On Mon, Jan 21, 2019 at 06:16:22PM +0100, Arnd Bergmann wrote:
> On Mon, Jan 21, 2019 at 4:40 AM Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
> >
> > Hi all,
> >
> > Today's linux-next merge of the pidfd tree got conflicts in:
> >
> > arch/x86/entry/syscalls/syscall_32.tbl
> > include/uapi/asm-generic/unistd.h
> >
> > between commit:
> >
> > 10a9a4dd92e6 ("arch: add split IPC system calls where needed")
> > b1788424aa2e ("y2038: add 64-bit time_t syscalls to all 32-bit architectures")
> >
> > from the y2038 tree and commit:
> >
> > 3d2991bc7a67 ("signal: add pidfd_send_signal() syscall")
> >
> > from the pidfd tree.
> >
> > I fixed it up (see below) and can carry the fix as necessary. This
> > is now fixed as far as linux-next is concerned, but any non trivial
> > conflicts should be mentioned to your upstream maintainer when your tree
> > is submitted for merging. You may also want to consider cooperating
> > with the maintainer of the conflicting tree to minimise any particularly
> > complex conflicts.
>
> Hi Stephen,
>
> > +385 i386 io_pgetevents sys_io_pgetevents_time32 __ia32_compat_sys_io_pgetevents
> > 386 i386 rseq sys_rseq __ia32_sys_rseq
> > + 387 i386 pidfd_send_signal sys_pidfd_send_signal __ia32_sys_pidfd_send_signal
> > +# room for arch specific syscalls
> > +393 i386 semget sys_semget __ia32_sys_semget
> > +394 i386 semctl sys_semctl __ia32_compat_sys_semctl
>
> > #define __NR_kexec_file_load 294
> > __SYSCALL(__NR_kexec_file_load, sys_kexec_file_load)
> > - /* 295 through 402 are unassigned to sync up with generic numbers */
> > + #define __NR_pidfd_send_signal 295
> > + __SYSCALL(__NR_pidfd_send_signal, sys_pidfd_send_signal)
> > ++/* 296 through 402 are unassigned to sync up with generic numbers */
> > +#if __BITS_PER_LONG == 32
> > +#define __NR_clock_gettime64 403
> > +__SYSCALL(__NR_clock_gettime64, sys_clock_gettime)
>
> If we merge my patches, then any other system calls should get numbers
> above 423 on all architectures. Part of what I did in my branch was to
> add all missing calls on architectures, and then move up to a common
> number for the newly added ones. Your merge works correctly, but
> it works against that idea by adding new numbers that conflict with
> existing ones on other architectures, e.g. 387 is already assigned
> on arm, microblaze, powerpc, and sh, and 294 is assigned almost
> everywhere other than the asm-generic version.
Hey Arnd,
I plan on sending the pidfd branch with the new pidfd_send_signal()
syscall for the 5.1 window. Should we somehow coordinate so that our
branches don't conflict? Any suggestions?
Christian
