[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <99ed92b7-c1b2-4e12-a7ee-776a7f890b47@redhat.com>
Date: Thu, 8 May 2025 17:03:46 +0200
From: David Hildenbrand <david@...hat.com>
To: Pantelis Antoniou <p.antoniou@...tner.samsung.com>,
Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org
Cc: linux-kernel@...r.kernel.org, Artem Krupotkin <artem.k@...sung.com>,
Charles Briere <c.briere@...sung.com>,
Wade Farnsworth <wade.farnsworth@...mens.com>, Peter Xu <peterx@...hat.com>
Subject: Re: [PATCH 1/1] Fix zero copy I/O on __get_user_pages allocated pages
On 07.05.25 17:41, Pantelis Antoniou wrote:
Hi,
> Recent updates to net filesystems enabled zero copy operations,
> which require getting a user space page pinned.
>
> This does not work for pages that were allocated via __get_user_pages
> and then mapped to user-space via remap_pfn_rage.
Right. Because the struct page of a VM_PFNMAP *must not be touched*. It
has to be treated like it doesn't exist.
>
> remap_pfn_range_internal() will turn on VM_IO | VM_PFNMAP vma bits.
> VM_PFNMAP in particular mark the pages as not having struct_page
> associated with them, which is not the case for __get_user_pages()
>
> This in turn makes any attempt to lock a page fail, and breaking
> I/O from that address range.
>
> This patch address it by special casing pages in those VMAs and not
> calling vm_normal_page() for them.
>
> Signed-off-by: Pantelis Antoniou <p.antoniou@...tner.samsung.com>
> ---
> mm/gup.c | 22 ++++++++++++++++++----
> 1 file changed, 18 insertions(+), 4 deletions(-)
>
> diff --git a/mm/gup.c b/mm/gup.c
> index 84461d384ae2..e185c18c0c81 100644
> --- a/mm/gup.c
> +++ b/mm/gup.c
> @@ -833,6 +833,20 @@ static inline bool can_follow_write_pte(pte_t pte, struct page *page,
> return !userfaultfd_pte_wp(vma, pte);
> }
>
> +static struct page *gup_normal_page(struct vm_area_struct *vma,
> + unsigned long address, pte_t pte)
> +{
> + unsigned long pfn;
> +
> + if (vma->vm_flags & (VM_MIXEDMAP | VM_PFNMAP)) {
> + pfn = pte_pfn(pte);
> + if (!pfn_valid(pfn) || is_zero_pfn(pfn) || pfn > highest_memmap_pfn)
> + return NULL;
> + return pfn_to_page(pfn);
> + }
> + return vm_normal_page(vma, address, pte);
I enjoy seeing vm_normal_page() checks in GUP code.
I don't enjoy seeing what you added before that :)
If vm_normal_page() tells you "this is not a normal", then we should not
touch it. There is one exception: the shared zeropage.
So, unfortunately, this is wrong.
--
Cheers,
David / dhildenb
Powered by blists - more mailing lists