Ubuntu 17.10's Laptop Issue Appears To Be Under Control, Fixable

[phoronix]

Phoronix: Ubuntu 17.10's Laptop Issue Appears To Be Under Control, Fixable

A week ago Ubuntu 17.10's ISO was pulled due to a show-stopping laptop bug whereby some UEFI-enabled laptops from multiple vendors were running into "BIOS corruption" where BIOS settings could no longer be changed, USB booting becoming non-functional, and similar UEFI-related issues. Fortunately, a fixed kernel is now available and some affected users are reporting a successful workaround for making their laptops full-functioning once again...

http://www.phoronix.com/scan.php?page=news_item&px=Ubuntu-17.10-Laptop-Fix

[Spooktra]

I am a devout Ubuntu user, consider it the best distro currently available with only Solus, Mageia, GhostBSD and OpenIndiana coming close but Canonical is a company that generated 126 million bucks last year, there is no excuse for the lack of Q&A testing that resulted in this mess, if Apple or Microsoft had this happen to them we would be ready to nail them to the nearest wall; if Linux is ever to become a credible player on the desktop we can't have these things happening in 2017.

[kaprikawn]

It hasn't been confirmed who's fault it is. And it is only limited to a particular bios. Do you really expect Canonical to test every configuration on the planet?

I'm no Canonical apologist, in fact I have a particular distaste for them. But that's no reason to be a reactionary and immediately blame them for this.

If you can show me something that says Canonical have been negligent then I'll shut up.

[Leopard]

I am a devout Ubuntu user, consider it the best distro currently available with only Solus, Mageia, GhostBSD and OpenIndiana coming close but Canonical is a company that generated 126 million bucks last year, there is no excuse for the lack of Q&A testing that resulted in this mess, if Apple or Microsoft had this happen to them we would be ready to nail them to the nearest wall; if Linux is ever to become a credible player on the desktop we can't have these things happening in 2017.

I don't know if you are really a devout user but at least i can say that you have no clue about what you're writing.

First; Apple won't get such troubles because MacOS is only limited to their machines. So they are just giving guarentee about small number of devices which they are fully controlled by Apple.

Second ; Microsoft won't get such troubles because device manufacturers ( like Lenovo or Asus ,Msi etc ) are testing their devices with Windows and they make it compatible with Windows . Microsoft is not testing their devices on their OS.

Third ; Manufacturers are only giving a shit about your problems when they appear on Windows. If you said that " I'm using Linux and i have bla bla bla problem on your firmware , Acpi wise " they said " Use Windows or gtfo "

So please don't post your opinions without thinking on them a little bit.

[Spooktra]

It hasn't been confirmed who's fault it is. And it is only limited to a particular bios. Do you really expect Canonical to test every configuration on the planet?

I'm no Canonical apologist, in fact I have a particular distaste for them. But that's no reason to be a reactionary and immediately blame them for this.

If you can show me something that says Canonical have been negligent then I'll shut up.

It's very easy to confirm it's Canonical's fault with some basic reasoning:

There are a multitude of base distributions and dozens of variants: Red Hat, Scientific Linux, Cent OS, all the Arch off shoots, Suse Enterprise, Suse Leap, Suse Tumbleweed, Mint, all the official Ubuntu flavors, Debian, Solus, the Mandriva off shoots, the list goes on and if you add in all the versions of each, such as Ubuntu 16.04 LTS, 16.10, 17.04, 17.10 on and on and on, and I'm not going to mention the BSD based distros.

Yet the only distro that caused issues with any laptop was Ubuntu's 17.10 offering and then it only showed up with the latest iso spins, when 17.10 betas were first released they didn't cause this issue, and I don't but the explanation involving the kernel because there are other distros that use the same kernel version and they don't have this problem.

I blame Canonical and they should be held accountable for this.

[vsteel]

It's very easy to confirm it's Canonical's fault with some basic reasoning:

There are a multitude of base distributions and dozens of variants: Red Hat, Scientific Linux, Cent OS, all the Arch off shoots, Suse Enterprise, Suse Leap, Suse Tumbleweed, Mint, all the official Ubuntu flavors, Debian, Solus, the Mandriva off shoots, the list goes on and if you add in all the versions of each, such as Ubuntu 16.04 LTS, 16.10, 17.04, 17.10 on and on and on, and I'm not going to mention the BSD based distros.

Yet the only distro that caused issues with any laptop was Ubuntu's 17.10 offering and then it only showed up with the latest iso spins, when 17.10 betas were first released they didn't cause this issue, and I don't but the explanation involving the kernel because there are other distros that use the same kernel version and they don't have this problem.

I blame Canonical and they should be held accountable for this.

But you don't know if they did test and Lenovo changed the firmware at the last moment or there was an undocumented update. You don't even know the reason why it happened.

You don't have enough information to say it was Canonical's fault or if they were the victim in this. All of the facts are needed before the finger pointing begins.

[Spooktra]

I don't know if you are really a devout user but at least i can say that you have no clue about what you're writing.

First; Apple won't get such troubles because MacOS is only limited to their machines. So they are just giving guarentee about small number of devices which they are fully controlled by Apple.

Second ; Microsoft won't get such troubles because device manufacturers ( like Lenovo or Asus ,Msi etc ) are testing their devices with Windows and they make it compatible with Windows . Microsoft is not testing their devices on their OS.

Third ; Manufacturers are only giving a shit about your problems when they appear on Windows. If you said that " I'm using Linux and i have bla bla bla problem on your firmware , Acpi wise " they said " Use Windows or gtfo "

So please don't post your opinions without thinking on them a little bit.

I thought about this much more than you have.

Why is it this problem only occurs with the latest Ubuntu 17.10 spins? The issue supposedly arises from Intel's SPI (Serial Peripheral Device) driver, code that supplied by Intel themselves and has been thoroughly tested and used by numerous laptops, so why doesn't it happen with that piece of shit offering known as Fedora 27?

Why doesn't it happen with Arch, Tumbleweed, Leap, Red Hat, Scientific Linux, Cent OS, Manjaro, Mageia, PCLinuxOS, GhostBSD, DragonFlyBSD, OpenIndiana, BSD proper, Debian, Mint, or any of the other Linux variants available?

Why doesn't it happen with Windows or OSX, which also used an Intel supplied SPI driver?

The issue can't be laid at the feet of Lenovo for not testing their product with Linux, the driver is coded and contributed by Intel, a long time code contributor to the Linux kernel and is thoroughly tested by a corporation worth billions. This driver is used extensively by numerous open source and closed source offerings yet the only variant that ran into problems was Ubuntu's latest 17.10 spins.

Clearly Canonical changed something in the way they initialize the driver during the boot process and they are fucking up people's computers. Of course they will never admit that it was their fault because they don't want to get hit with lawsuits from effected users but this is the type of thing that influences my decisions about what distro to use and it's likely that when/if the time comes for me to reinstall the OS due to an drive upgrade I will probably be reaching for a different distro.

[Spooktra]

But you don't know if they did test and Lenovo changed the firmware at the last moment or there was an undocumented update. You don't even know the reason why it happened.

You don't have enough information to say it was Canonical's fault or if they were the victim in this. All of the facts are needed before the finger pointing begins.

Here's what Canonical says right on their website:

https://wiki.ubuntu.com/ArtfulAardvark/ReleaseNotes?_ga=2.209292050.1801693164.1514395528-1013114896.1513906023#Known_issues

A bug in the Linux 4.13 kernel shipped in Ubuntu 17.10 can leave users unable to update any of their BIOS settings, including their system’s boot order, after booting this version of Ubuntu.

According to Canonical themselves the bug lies with the 4.13 kernel that they ship in Ubuntu 17.10. We all know that Canonical doesn't use the vanilla kernel, so they are obviously doing something that no one else is doing.

[kaprikawn]

It's very easy to confirm it's Canonical's fault with some basic reasoning:

There are a multitude of base distributions and dozens of variants: Red Hat, Scientific Linux, Cent OS, all the Arch off shoots, Suse Enterprise, Suse Leap, Suse Tumbleweed, Mint, all the official Ubuntu flavors, Debian, Solus, the Mandriva off shoots, the list goes on and if you add in all the versions of each, such as Ubuntu 16.04 LTS, 16.10, 17.04, 17.10 on and on and on, and I'm not going to mention the BSD based distros.

Yet the only distro that caused issues with any laptop was Ubuntu's 17.10 offering and then it only showed up with the latest iso spins, when 17.10 betas were first released they didn't cause this issue, and I don't but the explanation involving the kernel because there are other distros that use the same kernel version and they don't have this problem.

I blame Canonical and they should be held accountable for this.

https://www.youtube.com/watch?v=yUQmHFamb1Q - About 7:00

Someone who undoubtedly knows more about this than you saying it's a problem with an Intel module and it's NOT specific to Ubuntu, there's been reports of it affecting Antergos users too.

[GreatLord]

It's very easy to confirm it's Canonical's fault with some basic reasoning

it is not. I did read the tech issue.
To transate what the reason was some manufactor using a flash rom that was not design standard metod for flashig, and some area need be in read only or write only. That mean they did not follow best practics how things are done by all other vendor and standard that exists.

This count as special case and I did read the dev of the intel spidrv did start look into how they should add a fix and workaround for few bad flash chip that have been found now. I have not read how far they have come seen frist report.