On Fri, Oct 30, 2026, at 08:59, Simo Sorce via devel wrote: > On Fri, 2026-10-30 at 07:32 -0400, Stephen J Smoogen wrote: >> >> >> On Fri, Oct 30, 2026, at 07:19, Florian Weimer via devel wrote: >> > * Simon de Vlieger: >> > >> > > On Wed, Oct 28, 2026, at 6:47 PM, Florian Weimer via devel wrote: >> > > > The upstream Rust security team has not documented that Rust is >> > > > safe to >> > > > use in SUID binaries, and the standard library (and therefore >> > > > most >> > > > crates) do not take this special operating environment into >> > > > consideration. >> > > >> > > Aside from the dependencies; do languages for current SUID >> > > binaries document >> > > that they are safe to use in SUID contexts? Seems like a funky >> > > hurdle to >> > > impose. :) >> > >> > It's documented for the C library. Not in glibc manual, but for >> > example >> > in Garfinkel and Spafford, Partical UNIX and Internet Security. >> > The >> > second edition has an entire chapter, Writing Secure SUID and >> > Network >> > Programs. >> > >> >> That edition of the book is also older than many people writing Rust >> programs. The third and last edition is from 2003. I don't know what >> is in 2nd edition but third edition tips are at >> https://archive.org/details/practical_unix_and_internet_security-0596003234/page/516/mode/2up >> >> >> Frankly I put this above "It was on display in the bottom of a locked >> filing cabinet stuck in a disused lavatory with a sign on the door >> saying 'Beware of the Leopard.'" but not a lot more. Unless there is >> something similar in the glibc manual, or similar areas.. how is >> anyone in general going to know how to properly deal with suid/sgid >> or similar items? > > It is not like putting this search in duckduckgo: "is there some > documentation no how to write safe suid programs in unix ?" came up > with zero results ... >
And how is anyone to know which ones are 'trusted' and 'correct' and how many of them are 'this worked for me in 1993' or 'AI just barfed up this'? I have written some really bad suid code which was based on what people said was "good practice" in the 1980s and early 1990s. Which is what I thought Florian was referring to in the "The upstream Rust security team has not documented that Rust is safe to use in SUID binaries" and then pointing to a book from the 1990's as the reference that derailed me into a side rant which I should have avoided.. If the Rust team needs to document that either Rust is safe or how to write safe suid code in Rust.. then I would want the same is needed from the gcc or glibc security teams. Or if Fedora wants to put in a bar for proper suid code then it needs to have handy references to work with also to help people know what is expected currently. -- Stephen J Smoogen. Let us be kind to one another, for most of us are fighting a hard battle. -- Ian MacClaren -- _______________________________________________ devel mailing list -- [email protected] To unsubscribe send an email to [email protected] Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/[email protected] Do not reply to spam, report it: https://forge.fedoraproject.org/infra/tickets/issues/new
