From: Jason Gunthorpe <jgg@ziepe•ca>
To: Matthew Wilcox <willy@infradead•org>
Cc: Miaohe Lin <linmiaohe@huawei•com>,
David Hildenbrand <david@redhat•com>,
Peter Zijlstra <peterz@infradead•org>,
Yang Shi <shy828301@gmail•com>,
Qi Zheng <zhengqi.arch@bytedance•com>, Song Liu <song@kernel•org>,
sparclinux@vger•kernel.org,
Alexander Gordeev <agordeev@linux•ibm.com>,
Claudio Imbrenda <imbrenda@linux•ibm.com>,
Will Deacon <will@kernel•org>,
linux-s390@vger•kernel.org, Yu Zhao <yuzhao@google•com>,
Ira Weiny <ira.weiny@intel•com>,
Alistair Popple <apopple@nvidia•com>,
Hugh Dickins <hughd@google•com>,
Russell King <linux@armlinux•org.uk>,
Steven Price <steven.price@arm•com>,
Christoph Hellwig <hch@infradead•org>,
"Aneesh Kumar K.V" <aneesh.kumar@linux•ibm.com>,
Axel Rasmussen <axelrasmussen@google•com>,
Christian Borntraeger <borntraeger@linux•ibm.com>,
Thomas Hellstrom <thomas.hellstrom@linux•intel.com>,
Ralph Campbell <rcampbell@nvidia•com>,
Pasha Tatashin <pasha.tatashin@soleen•com>,
Anshuman Khandual <anshuman.khandual@arm•com>,
Heiko Carstens <hca@linux•ibm.com>, Peter Xu <peterx@redhat•com>,
Suren Baghdasaryan <surenb@google•com>,
linux-arm-kernel@lists•infradead.org,
SeongJae Park <sj@kernel•org>, Jann Horn <jannh@google•com>,
linux-mm@kvack•org, linuxppc-dev@lists•ozlabs.org,
"Kirill A. Shutemov" <kirill.shutemov@linux•intel.com>,
Naoya Horiguchi <naoya.horiguchi@nec•com>,
linux-kernel@vger•kernel.org, Minchan Kim <minchan@kernel•org>,
Mike Rapoport <rppt@kernel•org>,
Andrew Morton <akpm@linux-foundation•org>,
Mel Gorman <mgorman@techsingularity•net>,
"David S. Miller" <davem@davemloft•net>,
Zack Rusin <zackr@vmware•com>,
Mike Kravetz <mike.kravetz@oracle•com>
Subject: Re: [PATCH 05/12] powerpc: add pte_free_defer() for pgtables sharing page
Date: Fri, 2 Jun 2023 11:20:15 -0300 [thread overview]
Message-ID: <ZHn6n5eVTsr4Wl8x@ziepe.ca> (raw)
In-Reply-To: <ZHSwWgLWaEd+zi/g@casper.infradead.org>
On Mon, May 29, 2023 at 03:02:02PM +0100, Matthew Wilcox wrote:
> On Sun, May 28, 2023 at 11:20:21PM -0700, Hugh Dickins wrote:
> > +void pte_free_defer(struct mm_struct *mm, pgtable_t pgtable)
> > +{
> > + struct page *page;
> > +
> > + page = virt_to_page(pgtable);
> > + call_rcu(&page->rcu_head, pte_free_now);
> > +}
>
> This can't be safe (on ppc). IIRC you might have up to 16x4k page
> tables sharing one 64kB page. So if you have two page tables from the
> same page being defer-freed simultaneously, you'll reuse the rcu_head
> and I cannot imagine things go well from that point.
>
> I have no idea how to solve this problem.
Maybe power and s390 should allocate a side structure, sort of a
pre-memdesc thing to store enough extra data?
If we can get enough bytes then something like this would let a single
rcu head be shared to manage the free bits.
struct 64k_page {
u8 free_pages;
u8 pending_rcu_free_pages;
struct rcu_head head;
}
free_sub_page(sub_id)
if (atomic_fetch_or(1 << sub_id, &64k_page->pending_rcu_free_pages))
call_rcu(&64k_page->head)
rcu_func()
64k_page->free_pages |= atomic_xchg(0, &64k_page->pending_rcu_free_pages)
if (64k_pages->free_pages == all_ones)
free_pgea(64k_page);
Jason
next prev parent reply other threads:[~2023-06-02 14:21 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-29 6:11 [PATCH 00/12] mm: free retracted page table by RCU Hugh Dickins
2023-05-29 6:14 ` [PATCH 01/12] mm/pgtable: add rcu_read_lock() and rcu_read_unlock()s Hugh Dickins
2023-05-31 17:06 ` Jann Horn
2023-06-02 2:50 ` Hugh Dickins
2023-06-02 14:21 ` Jann Horn
2023-05-29 6:16 ` [PATCH 02/12] mm/pgtable: add PAE safety to __pte_offset_map() Hugh Dickins
2023-05-29 13:56 ` Matthew Wilcox
[not found] ` <ZHeg3oRljRn6wlLX@ziepe.ca>
2023-06-02 5:35 ` Hugh Dickins
2023-05-29 6:17 ` [PATCH 03/12] arm: adjust_pte() use pte_offset_map_nolock() Hugh Dickins
2023-05-29 6:18 ` [PATCH 04/12] powerpc: assert_pte_locked() " Hugh Dickins
2023-05-29 6:20 ` [PATCH 05/12] powerpc: add pte_free_defer() for pgtables sharing page Hugh Dickins
2023-05-29 14:02 ` Matthew Wilcox
2023-05-29 14:36 ` Hugh Dickins
2023-06-01 13:57 ` Gerald Schaefer
2023-06-02 6:38 ` Hugh Dickins
2023-06-02 14:20 ` Jason Gunthorpe [this message]
2023-06-06 3:40 ` Hugh Dickins
2023-06-06 18:23 ` Jason Gunthorpe
2023-06-06 19:03 ` Peter Xu
2023-06-06 19:08 ` Jason Gunthorpe
2023-06-07 3:49 ` Hugh Dickins
2023-05-29 6:21 ` [PATCH 06/12] sparc: " Hugh Dickins
2023-06-06 3:46 ` Hugh Dickins
2023-05-29 6:22 ` [PATCH 07/12] s390: add pte_free_defer(), with use of mmdrop_async() Hugh Dickins
2023-06-06 5:11 ` Hugh Dickins
2023-06-06 18:39 ` Jason Gunthorpe
2023-06-08 2:46 ` Hugh Dickins
2023-06-06 19:40 ` Gerald Schaefer
2023-06-08 3:35 ` Hugh Dickins
2023-06-08 13:58 ` Jason Gunthorpe
2023-06-08 15:47 ` Gerald Schaefer
2023-05-29 6:23 ` [PATCH 08/12] mm/pgtable: add pte_free_defer() for pgtable as page Hugh Dickins
2023-06-01 13:31 ` Jann Horn
[not found] ` <ZHekpAKJ05cr/GLl@ziepe.ca>
2023-06-02 6:03 ` Hugh Dickins
2023-06-02 12:15 ` Jason Gunthorpe
2023-05-29 6:25 ` [PATCH 09/12] mm/khugepaged: retract_page_tables() without mmap or vma lock Hugh Dickins
2023-05-29 23:26 ` Peter Xu
2023-05-31 0:38 ` Hugh Dickins
2023-05-31 15:34 ` Jann Horn
[not found] ` <ZHe0A079X9B8jWlH@x1n>
2023-05-31 22:18 ` Jann Horn
2023-06-01 14:06 ` Jason Gunthorpe
2023-06-06 6:18 ` Hugh Dickins
2023-05-29 6:26 ` [PATCH 10/12] mm/khugepaged: collapse_pte_mapped_thp() with mmap_read_lock() Hugh Dickins
2023-05-31 17:25 ` Jann Horn
2023-06-02 5:11 ` Hugh Dickins
2023-05-29 6:28 ` [PATCH 11/12] mm/khugepaged: delete khugepaged_collapse_pte_mapped_thps() Hugh Dickins
2023-05-29 6:30 ` [PATCH 12/12] mm: delete mmap_write_trylock() and vma_try_start_write() Hugh Dickins
[not found] ` <CAG48ez0pCqfRdVSnJz7EKtNvMR65=zJgVB-72nTdrNuhtJNX2Q@mail.gmail.com>
2023-06-02 4:37 ` [PATCH 00/12] mm: free retracted page table by RCU Hugh Dickins
2023-06-02 15:26 ` Jann Horn
2023-06-06 6:28 ` Hugh Dickins
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=ZHn6n5eVTsr4Wl8x@ziepe.ca \
--to=jgg@ziepe$(echo .)ca \
--cc=agordeev@linux$(echo .)ibm.com \
--cc=akpm@linux-foundation$(echo .)org \
--cc=aneesh.kumar@linux$(echo .)ibm.com \
--cc=anshuman.khandual@arm$(echo .)com \
--cc=apopple@nvidia$(echo .)com \
--cc=axelrasmussen@google$(echo .)com \
--cc=borntraeger@linux$(echo .)ibm.com \
--cc=davem@davemloft$(echo .)net \
--cc=david@redhat$(echo .)com \
--cc=hca@linux$(echo .)ibm.com \
--cc=hch@infradead$(echo .)org \
--cc=hughd@google$(echo .)com \
--cc=imbrenda@linux$(echo .)ibm.com \
--cc=ira.weiny@intel$(echo .)com \
--cc=jannh@google$(echo .)com \
--cc=kirill.shutemov@linux$(echo .)intel.com \
--cc=linmiaohe@huawei$(echo .)com \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
--cc=linux-kernel@vger$(echo .)kernel.org \
--cc=linux-mm@kvack$(echo .)org \
--cc=linux-s390@vger$(echo .)kernel.org \
--cc=linux@armlinux$(echo .)org.uk \
--cc=linuxppc-dev@lists$(echo .)ozlabs.org \
--cc=mgorman@techsingularity$(echo .)net \
--cc=mike.kravetz@oracle$(echo .)com \
--cc=minchan@kernel$(echo .)org \
--cc=naoya.horiguchi@nec$(echo .)com \
--cc=pasha.tatashin@soleen$(echo .)com \
--cc=peterx@redhat$(echo .)com \
--cc=peterz@infradead$(echo .)org \
--cc=rcampbell@nvidia$(echo .)com \
--cc=rppt@kernel$(echo .)org \
--cc=shy828301@gmail$(echo .)com \
--cc=sj@kernel$(echo .)org \
--cc=song@kernel$(echo .)org \
--cc=sparclinux@vger$(echo .)kernel.org \
--cc=steven.price@arm$(echo .)com \
--cc=surenb@google$(echo .)com \
--cc=thomas.hellstrom@linux$(echo .)intel.com \
--cc=will@kernel$(echo .)org \
--cc=willy@infradead$(echo .)org \
--cc=yuzhao@google$(echo .)com \
--cc=zackr@vmware$(echo .)com \
--cc=zhengqi.arch@bytedance$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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