Re: [PATCH v2 1/2] x86/MCE: Add function to allow filtering of MCA errors

[Borislav Petkov <bp@alien8.de>]

On Thu, Mar 21, 2019 at 08:25:17PM +0000, Ghannam, Yazen wrote:
> From: Yazen Ghannam <yazen.ghannam@amd.com>
> 
> Some systems may report spurious MCA errors. In general, spurious MCA
> errors may be disabled by clearing a particular bit in MCA_CTL. However,
> clearing a bit in MCA_CTL may not be recommended for some errors, so the
> only option is to ignore them.
> 
> An MCA error is printed and handled after it has been added to the MCE
> event pool. So an MCA error can be ignored by not adding it to the pool.
> 
> Create a function pointer to filter MCA errors and use this when adding
> an error to the MCE event pool.
> 
> Install a default function that does not filter any errors.
> 
> Cc: <stable@vger.kernel.org> # 4.14.x
> Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
> ---
> Link:
> https://lkml.kernel.org/r/20190307212552.8865-1-Yazen.Ghannam@amd.com
> 
> v1->v2:
> * This is a new patch replacing V1 Patch 1 which is no longer needed.
> 
>  arch/x86/include/asm/mce.h        | 3 +++
>  arch/x86/kernel/cpu/mce/core.c    | 6 ++++++
>  arch/x86/kernel/cpu/mce/genpool.c | 3 +++
>  3 files changed, 12 insertions(+)
> 
> diff --git a/arch/x86/include/asm/mce.h b/arch/x86/include/asm/mce.h
> index 22d05e3835f0..0b0b797a959c 100644
> --- a/arch/x86/include/asm/mce.h
> +++ b/arch/x86/include/asm/mce.h
> @@ -253,6 +253,9 @@ extern void mce_disable_bank(int bank);
>  extern void (*machine_check_vector)(struct pt_regs *, long error_code);
>  void do_machine_check(struct pt_regs *, long);
>  
> +/* Filter MCEs from the decoder chain. */

That should be something like:

/* Decides whether to add MCE records to the decoder chain or filter them out. */

> +extern bool (*filter_mce)(struct mce *m);
> +
>  /*
>   * Threshold handler
>   */
> diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c
> index b7fb541a4873..effb40581f08 100644
> --- a/arch/x86/kernel/cpu/mce/core.c
> +++ b/arch/x86/kernel/cpu/mce/core.c
> @@ -1771,6 +1771,12 @@ static void __mcheck_cpu_init_timer(void)
>  	mce_start_timer(t);
>  }
>  
> +/*
> + * Don't filter MCEs by default. Install a system-specific function, if needed.
> + */

That comment is kinda obvious.

> +static bool default_filter_mce(struct mce *m) { return false; }
> +bool (*filter_mce)(struct mce *) = default_filter_mce;
> +
>  /* Handle unconfigured int18 (should never happen) */
>  static void unexpected_machine_check(struct pt_regs *regs, long error_code)
>  {
> diff --git a/arch/x86/kernel/cpu/mce/genpool.c b/arch/x86/kernel/cpu/mce/genpool.c
> index 3395549c51d3..64d1d5a00f39 100644
> --- a/arch/x86/kernel/cpu/mce/genpool.c
> +++ b/arch/x86/kernel/cpu/mce/genpool.c
> @@ -99,6 +99,9 @@ int mce_gen_pool_add(struct mce *mce)
>  {
>  	struct mce_evt_llist *node;
>  
> +	if (filter_mce(mce))
> +		return -EINVAL;
> +
>  	if (!mce_evt_pool)
>  		return -EINVAL;
>  
> -- 
> 2.17.1
> 

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.