Announcement

Collapse
No announcement yet.

systemd 228 Had A Local Root Exploit

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

  • #61
    Originally posted by signals View Post

    Actually, from Windows I am using a UNC path to access a Samba share on the server, and it doesn't need to be mounted at all.

    In defense of systemd, whatever problem I had has been fixed and I have several Linux systems that now mount the NFS server at boot without issue. This was early on in the development of systemd and I'm sure it was probably just teething problems.

    I guess I should have expected all the hate when I posted... But, with all of the "modernization" of Linux distros now, I'm just not feeling the joy that I did with Linux back when it was just a Unix clone. I'm not saying Windows is better than Linux... All I'm saying is that for the first time in my life, Linux wasn't sufficiently attractive to make me reformat a new machine and Windows is doing just fine so far. Which is kind of sad actually...
    Ah, so like how you would open it as say "smb://server" in i.e Filer in Linux. Myself I don't see the problems that you see with the modernization of Linux since none of it makes it any more less of a Unix system. But each to their own and if you prefer Windows that is none of my business.

    Comment


    • #62
      Originally posted by starshipeleven View Post
      You already forgot the articles I linked that showed how Windows updates fucked up NVIDIA drivers and also Intel CPUs to answer your misguided attempt to throw shit on AMD by showing a forum thread where it was again MS's updates that fucked up AMD drivers?
      Yeah and those articles were exactly how relevant? As far as I recall, arguments ended after I pointed out that Nvidia had at the time 70% market share and so had proportionally considerably LESS driver issues than AMD had.

      Intel CPU issues were completely irrelevant because they were happening during overclocking or something. Which means, putting hardware out design specs.

      Comment


      • #63
        Originally posted by pal666 View Post
        there are no constant api/abi changes on linux. if you see some constant changes, then they are not api/abi or they are not linux
        Graph shows added/removed symbols, detailed changes in parameters and data types in the vmlinux binary and basic modules. It was handy and I grabbed it shamelessly from the net.

        Comment


        • #64
          Originally posted by F.Ultra View Post

          Ah, so like how you would open it as say "smb://server" in i.e Filer in Linux.
          More or less, I guess... Although it works everywhere, including in Powershell and in any config files on Windows. I don't have to be in Filer (or whatever) for it to work.

          Originally posted by F.Ultra View Post
          Myself I don't see the problems that you see with the modernization of Linux since none of it makes it any more less of a Unix system. But each to their own and if you prefer Windows that is none of my business.
          I'm not sure how I ended up defending my switch to Windows on the desktop in this thread. I guess I never should have mentioned it. It's a very uncomfortable position to be in for me because I was an absolute Microsoft hater in my younger years.

          But, I wouldn't say I "prefer" Windows. It's just that after 20 years of sparring with Linux, I decided it was time for a new sparring partner. Maybe I'll win more bouts, maybe I won't. Besides, Linux still outnumbers Windows 6 to 1 in my home environment. It's just not on my workstation anymore.

          Comment


          • #65
            Originally posted by ldo17 View Post
            Hands up all those who have actually used a system with this vulnerability ...

            .
            .
            .
            (crickets)
            .
            .
            .
            Now I don't use it but apparently SUSE used the vulnerable version: https://lwn.net/Articles/712488/

            Comment


            • #66
              No matter what it was lucky that someone found and identified the exploit for what it was. In the future maybe we and the systemd people should look closer at bugs that they assume aren't worthy of informing people and maintainers better about. A safe system should be more important than new features.

              Comment


              • #67
                Originally posted by sjukfan View Post
                In the future maybe we and the systemd people should look closer at bugs that they assume aren't worthy of informing people and maintainers better about.
                Indeed, though much easy to say than to do. Any developer on any project fixes a lot of bugs, most of them trivial. And security issues in particular are often a combination of multiple distinct bugs... each of which is only a small weakness on its own (or a serious one but with no way to exploit it), but when you put them together, they become a much more serious problem.

                So how do you know, when you're fixing one seemingly-trivial issue - say, an unchecked null-reference - that it's actually a small part of something big and scary?

                Comment


                • #68
                  Ok, let's see.

                  * There was a bug in an ancient version of systemd, which was fixed and correctly classified as a local DoS. Now it turns out that the combination of the systemd bug, and another bug known and unfixed since 2015, can get you a local root exploit. The internet explodes, systemd gets all the blame, lots of armchair developers tell the world how they would have designed systemd to avoid this bug.
                  Meanwhile, the setguid root exploit is still unfixed. Nobody cares.

                  * There is a bug in current versions of screen, which has been present for over a year. Anyone looking at the bug should immediately notice the potential for destroying the system and/or gaining root access. There is no phoronix article, there is barely any discussion, people just update and move on with their lives.

                  *sighs*

                  Comment


                  • #69
                    Originally posted by aht0 View Post
                    Yeah and those articles were exactly how relevant?
                    Because you were claiming that NVIDIA had less driver issues, which was plain bullshit.

                    As far as I recall, arguments ended after I pointed out that Nvidia had at the time 70% market share and so had proportionally considerably LESS driver issues than AMD had.
                    Sorry for that, must have missed that or I was too busy elsewhere. Here the answer: Sure, because there is only fictional AMD-only driver issues that never also affected NVIDIA that affected sales, while marketing, more powerful GPUs, more goddamn fanboys like you rooting for NVIDIA didn't also play a part.

                    Intel CPU issues were completely irrelevant because they were happening during overclocking or something. Which means, putting hardware out design specs.
                    No they were not issues due to overclocking, as it was a specific Windows update that fucked up a situation that was 100% fine before it, worldwide, and again was another windows update that fixed this specific issue for everyone at 100%, worldwide.

                    Comment


                    • #70
                      Originally posted by F.Ultra View Post
                      Now I don't use it but apparently SUSE used the vulnerable version: https://lwn.net/Articles/712488/
                      I can confirm this, I got the update yesterday on OpenSUSE.

                      Comment

                      Working...
                      X