[Info-vax] RMS record metadata, was: Re: Re; Spiralog, RMS Journaling (was Re: FREESPADRIFT)

[John Reagan]

On Monday, June 20, 2016 at 11:37:49 AM UTC-4, hb wrote:
> On 06/20/2016 03:47 PM, John Reagan wrote:
> 
> > You can put code into P2 space today on Itanium.  Fully supported.
> > Just link with /SEGMENT=CODE=P2.  The compilers don't care.  The
> > generated code is the same.
> 
> As far as I remember, C++ was different: you had (and still may have) to
> compile with POINTER_SIZE=64 to run the code from P2. Otherwise you may
> experience an unexpected ACCVIO. No, the linker doesn't check this: you
> say move code to P2 and the linker moves it. And yes, linker generated
> code is the same, as well.

I don't see anything in the current release notes or notes files but that isn't 100% proof.  The compiler should know that the FD is 32-bit addressable and the FD contains a 64-bit pointer to the code regardless of the current pointer size. It feels like a bug to me, it might have accidentally checked the current pointer size for the fetch from the FD?  Or maybe something in the STL had a 32-bit fetch?  Dunno.