| Message ID | 20210423105408.7265-1-roger.pau@citrix.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | x86/cpuid: do not expand max leaves on restore | expand |
On 23.04.2021 12:54, Roger Pau Monne wrote: > When restoring limit the maximum leaves to the ones supported by Xen > 4.13 in order to not expand the maximum leaf a guests sees. Note this > is unlikely to cause real issues. Why 4.13 (and not 4.12) here when ... > Guests restored from Xen versions 4.13 or greater will contain CPUID > data on the stream that will override the values set by > xc_cpuid_apply_policy. ... 4.13 already communicates the values? > --- a/tools/libs/guest/xg_cpuid_x86.c > +++ b/tools/libs/guest/xg_cpuid_x86.c > @@ -510,6 +510,11 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, > { > p->feat.mpx = test_bit(X86_FEATURE_MPX, host_featureset); > } > + > + /* Clamp maximum leaves to the supported ones on 4.13. */ Same aspect here then. (While not being a native speaker, it would still seem to me that flipping "supported" and "ones" would read slightly more clearly.) > + p->basic.max_leaf = min(p->basic.max_leaf, 0xdu); > + p->feat.max_subleaf = min(p->feat.max_subleaf, 1u); > + p->extd.max_leaf = min(p->extd.max_leaf, 0x1cu); > } With this I think the comment ahead of the enclosing if() wants either amending or moving immediately inside the if()'s body. Jan
On Fri, Apr 23, 2021 at 02:41:32PM +0200, Jan Beulich wrote: > On 23.04.2021 12:54, Roger Pau Monne wrote: > > When restoring limit the maximum leaves to the ones supported by Xen > > 4.13 in order to not expand the maximum leaf a guests sees. Note this > > is unlikely to cause real issues. > > Why 4.13 (and not 4.12) here when ... > > > Guests restored from Xen versions 4.13 or greater will contain CPUID > > data on the stream that will override the values set by > > xc_cpuid_apply_policy. > > ... 4.13 already communicates the values? Urg, yes, my bad. This should have been limited to the max leaves supported by 4.12, not 4.13. Let me send a new version also incorporating the feedback from below. Thanks, Roger.
diff --git a/tools/libs/guest/xg_cpuid_x86.c b/tools/libs/guest/xg_cpuid_x86.c index 5ea69ad3d51..9296fdc34bd 100644 --- a/tools/libs/guest/xg_cpuid_x86.c +++ b/tools/libs/guest/xg_cpuid_x86.c @@ -510,6 +510,11 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, { p->feat.mpx = test_bit(X86_FEATURE_MPX, host_featureset); } + + /* Clamp maximum leaves to the supported ones on 4.13. */ + p->basic.max_leaf = min(p->basic.max_leaf, 0xdu); + p->feat.max_subleaf = min(p->feat.max_subleaf, 1u); + p->extd.max_leaf = min(p->extd.max_leaf, 0x1cu); } if ( featureset )
When restoring limit the maximum leaves to the ones supported by Xen 4.13 in order to not expand the maximum leaf a guests sees. Note this is unlikely to cause real issues. Guests restored from Xen versions 4.13 or greater will contain CPUID data on the stream that will override the values set by xc_cpuid_apply_policy. Reported-by: Andrew Cooper <andrew.cooper3@citrix.com> Signed-off-by: Roger Pau Monné <roger.pau@citrix.com> --- tools/libs/guest/xg_cpuid_x86.c | 5 +++++ 1 file changed, 5 insertions(+)