Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
To:     Bjorn Helgaas <helgaas@kernel.org>, yu <kkndyu@gmail.com>
Cc:     linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org,
        Stephen Bates <sbates@raithlin.com>,
        Jonathan Cameron <jonathan.cameron@huawei.com>,
        Eric Wehage <Eric.Wehage@huawei.com>,
        Alexander Duyck <alexander.h.duyck@linux.intel.com>
References: <CADayL-pxr_dpJykyiTeQDO13XmKPoKUs544zVrB2taHTVEx9Pw@mail.gmail.com>
 <20181229022923.GA159477@google.com>
From:   Logan Gunthorpe <logang@deltatee.com>
Message-ID: <2aa4d13a-83c5-7e15-261f-b8f69def5224@deltatee.com>
Date:   Sat, 29 Dec 2018 12:18:03 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.3.0
MIME-Version: 1.0
In-Reply-To: <20181229022923.GA159477@google.com>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Subject: Re: How to force RC to forward p2p TLPs
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


On 2018-12-28 7:29 p.m., Bjorn Helgaas wrote:
>> We use the p2p DMA to transfer data between these two endpoint SOCs,
>> and if the host server is not enable ACS in BIOS, the p2p works well,
>> but when ACS is enabled in BIOS, the p2p is always failed. With the
>> help of a protocol analyzer, we can see that the TLP is redirected to
>> RC, and RC just discard it.
>>
>> I tried to find how to make RC forward redirected TLP to its original
>> target, but nothing found, it seems this is highly related to the RC
>> vendors.

As far as I know, this is not possible with any RCs I've worked with.
For our work, we ensure we disable ACS on paths between peers that want
to communicate. This is why we introduced the 'disable_acs_redir' kernel
parameter[1] to selectively disable ACS for bridges along P2P paths.

>> So is there some spec or document to describe how to set the RC? Any
>> suggestion is appreciated.

If there is, it would be the documentation for your specific RC. Odds
are, the RC simply does not support this.

Logan

[1] https://patchwork.kernel.org/patch/10549279/