linux-kernel - Re: [PATCH 2/4] i2c: nomadik: support Mobileye EyeQ6H I2C controller

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <fc17ef3d-7895-4ee8-bfa0-d31dd45f2f2c@kernel.org>
Date: Tue, 8 Oct 2024 18:03:36 +0200
From: Krzysztof Kozlowski <krzk@...nel.org>
To: Théo Lebrun <theo.lebrun@...tlin.com>
Cc: Linus Walleij <linus.walleij@...aro.org>,
 Andi Shyti <andi.shyti@...nel.org>, Rob Herring <robh@...nel.org>,
 Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley
 <conor+dt@...nel.org>, linux-arm-kernel@...ts.infradead.org,
 linux-i2c@...r.kernel.org, devicetree@...r.kernel.org,
 linux-kernel@...r.kernel.org,
 Vladimir Kondratiev <vladimir.kondratiev@...ileye.com>,
 Grégory Clement <gregory.clement@...tlin.com>,
 Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
 Tawfik Bayouk <tawfik.bayouk@...ileye.com>
Subject: Re: [PATCH 2/4] i2c: nomadik: support Mobileye EyeQ6H I2C controller

On 08/10/2024 16:43, Théo Lebrun wrote:
> Hello Krzysztof,
> 
> On Tue Oct 8, 2024 at 3:39 PM CEST, Krzysztof Kozlowski wrote:
>> On Tue, Oct 08, 2024 at 12:29:41PM +0200, Théo Lebrun wrote:
>>> Add EyeQ6H support to the nmk-i2c AMBA driver. It shares the same quirk
>>> as EyeQ5: the memory bus only supports 32-bit accesses. Avoid writeb()
>>> and readb() by reusing the same `priv->has_32b_bus` flag.
>>>
>>> It does NOT need to write speed-mode specific value into a register;
>>> therefore it does not depend on the mobileye,olb DT property.
>>>
>>> Refactoring is done using is_eyeq5 and is_eyeq6h boolean local
>>> variables. Sort variables in reverse christmas tree to try and
>>> introduce some logic into the ordering.
>>>
>>> Signed-off-by: Théo Lebrun <theo.lebrun@...tlin.com>
>>> ---
>>>  drivers/i2c/busses/i2c-nomadik.c | 22 +++++++++++-----------
>>>  1 file changed, 11 insertions(+), 11 deletions(-)
>>>
>>> diff --git a/drivers/i2c/busses/i2c-nomadik.c b/drivers/i2c/busses/i2c-nomadik.c
>>> index ad0f02acdb1215a1c04729f97bb14a4d93f88456..ea511d3a58073eaedb63850026e05b59427a69c6 100644
>>> --- a/drivers/i2c/busses/i2c-nomadik.c
>>> +++ b/drivers/i2c/busses/i2c-nomadik.c
>>> @@ -6,10 +6,10 @@
>>>   * I2C master mode controller driver, used in Nomadik 8815
>>>   * and Ux500 platforms.
>>>   *
>>> - * The Mobileye EyeQ5 platform is also supported; it uses
>>> + * The Mobileye EyeQ5 and EyeQ6H platforms are also supported; they use
>>>   * the same Ux500/DB8500 IP block with two quirks:
>>>   *  - The memory bus only supports 32-bit accesses.
>>> - *  - A register must be configured for the I2C speed mode;
>>> + *  - (only EyeQ5) A register must be configured for the I2C speed mode;
>>>   *    it is located in a shared register region called OLB.
>>>   *
>>>   * Author: Srinidhi Kasagar <srinidhi.kasagar@...ricsson.com>
>>> @@ -1046,8 +1046,6 @@ static int nmk_i2c_eyeq5_probe(struct nmk_i2c_dev *priv)
>>>  	struct regmap *olb;
>>>  	unsigned int id;
>>>  
>>> -	priv->has_32b_bus = true;
>>> -
>>>  	olb = syscon_regmap_lookup_by_phandle_args(np, "mobileye,olb", 1, &id);
>>>  	if (IS_ERR(olb))
>>>  		return PTR_ERR(olb);
>>> @@ -1070,13 +1068,15 @@ static int nmk_i2c_eyeq5_probe(struct nmk_i2c_dev *priv)
>>>  
>>>  static int nmk_i2c_probe(struct amba_device *adev, const struct amba_id *id)
>>>  {
>>> -	int ret = 0;
>>> -	struct nmk_i2c_dev *priv;
>>> -	struct device_node *np = adev->dev.of_node;
>>> -	struct device *dev = &adev->dev;
>>> -	struct i2c_adapter *adap;
>>>  	struct i2c_vendor_data *vendor = id->data;
>>> +	struct device_node *np = adev->dev.of_node;
>>> +	bool is_eyeq6h = of_device_is_compatible(np, "mobileye,eyeq6h-i2c");
>>> +	bool is_eyeq5 = of_device_is_compatible(np, "mobileye,eyeq5-i2c");
>>
>> You should use match data, not add compatibles in the middle of code.
>> That's preferred, scallable pattern. What you added here last time does
>> not scale and above change is a proof for that.
> 
> I would have used match data if the driver struct had a .of_match_table
> field. `struct amba_driver` does not. Are you recommending the approach
> below?
> 
> I don't see how it brings much to the driver but I do get the scaling
> issue as the number of support compatibles increases. This is a fear
> based on what *could* happen in the future though.

You still have adev->dev.of_node, which you can use for matching in
probe. See for example of_match_device() (and add a note so people will
not convert it to device_get_match_data() blindly).

Best regards,
Krzysztof