On Mon, 2019-09-02 at 15:47 +0200, Jan Beulich wrote:
> On 02.09.2019 14:51, David Woodhouse wrote:
> > On Mon, 2019-09-02 at 09:44 +0200, Jan Beulich wrote:
> > > Right, just one pair should survive. And seeing how things work before
> > > this series I think it indeed should be linker script symbols only.
> > > And then the ALIGN() ahead of the "start" ones should stay, but there's
> > > no need for one on the "end" ones (again as is currently the case).
> > 
> > If we don't align the end symbol then we need to go back to rounding up
> > the length with ((boot_trampoline_end - boot_trampoline_start) + 3) / 4
> > again though, right?
> 
> Wait - we've been talking about the *_rel sections / tables here,
> haven't we? All entries of these tables ought to be of equal size,
> and hence alignment of a table's "end" label automatically matches
> the size of the table entries.

The specific one we were taking about just then was
bootdata_{start,end} which is the data itself to be copied up/down, not
the relocations.

The _rel sections indeed need no alignment at the end, as you say.