* [PATCH] x86/VPMU: Check more carefully which bits are allowed to be written to MSRs
@ 2015-12-22 16:24 Boris Ostrovsky
2015-12-23 5:21 ` Tian, Kevin
0 siblings, 1 reply; 3+ messages in thread
From: Boris Ostrovsky @ 2015-12-22 16:24 UTC (permalink / raw)
To: jun.nakajima, kevin.tian, keir, jbeulich, andrew.cooper3
Cc: Boris Ostrovsky, dietmar.hahn, xen-devel
Current Intel VPMU emulation needs to perform more checks when writing
PMU MSRs on guest's behalf:
* MSR_CORE_PERF_GLOBAL_CTRL is not checked at all
* MSR_CORE_PERF_FIXED_CTR_CTRL has more reserved bits in PMU version 2
* MSR_CORE_PERF_GLOBAL_OVF_CTRL's bit 61 is allowed on versions greater
* than 2.
We can also use precomputed mask in core2_vpmu_do_interrupt().
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
---
xen/arch/x86/cpu/vpmu_intel.c | 18 ++++++++++++++++--
1 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/xen/arch/x86/cpu/vpmu_intel.c b/xen/arch/x86/cpu/vpmu_intel.c
index 3eff1ae..8dc0b48 100644
--- a/xen/arch/x86/cpu/vpmu_intel.c
+++ b/xen/arch/x86/cpu/vpmu_intel.c
@@ -87,7 +87,7 @@ static unsigned int __read_mostly arch_pmc_cnt, fixed_pmc_cnt;
/* Masks used for testing whether and MSR is valid */
#define ARCH_CTRL_MASK (~((1ull << 32) - 1) | (1ull << 21))
static uint64_t __read_mostly fixed_ctrl_mask, fixed_counters_mask;
-static uint64_t __read_mostly global_ovf_ctrl_mask;
+static uint64_t __read_mostly global_ovf_ctrl_mask, global_ctrl_mask;
/* Total size of PMU registers block (copied to/from PV(H) guest) */
static unsigned int __read_mostly regs_sz;
@@ -392,6 +392,8 @@ static int core2_vpmu_verify(struct vcpu *v)
if ( core2_vpmu_cxt->global_ovf_ctrl & global_ovf_ctrl_mask )
return -EINVAL;
+ if ( core2_vpmu_cxt->global_ctrl & global_ctrl_mask )
+ return -EINVAL;
fixed_ctrl = core2_vpmu_cxt->fixed_ctrl;
if ( fixed_ctrl & fixed_ctrl_mask )
@@ -627,6 +629,8 @@ static int core2_vpmu_do_wrmsr(unsigned int msr, uint64_t msr_content,
gdprintk(XENLOG_WARNING, "Guest setting of DTS is ignored.\n");
return 0;
case MSR_CORE_PERF_GLOBAL_CTRL:
+ if ( msr_content & global_ctrl_mask )
+ return -EINVAL;
core2_vpmu_cxt->global_ctrl = msr_content;
break;
case MSR_CORE_PERF_FIXED_CTR_CTRL:
@@ -820,7 +824,7 @@ static int core2_vpmu_do_interrupt(struct cpu_user_regs *regs)
if ( is_pmc_quirk )
handle_pmc_quirk(msr_content);
core2_vpmu_cxt->global_status |= msr_content;
- msr_content = 0xC000000700000000 | ((1 << arch_pmc_cnt) - 1);
+ msr_content = ~global_ovf_ctrl_mask;
wrmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, msr_content);
}
else
@@ -1001,10 +1005,20 @@ int __init core2_vpmu_init(void)
full_width_write = (caps >> 13) & 1;
fixed_ctrl_mask = ~((1ull << (fixed_pmc_cnt * FIXED_CTR_CTRL_BITS)) - 1);
+ if ( version == 2 )
+ fixed_ctrl_mask |= 0x444;
fixed_counters_mask = ~((1ull << core2_get_bitwidth_fix_count()) - 1);
+ global_ctrl_mask = ~((((1ULL << fixed_pmc_cnt) - 1) << 32) |
+ ((1ULL << arch_pmc_cnt) - 1));
global_ovf_ctrl_mask = ~(0xC000000000000000 |
(((1ULL << fixed_pmc_cnt) - 1) << 32) |
((1ULL << arch_pmc_cnt) - 1));
+ if ( version > 2)
+ /*
+ * Even though we don't support Uncore counters guests should be
+ * able to clear all available overflows.
+ */
+ global_ovf_ctrl_mask &= ~(1ULL << 61);
regs_sz = (sizeof(struct xen_pmu_intel_ctxt) - regs_off) +
sizeof(uint64_t) * fixed_pmc_cnt +
--
1.7.1
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH] x86/VPMU: Check more carefully which bits are allowed to be written to MSRs
2015-12-22 16:24 [PATCH] x86/VPMU: Check more carefully which bits are allowed to be written to MSRs Boris Ostrovsky
@ 2015-12-23 5:21 ` Tian, Kevin
2015-12-23 15:03 ` Boris Ostrovsky
0 siblings, 1 reply; 3+ messages in thread
From: Tian, Kevin @ 2015-12-23 5:21 UTC (permalink / raw)
To: Boris Ostrovsky, Nakajima, Jun, keir@xen.org, jbeulich@suse.com,
andrew.cooper3@citrix.com
Cc: dietmar.hahn@ts.fujitsu.com, xen-devel@lists.xen.org
> From: Boris Ostrovsky [mailto:boris.ostrovsky@oracle.com]
> Sent: Wednesday, December 23, 2015 12:25 AM
>
> Current Intel VPMU emulation needs to perform more checks when writing
> PMU MSRs on guest's behalf:
> * MSR_CORE_PERF_GLOBAL_CTRL is not checked at all
> * MSR_CORE_PERF_FIXED_CTR_CTRL has more reserved bits in PMU version 2
> * MSR_CORE_PERF_GLOBAL_OVF_CTRL's bit 61 is allowed on versions greater
> * than 2.
>
> We can also use precomputed mask in core2_vpmu_do_interrupt().
>
> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Acked-by: Kevin Tian <kevin.tian@intel.com>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH] x86/VPMU: Check more carefully which bits are allowed to be written to MSRs
2015-12-23 5:21 ` Tian, Kevin
@ 2015-12-23 15:03 ` Boris Ostrovsky
0 siblings, 0 replies; 3+ messages in thread
From: Boris Ostrovsky @ 2015-12-23 15:03 UTC (permalink / raw)
To: Tian, Kevin, Nakajima, Jun, keir@xen.org, jbeulich@suse.com,
andrew.cooper3@citrix.com
Cc: dietmar.hahn@ts.fujitsu.com, xen-devel@lists.xen.org
On 12/23/2015 12:21 AM, Tian, Kevin wrote:
>> From: Boris Ostrovsky [mailto:boris.ostrovsky@oracle.com]
>> Sent: Wednesday, December 23, 2015 12:25 AM
>>
>> Current Intel VPMU emulation needs to perform more checks when writing
>> PMU MSRs on guest's behalf:
>> * MSR_CORE_PERF_GLOBAL_CTRL is not checked at all
>> * MSR_CORE_PERF_FIXED_CTR_CTRL has more reserved bits in PMU version 2
>> * MSR_CORE_PERF_GLOBAL_OVF_CTRL's bit 61 is allowed on versions greater
>> * than 2.
>>
>> We can also use precomputed mask in core2_vpmu_do_interrupt().
>>
>> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
> Acked-by: Kevin Tian <kevin.tian@intel.com>
I think I missed one more register. Let me send another version.
-boris
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2015-12-23 15:03 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-12-22 16:24 [PATCH] x86/VPMU: Check more carefully which bits are allowed to be written to MSRs Boris Ostrovsky
2015-12-23 5:21 ` Tian, Kevin
2015-12-23 15:03 ` Boris Ostrovsky
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.