system/linux
system/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux synced 2024-09-19 18:46:35 +00:00
Code Issues Packages Projects Releases Wiki Activity
linux/arch/sparc/kernel/syscalls.S
Christian Brauner dcad2a62bc
sparc64: enable HAVE_COPY_THREAD_TLS 
This is part of a larger series that aims at getting rid of the
copy_thread()/copy_thread_tls() split that makes the process creation
codepaths in the kernel more convoluted and error-prone than they need
to be.
It also unblocks implementing clone3() on architectures not support
copy_thread_tls(). Any architecture that wants to implement clone3()
will need to select HAVE_COPY_THREAD_TLS and thus need to implement
copy_thread_tls(). So both goals are connected but independently
beneficial.

HAVE_COPY_THREAD_TLS means that a given architecture supports
CLONE_SETTLS and not setting it should usually mean that the
architectures doesn't implement it but that's not how things are. In
fact all architectures support CLONE_TLS it's just that they don't
follow the calling convention that HAVE_COPY_THREAD_TLS implies. That
means all architectures can be switched over to select
HAVE_COPY_THREAD_TLS. Once that is done we can remove that macro (yay,
less code), the unnecessary do_fork() export in kernel/fork.c, and also
rename copy_thread_tls() back to copy_thread(). At this point
copy_thread() becomes the main architecture specific part of process
creation but it will be the same layout and calling convention for all
architectures. (Once that is done we can probably cleanup each
copy_thread() function even more but that's for the future.)

Since sparc does support CLONE_SETTLS there's no reason to not select
HAVE_COPY_THREAD_TLS. This brings us one step closer to getting rid of
the copy_thread()/copy_thread_tls() split we still have and ultimately
the HAVE_COPY_THREAD_TLS define in general. A lot of architectures have
already converted and sparc is one of the few hat haven't yet. This also
unblocks implementing the clone3() syscall on sparc which I will follow
up later (if no one gets there before me). Once that is done we can get
of another ARCH_WANTS_* macro.

This patch just switches sparc64 over to HAVE_COPY_THREAD_TLS but not
sparc32 which will be done in the next patch. Once Any architecture that
supports HAVE_COPY_THREAD_TLS cannot call the do_fork() helper anymore.
This is fine and intended since it should be removed in favor of the
new, cleaner _do_fork() calling convention based on struct
kernel_clone_args. In fact, most architectures have already switched.
With this patch, sparc joins the other arches which can't use the
fork(), vfork(), clone(), clone3() syscalls directly and who follow the
new process creation calling convention that is based on struct
kernel_clone_args which we introduced a while back. This means less
custom assembly in the architectures entry path to set up the registers
before calling into the process creation helper and it is easier to to
support new features without having to adapt calling conventions. It
also unifies all process creation paths between fork(), vfork(),
clone(), and clone3(). (We can't fix the ABI nightmare that legacy
clone() is but we can prevent stuff like this happening in the future.)

Note that sparc can't easily call into the syscalls directly because of
its return value conventions when a new process is created which
needs to clobber the UREG_I1 register in copy_thread{_tls()} and it
needs to restore it if process creation fails. That's not a big deal
since the new process creation calling convention makes things simpler.

This removes sparc_do_fork() and replaces it with 3 clean helpers,
sparc_fork(), sparc_vfork(), and sparc_clone(). That means a little more
C code until the next patch unifies sparc 32bit and sparc64. It has the
advantage that we can remove quite a bit of assembler and it makes the
whole syscall.S process creation bits easier to read.
The follow-up patch will remove the custom sparc_do_fork() helper for
32bi sparc and move sparc_fork(), sparc_vfork(), and sparc_clone() into
a common process.c file. This allows us to remove quite a bit of
custom assembly form 32bit sparc's entry.S file too and allows to remove
even more code because now all helpers are shared between 32bit sparc
and sparc64 instead of having to maintain two separate sparc_do_fork()
implementations.

For some more context, please see:
commit 606e9ad200
Merge: ac61145a72 457677c70c
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Sat Jan 11 15:33:48 2020 -0800

    Merge tag 'clone3-tls-v5.5-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

    Pull thread fixes from Christian Brauner:
     "This contains a series of patches to fix CLONE_SETTLS when used with
      clone3().

      The clone3() syscall passes the tls argument through struct clone_args
      instead of a register. This means, all architectures that do not
      implement copy_thread_tls() but still support CLONE_SETTLS via
      copy_thread() expecting the tls to be located in a register argument
      based on clone() are currently unfortunately broken. Their tls value
      will be garbage.

      The patch series fixes this on all architectures that currently define
      __ARCH_WANT_SYS_CLONE3. It also adds a compile-time check to ensure
      that any architecture that enables clone3() in the future is forced to
      also implement copy_thread_tls().

      My ultimate goal is to get rid of the copy_thread()/copy_thread_tls()
      split and just have copy_thread_tls() at some point in the not too
      distant future (Maybe even renaming copy_thread_tls() back to simply
      copy_thread() once the old function is ripped from all arches). This
      is dependent now on all arches supporting clone3().

      While all relevant arches do that now there are still four missing:
      ia64, m68k, sh and sparc. They have the system call reserved, but not
      implemented. Once they all implement clone3() we can get rid of
      ARCH_WANT_SYS_CLONE3 and HAVE_COPY_THREAD_TLS.

Note that in the meantime, m68k has already switched to the new calling
convention.

Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Guo Ren <guoren@kernel.org>
Cc: linux-csky@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: sparclinux@vger.kernel.org
See: d95b56c77e ("openrisc: Cleanup copy_thread_tls docs and comments")
See: 0b9f386c4b ("csky: Implement copy_thread_tls")
Link: https://lore.kernel.org/r/20200512171527.570109-2-christian.brauner@ubuntu.com
2020-06-23 10:49:54 +02:00

302 lines
7.5 KiB
ArmAsm
Raw Blame History

/* SPDX-License-Identifier: GPL-2.0 */
/* SunOS's execv() call only specifies the argv argument, the
* environment settings are the same as the calling processes.
*/
sys64_execve:
set sys_execve, %g1
jmpl %g1, %g0
flushw
sys64_execveat:
set sys_execveat, %g1
jmpl %g1, %g0
flushw
#ifdef CONFIG_COMPAT
sunos_execv:
mov %g0, %o2
sys32_execve:
set compat_sys_execve, %g1
jmpl %g1, %g0
flushw
sys32_execveat:
set compat_sys_execveat, %g1
jmpl %g1, %g0
flushw
#endif
.align 32
#ifdef CONFIG_COMPAT
sys32_sigstack:
ba,pt %xcc, do_sys32_sigstack
mov %i6, %o2
#endif
.align 32
#ifdef CONFIG_COMPAT
sys32_sigreturn:
add %sp, PTREGS_OFF, %o0
call do_sigreturn32
add %o7, 1f-.-4, %o7
nop
#endif
sys_rt_sigreturn:
add %sp, PTREGS_OFF, %o0
call do_rt_sigreturn
add %o7, 1f-.-4, %o7
nop
#ifdef CONFIG_COMPAT
sys32_rt_sigreturn:
add %sp, PTREGS_OFF, %o0
call do_rt_sigreturn32
add %o7, 1f-.-4, %o7
nop
#endif
.align 32
1: ldx [%g6 + TI_FLAGS], %l5
andcc %l5, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0
be,pt %icc, rtrap
nop
call syscall_trace_leave
add %sp, PTREGS_OFF, %o0
ba,pt %xcc, rtrap
nop
/* This is how fork() was meant to be done, 8 instruction entry.
*
* I questioned the following code briefly, let me clear things
* up so you must not reason on it like I did.
*
* Know the fork_kpsr etc. we use in the sparc32 port? We don't
* need it here because the only piece of window state we copy to
* the child is the CWP register. Even if the parent sleeps,
* we are safe because we stuck it into pt_regs of the parent
* so it will not change.
*
* XXX This raises the question, whether we can do the same on
* XXX sparc32 to get rid of fork_kpsr _and_ fork_kwim. The
* XXX answer is yes. We stick fork_kpsr in UREG_G0 and
* XXX fork_kwim in UREG_G1 (global registers are considered
* XXX volatile across a system call in the sparc ABI I think
* XXX if it isn't we can use regs->y instead, anyone who depends
* XXX upon the Y register being preserved across a fork deserves
* XXX to lose).
*
* In fact we should take advantage of that fact for other things
* during system calls...
*/
.align 32
sys_vfork:
flushw
ba,pt %xcc, sparc_vfork
add %sp, PTREGS_OFF, %o0
.align 32
sys_fork:
flushw
ba,pt %xcc, sparc_fork
add %sp, PTREGS_OFF, %o0
.align 32
sys_clone:
flushw
ba,pt %xcc, sparc_clone
add %sp, PTREGS_OFF, %o0
.globl ret_from_fork
ret_from_fork:
/* Clear current_thread_info()->new_child. */
stb %g0, [%g6 + TI_NEW_CHILD]
call schedule_tail
mov %g7, %o0
ldx [%sp + PTREGS_OFF + PT_V9_I0], %o0
brnz,pt %o0, ret_sys_call
ldx [%g6 + TI_FLAGS], %l0
ldx [%sp + PTREGS_OFF + PT_V9_G1], %l1
call %l1
ldx [%sp + PTREGS_OFF + PT_V9_G2], %o0
ba,pt %xcc, ret_sys_call
mov 0, %o0
.globl sparc_exit_group
.type sparc_exit_group,#function
sparc_exit_group:
sethi %hi(sys_exit_group), %g7
ba,pt %xcc, 1f
or %g7, %lo(sys_exit_group), %g7
.size sparc_exit_group,.-sparc_exit_group
.globl sparc_exit
.type sparc_exit,#function
sparc_exit:
sethi %hi(sys_exit), %g7
or %g7, %lo(sys_exit), %g7
1: rdpr %pstate, %g2
wrpr %g2, PSTATE_IE, %pstate
rdpr %otherwin, %g1
rdpr %cansave, %g3
add %g3, %g1, %g3
wrpr %g3, 0x0, %cansave
wrpr %g0, 0x0, %otherwin
wrpr %g2, 0x0, %pstate
jmpl %g7, %g0
stb %g0, [%g6 + TI_WSAVED]
.size sparc_exit,.-sparc_exit
linux_sparc_ni_syscall:
sethi %hi(sys_ni_syscall), %l7
ba,pt %xcc, 4f
or %l7, %lo(sys_ni_syscall), %l7
linux_syscall_trace32:
call syscall_trace_enter
add %sp, PTREGS_OFF, %o0
brnz,pn %o0, 3f
mov -ENOSYS, %o0
/* Syscall tracing can modify the registers. */
ldx [%sp + PTREGS_OFF + PT_V9_G1], %g1
sethi %hi(sys_call_table32), %l7
ldx [%sp + PTREGS_OFF + PT_V9_I0], %i0
or %l7, %lo(sys_call_table32), %l7
ldx [%sp + PTREGS_OFF + PT_V9_I1], %i1
ldx [%sp + PTREGS_OFF + PT_V9_I2], %i2
ldx [%sp + PTREGS_OFF + PT_V9_I3], %i3
ldx [%sp + PTREGS_OFF + PT_V9_I4], %i4
ldx [%sp + PTREGS_OFF + PT_V9_I5], %i5
cmp %g1, NR_syscalls
bgeu,pn %xcc, 3f
mov -ENOSYS, %o0
sll %g1, 2, %l4
srl %i0, 0, %o0
lduw [%l7 + %l4], %l7
srl %i4, 0, %o4
srl %i1, 0, %o1
srl %i2, 0, %o2
ba,pt %xcc, 5f
srl %i3, 0, %o3
linux_syscall_trace:
call syscall_trace_enter
add %sp, PTREGS_OFF, %o0
brnz,pn %o0, 3f
mov -ENOSYS, %o0
/* Syscall tracing can modify the registers. */
ldx [%sp + PTREGS_OFF + PT_V9_G1], %g1
sethi %hi(sys_call_table64), %l7
ldx [%sp + PTREGS_OFF + PT_V9_I0], %i0
or %l7, %lo(sys_call_table64), %l7
ldx [%sp + PTREGS_OFF + PT_V9_I1], %i1
ldx [%sp + PTREGS_OFF + PT_V9_I2], %i2
ldx [%sp + PTREGS_OFF + PT_V9_I3], %i3
ldx [%sp + PTREGS_OFF + PT_V9_I4], %i4
ldx [%sp + PTREGS_OFF + PT_V9_I5], %i5
cmp %g1, NR_syscalls
bgeu,pn %xcc, 3f
mov -ENOSYS, %o0
sll %g1, 2, %l4
mov %i0, %o0
lduw [%l7 + %l4], %l7
mov %i1, %o1
mov %i2, %o2
mov %i3, %o3
b,pt %xcc, 2f
mov %i4, %o4
/* Linux 32-bit system calls enter here... */
.align 32
.globl linux_sparc_syscall32
linux_sparc_syscall32:
/* Direct access to user regs, much faster. */
cmp %g1, NR_syscalls ! IEU1 Group
bgeu,pn %xcc, linux_sparc_ni_syscall ! CTI
srl %i0, 0, %o0 ! IEU0
sll %g1, 2, %l4 ! IEU0 Group
srl %i4, 0, %o4 ! IEU1
lduw [%l7 + %l4], %l7 ! Load
srl %i1, 0, %o1 ! IEU0 Group
ldx [%g6 + TI_FLAGS], %l0 ! Load
srl %i3, 0, %o3 ! IEU0
srl %i2, 0, %o2 ! IEU0 Group
andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0
bne,pn %icc, linux_syscall_trace32 ! CTI
mov %i0, %l5 ! IEU1
5: call %l7 ! CTI Group brk forced
srl %i5, 0, %o5 ! IEU1
ba,pt %xcc, 3f
sra %o0, 0, %o0
/* Linux native system calls enter here... */
.align 32
.globl linux_sparc_syscall
linux_sparc_syscall:
/* Direct access to user regs, much faster. */
cmp %g1, NR_syscalls ! IEU1 Group
bgeu,pn %xcc, linux_sparc_ni_syscall ! CTI
mov %i0, %o0 ! IEU0
sll %g1, 2, %l4 ! IEU0 Group
mov %i1, %o1 ! IEU1
lduw [%l7 + %l4], %l7 ! Load
4: mov %i2, %o2 ! IEU0 Group
ldx [%g6 + TI_FLAGS], %l0 ! Load
mov %i3, %o3 ! IEU1
mov %i4, %o4 ! IEU0 Group
andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0
bne,pn %icc, linux_syscall_trace ! CTI Group
mov %i0, %l5 ! IEU0
2: call %l7 ! CTI Group brk forced
mov %i5, %o5 ! IEU0
nop
3: stx %o0, [%sp + PTREGS_OFF + PT_V9_I0]
ret_sys_call:
ldx [%sp + PTREGS_OFF + PT_V9_TSTATE], %g3
mov %ulo(TSTATE_XCARRY | TSTATE_ICARRY), %g2
sllx %g2, 32, %g2
cmp %o0, -ERESTART_RESTARTBLOCK
bgeu,pn %xcc, 1f
andcc %l0, (_TIF_SYSCALL_TRACE|_TIF_SECCOMP|_TIF_SYSCALL_AUDIT|_TIF_SYSCALL_TRACEPOINT|_TIF_NOHZ), %g0
ldx [%sp + PTREGS_OFF + PT_V9_TNPC], %l1 ! pc = npc
2:
/* System call success, clear Carry condition code. */
andn %g3, %g2, %g3
3:
stx %g3, [%sp + PTREGS_OFF + PT_V9_TSTATE]
bne,pn %icc, linux_syscall_trace2
add %l1, 0x4, %l2 ! npc = npc+4
stx %l1, [%sp + PTREGS_OFF + PT_V9_TPC]
ba,pt %xcc, rtrap
stx %l2, [%sp + PTREGS_OFF + PT_V9_TNPC]
1:
/* Check if force_successful_syscall_return()
* was invoked.
*/
ldub [%g6 + TI_SYS_NOERROR], %l2
brnz,pn %l2, 2b
ldx [%sp + PTREGS_OFF + PT_V9_TNPC], %l1 ! pc = npc
/* System call failure, set Carry condition code.
* Also, get abs(errno) to return to the process.
*/
sub %g0, %o0, %o0
stx %o0, [%sp + PTREGS_OFF + PT_V9_I0]
ba,pt %xcc, 3b
or %g3, %g2, %g3
linux_syscall_trace2:
call syscall_trace_leave
add %sp, PTREGS_OFF, %o0
stx %l1, [%sp + PTREGS_OFF + PT_V9_TPC]
ba,pt %xcc, rtrap
stx %l2, [%sp + PTREGS_OFF + PT_V9_TNPC]
Reference in a new issue View git blame Copy permalink