[oe] doing a serious "distclean"
Robert P. J. Day
rpjday at crashcourse.ca
Fri Nov 13 14:56:32 UTC 2009
as one of the posters mentioned earlier, bitbake's distclean doesn't
quite do a full cleaning. after bitbake'ing xterm, i have:
$ find . -name xterm-251*
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_rm_work
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_install
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_build
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_package_write
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_package_stage_all
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_package_write_ipk
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_unpack
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_compile
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_package
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_qa_staging
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_patch
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_fetch
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_configure
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_qa_configure
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_rm_work_all
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_populate_staging
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_package_stage
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_setscene
./stamps/armv7a-angstrom-linux-gnueabi/xterm-251-r0.do_distribute_sources
./deploy/glibc/sources/MIT-X/xterm/xterm-251.tgz
./deploy/glibc/sources/MIT-X/xterm/xterm-251.tgz.md5
./work/armv7a-angstrom-linux-gnueabi/xterm-251-r0
$
after "distclean"ing that package, what still remains is:
$ find . -name xterm-251*
./deploy/glibc/sources/MIT-X/xterm/xterm-251.tgz
./deploy/glibc/sources/MIT-X/xterm/xterm-251.tgz.md5
$
and, predictably, that tarball has been removed from the downloads
directory.
what would seem to be more useful is a bitbake clean command that
removes every single trace of a build from under the tmp directory,
but *leaves* the tarball in the download directory. that would
represent the best test for building a package from scratch, without
actually having to drag it down again, no? (or does such a bitbake
command already exist? "bitbake -c tmpclean"?)
rday
--
========================================================================
Robert P. J. Day Waterloo, Ontario, CANADA
Linux Consulting, Training and Kernel Pedantry.
Web page: http://crashcourse.ca
Twitter: http://twitter.com/rpjday
========================================================================
More information about the Openembedded-devel
mailing list