From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Thu, 28 Mar 2019 15:22:10 +0100 Subject: [Buildroot] [PATCH 3/3] package/s6-rc: Allow to integrate s6-rc services In-Reply-To: <7bb98b03-837a-2494-3de0-b8267a8b7d56@mind.be> References: <20190216212835.25503-1-vadim4j@gmail.com> <20190216212835.25503-4-vadim4j@gmail.com> <20190327195409.14931f34@windsurf> <7bb98b03-837a-2494-3de0-b8267a8b7d56@mind.be> Message-ID: <20190328152210.2bc05d0c@windsurf> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net On Wed, 27 Mar 2019 21:35:51 +0100 Arnout Vandecappelle wrote: > So, I've taken a quick look at all the finalize hooks we have, and there's > nothing that jumps out as "this you really want to do before the > overlay/post-build". I'm not sure though how it would interfere with the other > (non-hook) finalizations we're doing (i.e. removing static libs etc.). Every > hook would need to be reviewed in detail to be sure it's OK. > > If such a review is done, I wouldn't mind moving the hooks after the > overlay/post-build. I've checked the history, and I think the order of things is > mostly accidental. When post-build script was introduced (commit eed7d8737ad) it > was put at the end of target-finalize, without specific explanation. However, at > that time, purgelocales was still done after target-finalize (now it has moved > to a TARGET_FINALIZE_HOOK so it's done in the beginning). So I think there was > no real consideration at the time. When the overlay was added, it was put just > before the post-build script, which makes sense. And around that time, > everything had moved before the post-build script, so when the > TARGET_FINALIZE_HOOKS were introduced it made sense to keep that before the > post-build script as well. I agree that the order of the steps is definitely not the result of a lot of thinking, but is mostly accidental. > We do have to take into account, however, that this may break existing > overlays/post-build scripts in surprising ways... So it should definitely be > mentioned in the release notes! Indeed, and that makes the change potentially a bit annoying :/ Thomas -- Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com