Re: [PATCH v2 1/2] mm/memblock: expose only miminal interface to add/walk physmem

[Heiko Carstens]

On Wed, Jul 01, 2020 at 06:06:43PM +0300, Mike Rapoport wrote:
> Hi David,
> 
> On Wed, Jul 01, 2020 at 04:18:29PM +0200, David Hildenbrand wrote:
> > "physmem" in the memblock allocator is somewhat weird: it's not actually
> > used for allocation, it's simply information collected during boot, which
> > describes the unmodified physical memory map at boot time, without any
> > standby/hotplugged memory. It's only used on s390x and is currently the
> > only reason s390x keeps using CONFIG_ARCH_KEEP_MEMBLOCK.
> > 
> > Physmem isn't numa aware and current users don't specify any flags. Let's
> > hide it from the user, exposing only for_each_physmem(), and simplify. The
> > interface for physmem is now really minimalistic:
> > - memblock_physmem_add() to add ranges
> > - for_each_physmem() / __next_physmem_range() to walk physmem ranges
> > 
> > Don't place it into an __init section and don't discard it without
> > CONFIG_ARCH_KEEP_MEMBLOCK. As we're reusing __next_mem_range(), remove
> > the __meminit notifier to avoid section mismatch warnings once
> > CONFIG_ARCH_KEEP_MEMBLOCK is no longer used with
> > CONFIG_HAVE_MEMBLOCK_PHYS_MAP.
> > 
> > While fixing up the documentation, sneak in some related cleanups. We can
> > stop setting CONFIG_HAVE_MEMBLOCK_PHYS_MAP for s390x next.
> 
> As you noted in the previous version it should have been
> CONFIG_ARCH_KEEP_MEMBLOCK ;-)
> 
> > Cc: Heiko Carstens <heiko.carstens@xxxxxxxxxx>
> > Cc: Vasily Gorbik <gor@xxxxxxxxxxxxx>
> > Cc: Christian Borntraeger <borntraeger@xxxxxxxxxx>
> > Cc: Mike Rapoport <rppt@xxxxxxxxxxxxx>
> > Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> > Signed-off-by: David Hildenbrand <david@xxxxxxxxxx>
> 
> Reviewed-by: Mike Rapoport <rppt@xxxxxxxxxxxxx>
> 
> > ---
> >  arch/s390/kernel/crash_dump.c |  6 ++--
> >  include/linux/memblock.h      | 28 ++++++++++++++---
> >  mm/memblock.c                 | 57 ++++++++++++++++++-----------------
> >  3 files changed, 55 insertions(+), 36 deletions(-)

So I guess this should go via the s390 tree, since the second patch of
this series can go only upstream if both this patch and a patch which
is currently only on our features are merged before.

Any objections?