Boost 1.45
For a long time, the newest version of boost in openembedded has been 1.41, released over two years ago. As far as I can tell, the reason for this is that the recipe uses the cmake branch, which has been stagnant for the last several years. The reason for this is that the build system used by boost, bjam, is a travesty at best and a crime against packagers for the most (see the comment in the existing boost.inc for another poor soul's thoughts on the matter). Over the last few days, I've been trying to bring my embedded environment up-to-date with the current openembedded tree. Unfortunately, as it turns out boost 1.41 has some serious issues with gcc 4.5.
After quite a few exchanges with the boost folks and many hours of frustration, I think I finally have a working recipe. I'd appreciate it if people could try this and perhaps give some feedback. I can successfully build the package, but given the quirky nature of the build system it seems likely someone will find some configuration-specific issues.
Thanks,
- Ben
PATCH 1/1 boost: Initial commit of 1.45.0
--- .../01-fix-default-python-config.patch | 29 ++++ recipes/boost/boost-with-bjam.inc | 164 ++++++++++++++++++++ recipes/boost/boost_1.45.0.bb | 15 ++ 3 files changed, 208 insertions(+), 0 deletions(-) create mode 100644 recipes/boost/boost-1.45.0/01-fix-default-python-config.patch create mode 100644 recipes/boost/boost-with-bjam.inc create mode 100644 recipes/boost/boost_1.45.0.bb
recipes/boost/boost-1.45.0/01-fix-default-python-config.patch new file +29 -0
@@ -0,0 +1,29 @@ +Index: boost_1_45_0/tools/build/v2/tools/python.jam +=================================================================== +--- boost_1_45_0.orig/tools/build/v2/tools/python.jam 2010-12-27 16:10:56.340207790 -0500 ++++ boost_1_45_0/tools/build/v2/tools/python.jam 2010-12-27 16:11:53.751709203 -0500 +@@ -806,11 +806,7 @@ + } + + # Anything left to compute? +- if $(includes) && $(libraries) +- { +- .configured = true ; +- } +- else ++ if ! $(includes) || ! $(libraries) + { + version ?= $(fallback-version) ; + version ?= 2.5 ; +@@ -818,6 +814,11 @@ + compute-default-paths $(target-os) : $(version) : $(prefix:E=) ; + } + ++ if $(includes) && $(libraries) ++ { ++ .configured = true ; ++ } ++ + if ! $(interpreter-cmd) + { + fallback-cmd ?= python ;
recipes/boost/boost-with-bjam.inc new file +164 -0
164 lines not on this page · raw message
recipes/boost/boost_1.45.0.bb new file +15 -0
@@ -0,0 +1,15 @@ +require boost-with-bjam.inc + +PR = "${INC_PR}.1" + +SRC_URI = " \ + http://downloads.sourceforge.net/project/boost/boost/1.45.0/boost_1_45_0.tar.gz \ + file://01-fix-default-python-config.patch \ +" +S = "${WORKDIR}/boost_1_45_0" + +SRC_URI[md5sum] = "739792c98fafb95e7a6b5da23a30062c" +SRC_URI[sha256sum] = "7cd7a327ebe0f43db829eccce024b7f3b0d9d4d382422d59dc710793d55c6a12" + +DEPENDS += " icu " +
Signature · 1 line
1.7.0.4
2010/12/28 Ben Gamari <[email protected]>:
[...]
Quoted 5 lines
After quite a few exchanges with the boost folks and many hours of frustration, I think I finally have a working recipe. I'd appreciate it if people could try this and perhaps give some feedback. I can successfully build the package, but given the quirky nature of the build system it seems likely someone will find some configuration-specific issues.
Ben,
Thanks for your email, but....
If you do not attach or inline your recipe it is highly unlikely that you will get any feedback.
Best way is to make a patch and commit in git then use git format-patch and git send-email to send the patch to the list.
Frans.
A message that is not in the archive (sent off-list or not yet received): [email protected]
2010/12/28 Ben Gamari <[email protected]>:
Quoted 13 lines · Frans Meulenbroeks
Quoted 12 lines
2010/12/28 Ben Gamari <[email protected]>:
[...]
Quoted 5 lines
After quite a few exchanges with the boost folks and many hours of frustration, I think I finally have a working recipe. I'd appreciate it if people could try this and perhaps give some feedback. I can successfully build the package, but given the quirky nature of the build system it seems likely someone will find some configuration-specific issues.
Ben,
Quoted 9 lines
Quoted 5 lines
Thanks for your email, but....
If you do not attach or inline your recipe it is highly unlikely that you will get any feedback.
Well, this is odd. I actually did use git-send-email yet somehow it seems like the patch never made it. There must have been an error I missed.
Actually the patch made it, after I replied to you.
From the headers: trimmed some irrelevant fields
Received: by 10.231.200.140 with SMTP id ew12cs183342ibb;
Tue, 28 Dec 2010 02:29:25 -0800 (PST)
Received: by 10.150.92.21 with SMTP id p21mr5378669ybb.9.1293532164220;
Tue, 28 Dec 2010 02:29:24 -0800 (PST)
Return-Path: <[email protected]>
Received: from linuxtogo.org (linuxtogo.org [188.40.83.200])
by mx.google.com with ESMTPS id q34si33260779ybk.61.2010.12.28.02.21.52
(version=TLSv1/SSLv3 cipher=RC4-MD5);
Tue, 28 Dec 2010 02:29:24 -0800 (PST)
Received: from localhost ([127.0.0.1])
by linuxtogo.org with esmtp (Exim 4.72)
(envelope-from <[email protected]>)
id 1PXQv7-0008CO-JK; Tue, 28 Dec 2010 05:12:21 +0100
Received: from mail-qw0-f47.google.com ([209.85.216.47])
by linuxtogo.org with esmtp (Exim 4.72)
(envelope-from <[email protected]>) id 1PXMVU-0004pO-Lp
for [email protected];
Tue, 28 Dec 2010 00:30:54 +0100
Received: by qwg5 with SMTP id 5so8578239qwg.6
for <[email protected]>;
Mon, 27 Dec 2010 15:28:11 -0800 (PST)
Received: by 10.229.95.19 with SMTP id b19mr11363730qcn.179.1293492490412;
Mon, 27 Dec 2010 15:28:10 -0800 (PST)
Received: from ben-laptop (c-24-61-223-13.hsd1.nh.comcast.net [24.61.223.13])
by mx.google.com with ESMTPS id t7sm6133841qcs.40.2010.12.27.15.28.09
(version=SSLv3 cipher=RC4-MD5); Mon, 27 Dec 2010 15:28:10 -0800 (PST)From: Ben Gamari <[email protected]>
To: [email protected]
Date: Mon, 27 Dec 2010 18:28:04 -0500
Seems it took 11 hrs to get from you to me (if I interpret the data properly). your python post did not make it at all to here, but it did get to the gmane archive.
Frans
PATCH 1/1 boost: Initial commit of 1.45.0
In reply to Frans Meulenbroeks--- .../01-fix-default-python-config.patch | 29 ++++ recipes/boost/boost-with-bjam.inc | 164 ++++++++++++++++++++ recipes/boost/boost_1.45.0.bb | 15 ++ 3 files changed, 208 insertions(+), 0 deletions(-) create mode 100644 recipes/boost/boost-1.45.0/01-fix-default-python-config.patch create mode 100644 recipes/boost/boost-with-bjam.inc create mode 100644 recipes/boost/boost_1.45.0.bb
recipes/boost/boost-1.45.0/01-fix-default-python-config.patch new file +29 -0
@@ -0,0 +1,29 @@ +Index: boost_1_45_0/tools/build/v2/tools/python.jam +=================================================================== +--- boost_1_45_0.orig/tools/build/v2/tools/python.jam 2010-12-27 16:10:56.340207790 -0500 ++++ boost_1_45_0/tools/build/v2/tools/python.jam 2010-12-27 16:11:53.751709203 -0500 +@@ -806,11 +806,7 @@ + } + + # Anything left to compute? +- if $(includes) && $(libraries) +- { +- .configured = true ; +- } +- else ++ if ! $(includes) || ! $(libraries) + { + version ?= $(fallback-version) ; + version ?= 2.5 ; +@@ -818,6 +814,11 @@ + compute-default-paths $(target-os) : $(version) : $(prefix:E=) ; + } + ++ if $(includes) && $(libraries) ++ { ++ .configured = true ; ++ } ++ + if ! $(interpreter-cmd) + { + fallback-cmd ?= python ;
recipes/boost/boost-with-bjam.inc new file +164 -0
164 lines not on this page · raw message
recipes/boost/boost_1.45.0.bb new file +15 -0
@@ -0,0 +1,15 @@ +require boost-with-bjam.inc + +PR = "${INC_PR}.1" + +SRC_URI = " \ + http://downloads.sourceforge.net/project/boost/boost/1.45.0/boost_1_45_0.tar.gz \ + file://01-fix-default-python-config.patch \ +" +S = "${WORKDIR}/boost_1_45_0" + +SRC_URI[md5sum] = "739792c98fafb95e7a6b5da23a30062c" +SRC_URI[sha256sum] = "7cd7a327ebe0f43db829eccce024b7f3b0d9d4d382422d59dc710793d55c6a12" + +DEPENDS += " icu " +
Signature · 1 line
1.7.0.4