[PATCH 4/6] sysvipc: Return EINVAL for invalid msgctl commands

[Florian Weimer]

* Adhemerval Zanella:

> On 29/09/2020 04:58, Florian Weimer wrote:
>> * Adhemerval Zanella via Libc-alpha:
>> 
>>> It avoids regressions on possible future commands that might require
>>> additional libc support.  The downside is new commands added by newer
>>> kernels will need further glibc support.
>> 
>> Looks okay to me for now.
>> 
>> I think eventually we need to rethink what we want to do about these,
>> for the architectures that don't need translation.
>
> We already have __IPC_TIME64 to tell apart the ABI that require it, 
> we might to put the switch only if it is set.  However, I am not sure
> if it is really an improvement, since we will still need to revise
> it for architecture with __IPC_TIME64 == 1 anyway.

It's an improvement for users on architectures which can use this
facility.  I think we should not artificially hold back architectures
with full userspace/kernel alignment just because there are other
architectures out there which are more complicated.

Thanks,
Florian
-- 
Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill