Re: [RFC PATCH v3 2/2] tracing: add trace event for memory-failure

[Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>]

On Tue, Apr 07, 2015 at 07:05:31PM +0800, Xie XiuQi wrote:
> RAS user space tools like rasdaemon which base on trace event, could
> receive mce error event, but no memory recovery result event. So, I
> want to add this event to make this scenario complete.
> 
> This patch add a event at ras group for memory-failure.
> 
> The output like below:
> #  tracer: nop
> #
> #  entries-in-buffer/entries-written: 2/2   #P:24
> #
> #                               _-----=> irqs-off
> #                              / _----=> need-resched
> #                             | / _---=> hardirq/softirq
> #                             || / _--=> preempt-depth
> #                             ||| /     delay
> #            TASK-PID   CPU#  ||||    TIMESTAMP  FUNCTION
> #               | |       |   ||||       |         |
>        mce-inject-13150 [001] ....   277.019359: memory_failure_event: pfn 0x19869: recovery action for free buddy page: Delayed
> 
> 
> Cc: Tony Luck <tony.luck@intel.com>
> Cc: Steven Rostedt <rostedt@goodmis.org>
> Signed-off-by: Xie XiuQi <xiexiuqi@huawei.com>
> ---
>  include/ras/ras_event.h | 83 +++++++++++++++++++++++++++++++++++++++++++++++++
>  kernel/trace/trace.c    |  2 +-
>  mm/memory-failure.c     |  2 ++
>  3 files changed, 86 insertions(+), 1 deletion(-)
> 
> diff --git a/include/ras/ras_event.h b/include/ras/ras_event.h
> index 79abb9c..52c75f2 100644
> --- a/include/ras/ras_event.h
> +++ b/include/ras/ras_event.h
> @@ -11,6 +11,7 @@
>  #include <linux/pci.h>
>  #include <linux/aer.h>
>  #include <linux/cper.h>
> +#include <linux/mm.h>
>  
>  /*
>   * MCE Extended Error Log trace event
> @@ -232,6 +233,88 @@ TRACE_EVENT(aer_event,
>  		__print_flags(__entry->status, "|", aer_uncorrectable_errors))
>  );
>  
> +/*
> + * memory-failure recovery action result event
> + *
> + * unsigned long pfn -	Page Number of the corrupted page

Hi XiuQi,

I think "Page Frame Number" is better.

And embracing these definition code with #ifdef CONFIG_MEMORY_FAILURE might
be helpful if it reduces binary size for !CONFIG_MEMORY_FAILURE build?

Thanks,
Naoya Horiguchi

> + * int type	-	Page types of the corrupted page
> + * int result	-	Result of recovery action
> + */
> +
> +#define MF_ACTION_RESULT	\
> +	EM ( MF_IGNORED, "Ignord" )	\
> +	EM ( MF_FAILED,  "Failed" )	\
> +	EM ( MF_DELAYED, "Delayed" )	\
> +	EMe ( MF_RECOVERED, "Recovered" )
> +
> +#define MF_PAGE_TYPE		\
> +	EM ( MF_KERNEL, "reserved kernel page" )			\
> +	EM ( MF_KERNEL_HIGH_ORDER, "high-order kernel page" )		\
> +	EM ( MF_SLAB, "kernel slab page" )				\
> +	EM ( MF_DIFFERENT_COMPOUND, "different compound page after locking" ) \
> +	EM ( MF_POISONED_HUGE, "huge page already hardware poisoned" )	\
> +	EM ( MF_HUGE, "huge page" )					\
> +	EM ( MF_FREE_HUGE, "free huge page" )				\
> +	EM ( MF_UNMAP_FAILED, "unmapping failed page" )			\
> +	EM ( MF_DIRTY_SWAPCACHE, "dirty swapcache page" )		\
> +	EM ( MF_CLEAN_SWAPCACHE, "clean swapcache page" )		\
> +	EM ( MF_DIRTY_MLOCKED_LRU, "dirty mlocked LRU page" )		\
> +	EM ( MF_CLEAN_MLOCKED_LRU, "clean mlocked LRU page" )		\
> +	EM ( MF_DIRTY_UNEVICTABLE_LRU, "dirty unevictable LRU page" )	\
> +	EM ( MF_CLEAN_UNEVICTABLE_LRU, "clean unevictable LRU page" )	\
> +	EM ( MF_DIRTY_LRU, "dirty LRU page" )				\
> +	EM ( MF_CLEAN_LRU, "clean LRU page" )				\
> +	EM ( MF_TRUNCATED_LRU, "already truncated LRU page" )		\
> +	EM ( MF_BUDDY, "free buddy page" )				\
> +	EM ( MF_BUDDY_2ND, "free buddy page (2nd try)" )		\
> +	EMe ( MF_UNKNOWN, "unknown page" )
> +
> +/*
> + * First define the enums in MM_ACTION_RESULT to be exported to userspace
> + * via TRACE_DEFINE_ENUM().
> + */
> +#undef EM
> +#undef EMe
> +#define EM(a,b) TRACE_DEFINE_ENUM(a);
> +#define EMe(a,b)	TRACE_DEFINE_ENUM(a);
> +
> +MF_ACTION_RESULT
> +MF_PAGE_TYPE
> +
> +/*
> + * Now redefine the EM() and EMe() macros to map the enums to the strings
> + * that will be printed in the output.
> + */
> +#undef EM
> +#undef EMe
> +#define EM(a,b)		{ a, b },
> +#define EMe(a,b)	{ a, b }
> +
> +TRACE_EVENT(memory_failure_event,
> +	TP_PROTO(unsigned long pfn,
> +		 int type,
> +		 int result),
> +
> +	TP_ARGS(pfn, type, result),
> +
> +	TP_STRUCT__entry(
> +		__field(unsigned long, pfn)
> +		__field(int, type)
> +		__field(int, result)
> +	),
> +
> +	TP_fast_assign(
> +		__entry->pfn	= pfn;
> +		__entry->type	= type;
> +		__entry->result	= result;
> +	),
> +
> +	TP_printk("pfn %#lx: recovery action for %s: %s",
> +		__entry->pfn,
> +		__print_symbolic(__entry->type, MF_PAGE_TYPE),
> +		__print_symbolic(__entry->result, MF_ACTION_RESULT)
> +	)
> +);
>  #endif /* _TRACE_HW_EVENT_MC_H */
>  
>  /* This part must be outside protection */
> diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> index 25334e7..f5e8856 100644
> --- a/kernel/trace/trace.c
> +++ b/kernel/trace/trace.c
> @@ -6776,7 +6776,7 @@ static struct notifier_block trace_module_nb = {
>  };
>  #endif
>  
> -static __init int tracer_init_debugfs(void)
> +static __init int tracer_init_tracefs(void)
>  {
>  	struct dentry *d_tracer;
>  
> diff --git a/mm/memory-failure.c b/mm/memory-failure.c
> index 34e9c65..d118af8 100644
> --- a/mm/memory-failure.c
> +++ b/mm/memory-failure.c
> @@ -850,6 +850,8 @@ static struct page_state {
>   */
>  static void action_result(unsigned long pfn, int type, int result)
>  {
> +	trace_memory_failure_event(pfn, type, result);
> +
>  	pr_err("MCE %#lx: recovery action for %s: %s\n",
>  		pfn, action_page_type[type], action_name[result]);
>  }
> -- 
> 1.8.3.1
> 
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>