Re: [PATCH 4/6] x86, mce: Add mechanism to safely save information in MCE handler

[Ingo Molnar <mingo@elte.hu>]

* Tony Luck <tony.luck@intel.com> wrote:

> Machine checks on Intel cpus interrupt execution on all cpus, regardless
> of interrupt masking.  We have a need to save some data about the cause
> of the machine check (physical address) in the machine check handler that
> can be retrieved later to attempt recovery in a more flexible execution
> state.
> 
> Signed-off-by: Tony Luck <tony.luck@intel.com>
> ---
>  arch/x86/kernel/cpu/mcheck/mce.c |   51 ++++++++++++++++++++++++++++++++++++++
>  1 files changed, 51 insertions(+), 0 deletions(-)

Just some cleanliness nits:

> diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
> index 43f22c8..9b83b7d 100644
> --- a/arch/x86/kernel/cpu/mcheck/mce.c
> +++ b/arch/x86/kernel/cpu/mcheck/mce.c
> @@ -887,6 +887,57 @@ static void mce_clear_state(unsigned long *toclear)
>  }
>  
>  /*
> + * Need to save faulting physical address associated with a process
> + * in the machine check handler some place where we can grab it back
> + * later in mce_notify_process()
> + */
> +#define	MAX_MCE_INFO	16
> +struct mce_info {

please separate non-bulk definitons by newlines.

> +	atomic_t		inuse;
> +	struct task_struct	*t;
> +	__u64			paddr;
> +} mce_info[MAX_MCE_INFO];
> +
> +static void mce_save_info(__u64 addr)
> +{
> +	int	i;

that tab looks weird. [there's repeat occurances further below 
as well]

> +
> +	for (i = 0; i < MAX_MCE_INFO; i++)
> +		if (atomic_cmpxchg(&mce_info[i].inuse, 0, 1) == 0) {
> +			mce_info[i].t = current;
> +			mce_info[i].paddr = addr;
> +			return;
> +		}

We typically use curly braces for all multi-line statements - so 
two would be needed above.

> +
> +	mce_panic("Too many concurrent recoverable errors", NULL, NULL);
> +}
> +
> +static int mce_find_info(__u64 *paddr)
> +{
> +	int	i;
> +
> +	for (i = 0; i < MAX_MCE_INFO; i++)
> +		if (atomic_read(&mce_info[i].inuse) &&
> +		    mce_info[i].t == current) {
> +			*paddr = mce_info[i].paddr;
> +			return 1;
> +		}
> +	return 0;
> +}
> +
> +static void mce_clear_info(void)
> +{
> +	int	i;
> +
> +	for (i = 0; i < MAX_MCE_INFO; i++)
> +		if (atomic_read(&mce_info[i].inuse) &&
> +		    mce_info[i].t == current) {

the line-break shows that the code has complexit troubles. Doing 
this in the loop iterator:

	struct mce_info *mi = mce_info + i;

would help make it shorter and more readable.

Thanks,

	Ingo