Hello,
On Wed, Aug 07, 2024 at 09:11:01AM +0200, Arnd Bergmann wrote:
> On Wed, Aug 7, 2024, at 00:50, Namhyung Kim wrote:
> > And arch syscall tables to pick up changes from:
> >
> > b1e31c134a8a powerpc: restore some missing spu syscalls
> > d3882564a77c syscalls: fix compat_sys_io_pgetevents_time64 usage
> > 54233a425403 uretprobe: change syscall number, again
> > 63ded110979b uprobe: Change uretprobe syscall scope and number
> > 9142be9e6443 x86/syscall: Mark exit[_group] syscall handlers __noreturn
> > 9aae1baa1c5d x86, arm: Add missing license tag to syscall tables files
> > 5c28424e9a34 syscalls: Fix to add sys_uretprobe to syscall.tbl
> > 190fec72df4a uprobe: Wire up uretprobe system call
> >
> > This should be used to beautify syscall arguments and it addresses
> > these tools/perf build warnings:
> >
> > Warning: Kernel ABI header differences:
> > diff -u tools/arch/arm64/include/uapi/asm/unistd.h
> > arch/arm64/include/uapi/asm/unistd.h
> > diff -u tools/include/uapi/asm-generic/unistd.h
> > include/uapi/asm-generic/unistd.h
> > diff -u tools/perf/arch/x86/entry/syscalls/syscall_64.tbl
> > arch/x86/entry/syscalls/syscall_64.tbl
> > diff -u tools/perf/arch/powerpc/entry/syscalls/syscall.tbl
> > arch/powerpc/kernel/syscalls/syscall.tbl
> > diff -u tools/perf/arch/s390/entry/syscalls/syscall.tbl
> > arch/s390/kernel/syscalls/syscall.tbl
> >
> > Please see tools/include/uapi/README for details (it's in the first patch
> > of this series).
> >
> > Cc: Arnd Bergmann <arnd@xxxxxxxx>
> > Cc: linux-arch@xxxxxxxxxxxxxxx
> > Signed-off-by: Namhyung Kim <namhyung@xxxxxxxxxx>
> > ---
> > tools/arch/arm64/include/uapi/asm/unistd.h | 24 +------------------
> > tools/include/uapi/asm-generic/unistd.h | 2 +-
> > .../arch/powerpc/entry/syscalls/syscall.tbl | 6 ++++-
> > .../perf/arch/s390/entry/syscalls/syscall.tbl | 2 +-
> > .../arch/x86/entry/syscalls/syscall_64.tbl | 8 ++++---
> > 5 files changed, 13 insertions(+), 29 deletions(-)
> >
> > diff --git a/tools/arch/arm64/include/uapi/asm/unistd.h
> > b/tools/arch/arm64/include/uapi/asm/unistd.h
> > index 9306726337fe..df36f23876e8 100644
> > --- a/tools/arch/arm64/include/uapi/asm/unistd.h
> > +++ b/tools/arch/arm64/include/uapi/asm/unistd.h
> > -
> > -#define __ARCH_WANT_RENAMEAT
> > -#define __ARCH_WANT_NEW_STAT
> > -#define __ARCH_WANT_SET_GET_RLIMIT
> > -#define __ARCH_WANT_TIME32_SYSCALLS
> > -#define __ARCH_WANT_MEMFD_SECRET
> > -
> > -#include <asm-generic/unistd.h>
> > +#include <asm/unistd_64.h>
>
> This part won't work by itself, since you don't pick up
> the generated asm/unistd_64.h header but keep the old
> asm-generic/unistd.h header. Both have the same contents,
> so the easy way to do this is to just keep the existing
> version of the arm64 header for 6.11 and add a script to
> generate it in 6.12 the way we do for x86, using an
> architecture-independent script.
Thanks for the review, I'll drop the arm64 parts for now.
>
> > @@ -68,7 +69,7 @@
> > 57 common fork sys_fork
> > 58 common vfork sys_vfork
> > 59 64 execve sys_execve
> > -60 common exit sys_exit
> > +60 common exit sys_exit - noreturn
> > 61 common wait4 sys_wait4
> > 62 common kill sys_kill
> > 63 common uname sys_newuname
>
> Have you checked if this works correctly with the
> existing tools/perf/arch/x86/entry/syscalls/syscalltbl.sh?
Yep, the script only cares about the number and the name.
# the params are: nr abi name entry compat
# use _ for intentionally unused variables according to SC2034
while read nr _ name _ _; do
if [ $nr -ge 512 ] ; then # discard compat sycalls
break
fi
emit "$nr" "$name"
max_nr=$nr
done < $sorted_table
The only difference I see was the added uretprobe syscall.
$ tools/arch/x86/entry/syscalls/syscalltbl.sh tools/arch/x86/entry/syscalls/syscall_64.tbl x86_64 > b
(apply this series...)
$ tools/arch/x86/entry/syscalls/syscalltbl.sh tools/arch/x86/entry/syscalls/syscall_64.tbl x86_64 > a
$ diff -u b a
--- b 2024-08-07 10:28:04.267738574 -0700
+++ a 2024-08-07 10:28:18.543965369 -0700
@@ -334,6 +334,7 @@
[332] = "statx",
[333] = "io_pgetevents",
[334] = "rseq",
+ [335] = "uretprobe",
[424] = "pidfd_send_signal",
[425] = "io_uring_setup",
[426] = "io_uring_enter",
Thanks,
Namhyung
>
> Since not just table file contents but also the file format
> changed here, there is a good chance that the output
> is no longer what we need.
>
> Unfortunately, the format on x86 is now incompatible with
> the one on s390. I have a patch to change s390 in the future
> so we can use a single script for all of them.
>
> Arnd
