Three systemd vulnerabilities have been found

Three systemd vulnerabilities have been found.
Two are related to memory corruption vulnerabilities and the other one its "out-of-bounds error".

These vulnerabilities exists due to bad design or C can take responsability?

Source: bleepingcomputer.com/news/security/linux-systemd-affected-by-memory-corruption-vulnerabilities-no-patches-yet/

theregister.co.uk/2019/01/10/systemd_bugs_qualys/

Attached: systemd.jpg (640x372, 35K)

Other urls found in this thread:

twitter.com/isislovecruft/status/1084341028864192514
security-database.com/detail.php?alert=CVE-2017-16638
security-database.com/detail.php?alert=CVE-2017-18188
twitter.com/AnonBabble

"Three new security holes recently were discovered in Systemd by the Qualys security company. From ZDNet: "With any of these a local user can gain root privileges. Worse still, Qualys reports that 'To the best of our knowledge, all systemd-based Linux distributions are vulnerable.' Actually, that's not quite true, even Qualys admits. 'SUSE Linux Enterprise 15, openSUSE Leap 15.0, and Fedora 28 and 29 are not exploitable because their user space is compiled with GCC's -fstack-clash-protection.'" Red Hat has already released patches for 16864 and 16865. "

init.d when?

Fucking hell I have to get around to installing Gentoo before the poettering nigger creates a bug that wipes your disks and all firmware

Both. If he wasn't using C, these errors wouldn't be there. But these aren't complex software errors due to difficult architectures or hard engineering problems, they're problems caused by lack of even basic programming ability. Thus the issue is with lennart.

Wonder what the systemd shills are going to say about that. They kept claiming that systemd is not a security nightmare and it's opensource so it's clearly perfect.

People knows using alloca like SysyemD did was a bad thing since the 80s.
C is unsecure, but they were stupid ignoring best pratices.

>normal software has bugs
>...
>systemd has bugs
>OMG STOP THE PRESSES NSA NSA FUCK THIS

The only reason people get upset is because they're fucking retards who possess no actual technical knowledge. All they can do is parrot other people's opinions and pretend they know shit by getting outraged over "technical" topics. This isn't even systemd related. I see you retards do this all the time.
Consider getting euthanised and spare us the bandwidth of your retardation.

Ignorant fuck here, is there a way developers can fix those issues?

this. They probably never heard what a CVE is or what is a security patch since their shitty distro is unmaintained

People got upset because RedHat wanted systemd everywhere and it did it everywhere, despite their autistic screeching.

Let's not forget the DHCPv6 bug a while back

Attached: systemdhcp.jpg (3456x2304, 2.54M)

Butthurt shill is butthurt

come home, white man

Attached: devuan.png (1000x200, 13K)

is it a big deal for an average user since you need local access to do anything?

This.
Or MX Linux
Or /ourdistro/ guixsd
Or Gentoo
Or Void
Or Slackware
Or Artix

No, because it already got patched

Yeah, probably redteam has already fixed it. It's nothing serious and as said, when systemd has bugs, people go nuts just because of the polemic involving unix philosophy and how systemd doesn't follow it. So when a systemd vulnerability its discovered they announce it like "Yeah, I told you guys. Can we go back to SysV now?"

daily reminder that anyone who defends systemd is retarded. seriously use devuan, slackware, gentoo, funtoo, artix, or void

hey remember when a vulnerability in an OS was a reason to worry?
not it just doesn't. All OSs are vulnerable. And if the OS doesn't have holes, the hardware will. There's no escape from the horror at this point. I just live with it.

Attached: simplypain.png (468x431, 185K)

>Systemd sucks, here are some meme tier distros (except gentoo) that you should use instead.

Yeah but Rust evangelists be like:
"If you have used Rust, you would'nt have this vulnerabilities"
"Can we rewrite the linux kernel in Rust?"
"Can we rewrite systemd in Rust?"
"Can we rewrite LLVM in Rust? Why they used C++? It's an unsafe language, didn't you know that?"

it wouldn't be much of a problem to not use systemd, but the ones that don't use systemd are a fucking joke that don't address CVEs on time. So effectively you are more vulnerable on a distro without systemd ironically
this, it's retarded. I'm gonna stay CentOS AND C. Bad programmers want a scriptie solution for vulnerabilities like another language or distro, but that's just evading the problem so they don't do anything and Linux never improves

those distros espically slackware are really fucking good. the only one I actually don't like is void but some people do like that one so may as well include it

Butthurt systemd shill is butthurt

Why you don't like void user?

Not him but the package manager is a joke. Try to remove a package and it will remove half the system and brick itself without warning. It basically doesn't do correct dependency resolution for this. It also has very shitty package selection. Moreover, I find runit to be really bad: if anything bad happens to a service, your only way out is to boot from a live media and disable the service from there, whereas on other systems the service would just stay dead for you to deal with.

Openrc isn't bug free either.
ebin
twitter.com/isislovecruft/status/1084341028864192514

I want to use Slackware but there is no SELinux/Grsecurity/Apparmor support... So I'm in CentOS. I like packaging my own stuff.
this is why I don't use debian either, red hat and fedora dependencies always work the same way when installing and when removing. And if you don't want to use dependencies you use rpm

>Openrc isn't bug free either.
[citation needed]
Even then the real problem is that systemd is basically cancer, it encompasses everything whereas you can swap out openrc for anything you want without borking the rest of the system.

>whole ecosystem based around editing files
>nothing personal kid, we make all you 9000 commands useless by implementing obscure binary formats
Imagine still defending systemd in 2019

I just never liked using it. it was very buggy and I don't like xbps and I don't like how you manage services. it feels like I'm trying to use BSD as a desktop 10 years ago but if you like it that's fine

It isn't, but tries to actually minimize the attack surface, unlike systemd. Also, you can compare amount of detected vulnerabilities in systemd an OpenRC. Hint: OpenRC has significantly lower amount.

>Hint: OpenRC has significantly lower amount.
source?

>or C can take responsability?
You can't blame a language for bad programmers, that's absurd.

take your pills grandpa.
Slackware isn't gonna go anywhere
>nobody tracks openrc because nobody uses it
so it has no vulnerabilities
wew
it's obvious it has less since it is untracked and unmaintained, plus it's smaller

Yep. While rewriting Linux in Rust would be even more insecure. Pretty new language, in which you have to disable practically all garbage collection and security features in order to write self-hosting code.

My ubuntu 18.04 LTS already have it fixed

The entire notion of SystemD is bad design, and people have told you shit like this will do nothing but continue to occur on an ever more frequent basis.

But no, it's not SystemD at all, it's C.

You can definitely blame a language for making several classes of bugs harder to notice or test for, though. For example it's definitely python's fault that you can lose hours of work because there's a space too many somewhere or you made a typo (color vs colour), especially in the presence of other people's code. You can certainly also blame ruby for being too slow for your needs and java's jvm for creating uncontrollable GC pauses under your workloads.
There are two solution paths common in each of these problems:
- Work around these problems (use extensive static/simulation tooling and ridiculous test coverage in C, use an alloc-free style in java, write C modules in ruby, run everything in ipdb in python)
- Use a tool that actually fit your problem (i.e. change language to something that doesn't suck balls)
The theoretical solution "become a perfect machine that makes no mistakes and also be able to review everyone's code perfectly while never being tired" is not valid. If it was, you should be able to point to at least one competent C programmer. But there is no such thing (whereby competent is the ridiculous standard: has delivered non-trivial software with no low-level bugs).

OpenRC:
security-database.com/detail.php?alert=CVE-2017-16638
security-database.com/detail.php?alert=CVE-2017-18188
No more CVEs found, while the second one requires another piece of software to be installed.

There's no GC in rust, inbred.

Sry, was thinking about Go which has similiar problems.

Go has significant problems all the way through but gotards gave up on pretending it was a systems programming language unlike rust faggots. The real problem with rust is that the borrow checker is too primitive to allow non-trivial programs to be extended without turning it off (i.e. using unsafe).
Note however that unsafe regions in rust are safer than raw C/C++ (barely, but that is still true). The real gain is that you can minimize unsafe{} to make sure everything around these regions IS safe of these bugs. That means if a bug is found, you know where to look, and bugs are very unlikely to be found anyway (for memory-related and such).

It's smaller, that's why it's better. Tries to do one thing properly instead of creating a clusterfuck which systemd is.
Also, it's used enough to gain enough interest among security researchers. However, the difference from systemd is that its full code can be actually audited by a single security researcher, while systemd is so unmaintainable that even a team of programmers can't fully comprehend it.

This

According to pottering Itsnt his own fault, it's because of c.
When pottering rewrite kernel in rust all will be solved

>For example it's definitely python's fault that you can lose hours of work because there's a space too many somewhere or you made a typo
No, that's retarded.
You bring unable to use a tool correctly is not the tools fault.
Maybe it is designed badly, but that doesn't excuse you.
This is also not a python specific issue, the alternative is hunting for brackets.

>You can certainly also blame ruby for being too slow for your needs
No, you can blame programers choosing a language too slow for their needs.

>and java's jvm
That's not an implementation issue, not a program with the language itself.

>The theoretical solution "become a perfect machine that makes no mistakes and also be able to review everyone's code perfectly while never being tired" is not valid.
Of course it isn't. But your choice brought you there in the first place.
Using inappropriate languages is the fault of the programmer.

Just because you're cnile doesn't mean we enjoy seeing you shit your pants in public, gramps.

???