Re: [PATCH V2 17/17] MdeModulePkg: Add PEIM and lib to dsc file

[Jordan Justen <jordan.l.justen@intel.com>]

Whoops. It looks like this was specifically addressed in v3 of the
series. I just didn't see a follow up email discussion to Andrew's
email.

Please disregard.

-Jordan

On 2019-03-31 23:17:12, Jordan Justen wrote:
> On 2019-03-14 22:54:24, Andrew Fish via edk2-devel wrote:
> > I understand the motivation for this change as I've done something
> > much less portable that looks a lot like this to save the PEI XIP
> > space....
> > 
> > I seem to remember a long time ago we add a public VA_LIST to an API
> > and we ran into an issue due to the marker format being compiler
> > specific. You don't tend to see this on VC++ as it mostly uses a
> > pointer to the frame, but GCC and clang can use a data structure
> > especially not on IA32 (i386) for a VA_LIST (this is part of the
> > reason there are so many #ifdefs in the Base.h VA_LIST section).
> > 
> > In the past to fix this Mike Kinney added BASE_LIST. I seem to
> > remember MdeModulePkg/Library/PeiDxeDebugLibReportStatusCode passes
> > a BASE_LIST into ReportStatusCode vs. a VA_LIST for this reason. Was
> > this VA_LIST portability concern taken into consideration for this
> > new API?
> > 
> > In general VA_LIST is not an issue as long as all the code is
> > compiled with the same compiler but if binaries are in play then you
> > can have issues. So things like FSB, Option ROMs, OS Loaders, EFI
> > Shell, EFI Apps. are when you can see the issue.
> > 
> 
> I didn't see a reply to Andrew's concern. Was it considered, and
> addressed?
> 
> I remember having several issues in the past with var-args. I haven't
> seen it show up in quite a while.
> 
> I know we have to make sure to consistently apply EFIAPI to such
> function declarations.
> 
> I did think we still considered it forbidden to use var-args across a
> binary interface. (PPI, Protocol, Core functions)
> 
> I'm not sure if EFIAPI is enough to make it safe as long as all code
> is compiled by the same compiler, or if it truely makes it safe for a
> binary compiled with one compiler to be called from a binary compiled
> with any other compiler.
> 
> I'll add Ard and Laszlo to the Cc, as I seem to recall they may have
> looked at it before. (Although, I think they've been Cc'd on the
> series, so maybe they've already considered this, and don't expect
> there is a problem.)
> 
> -Jordan