From: Andrew Rybchenko <arybchenko@solarflare.com>
To: Yongseok Koh <yskoh@mellanox.com>, <adrien.mazarguil@6wind.com>,
<nelio.laranjeiro@6wind.com>, <bruce.richardson@intel.com>,
<konstantin.ananyev@intel.com>, <chaozhu@linux.vnet.ibm.com>,
<jerin.jacob@caviumnetworks.com>, <jianbo.liu@arm.com>
Cc: <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v3 1/8] eal: introduce DMA memory barriers
Date: Fri, 19 Jan 2018 10:16:17 +0300 [thread overview]
Message-ID: <805e1578-6e4d-6a1b-00ce-13e3b1c16be8@solarflare.com> (raw)
In-Reply-To: <20180119004430.15305-2-yskoh@mellanox.com>
On 01/19/2018 03:44 AM, Yongseok Koh wrote:
> This commit introduces rte_dma_wmb() and rte_dma_rmb(), in order to
> guarantee the ordering of coherent shared memory between the CPU and a DMA
> capable device.
>
> Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
It is already really good. Many thanks.
Maybe it would be useful to:
- avoid duplication of so long explanations (put in in one place and
add reference?)
- explain why it is bound to DMA or call it in a different way, since
right now it is bound
to coherent-mapped IO (rte_cio_rmb() ?). Yes, I see benefits to
follow Linux
terminology, but may be DPDK can do better :) I just add my
concerns, but let
EAL code maintainers to decide
- as I understand right now there is no control over DMA mapping since
mapping is done by UIO/VFIO drivers. Should documentation be updated
that
DMA mapping is assumed to be coherent?
- when it is applied, may be it makes sense to send HEADS UP to dev@,
it definitely deserves to be mentioned in the release notes
<...>
next prev parent reply other threads:[~2018-01-19 7:16 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-12-27 4:28 [dpdk-dev] [PATCH 1/2] eal/arm64: modify I/O device " Yongseok Koh
2017-12-27 4:28 ` [dpdk-dev] [PATCH 2/2] net/mlx5: fix synchonization on polling Rx completions Yongseok Koh
2018-01-04 12:58 ` [dpdk-dev] [PATCH 1/2] eal/arm64: modify I/O device memory barriers Jerin Jacob
2018-01-08 1:55 ` Jianbo Liu
2018-01-16 0:42 ` Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 0/8] introduce DMA " Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 1/8] eal: " Yongseok Koh
2018-01-16 2:47 ` Jianbo Liu
2018-01-16 7:49 ` Andrew Rybchenko
2018-01-16 9:10 ` Jianbo Liu
2018-01-17 13:46 ` Thomas Monjalon
2018-01-17 18:39 ` Yongseok Koh
2018-01-18 11:56 ` Andrew Rybchenko
2018-01-18 18:14 ` Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 2/8] eal/x86: define " Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 3/8] eal/ppc64: define DMA device " Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 4/8] eal/armv7: define DMA " Yongseok Koh
2018-01-16 2:48 ` Jianbo Liu
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 5/8] eal/arm64: " Yongseok Koh
2018-01-16 2:50 ` Jianbo Liu
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 6/8] net/mlx5: remove unnecessary memory barrier Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 7/8] net/mlx5: replace IO memory barrier with DMA " Yongseok Koh
2018-01-16 1:10 ` [dpdk-dev] [PATCH v2 8/8] net/mlx5: fix synchonization on polling Rx completions Yongseok Koh
2018-01-16 3:53 ` Jianbo Liu
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 0/8] introduce DMA memory barriers Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 1/8] eal: " Yongseok Koh
2018-01-19 7:16 ` Andrew Rybchenko [this message]
2018-01-22 18:29 ` Yongseok Koh
2018-01-22 20:59 ` Thomas Monjalon
2018-01-23 4:35 ` Jerin Jacob
2018-01-25 19:08 ` Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 2/8] eal/x86: define " Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 3/8] eal/ppc64: " Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 4/8] eal/armv7: " Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 5/8] eal/arm64: " Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 6/8] net/mlx5: remove unnecessary memory barrier Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 7/8] net/mlx5: replace IO memory barrier with DMA " Yongseok Koh
2018-01-19 0:44 ` [dpdk-dev] [PATCH v3 8/8] net/mlx5: fix synchonization on polling Rx completions Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 0/9] introduce coherent I/O memory barriers Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 1/9] eal: add Doxygen grouping for " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 2/9] eal: introduce coherent I/O " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 3/9] eal/x86: define " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 4/9] eal/ppc64: " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 5/9] eal/armv7: " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 6/9] eal/arm64: " Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 7/9] net/mlx5: remove unnecessary memory barrier Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 8/9] net/mlx5: replace I/O memory barrier with coherent version Yongseok Koh
2018-01-25 21:02 ` [dpdk-dev] [PATCH v4 9/9] net/mlx5: fix synchronization on polling Rx completions Yongseok Koh
2018-01-28 7:32 ` [dpdk-dev] [PATCH v4 0/9] introduce coherent I/O memory barriers Thomas Monjalon
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=805e1578-6e4d-6a1b-00ce-13e3b1c16be8@solarflare.com \
--to=arybchenko@solarflare.com \
--cc=adrien.mazarguil@6wind.com \
--cc=bruce.richardson@intel.com \
--cc=chaozhu@linux.vnet.ibm.com \
--cc=dev@dpdk.org \
--cc=jerin.jacob@caviumnetworks.com \
--cc=jianbo.liu@arm.com \
--cc=konstantin.ananyev@intel.com \
--cc=nelio.laranjeiro@6wind.com \
--cc=yskoh@mellanox.com \
/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).