mailarchive of the ptxdist mailing list
 help / color / mirror / Atom feed
From: Michael Olbrich <m.olbrich@pengutronix.de>
To: ptxdist@pengutronix.de, Roelf-Erik Carsjens <rca@thorsis.com>
Subject: Re: [ptxdist] [RFC PATCH 0/1] building qmake based target packages
Date: Mon, 27 Nov 2023 17:16:01 +0100	[thread overview]
Message-ID: <ZWTAwb8JjI4rn_Z_@pengutronix.de> (raw)
In-Reply-To: <20231120-stopped-bulginess-7410677f82a5@ifak-system.com>

On Mon, Nov 20, 2023 at 02:55:39PM +0100, Alexander Dahl wrote:
> Am Fri, Nov 10, 2023 at 08:51:01AM +0100 schrieb Michael Olbrich:
> > On Tue, Oct 17, 2023 at 01:35:14PM +0200, Alexander Dahl wrote:
> > > I currently have some kind of weird requirement: building a target
> > > package with qmake, but not linking against Qt.  It's a fork of RTKlib
> > > which is itself a university project (with hand written makefiles, but
> > > not for the part I'm interested in):
> > > 
> > > https://github.com/rtklibexplorer/RTKLIB/releases/tag/b34h
> > > 
> > > I actually achieved building it with the attached make rule.  Note it
> > > has a dependency to QT5, but I had to manually add the Qt mkspec file
> > > ptxdist creates as an option to the _CONF_OPT variable.
> > > 
> > > Two things are bad about this:
> > > 
> > > 1.) I don't think it should be necessary to let the make rule set the
> > > correct mkspec file.  There is only one and IMHO ptxdist should handle
> > > this internally if CONF_TOOL is qmake?
> > 
> > I'm pretty sure you're calling the wrong qmake here. You need to modify
> > PATH the same way e.g. rules/qwt5.make does.
> 
> This works, thanks.
> 
> (If not setting that, the qmake of my host system was taken.  Double
> checked that by removing that qmake for now.)
> 
> > > 2.) rather large Qt libraries are installed to the target now.  I guess
> > > I might be able to avoid this by setting 'select QT5 if BUILDTIME' in
> > > the in rule, but I would have expected HOST_QT5 to be sufficient.  This
> > > does not work however for two reasons: a) ptxdist complains it needs QT5
> > > if CONF_TOOL is qmake and b) if I only select HOST_QT5 that mkspec is
> > > not generated.
> > 
> > Do you need qmake from qt5 or would qt6 work as well? We already have a
> > host-qt6 package and if we build the binary there and split out the qmake
> > config stuff from qt6 into a new cross-qmake package.
> 
> Tested with 'select QT5' and 'select QT6' now and both work.
> Only doing 'select HOST_QT5' or 'HOST_QT6' yields this:
> 
>     [ 35/171] started : rtklib-demo5.prepare
>     
>     ptxdist: error: 'rtklib-demo5' uses 'qmake' but does not select 'qt5' or 'qt6'
> 
> (With ptxdist-2023.03.0 at least.)
> 
> > It would need to copy the mkspecs from the host package I think before
> > adding the custom mkspecs stuff.
> 
> This is not yet in place, right?  At least I could not see any related
> changes between ptxdist-2023.05.0 and master.

Correct. It's just an idea how this could work in the future. It's not
implemented at all.

Michael

> > > I saw ptxdist having nearly zero packages selecting QT5, and none
> > > requiring HOST_QT5 (?) so it's difficult to learn from examples here.
> > > Maybe this usecase is quite esoteric?  Don't know.
> > > 
> > > In the end I would only need qmake, nothing more.  That one generates a
> > > Makefile which is then used to compile the package.
> > > 
> > > Would take directions to improve ptxdist for qmake based packages.
> > > 
> > > Greets
> > > Alex
> > > 
> > > Alexander Dahl (1):
> > >   WIP: rtklib-demo5: Add new package
> > > 
> > >  rules/rtklib-demo5.in   | 23 ++++++++++++++++
> > >  rules/rtklib-demo5.make | 58 +++++++++++++++++++++++++++++++++++++++++
> > >  2 files changed, 81 insertions(+)
> > >  create mode 100644 rules/rtklib-demo5.in
> > >  create mode 100644 rules/rtklib-demo5.make
> > > 
> > > 
> > > base-commit: 90875f8af4f6f1dd9f8eda71df8e8a90866f9839
> > > -- 
> > > 2.30.2
> > > 
> > > 
> > > 
> > 
> > -- 
> > 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 |
> 
> 

-- 
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 |



      reply	other threads:[~2023-11-27 16:16 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-10-17 11:35 Alexander Dahl
2023-10-17 11:35 ` [ptxdist] [RFC PATCH 1/1] WIP: rtklib-demo5: Add new package Alexander Dahl
2023-11-10  7:51 ` [ptxdist] [RFC PATCH 0/1] building qmake based target packages Michael Olbrich
2023-11-20 13:55   ` Alexander Dahl
2023-11-27 16:16     ` 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=ZWTAwb8JjI4rn_Z_@pengutronix.de \
    --to=m.olbrich@pengutronix.de \
    --cc=ptxdist@pengutronix.de \
    --cc=rca@thorsis.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