public inbox for linux-media@vger.kernel.org
 help / color / mirror / Atom feed
From: Ariel D'Alessandro <ariel.dalessandro@collabora.com>
To: kieran.bingham@ideasonboard.com, linux-media@vger.kernel.org
Cc: hverkuil@xs4all.nl, sean@mess.org, p.zabel@pengutronix.de,
	laurent.pinchart@ideasonboard.com, ezequiel@collabora.com,
	nicolas@ndufresne.ca, gjasny@googlemail.com,
	xavier.claessens@collabora.com, nicolas.dufresne@collabora.com,
	user.vdr@gmail.com, sakari.ailus@iki.fi,
	mchehab+huawei@kernel.org, rosenp@gmail.com
Subject: Re: [v4l-utils v5 0/5] Add support for meson building
Date: Mon, 17 May 2021 17:55:03 -0300	[thread overview]
Message-ID: <008aec9c-c1e6-7f8b-dde5-8fa53c1e9ee6@collabora.com> (raw)
In-Reply-To: <f2f72ec5-e352-132f-b8d1-718589360bf0@ideasonboard.com>

Hi Kieran!

Thanks a lot for the feedback.

On 5/13/21 5:56 AM, Kieran Bingham wrote:
> Hi Ariel,
> 
> I'm still very much looking forward to this update!
> 
> I still think meson could be used in parallel for some cross over period
> while these issues are worked out, which would encourage and support
> wider testing before any consideration to swap altogether occurred.
> 
> 
> Furthermore, the performance improvement on this is really nice
> (intel i7 8th gen 4 core/8threads)
> 
> /usr/bin/time -v ./make-meson.sh
> 	User time (seconds): 99.38
> 	System time (seconds): 10.69
> 	Percent of CPU this job got: 577%
> 	Elapsed (wall clock) time (h:mm:ss or m:ss): 0:19.05
> 
> 
> /usr/bin/time -v ./make-autoconf.sh
> 	User time (seconds): 218.19
> 	System time (seconds): 21.46
> 	Percent of CPU this job got: 360%
> 	Elapsed (wall clock) time (h:mm:ss or m:ss): 1:06.54
> 
> however the build output directories are not comparable so I suspect
> that perhaps meson isn't (yet?) building all the same things?

Not sure what your scripts are doing, but the default autotools
behaviour should be equivalent to:

  $ meson configure build/ -Ddefault_library=both -Ddoxygen-doc=disabled

> 
> du -sh build-autoconf build-meson/
> 129M	build-autoconf
> 69M	build-meson/
> 
> 
> I've just tried testing it locally and found a smallish issue:
> 
> 
> On 12/05/2021 19:49, Ariel D'Alessandro wrote:
<snip>
>> Changes from v4:
>>
>>   * Dropped install_subdir() work-around to install empty dir. Supported
>>     now in meson v0.57.
>>   * Bump to meson v0.57. Replaced deprecated methods.
>>   * Added override_dependency() calls on libraries.
>>   * Properly check for bpf and SDL2 dependencies.
> 
> The SDL2 dependency is optional when I run ./bootstrap.sh; ./configure
> which gives the following output:
> 
> compile time options summary
> ============================
> 
>     Host OS                    : linux-gnu
>     X11                        : yes
>     GL                         : yes
>     glu                        : yes
>     libelf		       : yes
>     libjpeg                    : yes
>     libudev                    : yes
>     pthread                    : yes
>     QT version                 : v5.4 with QtGL
>     ALSA support               : yes
>     SDL support		       : no
> 
>     build dynamic libs         : yes
>     build static libs          : yes
> 
>     gconv                      : no
> 
>     dynamic libv4l             : yes
>     v4l_plugins                : yes
>     v4l_wrappers               : yes
>     libdvbv5                   : yes
>     dvbv5-daemon               : yes
>     v4lutils                   : yes
>     qv4l2                      : yes
>     qvidcap                    : yes
>     v4l2-ctl uses libv4l       : yes
>     v4l2-ctl-32                : no
>     v4l2-compliance            : yes
>     v4l2-compliance uses libv4l: yes
>     v4l2-compliance-32         : no
>     BPF IR Decoders:           : no
> 
> 
> However running meson build gives:
> 
> Has header "bpf/bpf.h" : NO
> Run-time dependency sdl2 found: YES 2.0.10
> Run-time dependency sdl2_image found: NO (tried pkgconfig and cmake)
> 
> meson.build:90:4: ERROR: C header 'SDL2/SDL_image.h' not found
> 
> Now I bet that's a corner case because SDL was found but SDL Image
> wasn't, but perhaps it needs updating so that both are optional, and the
> SDL support is only enabled when all SDL dependencies are met.
> 
> 
> 
> Of course I was able to 'solve/workaround' this locally with
>  sudo apt install libsdl2-image-dev
> 
> Note that the time measurements reported above were after installing SDL
> image so that both meson and autoconf should have been building the

That's a bug introducen in the latest v5 patchset, while addressing
comment [0] from Xavier. This will fix it:

$ git diff
diff --git a/meson.build b/meson.build
index bba8e030..f062856c 100644
--- a/meson.build
+++ b/meson.build
@@ -82,12 +82,12 @@ endif

 dep_sdl = dependency('SDL2', required: false)
 if not dep_sdl.found()
-    dep_sdl = cc.find_library('SDL2', has_headers: 'SDL2/SDL.h')
+    dep_sdl = cc.find_library('SDL2', has_headers: 'SDL2/SDL.h',
required: false)
 endif

 dep_sdl_image = dependency('SDL2_image', required: false)
 if not dep_sdl_image.found()
-    dep_sdl_image = cc.find_library('SDL2_image', has_headers:
'SDL2/SDL_image.h')
+    dep_sdl_image = cc.find_library('SDL2_image', has_headers:
'SDL2/SDL_image.h', required: false)
 endif

 dep_threads = dependency('threads')

[0]
https://gitlab.collabora.com/adalessandro/v4l-utils/-/commit/d6fc25fdd28227e88d5517342f045d5bda632272#note_82115

Thanks again!
Ariel

  reply	other threads:[~2021-05-17 20:55 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-12 18:49 [v4l-utils v5 0/5] Add support for meson building Ariel D'Alessandro
2021-05-12 18:49 ` [v4l-utils v5 1/5] Move README to markdown syntax Ariel D'Alessandro
2021-05-12 18:49 ` [v4l-utils v5 2/5] Add support for meson building Ariel D'Alessandro
2021-05-12 18:49 ` [v4l-utils v5 3/5] Copy Doxygen configuration file to doc/ Ariel D'Alessandro
2021-05-12 18:49 ` [v4l-utils v5 4/5] meson: Add support for doxygen documentation Ariel D'Alessandro
2021-05-12 18:49 ` [v4l-utils v5 5/5] Makefile.am: Distribute meson related files Ariel D'Alessandro
2021-05-13  8:56 ` [v4l-utils v5 0/5] Add support for meson building Kieran Bingham
2021-05-17 20:55   ` Ariel D'Alessandro [this message]
2021-05-17 22:13     ` Kieran Bingham
2021-05-18  7:23       ` Mauro Carvalho Chehab
2021-05-18 10:18         ` Kieran Bingham
2021-05-19 11:07           ` Mauro Carvalho Chehab
2021-06-16 13:36           ` Ariel D'Alessandro
2021-06-16 14:26       ` Ariel D'Alessandro
2021-06-16 14:59         ` Mauro Carvalho Chehab
2021-06-16 15:06           ` Xavier Claessens
2021-06-16 15:10           ` Nicolas Dufresne
2021-06-16 15:11           ` Laurent Pinchart
2021-10-04 14:24 ` Laurent Pinchart
2021-12-15 21:05   ` Laurent Pinchart
2021-11-18  9:03 ` Tomi Valkeinen
2021-11-18 10:09   ` Tomi Valkeinen
2021-12-15 21:07     ` Laurent Pinchart
2021-11-18 10:39   ` Laurent Pinchart
2021-12-15 21:10 ` Laurent Pinchart

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=008aec9c-c1e6-7f8b-dde5-8fa53c1e9ee6@collabora.com \
    --to=ariel.dalessandro@collabora.com \
    --cc=ezequiel@collabora.com \
    --cc=gjasny@googlemail.com \
    --cc=hverkuil@xs4all.nl \
    --cc=kieran.bingham@ideasonboard.com \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab+huawei@kernel.org \
    --cc=nicolas.dufresne@collabora.com \
    --cc=nicolas@ndufresne.ca \
    --cc=p.zabel@pengutronix.de \
    --cc=rosenp@gmail.com \
    --cc=sakari.ailus@iki.fi \
    --cc=sean@mess.org \
    --cc=user.vdr@gmail.com \
    --cc=xavier.claessens@collabora.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