Re: [edk2-devel] [PATCH v2 1/1] MdePkg: Add new JedecJep106Lib to fetch JEDEC JEP106 manufacturer

["Rebecca Cran" <rebecca@bsdio.com>]

Sorry, not the struct - the array:


STATIC CONST JEDEC_MANUFACTURERS *Jep106Manufacturers[] = {
   Jep106ManufacturersBank1,
   Jep106ManufacturersBank2,
   Jep106ManufacturersBank3,
   Jep106ManufacturersBank4,
   Jep106ManufacturersBank5,
   Jep106ManufacturersBank6,
   Jep106ManufacturersBank7,
   Jep106ManufacturersBank8,
   Jep106ManufacturersBank9,
   Jep106ManufacturersBank10,
   Jep106ManufacturersBank11,
   Jep106ManufacturersBank12,
   Jep106ManufacturersBank13,
   Jep106ManufacturersBank14
};

On 4/7/23 5:06 PM, Michael D Kinney wrote:
> Which struct?  That macro only applies to global variables, not structure declarations.
>
> Mike
>
>
>> -----Original Message-----
>> From: devel@edk2.groups.io <devel@edk2.groups.io> On Behalf Of Rebecca Cran
>> Sent: Friday, April 7, 2023 3:59 PM
>> To: Kinney, Michael D <michael.d.kinney@intel.com>; devel@edk2.groups.io; Liu, Zhiguang <zhiguang.liu@intel.com>; Gao, Liming
>> <gaoliming@byosoft.com.cn>
>> Cc: Rebecca Cran <rebecca@quicinc.com>
>> Subject: Re: [edk2-devel] [PATCH v2 1/1] MdePkg: Add new JedecJep106Lib to fetch JEDEC JEP106 manufacturer
>>
>> I guess GLOBAL_REMOVE_IF_UNREFERENCED should be added to the struct too?
>>
>>
>> On 4/7/23 4:42 PM, Rebecca Cran wrote:
>>> On 4/7/23 2:25 PM, Kinney, Michael D wrote:
>>>> Comments below.
>>>>
>>>> Hopefully this lib would only be used by modules that get compressed.
>>> I guess so, but that's for the user to decide.
>>>> Might add GLOBAL_REMOVE_IF_UNREFERENCED to the arrays of strings to
>>>> help the optimizer remove the data that is not referenced.
>>> Good idea - I'll add that.
>>>
>>>>> +CONST CHAR8 *
>>>>> +EFIAPI
>>>>> +Jep106GetManufacturerName (
>>>>> +  IN UINT8  Code,
>>>>> +  IN UINT8  ContinuationBytes
>>>>> +  )
>>>>> +{
>>>>> +  UINTN                      Index;
>>>>> +  CONST JEDEC_MANUFACTURERS  *ManufacturersBank;
>>>>> +
>>>>> +  Index = 0;
>>>>> +
>>>>> +  if (ContinuationBytes >= JEP106_MANUFACTURERS_NUM_BANKS) {
>>>>> +    ASSERT (0);
>>>> Do you really want an ASSERT() from this?  If this is data from a DIMM
>>>> I doubt we want platform to ASSERT().  Perhaps just return NULL?
>>> I'll remove it. I added it for validation when I was writing the SPD
>>> parsing library.
>>>>> +UINTN
>>>>> +EFIAPI
>>>>> +Jep106GetLongestManufacturerName (
>>>>> +  VOID
>>>>> +  )
>>>> Why is this API needed?  Wouldn’t you really just need the
>>>> longest of the ones present in the current boot to build the
>>>> SMBIOS record?.  Not the longest of all in the banks?
>>> I added it because it's useful to know the maximum possible size of
>>> the strings that could be in the SMBIOS Type 17 table before doing the
>>> SPD parsing.
>>>
>>> I'll remove it for now, and if it's really needed I can add it back in.
>>>
>>>
>>
>>
>>
>
>
> 
>
>