Re: [PATCH v2] x86: Fix x32 System V message queue syscalls

[Harald van Dijk]

On 01/08/2023 13:15, Harald van Dijk wrote:
> On 01/08/2023 08:15, Florian Weimer wrote:
> > * Harald van Dijk:
> >
> > > There is one complication that I think has not been mentioned yet:
> > > when _GNU_SOURCE is defined, glibc does provide a definition of struct
> > > msghdr in <sys/msg.h> with a field "__syscall_slong_t mtype;". This
> > > makes it slightly more likely that there is code out there in the wild
> > > that works fine with current kernels and would be broken by the
> > > fix. Given how rare x32 is, and how rare message queues are, this may
> > > still be acceptable, but I am mentioning it just in case this would
> > > cause a different approach to be preferred. And whatever is done, a
> > > fix should also be submitted to glibc.
> >
> > What should glibc do here?  Just change the definition in the header to
> > long and ignore the breakage?
>
> Yes, I believe so, but perhaps for glibc it might make sense to wait a 
> little bit to see if it does indeed go into the kernel in that form. If 
> the kernel ends up preferring something else, the change needed in glibc 
> might also be different.

With the patches sent just now, following the guidance from Andy in 
<https://lore.kernel.org/all/CALCETrUuBR3Pt_9NhRZTLzjZzwdsS2OPW4U2r31_1Uq-=poRDw@xxxxxxxxxxxxxx/>, 
assuming they get accepted, the right thing to do in glibc would indeed 
be to just change the header to use long. It will not affect anything 
other than x32, and most code avoids that header anyway and defines the 
type manually as per the documentation.

Cheers,
Harald van Dijk