Re: [PATCH 3/6] block: Create scsi_sense.h for SCSI and ATAPI

[Randy Dunlap]

On 05/23/2018 02:22 PM, Jens Axboe wrote:
> On 5/23/18 3:20 PM, Randy Dunlap wrote:
>> On 05/23/2018 02:14 PM, Jens Axboe wrote:
>>> On 5/23/18 2:52 PM, Kees Cook wrote:
>>>> On Wed, May 23, 2018 at 7:31 AM, Jens Axboe <axboe@xxxxxxxxx> wrote:
>>>>> On 5/23/18 8:25 AM, Christoph Hellwig wrote:
>>>>>> On Wed, May 23, 2018 at 08:13:56AM -0600, Jens Axboe wrote:
>>>>>>>> Should I move to code to a new drivers/scsi/scsi_sense.c and add it to
>>>>>>>> drivers/scsi/Makefile as:
>>>>>>>>
>>>>>>>> obj-$(CONFIG_BLK_SCSI_REQUEST)    += scsi_sense.o
>>>>>>>>
>>>>>>>> Every place I want to use the code is already covered by
>>>>>>>> CONFIG_BLK_SCSI_REQUEST, so it seems like I just need to know where to
>>>>>>>> put the .c file. :P
>>>>>>>
>>>>>>> I think this is so much saner than a SCSI select or dependency, so I'll
>>>>>>> have to disagree with Martin and Christoph. Just put it in drivers/scsi,
>>>>>>> if it's the location they care about.
>>>>>>
>>>>>> I actually plan to remove CONFIG_BLK_SCSI_REQUEST in a merge window
>>>>>> or two.  The only users are scsi and the ide layer, (virtio_blk
>>>>>> support has already been accidentally disabled for a while), and getting
>>>>>> rid of it allows to to shrink and simply the scsi data structures.
>>>>>>
>>>>>> But if you want this for now lets keep scsi_sense.c in drivers/scsi
>>>>>> but depend on CONFIG_BLK_SCSI_REQUEST, that is easy enough to fix up.
>>>>>
>>>>> It could be a stand-alone dependency, doesn't have to be BLK_SCSI_REQUEST.
>>>>> BLA_SCSI_SENSE or something would do. I don't care too much about that,
>>>>> mostly getting rid of the entire stack dependency.
>>>>
>>>> Aaand, I can't do this and leave it in drivers/scsi because of drivers/Makefile:
>>>>
>>>> obj-$(CONFIG_SCSI)              += scsi/
>>>>
>>>> So: this needs to live in block/ just like CONFIG_BLK_SCSI_REQUEST's
>>>> scsi_ioctl.c. I will split it into CONFIG_BLK_SCSI_SENSE, but I'll
>>>> still need to move the code from drivers/scsi/ to block/. Is this
>>>> okay?
>>>
>>> Ugh, so that would necessitate a change there too. As I said before,
>>> I don't really care where it lives. I know the SCSI folks seem bothered
>>> by moving it, but in reality, it's not like this stuff will likely ever
>>> really change. Of the two choices (select entire SCSI stack, or just move
>>> this little bit), I know what I would consider the saner option...
>>>
>>
>> or option 3:
>>
>> obj-y                           += scsi/
>>
>> so that make descends into drivers/scsi/ and then builds whatever is needed,
>> depending on individual kconfig options.
> 
> Right, that was the initial option, the later two are the other options.
> 

Sorry, I'm late to the party.

-- 
~Randy