From: Jason Gunthorpe <jgg@ziepe.ca>
To: Joao Martins <joao.m.martins@oracle.com>
Cc: Alistair Popple <apopple@nvidia.com>,
Dan Williams <dan.j.williams@intel.com>,
Felix Kuehling <Felix.Kuehling@amd.com>,
linux-mm@kvack.org, Vishal Verma <vishal.l.verma@intel.com>,
Dave Jiang <dave.jiang@intel.com>,
Naoya Horiguchi <naoya.horiguchi@nec.com>,
Matthew Wilcox <willy@infradead.org>,
John Hubbard <jhubbard@nvidia.com>,
Jane Chu <jane.chu@oracle.com>,
Muchun Song <songmuchun@bytedance.com>,
Mike Kravetz <mike.kravetz@oracle.com>,
Andrew Morton <akpm@linux-foundation.org>,
Jonathan Corbet <corbet@lwn.net>, Christoph Hellwig <hch@lst.de>,
nvdimm@lists.linux.dev, linux-doc@vger.kernel.org
Subject: Re: [PATCH v4 08/14] mm/gup: grab head page refcount once for group of subpages
Date: Thu, 30 Sep 2021 18:55:24 -0300 [thread overview]
Message-ID: <20210930215524.GI3544071@ziepe.ca> (raw)
In-Reply-To: <2ab374d1-4cc1-60a4-6663-81de7d59667b@oracle.com>
On Thu, Sep 30, 2021 at 06:54:05PM +0100, Joao Martins wrote:
> On 9/30/21 04:01, Alistair Popple wrote:
> > On Thursday, 30 September 2021 5:34:05 AM AEST Jason Gunthorpe wrote:
> >> On Wed, Sep 29, 2021 at 12:50:15PM +0100, Joao Martins wrote:
> >>
> >>>> If the get_dev_pagemap has to remain then it just means we have to
> >>>> flush before changing pagemap pointers
> >>> Right -- I don't think we should need it as that discussion on the other
> >>> thread goes.
> >>>
> >>> OTOH, using @pgmap might be useful to unblock gup-fast FOLL_LONGTERM
> >>> for certain devmap types[0] (like MEMORY_DEVICE_GENERIC [device-dax]
> >>> can support it but not MEMORY_DEVICE_FSDAX [fsdax]).
> >>
> >> When looking at Logan's patches I think it is pretty clear to me that
> >> page->pgmap must never be a dangling pointer if the caller has a
> >> legitimate refcount on the page.
> >>
> >> For instance the migrate and stuff all blindly calls
> >> is_device_private_page() on the struct page expecting a valid
> >> page->pgmap.
> >>
> >> This also looks like it is happening, ie
> >>
> >> void __put_page(struct page *page)
> >> {
> >> if (is_zone_device_page(page)) {
> >> put_dev_pagemap(page->pgmap);
> >>
> >> Is indeed putting the pgmap ref back when the page becomes ungettable.
> >>
> >> This properly happens when the page refcount goes to zero and so it
> >> should fully interlock with __page_cache_add_speculative():
> >>
> >> if (unlikely(!page_ref_add_unless(page, count, 0))) {
> >>
> >> Thus, in gup.c, if we succeed at try_grab_compound_head() then
> >> page->pgmap is a stable pointer with a valid refcount.
> >>
> >> So, all the external pgmap stuff in gup.c is completely pointless.
> >> try_grab_compound_head() provides us with an equivalent protection at
> >> lower cost. Remember gup.c doesn't deref the pgmap at all.
> >>
> >> Dan/Alistair/Felix do you see any hole in that argument??
> >
> > As background note that device pages are currently considered free when
> > refcount == 1 but the pgmap reference is dropped when the refcount transitions
> > 1->0. The final pgmap reference is typically dropped when a driver calls
> > memunmap_pages() and put_page() drops the last page reference:
> >
> > void memunmap_pages(struct dev_pagemap *pgmap)
> > {
> > unsigned long pfn;
> > int i;
> >
> > dev_pagemap_kill(pgmap);
> > for (i = 0; i < pgmap->nr_range; i++)
> > for_each_device_pfn(pfn, pgmap, i)
> > put_page(pfn_to_page(pfn));
> > dev_pagemap_cleanup(pgmap);
> >
> > If there are still pgmap references dev_pagemap_cleanup(pgmap) will block until
> > the final reference is dropped. So I think your argument holds at least for
> > DEVICE_PRIVATE and DEVICE_GENERIC. DEVICE_FS_DAX defines it's own pagemap
> > cleanup but I can't see why the same argument wouldn't hold there - if a page
> > has a valid refcount it must have a reference on the pagemap too.
>
> IIUC Dan's reasoning was that fsdax wasn't able to deal with
> surprise removal [1] so his patches were to ensure fsdax (or the
> pmem block device) poisons/kills the pages as a way to notify
> filesystem/dm that the page was to be kept unmapped:
Sure, but that has nothing to do with GUP, that is between the
filesytem and fsdax
> But if fsdax doesn't wait for all the pgmap references[*] on its
> pagemap cleanup callback then what's the pgmap ref in
> __gup_device_huge() pairs/protects us up against that is specific to
> fsdax?
It does wait for refs
It sets the pgmap.ref to:
pmem->pgmap.ref = &q->q_usage_counter;
And that ref is incr'd by the struct page lifetime - the unincr is in
__put_page() above
fsdax_pagemap_ops does pmem_pagemap_kill() which calls
blk_freeze_queue_start() which does percpu_ref_kill(). Then the
pmem_pagemap_cleanup() eventually does blk_mq_freeze_queue_wait()
which will sleep until the prefcpu ref reaches zero.
In other words fsdax cannot pass cleanup while a struct page exists
with a non-zero refcount, which answers Alistair's question about how
fsdax's cleanup work.
Jason
next prev parent reply other threads:[~2021-09-30 21:55 UTC|newest]
Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-27 14:58 [PATCH v4 00/14] mm, sparse-vmemmap: Introduce compound devmaps for device-dax Joao Martins
2021-08-27 14:58 ` [PATCH v4 01/14] memory-failure: fetch compound_head after pgmap_pfn_valid() Joao Martins
2021-08-27 14:58 ` [PATCH v4 02/14] mm/page_alloc: split prep_compound_page into head and tail subparts Joao Martins
2021-08-27 14:58 ` [PATCH v4 03/14] mm/page_alloc: refactor memmap_init_zone_device() page init Joao Martins
2021-08-27 14:58 ` [PATCH v4 04/14] mm/memremap: add ZONE_DEVICE support for compound pages Joao Martins
2021-08-27 15:33 ` Christoph Hellwig
2021-08-27 16:00 ` Joao Martins
2021-09-01 9:44 ` Christoph Hellwig
2021-09-09 9:38 ` Joao Martins
2021-08-27 14:58 ` [PATCH v4 05/14] device-dax: use ALIGN() for determining pgoff Joao Martins
2021-08-27 14:58 ` [PATCH v4 06/14] device-dax: ensure dev_dax->pgmap is valid for dynamic devices Joao Martins
2021-11-05 0:31 ` Dan Williams
2021-11-05 12:09 ` Joao Martins
2021-11-05 16:14 ` Joao Martins
2021-11-05 16:46 ` Dan Williams
2021-11-05 18:11 ` Joao Martins
2021-08-27 14:58 ` [PATCH v4 07/14] device-dax: compound devmap support Joao Martins
2021-11-05 0:38 ` Dan Williams
2021-11-05 14:10 ` Joao Martins
2021-11-05 16:41 ` Dan Williams
2021-08-27 14:58 ` [PATCH v4 08/14] mm/gup: grab head page refcount once for group of subpages Joao Martins
2021-08-27 16:25 ` Jason Gunthorpe
2021-08-27 18:34 ` Joao Martins
2021-08-30 13:07 ` Jason Gunthorpe
2021-08-31 12:34 ` Joao Martins
2021-08-31 17:05 ` Jason Gunthorpe
2021-09-23 16:51 ` Joao Martins
2021-09-28 18:01 ` Jason Gunthorpe
2021-09-29 11:50 ` Joao Martins
2021-09-29 19:34 ` Jason Gunthorpe
2021-09-30 3:01 ` Alistair Popple
2021-09-30 17:54 ` Joao Martins
2021-09-30 21:55 ` Jason Gunthorpe [this message]
2021-10-18 18:36 ` Jason Gunthorpe
2021-10-18 18:37 ` Jason Gunthorpe
2021-10-08 11:54 ` Jason Gunthorpe
2021-10-11 15:53 ` Joao Martins
2021-10-13 17:41 ` Jason Gunthorpe
2021-10-13 19:18 ` Joao Martins
2021-10-13 19:43 ` Jason Gunthorpe
2021-10-14 17:56 ` Joao Martins
2021-10-14 18:06 ` Jason Gunthorpe
2021-08-27 14:58 ` [PATCH v4 09/14] mm/sparse-vmemmap: add a pgmap argument to section activation Joao Martins
2021-08-27 14:58 ` [PATCH v4 10/14] mm/sparse-vmemmap: refactor core of vmemmap_populate_basepages() to helper Joao Martins
2021-08-27 14:58 ` [PATCH v4 11/14] mm/hugetlb_vmemmap: move comment block to Documentation/vm Joao Martins
2021-08-27 14:58 ` [PATCH v4 12/14] mm/sparse-vmemmap: populate compound devmaps Joao Martins
2021-08-27 14:58 ` [PATCH v4 13/14] mm/page_alloc: reuse tail struct pages for " Joao Martins
2021-08-27 14:58 ` [PATCH v4 14/14] mm/sparse-vmemmap: improve memory savings for compound pud geometry Joao Martins
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=20210930215524.GI3544071@ziepe.ca \
--to=jgg@ziepe.ca \
--cc=Felix.Kuehling@amd.com \
--cc=akpm@linux-foundation.org \
--cc=apopple@nvidia.com \
--cc=corbet@lwn.net \
--cc=dan.j.williams@intel.com \
--cc=dave.jiang@intel.com \
--cc=hch@lst.de \
--cc=jane.chu@oracle.com \
--cc=jhubbard@nvidia.com \
--cc=joao.m.martins@oracle.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mike.kravetz@oracle.com \
--cc=naoya.horiguchi@nec.com \
--cc=nvdimm@lists.linux.dev \
--cc=songmuchun@bytedance.com \
--cc=vishal.l.verma@intel.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).