[PATCH v1 01/10] mm/memory_hotplug: Don't allow to online/offline memory blocks with holes
dan.j.williams at intel.com
Tue Nov 5 01:30:25 UTC 2019
On Thu, Oct 24, 2019 at 5:10 AM David Hildenbrand <david at redhat.com> wrote:
> Our onlining/offlining code is unnecessarily complicated. Only memory
> blocks added during boot can have holes (a range that is not
> IORESOURCE_SYSTEM_RAM). Hotplugged memory never has holes (e.g., see
> add_memory_resource()). All boot memory is alread online.
...also perhaps clarify "already online" by what point in time and why
that is relevant. For example a description of the difference between
the SetPageReserved() in the bootmem path and the one in the hotplug
> Therefore, when we stop allowing to offline memory blocks with holes, we
> implicitly no longer have to deal with onlining memory blocks with holes.
Maybe an explicit reference of the code areas that deal with holes
would help to back up that assertion. Certainly it would have saved me
some time for the review.
> This allows to simplify the code. For example, we no longer have to
> worry about marking pages that fall into memory holes PG_reserved when
> onlining memory. We can stop setting pages PG_reserved.
...but not for bootmem, right?
> Offlining memory blocks added during boot is usually not guranteed to work
> either way (unmovable data might have easily ended up on that memory during
> boot). So stopping to do that should not really hurt (+ people are not
> even aware of a setup where that used to work
Maybe put a "Link: https://lkml.kernel.org/r/$msg_id" to that discussion?
> and that the existing code
> still works correctly with memory holes). For the use case of offlining
> memory to unplug DIMMs, we should see no change. (holes on DIMMs would be
However, less memory can be offlined than was theoretically allowed
previously, so I don't understand the "we should see no change"
comment. I still agree that's a price worth paying to get the code
cleanups and if someone screams we can look at adding it back, but the
fact that it was already fragile seems decent enough protection.
> Please note that hardware errors (PG_hwpoison) are not memory holes and
> not affected by this change when offlining.
> Cc: Andrew Morton <akpm at linux-foundation.org>
> Cc: Michal Hocko <mhocko at suse.com>
> Cc: Oscar Salvador <osalvador at suse.de>
> Cc: Pavel Tatashin <pasha.tatashin at soleen.com>
> Cc: Dan Williams <dan.j.williams at intel.com>
> Cc: Anshuman Khandual <anshuman.khandual at arm.com>
> Signed-off-by: David Hildenbrand <david at redhat.com>
> mm/memory_hotplug.c | 26 ++++++++++++++++++++++++--
> 1 file changed, 24 insertions(+), 2 deletions(-)
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 561371ead39a..8d81730cf036 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1447,10 +1447,19 @@ static void node_states_clear_node(int node, struct memory_notify *arg)
> node_clear_state(node, N_MEMORY);
> +static int count_system_ram_pages_cb(unsigned long start_pfn,
> + unsigned long nr_pages, void *data)
> + unsigned long *nr_system_ram_pages = data;
> + *nr_system_ram_pages += nr_pages;
> + return 0;
> static int __ref __offline_pages(unsigned long start_pfn,
> unsigned long end_pfn)
> - unsigned long pfn, nr_pages;
> + unsigned long pfn, nr_pages = 0;
> unsigned long offlined_pages = 0;
> int ret, node, nr_isolate_pageblock;
> unsigned long flags;
> @@ -1461,6 +1470,20 @@ static int __ref __offline_pages(unsigned long start_pfn,
> + /*
> + * Don't allow to offline memory blocks that contain holes.
> + * Consecuently, memory blocks with holes can never get onlined
> + * (hotplugged memory has no holes and all boot memory is online).
> + * This allows to simplify the onlining/offlining code quite a lot.
> + */
The last sentence of this comment makes sense in the context of this
patch, but I don't think it stands by itself in the code base after
the fact. The person reading the comment can't see the simplifications
because the code is already gone. I'd clarify it to talk about why it
is safe to not mess around with PG_Reserved in the hotplug path
because of this check.
After those clarifications you can add:
Reviewed-by: Dan Williams <dan.j.williams at intel.com>
More information about the devel