[PATCH] perf/x86/intel/pt: Test CPU vendor before loading

[PATCH] perf/x86/intel/pt: Test CPU vendor before loading

[Borislav Petkov]

From: Borislav Petkov <bp@suse.de>

This keeps poking at MSR 0x570 when loading, i.e. MSR_IA32_RTIT_CTL,
which is most likely not present on other vendors. It is using the _safe
variant but there's no need to try it at all, really. Besides, it causes
unhandled rdmsr warnings when booting in a guest:

 kvm [4414]: vcpu0 unhandled rdmsr: 0x570

So check CPU vendor before doing anything else.

Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: "Peter Zijlstra (Intel)" <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
---
 arch/x86/kernel/cpu/perf_event_intel_pt.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/arch/x86/kernel/cpu/perf_event_intel_pt.c b/arch/x86/kernel/cpu/perf_event_intel_pt.c
index 42169283448b..fa1262b025e7 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_pt.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_pt.c
@@ -1129,6 +1129,9 @@ static __init int pt_init(void)
 {
 	int ret, cpu, prior_warn = 0;
 
+	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
+		return 0;
+
 	BUILD_BUG_ON(sizeof(struct topa) > PAGE_SIZE);
 	get_online_cpus();
 	for_each_online_cpu(cpu) {
-- 
2.3.5

Re: [PATCH] perf/x86/intel/pt: Test CPU vendor before loading

[Peter Zijlstra]

On Thu, Nov 12, 2015 at 11:56:12AM +0100, Borislav Petkov wrote:
> diff --git a/arch/x86/kernel/cpu/perf_event_intel_pt.c b/arch/x86/kernel/cpu/perf_event_intel_pt.c
> index 42169283448b..fa1262b025e7 100644
> --- a/arch/x86/kernel/cpu/perf_event_intel_pt.c
> +++ b/arch/x86/kernel/cpu/perf_event_intel_pt.c
> @@ -1129,6 +1129,9 @@ static __init int pt_init(void)
>  {
>  	int ret, cpu, prior_warn = 0;
>  
> +	if (boot_cpu_data.x86_vendor != X86_VENDOR_INTEL)
> +		return 0;
> +

Does not apply; conflicts with:


commit 73fdeb66592ee80dffb16fb8a9b7378a00c1a826
Author: Huaitong Han <huaitong.han@intel.com>
Date:   Mon Aug 31 16:21:02 2015 +0800

    perf/x86/intel/pt: Fix KVM warning due to doing rdmsr() before the CPUID test
    
    If KVM does not support INTEL_PT, guest MSR_IA32_RTIT_CTL reading will
    produce host warning like "kvm [2469]: vcpu0 unhandled rdmsr: 0x570".
    
    Guest can determine whether the CPU supports Intel_PT according to CPUID,
    so test_cpu_cap function is added before rdmsr,and it is more in line with
    the code style.
    
    Signed-off-by: Huaitong Han <huaitong.han@intel.com>
    Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
    Reviewed-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
    Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
    Cc: Jiri Olsa <jolsa@redhat.com>
    Cc: Linus Torvalds <torvalds@linux-foundation.org>
    Cc: Peter Zijlstra <peterz@infradead.org>
    Cc: Stephane Eranian <eranian@google.com>
    Cc: Thomas Gleixner <tglx@linutronix.de>
    Cc: Vince Weaver <vincent.weaver@maine.edu>
    Cc: acme@kernel.org
    Link: http://lkml.kernel.org/r/1441009262-9792-1-git-send-email-huaitong.han@intel.com
    Signed-off-by: Ingo Molnar <mingo@kernel.org>

diff --git a/arch/x86/kernel/cpu/perf_event_intel_pt.c b/arch/x86/kernel/cpu/perf_event_intel_pt.c
index 42169283448b..868e1194337f 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_pt.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_pt.c
@@ -139,9 +139,6 @@ static int __init pt_pmu_hw_init(void)
 	long i;
 
 	attrs = NULL;
-	ret = -ENODEV;
-	if (!test_cpu_cap(&boot_cpu_data, X86_FEATURE_INTEL_PT))
-		goto fail;
 
 	for (i = 0; i < PT_CPUID_LEAVES; i++) {
 		cpuid_count(20, i,
@@ -1130,6 +1127,10 @@ static __init int pt_init(void)
 	int ret, cpu, prior_warn = 0;
 
 	BUILD_BUG_ON(sizeof(struct topa) > PAGE_SIZE);
+
+	if (!test_cpu_cap(&boot_cpu_data, X86_FEATURE_INTEL_PT))
+		return -ENODEV;
+
 	get_online_cpus();
 	for_each_online_cpu(cpu) {
 		u64 ctl;