Resume rpmbuild?

[Michael Hill]

I'm running...

rpmbuild -bb --noprep --with baseonly --without debuginfo
--target=i686 kernel.spec

...and after several hours the device hangs somewhere after this:

+ popd
~/rpmbuild/BUILD/kernel-4.5.fc23/linux-4.6.0-0.rc5.git1.1.mdh.fc23.i686
+ '[' '%{buildarch}' = noarch ']'
+ QA_CHECK_RPATHS=1
+ case "${QA_CHECK_RPATHS:-}" in
+ /usr/lib/rpm/check-rpaths

After I restart the device, is there a way to get rpmbuild to resume?

Mike

[Richard Z]

On Sat, Apr 30, 2016 at 08:59:21AM -0400, Michael Hill wrote:
> I'm running...
>
> rpmbuild -bb --noprep --with baseonly --without debuginfo
> --target=i686 kernel.spec
>
> ...and after several hours the device hangs somewhere after this:
> popd
>
> ~/rpmbuild/BUILD/kernel-4.5.fc23/linux-4.6.0-0.rc5.git1.1.mdh.fc23.i686
> '[' '%{buildarch}' = noarch ']'
> QA_CHECK_RPATHS=1
> case "${QA_CHECK_RPATHS:-}" in
> /usr/lib/rpm/check-rpaths
>
> After I restart the device, is there a way to get rpmbuild to resume?

people used to hack around this problem by modding the spec script to skip 
certain build steps based upon an arg passed as "RPM_COMMAND" shell env
variable.

Typical snippet would be

%build

case "${RPM_COMMAND:-all}" in
all|config)
%configure
;;
esac

case "${RPM_COMMAND:-all}" in
all|config|build)
make
;;
esac

Of course this is considered a hack and you loose all guarantees that the result 
is a clean build.

Richard

[Michael Hill]

On Sun, May 1, 2016 at 3:33 PM, Richard Z  wrote:
> Of course this is considered a hack and you loose all guarantees that the result
> is a clean build.

Thanks, Richard. Perfect.

Mike