From: Pasha Tatashin <Pavel.Tatashin@microsoft.com>
To: David Hildenbrand <david@redhat.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Cc: "linux-mm@kvack.org" <linux-mm@kvack.org>,
Andrew Morton <akpm@linux-foundation.org>,
Michal Hocko <mhocko@suse.com>, Vlastimil Babka <vbabka@suse.cz>,
Stephen Rothwell <sfr@canb.auug.org.au>,
Pavel Tatashin <pasha.tatashin@oracle.com>,
Kemi Wang <kemi.wang@intel.com>,
David Rientjes <rientjes@google.com>,
Jia He <jia.he@hxt-semitech.com>,
Oscar Salvador <osalvador@suse.de>,
Petr Tesarik <ptesarik@suse.com>,
Andrey Ryabinin <aryabinin@virtuozzo.com>,
Dan Williams <dan.j.williams@intel.com>,
Mathieu Malaterre <malat@debian.org>, Baoquan He <bhe@redhat.com>,
Wei Yang <richard.weiyang@gmail.com>,
Ross Zwisler <zwisler@kernel.org>,
"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Subject: Re: [PATCH v1 2/5] mm/memory_hotplug: enforce section alignment when onlining/offlining
Date: Thu, 30 Aug 2018 22:14:01 +0000 [thread overview]
Message-ID: <772774b8-77d8-c09b-f933-5ce29be58fa9@microsoft.com> (raw)
In-Reply-To: <20180816100628.26428-3-david@redhat.com>
Hi David,
I am not sure this is needed, because we already have a stricter checker:
check_hotplug_memory_range()
You could call it from online_pages(), if you think there is a reason to
do it, but other than that it is done from add_memory_resource() and
from remove_memory().
Thank you,
Pavel
On 8/16/18 6:06 AM, David Hildenbrand wrote:
> onlining/offlining code works on whole sections, so let's enforce that.
> Existing code only allows to add memory in memory block size. And only
> whole memory blocks can be onlined/offlined. Memory blocks are always
> aligned to sections, so this should not break anything.
>
> online_pages/offline_pages will implicitly mark whole sections
> online/offline, so the code really can only handle such granularities.
>
> (especially offlining code cannot deal with pageblock_nr_pages but
> theoretically only MAX_ORDER-1)
>
> Signed-off-by: David Hildenbrand <david@redhat.com>
> ---
> mm/memory_hotplug.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index 090cf474de87..30d2fa42b0bb 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -897,6 +897,11 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ
> struct memory_notify arg;
> struct memory_block *mem;
>
> + if (!IS_ALIGNED(pfn, PAGES_PER_SECTION))
> + return -EINVAL;
> + if (!IS_ALIGNED(nr_pages, PAGES_PER_SECTION))
> + return -EINVAL;
> +
> /*
> * We can't use pfn_to_nid() because nid might be stored in struct page
> * which is not yet initialized. Instead, we find nid from memory block.
> @@ -1600,10 +1605,9 @@ int offline_pages(unsigned long start_pfn, unsigned long nr_pages)
> struct zone *zone;
> struct memory_notify arg;
>
> - /* at least, alignment against pageblock is necessary */
> - if (!IS_ALIGNED(start_pfn, pageblock_nr_pages))
> + if (!IS_ALIGNED(start_pfn, PAGES_PER_SECTION))
> return -EINVAL;
> - if (!IS_ALIGNED(end_pfn, pageblock_nr_pages))
> + if (!IS_ALIGNED(nr_pages, PAGES_PER_SECTION))
> return -EINVAL;
> /* This makes hotplug much easier...and readable.
> we assume this for now. .*/
>
next prev parent reply other threads:[~2018-08-30 22:14 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-08-16 10:06 [PATCH v1 0/5] mm/memory_hotplug: online/offline_pages refactorings David Hildenbrand
2018-08-16 10:06 ` [PATCH v1 1/5] mm/memory_hotplug: drop intermediate __offline_pages David Hildenbrand
2018-08-16 11:44 ` Stephen Rothwell
2018-08-16 12:08 ` David Hildenbrand
2018-08-16 18:50 ` kbuild test robot
2018-08-30 20:17 ` Pasha Tatashin
2018-08-30 20:20 ` Pasha Tatashin
2018-08-16 10:06 ` [PATCH v1 2/5] mm/memory_hotplug: enforce section alignment when onlining/offlining David Hildenbrand
2018-08-16 12:34 ` Oscar Salvador
2018-08-30 22:14 ` Pasha Tatashin [this message]
2018-08-31 7:51 ` David Hildenbrand
2018-08-16 10:06 ` [PATCH v1 3/5] mm/memory_hotplug: check if sections are already online/offline David Hildenbrand
2018-08-16 10:47 ` Oscar Salvador
2018-08-16 11:00 ` David Hildenbrand
2018-08-30 22:17 ` Pasha Tatashin
2018-08-16 10:06 ` [PATCH v1 4/5] mm/memory_hotplug: onlining pages can only fail due to notifiers David Hildenbrand
2018-08-30 22:30 ` Pasha Tatashin
2018-08-16 10:06 ` [PATCH v1 5/5] mm/memory_hotplug: print only with DEBUG_VM in online/offline_pages() David Hildenbrand
2018-08-17 8:18 ` Oscar Salvador
2018-08-19 12:34 ` Wei Yang
2018-08-20 9:57 ` David Hildenbrand
2018-08-20 13:12 ` Wei Yang
2018-08-20 9:45 ` David Hildenbrand
2018-08-20 9:49 ` David Hildenbrand
2018-08-20 10:46 ` David Hildenbrand
2018-08-30 22:36 ` Pasha Tatashin
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=772774b8-77d8-c09b-f933-5ce29be58fa9@microsoft.com \
--to=pavel.tatashin@microsoft.com \
--cc=akpm@linux-foundation.org \
--cc=aryabinin@virtuozzo.com \
--cc=bhe@redhat.com \
--cc=dan.j.williams@intel.com \
--cc=david@redhat.com \
--cc=jia.he@hxt-semitech.com \
--cc=kemi.wang@intel.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=malat@debian.org \
--cc=mhocko@suse.com \
--cc=osalvador@suse.de \
--cc=pasha.tatashin@oracle.com \
--cc=ptesarik@suse.com \
--cc=richard.weiyang@gmail.com \
--cc=rientjes@google.com \
--cc=sfr@canb.auug.org.au \
--cc=vbabka@suse.cz \
--cc=zwisler@kernel.org \
/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
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox