| Message ID | 1509993206-26637-2-git-send-email-peter.maydell@linaro.org |
|---|---|
| State | Accepted |
| Headers | show |
| Series | linux-user: fix various SIGSEGV delivery bugs | expand |
Le 06/11/2017 à 19:33, Peter Maydell a écrit : > For s390x, the address passed to a signal handler in the > siginfo_t si_addr field is masked (in the kernel this is done in > do_sigbus() and do_sigsegv() in arch/s390/mm/fault.c). Implement > this architecture-specific oddity in linux-user. > > This is one of the issues described in > https://bugs.launchpad.net/qemu/+bug/1705118 > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> > --- > linux-user/main.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/linux-user/main.c b/linux-user/main.c > index aa02f25..b6dd9ef 100644 > --- a/linux-user/main.c > +++ b/linux-user/main.c > @@ -3238,6 +3238,10 @@ void cpu_loop(CPUAlphaState *env) > #endif /* TARGET_ALPHA */ > > #ifdef TARGET_S390X > + > +/* s390x masks the fault address it reports in si_addr for SIGSEGV and SIGBUS */ > +#define S390X_FAIL_ADDR_MASK -4096LL > + > void cpu_loop(CPUS390XState *env) > { > CPUState *cs = CPU(s390_env_get_cpu(env)); > @@ -3294,7 +3298,7 @@ void cpu_loop(CPUS390XState *env) > sig = TARGET_SIGSEGV; > /* XXX: check env->error_code */ > n = TARGET_SEGV_MAPERR; > - addr = env->__excp_addr; > + addr = env->__excp_addr & S390X_FAIL_ADDR_MASK; > goto do_signal; > case PGM_EXECUTE: > case PGM_SPECIFICATION: > Reviewed-by: Laurent Vivier <laurent@vivier.eu>
On 11/06/2017 03:33 PM, Peter Maydell wrote: > For s390x, the address passed to a signal handler in the > siginfo_t si_addr field is masked (in the kernel this is done in > do_sigbus() and do_sigsegv() in arch/s390/mm/fault.c). Implement > this architecture-specific oddity in linux-user. > > This is one of the issues described in > https://bugs.launchpad.net/qemu/+bug/1705118 > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> > --- > linux-user/main.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/linux-user/main.c b/linux-user/main.c > index aa02f25..b6dd9ef 100644 > --- a/linux-user/main.c > +++ b/linux-user/main.c > @@ -3238,6 +3238,10 @@ void cpu_loop(CPUAlphaState *env) > #endif /* TARGET_ALPHA */ > > #ifdef TARGET_S390X > + > +/* s390x masks the fault address it reports in si_addr for SIGSEGV and SIGBUS */ > +#define S390X_FAIL_ADDR_MASK -4096LL > + > void cpu_loop(CPUS390XState *env) > { > CPUState *cs = CPU(s390_env_get_cpu(env)); > @@ -3294,7 +3298,7 @@ void cpu_loop(CPUS390XState *env) > sig = TARGET_SIGSEGV; > /* XXX: check env->error_code */ > n = TARGET_SEGV_MAPERR; > - addr = env->__excp_addr; > + addr = env->__excp_addr & S390X_FAIL_ADDR_MASK; > goto do_signal; > case PGM_EXECUTE: > case PGM_SPECIFICATION: >
On 11/06/2017 07:33 PM, Peter Maydell wrote: > For s390x, the address passed to a signal handler in the > siginfo_t si_addr field is masked (in the kernel this is done in > do_sigbus() and do_sigsegv() in arch/s390/mm/fault.c). Implement > this architecture-specific oddity in linux-user. > > This is one of the issues described in > https://bugs.launchpad.net/qemu/+bug/1705118 > > Signed-off-by: Peter Maydell <peter.maydell@linaro.org> > --- > linux-user/main.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) Accurate, but really seems like a s390x kernel bug. Reviewed-by: Richard Henderson <richard.henderson@linaro.org> r~
On 8 November 2017 at 21:18, Richard Henderson <richard.henderson@linaro.org> wrote: > On 11/06/2017 07:33 PM, Peter Maydell wrote: >> For s390x, the address passed to a signal handler in the >> siginfo_t si_addr field is masked (in the kernel this is done in >> do_sigbus() and do_sigsegv() in arch/s390/mm/fault.c). Implement >> this architecture-specific oddity in linux-user. >> >> This is one of the issues described in >> https://bugs.launchpad.net/qemu/+bug/1705118 >> >> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> >> --- >> linux-user/main.c | 6 +++++- >> 1 file changed, 5 insertions(+), 1 deletion(-) > > Accurate, but really seems like a s390x kernel bug. The kernel code goes out of its way to do the masking, so if it's a bug presumably it's a "retain back compat with some older bug" thing... thanks -- PMM
diff --git a/linux-user/main.c b/linux-user/main.c index aa02f25..b6dd9ef 100644 --- a/linux-user/main.c +++ b/linux-user/main.c @@ -3238,6 +3238,10 @@ void cpu_loop(CPUAlphaState *env) #endif /* TARGET_ALPHA */ #ifdef TARGET_S390X + +/* s390x masks the fault address it reports in si_addr for SIGSEGV and SIGBUS */ +#define S390X_FAIL_ADDR_MASK -4096LL + void cpu_loop(CPUS390XState *env) { CPUState *cs = CPU(s390_env_get_cpu(env)); @@ -3294,7 +3298,7 @@ void cpu_loop(CPUS390XState *env) sig = TARGET_SIGSEGV; /* XXX: check env->error_code */ n = TARGET_SEGV_MAPERR; - addr = env->__excp_addr; + addr = env->__excp_addr & S390X_FAIL_ADDR_MASK; goto do_signal; case PGM_EXECUTE: case PGM_SPECIFICATION:
For s390x, the address passed to a signal handler in the siginfo_t si_addr field is masked (in the kernel this is done in do_sigbus() and do_sigsegv() in arch/s390/mm/fault.c). Implement this architecture-specific oddity in linux-user. This is one of the issues described in https://bugs.launchpad.net/qemu/+bug/1705118 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) -- 2.7.4