Re: [PATCH 04/12] libata-hp-prep: update ata_scsi_find_dev() and friends

[Jeff Garzik <jgarzik@pobox.com>]

zhao, forrest wrote:
> On Mon, 2006-05-29 at 15:25 +0900, Tejun Heo wrote:
>> 113b8fe60b97a6ce664d6098f23b56b75247de87
>> diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c
>> index 9e5cb9f..9adc5c1 100644
>> --- a/drivers/scsi/libata-scsi.c
>> +++ b/drivers/scsi/libata-scsi.c
>> @@ -52,8 +52,12 @@ #include "libata.h"
>>  #define SECTOR_SIZE	512
>>  
>>  typedef unsigned int (*ata_xlat_func_t)(struct ata_queued_cmd *qc, const u8 *scsicmd);
>> -static struct ata_device *
>> -ata_scsi_find_dev(struct ata_port *ap, const struct scsi_device *scsidev);
>> +
>> +static struct ata_device * __ata_scsi_find_dev(struct ata_port *ap,
>> +					const struct scsi_device *scsidev);
>> +static struct ata_device * ata_scsi_find_dev(struct ata_port *ap,
>> +					    const struct scsi_device *scsidev);
>> +
>>  
>>  #define RW_RECOVERY_MPAGE 0x1
>>  #define RW_RECOVERY_MPAGE_LEN 12
>> @@ -2308,6 +2312,23 @@ static unsigned int atapi_xlat(struct at
>>  	return 0;
>>  }
>>  
>> +static struct ata_device * ata_find_dev(struct ata_port *ap, int id)
>> +{
>> +	if (likely(id == 0 || (id == 1 && ap->flags & ATA_FLAG_SLAVE_POSS)))
>> +		return &ap->device[id];
>> +	return NULL;
>> +}
>> +
> 
>>From my understanding, the definition of function ata_find_dev() should
> be in libata-core.c instead of libata-scsi.c. Does it make sense to you?

You are correct from a modularity standpoint, but doing so will 
eliminate the possibility of the compiler noticing that this function -- 
a fast path function -- is small enough to inline.  That would 
completely eliminate the function call overhead.

	Jeff