Merge tag 'y2038-syscall-cleanup' of git://git.kernel.org:/pub/scm/linux/kernel/git/arnd/playground into timers/2038

Pull preparatory work for y2038 changes from Arnd Bergmann:

System call unification and cleanup

The system call tables have diverged a bit over the years, and a number of
the recent additions never made it into all architectures, for one reason
or another.

This is an attempt to clean it up as far as we can without breaking
compatibility, doing a number of steps:

 - Add system calls that have not yet been integrated into all architectures
   but that we definitely want there. This includes {,f}statfs64() and
   get{eg,eu,g,p,u,pp}id() on alpha, which have been missing traditionally.

 - The s390 compat syscall handling is cleaned up to be more like what we
   do on other architectures, while keeping the 31-bit pointer
   extension. This was merged as a shared branch by the s390 maintainers
   and is included here in order to base the other patches on top.

 - Add the separate ipc syscalls on all architectures that traditionally
   only had sys_ipc(). This version is done without support for IPC_OLD
   that is we have in sys_ipc. The new semtimedop_time64 syscall will only
   be added here, not in sys_ipc

 - Add syscall numbers for a couple of syscalls that we probably don't need
   everywhere, in particular pkey_* and rseq, for the purpose of symmetry:
   if it's in asm-generic/unistd.h, it makes sense to have it everywhere. I
   expect that any future system calls will get assigned on all platforms
   together, even when they appear to be specific to a single architecture.

 - Prepare for having the same system call numbers for any future calls. In
   combination with the generated tables, this hopefully makes it easier to
   add new calls across all architectures together.

All of the above are technically separate from the y2038 work, but are done
as preparation before we add the new 64-bit time_t system calls everywhere,
providing a common baseline set of system calls.

I expect that glibc and other libraries that want to use 64-bit time_t will
require linux-5.1 kernel headers for building in the future, and at a much
later point may also require linux-5.1 or a later version as the minimum
kernel at runtime. Having a common baseline then allows the removal of many
architecture or kernel version specific workarounds.

arch/xtensa/kernel/syscalls/syscall.tbl

@@ -103,7 +103,7 @@
 91	common	madvise				sys_madvise
 92	common	shmget				sys_shmget
 93	common	shmat				xtensa_shmat
-94	common	shmctl				sys_shmctl
+94	common	shmctl				sys_old_shmctl
 95	common	shmdt				sys_shmdt
 # Socket Operations
 96	common	socket				sys_socket
@@ -177,12 +177,12 @@
 161	common	semtimedop			sys_semtimedop
 162	common	semget				sys_semget
 163	common	semop				sys_semop
-164	common	semctl				sys_semctl
+164	common	semctl				sys_old_semctl
 165	common	available165			sys_ni_syscall
 166	common	msgget				sys_msgget
 167	common	msgsnd				sys_msgsnd
 168	common	msgrcv				sys_msgrcv
-169	common	msgctl				sys_msgctl
+169	common	msgctl				sys_old_msgctl
 170	common	available170			sys_ni_syscall
 # File System
 171	common	umount2				sys_umount
@@ -372,3 +372,4 @@
 349	common	pkey_alloc			sys_pkey_alloc
 350	common	pkey_free			sys_pkey_free
 351	common	statx				sys_statx
+352	common	rseq				sys_rseq