From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stefan Roese <sr@denx.de>
Date: Tue, 3 Nov 2015 08:22:56 +0100
Subject: [U-Boot] [PATCH v3 2/3] cfi_flash: convert to driver model
In-Reply-To: <56385371.3010305@wytron.com.tw>
References: <1444289667-23775-1-git-send-email-thomas@wytron.com.tw>
 <1446212027-20630-1-git-send-email-thomas@wytron.com.tw>
 <1446212027-20630-2-git-send-email-thomas@wytron.com.tw>
 <56371CE1.9000203@denx.de> <5637FE76.3050208@wytron.com.tw>
 <56384C9C.6030709@denx.de> <56385371.3010305@wytron.com.tw>
Message-ID: <563860D0.1070206@denx.de>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Hi Thomas,

On 03.11.2015 07:25, Thomas Chou wrote:
> On 2015?11?03? 13:56, Stefan Roese wrote:
>>> if (CONFIG_IS_ENABLED(CFI_FLASH)) {
>>>      struct udevice *dev;
>>>      cfi_flash_num_flash_banks = 0;
>>>      /* probe every MTD device to find cfi-flash device */
>>>      for (uclass_first_device(UCLASS_MTD, &dev);
>>>           dev;
>>>           uclass_next_device(&dev)) {
>>>      }
>>> }
>>
>> I'm still not an expert in the DM internals. I fail to see, where
>> the actual probing is happening in the loop. Is it buried in the
>> uclass_first_device() call?
>
> The probing is happening in both uclass_first_device() and
> uclass_next_device().
>
> The uclass_first_device() will probe the first device.
> While uclass_next_device() will probe the rest if they exist.

Okay. So cfi_flash_probe() will get called assigning the base
addresses that are available. The real "probing" (meaning
detection) is happening later in flash_init().

I think adding (or extending) a comment here to describe this
would be helpful in v2.

Thanks,
Stefan