Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Arnout Vandecappelle <arnout@mind.be>
To: buildroot@busybox.net
Subject: [Buildroot] [PATCH 08/20] packages: add infrastructure for virtual packages
Date: Fri, 04 Apr 2014 22:10:50 +0200	[thread overview]
Message-ID: <533F11CA.5040304@mind.be> (raw)
In-Reply-To: <19c0bbf58268a2133297840b8763d415ad5a4ef0.1394482605.git.yann.morin.1998@free.fr>

On 10/03/14 21:27, Yann E. MORIN wrote:
> From: "Yann E. MORIN" <yann.morin.1998@free.fr>
> 
> The virtual-package infrastructure allows to easily define a
> virtual package in a single line:
> 
>     $ cat package/some-virtual-package/some-virtual-package.mk
>     $(eval $(virtual-package))
> 
> And that's all. :-)
> 
> Thanks to ?ric for his work on the manual, that prompted the
> idea for this virtual-package infrastructure! ;-)
> 
> Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
> Cc: Eric Le Bihan <eric.le.bihan.dev@free.fr>
> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
> Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
> Cc: Mike Zick <minimod@morethan.org>
> Reviewed-by: Samuel Martin <s.martin49@gmail.com>
> ---
>  package/Makefile.in    |  1 +
>  package/pkg-virtual.mk | 66 ++++++++++++++++++++++++++++++++++++++++++++++++++
>  2 files changed, 67 insertions(+)
>  create mode 100644 package/pkg-virtual.mk
> 
> diff --git a/package/Makefile.in b/package/Makefile.in
> index 454f614..c5cdf5a 100644
> --- a/package/Makefile.in
> +++ b/package/Makefile.in
> @@ -362,4 +362,5 @@ include package/pkg-cmake.mk
>  include package/pkg-luarocks.mk
>  include package/pkg-perl.mk
>  include package/pkg-python.mk
> +include package/pkg-virtual.mk
>  include package/pkg-generic.mk
> diff --git a/package/pkg-virtual.mk b/package/pkg-virtual.mk
> new file mode 100644
> index 0000000..9b206ea
> --- /dev/null
> +++ b/package/pkg-virtual.mk
> @@ -0,0 +1,66 @@
> +################################################################################
> +# Virtual package infrastructure
> +#
> +# This file implements an infrastructure that eases development of
> +# package .mk files for virtual packages. It should be used for all
> +# virtual packages.
> +#
> +# See the Buildroot documentation for details on the usage of this
> +# infrastructure
> +#
> +# In terms of implementation, this virtual infrastructure requires
> +# the .mk file to only call the 'virtual-package' macro.
> +#
> +################################################################################
> +
> +
> +################################################################################
> +# inner-virtual-package -- defines the dependency rules of the virtual
> +# package against its provider.
> +#
> +#  argument 1 is the lowercase package name
> +#  argument 2 is the uppercase package name, including an HOST_ prefix
> +#             for host packages
> +#  argument 3 is the uppercase package name, without the HOST_ prefix
> +#             for host packages
> +#  argument 4 is the type (target or host)
> +################################################################################
> +
> +# Note: putting this comment here rather than in the define block, otherwise
> +# make would try to expand the $(error ...) in the comment, which is not
> +# really what we want.

 That could be avoided by $$-quoting it inside the comment as well :-)

> +# We need to use second-expansion for the $(error ...) call, below,
> +# so it is not evaluated now, but as part of the generated make code.
> +
> +define inner-virtual-package
> +
> +# Ensure the virtual package has an implementation defined.
> +ifeq ($(BR2_PACKAGE_HAS_$(2)),y)

 Even though it is not strictly necessary (it will anyway be eval'ed
immediately after expansion), I think it's better to consistently use $$
everywhere (except for $(1) etc. of course). Currently it's not done
consistently in package-generic but it would be better to make it
consistent here.

> +ifeq ($(call qstrip,$(BR2_PACKAGE_PROVIDES_$(2))),)
> +$$(error No implementation selected for virtual package $(1). Configuration error)

 I think "Configuration error" is a bit vague, but I can't think of
anything better.

> +endif
> +endif
> +
> +# A virtual package does not have any source associated
> +$(2)_SOURCE =
> +
> +# This must be repeated from inner-generic-package, otherwise we get an empty
> +# _DEPENDENCIES
> +$(2)_DEPENDENCIES ?= $(filter-out host-toolchain $(1),\
> +	$(patsubst host-host-%,host-%,$(addprefix host-,$($(3)_DEPENDENCIES))))
> +
> +# Add dependency against the provider
> +$(2)_DEPENDENCIES += $(call qstrip,$(BR2_PACKAGE_PROVIDES_$(2)))
> +
> +# Call the generic package infrastructure to generate the necessary
> +# make targets
> +$(call inner-generic-package,$(1),$(2),$(3),$(4))

 I don't really see the need to call inner-generic-package here. Isn't it
sufficient to do:

ifeq ($$(BR2_PACKAGE_HAS_$(2)),y)
ifeq ($$(call qstrip,$$(BR2_PACKAGE_PROVIDES_$(2))),)
$$(error No implementation selected for virtual package $(1).
Configuration error)
endif
endif

$(1): $$(call qstrip,$$(BR2_PACKAGE_PROVIDES_$(2))

 We don't need all the rest of the package infrastructure, do we? We
don't need to download any sources, we don't need legal info for the
virtual package, etc. etc. There's also no need to create stamp files


> +
> +endef
> +
> +################################################################################
> +# virtual-package -- the target generator macro for virtual packages
> +################################################################################
> +
> +virtual-package = $(call inner-virtual-package,$(pkgname),$(call UPPERCASE,$(pkgname)),$(call UPPERCASE,$(pkgname)),target)
> +host-virtual-package = $(call inner-virtual-package,host-$(pkgname),$(call UPPERCASE,host-$(pkgname)),$(call UPPERCASE,$(pkgname)),host)

 Note that with the simplification above, you only need the first two
arguments to the inner-virtual-package macro.

 Regards,
 Arnout


-- 
Arnout Vandecappelle                          arnout at mind be
Senior Embedded Software Architect            +32-16-286500
Essensium/Mind                                http://www.mind.be
G.Geenslaan 9, 3001 Leuven, Belgium           BE 872 984 063 RPR Leuven
LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle
GPG fingerprint:  7CB5 E4CC 6C2E EFD4 6E3D A754 F963 ECAB 2450 2F1F

  reply	other threads:[~2014-04-04 20:10 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-10 20:27 [Buildroot] [PATCH 0/20 v4] Add new virtual-package infrastructure (branch yem/virtual-packages) Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 01/20] package/libgles: rename the _HAS and _PROVIDES variables Yann E. MORIN
2014-04-04 16:41   ` Arnout Vandecappelle
2014-04-04 20:23     ` Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 02/20] package/libegl: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 03/20] package/libopenmax: " Yann E. MORIN
2014-04-04 16:41   ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 04/20] package/libopenvg: " Yann E. MORIN
2014-04-04 16:43   ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 05/20] package/luainterpreter: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 06/20] package/lua: rename config options Yann E. MORIN
2014-04-04 16:48   ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 07/20] manual: add virtual package tutorial Yann E. MORIN
2014-04-04 16:54   ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 08/20] packages: add infrastructure for virtual packages Yann E. MORIN
2014-04-04 20:10   ` Arnout Vandecappelle [this message]
2014-04-05 14:41     ` Yann E. MORIN
2014-04-06  9:01       ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 09/20] manual: update the virtual package section with the new infrastructure Yann E. MORIN
2014-03-12 11:59   ` Eric Le Bihan
2014-04-04 20:18   ` Arnout Vandecappelle
2014-03-10 20:27 ` [Buildroot] [PATCH 10/20] package/powervr: convert to the virtual-package infrastructure Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 11/20] package/libgles: " Yann E. MORIN
2014-03-12 12:15   ` Eric Le Bihan
2014-03-10 20:27 ` [Buildroot] [PATCH 12/20] package/libegl: " Yann E. MORIN
2014-03-12 12:18   ` Eric Le Bihan
2014-03-10 20:27 ` [Buildroot] [PATCH 13/20] package/libopenmax: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 14/20] package/libopenvg: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 15/20] package/luainterpreter: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 16/20] package/jpeg: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 17/20] package/cryptodev: " Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 18/20] package/udev: " Yann E. MORIN
2014-03-12 12:20   ` Eric Le Bihan
2014-03-10 20:27 ` [Buildroot] [PATCH 19/20] virtual-package: fake a version string for virtual packages Yann E. MORIN
2014-03-10 20:27 ` [Buildroot] [PATCH 20/20] FOO: tentative target and host virt packages Yann E. MORIN

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=533F11CA.5040304@mind.be \
    --to=arnout@mind.be \
    --cc=buildroot@busybox.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox