Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Korsgaard <peter@korsgaard.com>
To: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Adam Duskett <aduskett@gmail.com>,
	James Hilliard <james.hilliard1@gmail.com>,
	Julien Olivain <ju.o@free.fr>,
	Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	buildroot@buildroot.org
Subject: Re: [Buildroot] [PATCH] package/nftables: fix the build of the pyhon bindings
Date: Tue, 12 Sep 2023 23:21:20 +0200	[thread overview]
Message-ID: <87bke75bqn.fsf@48ers.dk> (raw)
In-Reply-To: <20230817191130.1022405-1-yann.morin.1998@free.fr> (Yann E. MORIN's message of "Thu, 17 Aug 2023 21:11:30 +0200")

>>>>> "Yann" == Yann E MORIN <yann.morin.1998@free.fr> writes:

 > nftables provides python bindings; it uses setuptools to install them.
 > We currently install those bindings by telling nftables buildsystem,
 > autotools, to install the python bindings.

 > However, we do not pass any of the environment variables that are needed
 > for setuptools packages. When host-python-setuptools is installed before
 > nftables is built [0], this breaks the system at runtime, as the
 > bindings are not installed; only the egg is, resulting in runtime errors
 > like:

 >     # python -c 'import nftables'
 >     Traceback (most recent call last):
 >       File "<string>", line 1, in <module>
 >     ModuleNotFoundError: No module named 'nftables'

 > Upstream has been doing some changes on their python handling, but it is
 > not in a released version, and we can't backport those changes either,
 > due to other big changes.

 > Instead, we split the pyhon bindings to their own package.

 > For legacy hadnling, we make that package default to y, so that existing
 > (def)config still work. The only novelty is that it can be disabled now.

 > Many thanks to Julien for testing and finding the offending dependency,
 > to James for suggesting the package split, and to Adam for, well,
 > trigerring the issue in the first place! ;-p

 > Note: a git bisect of the issue turns up 72 candidates for the breakage,
 > all around the time we dropped python2 support in early 2022; the last
 > known-good commit is 55df30f8b1fb (package/zfs: drop python2 support)
 > and the first known-bad commit is 697acda00d9f (package/pkg-python: drop
 > python2 host/setuptools support); everything in-between does not
 > configure (package/python/Config.in.host still sourced but already
 > removed), or does not build (host-python still in the dependency chain
 > but already removed), so had to be skipped during the bisect.

 > [0] This can happen when another python package using setuptools is
 >     built before nftables. However, with PPD, this never happens because
 >     host-python-setuptools is never in the dependency chain of nftables.

 > Reported-by: Julien Olivain <ju.o@free.fr>
 > Tested-by: Julien Olivain <ju.o@free.fr>
 > Suggested-by: James Hilliard <james.hilliard1@gmail.com>
 > Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
 > Cc: Adam Duskett <aduskett@gmail.com>
 > Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>

 > ---
 > Given the bisection result, it means that all the currently maintained
 > branches are affected by the issue.

Committed to 2023.02.x and 2023.05.x, thanks.

-- 
Bye, Peter Korsgaard
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  parent reply	other threads:[~2023-09-12 21:21 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-17 19:11 [Buildroot] [PATCH] package/nftables: fix the build of the pyhon bindings Yann E. MORIN
2023-08-20  9:46 ` Yann E. MORIN
2023-09-12 21:21 ` Peter Korsgaard [this message]
  -- strict thread matches above, loose matches on Subject: below --
2023-08-16 18:09 Yann E. MORIN
2023-08-16 19:09 ` Yann E. MORIN
2023-08-16 19:59 ` Julien Olivain

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=87bke75bqn.fsf@48ers.dk \
    --to=peter@korsgaard.com \
    --cc=aduskett@gmail.com \
    --cc=buildroot@buildroot.org \
    --cc=james.hilliard1@gmail.com \
    --cc=ju.o@free.fr \
    --cc=thomas.petazzoni@bootlin.com \
    --cc=yann.morin.1998@free.fr \
    /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