Thank you for the bug report, James.

The binary file fail-mbr.bin can be built only with a (real or emulated) 
x-86 machine; however it makes sense to install it in a non-x-86 server,
since partclone can use it while restoring an x-86 client.

So, what may I do? 

I suggest to include fail-mbr.bin in the package's source, and override
the lintian error, while documenting that this file comes with its
source.

On the other hand, such a method cannot be recommended : the same
argument might be used for any arbitrary binary file, and a boot sector
can be considered as a very sensitive piece of software.

Please have you some suggestion, do you know other methods to build this
file only with *one* architecture, and declare it as available for "all"
architectures? This would enforce its build from sources in the debian
build farm.

The closest issue which I could imagine among debian packages is the
build system of the package firmware-free; however I could not
understand, while inspecting loosely this package, whether there were
real source files in the package and whether there is any chance to
buils any binary from source. When I try to debuild this package with an
amd64 machine, nothing is built, binaries are just copied.

Best regards,                   Georges.

James Cowgill a écrit :
> Source: partclone
> Version: 0.2.78-1
> Severity: serious
> 
> Hi,
> 
> partclone failed to build on all non-x86 arches with this error:
> 
> > Making all in fail-mbr
> > make[3]: Entering directory '/«PKGBUILDDIR»/fail-mbr'
> > gcc -Wall -Werror -m32 -nostdlib -o fail-mbr.o fail-mbr.S
> > gcc: error: unrecognized command line option '-m32'
> > make[3]: *** [fail-mbr.o] Error 1
> > Makefile:481: recipe for target 'fail-mbr.o' failed
> > make[3]: Leaving directory '/«PKGBUILDDIR»/fail-mbr'
> > make[2]: *** [all-recursive] Error 1
> > Makefile:387: recipe for target 'all-recursive' failed
> > make[2]: Leaving directory '/«PKGBUILDDIR»'
> > make[1]: *** [all] Error 2
> > dh_auto_build: make -j1 returned exit code 2
> 
> Assembling fail-mbr.S with a non-x86 compiler clearly isn't going to work.
> 
> Thanks,
> James



-- 
Georges KHAZNADAR et Jocelyne FOURNIER
22 rue des mouettes, 59240 Dunkerque France.
Téléphone +33 (0)3 28 29 17 70

Attachment: signature.asc
Description: Digital signature

Reply via email to