Buildroot

From eLinux.org
Revision as of 10:30, 7 October 2015 by ThomasPetazzoni (talk | contribs) (Developer days)
Jump to: navigation, search

Buildroot is a nice, simple, and efficient embedded Linux build system.

Important links

Developer days

Future:

  • Buildroot Developer Days, 1-2 February 2016, Brussels, Belgium, after FOSDEM. Dates to be confirmed.
  • Buildroot Developer Days, 8-9 October 2016, Berlin, Germany, after ELCE. Dates to be confirmed.

Past:

Talks

This section gathers the list of talks given about Buildroot, as well as the slides and video when available.

Past:

  • "Buildroot: a deep dive into the core", Thomas Petazzoni, Embedded Linux Conference Europe, 13-15 October 2014, Düsseldorf, Germany. Slides.
  • Buildroot: what's new, Thomas Petazzoni, Embedded Linux Conference, 1 May 2014, San Jose, United States. Slides, HD video, Low-res video, Audio only
  • "Buildroot: what is new", Peter Korsgaard, Embedded Linux Conference Europe, 25 October 2013, Edinburgh, UK. Slides, Video.

Accounting

This section gathers all the income and expenses of the Buildroot project.

Current balance: + €423.14

  • 2015-01-08: + €423.14 : Google paid €423.14 ($500) for mentoring a student for the GSoC 2014

Notes: until we have a legal entity representing Buildroot, that money is held by Yann E. MORIN on behalf the Buildroot project. Accounting is handled in Euro.

List of forks

  • [1]. A Rasberry-Pi related fork.
  • [2]. Another RPi related fork, with a lot of focus on Qt5 and GStreamer.
  • [3]. Not a fork, but a convenience layer on top of buildroot.
  • [4]. Another wrapper around Buildroot, to help manage projects.

Todo list

This is a list of improvements that we would like to see in buildroot. Feel free to add suggestions here. If you're working on one of these items, put your name and the date behind it, to avoid duplicate work.

There are a number of patches that have been determined to be useful but for various reasons nobody currently has time to review or test them. Anybody, especially a person new to buildroot, is welcome to adopt these patches and resubmit them to the mailing list. These patches can be viewed by looking at the following link - http://patchwork.ozlabs.org/project/buildroot/list/?state=1&delegate=7151

Packages

Note: if you start working on any of these packages, please edit this section to indicate it. If the package is proposed in a bug report, please also update the bug report. Sending a mail to the mailing list also never hurts, you never know that someone else started working on it without following this guideline.

Documentation

  • Document how to contribute (patches to list instead of bugzilla, SOB, formatting rules, acked-by and tested-by, how often to repost, what to expect, CC's, ...) basic guide
  • Document how a package patch should be formatted (Comment, SOB, file naming rules, ...) + send upstream + CC sendpatches
  • In documentation, explain how to report a bug

Build/release infrastructure

  • Peter sets up a planet on whatever server and links to it from buildroot website

Core Buildroot infrastructure

  • Several improvements are possible in the download infrastructure (even after all the improvements that were already done):
    • Rename the downloaded files so they include the package name and version. Special care has to be taken for primary and secondary sites, and for extra downloads (including patches).
    • Split between FOO_SITE and FOO_SOURCE shouldn't be necessary. Or it could be made optional, i.e. make it possible to specify the full path in FOO_SOURCE.
    • Conserve downloaded git/hg trees, so that you can change the FOO_VERSION and avoid a re-download. This requires using 'git fetch URL' instead of 'git clone URL' when the .git directory exists already.
  • Add a 'make <package>-help' target.
  • It would be nice to add a br-configure script in host/usr/bin for autotools-based packages. Run ...BUILDROOTSDK/usr/bin/br-configure --enable-foo --disable-bar, and the br-configure script would call the ./configure script in the current directory passing all the right options (--host, and all environment variables CC, LD, AS, AR and such).
  • Make the HOST-directory a relocatable SDK:
    • Make sure that all binaries and libraries built for the host are built with a rpath pointing to host/usr/lib. Normally, this should already be the case, but it's worth checking.
    • Change the rpath value to $ORIGIN/../lib instead of the current absolute path $(O)/host/usr/lib.
    • Modify the compiler wrapper program of external toolchains so that instead of using a fixed location for the compiler tools, it deduces their location in a relative manner from its current location.
    • Modify/patch pkg-config so that instead of having a fixed location for the PKG_CONFIG_PATH and PKG_CONFIG_SYSROOT_DIR, those are deduced from the location of the pkg-config binary. This will allow a pkg-config binary that has been moved to still operate properly, without having to set any environment variable.
    • Write a shell script, installed in host/usr/bin, which would mungle the libtool .la files, the qmake.conf file and the CMake toolchain file to set the correct path. This script reads a file (can be host/usr/share/buildroot/location) which contains the original location of the SDK. This allows the script to do the right modifications on all the libtool, qmake.conf and cmake files. Once this is done, the script changes the host/usr/share/buildroot/location file so that it contains the new location.
    • Modify the external toolchain wrapper so that it bails out and warns the user if the directory it is executed in doesn't match the location of host/usr/share/buildroot/location. We haven't discussed how this could work with internal and crosstool-NG toolchains, though.
  • Properly detect thread and TLS support in external toolchains, or make TLS knob driven by thread availability in the toolchain. See the discussion in http://patchwork.ozlabs.org/patch/288051/ as reference)
  • Add intrumentation scripts to analyse package installed files:
    • identify what package installed what files, identify files overriden by a later package
    • find libraries with wrong RPATH/RUNPATH tags
    • detect unused .so libs (eg. shared libs that are not DT_NEEDED by anything - note: only detect those libs, don't remove: can be used as plugin (dlopen), or used by an application built outside Buildroot)
  • A checkpackage script that verifies a package coding style (e.g. 80 # in the .mk file, indentation with tabs, ...). It could also check consistency of depends/select though that's a bit more advance.

TODO items under discussion

Here are some nice-to-have's for which it is not entirely clear if and how they could be implemented:

  • Out-of-tree builds, which allows the package source to be shared between different output directories and between host and target compiles.
  • It would be nice if you could run a buildroot command that prepares a local copy of a package's source, and allows you to generate patches for it later. This could use git or quilt to keep track of the patches.
  • It would be nice if there was a make target to reinstall everything to the target (i.e. remove all the target-installed stamps, remove the root stamp, maybe remove the target too). However, what is missing is the copying of the toolchain support files (libc.so etc.). It's not obvious that this can be done in a reliable way.
  • It would be nice if there was some common infrastructure to combine the images into one final flash or SD card or whatever image. However, it is probably difficult to find the commonality of all the different use cases. And we don't even see all these use cases.
    • Yann E. Morin is working on this topic.
  • To facilitate debugging, all packages should be installed to the staging directory. The target directory should in fact be a subset of the staging directory. See the FOSDEM 2013 discussion at http://elinux.org/Buildroot:DeveloperDaysFOSDEM2013, and the discussion around patch http://patchwork.ozlabs.org/patch/252718/. This is however a significant change in Buildroot, so probably difficult to implement, and will raise a number of quite complicated questions.