mailarchive of the ptxdist mailing list
 help / color / mirror / Atom feed
From: Michael Olbrich <m.olbrich@pengutronix.de>
To: ptxdist@pengutronix.de
Subject: Re: [ptxdist] RFC: Make version selection.
Date: Fri, 25 Jun 2021 09:50:16 +0200	[thread overview]
Message-ID: <20210625075016.GB4006120@pengutronix.de> (raw)
In-Reply-To: <0dece245-050e-4334-0fdf-73b5ec14f0db@t2data.com>

On Sun, Jun 20, 2021 at 07:45:39AM +0200, Christian Melki wrote:
> On 6/19/21 11:44 PM, Roland Hieber wrote:
> > On Fri, Jun 18, 2021 at 02:29:07PM +0200, Christian Melki wrote:
> >> Make 4.3 introduced a new set of incompatibility.
> >>
> >> https://lwn.net/Articles/810071/
> >>
> >> This rarely affects building, but sometimes it will.
> >> How about adding something like tool version selection?
> >> (Building a specific version perhaps?).
> >> Ptxdist seem rather content with just "some make", but since host version
> >> will undeniably affect the build success in some cases, how about making it
> >> an option?
> >> Host cmake is a host-rule f.ex?
> > 
> > This could result in a bit of yak shaving, because PTXdist host tools
> > are built using the normal PTXdist build system, which first feeds all
> > rule files to make and uses it for dependency generation. So make would
> > still be needed on the host to build the new host-make.
> > 
> 
> Sure. It will need a real host make, regardless.
> But can't this be done like a ... host-make-old? Then your make variable
> for that obnoxious package gets treated with a "older make" and the rest
> of the world lives on like normal?
> Ptxdist doesn't care, neither does the vast majority of packages. They
> are going to be up to date and have fixed makefiles.

From what I can tell, the incompatibility is mostly on-way. If it works
with the new version, then it works with the old version. And PTXdist (and
all actively developed packages) have been fixed.

> > As an alternative, we could invent a special way to build the host-make
> > first (could be a static shell library function), and then restart
> > parsing the rules with the new host-make. But still then, at least one
> > make is needed on the build host to build PTXdist itself after the
> > ./configure … :)
> > 
> Yuck. :)

This is really not needed. PTXdist itself work with old and new versions.
It's slower and dependency handling is a bit degraded for _really_ old
versions but that's it.

> > As another alternative, we could test for known make incompatibilities
> > on PTXdist startup and try to set up our rule files accordingly so they
> > built with all make versions. Maybe add a linter function too, if
> > possible.
> 
> I got stuck on the new make with the really, really old gstreamer tools.
> I've patched them (will send those in if anybody cares). So keeping up
> with patches is also an alternative. But I'd rather be able to tell
> ptxdist "older make here please".

Well, you could create a host-old-make package, that installs it binary
into a separate path (like we do with qt5) and add a dependency and a
custom PATH to the packages that need it.

But I'm not sure if it is worth the effort. The necessary changes are not
that big and a few patches to fix the Makefiles are probably less work.

Regards,
Michael

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

_______________________________________________
ptxdist mailing list
ptxdist@pengutronix.de
To unsubscribe, send a mail with subject "unsubscribe" to ptxdist-request@pengutronix.de

      reply	other threads:[~2021-06-25  7:50 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-06-18 12:29 Christian Melki
2021-06-19 21:44 ` Roland Hieber
2021-06-20  5:45   ` Christian Melki
2021-06-25  7:50     ` Michael Olbrich [this message]

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=20210625075016.GB4006120@pengutronix.de \
    --to=m.olbrich@pengutronix.de \
    --cc=ptxdist@pengutronix.de \
    /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