[RFC v14-rc][PATCH 17/23] Record 'struct file' object instead of the file name for VMAs
orenl at cs.columbia.edu
Fri Mar 20 15:02:27 PDT 2009
Dave Hansen wrote:
> On Fri, 2009-03-20 at 17:20 -0400, Oren Laadan wrote:
>> What you suggest will reverse the order ... so while we expect to find
>> a vma, we'll find a file object. I'm wonder if it's worth changing the
>> format and logic for this.
> It makes logical sense to me to do this. You depend on the file to
> create the VMA. It makes sense to see dependencies before the object
> that is doing the depending.
> -- Dave
Or: you look at the vma header to know that you depend on the file...
(not that I really care how logic reads)
Anyway, the down side that I was hoping to avoid on restart, is to need
to expect "either...or" values of 'struct cr_hdr'->type instead of a
single deterministic value. That's the extra burden that I was referring
More information about the Containers