Nasty Linux systemd security bug reveale

They link indirectly to this bug: https://bugzilla.redhat.com/show_bug.cgi?id=1970887

Which has status "modified" (I guess that means "open"), and is blocked by a bunch of other open bugs that aren't viewable. So I guess the patches Qualys recommends would be bespoke, because it isn't fixed upstream.

"It works by enabling attackers to misuse the alloca() function in a way that would result in memory corruption. "

I thought alloca() had been banned? Well, the man page says "[..] Otherwise, its use is discouraged." (also, I suppose, due to its "interesting" consequences when used in functions which themselves are being in-lined)

Lately the systemd maintainers have been aggressive on HN, downvoting stories like this and critical comments. I wish they would instead spend the effort on addressing the community's most pressing concerns about the feature scope and creep of systemd's ever increasing size.

Mmm: https://www.qualys.com/2021/07/20/cve-2021-33909/sequoia-loc...

That’s why I am using OpenRC and S6 for my initd on all my Linux boxes.

Not impacted.