From mboxrd@z Thu Jan 1 00:00:00 1970 From: John Schmoller Date: Wed, 04 Nov 2009 14:37:44 -0600 Subject: [U-Boot] [RFC] env: Group environment variables In-Reply-To: <200911041536.02492.vapier@gentoo.org> References: <1257352452-11748-1-git-send-email-jschmoller@xes-inc.com> <200911041255.42023.vapier@gentoo.org> <1257358632.8937.1160.camel@johns> <200911041536.02492.vapier@gentoo.org> Message-ID: <1257367064.8937.1161.camel@johns> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On Wed, 2009-11-04 at 16:36 -0400, Mike Frysinger wrote: > On Wednesday 04 November 2009 13:17:12 John Schmoller wrote: > > On Wed, 2009-11-04 at 13:55 -0400, Mike Frysinger wrote: > > > On Wednesday 04 November 2009 11:34:12 John Schmoller wrote: > > > > This patch groups environment variables using a non-invasive protocol. > > > > Grouping is achieved by setting a "grouping" variable to a string of > > > > variables, and setting the master grouping variable, "env_groups" to > > > > the list of these grouping variables. > > > > > > > > For instance, > > > > setenv net ipaddr netmask gatewayip serverip > > > > setenv boot bootcmd bootdelay bootargs > > > > setenv env_groups net boot > > > > > > > > would print 4 variables grouped under net, 3 variables grouped under > > > > boot, and the rest of the variables grouped under "other". If > > > > env_groups is not defined, print behaves normally. > > > > > > > > Signed-off-by: John Schmoller > > > > --- > > > > I'm interesetd in seeing peoples opinions of this implementation of > > > > grouping environment variables. My major concerns about this > > > > implementation are > > > > > > my main concern is bloat. while i guess it would make nicer `printenv`, > > > i'm not going to use it. so please put it behind a config option so it > > > doesnt waste space on all boards. > > > > I can certainly do that. It is a requested feature on the U-Boot task > > list[1], so I didn't think that was needed, but I'll do it if that's > > what people want. > > sure, i'm not saying people other than me dont want it, just that i dont think > it's something you can call "core required functionality". Wolfgang's call of > course, but i'd like to see it behind a CONFIG (preferably disabled by > default). Will do. I'll wait to see if anyone has any other comments about implementation before resending. John