| Message ID | 146840102187.900.13546085522655887765.stgit@bahia.lan |
|---|---|
| State | New |
| Headers | show |
On Wed, 13 Jul 2016 11:10:21 +0200 Greg Kurz <groug@kaod.org> wrote: > Now that properties are set with -global, if we pass a bogus value for > compat mode, a warning is printed and the property is simply ignored. > For example, if the host is POWER8 and we pass compat=POWER7 instead of > compat=power7 the guest will see a POWER8 cpu. > > While here, this patch also changes the error message to be more explicit. > > Signed-off-by: Greg Kurz <groug@kaod.org> > --- Of course this causes qom-set to exit as well... not sure this is what we want. > target-ppc/translate_init.c | 6 +++++- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c > index 8f257fb74aa7..55688db9c9db 100644 > --- a/target-ppc/translate_init.c > +++ b/target-ppc/translate_init.c > @@ -8475,7 +8475,11 @@ static void powerpc_set_compat(Object *obj, Visitor *v, const char *name, > } else if (strcmp(value, "power8") == 0) { > *max_compat = CPU_POWERPC_LOGICAL_2_07; > } else { > - error_setg(errp, "Invalid compatibility mode \"%s\"", value); > + /* This is called from -global, which does not exit on error. Since we > + * don't want to start with a wrong compat mode, we exit now. > + */ > + error_report("PPC does not support compatibility mode \"%s\"", value); > + exit(1); > } > > g_free(value); > >
On Wed, Jul 13, 2016 at 02:34:32PM +0200, Greg Kurz wrote: > On Wed, 13 Jul 2016 11:10:21 +0200 > Greg Kurz <groug@kaod.org> wrote: > > > Now that properties are set with -global, if we pass a bogus value for > > compat mode, a warning is printed and the property is simply ignored. > > For example, if the host is POWER8 and we pass compat=POWER7 instead of > > compat=power7 the guest will see a POWER8 cpu. > > > > While here, this patch also changes the error message to be more explicit. > > > > Signed-off-by: Greg Kurz <groug@kaod.org> > > --- > > Of course this causes qom-set to exit as well... not sure this is > what we want. Yeah, that seems wrong. If I'm reading this right this is obsoleted by your later generic patch adding an &error_fatal to the generic code, yes? > > > target-ppc/translate_init.c | 6 +++++- > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c > > index 8f257fb74aa7..55688db9c9db 100644 > > --- a/target-ppc/translate_init.c > > +++ b/target-ppc/translate_init.c > > @@ -8475,7 +8475,11 @@ static void powerpc_set_compat(Object *obj, Visitor *v, const char *name, > > } else if (strcmp(value, "power8") == 0) { > > *max_compat = CPU_POWERPC_LOGICAL_2_07; > > } else { > > - error_setg(errp, "Invalid compatibility mode \"%s\"", value); > > + /* This is called from -global, which does not exit on error. Since we > > + * don't want to start with a wrong compat mode, we exit now. > > + */ > > + error_report("PPC does not support compatibility mode \"%s\"", value); > > + exit(1); > > } > > > > g_free(value); > > > > >
On Thu, 14 Jul 2016 12:05:42 +1000 David Gibson <david@gibson.dropbear.id.au> wrote: > On Wed, Jul 13, 2016 at 02:34:32PM +0200, Greg Kurz wrote: > > On Wed, 13 Jul 2016 11:10:21 +0200 > > Greg Kurz <groug@kaod.org> wrote: > > > > > Now that properties are set with -global, if we pass a bogus value for > > > compat mode, a warning is printed and the property is simply ignored. > > > For example, if the host is POWER8 and we pass compat=POWER7 instead of > > > compat=power7 the guest will see a POWER8 cpu. > > > > > > While here, this patch also changes the error message to be more explicit. > > > > > > Signed-off-by: Greg Kurz <groug@kaod.org> > > > --- > > > > Of course this causes qom-set to exit as well... not sure this is > > what we want. > > Yeah, that seems wrong. If I'm reading this right this is obsoleted > by your later generic patch adding an &error_fatal to the generic > code, yes? > Yes it is indeed. > > > > > target-ppc/translate_init.c | 6 +++++- > > > 1 file changed, 5 insertions(+), 1 deletion(-) > > > > > > diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c > > > index 8f257fb74aa7..55688db9c9db 100644 > > > --- a/target-ppc/translate_init.c > > > +++ b/target-ppc/translate_init.c > > > @@ -8475,7 +8475,11 @@ static void powerpc_set_compat(Object *obj, Visitor *v, const char *name, > > > } else if (strcmp(value, "power8") == 0) { > > > *max_compat = CPU_POWERPC_LOGICAL_2_07; > > > } else { > > > - error_setg(errp, "Invalid compatibility mode \"%s\"", value); > > > + /* This is called from -global, which does not exit on error. Since we > > > + * don't want to start with a wrong compat mode, we exit now. > > > + */ > > > + error_report("PPC does not support compatibility mode \"%s\"", value); > > > + exit(1); > > > } > > > > > > g_free(value); > > > > > > > > >
diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c index 8f257fb74aa7..55688db9c9db 100644 --- a/target-ppc/translate_init.c +++ b/target-ppc/translate_init.c @@ -8475,7 +8475,11 @@ static void powerpc_set_compat(Object *obj, Visitor *v, const char *name, } else if (strcmp(value, "power8") == 0) { *max_compat = CPU_POWERPC_LOGICAL_2_07; } else { - error_setg(errp, "Invalid compatibility mode \"%s\"", value); + /* This is called from -global, which does not exit on error. Since we + * don't want to start with a wrong compat mode, we exit now. + */ + error_report("PPC does not support compatibility mode \"%s\"", value); + exit(1); } g_free(value);
Now that properties are set with -global, if we pass a bogus value for compat mode, a warning is printed and the property is simply ignored. For example, if the host is POWER8 and we pass compat=POWER7 instead of compat=power7 the guest will see a POWER8 cpu. While here, this patch also changes the error message to be more explicit. Signed-off-by: Greg Kurz <groug@kaod.org> --- target-ppc/translate_init.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)