| Message ID | 20181220201502.12285-1-richard.henderson@linaro.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | linux-user: Add safe_syscall for riscv64 host | expand |
On Thu, 20 Dec 2018 at 20:16, Richard Henderson <richard.henderson@linaro.org> wrote: > > Signed-off-by: Richard Henderson <richard.henderson@linaro.org> > --- > > At some point we should make this routine be non-optional for > porting to a new host. Yes, I agree -- how many hosts do we still have which are missing support for it? thanks -- PMM
On 12/20/18 12:40 PM, Peter Maydell wrote: > On Thu, 20 Dec 2018 at 20:16, Richard Henderson > <richard.henderson@linaro.org> wrote: >> >> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> >> --- >> >> At some point we should make this routine be non-optional for >> porting to a new host. > > Yes, I agree -- how many hosts do we still have which are > missing support for it? Ignoring tci, mips (3 abis), ppc32, sparc (2 abis). r~
On Thu, Dec 20, 2018 at 12:15 PM Richard Henderson <richard.henderson@linaro.org> wrote: > > Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Alistair > --- > > At some point we should make this routine be non-optional for > porting to a new host. > > > r~ > > --- > linux-user/host/riscv64/hostdep.h | 23 +++++++ > linux-user/host/riscv64/safe-syscall.inc.S | 77 ++++++++++++++++++++++ > 2 files changed, 100 insertions(+) > create mode 100644 linux-user/host/riscv64/safe-syscall.inc.S > > diff --git a/linux-user/host/riscv64/hostdep.h b/linux-user/host/riscv64/hostdep.h > index 28467ba00b..865f0fb9ff 100644 > --- a/linux-user/host/riscv64/hostdep.h > +++ b/linux-user/host/riscv64/hostdep.h > @@ -8,4 +8,27 @@ > #ifndef RISCV64_HOSTDEP_H > #define RISCV64_HOSTDEP_H > > +/* We have a safe-syscall.inc.S */ > +#define HAVE_SAFE_SYSCALL > + > +#ifndef __ASSEMBLER__ > + > +/* These are defined by the safe-syscall.inc.S file */ > +extern char safe_syscall_start[]; > +extern char safe_syscall_end[]; > + > +/* Adjust the signal context to rewind out of safe-syscall if we're in it */ > +static inline void rewind_if_in_safe_syscall(void *puc) > +{ > + ucontext_t *uc = puc; > + unsigned long *pcreg = &uc->uc_mcontext.__gregs[REG_PC]; > + > + if (*pcreg > (uintptr_t)safe_syscall_start > + && *pcreg < (uintptr_t)safe_syscall_end) { > + *pcreg = (uintptr_t)safe_syscall_start; > + } > +} > + > +#endif /* __ASSEMBLER__ */ > + > #endif > diff --git a/linux-user/host/riscv64/safe-syscall.inc.S b/linux-user/host/riscv64/safe-syscall.inc.S > new file mode 100644 > index 0000000000..9ca3fbfd1e > --- /dev/null > +++ b/linux-user/host/riscv64/safe-syscall.inc.S > @@ -0,0 +1,77 @@ > +/* > + * safe-syscall.inc.S : host-specific assembly fragment > + * to handle signals occurring at the same time as system calls. > + * This is intended to be included by linux-user/safe-syscall.S > + * > + * Written by Richard Henderson <rth@twiddle.net> > + * Copyright (C) 2018 Linaro, Inc. > + * > + * This work is licensed under the terms of the GNU GPL, version 2 or later. > + * See the COPYING file in the top-level directory. > + */ > + > + .global safe_syscall_base > + .global safe_syscall_start > + .global safe_syscall_end > + .type safe_syscall_base, @function > + .type safe_syscall_start, @function > + .type safe_syscall_end, @function > + > + /* > + * This is the entry point for making a system call. The calling > + * convention here is that of a C varargs function with the > + * first argument an 'int *' to the signal_pending flag, the > + * second one the system call number (as a 'long'), and all further > + * arguments being syscall arguments (also 'long'). > + * We return a long which is the syscall's return value, which > + * may be negative-errno on failure. Conversion to the > + * -1-and-errno-set convention is done by the calling wrapper. > + */ > +safe_syscall_base: > + .cfi_startproc > + /* > + * The syscall calling convention is nearly the same as C: > + * we enter with a0 == *signal_pending > + * a1 == syscall number > + * a2 ... a7 == syscall arguments > + * and return the result in a0 > + * and the syscall instruction needs > + * a7 == syscall number > + * a0 ... a5 == syscall arguments > + * and returns the result in a0 > + * Shuffle everything around appropriately. > + */ > + mv t0, a0 /* signal_pending pointer */ > + mv t1, a1 /* syscall number */ > + mv a0, a2 /* syscall arguments */ > + mv a1, a3 > + mv a2, a4 > + mv a3, a5 > + mv a4, a6 > + mv a5, a7 > + mv a7, t1 > + > + /* > + * This next sequence of code works in conjunction with the > + * rewind_if_safe_syscall_function(). If a signal is taken > + * and the interrupted PC is anywhere between 'safe_syscall_start' > + * and 'safe_syscall_end' then we rewind it to 'safe_syscall_start'. > + * The code sequence must therefore be able to cope with this, and > + * the syscall instruction must be the final one in the sequence. > + */ > +safe_syscall_start: > + /* If signal_pending is non-zero, don't do the call */ > + lw t1, 0(t0) > + bnez t1, 0f > + scall > +safe_syscall_end: > + /* code path for having successfully executed the syscall */ > + ret > + > +0: > + /* code path when we didn't execute the syscall */ > + li a0, -TARGET_ERESTARTSYS > + ret > + .cfi_endproc > + > + .size safe_syscall_base, .-safe_syscall_base > -- > 2.17.2 > >
diff --git a/linux-user/host/riscv64/hostdep.h b/linux-user/host/riscv64/hostdep.h index 28467ba00b..865f0fb9ff 100644 --- a/linux-user/host/riscv64/hostdep.h +++ b/linux-user/host/riscv64/hostdep.h @@ -8,4 +8,27 @@ #ifndef RISCV64_HOSTDEP_H #define RISCV64_HOSTDEP_H +/* We have a safe-syscall.inc.S */ +#define HAVE_SAFE_SYSCALL + +#ifndef __ASSEMBLER__ + +/* These are defined by the safe-syscall.inc.S file */ +extern char safe_syscall_start[]; +extern char safe_syscall_end[]; + +/* Adjust the signal context to rewind out of safe-syscall if we're in it */ +static inline void rewind_if_in_safe_syscall(void *puc) +{ + ucontext_t *uc = puc; + unsigned long *pcreg = &uc->uc_mcontext.__gregs[REG_PC]; + + if (*pcreg > (uintptr_t)safe_syscall_start + && *pcreg < (uintptr_t)safe_syscall_end) { + *pcreg = (uintptr_t)safe_syscall_start; + } +} + +#endif /* __ASSEMBLER__ */ + #endif diff --git a/linux-user/host/riscv64/safe-syscall.inc.S b/linux-user/host/riscv64/safe-syscall.inc.S new file mode 100644 index 0000000000..9ca3fbfd1e --- /dev/null +++ b/linux-user/host/riscv64/safe-syscall.inc.S @@ -0,0 +1,77 @@ +/* + * safe-syscall.inc.S : host-specific assembly fragment + * to handle signals occurring at the same time as system calls. + * This is intended to be included by linux-user/safe-syscall.S + * + * Written by Richard Henderson <rth@twiddle.net> + * Copyright (C) 2018 Linaro, Inc. + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + */ + + .global safe_syscall_base + .global safe_syscall_start + .global safe_syscall_end + .type safe_syscall_base, @function + .type safe_syscall_start, @function + .type safe_syscall_end, @function + + /* + * This is the entry point for making a system call. The calling + * convention here is that of a C varargs function with the + * first argument an 'int *' to the signal_pending flag, the + * second one the system call number (as a 'long'), and all further + * arguments being syscall arguments (also 'long'). + * We return a long which is the syscall's return value, which + * may be negative-errno on failure. Conversion to the + * -1-and-errno-set convention is done by the calling wrapper. + */ +safe_syscall_base: + .cfi_startproc + /* + * The syscall calling convention is nearly the same as C: + * we enter with a0 == *signal_pending + * a1 == syscall number + * a2 ... a7 == syscall arguments + * and return the result in a0 + * and the syscall instruction needs + * a7 == syscall number + * a0 ... a5 == syscall arguments + * and returns the result in a0 + * Shuffle everything around appropriately. + */ + mv t0, a0 /* signal_pending pointer */ + mv t1, a1 /* syscall number */ + mv a0, a2 /* syscall arguments */ + mv a1, a3 + mv a2, a4 + mv a3, a5 + mv a4, a6 + mv a5, a7 + mv a7, t1 + + /* + * This next sequence of code works in conjunction with the + * rewind_if_safe_syscall_function(). If a signal is taken + * and the interrupted PC is anywhere between 'safe_syscall_start' + * and 'safe_syscall_end' then we rewind it to 'safe_syscall_start'. + * The code sequence must therefore be able to cope with this, and + * the syscall instruction must be the final one in the sequence. + */ +safe_syscall_start: + /* If signal_pending is non-zero, don't do the call */ + lw t1, 0(t0) + bnez t1, 0f + scall +safe_syscall_end: + /* code path for having successfully executed the syscall */ + ret + +0: + /* code path when we didn't execute the syscall */ + li a0, -TARGET_ERESTARTSYS + ret + .cfi_endproc + + .size safe_syscall_base, .-safe_syscall_base
Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- At some point we should make this routine be non-optional for porting to a new host. r~ --- linux-user/host/riscv64/hostdep.h | 23 +++++++ linux-user/host/riscv64/safe-syscall.inc.S | 77 ++++++++++++++++++++++ 2 files changed, 100 insertions(+) create mode 100644 linux-user/host/riscv64/safe-syscall.inc.S -- 2.17.2