mirror of https://github.com/torvalds/linux.git
c2debdb854
The uretprobe syscall is implemented as a performance enhancement on x86_64 by having the kernel inject a call to it on function exit; User programs cannot call this system call explicitly. As such, this syscall is considered a kernel implementation detail and should not be filtered by seccomp. Enhance the seccomp bpf test suite to check that uretprobes can be attached to processes without the killing the process regardless of seccomp policy. Signed-off-by: Eyal Birger <eyal.birger@gmail.com> Link: https://lore.kernel.org/r/20250202162921.335813-3-eyal.birger@gmail.com [kees: Skip archs without __NR_uretprobe] Signed-off-by: Kees Cook <kees@kernel.org> |
||
|---|---|---|
| .. | ||
| .gitignore | ||
| Makefile | ||
| config | ||
| seccomp_benchmark.c | ||
| seccomp_bpf.c | ||
| settings | ||