Re: TYPE_RBC cache fixes (sbp2.c affected)

[Douglas Gilbert <dougg@torque.net>]

Douglas Gilbert wrote:
> James Bottomley wrote:
> 
>> On Sat, 2005-05-21 at 17:22 +0100, Al Viro wrote:
>>
>>> Tell that to firmware authors, why don't you?
>>
>>
>>
>> I do ... but they don't listen ...
>>  
>>
>>>> Does anyone actually have one of these RBC devices and does it reject
>>>> the six byte mode sense commands?
>>>
>>>
>>> Yes, will check and do not expect the results to apply to other 
>>> devices...
>>
>>
>>
>> Thanks ... I'd be surprised if the entire class of RBC devices simply
>> ignored the standard; I wouldn't be surprised to find one or two that
>> are out of spec.
>>
>> James
> 
> James,
> Perhaps __scsi_mode_sense() could do a simple sanity
> check: for any valid mode page [after the header and
> block descriptor(s) are stepped over]:
>      ((mpage[0] & 3f) == page_num)
> 
> If a response fails that test, we don't believe it.

A bit more accurate sanity check:
     if ((page_num > 0) && (page_num < 0x3f) &&
         ((mpage[0] & 3f) == page_num))
	// page looks ok (even if subpage_num was 0xff)
     else
	// nah

Mode page number 0 is the unit attention vendor specific
page which is not necessarily in "mode page format".
The vendor that I know does use it, does at least follow
mode page format.
A requested page_num of 0x3f and a requested subpage_num
of 0xff are wildcards.

Doug Gilbert



-------------------------------------------------------
This SF.Net email is sponsored by Oracle Space Sweepstakes
Want to be the first software developer in space?
Enter now for the Oracle Space Sweepstakes!
http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click