[PATCH v5 05/21] arm: mvebu: x530: Disable smart hwcaps selection

[Pratyush Yadav]

On 05/06/20 09:50AM, Tan, Ley Foon wrote:
> 
> 
> > -----Original Message-----
> > From: Pratyush Yadav <p.yadav at ti.com>
> > Sent: Friday, June 5, 2020 5:48 PM
> > To: Chris Packham <Chris.Packham at alliedtelesis.co.nz>
> > Cc: Jagan Teki <jagan at amarulasolutions.com>; Vignesh R
> > <vigneshr at ti.com>; Ryder Lee <ryder.lee at mediatek.com>; Weijie Gao
> > <weijie.gao at mediatek.com>; GSS_MTK_Uboot_upstream
> > <GSS_MTK_Uboot_upstream at mediatek.com>; u-boot at lists.denx.de;
> > Sekhar Nori <nsekhar at ti.com>; Tan, Ley Foon <ley.foon.tan at intel.com>
> > Subject: Re: [PATCH v5 05/21] arm: mvebu: x530: Disable smart hwcaps
> > selection
> > 
> > Hi Chris,
> > 
> > On 04/06/20 08:47PM, Chris Packham wrote:
> > >
> > > On 5/06/20 12:44 am, Pratyush Yadav wrote:
> > > > The option SPI_FLASH_SMART_HWCAPS will be introduced in a future
> > commit.
> > > > It is enabled by default. It updates the hwcaps selection of SPI NOR
> > > > to use the SPI MEM's supports_op() hook. But this leads to a code
> > > > size increase and so the SPL binary exceeds the size limit.
> > > >
> > > > So, use the old hwcaps selection logic here to make sure the SPL
> > > > size does not exceed the limit.
> > > >
> > > > Signed-off-by: Pratyush Yadav <p.yadav at ti.com>
> > > > ---
> > > >   configs/x530_defconfig | 1 +
> > > >   1 file changed, 1 insertion(+)
> > > >
> > > > diff --git a/configs/x530_defconfig b/configs/x530_defconfig index
> > > > 67c8fbf663..35e2d53285 100644
> > > > --- a/configs/x530_defconfig
> > > > +++ b/configs/x530_defconfig
> > > > @@ -58,6 +58,7 @@ CONFIG_SYS_NAND_USE_FLASH_BBT=y
> > > >   CONFIG_NAND_PXA3XX=y
> > > >   CONFIG_SF_DEFAULT_BUS=1
> > > >   CONFIG_SF_DEFAULT_SPEED=50000000
> > > > +CONFIG_SPI_FLASH_SMART_HWCAPS=n
> > >
> > > Last time I checked =n wasn't valid Kconfig for this to have the
> > > desired effect wouldn't it need to be
> > 
> > So I've been doing it wrong all along, huh! Setting a config to 'n' does lead to
> > it getting converted to the "not set" form in .config, so that probably helped
> > in fooling me. Will fix. Thanks.
> > 
> Did you run "make savedefconfig" command to generate defconfig?
> Shouldn't edit defconfig manually.

No I didn't. Thanks for the tip.

-- 
Regards,
Pratyush Yadav
Texas Instruments India