Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   "Fabio M. De Francesco" <fmdefrancesco@gmail.com>
To:     linux-nvme@lists.infradead.org, Sagi Grimberg <sagi@grimberg.me>
Cc:     linux-kernel@vger.kernel.org, Christoph Hellwig <hch@lst.de>,
        Keith Busch <kbusch@kernel.org>,
        Chaitanya Kulkarni <Chaitanya.Kulkarni@wdc.com>,
        James Smart <james.smart@broadcom.com>,
        Ira Weiny <ira.weiny@intel.com>,
        Venkataramanan Anirudh <anirudh.venkataramanan@intel.com>,
        Al Viro <viro@zeniv.linux.org.uk>
Subject: Re: [PATCH] nvmet-tcp: Don't map pages which can't come from HIGHMEM
Date:   Tue, 30 Aug 2022 23:40:09 +0200
Message-ID: <5295597.Sb9uPGUboI@opensuse>
In-Reply-To: <20220829075401.868981-1-sagi@grimberg.me>
References: <20220829075401.868981-1-sagi@grimberg.me>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="utf-8"
Precedence: bulk

On luned=C3=AC 29 agosto 2022 09:54:01 CEST Sagi Grimberg wrote:
> From: "Fabio M. De Francesco" <fmdefrancesco@gmail.com>
>=20
> kmap() is being deprecated in favor of kmap_local_page().[1]
>=20
> There are two main problems with kmap(): (1) It comes with an overhead as
> mapping space is restricted and protected by a global lock for
> synchronization and (2) it also requires global TLB invalidation when the
> kmap=E2=80=99s pool wraps and it might block when the mapping space is fu=
lly
> utilized until a slot becomes available.
>=20
> The pages which will be mapped are allocated in nvmet_tcp_map_data(),
> using the GFP_KERNEL flag. This assures that they cannot come from
> HIGHMEM. This imply that a straight page_address() can replace the kmap()
> of sg_page(sg) in nvmet_tcp_map_pdu_iovec(). As a side effect, we might
> also delete the field "nr_mapped" from struct "nvmet_tcp_cmd" because,
> after removing the kmap() calls, there would be no longer any need of it.
>=20
> In addition, there is no reason to use a kvec for the command receive
> data buffers iovec, use a bio_vec instead and let iov_iter handle the
> buffer mapping and data copy.
>=20
> Test with blktests on a QEMU/KVM x86_32 VM, 6GB RAM, booting a kernel with
> HIGHMEM64GB enabled.
>=20
> [1] "[PATCH] checkpatch: Add kmap and kmap_atomic to the deprecated
> list" https://lore.kernel.org/all/20220813220034.806698-1-ira.weiny@intel=
=2Ecom/
>=20
> Cc: Chaitanya Kulkarni <chaitanyak@nvidia.com>
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Keith Busch <kbusch@kernel.org>
> Suggested-by: Ira Weiny <ira.weiny@intel.com>
> Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
> [sagi: added bio_vec plus minor naming changes]
> Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
> ---
>  drivers/nvme/target/tcp.c | 44 ++++++++++++---------------------------
>  1 file changed, 13 insertions(+), 31 deletions(-)

Hi Sagi,

Thanks for changing the code according to the suggestions from Christoph an=
d=20
Al.

Differently from what I just wrote while thanking Al, and after Ira made me=
=20
notice that you are one of the maintainers, I decided to leave everything l=
ike=20
you did.

The only exceptions we'll be about sending a v2 with the "Suggested-by" tag=
s=20
from Christoph and Al. In the meantime checkpatch warned that the alignment=
 of=20
"nr_pages" doesn't match the open parenthesis, so I'm changing it too.

Again thanks,

=46abio