Re: mpt2sas: BUG: using smp_processor_id() in preemptible

[Jan Schmidt <list.linux-scsi@jan-o-sch.net>]

Hi Eric,

On 21.02.2012 00:06, Moore, Eric wrote:
> Jan - the smp_processor_id() is part of NUMA support which I added over a year ago. I don't see any problem with the driver running on sles11sp2 which has btrfs support ( I'm not sure  if CONFIG_DEBUG_PREEMPT is turned on as I have not checked).   Any theory why its oopsing in smp_processor_id() ?

-- excerpt from include/linux/smp.h --
/*
 * smp_processor_id(): get the current CPU ID.
 *
 * if DEBUG_PREEMPT is enabled then we check whether it is
 * used in a preemption-safe way. (smp_processor_id() is safe
 * if it's used in a preemption-off critical section, or in
 * a thread that is bound to the current CPU.)
[...]
 */
#ifdef CONFIG_DEBUG_PREEMPT
  extern unsigned int debug_smp_processor_id(void);
# define smp_processor_id() debug_smp_processor_id()
#else
# define smp_processor_id() raw_smp_processor_id()
#endif
 
#define get_cpu()               ({ preempt_disable(); smp_processor_id(); })
#define put_cpu()               preempt_enable()
-- end --

... and ...

-- excerpt from drivers/scsi/mpt2sas/mpt2sas_base.c --
static inline u8 
_base_get_msix_index(struct MPT2SAS_ADAPTER *ioc)
{
        return ioc->cpu_msix_table[smp_processor_id()]; 
}
-- end --

Suggestion: Use get_cpu() and put_cpu().

-Jan