[Re: [meta-networking][PATCH 1/3] traceroute: resolve multilib issues] On 13.02.14 (Thu 11:34) Aws Ismail wrote:
> On 02/13/2013 10:34 PM, Joe MacDonald wrote: > >[[meta-networking][PATCH 1/3] traceroute: resolve multilib issues] On > >13.02.11 (Mon 13:16) Aws Ismail wrote: > > > >>Make recipe changes to enable successfull building > >>when multilib is being used. > >> > >>Signed-off-by: Aws Ismail <[email protected]> > >>--- > >> .../traceroute/traceroute_2.0.18.bb | 8 +++++--- > >> 1 files changed, 5 insertions(+), 3 deletions(-) > >> > >>diff --git > >>a/meta-networking/recipes-support/traceroute/traceroute_2.0.18.bb > >>b/meta-networking/recipes-support/traceroute/traceroute_2.0.18.bb > >>index 6eca731..f5f96b2 100644 > >>--- a/meta-networking/recipes-support/traceroute/traceroute_2.0.18.bb > >>+++ b/meta-networking/recipes-support/traceroute/traceroute_2.0.18.bb > >>@@ -11,6 +11,8 @@ LICENSE = "GPL-2.0 LGPL-2.1" > >> LIC_FILES_CHKSUM = "file://COPYING;md5=94d55d512a9ba36caa9b7df079bae19f \ > >> > >> file://COPYING.LIB;md5=bbb461211a33b134d42ed5ee802b37ff" > >>+PR = "r1" > >>+ > >> inherit update-alternatives > >> SRC_URI = > >> "${SOURCEFORGE_MIRROR}/traceroute/traceroute/traceroute-2.0.18/traceroute-2.0.18.tar.gz > >> \ > >>@@ -26,13 +28,13 @@ do_compile() { > >> do_install() { > >> install -d ${D}${bindir} > >>- install -m755 ${PN}/${PN} ${D}${bindir} > >>+ install -m755 ${BPN}/${BPN} ${D}${bindir} > >> install -m755 wrappers/tcptraceroute ${D}${bindir} > >> install -d ${D}${mandir} > >>- install -p -m644 ${PN}/${PN}.8 ${D}${mandir} > >>- ln -s ${PN}.8 ${D}${mandir}/${PN}6.8 > >>+ install -p -m644 ${BPN}/${BPN}.8 ${D}${mandir} > >>+ ln -s ${PN}.8 ${D}${mandir}/${BPN}6.8 > >> ln -s ${PN}.8 ${D}${mandir}/tcptraceroute.8 > >Are the targets actually installed as ${PN}.8 and not ${BPN}.8 in a > >multilib build scenario? What I'm asking is should the above be changed > >to this instead: > > > > ln -s ${BPN}.8 ${D}${mandir}/${BPN}6.8 > > ln -s ${BPN}.8 ${D}${mandir}/tcptraceroute.8 > > > I think you're right. It should be like that but the odd > thing is that even with ln -s ${PN}.8 no errors were > seen. Weird. But okay. So, so long as the BPN version is doing the right thing -- meaning it's actually creating a proper link and not a broken link, because ln is pretty happy to do whatever you tell it even if it doesn't make sense -- can you send out a revised patch set? -- -Joe MacDonald. :wq
signature.asc
Description: Digital signature
_______________________________________________ Openembedded-devel mailing list [email protected] http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
