From: Tom Rini Date: Mon, 9 Jul 2018 09:10:27 -0400 Subject: [U-Boot] [PATCH] cmd: ubi: change 'default y' for SUNXI to 'imply' in Kconfig In-Reply-To: References: <1531116757-5995-1-git-send-email-yamada.masahiro@socionext.com> <20180709072019.qemz7wx5gr57phwy@flea> Message-ID: <20180709131027.GI27264@bill-the-cat> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On Mon, Jul 09, 2018 at 08:14:28PM +0900, Masahiro Yamada wrote: > 2018-07-09 16:20 GMT+09:00 Maxime Ripard : > > On Mon, Jul 09, 2018 at 03:12:37PM +0900, Masahiro Yamada wrote: > >> It is not preferred to put SUNXI-specific code in the common place. > > > > I think Tom has been saying the exact opposite a number of times, > > since there's some circular dependencies issues that arise when using > > imply. > > > Platforms select/imply necessary drivers or features. > This is what we usually do. > > > ARCH_SUNXI already uses select / imply in arch/arm/Kconfig. > (Also ARCH_SUNXI does so in arch/arm/mach-sunxi/Kconfig in Linux Kernel) > > > On the other hand, no drivers or features select / imply ARCH_SUNXI. > > > To sum up, select or imply should be used > in the following direction. > > Platform ---(select or imply)---> features > > > If you follow this rule, > circular dependency will not happen. I think what ends up being confusing about this is that we must do: default 0xDEADBEEF if ARCH_FOO as we cannot: imply FOO_OPTION 0xDEADBEEF if FOO_DRIVER -- Tom -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 819 bytes Desc: not available URL: