On 3 September 2012 17:28, Yann E. MORIN <yann.morin.1...@free.fr> wrote:
> On Monday 03 September 2012 17:44:51 Peter Maydell wrote:
>> Personally I think it might indeed be a good idea to just say
>> "statically linked softmmu isn't supported" and forbid it, unless
>> somebody has a good use case for it...
>
> I personnally have such a use-case: provide run-everywhere qemu softmmu
> binaries that do have minimal requirements on the distro they'll be
> running on, especially when the distros can be very different (in kinds
> and in age).

...have you audited all of qemu's library usage to confirm that it
doesn't use any of the bits of glibc that don't work in the static
linking scenario? I'm reasonably confident that the linux-user binaries
are OK, much less so for the system binaries...

-- PMM

Reply via email to