Re: [PATCH] ide: Fix ata_id_has_dword_io to return DWORD I/O support properly

[Sergei Shtylyov <sshtylyov@ru.mvista.com>]

Hello.

Jeff Garzik wrote:

>> This patch fixes ata_id_has_dword_io to return 1 (supported) if
>> the drive is compliant to ATA2 or newer and evaluates the config
>> word for older drives.

>> Signed-off-by: Mario Schwalbe <schwalbe@inf.tu-dresden.de>

>> diff --git a/include/linux/ata.h b/include/linux/ata.h
>> index a53318b..d0d6a9c 100644
>> --- a/include/linux/ata.h
>> +++ b/include/linux/ata.h
>> @@ -700,12 +700,12 @@ static inline int ata_id_has_tpm(const u16 *id)

>>  static inline int ata_id_has_dword_io(const u16 *id)
>>  {
>> -    /* ATA 8 reuses this flag for "trusted" computing */
>> -    if (ata_id_major_version(id) > 7)
>> -        return 0;
>> -    if (id[ATA_ID_DWORD_IO] & (1 << 0))
>> -        return 1;
>> -    return 0;
>> +    /* This flag is defined up to ATA1 and deprecated since then
>> +       (ATA 8 reuses this flag for "trusted" computing). */
>> +    if (ata_id_major_version(id) <= 1)
>> +        return (id[ATA_ID_DWORD_IO] & (1 << 0)) != 0;
>> +    /* later revision drives support DWORD I/O just fine */
>> +    return 1;

> This seems like a risky assumption...

    Not at all, I think. It's total mystery how "DWORD I/O" could've been even 
(not) supported by a *drive* at all. Support of "DWORD I/O" is a property of 
the PATA controller. Frankly speaking, I don't think that this function is at 
all useful...

>     Jeff

MBR, Sergei