On Fri, 2021-03-05 at 09:59 +0000, Anton Ivanov wrote: > > This is proving very "interesting" to try to chase down, because the > "picking the wrong library" does not happen every time. > > F.E. yesterday my 5.10 builds were picking glibc memcpy and friends. > Today with the same config and everything else the same it is picking > built-ins.
Ouch. > I need to finds some better way to reproduce this. Maybe something like the original report? That caused sem_init() to be called, so we know libc will/may call something there. You and me probably don't have the nss setup to cause sem_init() to get called, but maybe simply putting void init_nss_interface(void) { panic("how did we get here"); } somewhere in the kernel image might already reproduce it? johannes