Torvalds Bemoans Size of RC7 For Linux Kernel 3.5

alphadogg writes "A host of small modifications and a large number of system-on-a-chip and PowerPC fixes inflated the size of release candidate No. 7 for Version 3.5 of the Linux kernel, according to curator Linus Torvalds' RC7 announcement, made on Saturday. Torvalds wasn't happy with the extensive changes, most of which he said he received Friday and Saturday, saying 'not cool, guys' in the announcement. However, the occasionally combustible kernel curator didn't appear to view this as a major setback. 'Now, admittedly, most of this is pretty small. The loadavg calculation fix patch is pretty big, but quite a lot of that is added comments,' he wrote, referring to the subroutine that measures system workload."

Comments

wow

[masternerdguy]

Linus is getting bitchy lately.

Re:wow

[PreparationH67]

Eh, at bit. But I will defend his ripping into the opensuse devs about using the root password for everything until my last breath.

Re:wow

[Trashcan Romeo]

There are few things more painful than a swollen kernel.

Re:wow

[morcego]

Linus is getting bitchy lately.

Yeah, and RMS was talking non-sense yesterday. What is the world coming to ...

Re:wow

[Anonymous Coward]

Obligatory.

Re:wow

[fahrbot-bot]

Yeah, and RMS was talking non-sense yesterday. What is the world coming to ...

Yesterday? I'm a big fan of RMS - since before the beard - but the day he doesn't talk non-sense will be news.

Re:wow

[osu-neko]

I think the point is not so much the swelling but the fact that this is a huge bunch of stuff to be thrown in during an RC cycle, between rc6 and rc7. You're not really supposed to be doing anything major to a release candidate...

Re:wow

[morcego]

Yeah, and RMS was talking non-sense yesterday. What is the world coming to ...

Yesterday? I'm a big fan of RMS - since before the beard - but the day he doesn't talk non-sense will be news.

Exactly my point. Just like the day Linus doesn't get bitchy :)

Geez, I figured we were all past the <sarcasm> tag already.

Re:wow

[Anonymous Coward]

Linus has always been bitchy.
It is why Linux is the way it is now.

If it wasn't for his bitchiness, it would be Windows. Yes, I am not kidding.
There'd be ENTERPRIIIISE CODING brilliance in there, AKA useless bloat for stuff nobody should EVER, IN THE HISTORY OF EVER, have access to, and countless other things. (up YOURS Microsoft! )
What's that, writing a driver are you? If it isn't fully descriptive in code, you're fired!
What's that? You saved a huge number of cycles by using a Goto there? FIRED, we want more lines! (I'm not even kidding, Linus had to defend a Goto in a driver-level file, this is how mad this anti-Goto retardedness is these days, kids man)
So on and so fourth.

Hey, at least he isn't a Ballmer. Nobody can beat ol' monkey boy.
Developers developers developers deve... oh go away developers we don't want you in Windows 8 anym... no sorry we were just kidding! ... honest!
Linus is always solid.
Without him, Linux would turn in to PHP. Look what happened to that. PHP is plain awful now. It started off with a good idea, then all the amateurs took control and ruined it. You don't want that now, do you?

Re:wow

[Tough Love]

There are few things more painful than a swollen kernel.

It's nothing an antibiosic shot wouldn't fix.

Re:wow

[Short Circuit]

I doubt Linus is getting more bitchy than normal. He's just had more 'popular' exposure and attention of and to his rants than normal. It's easy to guess why: Google+ gives him a lot more exposure and spread. Prior to his posting the rant against the root password requirement on Google+, I don't think I'd seen any of his opinions outside of near-fluff interview pieces or, possibly, LKML emails.

Certainly, people didn't care as much until they saw him lambast OpenSuSE developers. That got their attention and interest, and so folks like Slashdot and NetworkWorld are more likely to cover it. Heck, this kind of story is even out of character for /..

Linus only seems more bitchy because people are looking at him more.

Re:wow

[Anonymous Coward]

Gastrointestinal network address translation. The details are a bit mucky to me at this point.

Re:wow

[fatphil]

"Without [Linus], Linux would turn in to PHP. Look what happened to that. PHP is plain awful now. It started off with a good idea, then all the amateurs took control and ruined it."

There are a remarkable number of bodgers in the linux kernel too. The general code quality is not that high. In particular from some SoC vendors. He has to trust that if they're selling that code to customers who need it to work, then it probably works. And if it doesn't, then they are the maintainers responsible for fixing it. Unfortunately the customers are sometimes as incompetent as the chipset vendors, and don't know what they're being sold. I promise you, sometimes it really isn't pretty. Fortunately the core parts seem to be under much tighter rein, but they're still far from bulletproof.

Re:wow

[Fujisawa Sensei]

No shit.

But it's still not enough to make me switch back to fedora, ubuntu, kubuntu, or gentoo.

Re:wow

[asdf7890]

Unfortunately the customers are sometimes as incompetent as the chipset vendors, and don't know what they're being sold.

Or who to blame... If it doesn't work with Windows it is seen as the manufacturer's fault (as they provided the drivers) but if it doesn't work under Linux it is the kernel dev's fault (as the user doesn't know that the drivers there were written by the manufacturer too) and it is they who are expected to fix the problem. I do not envy them that position!

Re:wow

[TheInternetGuy]

Geez, I figured we were all past the <sarcasm> tag already.

</sarcasm>
Ah, there, that's better...
If you don't close your sarcasm tags, my sarcasm parser will get messed up and my whole day gets very confusing.

Re:wow

[X0563511]

It seems less like bitchiness to me than people writing shitty headlines to make it seem that way.

Re:wow

[knorthern knight]

> If it wasn't for his bitchiness, it would be Windows. Yes, I am not kidding. There'd be
> ENTERPRIIIISE CODING brilliance in there, AKA useless bloat for stuff nobody should
> EVER, IN THE HISTORY OF EVER, have access to, and countless other things.
> (up YOURS Microsoft! ) What's that, writing a driver are you? If it isn't fully descriptive
> in code, you're fired! What's that? You saved a huge number of cycles by using a Goto there?
> FIRED, we want more lines! (I'm not even kidding, Linus had to defend a Goto in a driver-level
> file, this is how mad this anti-Goto retardedness is these days, kids man) So on and so fourth.

Same problems for browsers. Remember how AOL wrecked Netscape trying to turn it into a pseudo-OS that ran on top of Windows/Linux/etc? Well, Firefox is following that same trail. Ordinary settings are being kept in f***ing SQL databases fercryinoutloud. Sqlite achieves atomic transactions and file integrity by doing an fsync() as required. This is what causes Firefox to freeze every so often. Idiots.

Chrome/Chromium is even worse. In linux it has hard-coded dependancies on dbus, udev, elfutils, and libXinerama (I only have 1 screen). More idiots.

Re:wow

[0100010001010011]

> The best package manager.

As someone who has dealt with dependency hell. I'm going to go ahead and say the deb and associated utilities (apt-get, aptitude) are the best I've ever dealt with.

Unless you're talking about Ubuntu. Where anyone and their mother can start their own PPA with no qualifications. I like the strict adherence to packaging standards that it takes to get into the debian repository.

Re:wow

[Simon Brooke]

Slackware is great for what it is. I remember working with it in the early 90s. I think Arch is also excellent. But none of these fill the same niche as RHEL.

I too used Slackware in the early nineties - after SLS on its fifty-mumble floppy disks, Then I used Red Hat, Mandrake, and even Caldera before I found Debian in about 1996. Once you've used Debian, and the Debian package manager, you're never going to want to use anything else.

Re:wow

[e70838]

Minix still exists. The purpose of Linus Torvalds was to learn assembly and in particular how intel cpu were supporting multitasking and memory protection. Linus then criticize minix idea of message passing saying it was incompatible with an optimal usage of cpu performances. They are still fighting.

Re:wow

[Truedat]

If you don't close your sarcasm tags, my sarcasm parser will get messed up and my whole day gets very confusing.

And we wouldn't want that, would we?

Why is this a story?

[Anonymous Coward]

Linus bitches and moans about the size of every release candidate. Better that broken stuff gets fixed now rather than with an ever-lengthenng string of point releases after the fact.

Re:Why is this a story?

[vlm]

Linus bitches and moans about the size of every release candidate. Better that broken stuff gets fixed now rather than with an ever-lengthenng string of point releases after the fact.

The kernel's always pushed the limits of memory, compilers... Here's a typical example from a little over 20 years ago from usenet

From: torvalds@klaava.Helsinki.FI (Linus Benedict Torvalds)
Newsgroups: comp.os.linux
Subject: Re: Help, can't compile 0.95a!
Date: 3 Apr 92 21:27:41 GMT
Organization: University of Helsinki

In article wjb@cogsci.cog.jhu.edu
(Bill Bogstad) writes:
>
> I have a 8 Meg system and also am having problems compiling fork.c.
>I would have thought that would have been sufficient....

Ok, the problem isn't memory: it's gcc-1.40. For some strange reason
the older gcc runs out of registers when optimizing some of the files in
the linux source distribution, and dies. This one isn't the same bug as
the "unknown insn" which was due to my hacks in the earlier 1.40 - this
one seems to be a genuine gcc bug.

Linux 0.95a is compileable with the older gcc if you just add the flag
"-fcombine-regs" to the command line. In fact, the only thing you need
to do is to remove a "#" from the makefiles: the line

#GCC_OPT = -fcombine-regs

should be uncommented, and gcc-1.40 will have no problems compiling the
source. This was documented in some of the release-notes for 0.95, but
I guess I forgot it for 0.95a.

Why remove the flag in the first place I hear you say? Simply because
gcc-2 doesn't understand -fcombine-regs, as it seems to do the
optimizations even without asking. There are other things I had to
change in the source to get gcc-2 to compile it, but this is the only
problem that made the old gcc choke.

With the advent of an official gcc-2.1 (this weekend?), people might
want to change to that one: note however that gcc-2.1 is about twice as
big as 1.40, so it's going to be slower on machines that swap... People
with just 2M of mem might not want to upgrade (*). I like the changes
to 2.1: the code quality seems to be a lot better (esp floating point).

On a slightly related note: the as-binary in newgcc has been reported by
several people to have problems. Getting as from the original
gcc-distribution by me (gccbin.tar.Z) might be a good idea if you have
problems with the newgcc version.

                Linus

(*) Even with only 2M of mem, using gcc-2 has it's good points. The
shared libraries should cut down on memory use as well as loading time
and disk-space use. Shared libraries work even with 1.40 if you know how
to build them, but 2.1 does it all automatically...

Re:Why is this a story?

[JimCanuck]

Are you sure you belong on /. ?

There are floppy disk Linux distributions. There has been for quite some time. Last I checked a floppy disk is only 1.44MB.

Let alone in 1992, a 8MB RAM system was on the higher end of a typical desktop.

Re:Why is this a story?

[Jah-Wren Ryel]

Read busy's other posts, he's a smart ass. It was even a little bit funny. I wouldn't have modded him down for it.

Re:Why is this a story?

[mwvdlee]

You seem to think a compiler inherently requires at least as much memory as the binary it is compiling will end up using.
Why?

Hold on a second.

[Millennium]

If I'm reading the article correctly, this isn't so much about file size as about the number of bugs fixed. Or rather, how many bugs still needed fixing in what was supposed to be the seventh release candidate of the kernel: something one would not expect to find so many bugs in very quickly.

Is this the case?

Re:Hold on a second.

[dacut]

In the actual e-mail, it's about both size and change velocity:

Because I last week I thought that making an -rc7 was not necessarily realy required, except perhaps mainly to check the late printk changes. But then today and yesterday, I got a ton of small pull requests, and now I find myself releasing an -rc7 that is actually bigger than rc6 was.

Re:Hold on a second.

[Anonymous Coward]

Linus is mainly complaining because he wants bugfixes to come in during the merge window. The RC's are then used to iron out bugs that got added by features that were added during the merge window OR to fix existing bugs that were too invasive to fix in a normal 3.x.x update. The idea is that the change from 3.4 to 3.5-rc1 is massive, 3.5rc-1 to 3.5rc2 is smaller, 3.5rc2 to rc3 is even smaller. And it keeps getting smaller until the number of commits is very low, and those commits are very small changes themselves. This SHOULD have been 3.5 release, but instead a ton of large commits were done after rc6 and that makes Linus uncomfortable about labeling 3.5 as Stable until people have a change to test out those new commits. The more commits people do past like rc2, the longer the delay until 3.5 is marked as stable and released, honestly unless im forgetting something, I havent seen a 7th release candidate for any kernel since the change to 3.0, most of them have been capping around 5. By a 7th RC there shouldnt be really anything going on unless an email comes in that is labeled "URGENT KERNEL PANIC FIX" and from the sounds of it...none of these were that, and could have all been saved for the merge window for 3.6. Instead we have the 3.5 kernel delayed by another week.

Re:Hold on a second.

[dkleinsc]

It seems like part of what he's trying to point out here is that there may be developers trying to cram in what are really new features into 3.5 by declaring them bugs and pushing them into RC's, rather than waiting until the next release. This behavior wouldn't surprise me in the least.

Re:Hold on a second.

[Cyrano de Maniac]

The way to achieve what you say Linus wants is for him to reject/postpone changes that fall outside RC criteria. "Sorry, the train has left the station. There's another one due to leave at 3.6." When developers learn that the development phase criteria are enforced they will adjust their behavior to fall in line, but contrapositively they will not adjust their behavior if the criteria are not enforced.

My sympathy is miniscule -- if RC-appropriate changes are what he wants then he should reject/postpone the changes in question as falling outside RC criteria instead of kvetching about them. It's a self-made and self-perpetuated problem; developers will abuse largesse only as long as they are allowed to.

Re:Hold on a second.

[Kjella]

The way to achieve what you say Linus wants is for him to reject/postpone changes that fall outside RC criteria. "Sorry, the train has left the station. There's another one due to leave at 3.6." When developers learn that the development phase criteria are enforced they will adjust their behavior to fall in line, but contrapositively they will not adjust their behavior if the criteria are not enforced.

He does. All the time. And people try bending the rules and stretching the definitions. All the time. You make it sound like Linus only had to tell them once and everybody'd go "well alright then" but it's more like a horny teenager with a girl on the back row of the cinema. No matter how many times those hands are pushed back they'll be back in a slightly different way or after another round of sweet talk. For those of you who have no idea what I'm talking about or what this "girl" thing is, you can imagine it's like the lobbyists in politics. No matter how many times a bill is defeated they'll keep pushing for new laws that amount to the same. In all three cases they just don't quit until they succeed.

WHO HE THINKS HE IS!

[Anonymous Coward]

Who the hell this Linus thinks he is by criticizing Linux development??!111?

Linus Says Something

[mwolfe38]

"Not cool guys." - linus OHMYGHOSH, front page news.

Re:Linus Says Something

[mwolfe38]

I actually am a huge linux and linus fan. I just thought this was a pretty "meh" article.

Re:Negative coding

[swillden]

Sounds like the kernel could use a good refactoring.

Because too many people contributed too many patches during a window in the development cycle when not many (or large) patches should be contributed?

Umm... I think you didn't understand what the problem is here. It's a violation of development process protocol that has nothing to do with the quality of the code. Someone trying to submit refactoring patches would have made it much worse, not better. Actually, it wouldn't have been worse, because Linus would just have rejected them at this point in time.

Re:Size in source or binary terms?

[rrohbeck]

No.

Re:Negative coding

[busyqth]

Sounds like the kernel could use a good refactoring.

Let's recode the whole thing, and this time, we'll do it RIGHT!

Re:Size in source or binary terms?

[jgrahn]

Are we talking about source code size, or the actual binary footprint on any individual supported system?

Neither. He's talking about the size of the diff from the previous release candidate (although it's impossible to tell from TFA).

slashverdicrap

[Anonymous Coward]

This networkworld.com article gets submitted to /.:

A host of small modifications and a large number of system-on-a-chip and PowerPC fixes inflated the size of release candidate No. 7 for Version 3.5 of the Linux kernel, according to curator Linus Torvalds' RC7 announcement, made on Saturday.

LAST TIME AROUND: Linux kernel 3.4 released

Torvalds wasn't happy with the extensive changes, most of which he said he received Friday and Saturday, saying "not cool, guys" in the announcement. However, the occasionally combustible kernel curator didn't appear to view this as a major setback.

"Now, admittedly, most of this is pretty small. The loadavg calculation fix patch is pretty big, but quite a lot of that is added comments," he wrote, referring to the subroutine that measures system workload.

However, he noted, there were also the assorted changes for SoCs, PowerPC compatibility, USB and audio to be folded in, forcing a comparatively large RC7.

"Ok, so it's still not *huge*, but it's bigger than -rc6 was. I had hoped for less," wrote Torvalds.

He also hopes that it won't be necessary to deploy an eighth release candidate before Version 3.5 of the kernel can be properly rolled out, and urged the community to "go forth and test."

Among the biggest new features expected in Linux 3.5 is enhanced compatibility with the ARM processor family, which are used in a wide array of low-cost computing devices. Several ARM-related fixes are part of 3.5-RC7, according to the official announcement email and changelog.

The H-Online reported earlier today that the final version of Linux 3.5 should be deployed next weekend, if all goes well with RC7.

The h-online.com article the networkworld one is a rehashing of:

Over the weekend, Linus Torvalds reluctantly published a seventh release candidate (RC7) for the 3.5 Linux kernel. In the LKML announcement email, the Linux creator says that he originally thought another RC would not necessarily be required; however, a large number of small pull requests submitted by developers late last week necessitated an additional RC for testing, leading Torvalds to tell the developers, "Not cool, guys. Not cool."

These changes include media fixes, random SOC fixes and PowerPC fixes, as well as patches for the leap second bug that caused Linux systems to freeze because of permanent high CPU loads that resulted in increased power consumption and wasted electricity. "Ok, so it's still not *huge*, but it's bigger than -rc6 was," said Torvalds, adding, "I had hoped for less."

Linus has asked the kernel developers to test the rc7 release to "make sure it's all good", and is hoping that he "won't have to do an -rc8". Barring any major problems over the coming week, Linux3.5 will likely be released next weekend. An overview of the changes made in the 3.5 kernel can be found in TheH's Kernel Log mini-series "Coming in 3.5" which examines the various subsystem developments in the upcoming release.

Review each article and notice what is and what is not a link, and where the links lead.

Re:Negative coding

[sconeu]

The HURD guys would like a word with you...

Re:Negative coding

[Anonymous Coward]

Like every large software project it deserves a rewrite from scratch because it's full of cruft, but nobody will ever find the time to do it.
At least some refactoring and de-crufting is done from time to time if some dev gets pissed off enough. Not something that happens in commercial SW development unless the code is hopelessly broken.

Every time someone says this they should be forced to sit in the corner and and copy this essay by Joel Spolsky on things you should never do 5000 times and give a copy to each of their friends together with an essay about what they have learned from this punishment.

Re:Negative coding

[dbIII]

No they wouldn't. IMHO the reason HURD has moved so slowly is because they told just about everyone who was interested in helping to fuck off. Linus was a bit more diplomatic even when people without much of a clue want to join in so it went rapidly from a small group to what we see today. Some of those clueless newbies he was not rude to and didn't scare away 15+ years ago are now a very long way from being clueless newbies.