speedup of Infiniband
messaging; 137 percent faster solid state disk access." Much of the
gain may come from using 2.6.32 as the base instead of 2.6.18, but it's
hard to tell for sure; source for this kernel does not seem to be available
for download as of this writing.
This is an interesting move in that it reintroduces competition at the kernel level - something distributors have not emphasized in recent years - and it demonstrates an intent to move a bit further away from the RHEL base that Oracle uses to build its offering. (Log in to post comments)
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:19 UTC (Mon) by rsidd (subscriber, #2582) [Link]
Do they define "unbreakable" anywhere? Are they sure there are no security holes?
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:22 UTC (Mon) by mmcgrath (subscriber, #44906) [Link]
> Do they define "unbreakable" anywhere?
It's just a marketing term they use. It has no foundation in engineering.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:39 UTC (Mon) by njd27 (subscriber, #5770) [Link]
Just in time for the Titanic centenary.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:13 UTC (Mon) by jordanb (guest, #45668) [Link]
Unbreakable Linux: God couldn't crack this kernel!
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 21, 2010 7:23 UTC (Tue) by salimma (subscriber, #34460) [Link]
Ah, but could God *write* such a kernel? :)
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 16:29 UTC (Mon) by eklitzke (subscriber, #36426) [Link] "Unbreakable" is just the name they use for their Linux support program; I think they've been using it since 2008. More info: Unbreakable Linux Whitepaper [pdf].
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:06 UTC (Mon) by Trelane (subscriber, #56877) [Link] "Unbreakable" refers to the fact that they've tried to patent all of the attacks they could think of.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 25, 2010 10:45 UTC (Sat) by Lennie (subscriber, #49641) [Link]
I always thought "unbreakable" did not refer to security but stability as I always thought it was mostly used in combination with their database clustering software on Linux.
If they use a cluster it could potentially mean greater stability if done right.
Competition at the kernel level Posted Sep 20, 2010 15:22 UTC (Mon) by tmassey (guest, #52228) [Link]
I hope that it inspires Red Hat to get back on a more regular upgrade schedule. The promise of a new major version every 18 months and a minor version every 3 or 4 months are *way* in the past. Maybe this will get them back on the horse.
Competition at the kernel level Posted Sep 21, 2010 1:16 UTC (Tue) by cventers (subscriber, #31465) [Link]
I have mixed opinions on that subject. When I was younger I once implemented some IT servers for a small company I was working at using Gentoo Linux, and I never stopped regretting it any time it came time to emerge software updates. (Would the server break, requiring me to spend hours manually diagnosing the problem?) Clearly, the bleeding edge is an absurdity... that experience taught me to appreciate the stable nature of enterprise Linux distributions when I'm using Linux for enterprise purposes.
At the same time, I'm deploying a CentOS 5.5 cluster right now, and there were a number of packages I had to backport from Fedora because I needed certain features in certain libraries, etc. In some cases it's really hard to maintain your own library fork and still stay within the RPM bailiwick. I'm eagerly awaiting 6, and hoping I don't have to spend as many weeks wearing my packager hat to prepare our system for it.
The thing is, RHEL 5 is at the very end of its life, which is why I had to backport as much as I did.
Perhaps 12 months is a more appropriate turnover time? If it is too frequent, it encourages deprecation of releases people may not be ready to stop using.
Competition at the kernel level Posted Sep 21, 2010 5:21 UTC (Tue) by tmassey (guest, #52228) [Link]
There's a difference between updating existing servers and installing new ones. That's why RHEL supports--and need to support--multiple major versions.
I'm virtually never going to upgrade from one major version to another on a production server: that's not why I want newer major versions faster. It's so that I don't have to install a multi-year-old kernel with a multi-year-old userspace on my brand new server which I'm going to want to run for the next half-decade.
RHEL 5 is past its prime. And if you want to stay with RHEL, where are you supposed to go? That's why I'm hoping that the competition from Oracle might motivate Red Hat to pick up the pace on their server platform.
Competition at the kernel level Posted Sep 21, 2010 9:47 UTC (Tue) by cventers (subscriber, #31465) [Link]
> There's a difference between updating existing servers and installing new
> ones. That's why RHEL supports--and need to support--multiple major > versions.
Indeed, but if you decide you're going to support X versions of a distribution, your support window for any one of them necessarily decreases with the period of its release cycle. Choosing X depends on how much manpower you're willing to spend on the problem.
Each major release in a stable distribution is effectively a huge set of forks. The older the software gets, the harder it can become to maintain, which is multiplied by how many major releases you're maintaining. Inevitably this has an effect on which packages, and how many, Red Hat is willing to support. (I really disliked RHEL until I found out about EPEL.)
Don't get me wrong... I wish RHEL 6 would have come around sooner. Just so long as that doesn't force users off "old" software too quickly.
Competition at the kernel level Posted Sep 21, 2010 13:41 UTC (Tue) by tmassey (guest, #52228) [Link]
Yes, but you would have to support multiple forks for *less* time if you had more of them.
Right now, Red Hat is looking to have to maintain RHEL5 for as long as 10 years or more. Realistically, they'll have to support it for at least 5 years after they release RHEL6, simply because brand new servers that were loaded with RHEL5 right before (and quite likely, for some time *after*) the release of the next version will have that type of life expectancy. So, the longer they go between releases, the longer they have to support the older versions. And as you said, the longer they have to do this, the more painful it becomes.
So, it's not as simple as "more versions = more work, less versions = less work"; by releasing major versions more frequently, they maintain more forks, but for less time.
But whether it's more work or less work, the fact remains: Red Hat has fallen *way* behind the standard they themselves set when they created the RHEL line. RHEL5 is crazy behind the times--especially the userspace. Any competition that inspires them to rectify this is, to me, a very good thing.
Competition at the kernel level Posted Sep 21, 2010 18:10 UTC (Tue) by mrjk (subscriber, #48482) [Link]
That isn't what a stable enterprise release is. You still have to support ALL of the releases as long as you say, could be 10 years or more. The more releases the more things you have to support and the more chance for issues. Keep the releases as they are and you can look elsewhere for your quick changes.
Competition at the kernel level Posted Sep 21, 2010 19:39 UTC (Tue) by dowdle (subscriber, #659) [Link] Keep the releases as they are and you can look elsewhere for your quick changes.Ok, how ARE the releases? Red Hat said they plan to come out with new releases every 18 - 24 months. RHEL 5 is 3.5 years old now. Their update schedule they've been doing pretty well with in my opinion... with an update about every 6 months. I don't recall them saying every 3 months but maybe I'm wrong on that. Should the schedule be kept as Red Hat said it would be... or should we use this new, as yet unknown, release cycle length? If you want to stick with RHEL 5 after RHEL 6 comes out, more power to you. Having a new release come out does not make RHEL 5 go away... so why there would be a downside to a new release for you makes no sense. I will give them some credit for re-basing some user apps... from both a usability stand point and an upstream-no-longer-supports-it stand point... and they do backport a ton of drivers and some features to their kernel so that it isn't so ancient.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:25 UTC (Mon) by ummmwhat (guest, #54087) [Link]
Anyone have found link to SRPM?
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:30 UTC (Mon) by fuhchee (subscriber, #40059) [Link]
Looking forward to the --author=foo@oracle.com patches, considering the "... based on the combined efforts of Oracle's Linux, Database, Middleware, and Hardware engineering teams ..." blurbage.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:32 UTC (Mon) by rvfh (subscriber, #31018) [Link]
And who's gonna sue them if they don't release the patches? Could they not get away with it?
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:35 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link]
Patches would fall under GPL and any copyright holder on the Linux kernel can enforce that license. I doubt Oracle wants to take that risk either.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:36 UTC (Mon) by madscientist (subscriber, #16861) [Link]
It's not illegal to announce that you've made changes to the kernel but not release any code. First they have to actually ship the kernel to someone. Then, if it's not available with the kernel, someone needs to ask them to provide the code. Then, they have to refuse to hand it out. Then, they can be sued by just about anyone holding copyrights on the kernel.
Personally I doubt me that it will ever get that far.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:48 UTC (Mon) by tnoo (subscriber, #20427) [Link]
> Then, if it's not available with the kernel, someone needs to ask them to > provide the code.
They only need to provide the source to the recipient of the kernel, not to the general public. If that person/entity doesn't redistribute the code, that's their choice.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 16:28 UTC (Mon) by madscientist (subscriber, #16861) [Link]
That's a common misconception. The GPLv2 clause 3 lists the obligations for redistribution and they allow you to EITHER provide the complete source code bindled with the distribution, OR accompany it with a written offer, good for at least 3 years, to give ANY third party (not just the recipient) the source. You're allowed to charge a reasonable fee for replication and distribution but that's it.
So if you don't hand out the source with the distribution, then you have to agree to give it to anyone who asks.
Easy evaluation Posted Sep 20, 2010 16:31 UTC (Mon) by dmarti (subscriber, #11625) [Link]
If you fill out your name and email address on their web form, you can get an evaluation copy, same as RHT.
You can always break out the GPL-licensed sources and use them under the GPL.
(Hmm, sounds like a doable business model--surf around and round up all the sources for stuff that might not be upstream, stick them in git repositories, and offer access--free of charge to email addresses that show up in "git log" for the upstream project, subscription-based to others.)
Easy evaluation Posted Sep 20, 2010 18:54 UTC (Mon) by blitzkrieg3 (subscriber, #57873) [Link]
I've talked about this with the guy who runs rhkernel.org, it seems like a good idea, just need a way to translate RPM/DEB->git
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 15:58 UTC (Mon) by ofirm (guest, #54632) [Link]
disclaimer - I'm an Oracle employee but I'm not part of the kernel team and have no internal knowledge about this project.
Oracle Linux binaries and source code are typically available at the main download site: http://edelivery.oracle.com/linux While Oracle Linux is free, the download site asks you to identify yourself and agree to some generic license. I didn't have a copy of the previous Oracle Linux 5 update 5 so I can't compare the checksums, but I guess that it is quite likely that updated binary and source-code CDs and DVDs will be upload over the next few days. Anyway, the main thing that is nice about this in my opinion is to have an updated, faster kernel on the existing RHEL/OEL 5 toolchain, keeping the existing user-space ABI. Compare this for the eternal wait for RHEL6... It will take at least a couple of years until most systems switch to RHEL6, so if we can enjoy a significantly faster kernel and still get "Enterprise" support, that seems pretty cool.
Re: ... and agree to some generic license Posted Sep 20, 2010 23:12 UTC (Mon) by ldo (subscriber, #40946) [Link]
Id say the agreement to export restrictions, at least, would seem to contravene the terms of the GPL.
Re: ... and agree to some generic license Posted Sep 21, 2010 7:28 UTC (Tue) by salimma (subscriber, #34460) [Link]
Oracle is bound by the same export restriction as any US-based company, but it's certainly interesting how more proactive they enforce this restriction.
Re: ... and agree to some generic license Posted Sep 21, 2010 8:56 UTC (Tue) by dlang (✭ supporter ✭, #313) [Link]
back when the restrictions were being enforced more stringently than they are now, Oracle was already a large company dealing with encryption and getting a lot of attention as a result. They probably setup standard practices and boilerplate that persists because it's easy to keep doing it, but deciding to stop doing something they are already doing would be a risk and draw attention to the thing where it's being stopped.
Re: ... and agree to some generic license Posted Sep 23, 2010 5:48 UTC (Thu) by brouhaha (subscriber, #1698) [Link] Oracle is bound by the same export restriction as any US-based companyYes, but section 7 of the GPL says "If [...] conditions are imposed on you [...] that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all." In other words, if they are not able to satisfy section 3b and give source code TO ANY THIRD PARTY, then they cannot comply with the GPL, and may not distribute Linux at all. US export law, or the laws of any other country, don't excuse anyone from complying with the GPL.
Re: ... and agree to some generic license Posted Sep 23, 2010 10:12 UTC (Thu) by spaetz (subscriber, #32870) [Link]
So I just have to found a state that disallows to own software at all and that would make all GPL applications invalid world wide? :-)
I wonder if Steve Ballmer/Jobs should not buy a small rock somewhere in some ocean.
Re: ... and agree to some generic license Posted Sep 23, 2010 16:24 UTC (Thu) by brouhaha (subscriber, #1698) [Link] So I just have to found a state that disallows to own software at all and that would make all GPL applications invalid world wide?No. If it is illegal to own software in Lower Slobovia, that doesn't make it illegal for me as a US citizen to sell software to someone there, because I am not within Lower Slobovia's legal jurisdiction. It is generally up to the citizens of Lower Slobovia to deal with their laws. But if Lower Slobovia allows its citizens to distribute software to France, but prohibits distribution to Italy, then its citizens can't redistribute GPL'd software while depending on section 3b of the GPL. They would have to rely on section 3a or 3c instead. If you think about this a little bit, you can probably figure out why it actually is relevant to the US, and not just the hypothetical Lower Slobovia.
Re: ... and agree to some generic license Posted Sep 23, 2010 20:02 UTC (Thu) by jds2001 (subscriber, #50765) [Link]
They distribute source under 3a. See http://public-yum.oracle.com/repo/OracleLinux/OL5/5/base/... which includes *both* source and binary distributions.
Also see http://blogs.oracle.com/wim/2010/09/some_quick_answers_on...
Hats off to Oracle on this, they've one-upped RHT on source availability.
Re: ... and agree to some generic license Posted Sep 23, 2010 20:15 UTC (Thu) by fuhchee (subscriber, #40059) [Link]
"Hats off to Oracle on this, they've one-upped RHT on source availability."
Doesn't that depend on whether the git tree they posted will
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:12 UTC (Mon) by masoncl (subscriber, #47138) [Link]
> Looking forward to the --author=foo@oracle.com patches, considering the "... based on the combined efforts of Oracle's Linux, Database, Middleware, and Hardware engineering teams ..." blurbage.
Grin, when you spin the rpm and hand it out for benchmarking, you usually get one of three answers:
1) This sucks
After you repeat #2 in a loop a bunch of times, I think its fair to call it a combined effort. You only get to #3 about once in a release, and then you stop reading their email because the answers can only get worse.
-chris
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:20 UTC (Mon) by mokki (subscriber, #33200) [Link]
This is just speculation, but my guess is that they have used a RHEL6 kernel (which is due to be released in few months) and test it against RHEL5 kernel. The announcement (which I have not yet read) propably is so vague that makes people think oracle has done major changes themselves to gain the performanced. Still, nice move from Oracle: even if RH would publish the same benchmark on RHEL6 (which oracle RDBMS contracts prohibit) with similar results they would still seem to be playing catch-up.
Oracle has done work on btrfs and cluster filesystems, which might explain some of the speed noost. But my gut feeling is that most of the speedup is from newer kernel and transparent hugepages (THP) which is peopably the largest single performance enhancement in RHEL6 for many a software. One published benchmark showed that virtualized kernel compile with THP was faster than bare metal without THP (and of course nonvirtualized+THP performed best).
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:04 UTC (Mon) by Wol (guest, #4433) [Link]
Hmm...
If that's what Oracle have done, they'd better not use it for advertising in the UK!
"Legal, decent, honest, truthful".
That would fail the "honest" and "truthful" tests, and get them a rap on the knuckles from the ASA. More than a couple such raps, and the press here are unlikely to want to take their advertising!
The ASA doesn't have to bark much - it rarely bites but when it does the results can be very nasty ...
Cheers,
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:43 UTC (Mon) by ummmwhat (guest, #54087) [Link]
"my guess is that they have used a RHEL6 kernel"
It seems that this is fully patched 2.6.32.16 http://oss.oracle.com/git/?p=linux-2.6-unbreakable.git;a=...
RedHat do not want to accept many -stable patches
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:46 UTC (Mon) by johill (subscriber, #25196) [Link]
It's missing the tags, but it seems to be based on 2.6.32.21.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 19:04 UTC (Mon) by ummmwhat (guest, #54087) [Link]
All the better for them and users :)
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 16:42 UTC (Mon) by SEJeff (subscriber, #51588) [Link]
Redhat's original response to the "Unbreakable Linux" marketing campaign was and still is great. They dub RHEL as "Unfakeable Linux".
I just bought a shirt:
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:33 UTC (Mon) by daniel (subscriber, #3181) [Link]
Hi Jeff,
At the time I regarded this Red Hat response as stylish, and in the event, highly accurate, with Oracle's fork largely ignored then and now by enterprise customers. With good reason as it turns out (see Oracle vs Google).
One small regression on Red Hat's part is that the original unfakeable site no longer exists.
Original:
http://www.redhat.com/promo/unfakeable/
as we can see:
http://web.archive.org/web/20061207054515/http://www.redh...
now redirects to:
http://www.redhat.com/promo/real/
By the way, unfakeable is properly spelled unfakable.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:51 UTC (Mon) by SEJeff (subscriber, #51588) [Link]
You're completely right Daniel.
Hopefully this will task Redhat with breaking their longstanding policy of never upgrading anything in an enterprise release. Novell did something similar in a recent SLES service pack[1]. Even though OEL is often seen as amusement, this is a good reason for some shops to check it out.
[1] http://www.h-online.com/open/news/item/SUSE-Linux-Enterpr...
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:25 UTC (Mon) by blitzkrieg3 (subscriber, #57873) [Link]
Red Hat has already done that with realtime kernel. If it's true that Oracle is running a .32 kernel as you speculate, than I don't think it's any big deal.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:39 UTC (Mon) by SEJeff (subscriber, #51588) [Link]
Read the other comments from oracle employees working on the kernel. It leads you to:
http://public-yum.oracle.com/repo/OracleLinux/OL5/5/base/... and the public git repo of their enhancements. Some of them look tasty.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 16:50 UTC (Mon) by arjan (subscriber, #36785) [Link]
I suspect some of this is forced.... 2.6.18 clearly is not competitive today, in terms of power and performance. (2.6.18 can consume 2x the power than a much newer kernel does... if you run a datacenter then, well, that REALLY REALLY matters).
In a way, by having the next release with a new kernel a long time apart, Red Hat has opened a window of vulnerability to this sort of competitive "jabs"...
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 16:52 UTC (Mon) by SEJeff (subscriber, #51588) [Link]
Didn't Novell do the same thing against Redhat by being the first to ship Xen when Redhat claimed Xen wasn't mission critical ready yet? Novell certainly trumpeted about it when Redhat just said they shipped a broken copy of Xen with patches to paper over issues. That didn't seem to work out so well for Novell.
What do you think?
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:05 UTC (Mon) by arjan (subscriber, #36785) [Link]
2.6.32 isn't exactly state of the art (2.6.36 would be that)...
if the gap in time was only a few months, I'd say you'd have a very good point.
If the gap is years... I can completely believe someone doing an enterprise ready kernel in such a time window.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 22:09 UTC (Mon) by dag- (subscriber, #30207) [Link]
Arjan, is this statement based on a vanilla 2.6.18 or on Red Hat's modified 2.6.18 kernel ? I ask this because I am running a 2.6.35 kernel on top of RHEL5 and the power-usage is only a little less than RHEL 2.6.18 kernel (shipping with RHEL 5.5).
It's somewhat more than 1W difference in idle state (about 14W with 2.6.35).
I remember Red Hat advertising twice that they made considerable power usage enhancements in their minor RHEL5 updates.
PS Regarding performance it's harder to measure as the 2.6.35 kernel is a preempted kernel with a kernel config targeting desktops.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 21, 2010 3:24 UTC (Tue) by arjan (subscriber, #36785) [Link]
on bigger servers the delta is actually quite huge (not such much on some laptops surprisingly)....
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:01 UTC (Mon) by masoncl (subscriber, #47138) [Link]
Hi everyone,
One of the goals of this kernel was to stay as close to 2.6.32.stable as we could. The sources are here in git, they won't be rebased:
http://oss.oracle.com/git/?p=linux-2.6-unbreakable.git;a=...
The main differences from mainline:
*) semtimedop optimizations. I posted these to the list a while ago, and Manfred took things in a less complex direction. He was waiting for me to fully benchmark the less complex version, but we ran out of time in the release cycle and had to focus on other things. Oracle hammers on the IPC lock, so these made a big difference, and now I finally have time to properly benchmark his approach against mine.
*) Ocfs2
There are other bits and pieces, but we resisted the urge to pile things in.
The solid state disk access number came on a huge machine, and the improvements came from getting rid a lock in the driver and enabling it for softirq affinity code without taking any of the request locks.
Over the next 12 months we'll be getting an update prepared to a new mainline version, and trying to hammer on upstream kernels as much as we can to reduce our patch count even more.
-chris
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:20 UTC (Mon) by jetsaredim (guest, #70230) [Link]
All,
I'm one of Chris' co-workers.
The associated rpms are freely available to all on the Oracle Public Yum server (http://public-yum.oracle.com). These repos mirror the content available on the distribution ISOs, but not bugfixes or errata. We even provide repo files to your convenience.
Specifically, the src rpm for the new kernel is located at http://public-yum.oracle.com/repo/OracleLinux/OL5/5/base/....
Enjoy. :)
Jared
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 17:43 UTC (Mon) by daney (subscriber, #24551) [Link]
A perfectly good flame-fest, and someone has to come along and ruin it with facts.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 19:41 UTC (Mon) by masoncl (subscriber, #47138) [Link]
Well, after things were basically working, we started fixing all the real bugs reported by an updated checkpatch.pl
Any lines that obviously should have been longer than 80 chars were joined back to their proper length, and curly braces were moved to their proper locations. Tabs were all changed to 3 spaces each, we couldn't quite decide on 2 vs 4.
After that was done, we forward ported devfs. Proper disc management code in Linux was long overdue.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 20:06 UTC (Mon) by ummmwhat (guest, #54087) [Link]
:) LOL
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 23, 2010 16:53 UTC (Thu) by walex (subscriber, #69836) [Link]
Nothing to laugh about 'devfs', it still has some crucial advantages over the abomination called 'udev', and as to that, please consider that 'udev' actually relies on something that is equivalent (but badly) to 'devfs', please check the output of 'cat /sys/block/sda/dev' for a laugh on how clever and nasty GKH is.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 23, 2010 21:32 UTC (Thu) by zlynx (subscriber, #2285) [Link]
The device numbers are critical information for creating a device node. It isn't equivalent to devfs at all.
Where devfs got into trouble was the creation and destruction of device nodes and the permissions on the nodes.
There is hardly any advantage to devfs over udev when a daemon has to be running to update device node permissions according to system policy. If you have that daemon it may as well be responsible for creating the nodes too.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 23, 2010 22:31 UTC (Thu) by njs (guest, #40338) [Link]
A picture of Greg with his goggles pushed up! You're right, that is clever.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 23, 2010 23:16 UTC (Thu) by bronson (subscriber, #4806) [Link]
That's quite a look of surprise on his face!
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:34 UTC (Mon) by ummmwhat (guest, #54087) [Link]
Thanks!
This is what I asked for!
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 18:46 UTC (Mon) by NightMonkey (subscriber, #23051) [Link]
FYI, you've got some 404'd links on the public-yum page:
Oracle Linux -> http://www.oracle.com/technologies/linux/technical.html -> 404
Also, that download page ( http://edelivery.oracle.com/EPD/GetUserInfo/get_form?call... ) has some seriously restrictive language which must be agreed to before downloading. Is that actually compatible with the GPL v2 to have all those additional restrictions attached before downloading the software?
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 19:15 UTC (Mon) by michaeljt (subscriber, #39183) [Link]
> Also, that download page ( http://edelivery.oracle.com/EPD/GetUserInfo/get_form?call... ) has some seriously restrictive language which must be agreed to before downloading. Is that actually compatible with the GPL v2 to have all those additional restrictions attached before downloading the software?
For Oracle Linux I see two restrictions - one that the person downloading must accept the GPLv2, and the other is the export restriction stuff that they are required to put there according to US law. I think that if the GPL doesn't allow that then it forbids any US company from distributing the Linux kernel altogether unless they rip out at least the cryptography bits.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 20, 2010 20:21 UTC (Mon) by nix (subscriber, #2304) [Link]
> the person downloading must accept the GPLv2
Only if they plan to redistribute it themselves (in which case the act of redistribution itself constitutes acceptance).
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 21, 2010 20:09 UTC (Tue) by butlerm (subscriber, #13312) [Link]
"in which case the act of redistribution itself constitutes acceptance"
Strictly speaking that is not true. No one "accepts" an ordinary license. The liability is entirely in terms of getting sued if you go beyond the rights granted to you by the license itself (or the other rights you have license or no license).
Of course this wanders into the bizarre world of shrinkwrap and other post-sale/post-delivery "licenses", and how and why they actually do what they claim to. At least there is substantial and identifiable consideration in most cases there though, making the arrangement look much more like a contract than a blanket grant to whoever and whomever.
Oracle's "Unbreakable Enterprise Kernel" Posted Sep 22, 2010 12:23 UTC (Wed) by nix (subscriber, #2304) [Link]
Yeah, you win the pedantry award. You don't accept a license: you choose to exercise the rights it grants you.
|
|||||||||||||