linux-kernel - Re: [PATCH v3 7/8] watchdog: max77714: add driver for the watchdog in the MAX77714 PMIC

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: <2af4073a-a5d9-e878-d622-b757b082359b@canonical.com>
Date:   Fri, 12 Nov 2021 17:07:56 +0100
From:   Krzysztof Kozlowski <krzysztof.kozlowski@...onical.com>
To:     Luca Ceresoli <luca@...aceresoli.net>,
        Guenter Roeck <linux@...ck-us.net>,
        linux-kernel@...r.kernel.org
Cc:     Lee Jones <lee.jones@...aro.org>, Rob Herring <robh+dt@...nel.org>,
        Chanwoo Choi <cw00.choi@...sung.com>,
        Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>,
        Alessandro Zummo <a.zummo@...ertech.it>,
        Alexandre Belloni <alexandre.belloni@...tlin.com>,
        Wim Van Sebroeck <wim@...ux-watchdog.org>,
        devicetree@...r.kernel.org, linux-rtc@...r.kernel.org,
        linux-watchdog@...r.kernel.org,
        Chiwoong Byun <woong.byun@...sung.com>,
        Laxman Dewangan <ldewangan@...dia.com>,
        Randy Dunlap <rdunlap@...radead.org>
Subject: Re: [PATCH v3 7/8] watchdog: max77714: add driver for the watchdog in
 the MAX77714 PMIC

On 12/11/2021 17:02, Luca Ceresoli wrote:
> Hi Guenter,
> 
> On 12/11/21 15:57, Guenter Roeck wrote:
>> On 11/11/21 2:58 PM, Luca Ceresoli wrote:
>>> Add a simple driver to support the watchdog embedded in the Maxim
>>> MAX77714
>>> PMIC.
>>>
>>> Signed-off-by: Luca Ceresoli <luca@...aceresoli.net>
>>>
>>
>> I just realized that this is effectively a rewrite of
>> drivers/watchdog/max77620_wdt.c.
>> The only difference I can see is is the register offsets (0x91 and 0x92
>> vs. 1 and 2) and some implementation details. Please add support for this
>> watchdog to the other driver or provide a _really_ good reason why that
>> is not possible.
> 
> I initially started developing MAX77714 watchdog support as an addition
> to max77620_wdt.c as the procedures look identical at least for the
> basic features.
> 
> But the register content seems completely different. Here are the notes
> I took at that time:
> 
> -------------------------8<-------------------------
> 
> MAX77620 has reg ONOFFCNFG1  at 0x41, ONOFFCNFG2  at 0x42.
> MAX77714 has reg CNFG1_ONOFF at 0x93, CNFG2_ONOFF at 0x94.
> OK, we can handle this with a register indirection table, indexed by
> chip model.
> 
> MAX77620 has MAX77620_REG_FPS_CFG0 register.
> On MAX77714 I was unable to find any such register (I haven't looked at
> FPS in detail though).
> OK, we can handle this with some if()s or entirely disable PM on the
> 77714 until anybody cares.
> 
> MAX77620 ONOFFCNFG1  has SFT_RST in bit 7.
> MAX77714 CNFG1_ONOFF has SFT_RST is bit 6.
> Uhm, should we have a _bit_ indirection table in addition to the
> _register_ indirection table?
> 
> MAX77620 ONOFFCNFG2  bit 5 is SLP_LPM_MSK, involved in FPS.
> MAX77620 ONOFFCNFG2  bit 6 is WD_RTS_WK, configures the watchdog timer.
> MAX77714 CNFG2_ONOFF bit 5 is WD_RTS_WK, configures the watchdog timer.
> On MAX77714 I haven't found SLP_LPM_MSK.
> 
> MAX77620 has 6 CID registers with "ES version" in CID5.
> MAX77714 has 5 CID registers with "DEVICE id" in CID3.
> CID registers would be useful to get the chip model directly from the
> chip, if only they had the same structure.
> 
> Almost all of the registers I have been looking into have similar
> differences.
> 
> -------------------------8<-------------------------
> 
> When I started adding indirection tables the driver started growing
> bigger and uglier, and that little simple driver started being big and
> complex. So I opted to add a new driver.
> 

The register offset differences are trivial and we do it in several
drivers. Also in rtc-max77686 used by you here.
Lack of features as well - just have a variant/driver data which defines
certain features (true/false) or quirk bits (see s3c2410_wdt).

The second driver - s3c2410_wdt - also customizes the bits.

Therefore if the generic device operating configuration is similar (same
generic control flow) and differences are in bits and offsets, then it
should be one driver.

Best regards,
Krzysztof