Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Yann E. MORIN" <yann.morin.1998@free.fr>
To: buildroot@buildroot.org
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>,
	Julien Olivain <ju.o@free.fr>,
	James Hilliard <james.hilliard1@gmail.com>,
	Adam Duskett <aduskett@gmail.com>
Subject: Re: [Buildroot] [PATCH] package/nftables: fix the build of the pyhon bindings
Date: Wed, 16 Aug 2023 21:09:14 +0200	[thread overview]
Message-ID: <20230816190914.GC1340200@scaer> (raw)
In-Reply-To: <20230816180916.2608831-1-yann.morin.1998@free.fr>

All,

There is still an unkown part to the tory, which is when it broke, or if
it always was broken. I'm stil invetigating this, and this will be added
to the commit log shortly (thus the patch is "changes-requested" in
patchwork).

Still, feedback on the changes are still appreciated.

Regards,
Yann E. MORIN.

On 2023-08-16 20:09 +0200, Yann E. MORIN spake thusly:
> 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
> 
> [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>
> 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>
> ---
>  package/nftables/Config.in                    |  5 +++++
>  package/nftables/nftables-python/Config.in    |  4 ++++
>  .../nftables-python/nftables-python.hash      |  1 +
>  .../nftables-python/nftables-python.mk        | 22 +++++++++++++++++++
>  package/nftables/nftables.mk                  | 10 +++------
>  5 files changed, 35 insertions(+), 7 deletions(-)
>  create mode 100644 package/nftables/nftables-python/Config.in
>  create mode 120000 package/nftables/nftables-python/nftables-python.hash
>  create mode 100644 package/nftables/nftables-python/nftables-python.mk
> 
> diff --git a/package/nftables/Config.in b/package/nftables/Config.in
> index 8d172b7256..833a3a38a2 100644
> --- a/package/nftables/Config.in
> +++ b/package/nftables/Config.in
> @@ -13,5 +13,10 @@ config BR2_PACKAGE_NFTABLES
>  
>  	  http://www.netfilter.org/projects/nftables/index.html
>  
> +# Legacy: this used to be handled in nftables.mk
> +if BR2_PACKAGE_NFTABLES
> +source "package/nftables/nftables-python/Config.in"
> +endif
> +
>  comment "nftables needs a toolchain w/ wchar, headers >= 3.12"
>  	depends on !BR2_USE_WCHAR || !BR2_TOOLCHAIN_HEADERS_AT_LEAST_3_12
> diff --git a/package/nftables/nftables-python/Config.in b/package/nftables/nftables-python/Config.in
> new file mode 100644
> index 0000000000..b16e2d0ad1
> --- /dev/null
> +++ b/package/nftables/nftables-python/Config.in
> @@ -0,0 +1,4 @@
> +config BR2_PACKAGE_NFTABLES_PYTHON
> +	bool "python bindings"
> +	default y # legacy
> +	depends on BR2_PACKAGE_PYTHON3
> diff --git a/package/nftables/nftables-python/nftables-python.hash b/package/nftables/nftables-python/nftables-python.hash
> new file mode 120000
> index 0000000000..9ac74580e7
> --- /dev/null
> +++ b/package/nftables/nftables-python/nftables-python.hash
> @@ -0,0 +1 @@
> +../nftables.hash
> \ No newline at end of file
> diff --git a/package/nftables/nftables-python/nftables-python.mk b/package/nftables/nftables-python/nftables-python.mk
> new file mode 100644
> index 0000000000..908bacd99a
> --- /dev/null
> +++ b/package/nftables/nftables-python/nftables-python.mk
> @@ -0,0 +1,22 @@
> +################################################################################
> +#
> +# nftables-python
> +#
> +################################################################################
> +
> +# The following assignments work only because nftables.mk is included before
> +# this file is.
> +NFTABLES_PYTHON_VERSION = $(NFTABLES_VERSION)
> +NFTABLES_PYTHON_SOURCE = $(NFTABLES_SOURCE)
> +NFTABLES_PYTHON_SITE = $(NFTABLES_SITE)
> +NFTABLES_PYTHON_LICENSE = $(NFTABLES_LICENSE)
> +NFTABLES_PYTHON_LICENSE_FILES = $(NFTABLES_LICENSE_FILES)
> +
> +# We share the same source code as nftables
> +NFTABLES_PYTHON_DL_SUBDIR = nftables
> +
> +NFTABLES_PYTHON_SUBDIR = py
> +
> +NFTABLES_PYTHON_SETUP_TYPE = setuptools
> +
> +$(eval $(python-package))
> diff --git a/package/nftables/nftables.mk b/package/nftables/nftables.mk
> index b0a14bd429..aa88203ab5 100644
> --- a/package/nftables/nftables.mk
> +++ b/package/nftables/nftables.mk
> @@ -42,13 +42,6 @@ else
>  NFTABLES_CONF_OPTS += --without-json
>  endif
>  
> -ifeq ($(BR2_PACKAGE_PYTHON3),y)
> -NFTABLES_CONF_OPTS += --enable-python
> -NFTABLES_DEPENDENCIES += python3
> -else
> -NFTABLES_CONF_OPTS += --disable-python
> -endif
> -
>  NFTABLES_CONF_ENV = LIBS="$(NFTABLES_LIBS)"
>  
>  define NFTABLES_LINUX_CONFIG_FIXUPS
> @@ -58,3 +51,6 @@ define NFTABLES_LINUX_CONFIG_FIXUPS
>  endef
>  
>  $(eval $(autotools-package))
> +
> +# Legacy: we used to handle it in this .mk
> +include package/nftables/nftables-python/nftables-python.mk
> -- 
> 2.25.1
> 

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  reply	other threads:[~2023-08-16 19:09 UTC|newest]

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

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=20230816190914.GC1340200@scaer \
    --to=yann.morin.1998@free.fr \
    --cc=aduskett@gmail.com \
    --cc=buildroot@buildroot.org \
    --cc=james.hilliard1@gmail.com \
    --cc=ju.o@free.fr \
    --cc=thomas.petazzoni@bootlin.com \
    /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