linux-kernel - Re: [PATCH 1/2] platform/chrome: skip old cros

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: <bb66112e-5830-878c-4ce7-ac462833d61b@collabora.com>
Date:   Fri, 10 Apr 2020 00:42:10 +0200
From:   Enric Balletbo i Serra <enric.balletbo@...labora.com>
To:     Mathew King <mathewk@...omium.org>, linux-kernel@...r.kernel.org
Cc:     Jett Rink <jettrink@...omium.org>,
        Benson Leung <bleung@...omium.org>,
        Guenter Roeck <groeck@...omium.org>
Subject: Re: [PATCH 1/2] platform/chrome: skip old cros_ec responses

Hi Mathew,

Thanks for your patch.

On 8/4/20 20:16, Mathew King wrote:
> From: Jett Rink <jettrink@...omium.org>
> 
> The ISHTP layer can give us old responses that we already gave up on. We
> do not want to interpret these old responses as the current response we
> are waiting for.
> 

Looking at the code and with the above explanation I am not sure I get what is
doing this patch, could you explain a bit more, thanks.

> Signed-off-by: Jett Rink <jettrink@...omium.org>
> Signed-off-by: Mathew King <mathewk@...omium.org>
> ---
>  drivers/platform/chrome/cros_ec_ishtp.c | 32 ++++++++++++++++++-------
>  1 file changed, 23 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/platform/chrome/cros_ec_ishtp.c b/drivers/platform/chrome/cros_ec_ishtp.c
> index 93a71e93a2f1..6f90deb5cf55 100644
> --- a/drivers/platform/chrome/cros_ec_ishtp.c
> +++ b/drivers/platform/chrome/cros_ec_ishtp.c
> @@ -48,7 +48,8 @@ static const guid_t cros_ish_guid =
>  struct header {
>  	u8 channel;
>  	u8 status;
> -	u8 reserved[2];
> +	u8 id;
> +	u8 reserved;
>  } __packed;
>  
>  struct cros_ish_out_msg {
> @@ -90,6 +91,7 @@ static DECLARE_RWSEM(init_lock);
>   * data exceeds this value, we log an error.
>   * @size: Actual size of data received from firmware.
>   * @error: 0 for success, negative error code for a failure in process_recv().
> + * @expected_id: Expected id for response that we are waiting on.
>   * @received: Set to true on receiving a valid firmware	response to host command
>   * @wait_queue: Wait queue for host to wait for firmware response.
>   */
> @@ -98,6 +100,7 @@ struct response_info {
>  	size_t max_size;
>  	size_t size;
>  	int error;
> +	u8 expected_id;
>  	bool received;
>  	wait_queue_head_t wait_queue;
>  };
> @@ -162,6 +165,7 @@ static int ish_send(struct ishtp_cl_data *client_data,
>  		    u8 *out_msg, size_t out_size,
>  		    u8 *in_msg, size_t in_size)
>  {
> +	static u8 current_id;
>  	int rv;
>  	struct header *out_hdr = (struct header *)out_msg;
>  	struct ishtp_cl *cros_ish_cl = client_data->cros_ish_cl;
> @@ -174,8 +178,11 @@ static int ish_send(struct ishtp_cl_data *client_data,
>  	client_data->response.data = in_msg;
>  	client_data->response.max_size = in_size;
>  	client_data->response.error = 0;
> +	client_data->response.expected_id = ++current_id;

So on every ish_send call this variable is increased in a range 1 to 255 first,
then overflows and goes from 0 to 255. Is this what you want to do?

>  	client_data->response.received = false;
>  
> +	out_hdr->id = client_data->response.expected_id;
> +
>  	rv = ishtp_cl_send(cros_ish_cl, out_msg, out_size);
>  	if (rv) {
>  		dev_err(cl_data_to_dev(client_data),
> @@ -249,17 +256,23 @@ static void process_recv(struct ishtp_cl *cros_ish_cl,
>  
>  	switch (in_msg->hdr.channel) {
>  	case CROS_EC_COMMAND:
> -		/* Sanity check */
> -		if (!client_data->response.data) {
> +		if (client_data->response.received) {
>  			dev_err(dev,
> -				"Receiving buffer is null. Should be allocated by calling function\n");
> -			client_data->response.error = -EINVAL;
> -			goto error_wake_up;
> +				"Previous firmware message not yet processed\n");
> +			goto end_error;
>  		}
>  
> -		if (client_data->response.received) {
> +		if (client_data->response.expected_id != in_msg->hdr.id) {

And here you compare that the response received matches with the message id. I
assume the ISH is sending a sequential id on every message?

>  			dev_err(dev,
> -				"Previous firmware message not yet processed\n");
> +				"Dropping old response id %d\n",
> +				in_msg->hdr.id);


How often this message appears?

> +			goto end_error;
> +		}
> +
> +		/* Sanity check */
> +		if (!client_data->response.data) {
> +			dev_err(dev,
> +				"Receiving buffer is null. Should be allocated by calling function\n");
>  			client_data->response.error = -EINVAL;
>  			goto error_wake_up;
>  		}
> @@ -289,9 +302,10 @@ static void process_recv(struct ishtp_cl *cros_ish_cl,
>  		memcpy(client_data->response.data,
>  		       rb_in_proc->buffer.data, data_len);
>  
> +error_wake_up:
>  		/* Set flag before waking up the caller */
>  		client_data->response.received = true;
> -error_wake_up:
> +
>  		/* Wake the calling thread */
>  		wake_up_interruptible(&client_data->response.wait_queue);
>  
>