Re: [PATCH v3 05/10] x86/fpu/xstate: Define new functions for clearing fpregs and xstates

[Borislav Petkov]

On Sat, Mar 28, 2020 at 09:43:02AM -0700, Yu-cheng Yu wrote:
> @@ -318,18 +313,40 @@ static inline void copy_init_fpstate_to_fpregs(void)
>   * Called by sys_execve(), by the signal handler code and by various
>   * error paths.
>   */
> -void fpu__clear(struct fpu *fpu)
> +static void fpu__clear(struct fpu *fpu, int clear_user_only)

I said:

"fpu__clear(struct fpu *fpu, bool user_only)"
                             ^^^^^^^^^^^^^^

you made it

                     ..., int clear_user_only)

Why?

If you agree with the review comment, then please do it as suggested. If
you don't, then say why you don't.

Why would you do something in-between?

>  {
> -     WARN_ON_FPU(fpu != &current->thread.fpu); /* Almost certainly an 
> anomaly */

Why are you moving this into the callers when *both* do it?

> +     if (static_cpu_has(X86_FEATURE_FPU)) {

Flip this logic:

        if (!static_cpu_has(X86_FEATURE_FPU)) {
                fpu__drop(fpu);
                fpu__initialize(fpu);
                return;
        }

        fpregs_lock();
        ...

to save an indentation level and make the important case more readable
and locking more prominent.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette