| Message ID | 20200225173133.3632-1-andrew.cooper3@citrix.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | x86/sysctl: Don't return cpu policy data for compiled-out support | expand |
On 25.02.2020 18:31, Andrew Cooper wrote: > Policy objects aren't tiny, and the derivation logic isn't trivial. We are > about to increase the number of policy objects, so take this opportunity to > drop logic and storage space based on CONFIG_{PV,HVM}. It doesn't look as if you were dropping either logic or storage space. With this aspect taken care of (by adjusting wording, or by clarification of what is meant) ... > Start by causing XEN_SYSCTL_get_cpu_policy to fail with -EOPNOTSUPP when > requesting data for a compiled-out subsystem. Update xen-cpuid to cope and > continue to further system policies, seeing as the indicies are interleaved. > > Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> Reviewed-by: Jan Beulich <jbeulich@suse.com>
On 26/02/2020 09:32, Jan Beulich wrote: > On 25.02.2020 18:31, Andrew Cooper wrote: >> Policy objects aren't tiny, and the derivation logic isn't trivial. We are >> about to increase the number of policy objects, so take this opportunity to >> drop logic and storage space based on CONFIG_{PV,HVM}. > It doesn't look as if you were dropping either logic or storage space. With > this aspect taken care of (by adjusting wording, or by clarification of what > is meant) ... > >> Start by That is this is here to signify. No logic or storage space can be dropped until some #ifdef-ary is sprinkled around cpuid.c and msr.c, but I can't do any of that while there are unguarded external references to the objects. ~Andrew
On 26.02.2020 10:58, Andrew Cooper wrote: > On 26/02/2020 09:32, Jan Beulich wrote: >> On 25.02.2020 18:31, Andrew Cooper wrote: >>> Policy objects aren't tiny, and the derivation logic isn't trivial. We are >>> about to increase the number of policy objects, so take this opportunity to >>> drop logic and storage space based on CONFIG_{PV,HVM}. >> It doesn't look as if you were dropping either logic or storage space. With >> this aspect taken care of (by adjusting wording, or by clarification of what >> is meant) ... >> >>> Start by > > That is this is here to signify. How about s/take this/will have the/ in the sentence further up then? To me, "Start by" in no way indicates that really there aren't any savings yet. Might be my non-native English understanding of this phrase, of course. > No logic or storage space can be dropped until some #ifdef-ary is > sprinkled around cpuid.c and msr.c, but I can't do any of that while > there are unguarded external references to the objects. Of course. Jan
diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 7726c4ed3c..f55b67640a 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -480,7 +480,16 @@ int main(int argc, char **argv) if ( xc_get_system_cpu_policy(xch, i, &nr_leaves, leaves, &nr_msrs, msrs) ) + { + if ( errno == EOPNOTSUPP ) + { + printf("%s policy not supported by Xen\n", + sys_policies[i]); + continue; + } + err(1, "xc_get_system_cpu_policy(, %s,,)", sys_policies[i]); + } print_policy(sys_policies[i], leaves, nr_leaves, msrs, nr_msrs); diff --git a/xen/arch/x86/sysctl.c b/xen/arch/x86/sysctl.c index 4a76f0f47f..59a384023b 100644 --- a/xen/arch/x86/sysctl.c +++ b/xen/arch/x86/sysctl.c @@ -33,7 +33,7 @@ #include <asm/psr.h> #include <asm/cpuid.h> -const struct cpu_policy system_policies[] = { +const struct cpu_policy system_policies[6] = { [ XEN_SYSCTL_cpu_policy_raw ] = { &raw_cpuid_policy, &raw_msr_policy, @@ -42,22 +42,26 @@ const struct cpu_policy system_policies[] = { &host_cpuid_policy, &host_msr_policy, }, +#ifdef CONFIG_PV [ XEN_SYSCTL_cpu_policy_pv_max ] = { &pv_max_cpuid_policy, &pv_max_msr_policy, }, - [ XEN_SYSCTL_cpu_policy_hvm_max ] = { - &hvm_max_cpuid_policy, - &hvm_max_msr_policy, - }, [ XEN_SYSCTL_cpu_policy_pv_default ] = { &pv_max_cpuid_policy, &pv_max_msr_policy, }, +#endif +#ifdef CONFIG_HVM + [ XEN_SYSCTL_cpu_policy_hvm_max ] = { + &hvm_max_cpuid_policy, + &hvm_max_msr_policy, + }, [ XEN_SYSCTL_cpu_policy_hvm_default ] = { &hvm_max_cpuid_policy, &hvm_max_msr_policy, }, +#endif }; struct l3_cache_info { @@ -426,6 +430,12 @@ long arch_do_sysctl( array_index_nospec(sysctl->u.cpu_policy.index, ARRAY_SIZE(system_policies))]; + if ( !policy->cpuid || !policy->msr ) + { + ret = -EOPNOTSUPP; + break; + } + /* Process the CPUID leaves. */ if ( guest_handle_is_null(sysctl->u.cpu_policy.cpuid_policy) ) sysctl->u.cpu_policy.nr_leaves = CPUID_MAX_SERIALISED_LEAVES; diff --git a/xen/include/public/sysctl.h b/xen/include/public/sysctl.h index 7e43bfe1bd..4dfba39ed8 100644 --- a/xen/include/public/sysctl.h +++ b/xen/include/public/sysctl.h @@ -1051,6 +1051,8 @@ struct xen_sysctl_set_parameter { * experimental features outside of security support. * - Default_*: Default set of features a PV or HVM guest can use. This is * the security supported set. + * May fail with -EOPNOTSUPP if querying for PV or HVM data when support is + * compiled out of Xen. */ struct xen_sysctl_cpu_policy { #define XEN_SYSCTL_cpu_policy_raw 0
Policy objects aren't tiny, and the derivation logic isn't trivial. We are about to increase the number of policy objects, so take this opportunity to drop logic and storage space based on CONFIG_{PV,HVM}. Start by causing XEN_SYSCTL_get_cpu_policy to fail with -EOPNOTSUPP when requesting data for a compiled-out subsystem. Update xen-cpuid to cope and continue to further system policies, seeing as the indicies are interleaved. Signed-off-by: Andrew Cooper <andrew.cooper3@citrix.com> --- CC: Jan Beulich <JBeulich@suse.com> CC: Wei Liu <wl@xen.org> CC: Roger Pau Monné <roger.pau@citrix.com> --- tools/misc/xen-cpuid.c | 9 +++++++++ xen/arch/x86/sysctl.c | 20 +++++++++++++++----- xen/include/public/sysctl.h | 2 ++ 3 files changed, 26 insertions(+), 5 deletions(-)