Re: [PATCH v5 1/6] arm64/mm: Introduce the init_pg_dir.

From: Mark Rutland
Date: Mon Sep 24 2018 - 10:03:44 EST


On Mon, Sep 24, 2018 at 02:01:37PM +0100, Mark Rutland wrote:
> On Mon, Sep 17, 2018 at 12:43:28PM +0800, Jun Yao wrote:
> > To make the swapper_pg_dir read only, we will move it to the rodata
> > section. And force the kernel to set up the initial page table in
> > the init_pg_dir. After generating all levels page table, we copy
> > only the top level into the swapper_pg_dir during paging_init().

> > +#define INIT_PG_TABLES \
> > + . = ALIGN(PAGE_SIZE); \
> > + init_pg_dir = .; \
> > + . += SWAPPER_DIR_SIZE; \
> > + init_pg_end = .;
>
> This will allocate a full set of PGD + PUD + PMD (+ PTE) levels, but the
> commit message says we only copy the top level table.
>
> The memory for otjher levels of table can be re-allocated once the
> kernel init phase is over, so that seems a bit worrying.
>
> Do we switch over to new copies of those levels, or do we continue using
> the entries from INIT_PG_TABLES?

I now see that we already allocate copies of those levels of table (and
free the originals in paging_init), so this is not a problem.

Sorry for the noise!

Mark.