#1218 "make lint" is broken in packages cvs
Closed: Fixed None Opened 15 years ago by chkr.

running "make lint" in the packages CVS doesn't check all rpm files due to an additional space in the generated rpmlint command line

how to reproduce:

cvs -d :pserver:anonymous@cvs.fedoraproject.org:/cvs/pkgs co lsscsi
cd lsscsi/devel
make local
make lint
rpmlint lsscsi-0.22-2.fc11.src.rpm i686/-0.22-2.fc11. i686.rpm
Error: no installed packages by name i686/
-0.22-2.fc11.
Error: no installed packages by name i686.rpm
1 packages and 0 specfiles checked; 0 errors, 0 warnings.

The 4th line from the bottom is wrong: "i686/*-0.22-2.fc11. i686.rpm"

It looks like that the problem is caused in common/Makefile.common when the "LOCALARCH" variable is created:

LOCALARCH := $(if $(shell grep -i '^BuildArch:.*noarch' $(SPECFILE)), noarch, $(shell uname -m))

The attached patch seems to fix the problem. Since I'm not yet completely familiar with the complete build system, I'm not sure whether the patch has any bad side-effects.

Please note that the attached patch fixes also a very minor UI glitch: when running "make help" the rpmlint line is not correctly indented.


ping

Any news regarding this bug? ;-) If there is anything I can do to help to get this fixed, please let me know...

Sorry this fell between the cracks, I don't think anybody was CCed to the ticket by default. Assigning this to Dennis.

this patch will not be applied. because we now support noarch sub packages the determination of what is noarch is broken in Makefile.common and needs to be reworked.

Hey, it looks like Toshio forwarded this ticket to fedora-buildsys-list (https://www.redhat.com/archives/fedora-buildsys-list/2009-May/msg00049.html). That's probably the best place to discuss getting the patch/Makefile.common into a state where it be fixed. Thanks for working on this!

Replying to [comment:3 ausil]:

this patch will not be applied. because we now support noarch sub packages the determination of what is noarch is broken in Makefile.common and needs to be reworked.

Thanks for having a look at this.

It's ok that the patch is not accepted and needs rework.

But I don't understand why you've closed the ticket with "wontfix" since the problem itself is still present.

Login to comment on this ticket.

Metadata