From: John Hubbard <jhubbard@nvidia.com>
To: Logan Gunthorpe <logang@deltatee.com>,
linux-kernel@vger.kernel.org, linux-nvme@lists.infradead.org,
linux-block@vger.kernel.org, linux-pci@vger.kernel.org,
linux-mm@kvack.org
Cc: "Christoph Hellwig" <hch@lst.de>,
"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
"Dan Williams" <dan.j.williams@intel.com>,
"Jason Gunthorpe" <jgg@ziepe.ca>,
"Christian König" <christian.koenig@amd.com>,
"Don Dutile" <ddutile@redhat.com>,
"Matthew Wilcox" <willy@infradead.org>,
"Daniel Vetter" <daniel.vetter@ffwll.ch>,
"Minturn Dave B" <dave.b.minturn@intel.com>,
"Jason Ekstrand" <jason@jlekstrand.net>,
"Dave Hansen" <dave.hansen@linux.intel.com>,
"Xiong Jianxin" <jianxin.xiong@intel.com>,
"Bjorn Helgaas" <helgaas@kernel.org>,
"Ira Weiny" <ira.weiny@intel.com>,
"Robin Murphy" <robin.murphy@arm.com>,
"Martin Oliveira" <martin.oliveira@eideticom.com>,
"Chaitanya Kulkarni" <ckulkarnilinux@gmail.com>,
"Ralph Campbell" <rcampbell@nvidia.com>,
"Stephen Bates" <sbates@raithlin.com>
Subject: Re: [PATCH v9 2/8] iov_iter: introduce iov_iter_get_pages_[alloc_]flags()
Date: Mon, 5 Sep 2022 16:21:13 -0700 [thread overview]
Message-ID: <24ee706f-b70c-28d0-15a7-13d0dc9254bb@nvidia.com> (raw)
In-Reply-To: <20220825152425.6296-3-logang@deltatee.com>
On 8/25/22 08:24, Logan Gunthorpe wrote:
> Add iov_iter_get_pages_flags() and iov_iter_get_pages_alloc_flags()
> which take a flags argument that is passed to get_user_pages_fast().
>
> This is so that FOLL_PCI_P2PDMA can be passed when appropriate.
>
> Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
> ---
> include/linux/uio.h | 6 ++++++
> lib/iov_iter.c | 40 ++++++++++++++++++++++++++++++++++++----
> 2 files changed, 42 insertions(+), 4 deletions(-)
>
Yes.
I also don't object to Christoph's additional request to refactor and
rename around iov_iter_get_pages2() and such, but this version here is
perfectly good, so please feel free to add:
Reviewed-by: John Hubbard <jhubbard@nvidia.com>
Looking ahead, interestingly enough, it turns out that this approach to
the wrappers is really pretty close to what I posted in patch 4/7 of my
"convert most filesystems to pin_user_pages_fast()" series [1]. Instead
of passing gup_flags through, I tried to avoid that (because FOLL_PIN
is, as a mild design preference, supposed to remain internal to gup.c),
but given that you need to do it, I think I can just build on top of
your approach, and pass in FOLL_PIN via your new gup_flags arg.
Along those lines, I've copied you in on "New topic branch for block +
gup work?", let's see what happens over there.
[1] https://lore.kernel.org/r/20220831041843.973026-1-jhubbard@nvidia.com
thanks,
--
John Hubbard
NVIDIA
> diff --git a/include/linux/uio.h b/include/linux/uio.h
> index 5896af36199c..76ba69edb8c5 100644
> --- a/include/linux/uio.h
> +++ b/include/linux/uio.h
> @@ -247,8 +247,14 @@ void iov_iter_pipe(struct iov_iter *i, unsigned int direction, struct pipe_inode
> void iov_iter_discard(struct iov_iter *i, unsigned int direction, size_t count);
> void iov_iter_xarray(struct iov_iter *i, unsigned int direction, struct xarray *xarray,
> loff_t start, size_t count);
> +ssize_t iov_iter_get_pages_flags(struct iov_iter *i, struct page **pages,
> + size_t maxsize, unsigned maxpages, size_t *start,
> + unsigned int gup_flags);
> ssize_t iov_iter_get_pages2(struct iov_iter *i, struct page **pages,
> size_t maxsize, unsigned maxpages, size_t *start);
> +ssize_t iov_iter_get_pages_alloc_flags(struct iov_iter *i,
> + struct page ***pages, size_t maxsize, size_t *start,
> + unsigned int gup_flags);
> ssize_t iov_iter_get_pages_alloc2(struct iov_iter *i, struct page ***pages,
> size_t maxsize, size_t *start);
> int iov_iter_npages(const struct iov_iter *i, int maxpages);
> diff --git a/lib/iov_iter.c b/lib/iov_iter.c
> index 4b7fce72e3e5..dedb78f3c655 100644
> --- a/lib/iov_iter.c
> +++ b/lib/iov_iter.c
> @@ -1427,7 +1427,8 @@ static struct page *first_bvec_segment(const struct iov_iter *i,
>
> static ssize_t __iov_iter_get_pages_alloc(struct iov_iter *i,
> struct page ***pages, size_t maxsize,
> - unsigned int maxpages, size_t *start)
> + unsigned int maxpages, size_t *start,
> + unsigned int gup_flags)
> {
> unsigned int n;
>
> @@ -1439,7 +1440,6 @@ static ssize_t __iov_iter_get_pages_alloc(struct iov_iter *i,
> maxsize = MAX_RW_COUNT;
>
> if (likely(user_backed_iter(i))) {
> - unsigned int gup_flags = 0;
> unsigned long addr;
> int res;
>
> @@ -1497,10 +1497,24 @@ ssize_t iov_iter_get_pages2(struct iov_iter *i,
> return 0;
> BUG_ON(!pages);
>
> - return __iov_iter_get_pages_alloc(i, &pages, maxsize, maxpages, start);
> + return __iov_iter_get_pages_alloc(i, &pages, maxsize, maxpages,
> + start, 0);
> }
> EXPORT_SYMBOL(iov_iter_get_pages2);
>
> +ssize_t iov_iter_get_pages_flags(struct iov_iter *i, struct page **pages,
> + size_t maxsize, unsigned maxpages, size_t *start,
> + unsigned int gup_flags)
> +{
> + if (!maxpages)
> + return 0;
> + BUG_ON(!pages);
> +
> + return __iov_iter_get_pages_alloc(i, &pages, maxsize, maxpages,
> + start, gup_flags);
> +}
> +EXPORT_SYMBOL_GPL(iov_iter_get_pages_flags);
> +
> ssize_t iov_iter_get_pages_alloc2(struct iov_iter *i,
> struct page ***pages, size_t maxsize,
> size_t *start)
> @@ -1509,7 +1523,7 @@ ssize_t iov_iter_get_pages_alloc2(struct iov_iter *i,
>
> *pages = NULL;
>
> - len = __iov_iter_get_pages_alloc(i, pages, maxsize, ~0U, start);
> + len = __iov_iter_get_pages_alloc(i, pages, maxsize, ~0U, start, 0);
> if (len <= 0) {
> kvfree(*pages);
> *pages = NULL;
> @@ -1518,6 +1532,24 @@ ssize_t iov_iter_get_pages_alloc2(struct iov_iter *i,
> }
> EXPORT_SYMBOL(iov_iter_get_pages_alloc2);
>
> +ssize_t iov_iter_get_pages_alloc_flags(struct iov_iter *i,
> + struct page ***pages, size_t maxsize,
> + size_t *start, unsigned int gup_flags)
> +{
> + ssize_t len;
> +
> + *pages = NULL;
> +
> + len = __iov_iter_get_pages_alloc(i, pages, maxsize, ~0U, start,
> + gup_flags);
> + if (len <= 0) {
> + kvfree(*pages);
> + *pages = NULL;
> + }
> + return len;
> +}
> +EXPORT_SYMBOL_GPL(iov_iter_get_pages_alloc_flags);
> +
> size_t csum_and_copy_from_iter(void *addr, size_t bytes, __wsum *csum,
> struct iov_iter *i)
> {
next prev parent reply other threads:[~2022-09-05 23:21 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-25 15:24 [PATCH v9 0/8] Userspace P2PDMA with O_DIRECT NVMe devices Logan Gunthorpe
2022-08-25 15:24 ` [PATCH v9 1/8] mm: introduce FOLL_PCI_P2PDMA to gate getting PCI P2PDMA pages Logan Gunthorpe
2022-09-05 22:27 ` John Hubbard
2022-08-25 15:24 ` [PATCH v9 2/8] iov_iter: introduce iov_iter_get_pages_[alloc_]flags() Logan Gunthorpe
2022-09-05 14:33 ` Christoph Hellwig
2022-09-05 23:21 ` John Hubbard [this message]
2022-09-06 16:52 ` Logan Gunthorpe
2022-08-25 15:24 ` [PATCH v9 3/8] block: add check when merging zone device pages Logan Gunthorpe
2022-09-05 14:34 ` Christoph Hellwig
2022-09-05 23:58 ` John Hubbard
2022-08-25 15:24 ` [PATCH v9 4/8] lib/scatterlist: " Logan Gunthorpe
2022-09-05 14:34 ` Christoph Hellwig
2022-09-06 0:21 ` John Hubbard
2022-08-25 15:24 ` [PATCH v9 5/8] block: set FOLL_PCI_P2PDMA in __bio_iov_iter_get_pages() Logan Gunthorpe
2022-09-05 14:36 ` Christoph Hellwig
2022-09-06 0:48 ` John Hubbard
2022-08-25 15:24 ` [PATCH v9 6/8] block: set FOLL_PCI_P2PDMA in bio_map_user_iov() Logan Gunthorpe
2022-09-05 14:36 ` Christoph Hellwig
2022-09-06 0:54 ` John Hubbard
2022-08-25 15:24 ` [PATCH v9 7/8] PCI/P2PDMA: Allow userspace VMA allocations through sysfs Logan Gunthorpe
2022-09-01 16:20 ` Greg Kroah-Hartman
2022-09-01 16:32 ` Logan Gunthorpe
2022-09-01 16:42 ` Greg Kroah-Hartman
2022-09-01 18:14 ` Logan Gunthorpe
2022-09-01 18:36 ` Greg Kroah-Hartman
2022-09-01 19:16 ` Logan Gunthorpe
2022-09-02 5:53 ` Greg Kroah-Hartman
2022-09-02 18:46 ` Logan Gunthorpe
2022-09-20 6:46 ` Christoph Hellwig
2022-09-22 8:38 ` Greg Kroah-Hartman
2022-09-22 14:58 ` Logan Gunthorpe
2022-08-25 15:24 ` [PATCH v9 8/8] ABI: sysfs-bus-pci: add documentation for p2pmem allocate Logan Gunthorpe
2022-09-01 16:18 ` Greg Kroah-Hartman
2022-09-01 16:33 ` Logan Gunthorpe
2022-09-06 1:03 ` John Hubbard
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=24ee706f-b70c-28d0-15a7-13d0dc9254bb@nvidia.com \
--to=jhubbard@nvidia.com \
--cc=christian.koenig@amd.com \
--cc=ckulkarnilinux@gmail.com \
--cc=dan.j.williams@intel.com \
--cc=daniel.vetter@ffwll.ch \
--cc=dave.b.minturn@intel.com \
--cc=dave.hansen@linux.intel.com \
--cc=ddutile@redhat.com \
--cc=gregkh@linuxfoundation.org \
--cc=hch@lst.de \
--cc=helgaas@kernel.org \
--cc=ira.weiny@intel.com \
--cc=jason@jlekstrand.net \
--cc=jgg@ziepe.ca \
--cc=jianxin.xiong@intel.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-nvme@lists.infradead.org \
--cc=linux-pci@vger.kernel.org \
--cc=logang@deltatee.com \
--cc=martin.oliveira@eideticom.com \
--cc=rcampbell@nvidia.com \
--cc=robin.murphy@arm.com \
--cc=sbates@raithlin.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).