[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <26a62e8c-b3cd-bf61-2113-1e8375f2cc9c@huawei.com>
Date: Tue, 20 Sep 2022 15:35:51 +0800
From: taolan <taolan@...wei.com>
To: Wolfram Sang <wsa@...nel.org>, <linux-i2c@...r.kernel.org>,
<linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] i2c: hix5hd2: Add some debug enhancement for register
access
-------- Original Message --------
Title: Re: [PATCH] i2c: hix5hd2: Add some debug enhancement for register access
From: wsa@...nel.org
To: taolan@...wei.com; Cc: linux-i2c@...r.kernel.org, linux-kernel@...r.kernel.org
Date: 2022/9/18 04:29
> On Tue, Sep 13, 2022 at 12:48:20PM +0000, Tao Lan wrote:
>> From: taolan <taolan@...wei.com>
>>
>> Sometimes, to locate a fault, we need to know how the register is
>> configured and whether the configuration is incorrect. Currently, no
>> better method is available for analysis.
>>
>> This patch tries to solve the problem by naming the registers and
>> printing the accessed values.
>>
>> Signed-off-by: taolan <taolan@...wei.com>
>
> Such debug helpers are not suitable for upstream because they need quite
> some resources for rare cases...
Yeah, how about writing to proc as an event, or any suggestions?
>
>> + pr_notice("write 0x%04x to %s\n", val, g_i2c_reg_name[idx >> 0x2].name);
>
> ... and using this loglevel you will flood logs of users.
The loglevel can be adjusted to DEBUG level,and this debug can be off by default. is this ok?
>
>
Powered by blists - more mailing lists