From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Le Bihan Date: Mon, 10 Feb 2014 14:44:15 +0100 Subject: [Buildroot] [PATCH v7 0/4] udev is now provided by systemd or eudev In-Reply-To: <52F5435A.3080700@mind.be> References: <1391779295-19140-1-git-send-email-eric.le.bihan.dev@free.fr> <52F5435A.3080700@mind.be> Message-ID: <20140210134415.GD22843@pc-eric> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Hi! On Fri, Feb 07, 2014 at 09:34:34PM +0100, Arnout Vandecappelle wrote: > > I have one more remark about the way the udev virtual package is > implemented. We found out for the opengl packages that there is a better > way to do it, something like: > > udev/Config.in: > > config BR2_PACKAGE_HAS_UDEV > bool > > config BR2_PACKAGE_PROVIDES_UDEV > depends on BR2_PACKAGE_HAS_UDEV > string > > > udev/udev.mk: > > UDEV_SRC = > UDEV_DEPENDENCIES = $(call qstrip,$(BR2_PACKAGE_PROVIDES_UDEV)) > > ifeq ($(UDEV_DEPENDENCIES),) > define UDEV_CONFIGURE_CMDS > echo "No Udev implementation selected. Configuration error." > exit 1 > endef > endif > > > eudev/Config.in: > > ... > if BR2_PACKAGE_EUDEV > config BR2_PACKAGE_PROVIDES_UDEV > default "eudev" > endif > > > > That moves all the provider logic to the provider itself, which for > instance makes it possible for BR2_EXTERNAL to provide an alternative > udev implementation (e.g. a different udev version, which could be _very_ > relevant). > > > I repeat, though: this can be done as a follow-up patch. Yes, the word 'provides' is more explicit. I'll put this in a follow-up patch. If this is the definitive way of implementing a virtual package, shouldn't it be added to docs/manual/adding-packages-tips.txt? Thanks for the review. Best regards, ELB