On Wed, May 19, 2021 at 11:00 PM Arnd Bergmann [off-list ref] wrote:

On Wed, May 19, 2021 at 10:33 PM Thomas Gleixner [off-list ref] wrote:

quoted

On Mon, May 17 2021 at 22:33, Arnd Bergmann wrote:

quoted

From: Arnd Bergmann <arnd@arndb.de>

These are all handled correctly when calling the native
system call entry point, so remove the special cases.
 arch/x86/entry/syscall_x32.c              |  2 ++
 arch/x86/entry/syscalls/syscall_32.tbl    |  6 ++--
 arch/x86/entry/syscalls/syscall_64.tbl    |  4 +--

That conflicts with

  https://lore.kernel.org/lkml/20210517073815.97426-1-masahiroy@kernel.org/ (local)

which I'm picking up. We have more changes in that area coming in.

Ok, thanks for the heads-up. I'll try a merge or rebase to see how this can be
handled. If both the drivers/net and drivers/media get picked up for 5.14, maybe
the rebased patches can go through -mm on top, along with the final
removal of compat_alloc_user_space()/copy_in_user(). If not, I suppose these
four patches can also wait another release.

On second thought, this patch 4/4 is not even required here to kill off
compat_alloc_user_space, so the easiest alternative might be to merge the
other patches first, and then do this part together with the removal of
the unused functions in a follow-up series.

        Arnd