Re: [edk2-devel] [PATCH v3 1/5] UefiCpuPkg/SmmBaseHob.h: Add SMM Base HOB Data

["Ni, Ray" <ray.ni@intel.com>]

I think we need to define a new chunk hob type for solution proposed by Gerd.
I will leave that to PIWG (PI spec working group) for discussion.

In this topic, I prefer to solve the problem in the CPU domain. 

> -----Original Message-----
> From: Wu, Jiaxin <jiaxin.wu@intel.com>
> Sent: Friday, February 3, 2023 9:23 PM
> To: kraxel@redhat.com; Ni, Ray <ray.ni@intel.com>
> Cc: Johnson, Brian <brian.johnson@hpe.com>; devel@edk2.groups.io; Laszlo
> Ersek <lersek@redhat.com>; Dong, Eric <eric.dong@intel.com>; Zeng, Star
> <star.zeng@intel.com>; Kumar, Rahul R <rahul.r.kumar@intel.com>; Kinney,
> Michael D <michael.d.kinney@intel.com>; Zimmer, Vincent
> <vincent.zimmer@intel.com>
> Subject: RE: [edk2-devel] [PATCH v3 1/5] UefiCpuPkg/SmmBaseHob.h: Add SMM
> Base HOB Data
> 
> Hi Gerd,
> 
> Is it still the solution with multiple hobs created for big data but you want the
> hob splitting and reassembling can be encapsulated in the Hoblib?
> 
> Does it need define the new hob type for chunk?
> 
> Thanks,
> Jiaxin
> 
> > -----Original Message-----
> > From: kraxel@redhat.com <kraxel@redhat.com>
> > Sent: Friday, February 3, 2023 3:55 PM
> > To: Ni, Ray <ray.ni@intel.com>
> > Cc: Johnson, Brian <brian.johnson@hpe.com>; devel@edk2.groups.io; Laszlo
> > Ersek <lersek@redhat.com>; Wu, Jiaxin <jiaxin.wu@intel.com>; Dong, Eric
> > <eric.dong@intel.com>; Zeng, Star <star.zeng@intel.com>; Kumar, Rahul R
> > <rahul.r.kumar@intel.com>; Kinney, Michael D <michael.d.kinney@intel.com>;
> > Zimmer, Vincent <vincent.zimmer@intel.com>
> > Subject: Re: [edk2-devel] [PATCH v3 1/5] UefiCpuPkg/SmmBaseHob.h: Add
> > SMM Base HOB Data
> >
> > On Fri, Feb 03, 2023 at 03:14:42AM +0000, Ni, Ray wrote:
> > > Gerd,
> > > Can you please explain a bit more on the chunk idea?
> >
> > > > > to introduce a generic and reusable concept of chunked HOBs, so you
> can
> > > > > add helper functions to HobLib for splitting and reassembling, with a
> > > > > struct along the lines of:
> > > > >
> > > > > typedef struct {
> > > > > 	// offset and size of this particular chunk
> > > > > 	UINT32	ChunkOffset;
> > > > > 	UINT32	ChunkSize;
> > > > >
> > > > > 	// number of chunks and size of all chunks combined.
> > > > > 	UINT32	ChunkCount;
> > > > > 	UINT32	TotalSize;
> > > > >
> > > > > 	// chunk data
> > > > > 	UINT8   Data[0];
> > > > > } EFI_HOB_CHUNK;
> >
> > Reassembling would work like this:
> >
> >   // once
> >   AssembledHob = malloc(HobChunk->TotalSize);
> >
> >   // for each chunk
> >   memcpy(AssembledHob + HobChunk->ChunkOffset, HobChunk->Data,
> > HobChunk->ChunkSize);
> >
> > Possible shortcut:
> >
> >   if (HobChunk->ChunkSize == HobChunk->TotalSize)
> >     // data is not splitted into multiple chunks
> >     AssembledHob = HobChunk->Data
> >
> > Advantage: you avoid the allocation in case the data fits into a single
> > HOB.  Disadvantage: you need to track whenever AssembledHob is allocated
> > (and must eventually be freed) or not.
> >
> > HobChunk->ChunkCount is not really needed but might be useful for sanity
> > checking.
> >
> > take care,
> >   Gerd