LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: David Hildenbrand <david@redhat.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: Matthew Wilcox <willy@infradead.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	linux-mm@kvack.org, Steven Rostedt <rostedt@goodmis.org>,
	Ingo Molnar <mingo@redhat.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Huang Ying <ying.huang@intel.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Pavel Tatashin <pasha.tatashin@oracle.com>,
	Miles Chen <miles.chen@mediatek.com>,
	Mel Gorman <mgorman@techsingularity.net>,
	Rik van Riel <riel@redhat.com>, James Hogan <jhogan@kernel.org>,
	"Levin, Alexander (Sasha Levin)" <alexander.levin@verizon.com>,
	open list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH RFC 2/8] mm: introduce PG_offline
Date: Mon, 30 Apr 2018 08:31:00 +0200	[thread overview]
Message-ID: <a0748efe-60e1-bc85-dbf3-a2352ab2c5b1@redhat.com> (raw)
In-Reply-To: <20180429210850.GB26305@dhcp22.suse.cz>

On 29.04.2018 23:08, Michal Hocko wrote:
> On Sun 22-04-18 17:13:52, David Hildenbrand wrote:
>> On 22.04.2018 16:02, Matthew Wilcox wrote:
>>> On Sun, Apr 22, 2018 at 10:17:31AM +0200, David Hildenbrand wrote:
>>>> On 22.04.2018 05:01, Matthew Wilcox wrote:
>>>>> On Sat, Apr 21, 2018 at 06:52:18PM +0200, Vlastimil Babka wrote:
>>>>>> Sounds like your newly introduced "page types" could be useful here? I
>>>>>> don't suppose those offline pages would be using mapcount which is
>>>>>> aliased there?
>>>>>
>>>>> Oh, that's a good point!  Yes, this is a perfect use for page_type.
>>>>> We have something like twenty bits available there.
>>>>>
>>>>> Now you've got me thinking that we can move PG_hwpoison and PG_reserved
>>>>> to be page_type flags too.  That'll take us from 23 to 21 bits (on 32-bit,
>>>>> with PG_UNCACHED)
>>>>
>>>> Some things to clarify here. I modified the current RFC to also allow
>>>> PG_offline on allocated (ballooned) pages (e.g. virtio-balloon).
>>>>
>>>> kdump based dump tools can then easily identify which pages are not to
>>>> be dumped (either because the content is invalid or not accessible).
>>>>
>>>> I previously stated that ballooned pages would be marked as PG_reserved,
>>>> which is not true (at least not for virtio-balloon). However this allows
>>>> me to detect if all pages in a section are offline by looking at
>>>> (PG_reserved && PG_offline). So I can actually tell if a page is marked
>>>> as offline and allocated or really offline.
>>>>
>>>>
>>>> 1. The location (not the number!) of PG_hwpoison is basically ABI and
>>>> cannot be changed. Moving it around will most probably break dump tools.
>>>> (see kernel/crash_core.c)
>>>
>>> It's not ABI.  It already changed after 4.9 when PG_waiters was introduced
>>> by commit 62906027091f.
>>
>> It is, please have a look at the file I pointed you to.
>>
>> We export the *value* of PG_hwpoison in the ELF file, therefore the
>> *value* can change, but the *location* (page_flags, mapcount, whatever)
>> must not change. Or am I missing something here? I don't think we can
>> move PG_hwpoison that easily.
>>
>> Also, I can read "For pages that are never mapped to userspace,
>> page->mapcount may be used for storing extra information about page
>> type" - is that true for PG_hwpoison/PG_reserved? I am skeptical.
>>
>> And we need something similar for PG_offline, because it will become
>> ABI. (I can see that PAGE_BUDDY_MAPCOUNT_VALUE is also exported in an
>> ELF file, so maybe a new page type might work for marking a page offline
>> - but I have to look at the details first tomorrow)
> 
> Wait wait wait. Who is relying on this? Kdump? Page flags have always
> been an internal implementation detail and _nobody_ outside of the
> kernel should ever rely on the specific value. Well, kdump has been
> cheating but that is because kdump is inherently tight to a specific
> kernel implementation but that doesn't make it a stable ABI IMHO.
> Restricting the kernel internals because of a debugging tool would be
> quite insane.
> 

kdump tools (makedumptool) don't rely on any specific value or assume
anything.

Using the example of musing PG_hwpoison to mapcount:

If it sees PG_hwpoison:
- it knows the right bit number to use
- it knows the kernel uses it

If it doesn't see PG_hwpoison (in the ELF info) anymore:
- it cannot exclude poisoned pages anymore, potentially crashing the
system during a dump


If you have a better fitting name for "requires a interlocked update
with tools to keep it working" than ABI, please let me know :)

Anyhow, I have a new prototype based on PAGE_OFFLINE_MAPCOUNT_VALUE that
I will share briefly.

-- 

Thanks,

David / dhildenb

  reply	other threads:[~2018-04-30  6:31 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20180413131632.1413-1-david@redhat.com>
2018-04-13 13:16 ` [PATCH RFC 1/8] mm/memory_hotplug: Revert "mm/memory_hotplug: optimize memory hotplug" David Hildenbrand
2018-04-13 13:16 ` [PATCH RFC 2/8] mm: introduce PG_offline David Hildenbrand
2018-04-13 13:40   ` Michal Hocko
2018-04-13 13:46     ` David Hildenbrand
2018-04-17 11:50     ` David Hildenbrand
2018-04-13 17:11   ` Matthew Wilcox
2018-04-16  8:31     ` David Hildenbrand
2018-04-21 16:52     ` Vlastimil Babka
2018-04-22  3:01       ` Matthew Wilcox
2018-04-22  8:17         ` David Hildenbrand
2018-04-22 14:02           ` Matthew Wilcox
2018-04-22 15:13             ` David Hildenbrand
2018-04-29 21:08               ` Michal Hocko
2018-04-30  6:31                 ` David Hildenbrand [this message]
2018-04-20  7:30   ` David Hildenbrand
2018-04-13 13:16 ` [PATCH RFC 3/8] mm: use PG_offline in online/offlining code David Hildenbrand
2018-04-13 13:31 ` [PATCH RFC 4/8] kdump: expose PG_offline David Hildenbrand
2018-04-13 13:33 ` [PATCH RFC 5/8] mm: only mark section offline when all pages are offline David Hildenbrand
2018-04-13 13:33 ` [PATCH RFC 6/8] mm: offline_pages() is also limited by MAX_ORDER David Hildenbrand
2018-04-13 13:33 ` [PATCH RFC 7/8] mm: allow to control onlining/offlining of memory by a driver David Hildenbrand
2018-04-13 15:59   ` Michal Hocko
2018-04-13 16:32     ` David Hildenbrand
2018-04-13 13:33 ` [PATCH RFC 8/8] mm: export more functions used to online/offline memory David Hildenbrand

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=a0748efe-60e1-bc85-dbf3-a2352ab2c5b1@redhat.com \
    --to=david@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=alexander.levin@verizon.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jhogan@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mgorman@techsingularity.net \
    --cc=mhocko@kernel.org \
    --cc=miles.chen@mediatek.com \
    --cc=mingo@redhat.com \
    --cc=pasha.tatashin@oracle.com \
    --cc=riel@redhat.com \
    --cc=rostedt@goodmis.org \
    --cc=vbabka@suse.cz \
    --cc=willy@infradead.org \
    --cc=ying.huang@intel.com \
    --subject='Re: [PATCH RFC 2/8] mm: introduce PG_offline' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).