On Jan 22, 2018, at 8:35 PM, Jerin Jacob [off-list ref] wrote:

-----Original Message-----

quoted

Date: Mon, 22 Jan 2018 18:29:31 +0000
From: Yongseok Koh <redacted>
To: Andrew Rybchenko <redacted>, Thomas Monjalon
[off-list ref], "jianbo.liu@arm.com" [off-list ref], Jerin
Jacob [off-list ref]
CC: Adrien Mazarguil <redacted>, Nélio Laranjeiro
[off-list ref], "bruce.richardson@intel.com"
[off-list ref], "Ananyev, Konstantin"
[off-list ref], Chao Zhu [off-list ref],
"dev@dpdk.org" [off-list ref]
Subject: Re: [dpdk-dev] [PATCH v3 1/8] eal: introduce DMA memory barriers

quoted

On Jan 18, 2018, at 11:16 PM, Andrew Rybchenko [off-list ref] wrote:

On 01/19/2018 03:44 AM, Yongseok Koh wrote:

quoted

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 
[off-list ref]

Acked-by: Andrew Rybchenko <redacted>

It is already really good. Many thanks.

Thank you!

quoted

Maybe it would be useful to:
- avoid duplication of so long explanations (put in in one place and add reference?)

May have to ask Thomas how to do this. Thomas?

quoted

- 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

Good idea. Like to hear from other people. But, following linux terms sometime
could be good to welcome developers from kernel community to DPDK world. :-)

To people in the cc list, any other concerns?
Especially ARM users - Jianbo and Jerin?

I like Andrew's suggestion. IMO, rte_cio_?mb() makes more sense.

If there's no more suggestion or objection, will send out v3 with the changes requested here by Andrew.

Thank you for reviews and comments.

Yongseok