On Thu, 11 Sep 2025 14:37:58 -0400 Steven Rostedt <[email protected]> wrote:
> > Probably it may worth moving cond_resched under within_module check, but > > before test_for_valid_rec call. > > Hmm, do modules even have weak functions? If they don't then that test is > useless. Doing a quick search, I see that they do :-/ drivers/tty/tty_ioctl.c:__weak int kernel_termios_to_user_termios(struct termios2 __user *u, Looks like sparc can overwrite it. Thus, I'm guessing a module could add a function and an arch may make it different. That means we still need this for now. Yeah, just send a v2 with the cond_resched() after the check for module. -- Steve
