From: Felix Kuehling <felix.kuehling@amd.com>
To: Jason Gunthorpe <jgg@nvidia.com>
Cc: Christoph Hellwig <hch@lst.de>,
David Hildenbrand <david@redhat.com>,
Alex Sierra <alex.sierra@amd.com>,
akpm@linux-foundation.org, linux-mm@kvack.org,
rcampbell@nvidia.com, linux-ext4@vger.kernel.org,
linux-xfs@vger.kernel.org, amd-gfx@lists.freedesktop.org,
dri-devel@lists.freedesktop.org, jglisse@redhat.com,
apopple@nvidia.com, willy@infradead.org
Subject: Re: [PATCH v6 01/10] mm: add zone device coherent type memory support
Date: Tue, 15 Feb 2022 17:49:07 -0500 [thread overview]
Message-ID: <002ad572-4d32-7133-06f3-aa680c297be2@amd.com> (raw)
In-Reply-To: <20220215214749.GA4160@nvidia.com>
On 2022-02-15 16:47, Jason Gunthorpe wrote:
> On Tue, Feb 15, 2022 at 04:35:56PM -0500, Felix Kuehling wrote:
>> On 2022-02-15 14:41, Jason Gunthorpe wrote:
>>> On Tue, Feb 15, 2022 at 07:32:09PM +0100, Christoph Hellwig wrote:
>>>> On Tue, Feb 15, 2022 at 10:45:24AM -0400, Jason Gunthorpe wrote:
>>>>>> Do you know if DEVICE_GENERIC pages would end up as PageAnon()? My
>>>>>> assumption was that they would be part of a special mapping.
>>>>> We need to stop using the special PTEs and VMAs for things that have a
>>>>> struct page. This is a mistake DAX created that must be undone.
>>>> Yes, we'll get to it. Maybe we can do it for the non-DAX devmap
>>>> ptes first given that DAX is more complicated.
>>> Probably, I think we can check the page->pgmap type to tell the
>>> difference.
>>>
>>> I'm not sure how the DEVICE_GENERIC can work without this, as DAX was
>>> made safe by using the unmap_mapping_range(), which won't work
>>> here. Is there some other trick being used to keep track of references
>>> inside the AMD driver?
>> Not sure I'm following all the discussion about VMAs and DAX. So I may be
>> answering the wrong question: We treat each ZONE_DEVICE page as a reference
>> to the BO (buffer object) that backs the page. We increment the BO refcount
>> for each page we migrate into it. In the dev_pagemap_ops.page_free callback
>> we drop that reference. Once all pages backed by a BO are freed, the BO
>> refcount reaches 0 [*] and we can free the BO allocation.
> Userspace does
> 1) mmap(MAP_PRIVATE) to allocate anon memory
> 2) something to trigger migration to install a ZONE_DEVICE page
> 3) munmap()
>
> Who decrements the refcout on the munmap?
>
> When a ZONE_DEVICE page is installed in the PTE is supposed to be
> marked as pte_devmap and that disables all the normal page refcounting
> during munmap().
>
> fsdax makes this work by working the refcounts backwards, the page is
> refcounted while it exists in the driver, when the driver decides to
> remove it then unmap_mapping_range() is called to purge it from all
> PTEs and then refcount is decrd. munmap/fork/etc don't change the
> refcount.
Hmm, that just means, whether or not there are PTEs doesn't really
matter. It should work the same as it does for DEVICE_PRIVATE pages. I'm
not sure where DEVICE_PRIVATE page's refcounts are decremented on unmap,
TBH. But I can't find it in our driver, or in the test_hmm driver for
that matter.
Regards,
Felix
>
> Jason
next prev parent reply other threads:[~2022-02-15 22:49 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-01 15:48 [PATCH v6 00/10] Add MEMORY_DEVICE_COHERENT for coherent device memory mapping Alex Sierra
2022-02-01 15:48 ` [PATCH v6 01/10] mm: add zone device coherent type memory support Alex Sierra
2022-02-11 16:15 ` David Hildenbrand
2022-02-11 16:39 ` David Hildenbrand
2022-02-11 16:52 ` Sierra Guiza, Alejandro (Alex)
2022-02-11 17:07 ` Felix Kuehling
2022-02-15 12:16 ` David Hildenbrand
2022-02-15 14:45 ` Jason Gunthorpe
2022-02-15 18:32 ` Christoph Hellwig
2022-02-15 19:41 ` Jason Gunthorpe
2022-02-15 21:35 ` Felix Kuehling
2022-02-15 21:47 ` Jason Gunthorpe
2022-02-15 22:49 ` Felix Kuehling [this message]
2022-02-16 2:01 ` Jason Gunthorpe
2022-02-16 16:56 ` Felix Kuehling
2022-02-16 17:28 ` Jason Gunthorpe
2022-02-16 1:23 ` Alistair Popple
2022-02-16 2:03 ` Jason Gunthorpe
2022-02-16 2:36 ` Alistair Popple
2022-02-16 8:31 ` David Hildenbrand
2022-02-16 12:26 ` Jason Gunthorpe
2022-02-17 1:05 ` Alistair Popple
2022-02-17 21:12 ` Felix Kuehling
2022-02-18 0:19 ` Jason Gunthorpe
2022-02-18 19:20 ` Felix Kuehling
2022-02-18 19:26 ` Jason Gunthorpe
2022-02-18 19:37 ` Felix Kuehling
2022-02-28 20:34 ` [PATCH] mm: split vm_normal_pages for LRU and non-LRU handling Alex Sierra
2022-02-28 22:41 ` Felix Kuehling
2022-03-01 8:03 ` David Hildenbrand
2022-03-01 16:08 ` Felix Kuehling
2022-03-01 16:22 ` David Hildenbrand
2022-03-01 16:30 ` Felix Kuehling
2022-03-01 16:32 ` David Hildenbrand
2022-02-18 0:59 ` [PATCH v6 01/10] mm: add zone device coherent type memory support Alistair Popple
2022-02-11 16:45 ` Jason Gunthorpe
2022-02-11 16:49 ` David Hildenbrand
2022-02-11 16:56 ` Jason Gunthorpe
2022-02-15 12:15 ` David Hildenbrand
2022-02-15 18:52 ` Felix Kuehling
2022-02-11 17:05 ` Felix Kuehling
2022-02-14 2:04 ` Alistair Popple
2022-02-01 15:48 ` [PATCH v6 02/10] mm: add device coherent vma selection for memory migration Alex Sierra
2022-02-01 15:48 ` [PATCH v6 03/10] mm/gup: fail get_user_pages for LONGTERM dev coherent type Alex Sierra
2022-02-01 15:48 ` [PATCH v6 04/10] drm/amdkfd: add SPM support for SVM Alex Sierra
2022-02-01 15:48 ` [PATCH v6 05/10] drm/amdkfd: coherent type as sys mem on migration to ram Alex Sierra
2022-02-01 15:48 ` [PATCH v6 06/10] lib: test_hmm add ioctl to get zone device type Alex Sierra
2022-02-01 15:48 ` [PATCH v6 07/10] lib: test_hmm add module param for " Alex Sierra
2022-02-01 15:48 ` [PATCH v6 08/10] lib: add support for device coherent type in test_hmm Alex Sierra
2022-02-01 15:49 ` [PATCH v6 09/10] tools: update hmm-test to support device coherent type Alex Sierra
2022-02-01 15:49 ` [PATCH v6 10/10] tools: update test_hmm script to support SP config Alex Sierra
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=002ad572-4d32-7133-06f3-aa680c297be2@amd.com \
--to=felix.kuehling@amd.com \
--cc=akpm@linux-foundation.org \
--cc=alex.sierra@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=apopple@nvidia.com \
--cc=david@redhat.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=hch@lst.de \
--cc=jgg@nvidia.com \
--cc=jglisse@redhat.com \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-xfs@vger.kernel.org \
--cc=rcampbell@nvidia.com \
--cc=willy@infradead.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;
as well as URLs for NNTP newsgroup(s).