Re: [PATCH RFC 1/4] syscalls: Create unified partial table for all archs
André Almeida <[email protected]> Fri, 24 Jul 2026 17:18:48 -0300
| Newsgroups | org.kernel.vger.linux-kbuild,org.kernel.vger.linux-kernel |
|---|---|
| Message-ID | <[email protected]> |
Em 24/07/2026 17:09, Arnd Bergmann escreveu: > On Fri, Jul 24, 2026, at 22:00, André Almeida wrote: >> To take advantage of the subset of syscall numbers that are guaranteed to >> be shared, create a new table and adapt generation scripts to use it. In >> that way, every new syscall can be added to a single file. >> >> Signed-off-by: André Almeida <[email protected]> > > Thanks a lot for working on this, I had planned to do this myself > for a while but had never quite gotten around to finish it. > >> +# >> +# syscall numbers shared by architectures >> + >> +442 common mount_setattr sys_mount_setattr >> +443 common quotactl_fd sys_quotactl_fd >> +444 common landlock_create_ruleset sys_landlock_create_ruleset >> +445 common landlock_add_rule sys_landlock_add_rule > > It should really be anything starting at 403, not just 442 and higher. > > There are some trivial differences that have crept in over time, > but we should be able to work around those. > The thing that prevent me starting before 442 was how to deal with compat. If I add to the table 441 common epoll_pwait2 sys_epoll_pwait2 compat_sys_epoll_pwait2 It will wrongly add __SYSCALL_WITH_COMPAT(441, sys_epoll_pwait2, compat_sys_epoll_pwait2) to arch/x86/include/generated/asm/syscalls_64.h. Maybe I could add a --ignore-compat to scripts/syscalltbl.sh, and add this flag for 64 builds. What do you think? >> index 22e34cd46b9b..77934a2d0f51 100755 >> --- a/scripts/syscallhdr.sh >> +++ b/scripts/syscallhdr.sh >> @@ -26,6 +26,7 @@ usage() { >> echo >&2 " --emit-nr Emit the macro of the number of >> syscalls (__NR_syscalls)" >> echo >&2 " --offset OFFSET The offset of syscall numbers" >> echo >&2 " --prefix PREFIX The prefix to the macro like >> __NR_<PREFIX><NAME>" >> + echo >&2 " --common-tbl Use the common number table" >> exit 1 > > Ideally we shouldn't even need this bit either, though it may > help for the transition phase. > > Arnd