From: "Mi, Dapeng" <dapeng1.mi@linux.intel.com>
To: "Chen, Zide" <zide.chen@intel.com>,
qemu-devel@nongnu.org, kvm@vger.kernel.org,
Paolo Bonzini <pbonzini@redhat.com>,
Zhao Liu <zhao1.liu@intel.com>, Peter Xu <peterx@redhat.com>,
Fabiano Rosas <farosas@suse.de>
Cc: Xiaoyao Li <xiaoyao.li@intel.com>,
Dongli Zhang <dongli.zhang@oracle.com>
Subject: Re: [PATCH V2 11/11] target/i386: Disable guest PEBS capability when not enabled
Date: Wed, 11 Feb 2026 09:20:44 +0800 [thread overview]
Message-ID: <4e523f23-a84a-4a93-95fe-c1c00442e2c2@linux.intel.com> (raw)
In-Reply-To: <488a2ea7-062d-4561-89ec-f053af234cdd@intel.com>
On 2/11/2026 3:05 AM, Chen, Zide wrote:
>
> On 2/9/2026 11:30 PM, Mi, Dapeng wrote:
>> On 1/29/2026 7:09 AM, Zide Chen wrote:
>>> When PMU is disabled, guest CPUID must not advertise Debug Store
>>> support. Clear both CPUID.01H:EDX[21] (DS) and CPUID.01H:ECX[2]
>>> (DS64) in this case.
>>>
>>> Set IA32_MISC_ENABLE[12] (PEBS_UNAVAILABLE) when Debug Store is not
>>> exposed to the guest.
>>>
>>> Note: Do not infer that PEBS is unsupported from
>>> IA32_PERF_CAPABILITIES[11:8] (PEBS_FMT) being 0. A value of 0 is a
>>> valid PEBS record format on some CPUs.
>>>
>>> Signed-off-by: Zide Chen <zide.chen@intel.com>
>>> ---
>>> V2:
>>> - New patch.
>>>
>>> target/i386/cpu.c | 6 ++++++
>>> target/i386/cpu.h | 1 +
>>> 2 files changed, 7 insertions(+)
>>>
>>> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
>>> index ec6f49916de3..445361ab7a06 100644
>>> --- a/target/i386/cpu.c
>>> +++ b/target/i386/cpu.c
>>> @@ -9180,6 +9180,10 @@ static void x86_cpu_reset_hold(Object *obj, ResetType type)
>>> env->msr_ia32_misc_enable |= MSR_IA32_MISC_ENABLE_MWAIT;
>>> }
>>>
>>> + if (!(env->features[FEAT_1_EDX] & CPUID_DTS)) {
>>> + env->msr_ia32_misc_enable |= MSR_IA32_MISC_ENABLE_PEBS_UNAVAIL;
>>> + }
>>> +
>>> memset(env->dr, 0, sizeof(env->dr));
>>> env->dr[6] = DR6_FIXED_1;
>>> env->dr[7] = DR7_FIXED_1;
>>> @@ -9474,6 +9478,8 @@ void x86_cpu_expand_features(X86CPU *cpu, Error **errp)
>>> env->features[FEAT_1_ECX] &= ~CPUID_EXT_PDCM;
>>> }
>>>
>>> + env->features[FEAT_1_ECX] &= ~CPUID_EXT_DTES64;
>>> + env->features[FEAT_1_EDX] &= ~CPUID_DTS;
>> Strictly speaking, we need to check BTS as well before clearing DS. BTS
>> also depends on DS.
> But BTS is already unconditionally disabled from the guest in patch 1/11.
Yes, but from code logic, it's incomplete. We need to check both PEBS and
BTS are unavailable, and then disable DS.
>
>
>>> env->features[FEAT_7_0_EDX] &= ~CPUID_7_0_EDX_ARCH_LBR;
>>> }
>>>
>>> diff --git a/target/i386/cpu.h b/target/i386/cpu.h
>>> index 5ab107dfa29f..0fecf561173e 100644
>>> --- a/target/i386/cpu.h
>>> +++ b/target/i386/cpu.h
>>> @@ -483,6 +483,7 @@ typedef enum X86Seg {
>>> /* Indicates good rep/movs microcode on some processors: */
>>> #define MSR_IA32_MISC_ENABLE_FASTSTRING (1ULL << 0)
>>> #define MSR_IA32_MISC_ENABLE_BTS_UNAVAIL (1ULL << 11)
>>> +#define MSR_IA32_MISC_ENABLE_PEBS_UNAVAIL (1ULL << 12)
>>> #define MSR_IA32_MISC_ENABLE_MWAIT (1ULL << 18)
>>> #define MSR_IA32_MISC_ENABLE_DEFAULT (MSR_IA32_MISC_ENABLE_FASTSTRING |\
>>> MSR_IA32_MISC_ENABLE_BTS_UNAVAIL)
prev parent reply other threads:[~2026-02-11 1:20 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-28 23:09 [PATCH V2 00/11] target/i386: Misc PMU, PEBS, and MSR fixes and improvements Zide Chen
2026-01-28 23:09 ` [PATCH V2 01/11] target/i386: Disable unsupported BTS for guest Zide Chen
2026-02-10 6:31 ` Mi, Dapeng
2026-02-11 6:14 ` Xiaoyao Li
2026-03-04 18:22 ` Chen, Zide
2026-01-28 23:09 ` [PATCH V2 02/11] target/i386: Don't save/restore PERF_GLOBAL_OVF_CTRL MSR Zide Chen
2026-01-28 23:09 ` [PATCH V2 03/11] target/i386: Gate enable_pmu on kvm_enabled() Zide Chen
2026-01-28 23:09 ` [PATCH V2 04/11] target/i386: Support full-width writes for perf counters Zide Chen
2026-01-28 23:09 ` [PATCH V2 05/11] target/i386: Increase MSR_BUF_SIZE and split KVM_[GET/SET]_MSRS calls Zide Chen
2026-02-10 6:57 ` Mi, Dapeng
2026-02-10 17:23 ` Chen, Zide
2026-01-28 23:09 ` [PATCH V2 06/11] target/i386: Save/Restore DS based PEBS specfic MSRs Zide Chen
2026-01-28 23:09 ` [PATCH V2 07/11] target/i386: Make some PEBS features user-visible Zide Chen
2026-02-10 7:02 ` Mi, Dapeng
2026-01-28 23:09 ` [PATCH V2 08/11] target/i386: Clean up LBR format handling Zide Chen
2026-02-10 7:07 ` Mi, Dapeng
2026-01-28 23:09 ` [PATCH V2 09/11] target/i386: Refactor " Zide Chen
2026-02-10 7:14 ` Mi, Dapeng
2026-01-28 23:09 ` [PATCH V2 10/11] target/i386: Add pebs-fmt CPU option Zide Chen
2026-01-28 23:09 ` [PATCH V2 11/11] target/i386: Disable guest PEBS capability when not enabled Zide Chen
2026-02-10 7:30 ` Mi, Dapeng
2026-02-10 19:05 ` Chen, Zide
2026-02-11 1:20 ` Mi, Dapeng [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4e523f23-a84a-4a93-95fe-c1c00442e2c2@linux.intel.com \
--to=dapeng1.mi@linux.intel.com \
--cc=dongli.zhang@oracle.com \
--cc=farosas@suse.de \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=peterx@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=xiaoyao.li@intel.com \
--cc=zhao1.liu@intel.com \
--cc=zide.chen@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox