| Message ID | 1485868319-16151-2-git-send-email-thuth@redhat.com |
|---|---|
| State | New |
| Headers | show |
On Tue, Jan 31, 2017 at 02:11:58PM +0100, Thomas Huth wrote: > When running with KVM on POWER, we are registering a "family" CPU > type for the host CPU that we are running on. For example, on all > POWER8-compatible hosts, we register a "POWER8" CPU type, so that > you can always start QEMU with "-cpu POWER8" there, without the > need to know whether you are running on a POWER8, POWER8E or POWER8NVL > host machine. > However, we also have a "POWER8" CPU alias in the ppc_cpu_aliases list > (that is mainly useful for TCG). This leads to two cosmetical drawbacks: > If the user runs QEMU with "-cpu ?", we always claim that POWER8 is an > "alias for POWER8_v2.0" - which is simply not true when running with > KVM on POWER. And when using the 'query-cpu-definitions' QMP call, > there are currently two entries for "POWER8", one for the alias, and > one for the additional registered type. > To solve the two problems, we should rather update the "family" alias > instead of registering a new types. We then only have one "POWER8" > CPU definition around, an alias, which also points to the right > destination. > > Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1396536 > Signed-off-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Updating the otherwise static table of aliases is kind of ugly, but then so is registering an extra full type as we do now. Is this safe to apply without the follow up patch to vl.c. > --- > target/ppc/kvm.c | 36 +++++++++++++++++++++++------------- > 1 file changed, 23 insertions(+), 13 deletions(-) > > diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c > index ec92c64..f58c260 100644 > --- a/target/ppc/kvm.c > +++ b/target/ppc/kvm.c > @@ -24,6 +24,7 @@ > #include "qemu-common.h" > #include "qemu/error-report.h" > #include "cpu.h" > +#include "cpu-models.h" > #include "qemu/timer.h" > #include "sysemu/sysemu.h" > #include "sysemu/hw_accel.h" > @@ -2412,6 +2413,7 @@ static int kvm_ppc_register_host_cpu_type(void) > }; > PowerPCCPUClass *pvr_pcc; > DeviceClass *dc; > + int i; > > pvr_pcc = kvm_ppc_get_host_cpu_class(); > if (pvr_pcc == NULL) { > @@ -2420,13 +2422,6 @@ static int kvm_ppc_register_host_cpu_type(void) > type_info.parent = object_class_get_name(OBJECT_CLASS(pvr_pcc)); > type_register(&type_info); > > - /* Register generic family CPU class for a family */ > - pvr_pcc = ppc_cpu_get_family_class(pvr_pcc); > - dc = DEVICE_CLASS(pvr_pcc); > - type_info.parent = object_class_get_name(OBJECT_CLASS(pvr_pcc)); > - type_info.name = g_strdup_printf("%s-"TYPE_POWERPC_CPU, dc->desc); > - type_register(&type_info); > - > #if defined(TARGET_PPC64) > type_info.name = g_strdup_printf("%s-"TYPE_SPAPR_CPU_CORE, "host"); > type_info.parent = TYPE_SPAPR_CPU_CORE, > @@ -2436,14 +2431,29 @@ static int kvm_ppc_register_host_cpu_type(void) > type_info.class_data = (void *) "host"; > type_register(&type_info); > g_free((void *)type_info.name); > - > - /* Register generic spapr CPU family class for current host CPU type */ > - type_info.name = g_strdup_printf("%s-"TYPE_SPAPR_CPU_CORE, dc->desc); > - type_info.class_data = (void *) dc->desc; > - type_register(&type_info); > - g_free((void *)type_info.name); > #endif > > + /* > + * Update generic CPU family class alias (e.g. on a POWER8NVL host, > + * we want "POWER8" to be a "family" alias that points to the current > + * host CPU type, too) > + */ > + dc = DEVICE_CLASS(ppc_cpu_get_family_class(pvr_pcc)); > + for (i = 0; ppc_cpu_aliases[i].alias != NULL; i++) { > + if (strcmp(ppc_cpu_aliases[i].alias, dc->desc) == 0) { > + ObjectClass *oc = OBJECT_CLASS(pvr_pcc); > + char *suffix; > + > + ppc_cpu_aliases[i].model = g_strdup(object_class_get_name(oc)); > + suffix = strstr(ppc_cpu_aliases[i].model, "-"TYPE_POWERPC_CPU); > + if (suffix) { > + *suffix = 0; > + } > + ppc_cpu_aliases[i].oc = oc; > + break; > + } > + } > + > return 0; > } >
On 01.02.2017 01:10, David Gibson wrote: > On Tue, Jan 31, 2017 at 02:11:58PM +0100, Thomas Huth wrote: >> When running with KVM on POWER, we are registering a "family" CPU >> type for the host CPU that we are running on. For example, on all >> POWER8-compatible hosts, we register a "POWER8" CPU type, so that >> you can always start QEMU with "-cpu POWER8" there, without the >> need to know whether you are running on a POWER8, POWER8E or POWER8NVL >> host machine. >> However, we also have a "POWER8" CPU alias in the ppc_cpu_aliases list >> (that is mainly useful for TCG). This leads to two cosmetical drawbacks: >> If the user runs QEMU with "-cpu ?", we always claim that POWER8 is an >> "alias for POWER8_v2.0" - which is simply not true when running with >> KVM on POWER. And when using the 'query-cpu-definitions' QMP call, >> there are currently two entries for "POWER8", one for the alias, and >> one for the additional registered type. >> To solve the two problems, we should rather update the "family" alias >> instead of registering a new types. We then only have one "POWER8" >> CPU definition around, an alias, which also points to the right >> destination. >> >> Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1396536 >> Signed-off-by: Thomas Huth <thuth@redhat.com> > > Reviewed-by: David Gibson <david@gibson.dropbear.id.au> > > Updating the otherwise static table of aliases is kind of ugly, but > then so is registering an extra full type as we do now. > > Is this safe to apply without the follow up patch to vl.c. Yes. It fixes the problem with "query-cpu-definitions" already. You just need the other patch to get the output of "-cpu ?" right, too. Thomas
On Wed, Feb 01, 2017 at 08:39:47AM +0100, Thomas Huth wrote: > On 01.02.2017 01:10, David Gibson wrote: > > On Tue, Jan 31, 2017 at 02:11:58PM +0100, Thomas Huth wrote: > >> When running with KVM on POWER, we are registering a "family" CPU > >> type for the host CPU that we are running on. For example, on all > >> POWER8-compatible hosts, we register a "POWER8" CPU type, so that > >> you can always start QEMU with "-cpu POWER8" there, without the > >> need to know whether you are running on a POWER8, POWER8E or POWER8NVL > >> host machine. > >> However, we also have a "POWER8" CPU alias in the ppc_cpu_aliases list > >> (that is mainly useful for TCG). This leads to two cosmetical drawbacks: > >> If the user runs QEMU with "-cpu ?", we always claim that POWER8 is an > >> "alias for POWER8_v2.0" - which is simply not true when running with > >> KVM on POWER. And when using the 'query-cpu-definitions' QMP call, > >> there are currently two entries for "POWER8", one for the alias, and > >> one for the additional registered type. > >> To solve the two problems, we should rather update the "family" alias > >> instead of registering a new types. We then only have one "POWER8" > >> CPU definition around, an alias, which also points to the right > >> destination. > >> > >> Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1396536 > >> Signed-off-by: Thomas Huth <thuth@redhat.com> > > > > Reviewed-by: David Gibson <david@gibson.dropbear.id.au> > > > > Updating the otherwise static table of aliases is kind of ugly, but > > then so is registering an extra full type as we do now. > > > > Is this safe to apply without the follow up patch to vl.c. > > Yes. It fixes the problem with "query-cpu-definitions" already. You just > need the other patch to get the output of "-cpu ?" right, too. Great. Applied to ppc-for-2.9.
diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c index ec92c64..f58c260 100644 --- a/target/ppc/kvm.c +++ b/target/ppc/kvm.c @@ -24,6 +24,7 @@ #include "qemu-common.h" #include "qemu/error-report.h" #include "cpu.h" +#include "cpu-models.h" #include "qemu/timer.h" #include "sysemu/sysemu.h" #include "sysemu/hw_accel.h" @@ -2412,6 +2413,7 @@ static int kvm_ppc_register_host_cpu_type(void) }; PowerPCCPUClass *pvr_pcc; DeviceClass *dc; + int i; pvr_pcc = kvm_ppc_get_host_cpu_class(); if (pvr_pcc == NULL) { @@ -2420,13 +2422,6 @@ static int kvm_ppc_register_host_cpu_type(void) type_info.parent = object_class_get_name(OBJECT_CLASS(pvr_pcc)); type_register(&type_info); - /* Register generic family CPU class for a family */ - pvr_pcc = ppc_cpu_get_family_class(pvr_pcc); - dc = DEVICE_CLASS(pvr_pcc); - type_info.parent = object_class_get_name(OBJECT_CLASS(pvr_pcc)); - type_info.name = g_strdup_printf("%s-"TYPE_POWERPC_CPU, dc->desc); - type_register(&type_info); - #if defined(TARGET_PPC64) type_info.name = g_strdup_printf("%s-"TYPE_SPAPR_CPU_CORE, "host"); type_info.parent = TYPE_SPAPR_CPU_CORE, @@ -2436,14 +2431,29 @@ static int kvm_ppc_register_host_cpu_type(void) type_info.class_data = (void *) "host"; type_register(&type_info); g_free((void *)type_info.name); - - /* Register generic spapr CPU family class for current host CPU type */ - type_info.name = g_strdup_printf("%s-"TYPE_SPAPR_CPU_CORE, dc->desc); - type_info.class_data = (void *) dc->desc; - type_register(&type_info); - g_free((void *)type_info.name); #endif + /* + * Update generic CPU family class alias (e.g. on a POWER8NVL host, + * we want "POWER8" to be a "family" alias that points to the current + * host CPU type, too) + */ + dc = DEVICE_CLASS(ppc_cpu_get_family_class(pvr_pcc)); + for (i = 0; ppc_cpu_aliases[i].alias != NULL; i++) { + if (strcmp(ppc_cpu_aliases[i].alias, dc->desc) == 0) { + ObjectClass *oc = OBJECT_CLASS(pvr_pcc); + char *suffix; + + ppc_cpu_aliases[i].model = g_strdup(object_class_get_name(oc)); + suffix = strstr(ppc_cpu_aliases[i].model, "-"TYPE_POWERPC_CPU); + if (suffix) { + *suffix = 0; + } + ppc_cpu_aliases[i].oc = oc; + break; + } + } + return 0; }
When running with KVM on POWER, we are registering a "family" CPU type for the host CPU that we are running on. For example, on all POWER8-compatible hosts, we register a "POWER8" CPU type, so that you can always start QEMU with "-cpu POWER8" there, without the need to know whether you are running on a POWER8, POWER8E or POWER8NVL host machine. However, we also have a "POWER8" CPU alias in the ppc_cpu_aliases list (that is mainly useful for TCG). This leads to two cosmetical drawbacks: If the user runs QEMU with "-cpu ?", we always claim that POWER8 is an "alias for POWER8_v2.0" - which is simply not true when running with KVM on POWER. And when using the 'query-cpu-definitions' QMP call, there are currently two entries for "POWER8", one for the alias, and one for the additional registered type. To solve the two problems, we should rather update the "family" alias instead of registering a new types. We then only have one "POWER8" CPU definition around, an alias, which also points to the right destination. Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1396536 Signed-off-by: Thomas Huth <thuth@redhat.com> --- target/ppc/kvm.c | 36 +++++++++++++++++++++++------------- 1 file changed, 23 insertions(+), 13 deletions(-)