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;
Subject: Re: [PATCH v2] rpmsg: core: add API to get MTU
To:     Bjorn Andersson <bjorn.andersson@linaro.org>
CC:     Ohad Ben-Cohen <ohad@wizery.com>, <linux-kernel@vger.kernel.org>,
        <linux-remoteproc@vger.kernel.org>,
        <linux-arm-msm@vger.kernel.org>, Suman Anna <s-anna@ti.com>,
        Fabien DESSENNE <fabien.dessenne@st.com>,
        <linux-stm32@st-md-mailman.stormreply.com>
References: <20191113172249.32412-1-arnaud.pouliquen@st.com>
 <20200113172453.GQ738324@yoga>
From:   Arnaud POULIQUEN <arnaud.pouliquen@st.com>
Message-ID: <c6ecd3b6-2a3b-11d8-6d1c-a531c73bc388@st.com>
Date:   Tue, 14 Jan 2020 10:06:38 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.2.2
MIME-Version: 1.0
In-Reply-To: <20200113172453.GQ738324@yoga>
Content-Type: text/plain; charset="utf-8"
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi Bjorn

On 1/13/20 6:24 PM, Bjorn Andersson wrote:
> On Wed 13 Nov 09:22 PST 2019, Arnaud Pouliquen wrote:
> 
>> Return the rpmsg buffer MTU for sending message, so rpmsg users
>> can split a long message in several sub rpmsg buffers.
>>
> 
> I won't merge this new api without a client, and I'm still concerned
> about the details.
The client exists: it is the rpmsg tty that i 've been rying to upstream since for a while.
https://patchwork.kernel.org/cover/11130213/
This patch is the result of some comments you did on rpmsg tty thread. 
Suman was also interested in and request to merge it independently
(https://lkml.org/lkml/2019/9/3/774).
That's why i'm trying to do it in 2 steps.

> 
>> Signed-off-by: Arnaud Pouliquen <arnaud.pouliquen@st.com>
>> ---
>>  V1 to V2
>>
>>   V1 patch:https://lore.kernel.org/patchwork/patch/1124684/
>>   - Change patch title,
>>   - as not solution today to support MTU on GLINK make ops optional,
>>     RPMsg client API returns -ENOTSUPP in this case,
>>   - suppress smd and glink patches.
> 
> That's ok.
> 
>> ---
>>  drivers/rpmsg/rpmsg_core.c       | 21 +++++++++++++++++++++
>>  drivers/rpmsg/rpmsg_internal.h   |  2 ++
>>  drivers/rpmsg/virtio_rpmsg_bus.c | 10 ++++++++++
>>  include/linux/rpmsg.h            | 10 ++++++++++
>>  4 files changed, 43 insertions(+)
>>
>> diff --git a/drivers/rpmsg/rpmsg_core.c b/drivers/rpmsg/rpmsg_core.c
>> index e330ec4dfc33..a6ef54c4779a 100644
>> --- a/drivers/rpmsg/rpmsg_core.c
>> +++ b/drivers/rpmsg/rpmsg_core.c
>> @@ -283,6 +283,27 @@ int rpmsg_trysend_offchannel(struct rpmsg_endpoint *ept, u32 src, u32 dst,
>>  }
>>  EXPORT_SYMBOL(rpmsg_trysend_offchannel);
>>  
>> +/**
>> + * rpmsg_get_mtu() - get maximum transmission buffer size for sending message.
>> + * @ept: the rpmsg endpoint
>> + *
>> + * This function returns maximum buffer size available for a single message.
>> + *
>> + * Return: the maximum transmission size on success and an appropriate error
>> + * value on failure.
> 
> Is the expectation that a call to rpmsg_send() with this size will
> eventually succeed?
yes, this should be the role of the transport layer
(e.g. RPMsg VirtIO bus) to ensure this.

> 
>> + */
> [..]
>> +static ssize_t virtio_rpmsg_get_mtu(struct rpmsg_endpoint *ept)
>> +{
>> +	struct rpmsg_device *rpdev = ept->rpdev;
>> +	struct virtio_rpmsg_channel *vch = to_virtio_rpmsg_channel(rpdev);
>> +
>> +	return vch->vrp->buf_size - sizeof(struct rpmsg_hdr);
> 
> I'm still under the impression that the rpmsg protocol doesn't have to
> operate on fixed size messages. Would this then return vrp->num_bufs *
> vrp->buf_size / 2 - sizeof(rpmsg_hdr)?
it depends on the transport layer. For RPMsg over virtio, this is the size
of the payload of a buffer so vrp->buf_size  - sizeof(rpmsg_hdr)

Regards,
Arnaud

> 
>> +}
>> +
> 
> Regards,
> Bjorn
>