Re: RFS: python-cartopy/0.14.2-2
On Thu, 2017-01-19 at 15:17 +0100, Jerome Kieffer wrote:
> On Thu, 19 Jan 2017 14:13:24 +0000
> James Clarke <jrtc27@debian.org> wrote:
>
> > On 19 Jan 2017, at 13:29, Jerome Kieffer <Jerome.Kieffer@esrf.fr> wrote:
> > >
> > > On Thu, 19 Jan 2017 11:50:18 +0000
> > > Ghislain Vaillant <ghisvail@gmail.com> wrote:
> > >
> > >
> > > > > > Then there's the additional complication that x32 is a 32-bit version of
> > > > > > amd64, so it shouldn't fail the test. Therefore, I would propose
> > > > > > something like the following Debian-specific hack:
> > > > > >
> > > > > > host_cpu = subprocess.check_output(['dpkg-architecture', '-qDEB_HOST_ARCH_CPU'])
> > > > > > if host_cpu != 'i386':
> > > > > > # Do the assert
> > >
> > > Have you tried platfrom.machine ? it return either "i386" or "i686"
> > >
> > > It is cleaner but I don't know how it behaves with an "x32" kernel.
> >
> > That doesn't help with i386 (or x32) chroots on an amd64 kernel.
> >
> > root@debian:/# uname -a
> > Linux debian 4.9.0-1-amd64 #1 SMP Debian 4.9.2-2 (2017-01-12) x86_64 GNU/Linux
> > root@debian:/# dpkg-architecture -qDEB_HOST_ARCH_CPU
> > i386
> > root@debian:/# python3 -c "import platform; print(platform.machine())"
> > x86_64
>
> If one wants to know if you are running a python in 32 or 64 bits:
>
> python -c "print(tuple.__itemsize__*8)"
>
> This is independent on the operating system.
> Cheers,
We are interested in a *subset* of all 32-bit architectures (i386
family).
Ghis
Reply to: