Re: [PATCH 1/4] hw/ufs: minor bug fixes related to ufs-test

[Jeuk Kim <jeuk20.kim@gmail.com>]

On 8/22/2024 3:00 PM, Yoochan Jeong wrote:
> On 8/22/2024, Jeuk Kim wrote:
>> On 8/21/2024 11:32 AM, 정유찬 wrote:
>>>    static void ufs_send_query(QUfs *ufs, uint8_t slot, uint8_t query_function,
>>>                                uint8_t query_opcode, uint8_t idn, uint8_t index,
>>> +                          uint8_t selector, uint32_t attr_value,
>> We use ufs_send_query() not only for attributes, but also descriptors
>> and flags.
>>
>> Please rename `attr_value` to `value`.
>>
>>
>   
> I think this parameter name is okay, because this "value" in UPIU is
> only used when writing attributes. Writing flags do not require an
> actual value, and descriptor data will be stored in data segmentation
> area.

Okay. That's reasonable.


>   
>>>                            
>> We might need to check condition here that `query_opcode !=
>> UFS_UPIU_QUERY_OPCODE_WRITE_DESC`, since
>>
>> it is not implemented yet.
>>
> in ufs_exec_query_write function in ufs.c, it already checks
> if it is trying to write a descriptor. Is there any particular
> reason that we should check it here in advance?

You're right. We don't need to check it here.