From: slash.tmp@free•fr (Mason)
To: linux-arm-kernel@lists•infradead.org
Subject: arasan,sdhci.txt "compatibility" DT binding
Date: Fri, 18 Nov 2016 11:37:20 +0100 [thread overview]
Message-ID: <582ED9E0.5090506@free.fr> (raw)
In-Reply-To: <56B21DDB.6050708@free.fr>
On 03/02/2016 16:33, Mason wrote:
> On 03/02/2016 16:21, S?ren Brinkmann wrote:
>> On Wed, 2016-02-03 at 10:58:24 +0100, Michal Simek wrote:
>>> On 3.2.2016 09:31, Mason wrote:
>>>> On 03/02/2016 08:20, Michal Simek wrote:
>>>>
>>>>> On 3.2.2016 03:33, Shawn Lin wrote:
>>>>>> + Michal, S?ren Brinkmann
>>>>>>
>>>>>> On 2016/2/2 17:49, Mason wrote:
>>>>>>> Hello everyone,
>>>>>>>
>>>>>>> Documentation/devicetree/bindings/mmc/arasan,sdhci.txt states:
>>>>>>>
>>>>>>> Required Properties:
>>>>>>> - compatible: Compatibility string. Must be 'arasan,sdhci-8.9a' or
>>>>>>> 'arasan,sdhci-4.9a' or 'arasan,sdhci-5.1'
>>>>>>>
>>>>>>> What do 8.9a, 4.9a, and 5.1 refer to?
>>>>>>>
>>>>>>
>>>>>> Good question.
>>>>>>
>>>>>> Michal told me that 8.9a and 4.9a came from Xilinx
>>>>>> databook which define their available arasan controller to be version
>>>>>> 4.9a and 8.9a.
>>>>>
>>>>> Our version is coming from here.
>>>>> http://www.xilinx.com/support/documentation/user_guides/ug585-Zynq-7000-TRM.pdf
>>>>> page 28
>>>>>
>>>>> Datasheets from 2012 and 2010 doesn't look too recent and I expect that
>>>>> Arasan did a lot of work from that time that's why I am not surprised
>>>>> that you are not able to see that versions.
>>>>
>>>> Hello Michal,
>>>>
>>>> I'm even more confused.
>>>>
>>>> Arasan's 2010-02-19 data sheet is titled
>>>> SD3.0 / SDIO3.0 / eMMC4.4 AHB Host Controller
>>>>
>>>> Page 28 of the Xilinx data sheet mentions
>>>> SD2.0 / SDIO2.0 / MMC3.31 AHB Host Controller
>>>> Version 8.9A_apr02nd_2010
>>>>
>>>> It does not make sense that Arasan would support SD3.0 in February,
>>>> then go back to SD 2.0 in April.
>>>>
>>>> I do note eMMC4.4 vs MMC3.31 => perhaps these are two *different*
>>>> IP blocks?
>>>>
>>>> Do your data sheets come with revision history?
>>>
>>> I don't have datasheet for this IP in my hand that's why I can't check it.
>>> But I can't see any problem with it. Our zynq SoC supports SD2.0 and it
>>> was requirement at that time. Bugs can happen. Arasan fixed it and
>>> create new version.
>>> At the same time can have 3.0 versions but vendor is just decide not to
>>> use it for whatever reason.
>>>
>>> That's why timing of features and versions can upgrade any time and
>>> unfortunately bugs happen.
>>
>> We have several Arasan data sheets here. The document names are:
>> "SD2.0 / SDIO2.0 / MMC3.31 AHB Userguide" and the revisions are 9.2a,
>> 4.4a and 5.4a. I have the feeling that the document revisions have
>> been mistaken as the IP revision. I cannot find any other indicator
>> for the IP revision though. Does the IP have a way to discover its
>> revision?
>
> To summarize, it looks to me like
> 4.9a and 8.9a are documentation revision numbers for this IP:
> "SD2.0 / SDIO2.0 / MMC3.31 AHB Host Controller"
>
> and 5.1 seems to be the eMMC standard, so one of these IP:
> http://arasan.com/products/emmc51/sd3-emmc-5-1-host/
> http://arasan.com/products/emmc51/emmc-5-1-sd-4-1-host/
>
> Whereas my board is using still another IP:
> "SD3.0 / SDIO3.0 / eMMC4.4 AHB Host Controller"
>
>
> If that is correct, then I should be able to use the 8.9a
> compatible string, and perhaps my hardware will work in
> slightly degraded performance mode, from not using the
> newest protocol gizmos available.
Resurrecting this thread after a chat with Michal on IRC.
The driver now supports a few more compatible strings.
Tracing the history...
Soren/Xilinx defined "arasan,sdhci-8.9a" in e3ec3a3d11ad
They have the "SD2.0/SDIO2.0/MMC3.31 AHB Host Controller" version
(8.9a might be a document revision number, dated 2010-04-02)
Suman/APM added "arasan,sdhci-4.9a" in 308f3f8d8112
@Suman, @Rameshwar: what specific IP block does your SoC embed?
What does 4.9a refer to? The documentation revision number?
Shawn/Rockchip added "arasan,sdhci-5.1" in da795ec26e25
This seems to be for an *actual* eMMC 5.1 version
(instead of a documentation version) as mentioned in
the commit message.
Douglas added "rockchip,rk3399-sdhci-5.1"
and made several other improvements.
I have reached out to Arasan support. Hopefully they can also
help clear up the confusion, assuming they care about the
Linux driver.
Regards.
next parent reply other threads:[~2016-11-18 10:37 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <56B07BA5.4060407@free.fr>
[not found] ` <56B166F1.5080605@rock-chips.com>
[not found] ` <56B1AA44.1010108@xilinx.com>
[not found] ` <56B1BAF8.2050907@free.fr>
[not found] ` <56B1CF40.1030203@xilinx.com>
[not found] ` <20160203152104.GP4215@xsjsorenbubuntu>
[not found] ` <56B21DDB.6050708@free.fr>
2016-11-18 10:37 ` Mason [this message]
2016-11-18 10:49 ` arasan,sdhci.txt "compatibility" DT binding Rameshwar Sahu
2016-11-18 12:42 ` Mason
2016-11-18 13:22 ` Michal Simek
2016-11-28 15:44 ` Mason
2016-11-28 16:15 ` Arnd Bergmann
2016-11-28 16:52 ` Mason
2016-11-29 7:29 ` Rameshwar Sahu
2016-11-30 10:51 ` Sebastian Frias
2016-11-30 13:17 ` Michal Simek
2016-11-28 16:23 ` Sebastian Frias
2016-12-01 4:09 ` Shawn Lin
2016-12-01 12:29 ` Mason
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=582ED9E0.5090506@free.fr \
--to=slash.tmp@free$(echo .)fr \
--cc=linux-arm-kernel@lists$(echo .)infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox