Announcement

Collapse
No announcement yet.

Intel Laptop Users Should Avoid Linux 5.19.12 To Avoid Potentially Damaging The Display

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • macemoneta
    replied
    Originally posted by Luke View Post
    I found out the hard way that some Chromebooks once ChromeOS is removed and a real Linux distro installed have a nasty firmware bug/feature: if you accidently press the spacebar pre-boot, the firmware will re-lock the boot loader and you cannot boot.
    That is, of course, an "off label" use of a Chromebook; you have to disable the write protection and security to install a new BIOS and OS. However, that also means your device only supports the legacy BIOS, not the replacement UEFI, which doesn't have that issue. Many Chromebooks (50 new models) were added to UEFI support recently, so you may want to update to see if you now have the option of UEFI installation.

    Leave a comment:


  • macemoneta
    replied
    Originally posted by pWe00Iri3e7Z9lHOX2Qx View Post
    Well that's fun. Fedora and Tumbleweed are both on 5.19.12.
    Fedora has 5.19.13 in testing, or you can pull it from Koji.

    Leave a comment:


  • CTTY
    replied
    On my arch linux machine I use A/B-style squashfs images. So if there is a problem, I can go back to the last working state.

    Leave a comment:


  • CTTY
    replied
    Remember when these good old stable distros made systems unbootable?
    https://lists.ubuntu.com/archives/ub...st/005549.html

    I had luck, because my server (and machines I administrate) did not automatically install these update and reboot in that short period, but it would have been a disaster because I was miles away.
    Last edited by CTTY; 04 October 2022, 06:18 PM.

    Leave a comment:


  • Chugworth
    replied
    I wonder what are the specifics of the chips or screens that were affected. I was running that version for a couple of days on my ThinkPad X1 Carbon Gen 9 (Core i7-1165G7) and didn't notice anything odd.

    Leave a comment:


  • Luke
    replied
    There is a reason Debian Unstable is named "Sid." It's right there on their website, with all the Toy Story names Sid is the kid that breaks things. Also the very name "Unstable" is a warning to new users. It's for folks who know what they are doing-and if it didn't exist there'd be nobody to test all this stuff before it gets to stable.

    I have a known good backup snapshot of my OS on my main system drive, and multiple data backups.Never had a kernel bug eat a filesystem, but had multiple disk failures and a wrong-device dd eat a few filesystems. Copying back a BIG filesystem can be quite time consuming after such an incident but you don't lose data because you never have it all mounted at once and never have all of it on the same device.

    I always tell windows users to treat ransomware as though a power surge ate their system and data discs and someone put new ones in: in the end, the results are the same and SHIT HAPPENS. One person gets ransomware, another gets a bad HDD from the vendor, yet another gets kernel issues. Displays too get trashed from many things. You can be very careful to use only stable kernels, only to sit on the bag containing your laptop by mistake.

    I found out the hard way that some Chromebooks once ChromeOS is removed and a real Linux distro installed have a nasty firmware bug/feature: if you accidently press the spacebar pre-boot, the firmware will re-lock the boot loader and you cannot boot. If you don't have and cannot find(or cannot download due to bandwidth) a recovery image, the system can be consided "soft-bricked" from ONE accidental keypress. Even if you do, setup is a total do-over unless you keep the OS and everything else on external media and don't use the internal drive at all. That's because it will wipe (or is supposed to wipe) everything you wrote to the system drive. Those Chromebooks that act that way really should be reflashed with vanilla coreboot before use.

    Leave a comment:


  • stormcrow
    replied
    Originally posted by Anux View Post
    You're partly right, but there will always be people that will never have a problem. It's like HDDs many will run into a defect but I know also many people that never had a bad one.
    As a former QA person back in the day, generally speaking a "I don't have a problem" was less important to me than how many other people are reporting the same problem (but not just saying me too). The question is usually NOT "Who doesn't have this problem?" because it can be assumed the people that pushed the update out the door don't have the problem or they would have caught it. It's "Who has this problem and what's different (or similar) with their system (from mine)?" The first report of "I don't have the problem" might be useful from a testing point of view assuming the person reporting is qualified and really does have a similar setup to the believed problematic hardware - definitely not a given from random posts on the Internet even if they give a model number. After that people posting "me too" without any other information or slight information, which is what's going on here, is unhelpful. It's extremely risky to take people "I don't have a problem" at their word even if they supposedly have similar hardware (hardware model & revision numbers don't tell you everything).

    Those that are doing this when there's a valid - dangerous - problem report with something as important with the Linux kernel need to stop. You're as unhelpful as the people posting "me too" to problem reports.
    Last edited by stormcrow; 04 October 2022, 06:13 PM.

    Leave a comment:


  • Viki Ai
    replied
    OTOH, I run a testing branch on my home system specifically because I want to take part in testing and improve the stable branch for those (including me) using stable on more critical machines. So encountering bugs, while annoying, is kind of the point with a 'testing' branch.

    Leave a comment:


  • piorunz
    replied
    Originally posted by Weasel View Post
    Don't worry, it's always a "first time" for anyone when it fucks up their setup.

    And then you'll have some smartass who's gonna say "works for me, never had issues" and others believe him... until a "first time" happens to them. Then they complain, and another smartass says the same thing, ad infinitum.

    Rolling release distros will break your setup at some point. It's not a question of if, but when.
    At some point, almost all bleeding-edge wannabes (read: beta testers) will get tired of ridiculous errors no one else have and get back to stable distros with tail tucked between their legs.

    Leave a comment:


  • NotMine999
    replied
    Originally posted by kvuj View Post
    I thought it was my laptop dying or something.

    How the hell did Arch let this kernel get to stable without testing it on Intel hardware? It's the first time Arch broke my install with an update and it seems oddly obvious of a bug.
    The use of the words Arch and testing in the same sentence

    My own experiences with Arch lasted about 3 or 4 months. Then I realized it's release program was only a bit more stable than a druggie on crack.

    Those experiences were a few years ago so I guess Arch has not improved that much based on the quoted comments.

    Leave a comment:

Working...
X