linux-kernel - Re: [PATCH 4/8] platform/chrome: cros_ec_chardev: Use cros_ec_cmd_xfer

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: <4e441ead-44f3-1787-28a3-3e71893105f9@collabora.com>
Date:   Wed, 26 Feb 2020 15:59:59 +0100
From:   Enric Balletbo i Serra <enric.balletbo@...labora.com>
To:     Prashant Malani <pmalani@...omium.org>
Cc:     linux-kernel@...r.kernel.org,
        Collabora Kernel ML <kernel@...labora.com>,
        groeck@...omium.org, bleung@...omium.org, dtor@...omium.org,
        gwendal@...omium.org
Subject: Re: [PATCH 4/8] platform/chrome: cros_ec_chardev: Use
 cros_ec_cmd_xfer_status helper

Hi Prashant,

On 25/2/20 20:55, Prashant Malani wrote:
> Hi Enric,
> 
> On Thu, Feb 20, 2020 at 04:58:55PM +0100, Enric Balletbo i Serra wrote:
>> This patch makes use of cros_ec_cmd_xfer_status() instead of
>> cros_ec_cmd_xfer(). In this case the change is trivial and the only
>> reason to do it is because we want to make cros_ec_cmd_xfer() a private
>> function for the EC protocol and let people only use the
>> cros_ec_cmd_xfer_status() to return Linux standard error codes.
>>
>> Signed-off-by: Enric Balletbo i Serra <enric.balletbo@...labora.com>
>> ---
>>
>>  drivers/platform/chrome/cros_ec_chardev.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/platform/chrome/cros_ec_chardev.c b/drivers/platform/chrome/cros_ec_chardev.c
>> index c65e70bc168d..b51ab24055f3 100644
>> --- a/drivers/platform/chrome/cros_ec_chardev.c
>> +++ b/drivers/platform/chrome/cros_ec_chardev.c
>> @@ -301,7 +301,7 @@ static long cros_ec_chardev_ioctl_xcmd(struct cros_ec_dev *ec, void __user *arg)
>>  	}
>>  
>>  	s_cmd->command += ec->cmd_offset;
>> -	ret = cros_ec_cmd_xfer(ec->ec_dev, s_cmd);
>> +	ret = cros_ec_cmd_xfer_status(ec->ec_dev, s_cmd);
> 
> One issue I see here is that if we were to later convert
> cros_ec_cmd_xfer_status() to cros_ec_cmd(), we would lose the
> s_cmd->result value, since I was hoping to avoid returning msg->result
> via a pointer parameter. I don't know if userspace actually uses that, but I
> am assuming it does.
> 

I'd like to avoid returning msg->result via a pointer parameter too. I didn't
analyze all the cases but I suspect that in most cases it is not really needed,
so let's start by converting to the cros_ec_cmd the cases that are clear and
let's go one by one for the ones that are not clear.

IMO cros_ec_cmd should return the same as cros_ec_cmd_xfer_status. So you should
use cros_ec_cmd_xfer_status inside cros_ec_cmd.

Regards,
 Enric

> So, should cros_ec_cmd() keep the *result pointer like so ?:
> 
> int cros_ec_cmd(struct cros_ec_device *ec, u32 version, u32 command,
>                void *outdata, u32 outsize, void *indata, u32 insize, u32 *result);
> 
> This way, we can manually re-populate s_cmd->result with |*result|.
> 
> Or, should we drop msg->result while returning s_cmd to userspace? I am
> guessing the answer is no, but thought I'd check with you first.
> 
> Best regards,
> 
> 
>>  	/* Only copy data to userland if data was received. */
>>  	if (ret < 0)
>>  		goto exit;
>> -- 
>> 2.25.0
>>