Hello all, I have recently encountered a case where a VIA C3 Nehemaiah CPU returns "Illegal Instruction" when trying to run 'sudo' or 'visudo'.
After some poking around, I discovered that the FreeBSD folks have encountered this as well, and that it appears to be an issue with GCC where the --fcf-protection option results in the use of the ENDBR32 instruction, which is not supported on the VIA C3 Nehemaiah processor (despite being otherwise i686-compatible). Here are the relevant threads that I found: Similar issue in Gentoo: https://bugs.gentoo.org/show_bug.cgi?id=862201 <https://bugs.gentoo.org/show_bug.cgi?id=862201> Similar issue in FreeBSD, more recent, but different processor: https://forums.freebsd.org/threads/illegal-instruction-after-12-4-upgrade-i386.89353/ <https://forums.freebsd.org/threads/illegal-instruction-after-12-4-upgrade-i386.89353/> Relevant GCC commit: https://gcc.gnu.org/git/gitweb.cgi?p=gcc.git;h=77d372abec0fbf2cfe922e3140ee3410248f979e <https://gcc.gnu.org/git/gitweb.cgi?p=gcc.git;h=77d372abec0fbf2cfe922e3140ee3410248f979e> I did try rebuilding sudo from source, though it has been so long since I did that - I'm much more "sysadmin" than anything else these days - that I haven't had a chance to dive into the source and force it to build as "i486" or remove the --fcf-protection option. The VIA C3 Nehemaiah chip is an i686-class chip, though it's a VIA (a.k.a. Cyrix) so I'm not entirely surprised that there are differences in its instruction set from the Intel chips. I had a working bullseye installation and attempted to upgrade to bookworm when I ran into this issue. Any other information I can provide that would help track this down? I assume that the build environment used for creating the 32-bit packages just doesn't take this case into account. Please advise. Thanks in advance! --J
