Re: [PATCH] x86/nmi: Enable nested do_nmi() handling for 64-bit kernels

[Qixuan Wu <qixuan.wu@linux.alibaba.com>]

Hi Chenggang,

   Have a doubt about the old code.

On 2018/4/12 PM 12:58, chenggang.qin@linux.alibaba.com Wrote:
> From: Andy Lutomirski <luto@kernel.org>
> 
> commit: 9d05041679904b12c12421cbcf9cb5f4860a8d7b upstream
> 
> 32-bit kernels handle nested NMIs in C.  Enable the exact same
> handling on 64-bit kernels as well.  This isn't currently
> necessary, but it will become necessary once the asm code starts
> allowing limited nesting.
> 
> ......
> -static inline void nmi_nesting_preprocess(struct pt_regs *regs)
> +dotraplinkage notrace void
> +do_nmi(struct pt_regs *regs, long error_code)
>   {
> +	if (this_cpu_read(nmi_state) != NMI_NOT_RUNNING) {
> +		this_cpu_write(nmi_state, NMI_LATCHED);
> +		return;
> +	}
> +	this_cpu_write(nmi_state, NMI_EXECUTING);
> +	this_cpu_write(nmi_cr2, read_cr2());
> +nmi_restart:
> +

Here if there are more than 2 NMIs nested, but the nmi_state is always
NMI_LATCHED.

>   
> -	/* On i386, may loop back to preprocess */
> -	nmi_nesting_postprocess();
> +	if (unlikely(this_cpu_read(nmi_cr2) != read_cr2()))
> +		write_cr2(this_cpu_read(nmi_cr2));
> +	if (this_cpu_dec_return(nmi_state))
> +		goto nmi_restart;
>   }

But here at most re-execute 2 NMIs, so some nmi lost ?
And cr2 is always the first NMI's cr2. CR2 is wrong for the later NMIs.

>   void stop_nmi(void)
> 

Thanks & Regards
Qixuan Wu.