All of lore.kernel.org
 help / color / mirror / Atom feed
From: Rob Herring <robherring2@gmail.com>
To: Kumar Gala <galak@codeaurora.org>
Cc: Olof Johansson <olof@lixom.net>,
	David Brown <davidb@codeaurora.org>,
	Pawel Moll <pawel.moll@arm.com>,
	Mark Rutland <mark.rutland@arm.com>,
	Stephen Warren <swarren@wwwdotorg.org>,
	Ian Campbell <ijc+devicetree@hellion.org.uk>,
	Russell King <linux@arm.linux.org.uk>,
	"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
	"linux-kernel@vger.kernel.org list"
	<linux-kernel@vger.kernel.org>,
	Rohit Vaswani <rvaswani@codeaurora.org>,
	linux-arm-msm@vger.kernel.org,
	"linux-arm-kernel@lists.infradead.org"
	<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH] ARM: msm: Move msm devicetrees under a Qualcomm dir
Date: Thu, 12 Sep 2013 13:20:17 -0500	[thread overview]
Message-ID: <523205E1.2040704@gmail.com> (raw)
In-Reply-To: <9DD9D9A7-ED55-412F-A18D-599D1E365003@codeaurora.org>

On 09/12/2013 01:15 PM, Kumar Gala wrote:
> 
> On Sep 12, 2013, at 1:04 PM, Olof Johansson wrote:
> 
>> On Thu, Sep 12, 2013 at 10:55 AM, Kumar Gala <galak@codeaurora.org>
>> wrote:
>>> 
>>> On Sep 12, 2013, at 12:06 PM, Olof Johansson wrote:
>>> 
>>>> On Thu, Sep 12, 2013 at 9:53 AM, Kumar Gala
>>>> <galak@codeaurora.org> wrote:
>>>>> 
>>>>> On Sep 12, 2013, at 11:46 AM, Olof Johansson wrote:
>>>>> 
>>>>>> On Thu, Sep 12, 2013 at 9:37 AM, Kumar Gala
>>>>>> <galak@codeaurora.org> wrote:
>>>>>>> Use the Qualcomm vendor prefix (qcom) as the directory
>>>>>>> name for Qualcomm MSM devicetrees going forward.
>>>>>>> 
>>>>>>> Signed-off-by: Kumar Gala <galak@codeaurora.org>
>>>>>> 
>>>>>> Let's not move just one platform like this. If we are to do
>>>>>> this, we should move everything, and that will be really
>>>>>> painful and needs to be done in a controlled manner,
>>>>>> probably scripted and right before a -rc1 or such.
>>>>> 
>>>>> 
>>>>> Than I suggest we deal with it when we pull the device trees
>>>>> out of the kernel tree.
>>>>> 
>>>>> As such, I'd tell Rohit to go forward with the file being
>>>>> named apq8074-dragonboard.dtb for the time being.
>>>> 
>>>> My original request to please use a common prefix for your
>>>> product families stands. Please prefix with msm-*, or if you
>>>> have to, qcom-* instead, since you guys can't seem to make your
>>>> mind up on standard prefixes (msm, apq, etc).
>>> 
>>> This is silly, I dont see the reason to go with
>>> qcom-apq<SOC>-<BOARD>.dts and than in the future drop qcom- when
>>> we mostly likely shift to a dir structure.
>>> 
>>> As engineers we are all too aware of the lack of sanity in
>>> marketing names, but its what we have so we have to live with
>>> it.
>> 
>> And we all have a choice whether we let the marketing people's 
>> insanity spread into our engineering projects, or if we keep it as 
>> sane as possible in spite of them.
>> 
>> I wouldn't have an objection here if there was some sort of
>> rationale between what "apq" and "msm" means. But it seems like
>> qualcomm rolls a dice and decides if a platform will have one name
>> or the other. Dragonboard dmesg says msm<foo>. DTS file for the
>> same board says apq. DTSI file says one thing, overridden by the
>> dts to something else. Total chaos.
>> 
>> I would be fine with adding two instead of one (after all,
>> platforms like TI has this for AM* vs OMAP*, etc), but there _has_
>> to be some sort of consistency or you might just as well assign a
>> random string as name.
>> 
>> So, if you can't come up with a reasonable, rational and
>> consistent naming scheme (which, apparantly, you can't since your
>> marketing guys are in control of this and they don't get it right),
>> then at least prefix with a common string for the platform. That's
>> all I'm asking.
>> 
>> 
>> -Olof
> 
> I'm not sure get why the insanity of naming impacts anything, those
> people that care about the platforms know what names mean what.  Plus
> we have this same set of having to be in the know for other SoC
> already.  Freescale does this all the time in PPC land, they just do
> it with numbers and not prefixes.
> 
> I suggested doing a qcom/ dir, as it addresses your concern, of
> putting all of our mess in one place.  Just because every other
> SoC/mach/plat hasn't done that should stop us from at least doing it
> for qcom.
> 

As Stephen W has pointed out previously, the dtb filename itself is or
may be an ABI and the bootloader may be hardcoded to a name. So we
should avoid future renames.

Rob

WARNING: multiple messages have this Message-ID (diff)
From: robherring2@gmail.com (Rob Herring)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH] ARM: msm: Move msm devicetrees under a Qualcomm dir
Date: Thu, 12 Sep 2013 13:20:17 -0500	[thread overview]
Message-ID: <523205E1.2040704@gmail.com> (raw)
In-Reply-To: <9DD9D9A7-ED55-412F-A18D-599D1E365003@codeaurora.org>

On 09/12/2013 01:15 PM, Kumar Gala wrote:
> 
> On Sep 12, 2013, at 1:04 PM, Olof Johansson wrote:
> 
>> On Thu, Sep 12, 2013 at 10:55 AM, Kumar Gala <galak@codeaurora.org>
>> wrote:
>>> 
>>> On Sep 12, 2013, at 12:06 PM, Olof Johansson wrote:
>>> 
>>>> On Thu, Sep 12, 2013 at 9:53 AM, Kumar Gala
>>>> <galak@codeaurora.org> wrote:
>>>>> 
>>>>> On Sep 12, 2013, at 11:46 AM, Olof Johansson wrote:
>>>>> 
>>>>>> On Thu, Sep 12, 2013 at 9:37 AM, Kumar Gala
>>>>>> <galak@codeaurora.org> wrote:
>>>>>>> Use the Qualcomm vendor prefix (qcom) as the directory
>>>>>>> name for Qualcomm MSM devicetrees going forward.
>>>>>>> 
>>>>>>> Signed-off-by: Kumar Gala <galak@codeaurora.org>
>>>>>> 
>>>>>> Let's not move just one platform like this. If we are to do
>>>>>> this, we should move everything, and that will be really
>>>>>> painful and needs to be done in a controlled manner,
>>>>>> probably scripted and right before a -rc1 or such.
>>>>> 
>>>>> 
>>>>> Than I suggest we deal with it when we pull the device trees
>>>>> out of the kernel tree.
>>>>> 
>>>>> As such, I'd tell Rohit to go forward with the file being
>>>>> named apq8074-dragonboard.dtb for the time being.
>>>> 
>>>> My original request to please use a common prefix for your
>>>> product families stands. Please prefix with msm-*, or if you
>>>> have to, qcom-* instead, since you guys can't seem to make your
>>>> mind up on standard prefixes (msm, apq, etc).
>>> 
>>> This is silly, I dont see the reason to go with
>>> qcom-apq<SOC>-<BOARD>.dts and than in the future drop qcom- when
>>> we mostly likely shift to a dir structure.
>>> 
>>> As engineers we are all too aware of the lack of sanity in
>>> marketing names, but its what we have so we have to live with
>>> it.
>> 
>> And we all have a choice whether we let the marketing people's 
>> insanity spread into our engineering projects, or if we keep it as 
>> sane as possible in spite of them.
>> 
>> I wouldn't have an objection here if there was some sort of
>> rationale between what "apq" and "msm" means. But it seems like
>> qualcomm rolls a dice and decides if a platform will have one name
>> or the other. Dragonboard dmesg says msm<foo>. DTS file for the
>> same board says apq. DTSI file says one thing, overridden by the
>> dts to something else. Total chaos.
>> 
>> I would be fine with adding two instead of one (after all,
>> platforms like TI has this for AM* vs OMAP*, etc), but there _has_
>> to be some sort of consistency or you might just as well assign a
>> random string as name.
>> 
>> So, if you can't come up with a reasonable, rational and
>> consistent naming scheme (which, apparantly, you can't since your
>> marketing guys are in control of this and they don't get it right),
>> then at least prefix with a common string for the platform. That's
>> all I'm asking.
>> 
>> 
>> -Olof
> 
> I'm not sure get why the insanity of naming impacts anything, those
> people that care about the platforms know what names mean what.  Plus
> we have this same set of having to be in the know for other SoC
> already.  Freescale does this all the time in PPC land, they just do
> it with numbers and not prefixes.
> 
> I suggested doing a qcom/ dir, as it addresses your concern, of
> putting all of our mess in one place.  Just because every other
> SoC/mach/plat hasn't done that should stop us from at least doing it
> for qcom.
> 

As Stephen W has pointed out previously, the dtb filename itself is or
may be an ABI and the bootloader may be hardcoded to a name. So we
should avoid future renames.

Rob

  reply	other threads:[~2013-09-12 18:20 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-09-12 16:37 [PATCH] ARM: msm: Move msm devicetrees under a Qualcomm dir Kumar Gala
2013-09-12 16:37 ` Kumar Gala
2013-09-12 16:46 ` Olof Johansson
2013-09-12 16:46   ` Olof Johansson
2013-09-12 16:53   ` Kumar Gala
2013-09-12 16:53     ` Kumar Gala
2013-09-12 17:06     ` Olof Johansson
2013-09-12 17:06       ` Olof Johansson
2013-09-12 17:55       ` Kumar Gala
2013-09-12 17:55         ` Kumar Gala
2013-09-12 18:04         ` Olof Johansson
2013-09-12 18:04           ` Olof Johansson
2013-09-12 18:15           ` Kumar Gala
2013-09-12 18:15             ` Kumar Gala
2013-09-12 18:15             ` Kumar Gala
2013-09-12 18:20             ` Rob Herring [this message]
2013-09-12 18:20               ` Rob Herring
2013-09-12 18:46               ` Olof Johansson
2013-09-12 18:46                 ` Olof Johansson
2013-09-12 19:26                 ` Kumar Gala
2013-09-12 19:26                   ` Kumar Gala
     [not found]         ` <3E38F48E-EE41-4564-84A1-36CF07B08811-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
2013-09-12 22:47           ` David Brown
2013-09-12 22:47             ` David Brown
2013-09-12 22:47             ` David Brown
2013-09-13  2:05             ` Kumar Gala
2013-09-13  2:05               ` Kumar Gala
     [not found]               ` <9EA264FE-7621-458D-8333-9340DDBFC4EE-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
2013-09-16 18:28                 ` Rohit Vaswani
2013-09-16 18:28                   ` Rohit Vaswani
2013-09-16 18:28                   ` Rohit Vaswani
     [not found]                   ` <52374DD1.9040209-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org>
2013-09-16 18:46                     ` Kumar Gala
2013-09-16 18:46                       ` Kumar Gala
2013-09-16 18:46                       ` Kumar Gala
2013-09-16 18:50                       ` Olof Johansson
2013-09-16 18:50                         ` Olof Johansson
     [not found]                         ` <CAOesGMgCvnuX918pt6vT0_QsHdfJ8b813iu2tXRHvdpL8GZ3dg-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2013-09-17 15:52                           ` Kumar Gala
2013-09-17 15:52                             ` Kumar Gala
2013-09-17 15:52                             ` Kumar Gala

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=523205E1.2040704@gmail.com \
    --to=robherring2@gmail.com \
    --cc=davidb@codeaurora.org \
    --cc=devicetree@vger.kernel.org \
    --cc=galak@codeaurora.org \
    --cc=ijc+devicetree@hellion.org.uk \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux@arm.linux.org.uk \
    --cc=mark.rutland@arm.com \
    --cc=olof@lixom.net \
    --cc=pawel.moll@arm.com \
    --cc=rvaswani@codeaurora.org \
    --cc=swarren@wwwdotorg.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.