linux-kernel - Re: [PATCH 3/4] rpmsg: char: Introduce the "rpmsg-raw" channel

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <b55cd4e5-fb9d-a0ab-03a9-3a771898db04@foss.st.com>
Date:   Wed, 16 Jun 2021 14:38:26 +0200
From:   Arnaud POULIQUEN <arnaud.pouliquen@...s.st.com>
To:     Mathieu Poirier <mathieu.poirier@...aro.org>
CC:     Bjorn Andersson <bjorn.andersson@...aro.org>,
        Ohad Ben-Cohen <ohad@...ery.com>,
        <linux-remoteproc@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
        <linux-stm32@...md-mailman.stormreply.com>, <julien.massot@....bzh>
Subject: Re: [PATCH 3/4] rpmsg: char: Introduce the "rpmsg-raw" channel

Hi Mathieu,

On 6/15/21 10:01 PM, Mathieu Poirier wrote:
> On Mon, Jun 07, 2021 at 07:30:31PM +0200, Arnaud Pouliquen wrote:
>> Allows to probe the endpoint device on a remote name service announcement,
>> by registering a rpmsg_driverfor the "rpmsg-raw" channel.
>>
>> With this patch the /dev/rpmsgX interface can be instantiated by the remote
>> firmware.
>>
>> Signed-off-by: Arnaud Pouliquen <arnaud.pouliquen@...s.st.com>
>> ---
>>  drivers/rpmsg/rpmsg_char.c | 54 ++++++++++++++++++++++++++++++++++++--
>>  1 file changed, 52 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/rpmsg/rpmsg_char.c b/drivers/rpmsg/rpmsg_char.c
>> index 4199ac1bee10..3b850b218eb0 100644
>> --- a/drivers/rpmsg/rpmsg_char.c
>> +++ b/drivers/rpmsg/rpmsg_char.c
>> @@ -25,6 +25,8 @@
>>  
>>  #include "rpmsg_char.h"
>>  
>> +#define RPMSG_CHAR_DEVNAME "rpmsg-raw"
>> +
>>  static dev_t rpmsg_major;
>>  static struct class *rpmsg_class;
>>  
>> @@ -416,6 +418,40 @@ int rpmsg_chrdev_eptdev_create(struct rpmsg_device *rpdev, struct device *parent
>>  }
>>  EXPORT_SYMBOL(rpmsg_chrdev_eptdev_create);
>>  
>> +static int rpmsg_chrdev_probe(struct rpmsg_device *rpdev)
>> +{
>> +	struct rpmsg_channel_info chinfo;
>> +
>> +	memcpy(chinfo.name, RPMSG_CHAR_DEVNAME, sizeof(RPMSG_CHAR_DEVNAME));
>> +	chinfo.src = rpdev->src;
>> +	chinfo.dst = rpdev->dst;
>> +
>> +	return __rpmsg_chrdev_eptdev_create(rpdev, &rpdev->dev, chinfo, true);
> 
> I am a little puzzled here as to why we need different modes... Why can't we
> simply call rpmsg_chrdev_eptdev_create() and let the endpoint be created on
> open() and destroyed on release() as per the current implementation?

The main reason is the support of the NS announcement
a NS announcement is received from the remote processor:
channel name: "rpmsg-raw"
remote address (dst address): 0x400
local address (scr address) : RPMSG_ADDR_ANY
=> no default endpoint, and not local address.

case 1) if we use legacy implementation ( no default endpoint)
=> create/destroy endpoint on open/stop
- on first open: created endpoint is bound to scr address 0x406
- a first message is sent to the remote side, the address 0x406 is stored as
default channel dst address on remote side.
- on close: endpoint is closed and associated address 0x406 is free.
- another driver create an enpoint the address 0x406 is reserved for this new
endpoint.
- on new open:  scr address is set to next value 0x407
=> how to inform remote processor that the address has changed?
=> no reservation mechanism that ensure that you can reuse the same address

case 2) relying on use_default_ept
=> Ensure that both side have always the same addresses to communicate.

> 
> I'd rather keep things simple for the refactoring and introduce new features
> later if need be.

Yes I agree with you, but here it could become a nightmare for the remote
processor if the Linux endpoint address is not stable.

Anyway we can consider this as a workaround waiting the extension of the NS
announcement to have a better management of the address exchange on channel
initialization.

Thanks
Arnaud

> 
> As I said, it may be that I don't understand the usecase.
> 
> Thanks,
> Mathieu
> 
>> +}
>> +
>> +static void rpmsg_chrdev_remove(struct rpmsg_device *rpdev)
>> +{
>> +	int ret;
>> +
>> +	ret = device_for_each_child(&rpdev->dev, NULL, rpmsg_chrdev_eptdev_destroy);
>> +	if (ret)
>> +		dev_warn(&rpdev->dev, "failed to destroy endpoints: %d\n", ret);
>> +}
>> +
>> +static struct rpmsg_device_id rpmsg_chrdev_id_table[] = {
>> +	{ .name	= RPMSG_CHAR_DEVNAME },
>> +	{ },
>> +};
>> +
>> +static struct rpmsg_driver rpmsg_chrdev_driver = {
>> +	.probe = rpmsg_chrdev_probe,
>> +	.remove = rpmsg_chrdev_remove,
>> +	.id_table = rpmsg_chrdev_id_table,
>> +	.drv = {
>> +		.name = "rpmsg_chrdev",
>> +	},
>> +};
>> +
>>  static int rpmsg_chrdev_init(void)
>>  {
>>  	int ret;
>> @@ -429,16 +465,30 @@ static int rpmsg_chrdev_init(void)
>>  	rpmsg_class = class_create(THIS_MODULE, "rpmsg");
>>  	if (IS_ERR(rpmsg_class)) {
>>  		pr_err("failed to create rpmsg class\n");
>> -		unregister_chrdev_region(rpmsg_major, RPMSG_DEV_MAX);
>> -		return PTR_ERR(rpmsg_class);
>> +		ret = PTR_ERR(rpmsg_class);
>> +		goto free_region;
>> +	}
>> +
>> +	ret = register_rpmsg_driver(&rpmsg_chrdev_driver);
>> +	if (ret < 0) {
>> +		pr_err("rpmsg: failed to register rpmsg raw driver\n");
>> +		goto free_class;
>>  	}
>>  
>>  	return 0;
>> +
>> +free_class:
>> +	class_destroy(rpmsg_class);
>> +free_region:
>> +	unregister_chrdev_region(rpmsg_major, RPMSG_DEV_MAX);
>> +
>> +	return ret;
>>  }
>>  postcore_initcall(rpmsg_chrdev_init);
>>  
>>  static void rpmsg_chrdev_exit(void)
>>  {
>> +	unregister_rpmsg_driver(&rpmsg_chrdev_driver);
>>  	class_destroy(rpmsg_class);
>>  	unregister_chrdev_region(rpmsg_major, RPMSG_DEV_MAX);
>>  }
>> -- 
>> 2.17.1
>>