Permalink
Browse files

[ios] dpkg-deb -info can seq fault and abort the buildbot, info is no…

…t really required so comment it out
  • Loading branch information...
1 parent ecadaf4 commit c31aca24e63b1b70a6e207f360f134f460cbe2d7 @davilla davilla committed Sep 20, 2012
@@ -112,8 +112,8 @@ export COPYFILE_DISABLE=true
export COPY_EXTENDED_ATTRIBUTES_DISABLE=true
#
dpkg-deb -b $DIRNAME/$PACKAGE $DIRNAME/$ARCHIVE
-dpkg-deb --info $DIRNAME/$ARCHIVE
-dpkg-deb --contents $DIRNAME/$ARCHIVE
+#dpkg-deb --info $DIRNAME/$ARCHIVE
+#dpkg-deb --contents $DIRNAME/$ARCHIVE
# clean up by removing package dir
${SUDO} rm -rf $DIRNAME/$PACKAGE
@@ -96,8 +96,8 @@ export COPYFILE_DISABLE=true
export COPY_EXTENDED_ATTRIBUTES_DISABLE=true
#
dpkg-deb -b $DIRNAME/$PACKAGE $DIRNAME/$ARCHIVE
-dpkg-deb --info $DIRNAME/$ARCHIVE
-dpkg-deb --contents $DIRNAME/$ARCHIVE
+#dpkg-deb --info $DIRNAME/$ARCHIVE
+#dpkg-deb --contents $DIRNAME/$ARCHIVE
# clean up by removing package dir
${SUDO} rm -rf $DIRNAME/$PACKAGE

7 comments on commit c31aca2

Owner

Memphiz replied Sep 20, 2012

The odyssey ends here? ;)

Contributor

davilla replied Sep 20, 2012

@Memphiz, do you see any seg-fault errors running the older mk.... when packaging for ios/atv ?

Contributor

davilla replied Sep 20, 2012

/var/lib/buildbot/buildbot-work/sandbox/lib/python2.5 is not me :)

Owner

Memphiz replied Sep 20, 2012

I have never seen seg-faults tbh. But i also do not run from launchd of course. The signing thing is fixed now at least as it seems ;)

Member

theuni replied Sep 20, 2012

buildbot uses the info from the dpkg-deb --info $DIRNAME/$ARCHIVE to name the resulting file. Without that, it gets confused and dies.

We'll have to hunt down the dpkg-deb bug, or grep something else for the version.

Contributor

davilla replied Sep 21, 2012

Grrrr :)

Please sign in to comment.