> For the complete documentation index, see [llms.txt](https://lagnos.gitbook.io/linux-kernel-notes/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://lagnos.gitbook.io/linux-kernel-notes/misc/noinstr.md).

# noinstr

<https://docs.kernel.org/core-api/entry.html>

* syscall entry must not be instrumented because :
  * starts in assembly and interacts with love-level C and stack frames (registers should not get overwritten)
  * this function set up `lockdep`, `RCU/Context tracking` and tracing for the call using `syscall_eter_from_user_mode()`. It then invokes various entry functions like ptrace, seccomp, audit, syscall tracing, etc. (RCU is not yet setup at the beginning, hence must not be instrumented)

```c
noinstr void syscall(struct pt_regs *regs, int nr)
{
      arch_syscall_enter(regs);
      nr = syscall_enter_from_user_mode(regs, nr);

      instrumentation_begin();
      if (!invoke_syscall(regs, nr) && nr != -1)
              result_reg(regs) = __sys_ni_syscall(regs);
      instrumentation_end();

      syscall_exit_to_user_mode(regs);
}
```

> Do not nest syscalls. Nested systcalls will cause RCU and/or context tracking to print a warning.