From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-vs1-f49.google.com (mail-vs1-f49.google.com [209.85.217.49]) by mx.groups.io with SMTP id smtpd.web09.15927.1582135540523878316 for ; Wed, 19 Feb 2020 10:05:41 -0800 Authentication-Results: mx.groups.io; dkim=pass header.i=@soccasys-com.20150623.gappssmtp.com header.s=20150623 header.b=aU28xJbp; spf=pass (domain: soccasys.com, ip: 209.85.217.49, mailfrom: laurent.gauthier@soccasys.com) Received: by mail-vs1-f49.google.com with SMTP id 7so886607vsr.10 for ; Wed, 19 Feb 2020 10:05:40 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=soccasys-com.20150623.gappssmtp.com; s=20150623; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc:content-transfer-encoding; bh=pBLb9wRveT2vh4j92M98QmulTznR4gOKoXjp2eUCtHs=; b=aU28xJbpZvtv5aOEv9IOIVikmVJxfCmbnr/VYgjlY2zqcbKZS3W3UkqB3p+j4sHgol KNuwd9fFYY2YrkjEJVXzRm6/gncT7+JeqrGT09AvtqPT/1bvwol1so4xHqV2s6fIxAaj 2GaQ/M9CWDynOsDCtlxeSzskwy3+sRmbfT6TN30K+Y4xvyYOTB7j8UgHWJ+BHkmTRzDr 3RDrfn5A+QDAKbmvyyerK6S0bdrZ+CGAfsgOuGN12od67AVG10d8JHHc+YiLV9qB0rqC HwWc6Ljo7vU2y2GoOfNtSrlzVWLcPTMhWjo35T7xrgNdPTNv5B88/HrVs8l5Nsa/kDKT pEnQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc:content-transfer-encoding; bh=pBLb9wRveT2vh4j92M98QmulTznR4gOKoXjp2eUCtHs=; b=oRlydNZci9MbVxqOyz8LDMYqcsaMfPiyp4Z7hqQX/kqK82XezRX59d3nn3E9Y+WBUL H4kwhoqXFrC8vEV1zA7XdiJqjiP0vdXztZBm+LTnfAnEm88kq6KMPa1syUGiBTO0/veU NSnbfGGlXEji1RqkelQiM09EWUc7CUvhU8Xg2NWxZd5FMr8NuoX3isnyXsvwsnbnYaE/ 3jloI0LF0tOCD14jEbW1tY+3UrDYUrdY7/aFhf2bWjODkcV+3KGG+k7Mwmv6jk3TceCv Nw8M325/zdtYnBK63SN4LnK4o3Vle3YcFYHFXt/n9oLiyiLsAYJwsJSEXDKWNLFzw38O o5nw== X-Gm-Message-State: APjAAAVBLR2kZElxk2CQe3lKml9iIeRRYDV/+DJBoY3kChGYhh6JAzFA D+7afJ6UI4CGaRBqAd+pS6xqxfYKoc4ahaHb6RqLwQ== X-Google-Smtp-Source: APXvYqxov38kdoAf1CCV1AHiuSr93DVpvrP8kInPLd9EPhxeG0iRFJiGwy+Z14LQssGpBKJI/LHIrRCttev8X570ELU= X-Received: by 2002:a67:7950:: with SMTP id u77mr14871019vsc.209.1582135539163; Wed, 19 Feb 2020 10:05:39 -0800 (PST) MIME-Version: 1.0 References: In-Reply-To: From: "Laurent Gauthier" Date: Wed, 19 Feb 2020 19:05:03 +0100 Message-ID: Subject: Re: [yocto] Modified GENIVI Cannelloni recipe with strange side effects To: Zoran Stojsavljevic Cc: yocto@lists.yoctoproject.org Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Hi Zoran, I just saw your reply now. I think that you might want to remove the INHIBIT_SYSROOT_STRIP and other INHIBIT_* options from your recipe. For reference a message from Khem warning that this option should be used sparingly: * https://www.yoctoproject.org/pipermail/yocto/2019-March/044415.html My best guess is that the use of this option is directly linked to chrpath being needed. As this recipe is being built with a rather clean looking CMakeLists.txt none of these weird options are needed. Kind regards, Laurent. On Mon, Feb 17, 2020 at 8:01 AM Zoran Stojsavljevic wrote: > > > The issue I see is that the following files have been build but NOT in= stalled: > > > > * libcannelloni-common.so.0 > > * libcannelloni-common.so.0.0.1 > > Not quite... The solution is outlined here (in function do_install): > + ## ERROR: QA Issue: package cannelloni contains bad RPATH > + ## quick fix is in a do_install or do_install_append do > + chrpath -d ${D}${bindir}/cannelloni > > https://github.com/ZoranStojsavljevic/meta-socketcan/blob/master/recipes= -can/cannelloni/cannelloni.bb > https://github.com/ZoranStojsavljevic/meta-socketcan/blob/master/recipes= -can/cannelloni/cannelloni.bb_GENIVI > > I admit, your first email has shaken my head, so I can see things much > more clear. :-) > > My best guess, this solution is just a workaround (not the final one), > since I have in ${D} the following: > > cannelloni-1.0: package cannelloni contains bad RPATH > /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/cortexa8h= f-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build: > in file /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/c= ortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/packages-split/cannello= ni/usr/bin/cannelloni > [rpaths] > > So, since my limited knowledge about bitbake build systems ends here, > somebody from YOCTO primes (potentially Khem Raj, Ross Burton, maybe > even Richard Purdie) should look more closely into this issue > (apologies for my unsolicited suggestions). > > Laurent, > > Once again, thank you for unselfish help, > Zoran > _______ > > > On Fri, Feb 14, 2020 at 2:20 PM Laurent Gauthier > wrote: > > > > Hi Zoran, > > > > You are almost there! I can feel it... :-) > > > > The issue I see is that the following files have been build but NOTins= talled: > > > > * libcannelloni-common.so.0 > > * libcannelloni-common.so.0.0.1 > > > > If you make sure that they are installed that should fix your issue. > > > > Based on the info you provided no RDEPENDS seems to be required as it > > all appears that everything is in one package named "cannelloni", > > rather than a package for the main executable and then packages for > > libraries. > > > > Kind regards, Laurent. > > > > On Fri, Feb 14, 2020 at 12:43 PM Zoran Stojsavljevic > > wrote: > > > > > > Hello Laurent, > > > > > > Many thanks to you for the help. :-) > > > > > > I did some modifications, and now I have all the elements in there/i= n place: > > > > > > [user@fedora31-ssd cannelloni]$ cd ../../../build/tmp > > > [user@fedora31-ssd tmp]$ find . -name libcannelloni* > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/image/us= r/lib/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/sysroot-= destdir/usr/lib/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/package/= usr/lib/.debug/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/package/= usr/lib/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/packages= -split/cannelloni/usr/lib/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/packages= -split/cannelloni-dbg/usr/lib/.debug/libcannelloni-common.so > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build/li= bcannelloni-common.so.0 > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build/li= bcannelloni-common.so.0.0.1 > > > ./work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build/li= bcannelloni-common.so > > > ./sysroots-components/cortexa8hf-neon/cannelloni/usr/lib/libcannello= ni-common.so > > > > > > I miss the very end of your thoughts. Namely: > > > > > > > The name of the package containing the shared library is name of t= he > > > > xxx first-level directory "packages-split/xxx". > > > > > > So, how should I write the RDEPENDS command? > > > > > > Something as: RDEPENDS_${PN} =3D "???" > > > > > > What should I put on the right side of the equation (according to th= e above traces)? > > > > > > Thank you, > > > Zoran > > > _______ > > > > > > On Fri, Feb 14, 2020 at 11:49 AM Laurent Gauthier wrote: > > >> > > >> Hi Zoran, > > >> > > >> The issue seems to be that the executable /usr/bin/cannelloni has a > > >> reference to a shared library (libcannelloni-common.so.0) for which > > >> the Yocto build system is not able to determine automatically which > > >> package provides it. > > >> > > >> Based on the name I would assume that this package should be create= d > > >> by the same recipe that produces this executable (one recipe produc= es > > >> multiple packages). > > >> > > >> The most probable reason for this is that the new version of the > > >> package you are trying to build does not install the "missing" shar= ed > > >> library properly. But here are some steps you could follow to try t= o > > >> determine the stage of build/install/package where the shared libra= ry > > >> goes missing. > > >> > > >> To debug this I would suggest that you check that this > > >> "libcannelloni-common.so.0" shared library is present in several > > >> directories. > > >> > > >> First in the build directory: > > >> > > >> * /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/co= rtexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build > > >> > > >> If it is not there that would be very surprising. I will assume tha= t > > >> it is present. Let us know if it is not. > > >> > > >> Then the next location to check for this shared library is the foll= owing: > > >> > > >> * /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/co= rtexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/package > > >> > > >> If the file is not there, then it means that the recipe did not > > >> "install" it (as this directory is populated by do_install). > > >> > > >> If the file is there then you can check if it is correctly assigned= in > > >> a package by determining if it is also found in: > > >> > > >> * /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/co= rtexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/packages-split > > >> > > >> If the file is not there, then it means that the recipe did not > > >> "package" it properly (as this directory is populated by do_package= ). > > >> You should review the recipe for any anomaly in assigning installed > > >> files to individual packages. > > >> > > >> If the file is there then you probably should add the package that > > >> contains the shared library in the RDEPENDS for the "cannelloni" > > >> package. > > >> > > >> The name of the package containing the shared library is name of th= e > > >> xxx first-level directory "packages-split/xxx". > > >> > > >> Not sure if that will solve your issue, but hopefully that will hel= p. > > >> > > >> Kind Regards, Laurent. > > >> > > >> On Fri, Feb 14, 2020 at 11:27 AM Zoran wrote: > > >> > > > >> > Hello List, > > >> > > > >> > I am trying to solve very interesting ERROR I am getting with sli= ghtly modified GENIVI Canneloni recipe: > > >> > https://github.com/ZoranStojsavljevic/meta-socketcan/blob/master/= recipes-can/cannelloni/cannelloni.bb > > >> > > > >> > If I take the recipe as is, everything works fine, with: > > >> > ## SRCREV =3D "${AUTOREV}" > > >> > SRCREV =3D "0fb6880b719b8acf2b4210b264b7140135e4be8a" > > >> > > > >> > Everything works fine, but if I swap the static hash with auto la= test hash (SRCREV =3D "${AUTOREV}": > > >> > SRCREV =3D "${AUTOREV}" > > >> > ## SRCREV =3D "0fb6880b719b8acf2b4210b264b7140135e4be8a" > > >> > > > >> > I am getting these ERRORS, which seems to me very strange?! > > >> > _______ > > >> > > > >> > Sstate summary: Wanted 11 Found 6 Missed 5 Current 1398 (54% matc= h, 99% complete) > > >> > NOTE: Executing Tasks > > >> > NOTE: Setscene tasks completed > > >> > ERROR: cannelloni-1.0-r0 do_package_qa: QA Issue: package cannell= oni contains bad RPATH /home/user/projects2/beaglebone-black/bbb-yocto/buil= d/tmp/work/cortexa8hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/build: in f= ile /home/user/projects2/beaglebone-black/bbb-yocto/build/tmp/work/cortexa8= hf-neon-poky-linux-gnueabi/cannelloni/1.0-r0/packages-split/cannelloni/usr/= bin/cannelloni [rpaths] > > >> > ERROR: cannelloni-1.0-r0 do_package_qa: QA Issue: /usr/bin/cannel= loni contained in package cannelloni requires libcannelloni-common.so.0, bu= t no providers found in RDEPENDS_cannelloni? [file-rdeps] > > >> > ERROR: cannelloni-1.0-r0 do_package_qa: QA run found fatal errors= . Please consider fixing them. > > >> > ERROR: Logfile of failure stored in: /home/user/projects2/beagleb= one-black/bbb-yocto/build/tmp/work/cortexa8hf-neon-poky-linux-gnueabi/canne= lloni/1.0-r0/temp/log.do_package_qa.255490 > > >> > ERROR: Task (/home/user/projects2/beaglebone-black/bbb-yocto/meta= -socketcan/recipes-can/cannelloni/cannelloni.bb:do_package_qa) failed with = exit code '1' > > >> > NOTE: Tasks Summary: Attempted 3791 tasks of which 3788 didn't ne= ed to be rerun and 1 failed. > > >> > _______ > > >> > > > >> > Any advise how to make GENIVI Cannelloni recipe to work with: SRC= REV =3D "${AUTOREV}" ??? > > >> > > > >> > Thank you, > > >> > Zoran > > >> > > > >> >=20 > > >> > > >> > > >> > > >> -- > > >> Laurent Gauthier > > >> Phone: +33 630 483 429 > > >> http://soccasys.com > > > > > > > > -- > > Laurent Gauthier > > Phone: +33 630 483 429 > > http://soccasys.com --=20 Laurent Gauthier Phone: +33 630 483 429 http://soccasys.com