How to destroy a brand-new Samsung laptop: Boot Linux on it Linux users accidentally bricked their new Samsung laptops by booting their favourite open-source OS on the shiny computers. A kernel driver crashes on Sammy machines when users start up from an Ubuntu 12 USB key - although other distributions may be at risk - giving them the dreaded black screen of no activity whatsoever. …
> Well if you will run freeware crap, you get what you pay for....
Yes you're right, they should have used a closed source proprietary EFI solution and charged the users thousands per seat shouldn't they?
We know that you only get quality from paying thousands to highly paid consultants.
This post has been deleted by its author
I have a Samsung VCR (yes the kind that plays VHS tapes) from 1995 that just keeps going and going... A few years ago I thought its time had come due to being replaced by DVD players, but now it happily plays children's films (obtained for free, or nearly free, from fleamarkets and recycling centers) for my son.
OTOH my Samsung LCD TV died after 14 months of very light use - it worked.. locked to one channel only and at one volume only - refusing to accept inputs from either the supplied remote, the correctly programmed Sky remote, or the few buttons mounted on the side of the set. According to the service agents it was a motherboard fault - the Warranty company just wrote me a cheque and I rectified my mistake with a nice Panny Smart Plasma that has performed faultlessly for 18 months now.
I have a galaxy s3, but I wouldn't touch their other stuff...
Just give it a little while longer and your Galaxy S3 might well give up the ghost as well. Seems to be a known problem after about 6-9 months whereby there's a serious mainboard failure and the device is completely dead. Happened to mine 2 days ago (7 months old).
And isn't that mainboard problem because of the flex of the phone? That gives the phone that plasticky feeling and seriously detracts from perceived quality, and puts additional strain on the mobo and all soldering.
Sometimes the perceived quality indicates the real quality. The components are great, but they seem put together in a way that seriously reduces the quality.
Well, we're still using ~60 Samsung 1600x1200 monitors bought about six years ago and format-wise preferable to the 1920x1080 ones you get today. (They also pivot for 1200x1600 if you like that).
None failed in warranty. Two have failed since. I blame London Electricity for frying their power supplies.
Don't generalise from a small sample, or from a single product. All manufacturers occasionally ship lemons. The real test is how much hassle the warranty service puts you through. I don't have mucjh experience with Samsung because their products don't seem to fail during warranty. (And they stay reliable after). IIyama are very good. Acer and Dell were once so bad that we stopped buying from them.
This post has been deleted by its author
Your case in point of Jap cars being quality British built, (often regarded as better built & in higher productive factories than Japan) illustrates what happens when you remove the cronied, nepotist, corrupt. arrogant, Ox bridged BA'd old style British management.
Now all we have to do is remove them from the expensive inefficient legal, banking, city "investment", parliamentary systems & the recession will be sorted!
On the other hand, I have a Samsung Freesat STB, and its a pile of crap (an SMT-S7800). Just do a search for it.
Regularly loses its library of recorded programmes, and required a hard reset every moth or so to reset the increasingly erratic record timer. If you hit record on an SD programme, it will ask you if you want to record the HD version. Answer yes and it will record whatever is on the HD channel at the same time, even if it isn't the programme you want. Freezes about once a month, and every time you have to reset the thing (which is frequently) you have to re-schedule all of your recordings as the timer settings won't survive a reset.
So. while Samsung make good stuff (I bought the box on the name). they are just as cabable of producing a pup as anyone else.
@ Eponymous Cowherd: All the faults you describe pretty much match issues I had with a Sky Digibox back circa 2004ish. I couldn't believe the times when looking at the EPG caused the thing to freeze. Eventually I gave up on it and got myself a cheap generic satellite receiver and a DiSEqC switch. Haven't looked back since / Good riddance.
No prizes for guessing which cult you're a member of.
FWIW, in the interests of full disclosure, I own a Samsung TV and a Samsung monitor at home, both of which are sturdy, attractive, and function well. I don't own a Samsung phone, because I tihnk there are actually better alternatives out there. Those alternatives aren't made by Apple.
Samsung produces rather all right products on multiple IT segments. Yes, sometimes they feel cheap and plastic-y (printers for example), but their display products and SSD's are top of the pops.
What's bothering me is that Samsung support on all those segments are pretty weak and documentation is your standard far-east fare, i.e. if the product even is documented it is riddled with bad English translations and spellign mitsakes, and I have never seen any quality software come from Samsung. I fucking hate with passion Samsung's Digital Signage software (doesn't work well, documentation provided on Powerpoint files?) and don't get me started on the laptop utilities.
IBM, HP and other big houses are partly more expensive because they usually document everything ad nauseam and they support their products very well, producing bug fixes, driver/firmware updates to satisfactory levels. Samsung doesn't.
Much experience in French cars then. As far as I know Citroen has been the number one in the Rally world for some seven years. I would rather consider US cars as junk due to poor fuel economy and as ever very poor driving characteristics. Sorry, prefer what ever you like, and why not cars produced in GB, but this was about loading Linux on a new Samsung laptop. If a problem, annoying indeed, but how in hell has this anything to do with people buying French cars. And if they do, why in hell, do you think they do it if they think they are junk.
The French have been doing well, with cars, ever since they invented them.
This article was about software, some kernel driver, that Samsung fucked up.
I drove my Pug for nearly 16 years, so I can safely comment on how good French cars were. I'm not keen on the latest models, but my 1995 Peugeot 306, owned from new, rarely let me down. With 15 years past and well over 180000 miles on the clock, yes, I had problems, but I defy anyone to have a car that old and that travelled that hasn't had problems.
Mind you, the successors to the 306 look pretty poor, IMHO!
I've been driving French cars since 1990...
First my mother's old Citroën GS, then I bought a used Citroën CX which lasted me many years(until my shoddy servicing killed the engine).
The BX which replaced it had a slight fault. Seems it can't take being backed over by a 6WD dump-truck.
My first Berlingo was killed in a parking lot because of a SUV with clueless owner.
My second Berlingo... Well... the sunroof(ragtop type) doesn't work... yet...
(All my cars have been bought secondhand. )
There's no problem keeping French cars running, as long as you show them the same attention as you should your wife. If you don't.. (if you skip servicing...) you suffer endless agony...
I have had 3 French cars in my time, and one of them was the most reliable car I have ever owned (while another one was the 2nd worst).
I had a Renault 25 (widely regarded as being one of the biggest piles of crap ever to take to the road). I had it for 12 years and, in that time, never needed anything other than servicing. It eventually gave up the ghost when the gearbox went, but it was 16 years old, so pretty good going.
The worst car was a Ford Granada. 6 years old, had it 18 months and it contrived a simultaneous failure of both the cylinder head and auto box.
The best car was a Nissan Primera. Rock solid, sold it because I needed something bigger. Now have a Peugeot 406 1.8 petrol estate (station wagon to those in the US). Had it just over 3 years and has been solidly reliable in that time.
"The French have been doing well, with cars, ever since they invented them." WTF?
http://wiki.answers.com/Q/Who_invented_the_car
French Jesuit Ferdinand Verbiest invented the first steam car in China in the year 1672.
Nicolas Joseph Cugnot of France built a steam automobile in 1769.
Gustave Trouve of France designed and invented the first electric car in 1881.
(Karl Benz was the German mechanical engineer who designed, patented and in 1885 built the world's first
practical automobile to be powered by a gasoline engine. )
http://www.infoplease.com/ipa/A0004637.html
first true automobile, not carriage with motor) René Panhard, Emile Lavassor, France, 1891
Panhard et Lavassor http://en.wikipedia.org/wiki/Panhard
They were one of the biggest manufacturers very early on in Europe, and invented many technologies which are still used today in other cars. Like Panhard rods, putting the motor in front with rear wheel drive (Systeme Panhard), modern transmission with sliding gears etc.
"The French have been doing well, with cars, ever since they invented them."
Well, Mr. Benz might have something to say about that!
That aside, every now and again a comment saying "...like people buying <xyz/> cars, not realizing how overrated they are...". Best thing to do is not feed the trolls, especially not after midnight...
Have owned two albeit, BL tinged, British cars, three French cars (but not Renault), and one Japanese car and have driven many of them around the clock. The French cars were agreeably reliable, the Japanese one almost faultless and the British cars . . . . ask Red Robbo as he should be very proud of what he did to the wonderful British car industry and can probably tell you why the cars made in his era were so wonderful too.
Red Robbo destroyed the "wonderful British car industry"? What about poor management, uninspired design, and poor quality control?
Managers control businesses, not unions and workers. Rover had 12 years of private ownership to turn into a quality car maker, but the only bit that survives is Land rover, which went to Ford when BMW were in charge.
Jaguar's quality was useless until Ford bought them and started putting in some real QA, now they're a benchmark for quality, and JLR are storming ahead.
The "wonderful British car industry" was complacent and badly managed; they thought that, after WW2, the world owed them a living and that the upstarts from Germany and Japan were a joke. Thirty years on Britain makes more cars than ever, but for American, Japanese, German, and Indian companies.
"People buy French cars because they think they're good, when in actual fact they're total junk."
You may have a point, but the same could be said for VW and Volvo, who appear to be much the same as Citroen/Peugeot, i.e. slightly below average reliability.
BMW/Audi are a little better, about the same as Skoda, but not as good as Hyundai.
It'š not looking too good for LR or Renault though.....
Top 7 companies listed are all Japanese.
http://www.thisismoney.co.uk/money/cars/article-1633078/Japanese-cars-most-reliable.html
Last three cars (Renault, Citroen, Citroen) have been reliable. Might not win rallies, might be crap as a getaway car from that legendary bank heist everybody dreams about (you apparently need a Mini for that, saw this film once...); however on a freezing cold damp horrible winter morning, you want to get into your car, turn the key, and have something happen.
Especially when you have a job to go to. Reliability is important.
Aren't Dell, HP and Asus/Acer better bets for standards-compliant UEFIs and BIOSes ? What's a good recommend on the Linux HCLs?
I won't say anything bad about Toshiba (today). It would be in shards all over the floor if I hadn't been brought up to not waste things. "Och aye wee man, ye've nay finishit yer parritch?".
Clenches fist and grits teeth and starts frothing like in "Downfall" ;)
But surely we should all be buying nice tablets by now? And I don't mean Valium, although that might be appropriate in this case ;)
I thought that 'bricked' meant that your shiny item of kit had assumed the functionality of a brick, but that it was recoverable if you knew how to do the necessary digital incantations.
You seem to think it means that your shiny item of kit has suffered the equivalent of having a brick thrown through it.
We need a formal and agreed glossary of terms.
I thought that 'bricked' meant that your shiny item of kit had assumed the functionality of a brick, but that it was recoverable if you knew how to do the necessary digital incantations.
You're half right. 'Bricked' does imply the functionality of a brick, but it's not always possible to recover from such a state. There are device that when bricked stay that way and there's no way to fix them short of replacing parts. Even with devices that can normally be unbricked it's sometimes possible to brick them so badly that the unbricking techniques won't work.
There are different levels of bricking. Each start with the same problem: useless machine. But "soft" bricks can usually be remedied by switching the machine to different or alternate modes of operation that go around the piece of software/firmware causing the brick. Harder bricks usually affect some piece of software that can't be avoided. In this case, we're dealing with something worse, a "baked" brick in which some hardware or firmware component has been damaged, instantly rendering any fallback worthless because the failure is either in the fallback or affect something above it.
Its a boot-time firmware problem unrelated to the OS. Use of the technical term "bricked"...
O RLY? You may want to take that up with the author of the article, the title of which is:
How to destroy a brand-new Samsung laptop: Boot Linux on it
I don't see anything in the article that indicates Windows users are having the same problem.
"I don't see anything in the article that indicates Windows users are having the same problem."
I laughed too - at you. Ross K, sir, obviously lack knowledge on the subject matter and therefore have to rely solely on the author's explanation. Perhaps you should consider reading about something you know about so you can engage your brain. Or perhaps when encountering something you do not understand, rather than making assumptions, actually study about it.
I laughed too - at you. Ross K, sir, obviously lack knowledge on the subject matter and therefore have to rely solely on the author's explanation. Perhaps you should consider reading about something you know about so you can engage your brain. Or perhaps when encountering something you do not understand, rather than making assumptions, actually study about it.
Wow. You sir, posting as an Anonymous Coward, really stuck it to me.
Instead of relying on what the author of an article says, I will do as you say instead and "actually study about it."
I now realise the error of my ways, and won't criticise a crowd-sourced operating system again
It's not unrelated to the OS true, but as has been said the blame lies squarely with Samsung.
It should not be possible to kill a UEFI firmware simply due to loading a bad driver.
In fact the driver isn't even bad, it works fine on other models of laptop.
But whether you point to the UEFI or the kernel module as the cause, I'm pretty sure it's Samsung's fault since as far as I'm aware the samsung-laptop module is code actually code contributed to Linux by Samsung.
Well that's the thing of it isn't it? When I choose an OS I choose one I know will work.
This issue aside, I still have problems installing Linux on my Sony Vaio. Every year I rebuild it, and every year I start with a version of Linux (last was Ubuntu) to see how it fares. Never ever has a Linux install !just worked!. There's forever a %&$£ problem with drivers for ^$%£ something. And that Vaio is 5 years old now, so there shouldn't be.
Issues like this don't help Linux. Pointing fingers and blaming someone, irrespective of who that might be, doesn't help Linux.
every year I start with a version of Linux (last was Ubuntu) to see how it fares. Never ever has a Linux install !just worked!.
Next time try Mint or Mepis. Ubuntu's always been a steaming pile of crap despite all the press it gets. I've never seen it just work on anything, and on most of the machines that I've tried it on even Debian took less effort to get up. Mint and Mepis, on the other hand, have never let me down when 'just works' is the goal (though I do still prefer Debian).
In my experience, the various Ubuntu releases work on everything I've put it on with few or no problems.
In the last 8 years since I started using it, I've put it on lots of Thinkpads and other laptops, netbooks, and desktop systems, and while I won't say that I've never had problems, none of them have been show-stoppers.
OK, when I first used my eeePC 701, I had wireless problems until the slightly strange Atheros chipset gained a Linux driver. My Thinkpad T30 does not reset the sound correctly after suspend, and the Mobile Radeon graphics adapter is too old to work with Compiz well, and I came across a wireless card for which there was no Linux support on a Shuttle XPS (which, incidental, did not work in Windows very well either).
I suspect that your Vaio must have some very specific hardware in it, and only works on Windows because you have a system restore image prepared by Sony that contains the right drivers. I would be interested in seeing how well you managed to get it working with a retail windows install disk, and what would not work.
Windows users think that their systems 'just work', but this is mainly because the PC manufacturer has taken the necessary background steps of identifying the drivers and building a bundle of Windows and drivers specifically for their systems. If they went to the same lengths for Linux, it would be the same.
What is amazing in my view is that a single build (one CD, not even a DVD) of, say, Ubuntu will 'just work' on a huge number of different systems without all of the behind the scenes customisations that happen for Windows, because they are done for you.
One of the problems is that Windows drivers are specific to a particular instance of hardware, so a Atheros card from say Netgear would not work with the drivers supplied by Belkin for a card with the same chipset, and often not even with the driver for another card from the same supplier with the same chipset.
Linux is different here, because its drivers are largely manufacturer agnostic. It identifies something like an Atheros chipset, and it configures the driver regardless of the manufacturer (OK, I know Atheros is an old chipset not used much now, but it's the one that came to mind first).
Occasionally, you will come across some hardware for which the PCI or USB ID's are not in the database, so the module code cannot identify the required driver correctly, but this becomes less and less frequent as time goes by, and is usually fixed for what must be regarded as non-mainstream hardware (if it were mainstream, the ID's would be in the database) after a little Googling. Not everyone's forte, I accept, but you cant expect the distro maintainers to be omnipotent!
You're insisting on using hardware from Sony, one of the most anti-FOSS, anti-customer company on Earth and you're blaming Linux ? Heck, even Windows service packs have to be downloaded from Sony because they've been heavily modified. Stop using Linux, you're wasting your time. And ours too.
One was a tenny tiny pocket sized thing that just about managed to run the Win 2000 that came with it and you could have a network cable, or a USB port, but not both at the same time
THe other is an old A series Vaoi. It's been running Xubuntu (and Kubuntu) quite happily for a couple of years now
"Indeed. These freetards write a poor quality driver that crashes hard enough the damage the firmware, and then blame the hardware manufacturer."
A few points:
1) The driver that bricks the Samsung laptops wasnt written by 'freetards', from what I read, it was written by Samsung. Strike one
2) If the hardware is properly designed, it shoudn't be possible to cause it to break using software. Samsung aren't the first to be guilty of this particular crime, but it's still pretty damn stupid, and no doubt embarrassing for the hardware manufacturer. Strike two
3) Calling Linux users 'freetards' is just plain ignorant. I assume you are not aware that much of the internet runs on Linux servers running Apache, or that many universities and research establishments use this operating system for various purposes, or indeed that it is a good way for people with a serious interest in many aspects of computing to gain some knowledge and experience of unix-type operating systems and their quirks. Use of the term to describe such people is... well, it's retarded. Strike three, you're out.
1) It was written by Greg Kroah-Hartman who is not a Samsung employee.
2) Ah, so Windows causes the same issue? Oh yeah, it doesn't.
3) They are freetards. They get their OS for free, the applications for free and when they break they expect and unrelated hardware manufacturer to replace it for free and then fix the problem the freetards caused. For free.
1) Greg Kroah-Hartman has written it based on Samsung provided API and specs
2) Now it will because the news is rapidly spreading. There is a precise sequence of binary instructions that will brick the machine. Do you still think Windows can't be "upgraded" to do the same ?
3) It's your problem if you pay for Windows, not ours. And just to upset your stomach, Samsung will fix it for free because of 2) above.
Continue laughing.
"3) They are freetards. They get their OS for free, the applications for free and when they break they expect and unrelated hardware manufacturer to replace it for free and then fix the problem the freetards caused. For free."
But you or the other AC said they wrote it. I agree that freeloaders (and that is the appropriate term because the same folks will happily use pirated software) are a pain. However, they don't write code. That's what's the problem with them. But at least they make good testers. ;)
"Indeed. These freetards write a poor quality driver that crashes hard enough the damage the firmware, and then blame the hardware manufacturer."
Engineers often rely on pre-built components. They expect them to work as advertised. When they don't it's their fault, by your definition.
Isn't this a shining example of why those damned Eulas/Licence agreements exist.
--->>>> Install your own software at your own risk..... <<<<--------
I am not critisizing Linux or Samsung, I am merely pointing out that installing your own ROMS, OSes etc is normally something which is done at your own risk.
it's the equivalant of me chipping the engine of my car, BMW refuse to garauntee the engine if I make mods...
By the way the article does not mention if the problem also exists when reinstalling windows or any other OSes....It wold be more interesting to know if this was a generic problem or a UbuntuLinux problem.
If it is a generic problem then the Ubuntu installers have just cause, if not well ........that's life.
Quote: "No actually, it isn't the equivalent of you chipping your BMW. If anything it's more the equivalent of you fitting non BMW tyres and them blaming this [insert option here]".
Which they do. BMW, Porche and some other from the list of our "favorite" car companies have homologation programs and will refuse warranties if you have fitted non-homologated tyres, used non-homologated oil, etc.
This post has been deleted by its author
> it's the equivalant of me chipping the engine of my car, BMW refuse to garauntee the engine if I make mods...
No it isn't. You are not adding or removing anything from the laptop.
You are plugging a USB stick into a USB port and rebooting. The laptop should be able to handle this without bricking.
"And it can - so long as some dodgy driver (not written, endorsed or supported by the manufacturer) doesn't barf all over the firmware."
The point is no matter what is on the USB stick, it shouldn't be able to barf all over the firmware as you put it. Modifications to the firmware should only be allowed by trusted means and not a random bit of code on a USB stick. (e.g. my BIOS can only be updated through itself, where it makes a backup before the operation to prevent bricking).
@Condiment
You are adding to the machine.
You are adding a set of instructions, from teh Linux Kernel, that only come into existance under certain circumstances.. These instructions do not exist without the USB key - as I mentioned before it is the equivalant of modding your machine.
Without the USB key no problem.
With the USB key problem.
Please note the need for the "additional" USB key before the problems are generated.
Interesting (but wrong) analogy to chipping cars, and how ironic BMW was chosen as an example, because if anything this is more like the 'Nikasil' engine liner fiasco that that very company was eventually forced to accept responsibility for after initial denial.
BMW used Nikasil as a strengthener in its engines. Then owners in certain territories (including the UK) started suffering premature engine failure. BMW denied there was any issue as the majority of cars with the material were unaffected. When it was established that higher sulphur content in the petrol sold in these regions was the cause of the eroding of the Nikasil plating BMW claimed they were vindicated, until it was pointed out to them that the maximum permissible sulphur content in petrol is laid down in International standards and that BMW were obliged to make sure their car could run on it or state clearly in their literature otherwise.The fuel the afflicted owners had been using was still within those standards and BMW had never bothered to test the car at the extreme levels - just ran it on the nice low sulphur fuel they had available in their domestic market during testing.
Cue u-turn and costly engine replacement program, even many years after end of warranty.
Can't see this one ending nicely though.
well as with all thg8ings it seems to be a combination...
of a Linux driver based on Samsung code which may, or may not, get loaded depending on which boot method you use, and which may, or may not write to certain RAM areas that may, or may not, brick the machine beyond the ability to boot anything. Even repair software.
To my mind whatever else may be said, a machine that cant even boot a bios screen, is a machine whose design is faulty.
But I suppose that with bios in flash, suitable for upgrades (or malicious attacks, or sheer murphyisms) is always gonna be vulnerable.
In short if you send a machine out with even the boot loader in EAROM you will inevitably have failures and returns.
>Don't you mean Obviously!.... or am I getting my trolls mixed up?
You're getting your trolls mixed up. Obviously! thinks every Apple user is a mindless moron, it is Eadon that thinks people only use Windows because they haven't yet seen the light. Neither get much deeper than that, though Eadon usually does stick around to defend his naivety.
Strangely, I haven't yet seen them spar off each other.
Good idea! And we should all dress in blue all buttoned-up working clothes, and ride the same bike model and have the same hair cut , be allowed to have a single child that will grow up into someone simple like you. Yes, life would be so simple.
For a moment I thought you are joking but using the word freetard made me doubt it.
If Samsung missed this, then what else did they miss? What other ticking time bomb is lurking in their products. They get to go straight to the top of the PC sh*t list. There they will stay until they get their act together or someone screws up worse.
...not that they would have been the first name to come to mind even before this debacle.
I'm sorry, please explain to me why Samsung now have to fix a buggy Linux kernel driver.
Did Samsung write it?
If you don't want to run Windows, don't but a computer with Windows installed. There's loads of companies selling alternatives...no, wait; there isn't because Linux is an epic desktop fail. It destrors your computer for one.
Dear AC
"I'm sorry, please explain to me why Samsung now have to fix a buggy Linux kernel driver." Because a) they wrote it, and b) it's not that they need to fix, it's the crappy laptop that allows its firmware to be vaped by a software error.
"Did Samsung write it?"
As stated many times, yes, in this insance, but even if they hadn't they would still be responsible for building a laptop that doesn't fail to protect its firmware.
Is that so hard to understand?
Signed,
a non-Linux user.
I have to agree. I have no problem with a product failing when someone throws something at it that it isn't designed for - so if it was designed to run Windows (and just Windows), and they're clear about this, then bricking it with A.N.Other OS is not Sammy's fault.
However, as dear old Sammy then offered code that is supposed to support this functionality on a given product, then they're in the wrong. I wonder if they put a disclaimer with the code to cover them on this....?
In other words, Sammy should've offered no code rather than dangerous code.
"Secure Boot could actually improve security, unless there are bugs in EFI".
I think this settles it. If there are bugs in EFI implementations that brick your hardware... there are most likely security critical bugs in there, too. (In fact bricking the hardware might be a security problem by itself depending on the scenario)
"A kernel driver........crashes....giving them the dreaded black screen of no activity whatsoever."
Shit happens.
"...the firmware.......refuses to start up the laptop..."
Shit happens.
"....effectively ruining the product." / "...destroyed two motherboards...."
This is the sort of shit that should never happen. Crapping on the firmware I can sort of understand, but having that as even a possibility in software and not having a recovery process for it? What a shite product!
"Were there any stickers on those machines with "Designed For Linux" on them?"
Probably not, were there any stickers on it that said "designed for Windows" on it? I've not seen them for some time. The only ones I see are "Windows 7", telling you what software is on it at shipping, or "certified for Windows 8". None of them say "Can only run Windows 7" which would be a strange thing to put on an x86-64 machine. Presumably your understanding would be that it can't run Windows Vista either?
"Yeah, I put diesel in my petrol tank ... guess what?"
No, the closest analogy is you put Shell petrol in a Ford Focus, when the sticker on the cap now says "Ford recommends BP"
So why is Linux's crashy driver suddenly Samsung's problem?
Because the FW should not be able to be left in a state where it can't reboot. If you can't turn it back on, its a FW bug.
In the same way that playing different moves in a chess game shouldn't be able to leave the OS unable to boot.
@AC Let me just fix that for you?
Does Samsung support Linux?
Yes, not specifically on this product but they contribute kernel code.
Did Samsung write the Linux driver that crashes?
The code used to compile the module came from Samsung.
Did Samsung install Linux?
No. But given they contribute code to the Linux Kernel to make it work with their laptops, perhaps they should of done so to test it?
If this happened with when installing Windows XP on the laptop, because XP contained an old Samsung driver that broke the UEFI firmware would you be blaming Microsoft or Samsung?
Does Samsung support Linux?
Yes, not specifically on this product but they contribute kernel code.
So...no.
Did Samsung write the Linux driver that crashes?
The code used to compile the module came from Samsung.
And naturally Samsung had the massive vision to write code into this module for a product it probably never even knew it was going to build at the time, or even that it had Linux visions for.
But given they contribute code to the Linux Kernel to make it work with their laptops, perhaps they should of done so to test it?
Newsflash: Contributing to the kernel != automatic, full support for all Linux builds on all products.
If this happened with when installing Windows XP on the laptop, because XP contained an old Samsung driver that broke the UEFI firmware would you be blaming Microsoft or Samsung?
Neither. XP isn't supported either, from what I understand.
At the end of the day...regardless of what Samsung does on the Linux front, if it says LINUX is not supported on this hardware (or alternatively *only* OS xxxx is), then the risk is entirely yours.
Apart from the title. Is there no way of reflashing the chip? And if there is surely it would be in Samsungs best interest to provide this help.
Customer service seems to be dead these days, with most companies only interested in taking your money there and then. Surely it would make more sense to give continued support/advise so that customer comes back and buys their next product from you. Or is this wishful thinking.
Final thing where is the rest of the story, we have comments from a Senior Ubunto developer, an Intel developer, but where is the comment from Samsung??
Quoting an Ubunto developer about what someone, somewhere else who spoke to someone, somewhere at Samsung told him isnt really getting the Samsung story is it??
Why should Samsung do anything? It's not their code, it's not an OS they support and they sell the computers with a perfectly good OS pre-installed.
What next? WAH! My laptop broke when I used is scuba diving. WAH! Samsung must make it waterproof. FOR FREE! WAH!
Get a grip, feeetards. YOUR code broke it, YOU have to fix and you should pay for a new laptop.
Not such a good comparison. Better would be that your aqualung broke because you swam in one sea instead of another.
The big, screaming problem that SAMSUNG (yes, the supplier) has is that it has now created a laptop that can be totally bricked by software alone. Given that the OS it normally supports it's Windows, it can't take long before virus infections appear that will use that vulnerability to screw over even a Windows install.
There is no way you can start blaming users for wanting to install Linux. Given that users have done this for as long as Linux exists, there is a de facto recovery strategy called "reboot" that Samsung has now broken. If Samsung wants to prevent warranty claims it should fit a VERY large sticker on the machine stating "DOES NOT SUPPORT LINUX"..
Personally, I would not touch this laptop with a 10 foot pole.
It is doubtful that Samsung "support" Windows in any meaningful sense beyond putting it on the disk.
UEFI secure boot, I thought, was intended to prevent corruption of the boot process in addition, of course, to the primary goal of preventing installation of alternative operating systems.
"Why should Samsung do anything? It's not their code, it's not an OS they support and they sell the computers with a perfectly good OS pre-installed."
Soerry you didn't manage to read the comments before writing your bollocks about not being their code.
I know they sell their computers with an operating system) because UNLESS you pay over the odds or build your own (DIY laptops) we have to pay for the shite that's sold with it.
Whether or not is is 'perfectly good' is something only an MS user (I use the word 'user' advisedly) would claim.
MS astroturfing. Nice.
Now let's see, what would these shills say if a virus were to reweite the Win8 EFI loader, taking advantage of this particular bug and brick the Samsung laptops when booting Win8?
I'm old enough to remember the Chernobyl/CIH virus. Motherboard designs were changed after that, so why should Samsung dismiss this as "freetard tinkering"?
I've posted a few replies so far, but I wanted to summarise in a post of my own.
The code for linux kernel module that causes this was supplied by Samsung.
This could just as easily have happened when installing a different edition Windows than the one supplied, if there were drivers included from Samsung from a previous product that broke a current one.
Apart from in a small number of circumstances, driver software should not be able to break hardware. Not because of the driver code but because the design of the firmware.
A few years back there was a similar incident with Intel network cards that used the e1000e module, again code from the manufacturer.
You're talking far too much sense for this comments page.
Yes, there should be a way to update firmware using a booted operating system of some sort. Ideally booting from the firmware itself should be the only way.
The Linux haters seem to have their daft opinions, but there's no reason why an MS driver couldn't have done this, or malware for that matter. Someone at Samsung needs to rethink their implementation.
The Linux haters in this case are idiots. They're no better than the Linux lovers and Apple bitches who can't see through their own prejudice and insist that malware and bugs are only found on Windows.
The problem's the firmware. It will be fixed.
I don't generally like Linux (jack of all trades, master of none), but I generally agree with what's in your post. Note that the driver was written by someone on the Linux kernel team based on code from Samsung (see post #3): http://forums.opensuse.org/english/get-technical-help-here/install-boot-login/482848-booting-linux-uefi-bricks-samsung-laptops-how-do-i-get-around.html . So now it's a game of finger pointing.
But I can't help but think this problem would have been caught in testing and immediately fixed had Samsung fully supported Linux. Samsung would have tested and caught this bug if it was happening in Windows. That's what "supported OS" means. It's highly doubtful that a released MS driver would have done this -- this bug is so obvious that any test would have caught it.
I think everyone is shocked because the idea that you can try out Linux (or any other non-supported OS) on your PC without harming anything just went out the window.
Come on Phil, You've been round here long enough to know that the Anti-linux crowd on El Reg would have read these words from the article "Linux...... Bricked.......... Laptop" and that would have been enough, no need to read the whole thing to find out what I call "The facts"
I wish we could all just get along, why cant we all just be nice to each other?
Have a beer on me
This is why most laptop BIOS updaters insist on high battery charge and AC supply to perform an update, they are doing everything they can to avoid a power glitch at an inappropriate moment.
Old Intel PC motherboards used to come with a jumper that could put you into BIOS recovery mode with an 8K read-only part of the flash chip that could boot to the point where it could reload the BIOS image from a floppy. Not sure why there isn't some emergency mode available to allow a reflash of Samsung's laptop firmware.
Though not a laptop, my mate's recent Gigabyte motherboard boasts two BIOSs, one at least is UEFI (don't know what the backup BIOS is, we haven't had need of it).
[As if to take a pop at the Koreans, the motherboard has written on it "Japanese Capacitors"- reassuring, since the aged Dell Optiplex it replaced died of blown MB capacitors, allegedly the result of incomplete industrial espionage]
I bought a new Samsung 305 laptop to replace a 2003 Dell with a lowgrade Nvidia graphics controller that can run neither recent Ubuntu nor Fedora desktops (except for XFCE).
I installed Fedora 18, after pushing the Win7 installed on it out of the way. No problems so far (2 days), but UEFI is not turned on or needed. I even found good artwork for the desktop, a spherical cow at the origin, from Berkley.
Anybody want a old Dell, runs XP real well ?
Consider that a Windows 9 could have the same problem. If Samsung didn't lock out the BIOS from an OS incursion, then your re-install disc/usb drive could cause problems because all of the hdd's are imaged and there is always the possibility that the original could have problems if the BIOS were ever updated. Would updating the BIOS negate an older version of the OS that came with the machine and require an install of an SP version?
This post has been deleted by its author
This post has been deleted by its author
I'm surprised how many people are so willing to accept this Closed ecosystem, the Walled Garden as I have heard it termed.
Peronally I blame Apple, and to a slighly lesser extent Microsoft, oh and Sony too.
I remember when you bought hardware and you had to either boot off a floppy, or install the OS yourself to a 10Mb hard disk, you might have been supplied with an OS, but you could pick which one you wanted.
A lot of people here expect that when you buy a computer that you have to use the OS which is supplied at the time and that if it bricks when you try a different OS then it's your fault, my mind screams WTF?
What happened to standards, interoperability and choice for the consumer, what has happened with the Samsung product seems to me to be a mistake and a bug in the firmware yet there are a lot of people who seem to think that you shouldn't be able to run a different OS on the laptop and should you attempt to do so then it's acceptable for the motherboard to be damaged.
The OS won't be overwriting the firmware, just trying to call it's functions, it shouldn't be possible to brick it like that, the firmware should balk at self harm.
Especially given that this is a technical site, with a technical audience, it's amazing how many people completely fail to understand what is expected by many purchasers of a PC, be it desktop or laptop.
Believe it or not, you guys, one of those things is the freedom to run a different OS. Yes, that is a reasonable expectation, because more than enough people want to, and do, just that. It might not work. We all know that laptops come with devices that might not be supported --- but it should not kill it. And it certainly should not affect the warranty, because this is a computer, and hey, that's what computers are for.
To all the ignorant people who love to rant about freetards, and invite flames in return, here's mine: I hope all your linux devices fail today. You'll probably be shocked to discover just how much of a freetard you are. At least it will educate you. Assuming that's possible. Perhaps your brains no longer actually process anythinng that is not said by a consultant or a microsoft salesman. Perhaps you are consultants or MS salesmen.
Perhaps you'd also all like to consider that all of your "freetards" buying a machine like this have paid for the Microsoft operating system that comes on the machine. Whether they want it or not. Not so "freetarded," eh?
It's also amazing how many of the people calling Linux users Freetards do not understand that the Free we embrace is free as in freedom, not free as in free beer. Quite how that makes you retarded is beyond me.
If you look at the Humble Indie games takings, the biggest payers are the Linux users, a pretty good indication that Linux users are not cheapskates. I contribute to KDE and my distros of choice and many others do too because I am incredibly grateful that I have an operating system where there is no chance of hiding spyware, no crappy licences and I have the freeDOM to with it what I want.
This seems to follow a general trend.
If you ask people why they use Windows, you typically don't get very good answers. The best you get is that they want to "rely on a strong partner", or that they are vendor locked. In many cases you even get the verbal equivalent of flung shit. There are quite a few people proud of never having seen anything else than the last few versions of Windows.
Linux people usually have more in depth reasons. For example the higher quality of open source software, the longer commitment to maintenance, or at least the lower cost. For technical people Linux has become the default solution. If in doubt install Linux, it'll do the job. Most Linux people also have more knowledge about Windows than most Windows people, even though they usually claim that they just have a very basic understanding of Windows.
Of course I'm generalizing here, there are also probably Linux fanbois flinging around shit, and Windows geeks who actually love their system for good reasons.
I haven't met many proud Windows users - either they aren't that technical, and Windows is just what their computer comes with, and they use in their workplace - or they use Windows because the software they use isn't readily available for Linux. Most Windows users i know who are interested enough to actually talk about their OS can recount a thousand frustrations - from bugs to poor design choices.
I started using PCs when they were still 'IBM compatibles' not 'Windows Machines' and back then you got used to things not working just because they felt like it. Windows has usually improved over different versions- so I guess with each new version us Win users hope that "maybe this MS have nailed it". Oh well.
I have dabbled in desktop Linux and have found it is not without its issues either, though.
Here's a thought- if all OSs were fast, reliable, secure and ran all software the user wanted- the main differentiation would be the GUI.
There is no doubt at all that Samung makes some good stuff. I've been sitting in front of a Samsung monitor for the past three years. I have Samsung Air Conditioning machines which, in this tropical part of the world are considered mid-priced, mid-market, but, with the occasional (well, ok, annual, but they are about six years old) repair they work fine.
Judges may not think Samsung phones are "cool," but I certainly have nothing against them. I don't own one, but wouldn't mind. There are a vast number of people who prefer them (and/or find them more affordable) over the "cool" phone.
Given their size as parts and screen manufacturers, probably most people have at least a bit of Samsung in their lives.
I have nothing against Samsung. In this instance, they have appear to have screwed up. They deserve stick for that, and they are getting it. Nobody is vowing to never buy Samsung again. The fact that they are sharing code with linux developers is, in itself, points.
To my mind 'Bricked' means no recovery no functionality. If you can recover it it wasn't a brick it was merely a broken laptop.
If you can brick a laptop by plugging something into a USB port then it's a big problem. It doesn't matter how it was discovered. What we now know is that there is a system call on board that will fuck it right up. Bad news. For one thing it's an attack vector for bad people that do bad things and secondly it's a huge seagoing battleship mine whenever the user upgrades or installs anything.
Actually thats quite funny, you could make a bootable USB stick and pass it around, with say the latest films on it or something, and wait for the screams as it bricks something bought by a windoze user.
I am totally doing this! I love it when windoze people get a virus or something, I very rarely help them, I usually say what do you expect....
This thread is a monument to the shoddy state of IT.
Seriously. On a website that is supposed to tbe tech savy it would appear that vast numbers of commenters have no clue.
I guess we can just do away with NMIs and execution rings. From the perspective of this thread those things are less useful than product branding.
What happened to informatics?
So my guess is that Samsung now owes everyone a new motherboard? Dear me. Perhaps it's time to consider socketing the eprom, NVRAM style, lest a class action suit bring down the manufacturer.
My sympathies to all those affected, even though I don't have an evangelical stance on any particular O/S*. No-one should be subjected to this level of incompetence.
* except the pile of steaming crap that is OS9, the most mendaciously oversold OS in the history of OSes.
It just does, doesn't matter what the box is, it will be littered with bugs if it has an EFI BIOS.
I can't actually get into the BIOS setup on my Gigabyte Z77 board if the SATA mode is RAID, I've wired the clear cmos jumper to the front panel reset button to get around this problem.
This post has been deleted by its author
Well, it's obviously NOT clear what has happened - especially to you it appears.
Firmware is indeed software up-datable. BUT, its should still NEVER be possible to do something in SW that can brick the device completely. Even if you flash with the wrong software there must ALWAYS be a mechanism to get back to a state where you can flash the RIGHT software. This is a known problem with a known solution. Samsung appears to have a bug in their system, since it doesn't happen (as far as we know) on ANY OTHER DEVICE.
Actually it may be quite easy. Just walk into a friendly computer store and ask to see the products. With a nice flashy USB key in hand (attempt) to boot from it. When the device bricks up, mention it to the management "That's funny, I can't get this to work, do you have another?".
Repeat as often as necessary.
When you run through a chunk of inventory (at the store), and they no longer have items to display/sell, they will complain to Samsung and a solution WILL happen.
Being able to "brick" a computer this easily is a major failure mode. I guess it would be similar to putting the wrong brand of gasoline (petrol) in a vehicle, and causing the car to never start again at the gas station. Bad, very bad!
I'd imagine the BIOS chip can be replaced or re-flashed but it certainly is a PITA when you need your laptop and data.
Dunno about you, but I like to back up my data before doing an OS upgrade (or downgrade in this case). Anybody who doesn't, and then bitches about losing data is a moron.
I'd imagine most linux users have more than one computer, given how technically proficient they are. Restoring the previously backed up data shouldn't be a "PITA" at all, should it?
Those that are saying "it works with Windows, therefore It Works" are behind the times.
Microsoft used to have that attitude. But about a dozen years ago they very suddenly understood the flaw with that approach. Most of the things that "worked with Windows" and didn't work with Linux, actually didn't work.
They just happened to not crash immediately with Windows 95 and perhaps Windows 98.
Once Microsoft put a priority on an OS that didn't require a daily reboot, and tried to move to improved API (e.g. 32 bit, eliminate the need to rely on undocumented state in registers and variables) they found that almost every hardware issue that Linux faced also bit them when they tried to introduce an updated OS.
Almost certainly this Linux-triggered bug would have otherwise lay hidden, waiting to bite them with Windows++.
This post has been deleted by its author
Wow, you lot take your operating systems seriously, don't you?
I haven't seen so much downvoting of comments since... Well actually I don't know when I last saw a group of people get their knickers in such a knot.
Even Apple zealots don't get as worked up as a linux user on his high horse.
Two points:
Yes, OS's are serious. If you don't think OS's are serious, then perhaps you had better retire from the IT scene.
There are OS's that are used to service huge amounts of websites, super computers, banks, phones, tablets cars, planes etc. YOU rely of OS's (of whatever flavours) to do almost everything in your day to day life. And you don't think that is 'serious' ??
As to this particular point, the downvotes are because people like yourself are entirely missing why this Samsung issue (yes SS, not Linux) is a big thing. I personally don't care which OS you use (I use Windows, Ubuntu, Raspbian etc), and have the understanding and knowledge to know this is a firmware issue that could affect ANY OS you run on the device in question. Not just Linux.
I made the mistake of purchasing a 14" Series NP530U39 a few months ago. They are terrible bits of kit.
The first one arrived with a dead drive.
The second one kind of works
The ExpressCache means that the blasted thing keeps corrupting the filesystem randomly and I need to run a system restore because Chkdsk won't run (autochk.exe is corrupted.)
After a few months the case is starting to warp (I treat the thing carefully although I do need to carry it around.)
You can't turn off the Bluetooth chip without turning off the Wireless network.
The screen is a cheap job that has a terrible viewing angle.
The function keys don't work most of the time.
They keyboard keys are getting loose.
The wireless network adapter just dies at random.
The thing is terrible. I really miss the Asus kit I used to stick with in the past.
I wonder how many of the people slagging off Linux users have ever considered:
1: If anything they use is Linux based or based on OpenSource work such as Android phones and Mac OSX or their home router
2: If they have ever visited a site run on an Apache web server - oh wait what does the reg use?!
3: If they have ever used an application like OpenOffice
4: If they realise that RedHat most certainly do charge for 'linux' and it is therefore not the 'free' they think we all want but a different meaning to 'free'
5: Do they really think we are all cost motivated?
6: I wonder if they would use Windows for their corporate firewall?
Although there's news in it this time, I still think this is likely to be the special Friday Favourite - the story written to make sure there's a slew of rabid fanbois, haters, likers and appeasers spouting out on the forums.
It's an unusual one, but Linux failing will usually do it, along with any story containing Apple suing, Microsoft succeeding or failing, Google being evil, or the US and the UK having a disagreement over language.
I have a 530U3C, it's actually a nice piece of kit and very good value.
First thing I did was kill all the weird stuff in the BIOS, swap out the hard disk and put in a 256GB SSD and install Mint and plain Windows 7 in a dual-boot configuration. Windows is on the new SSD and Linux is on the 22GB internal SSD that's there as some sort of cache by default.
Extreme distrust of anything new or unusual continues to pay dividends for me.
Yep, "learning from our mistakes" is one thing, "learning from other people's mistakes" is another. Never use version x.1 of anything, search forums for other people's experiences before committing time and money you can't spare.
On occasion it is good to experiment, but don't be heartbroken if it doesn't work- and post your experiences - good and bad - where people can find them. It' not all wasted- getting into a jam causes you how to get out of it, or avoid it in future.
[300 comments... wow!]
I've been using Linux sine 1994, and Ubuntu is a travesty. Never once has an install worked or upgraded correctly the first time using it on any of my equipment. Mice not working, incorrect video drivers, dual booting incorrect, never ending file sharing problems. For a new user, those problems are SHOW STOPPERS. Ubuntu is probably responsible for turning away vast amounts of new desktop users from Linux. Bug reports on Ubuntu are ABSOLUTELY useless. Nothing ever gets fixed. I've had bug reports in for years and nothing gets done. I just stopped making bug reports, it WASTES my valuable time.
I did put in many hours of time helping other newbies on Linux sites, but that got old pretty quick when Ubuntu never fixed anything. I've got pretty common equipment and tried once again with the latest version, 12.10, and once again it failed. video on my Nvidia 480 GTX's was all messed up. Install was not clear on where the GRUB install goes (although I know where it should go), my R.A.T.7 gaming mouse did not work on install, even though there are bug reports all over the net on what to fix in the install. I quickly installed Linux MInt 14 and all was well again. Ubuntu fails on so many levels. They are actually hurting the spread of Linux rather than helping. Sure I'll get flamed from the Ubuntu robots, but I'm not saying Linux is bad, it's great for what it is, but Ubuntu should just shut down and let people that actually give a hoot about fixing problems take over. (I switched to anonymous for this post, since the accusations will fly about this and that, and not relevant to the real problem, Ubuntu itself.)
Yes, this may impact us “normal” users. I’ve multi-booted my main i7 PC – blah blah; when I turn off and go to bed, then I boot again in the morning – the bios is set to force subsequent boot-ups to Win 8 only. I have to restart Win 8 to get my BIOS boot up menu again; since I use that to multi-boot each OS drive. This is a presumption from Microsoft – they’re insisting we boot Win8! I DON’T LIKE IT MS!!
Is this Samsung firmware brick problem restricted to Ubuntu 12.10 triggering it.or will other distro's do it?
It seams especially ironic as Ubuntu 12.10 (& Fedora) are supposed to come with a MS generated EUFI key, (with great lot of criticism from the Free Software Foundation & Linux community !)
I agree with AC, stop using Ubuntu, its getting too bloated to run on anything but the latest machines, bugs remain unfixed. I too have moved to Mint.
http://www.h-online.com/open/news/item/Protection-against-Samsung-UEFI-bug-merged-into-Linux-kernel-1795332.html
[Quote] "According to his analysis, entries for calling the UEFI setup get overwritten because one of the UEFI firmware's functions returns a value which is not in accordance with the UEFI specification."
This is clearly an issue with the firmware and not the kernel driver hence the reason why linux doesn't crash every other UEFI laptop. In IT we develop standards for reasons like this and when manufacturers don't follow those standards, things go wrong. Those of you who use this as an opportunity to bash linux clearly aren't to tech savvy and you look like idiots to those of us who actually understand how a computer is supposed to work.
"This is clearly an issue with the firmware and not the kernel"
If the firmware returns a wild value to a parameter search shouldn't the kernel respond defensively rather than blat over the boot entries?
Let's say malloc() returns NULL and the app explodes when it tries to use that as a pointer. Whose fault, the app or the heap manager? Think about it before you call people idiots: nothing in this story is black and white.
C.
Well the app didn't explode here, the firmware did. The app works fine on other firmware. Who's fault is it? The fault of the person who didn't follow standards that were set for UEFI. When you don't follow guidelines for a commonly used technology, unexpected things happen.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
