×

Announcing: Slashdot Deals - Explore geek apps, games, gadgets and more. (what is this?)

Thank you!

We are sorry to see you leave - Beta is different and we value the time you took to try it out. Before you decide to go, please take a look at some value-adds for Beta and learn more about it. Thank you for reading Slashdot, and for making the site better!

Linus Torvalds: "GCC 4.9.0 Seems To Be Terminally Broken"

timothy posted about 4 months ago | from the you'll-never-believe-what-he-actually-said dept.

Bug 739

hypnosec (2231454) writes to point out a pointed critique from Linus Torvalds of GCC 4.9.0. after a random panic was discovered in a load balance function in Linux 3.16-rc6. in an email to the Linux kernel mailing list outlining two separate but possibly related bugs, Linus describes the compiler as "terminally broken," and worse ("pure and utter sh*t," only with no asterisk). A slice: "Lookie here, your compiler does some absolutely insane things with the spilling, including spilling a *constant*. For chrissake, that compiler shouldn't have been allowed to graduate from kindergarten. We're talking "sloth that was dropped on the head as a baby" level retardation levels here .... Anyway, this is not a kernel bug. This is your compiler creating completely broken code. We may need to add a warning to make sure nobody compiles with gcc-4.9.0, and the Debian people should probably downgrate their shiny new compiler."

Sorry! There are no comments related to the filter you selected.

Great. Now the sloth community... (5, Funny)

jpellino (202698) | about 4 months ago | (#47544467)

is going to be gunning for an apology...

Re:Great. Now the sloth community... (5, Funny)

Anonymous Coward | about 4 months ago | (#47544481)

I'm a sloth and I thought it was funny.

Re:Great. Now the sloth community... (2, Funny)

Anonymous Coward | about 4 months ago | (#47544621)

^^^ Self-hating sloth ^^^

I know you're trying to be funny, but... (3, Insightful)

SuperBanana (662181) | about 4 months ago | (#47544709)

...the people who deserve the apology are the people who were subject to an abusive tirade.

You can point out someone made a mistake. There's no obligation to be "nice" when doing so. There is an obligation to not be abusive, which is what Linus repeatedly does. Abuse includes mockery, ridicule, name calling, etc.

He's being a bully, pure and simple - using his popularity to shove around others. That should not be tolerated, full stop.

Re:I know you're trying to be funny, but... (5, Funny)

HornWumpus (783565) | about 4 months ago | (#47544729)

This is abuse, arguments are in the next article.

Re:I know you're trying to be funny, but... (4, Insightful)

SuricouRaven (1897204) | about 4 months ago | (#47544763)

His manner is coarse, but you must admit that he's gotten the job done. Linux advances on schedule, patches get incorporated, code gets tested, and all proceeds smoothly.

Re:I know you're trying to be funny, but... (4, Insightful)

Concerned Onlooker (473481) | about 4 months ago | (#47544841)

And Mussolini made the trains run on time?

Re:I know you're trying to be funny, but... (1)

jeffb (2.718) (1189693) | about 4 months ago | (#47544843)

So you're saying that the trains run on time?

Re:I know you're trying to be funny, but... (1, Insightful)

SuperBanana (662181) | about 4 months ago | (#47544867)

His manner is coarse

It's not "coarse", it's abusive. Namecalling, mocking, ridicule, hyperbole. That's abuse.

you must admit that he's gotten the job done. Linux advances on schedule, patches get incorporated, code gets tested, and all proceeds smoothly.

I sacrificed a chicken yesterday and successfully committed code. You must admit that the ritualistic sacrifice got the job done.

("Getting the job done" does not, and has never required being abusive to others. Getting the job done while being abusive is not proof that being abusive is required or even was part of, "getting the job done.")

Re:I know you're trying to be funny, but... (4, Insightful)

mark_reh (2015546) | about 4 months ago | (#47544769)

I disagree. Political correctness of the 90s has led to today's "rule" that all opinions are valid and of equal value when anyone with a functional brain knows they are not. Linus is merely saying what ought to be said. Maybe a little public ridicule will be the motivating force to ensure that this sort of thing doesn't happen again.

Would you prefer we hand the developers an award for "participation"?

Strawman argument (0, Troll)

SuperBanana (662181) | about 4 months ago | (#47544833)

Strawman argument. Nobody except you has posited that "all opinions are valid", and nobody suggested that criticisms can't be made. You invented that position to attack it.

I specifically said: it's fine to tell people they did something wrong. What you may not do is be abusive.

Re:I know you're trying to be funny, but... (1)

ganjadude (952775) | about 4 months ago | (#47544777)

and you know what? sometimes that what it takes to get things done. If you dont like it? stop doing it! You are working for *FREE*, you are right, you DONT have to take the abuse if you dont want. no one is holding a gun to anyones head over it.

Re:I know you're trying to be funny, but... (1)

Anonymous Coward | about 4 months ago | (#47544797)

"There is an obligation to not be abusive, which is what Linus repeatedly does. "

If GCC 4.9.0 is supposed to be a stale release and it doesn't work, why can't Linus call it shit?

I was moderately amused by the typos (1)

jpellino (202698) | about 4 months ago | (#47544827)

in the ginned-up outrage over the mis-typed code.

Re:I know you're trying to be funny, but... (-1, Troll)

hairyfeet (841228) | about 4 months ago | (#47544857)

Torvalds has shown repeatedly he has ZERO professionalism and the fact that so many in the FOSS community make excuses for his childish amateur hour behavior is truly sad. There is NO EXCUSE for acting like a spoiled rotten 14 year old Halo player and the fact that Torvalds can continue acting this childish and get away with it is a pretty damning commentary on the Linux community and its support of diva behavior.

Re:I know you're trying to be funny, but... (5, Insightful)

epyT-R (613989) | about 4 months ago | (#47544879)

The only way to cajole passive aggressives into taking action is to shame them publicly. That's why they whine like little babies and play the victim role. Unfortunately, passive aggressive dynamics run so much of society nowadays that they've been codified into law as 'rights', and the bluntness/truth that contradicts them, 'hate crimes.'

A well placed 'Fuck you!' can save years and years of mountainous, expensive political tugs of war.

Re:I know you're trying to be funny, but... (-1)

Anonymous Coward | about 4 months ago | (#47544895)

...the people who deserve the apology are the people who were subject to an abusive tirade.

You can point out someone made a mistake. There's no obligation to be "nice" when doing so. There is an obligation to not be abusive, which is what Linus repeatedly does. Abuse includes mockery, ridicule, name calling, etc.

He's being a bully, pure and simple - using his popularity to shove around others. That should not be tolerated, full stop.

I see. So, I suppose it would be unfair to return the favor and call him an immature fucking shit-tard who doesn't deserve to have a voice in the Linux community anymore regardless of his contributions?

Sorry, shit-tard may be demeaning to dung beetles. I meant fucktard.

Point is anyone can voice their opinion. Whether or not the community allows the rantings of a fucktard to be the voice of the community is perhaps more the fault of the community who continues to grant an immature boy king his crown.

Re:Great. Now the sloth community... (3, Funny)

FatdogHaiku (978357) | about 4 months ago | (#47544753)

is going to be gunning for an apology...

Maybe, but it take a while for them to get worked up enough to do anything...

Job Vacancies (0)

Anonymous Coward | about 4 months ago | (#47544469)

"sloth that was dropped on the head as a baby"

I don't know about coding but I have part-time vacancies in trolling CNN articles if he's interested.

gcc is like congress (5, Funny)

NemoinSpace (1118137) | about 4 months ago | (#47544471)

It works better when they do nothing

"pointed" (5, Funny)

Arancaytar (966377) | about 4 months ago | (#47544479)

From Torvalds, that's uncharacteristically friendly.

Or upgrade to llvm ... (4, Insightful)

perpenso (1613749) | about 4 months ago | (#47544487)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being above to compile with either gcc or llvm would be a nice option.

Being able ... (1)

perpenso (1613749) | about 4 months ago | (#47544495)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being above to compile with either gcc or llvm would be a nice option.

That should have been "being able". One day I will have to start proofreading.

Re:Or upgrade to llvm ... (-1)

Anonymous Coward | about 4 months ago | (#47544553)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being above to compile with either gcc or llvm would be a nice option.

How could you _upgrade_ from GPLv3 to BSD? Sounds like the reverse.

Re:Or upgrade to llvm ... (0, Funny)

Anonymous Coward | about 4 months ago | (#47544633)

Yeah, because we all know that being a license bitch is much more important than using the right tool for the job. Now that the shoe is on the other foot, GPL zealots can't handle it.

Compiler doesn't change the license ... (5, Informative)

perpenso (1613749) | about 4 months ago | (#47544637)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being [able] to compile with either gcc or llvm would be a nice option.

How could you _upgrade_ from GPLv3 to BSD? Sounds like the reverse.

Compiling with a BSD licensed compiler does not change the license of the software being compiled. Linux would remain gpl regardless of whether gcc or llvm is used.

Re:Compiler doesn't change the license ... (-1)

Anonymous Coward | about 4 months ago | (#47544685)

The main problem is that Linus did not take copyright assignments, so it's practially impossible now to relicense the Linux kernel or upgrade it to GPLv3.

Therefore, always remember to use "GPLv3 or later" when you release software. The "or later" is really important.

Re:Compiler doesn't change the license ... (1)

HornWumpus (783565) | about 4 months ago | (#47544747)

You make that sound like a bad thing.

The problem is Linux can't upgrade to BSD, but you can...

Re:Compiler doesn't change the license ... (0)

armanox (826486) | about 4 months ago | (#47544755)

The main problem is that Linus did not take copyright assignments, so it's practially impossible now to relicense the Linux kernel or upgrade it to GPLv3.

Therefore, always remember to use "GPLv3 or later" when you release software. The "or later" is really important.

You make it sound like we care about GPLv3. Sorry, I'll stick to licenses like the BSD ones when I can.

Re:Compiler doesn't change the license ... (1, Funny)

Anonymous Coward | about 4 months ago | (#47544799)

A lot of people care about GPLv3, it's the current version of what is essentially regarded as the default free software license.

Re:Compiler doesn't change the license ... (0)

Anonymous Coward | about 4 months ago | (#47544757)

If he did, and tried to change the license to GPLv3 linux would be forked to hell.

"or later" ... its a blank check ... its a trap ! (1)

perpenso (1613749) | about 4 months ago | (#47544813)

The main problem is that Linus did not take copyright assignments, so it's practially impossible now to relicense the Linux kernel or upgrade it to GPLv3. Therefore, always remember to use "GPLv3 or later" when you release software. The "or later" is really important.

No, Linus did the right thing. "Or later" is very dangerous, its a blank check, its an unknown, ... its a trap! We have no idea what some future GPL license may include. It may include things that we do not want. As some developers who are staying with gpl 2 intentionally have said about gpl 3.

That said, the above is off topic. Compiling the kernel with llvm does not change the license of the kernel. A BSD licensed compiler has no effect on GPL licensed source code. The resulting binary derives its license from its source, not the compiler used.

Re:Or upgrade to llvm ... (0)

Anonymous Coward | about 4 months ago | (#47544619)

- Upgrade to llvm.
- The kernel doesn't compile.
- Upgrade to GCC.
- Hit some bug.
- Upgrade to llvm ...

When do you stop the loop?

Re:Or upgrade to llvm ... (1)

Anonymous Coward | about 4 months ago | (#47544635)

At one point someone will inevitably try to build it with Visual Studio, and then it's over.

Re:Or upgrade to llvm ... (0)

Anonymous Coward | about 4 months ago | (#47544679)

Not likely. It will take them until the Heat Death of the Universe to rewrite all those Makefiles as Visual Studio project files, and then convert the C source into that subset of C++ that Visual Studio calls C these days.

Need many compilers? Try Gentoo... (1)

gentryx (759438) | about 4 months ago | (#47544713)

On my machine, I have clang 3.4.2-r100, gcc 4.9.0, 4.8.3, 4.7.4, 4.6.4 and icc 14.0.3.174 installed. All simultaneously, no hassle.

Re:Or upgrade to llvm ... (5, Insightful)

maccodemonkey (1438585) | about 4 months ago | (#47544783)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being above to compile with either gcc or llvm would be a nice option.

It's amazed me how attitudes from LLVM in the last few years has gone from:
"OMG WTF why is Apple doing their own compiler why can't they contribute to GCC they want to make everything proprietary"
To:
"Hey, LLVM is a pretty great tool."

I'm happy to see LLVM making such great strides.

Re:Or upgrade to llvm ... (-1)

Anonymous Coward | about 4 months ago | (#47544851)

Well, when clang (which is what I presume you mean) can actually compile a programming language (rather than c/c++), I'll give it a try.

(Snarky is as snarky says.)

Or upgrade to llvm ... (1)

voltorb (2668983) | about 4 months ago | (#47544883)

Debian people should probably downgrate their shiny new compiler.

Or upgrade to llvm. Being above to compile with either gcc or llvm would be a nice option.

"Update to to icc", that I would understand (for Intel platforms). "Upgrade to LLVM" sounds like this is not coming from a C++ programmer who really cares about the final binary (yes, GCC still produces better binaries is most situations). And given the issue at hand, it seems you're under the delusion that LLVM is bug-free and doesn't miscompile stuff. Yes, LLVM gives prettier error messages. Yes, LLVM code-base is module. Yes, Apple fan boys love it. Yes, it's BSD licensed and plays along nice with people who don't really care about software freedom. Nothing I care about here. But, LLVM codebase is as bug-ridden as GCC or any other codebase, don't pretend otherwise.

Surprise, surprise... (-1)

QuietLagoon (813062) | about 4 months ago | (#47544493)

Linus is having another meltdown, complaining about something he doesn't like.

Re:Surprise, surprise... (1, Insightful)

amiga3D (567632) | about 4 months ago | (#47544535)

He doesn't complain, he blows the fuck up.

Re:Surprise, surprise... (1)

BitZtream (692029) | about 4 months ago | (#47544623)

And every time he does ... he's right to do so.

A meltdown is when some candy ass can't deal with reality and blows up emotionally without provocation or justification.

When Linus blows up, he's pretty much always right and its pretty much always after the other guy(s) repeatedly denied being wrong or acted like an asshole who couldn't possibly be wrong.

When has anyone seen Linus blow up on someone who didn't actually deserve it?

GCC deserves to die.

Re:Surprise, surprise... (0)

Sun (104778) | about 4 months ago | (#47544645)

Not that discussion again....

Linus blowing up at Andrew Tridgdell for "reverse engineering" the bitkeeper protocol comes to mind.

I will agree that this time around, the complaints are grounded. It does, indeed, seem like a compiler bug. Whether that is a reason to be so critical of gcc 4.9.0, I don't know. It's obvious a serious problem for the kernel.

Shachar

Re:Surprise, surprise... (0)

Anonymous Coward | about 4 months ago | (#47544707)

Exactly this. Linus has done far more than the no talent N00bs here on slashdot that whine every time. Those people wish they could do 1/10th what he does.

And honestly, I agree with Linus, most of the time when crap goes sideways it's because of very stupid things and people need to be called stupid to their face when thy do something stupid.

GCC 4.9.0 is a steaming pile of crap, and they know they should have never released it.

Re:Surprise, surprise... (3, Funny)

PlusFiveTroll (754249) | about 4 months ago | (#47544551)

I don't 'like' lima beans, that doesn't mean there is anything wrong with them. In this case your lima beans have been switched out with castor beans and you're dying. Pretty big difference.

Re:Surprise, surprise... (2)

QuietLagoon (813062) | about 4 months ago | (#47544601)

I don't 'like' lima beans, that doesn't mean there is anything wrong with them. In this case your lima beans have been switched out with castor beans and you're dying. Pretty big difference.

It's a compiler bug, for Pete's sake, not the end of the world.

.
Revert to a known working version of the compiler, submit a bug report and move on. Why the temper tantrums? What is with all the drama?

Re:Surprise, surprise... (1)

jones_supa (887896) | about 4 months ago | (#47544615)

These days I interpret Linus' "meltdowns" just as some funny nerd rage. He uses that technique to strongly underline the importance of his point, it's never real anger. Often there's a dash of humor in the mix, such as in this case the comment "that compiler shouldn't have been allowed to graduate from kindergarten".

Re:Surprise, surprise... (0)

Anonymous Coward | about 4 months ago | (#47544789)

That's not humor. It's hyperbole intended to inflame and insult people behind what he's complaining about. I'm much more impressed by people who simply state the problem clearly and concisely.

Re:Surprise, surprise... (0)

Anonymous Coward | about 4 months ago | (#47544887)

I'm much more impressed by people who don't post anonymously when critiquing other people's means of communicating.

(Yes, I know I'm posting anonymously.)

Re:Surprise, surprise... (1)

philip.paradis (2580427) | about 4 months ago | (#47544889)

You seem to be making the implication that it's not okay for Linus to loudly complain about a compiler that produces a broken Linux kernel. Why is that?

Not just version 4.9 (4, Informative)

kthreadd (1558445) | about 4 months ago | (#47544503)

According to the actual bug report this problem seams to have started in 4.5. They only triggered it in 4.9.

https://gcc.gnu.org/bugzilla/s... [gnu.org]

Already fixed. (4, Informative)

Bananenrepublik (49759) | about 4 months ago | (#47544653)

... and they already fixed the bug two weeks ago, see the bug report [gnu.org]

Re:Not just version 4.9 (1)

jones_supa (887896) | about 4 months ago | (#47544655)

Another worrying thing which Linus says there is:

"The problem is that these things are a bitch to debug - they turn into these completely impossible kernel oopses or corruption, and we were just very lucky that this one case happened to be repeatable and pinpoint for two people. Are there others? We have no way of knowing.."

Each new GCC version has regressions (2)

loufoque (1400831) | about 4 months ago | (#47544517)

4.8 was even worse.
That's hardly news.

Oe noes! A compiler bug! (5, Interesting)

serviscope_minor (664417) | about 4 months ago | (#47544527)

Oh gosh a compiler bug! The world is going to end and GCC is terminally broken for ever and ever and ever. Life happens, and occasionally that includes compiler bugs. I've seen fewer bugs in GCC than any other production compiler ever.

Anyway, it seems like GCC is implementing a very obscure compiler option incorrectly in some circumstances which causes a crash.

But of course this is cue for lamentations of how awful and braindead GCC is and so much drama.

End result, the GCC people will fix this bug in short order (what are GCC point releases for anyway), and distributers will probably have a patch package out for 4.9.0 before 4.9.1 ships (what are distributors for anyway?) and the world will keep turning and GCC will go back from being the buggy broken braindead piece of shit to yet again being the most solid production compiler in existence.

It's a little ironic that the he's so quick to attack the GCC people. The success of Linux is 100% built off the success of GCC. There have been no other credible compilers for Linux throughout the majority of its existence and without GCC being bulletproof, Linux would never have been solid.

Re:Oe noes! A compiler bug! (0, Offtopic)

amiga3D (567632) | about 4 months ago | (#47544575)

I'm just wishing the Debian folks would do something about putting ffmpeg back in the repo. That shit about the ffmpeg that wasn't really ffmpeg caused a lot of trouble with breakage. I'm not even going to mention that buggy, smelly pile of shit called wodim that made burning optical media on linux go from easy to massive frustration. One silver lining though, I've learned to compile from source.

Re:Oe noes! A compiler bug! (2)

QuietLagoon (813062) | about 4 months ago | (#47544617)

...It's a little ironic that the he's so quick to attack the GCC people. ...

Especially since the Linux kernel has always been bug-free.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544793)

Does Linus hinder the GCC development?

What happened here is that someone compiled the kernel with GCC. The result was buggy and crashed. It landed on Linus table when it should have been reported to the GCC developers.
Linus spend time tracking down a bug that he wasn't responsible for and out of his reach to fix.
He has the right to be cranky.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544647)

torvalds is a drama queen. anything that someone else broke is a huge fire, while his mistakes are also somehow your fault.

Re:Oe noes! A compiler bug! (0, Flamebait)

BitZtream (692029) | about 4 months ago | (#47544651)

End result, the GCC people will fix this bug in short order (what are GCC point releases for anyway)

The bug was reported 4 point releases ago. It just now started effecting the kernel.

Claiming the GCC crew will 'fix this bug in short order' is like claiming Obama is leading the charge in transparent government.

GCC has never been a solid production compiler.

The success of Linux is 100% built off the success of GCC.

You have that pretty much backwards. Without Linux, GCC wouldn't matter to anyone. Linux can be built with other compilers with a little effort, ask Intel about it.

There have been no other credible compilers for Linux throughout the majority of its existence

You're pretty clueless. Intel would beg to differ. No one that matters compiles high performance code on GCC, they use the Intel compiler.

Re:Oe noes! A compiler bug! (4, Insightful)

NotInHere (3654617) | about 4 months ago | (#47544721)

End result, the GCC people will fix this bug in short order (what are GCC point releases for anyway)

The bug was reported 4 point releases ago. It just now started effecting the kernel.

In fact, it has been fixed in trunk even before Linus' rant.
https://gcc.gnu.org/bugzilla/s... [gnu.org]

Re:Oe noes! A compiler bug! (5, Funny)

mvdwege (243851) | about 4 months ago | (#47544743)

Claiming the GCC crew will 'fix this bug in short order' is like claiming Obama is leading the charge in transparent government.

Since the bug has already been fixed [gnu.org] , I suppose this means you'll be wholeheartedly endorsing Obama now?

Re:Oe noes! A compiler bug! (4, Interesting)

serviscope_minor (664417) | about 4 months ago | (#47544749)

Claiming the GCC crew will 'fix this bug in short order' is like claiming Obama is leading the charge in transparent government.

Care to take a wager that it will be fixed in 4.9.1?

GCC has never been a solid production compiler.

Utter crap, bordering on an outright lie.

The list of systems I've developed on is something like: PCs, Sun, SGI, HP, AIX, PICs, 8051, Blackfin, ARM, AVR and probably a bunch I've forgotten about. I've used very many compilers over the years. There has not been one that can match GCC in solidity and general lack of bugs.

Not. A. Single. One.

But apparently "production compiler" in your world means something completely different.

You have that pretty much backwards. Without Linux, GCC wouldn't matter to anyone. Linux can be built with other compilers with a little effort, ask Intel about it.

As far as I know, it's been GCC, Intel (which is useless for Linux's most popular platforms) and TCC. I don't know of anyone who actually uses a non GCC compiled kernel.

You're pretty clueless. Intel would beg to differ. No one that matters compiles high performance code on GCC, they use the Intel compiler.

Ah no TRUE scotsman would use GCC. Got it.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544765)

You're pretty clueless. Intel would beg to differ. No one that matters compiles high performance code on GCC, they use the Intel compiler.

Unless they want to not use Intel processors, because the Intel compiler was known to cripple the performance [agner.org] on non-Intel processors. I'm also wondering now who is buying the PathScale [wikipedia.org] compiler?

Re:Oe noes! A compiler bug! (1)

cheesybagel (670288) | about 4 months ago | (#47544817)

The bug happens to manifest on version 4.9.0. How many Linux kernel versions have you know to be bugfree.

GCC is the most robust compiler around. Don't compare it with ICC which breaks compiling anything other than microbenchmarks.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544699)

Linus is very consistent with attacking people. Creating buggy code is fine. Pushing it upstream is not. Releasing it as stable is a mortal sin.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544703)

Oh gosh a compiler bug! The world is going to end and GCC is terminally broken for ever and ever and ever. Life happens, and occasionally that includes compiler bugs. I've seen fewer bugs in GCC than any other production compiler ever.

Anyway, it seems like GCC is implementing a very obscure compiler option incorrectly in some circumstances which causes a crash.

But of course this is cue for lamentations of how awful and braindead GCC is and so much drama.

End result, the GCC people will fix this bug in short order (what are GCC point releases for anyway), and distributers will probably have a patch package out for 4.9.0 before 4.9.1 ships (what are distributors for anyway?) and the world will keep turning and GCC will go back from being the buggy broken braindead piece of shit to yet again being the most solid production compiler in existence.

It's a little ironic that the he's so quick to attack the GCC people. The success of Linux is 100% built off the success of GCC. There have been no other credible compilers for Linux throughout the majority of its existence and without GCC being bulletproof, Linux would never have been solid.

Well, that didn't take long.

Oe noes! "Naughty" language! (0)

Anonymous Coward | about 4 months ago | (#47544739)

Linus describes the compiler as "terminally broken," and worse ("pure and utter sh*t," only with no asterisk).

So he said "shit". Big deal. It's just a word. Not like the "replacement" obscures it any, or even gets pronounced differently. Seriously, shit without the i (sht) sounds pretty much identical. But somehow, that maintains the "pureness" of the quoter, or something.

Some people really need to learn to grow the fuck up when it comes to language.

Re:Oe noes! A compiler bug! (0)

armanox (826486) | about 4 months ago | (#47544791)

I've seen fewer bugs in GCC than any other production compiler ever.

Anyway, it seems like GCC is implementing a very obscure compiler option incorrectly in some circumstances which causes a crash.

But of course this is cue for lamentations of how awful and braindead GCC is and so much drama.

End result, the GCC people will fix this bug in short order (what are GCC point releases for anyway), and distributers will probably have a patch package out for 4.9.0 before 4.9.1 ships (what are distributors for anyway?) and the world will keep turning and GCC will go back from being the buggy broken braindead piece of shit to yet again being the most solid production compiler in existence.

It's a little ironic that the he's so quick to attack the GCC people. The success of Linux is 100% built off the success of GCC. There have been no other credible compilers for Linux throughout the majority of its existence and without GCC being bulletproof, Linux would never have been solid.

Really? I hate GCC with a passion. Otherwise, we wouldn't have 'gccisms' that affect code portability to other POSIX platforms. GCC can go DIAF for all I care.

Re:Oe noes! A compiler bug! (2)

serviscope_minor (664417) | about 4 months ago | (#47544823)

Really? I hate GCC with a passion. Otherwise, we wouldn't have 'gccisms' that affect code portability to other POSIX platforms. GCC can go DIAF for all I care.

So, tell me mr GCC hater, which compilers don't have non-portable extensions?

Re:Oe noes! A compiler bug! (0)

evilviper (135110) | about 4 months ago | (#47544801)

GCC is terminally broken for ever and ever and ever.

GCC is a mess that has been getting consistently worse since 3.0. It's so bad that compiling GCC with GCC, with any CPU optimizations enabled, produces a non-working compiler. It just keeps getting bigger and slower, and has a great many proprietary GCC-isms that open source developers keep using, not even realizing they're bugs.

The crappiness of GCC has driven tons of people away, and spurred the development of LLVM, tcc, and others.

I've seen fewer bugs in GCC than any other production compiler ever.

Either you're not looking (myopia is fun), or you have very little experience with other modern compilers.

End result, the GCC people will fix this bug in short order

With this much publicity, they might... But major bugs that get reported, but don't hit the /. front page, and often linger for year after year.

The success of Linux is 100% built off the success of GCC. There have been no other credible compilers for Linux throughout the majority of its existence

Only true if you drink rms' kool-aid... Otherwise, any of the proprietary compilers out there would have done the job just fine. Or Linux developers would have put some effort into getting another compiler up-to-par for their purposes if nothing had been available, kinda like they did with the kernel...

And now, the competition is just waiting to break-through and rid us of all the GCC nonsense.

Re:Oe noes! A compiler bug! (4, Insightful)

serviscope_minor (664417) | about 4 months ago | (#47544847)

GCC is a mess that has been getting consistently worse since 3.0. It's so bad that compiling GCC with GCC, with any CPU optimizations enabled, produces a non-working compiler.

[citation needed]

It just keeps getting bigger and slower, and has a great many proprietary GCC-isms that open source developers keep using, not even realizing they're bugs.

Every single compiler out there offers nonportable extensions.

Either you're not looking (myopia is fun), or you have very little experience with other modern compilers.

I could level the same complaint at you. The other compilers have more. Like full up ICE crashes.

Only true if you drink rms' kool-aid...

Ah and now we get to the root of it. You've decided to throw logic and facts to the wind and just go off on an I-hate-RMS-so-I-hate-GCC-by-proxy rant.

Grow up.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544899)

It's a little ironic that the he's so quick to attack the GCC people.

Most likely he prepares the community for the eventual upgrade to pcc compiler.

Re:Oe noes! A compiler bug! (0)

Anonymous Coward | about 4 months ago | (#47544903)

I agree with this sentiment. Sadly with clang/llvm there is now a compiler shoot-out where correctness and developer features are secondary. They get fixed when exposed in embarrassing ways like this, so good on Linus for having a rant and making GCC escalate this bug.

Take this example with clearly unreachable dead code:

int foo(int x) {
    return x;
    return x+x;
}

Any programmer would expect this to choke the compiler. However, with -Wall both GCC and clang will compile with no warnings. Look at the "goto fail" fiasco.

C compiler writers are locked into a compiler performance war and making the community suffer. I doubt I will notice the marginal speed ups from a GCC upgrade, most performance issues are tied into software and the libraries they use. I do notice when things break and when my development time is increased by the compiler not having warnings that safe guard me from myself. I feel this is where Linus' rant is coming from.

Fwiw, the example above results in an unreachable statement warning when the code is written in Java.

The reddit thread... (0, Interesting)

Anonymous Coward | about 4 months ago | (#47544537)

...which has been going a lot longer, has good discussion. The problem occurs when an uncommon combination of flags is selected in specific versions. Most people don't have to worry about this too much.

Re:The reddit thread... (1)

HornWumpus (783565) | about 4 months ago | (#47544781)

I don't believe it, reddit is full of grammarian twits.

This could be good... (1)

freeze128 (544774) | about 4 months ago | (#47544571)

Linus developed the linux kernel, and it's obviously very important to him. He didn't like the version control system that he was using so he invented a better one (GIT). Perhaps this will encourage him to create a better, less "retarded" compiler.

Re:This could be good... (1)

TechyImmigrant (175943) | about 4 months ago | (#47544599)

If he built gcc like git, each compiler would have it's own local language that other compilers could mix and match with the languages from other gccs.

Re:This could be good... (1, Flamebait)

BitZtream (692029) | about 4 months ago | (#47544663)

Perhaps this will encourage him to create a better, less "retarded" compiler.

Or, like everyone else, he'll just switch to clang/llvm. GCCs days are numbered. RMS's political agenda has ensured its demise. It'll be a slow painful death, but its a death none the less.

Re: This could be good... (0)

Anonymous Coward | about 4 months ago | (#47544733)

He used to call sparse ( https://sparse.wiki.kernel.org/index.php/Main_Page ) a compiler. But the idea was to get good compiler warnings and not optimized code.

What is rbp anyway? (1)

jones_supa (887896) | about 4 months ago | (#47544577)

Can you explain what is the %rbp, and why are its offsets negative? If I set values 1, 2, 3, 4 to a bunch of variables, I get this assembly code:

movl $1, -4(%rbp)
movl $2, -8(%rbp)
movl $3, -12(%rbp)
movl $4, -16(%rbp)

Re: What is rbp anyway? (5, Informative)

Anonymous Coward | about 4 months ago | (#47544669)

rbp is a stack pointer [hackerschool.com] it is negative because it grows down.

Re: What is rbp anyway? (1)

jones_supa (887896) | about 4 months ago | (#47544727)

Nice article, that should explain a bunch of things to me.

Re:What is rbp anyway? (0)

Anonymous Coward | about 4 months ago | (#47544677)

at&t syntax assembler
64bit base pointer, i'm sure you're a troll
and the negative values on there represent stack variables
as + values from rbp [assuming a correct stack frame] would be +(%rbp)

troll?

Re:What is rbp anyway? (0)

Anonymous Coward | about 4 months ago | (#47544687)

at&t syntax intel assembler*

Re:What is rbp anyway? (1)

jones_supa (887896) | about 4 months ago | (#47544693)

Troll, troll, yeah, I'm so troll...

Re:What is rbp anyway? (1)

cheesybagel (670288) | about 4 months ago | (#47544835)

RBP is the 64-bit version of the X86 BP register. It is used to point to the beginning of the stack frame. Are those 'variables' function arguments?

It is just fetching values directly behind the end of the stack frame.

Re:What is rbp anyway? (3, Informative)

profplump (309017) | about 4 months ago | (#47544849)

%rpb is the base pointer. The offset is negative because the local variables live below the base pointer (being in-scope only after the base pointer is moved to the current function); -X(%rbp) points to a local variable.

http://stackoverflow.com/quest... [stackoverflow.com]

Why the asterisk? (4, Insightful)

Stormwatch (703920) | about 4 months ago | (#47544589)

If you want to say shit, say shit. We're all grown-ups here.

Re:Why the asterisk? (4, Insightful)

BitZtream (692029) | about 4 months ago | (#47544675)

Yea, my first thought when I read that was 'How retarded is timothy to think that implying a word without using the actual word is any different'. Its stupid. Its like saying 'the n word'. We all know EXACTLY what you mean, you're just too chicken shit to actually say it.

Oh, and he had to tell us that Linus didn't use the asterisks!

Re:Why the asterisk? (1)

michelcolman (1208008) | about 4 months ago | (#47544785)

On some internet connections with restrictive filters, pages containing "dirty words" are automatically blocked. Happens to me every now and then when I'm surfing at work, even on innocent sites. I did get them to whitelist Slashdot, though. Other people might not be so lucky.

Linux, a miracle (1, Troll)

demon driver (1046738) | about 4 months ago | (#47544609)

I keep wondering how Linux could become as good as it is, with a coordinator being a person like Torvalds. How many capable developers would put up with a boss like that in their day job? Yet they do working for Torvalds in their spare time...

Re:Linux, a miracle (1)

michelcolman (1208008) | about 4 months ago | (#47544691)

And how could Apple become as good as it is? Hmmm...

Re:Linux, a miracle (1)

AmazingRuss (555076) | about 4 months ago | (#47544711)

All the ones that aren't being yelled at love watching him yell at somebody else.

Re:Linux, a miracle (0)

Anonymous Coward | about 4 months ago | (#47544717)

You're quite a moron, probably washing toilets at McDonalds.
Hard things are obtained by effort and this effort is leaving its marks on people's behaviour.
You cannot be nice when your product(Linux) which runs on m(b)illions equipments may end in
failure because of other people mistakes.

Btw , all other people working on kernel are thankfull to Torvalds for his continous effort.
Look on Linux kernel mailing list , Torvalds was the only one who figured out the problem
(ie that panic in load_balance() is a gcc problem ).

Re:Linux, a miracle (1, Flamebait)

hackus (159037) | about 4 months ago | (#47544767)

The reason why LINUX is as good as it is, because people who are involved with LINUX are passionate about great software. The reason why LINUX attracts people who are passionate about good software and world class engineering is because you can actually do something about these sorts of errors.

With proprietary software like Windows, that passion is crushed because you can never ever fix these sorts of things, without going to jail in the f*cking fascist corporate state.

Tolerance is lower, because YOU as a software developer have EPIC amounts of control over the software stack. You are suppose to test these things before deployment. On Windows or some other piece of shit operating system with no source code this sort of thing is tolerated much more frequently because you do not have that level of control and more importantly: Defects in proprietary software are thought of as a REVENUE STREAM, not as a PIECE OF SHIT SOFTWARE.

AS A RESULT OF THAT LOSS OF CONTROL THE SOFTWARE QUALITY EXPECTATIONS ARE VASTY REDUCED IN PROPRIETARY SOFTWARE SHIT.

Hence Windows shit, vs LINUX EPIC engineering.

That EPIC engineering runs the fastest machines, the largest computers, the smallest computers do you know why?

The open source community have passionate people who don't put up with BULLSHIT BOSSES. THE KIND YOU LIKE TO HAVE AROUND AT WORK THINKING ABOUT NEXT QUARTERS NUMBERS INSTEAD OF WHY HIS CUSTOMER STAYED UP LAST NIGHT TILL 2AM IN THE MORNING ON A TECHNICAL SUPPORT LINE TRYING TO RESTORE THEIR SYSTEMS.

I figure Windows shit owes me like 5 years of sleep as a IT Monkey going on 25 years in the industry. I always think, why didn't someone yell at the guy like LINUS who wrote this shit, to fix it?

Keep yelling LINUS!!

Strange censorship (5, Funny)

evilviper (135110) | about 4 months ago | (#47544715)

"pure and utter sh*t," only with no asterisk

So he actually called it, "pure and utter sht"?

Re:Strange censorship (4, Insightful)

MindPrison (864299) | about 4 months ago | (#47544863)

"pure and utter sh*t," only with no asterisk

So he actually called it, "pure and utter sht"?

Remember, that's the guy that said "Nvidia, F*CK YOU!". http://www.youtube.com/watch?v... [youtube.com]
He's a passionate, colorful character...he's been like that for as long as I can remember him. Thank god there's still awesome nerds out there that just can't be bought.

Why does anyone work with this guy? (-1)

Anonymous Coward | about 4 months ago | (#47544741)

Seriously why does anyone work on this project? I'm not a programmer, so this isn't my thing.

But Linux is not seeing serious adoption outside of programmers/technical folks. I'm sorry, but it's true; Linux marketshare is nothing.

The product is all over the board. There's so many developers doing so many things, but no one simple thing to enable adoption.

The collaborative nature of the project is held in check because of the attitude of the one guy leading it, which amounts to little more than nerd rage and bullying people he doesn't like or agree with. His approach to handling the Linux community is entirely antagonistic, cronyist, and unprofessional, which is directly resulting in the projects lack of adoption.

At least this is how it looks like to an outsider, non-programmer, so is there something I'm missing? Why work on a project that is not succeeding simply because of it's tyrannical project lead?

Oh look! Really? (0)

Anonymous Coward | about 4 months ago | (#47544837)

Linus has become an unbalanced psycho that needs his meds. Really. Grow up, or at least have a nappy change and another prozac
before turning on the tablet dude.

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?