Am 2022-03-07 12:33, schrieb Gaurav Jain:
-----Original Message-----
From: Michael Walle <mich...@walle.cc>
Sent: Monday, March 7, 2022 4:39 PM
To: Gaurav Jain <gaurav.j...@nxp.com>
Cc: sba...@denx.de; Varun Sethi <v.se...@nxp.com>; Adrian Alonso
<adrian.alo...@nxp.com>; Alison Wang <alison.w...@nxp.com>; Andy Tang
<andy.t...@nxp.com>; feste...@gmail.com; Franck Lenormand
<franck.lenorm...@nxp.com>; Horia Geanta <horia.gea...@nxp.com>; Ji
Luo
<ji....@nxp.com>; ma...@denx.de; Meenakshi Aggarwal
<meenakshi.aggar...@nxp.com>; Mingkai Hu <mingkai...@nxp.com>;
olte...@gmail.com; Pankaj Gupta <pankaj.gu...@nxp.com>; Peng Fan
<peng....@nxp.com>; Pramod Kumar <pramod.kuma...@nxp.com>; Priyanka
Jain <priyanka.j...@nxp.com>; Rajesh Bhagat <rajesh.bha...@nxp.com>;
Sahil
Malhotra <sahil.malho...@nxp.com>; Shengzhou Liu
<shengzhou....@nxp.com>;
Silvano Di Ninno <silvano.dini...@nxp.com>; s...@chromium.org; u-
b...@lists.denx.de; dl-uboot-imx <uboot-...@nxp.com>; Wasim Khan
<wasim.k...@nxp.com>; Ye Li <ye...@nxp.com>
Subject: Re: [EXT] Re: [PATCH v8 01/15] crypto/fsl: Add support for
CAAM Job
ring driver model
Caution: EXT Email
Am 2022-03-07 11:56, schrieb Gaurav Jain:
>> -----Original Message-----
>> From: Michael Walle <mich...@walle.cc>
>> Sent: Monday, March 7, 2022 3:28 PM
>> To: sba...@denx.de
>> Cc: Varun Sethi <v.se...@nxp.com>; Adrian Alonso
>> <adrian.alo...@nxp.com>; Alison Wang <alison.w...@nxp.com>; Andy
Tang
>> <andy.t...@nxp.com>; feste...@gmail.com; Franck Lenormand
>> <franck.lenorm...@nxp.com>; Gaurav Jain <gaurav.j...@nxp.com>; Horia
>> Geanta <horia.gea...@nxp.com>; Ji Luo <ji....@nxp.com>;
>> ma...@denx.de; Meenakshi Aggarwal <meenakshi.aggar...@nxp.com>;
>> Mingkai Hu <mingkai...@nxp.com>; olte...@gmail.com; Pankaj Gupta
>> <pankaj.gu...@nxp.com>; Peng Fan <peng....@nxp.com>; Pramod Kumar
>> <pramod.kuma...@nxp.com>; Priyanka Jain <priyanka.j...@nxp.com>;
>> Rajesh Bhagat <rajesh.bha...@nxp.com>; Sahil Malhotra
>> <sahil.malho...@nxp.com>; Shengzhou Liu <shengzhou....@nxp.com>;
>> Silvano Di Ninno <silvano.dini...@nxp.com>; s...@chromium.org; u-
>> b...@lists.denx.de; dl-uboot-imx <uboot-...@nxp.com>; Wasim Khan
>> <wasim.k...@nxp.com>; Ye Li <ye...@nxp.com>; Michael Walle
>> <mich...@walle.cc>
>> Subject: Re: [EXT] Re: [PATCH v8 01/15] crypto/fsl: Add support for
>> CAAM Job ring driver model
>>
>> Caution: EXT Email
>>
>> > On 03.03.22 14:41, Gaurav Jain wrote:
>> >> As we have not received any response from imx6dl_mamoj board
maintainer.
>> >> I propose the below solution
>> >>
>> >> --- a/arch/arm/mach-imx/Kconfig
>> >> +++ b/arch/arm/mach-imx/Kconfig
>> >> @@ -49,8 +49,8 @@ config USE_IMXIMG_PLUGIN config IMX_HAB
>> >> - select FSL_CAAM if HAS_CAAM
>> >> - imply CMD_DEKBLOB if HAS_CAAM
>> >> + imply FSL_CAAM if HAS_CAAM
>> >> + imply CMD_DEKBLOB if FSL_CAAM
>> >> Help
>> >>
>> >
>> > IMO this is ok, I was also wrong, Marek is not the maintainer of
>> > this board. This was the only board with broken build - let's say,
>> > I will still wait a couple of days, and if there is no comments, I
>> > will apply your series (but then V10). I can apply this fix myself,
>> > no need to post the series again (I have not seen any other comment
>> > or request to
>> change).
>>
>> I don't understand why the solution isn't the same one as for the
>> layerscape part in this series[1]: enable the config per board (that
>> is your boards) and leave all others the same as before?
>
> imx6dl_mamoj caam driver is not enabled by any of my changes.
> This board is enabling IMX_HAB which select FSL_CAAM.
> Proposed changes making it imply so that FSL_CAAM can be disabled in
> board defconfig.
Ahh it was already selected before. But mhh, does IMX_HAB even makes
sense
without FSL_CAAM? Why was is a hard dependency before?
With imply, this will still enable FSL_CAAM unless it is explicitly
disabled in defconfig. With select I do not have choice of disabling
FSL_CAAM in defconfig. Now I have disabled FSL_CAAM only for
imx6dl_mamoj_defconfig, as it is reporting spl size issues with caam
driver model approach.
I can see *what* you are doing, but that doesn't answer *why* it was
a hard dependency before and why now of a sudden can be a soft
dependency.
-michael
