Received-SPF: Pass (protection.outlook.com: domain of amd.com designates
 165.204.84.17 as permitted sender) receiver=protection.outlook.com;
 client-ip=165.204.84.17; helo=SATLEXMB04.amd.com;
Date:   Fri, 7 Jan 2022 12:05:45 +0100
From:   Robert Richter <rrichter@amd.com>
To:     Terry Bowman <Terry.Bowman@amd.com>
CC:     Guenter Roeck <linux@roeck-us.net>,
        <linux-watchdog@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
        <wim@linux-watchdog.org>, <ssg.sos.patches@amd.com>,
        <sudheesh.mavila@amd.com>,
        "Lendacky, Thomas" <thomas.lendacky@amd.com>
Subject: Re: [PATCH v2 3/4] Watchdog: sp5100_tco: Add EFCH SMBus controller
 initialization using MMIO
Message-ID: <YdgeU2p+i5hN1XCU@rric.localdomain>
References: <20211103161521.43447-1-terry.bowman@amd.com>
 <20211103161521.43447-4-terry.bowman@amd.com>
 <20220106181809.GA240027@roeck-us.net>
 <9afabe55-6429-2284-cafd-d59ce481f067@amd.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <9afabe55-6429-2284-cafd-d59ce481f067@amd.com>
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 07 Jan 2022 11:05:51.2434
 (UTC)
X-MS-Exchange-CrossTenant-Network-Message-Id: e8ec5a07-0946-4bc8-8f53-08d9d1cdaad1
X-MS-Exchange-CrossTenant-Id: 3dd8961f-e488-4e60-8e11-a82d994e183d
X-MS-Exchange-CrossTenant-OriginalAttributedTenantConnectingIp: TenantId=3dd8961f-e488-4e60-8e11-a82d994e183d;Ip=[165.204.84.17];Helo=[SATLEXMB04.amd.com]
X-MS-Exchange-CrossTenant-AuthSource: CO1NAM11FT034.eop-nam11.prod.protection.outlook.com
X-MS-Exchange-CrossTenant-AuthAs: Anonymous
X-MS-Exchange-CrossTenant-FromEntityHeader: HybridOnPrem
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MWHPR12MB1789
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 06.01.22 13:07:11, Terry Bowman wrote:
> On 1/6/22 12:18 PM, Guenter Roeck wrote:
> > On Wed, Nov 03, 2021 at 11:15:20AM -0500, Terry Bowman wrote:

> >> diff --git a/drivers/watchdog/sp5100_tco.c b/drivers/watchdog/sp5100_tco.c
> >> index 80ae42ae7aaa..4777e672a8ad 100644
> >> --- a/drivers/watchdog/sp5100_tco.c
> >> +++ b/drivers/watchdog/sp5100_tco.c
> >> @@ -48,12 +48,14 @@
> >>  /* internal variables */
> >>  
> >>  enum tco_reg_layout {
> >> -	sp5100, sb800, efch
> >> +	sp5100, sb800, efch, efch_mmio
> >>  };
> >>  
> >>  struct sp5100_tco {
> >>  	struct watchdog_device wdd;
> >>  	void __iomem *tcobase;
> >> +	void __iomem *addr;
> >> +	struct resource *res;
> > 
> > I must admit that I really don't like this code. Both res and
> > addr are only used during initialization, yet their presence suggests
> > runtime usage. Any chance to reqork this to not require those variables ?

We did that in an earlier version, see struct efch_cfg of:

 https://patchwork.kernel.org/project/linux-watchdog/patch/20210813213216.54780-1-Terry.Bowman@amd.com/

The motivation of it was the same as you suggested to only use it
during init.

Having it in struct sp5100_tco made things simpler esp. in the
definition of the function interfaces where those new members are
used.

If that init vars are no longer in struct sp5100_tco then callers of
efch_read_pm_reg8() and efch_update_pm_reg8() will need to carry a
pointer to them. To avoid this I see those options:

1) Implement them as global (or a single global struct) and possibly
protect it by a mutex. There is only a single device anyway and we
wouldn't need a protection.

2) Have an own mmio implementation of tco_timer_enable() and/or
sp5100_tco_timer_init().

> Yes, v3 will include refactoring to remove 'res' and 'addr'. I will also 
> correct the trailing newline you mentioned in an earlier email.
> 
> Regards,
> Terry
> 
> >>  	enum tco_reg_layout tco_reg_layout;

While at it, tco_reg_layout is also only used during initialization
and can be moved there too. This would raise option 3:

3) Add a pointer of struct sp5100_tco to struct efch_cfg and use that
struct instead in init funtions only. But that causes the most rework
(which would be ok to me).

Going with 3) looks the cleanest way, I would try that. But all
options have its advantages.

-Robert

> >>  };