From: "Huang, Kai" <kai.huang@intel.com>
To: "kvm@vger.kernel.org" <kvm@vger.kernel.org>,
"pbonzini@redhat.com" <pbonzini@redhat.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 3/3] KVM: VMX: Initialize TDX during KVM module load
Date: Thu, 28 Nov 2024 03:34:24 +0000 [thread overview]
Message-ID: <7109da789f52dac408f1fb9e0440ad418af6e3ab.camel@intel.com> (raw)
In-Reply-To: <20241127201019.136086-4-pbonzini@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 3323 bytes --]
Hi Paolo,
Thanks for doing this!
> +
> +static void __do_tdx_cleanup(void)
> +{
> + /*
> + * Once TDX module is initialized, it cannot be disabled and
> + * re-initialized again w/o runtime update (which isn't
> + * supported by kernel). Only need to remove the cpuhp here.
> + * The TDX host core code tracks TDX status and can handle
> + * 'multiple enabling' scenario.
> + */
> + WARN_ON_ONCE(!tdx_cpuhp_state);
> + cpuhp_remove_state_nocalls(tdx_cpuhp_state);
> + tdx_cpuhp_state = 0;
> +}
As Gao, Chao pointed out, there's bug here since it is also called by
__do_tdx_bringup() which is called with cpus_read_lock() hold. We need the
_cpuslocked() version here.
I pasted the fixup at the bottom of this reply for your reference and please
merge if it is fine to you. The diff is also attached if that's easier.
[...]
> +int __init tdx_bringup(void)
> +{
> + int r;
> +
> + if (!enable_tdx)
> + return 0;
> +
> + if (!kvm_can_support_tdx()) {
> + pr_err("tdx: no TDX private KeyIDs available\n");
> + goto success_disable_tdx;
> + }
> +
> + if (!enable_virt_at_load) {
> + pr_err("tdx: tdx requires kvm.enable_virt_at_load=1\n");
> + goto success_disable_tdx;
> + }
The intention of kvm_can_support_tdx() is to put all dependency checks to it. I
think we should just put the check of 'enable_virt_at_load' inside. And there
will be more checks in the later patches such as checking 'tdp_mmu_enabled' and
'enable_mmio_caching' so it doesn't make sense to check 'enable_virt_at_load'
outside.
diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
index 0666dfbe0bc0..d8c008437e79 100644
--- a/arch/x86/kvm/vmx/tdx.c
+++ b/arch/x86/kvm/vmx/tdx.c
@@ -38,10 +38,17 @@ static void __do_tdx_cleanup(void)
* 'multiple enabling' scenario.
*/
WARN_ON_ONCE(!tdx_cpuhp_state);
- cpuhp_remove_state_nocalls(tdx_cpuhp_state);
+ cpuhp_remove_state_nocalls_cpuslocked(tdx_cpuhp_state);
tdx_cpuhp_state = 0;
}
+static void __tdx_cleanup(void)
+{
+ cpus_read_lock();
+ __do_tdx_cleanup();
+ cpus_read_unlock();
+}
+
static int __init __do_tdx_bringup(void)
{
int r;
@@ -68,7 +75,17 @@ static int __init __do_tdx_bringup(void)
static bool __init kvm_can_support_tdx(void)
{
- return cpu_feature_enabled(X86_FEATURE_TDX_HOST_PLATFORM);
+ if (!cpu_feature_enabled(X86_FEATURE_TDX_HOST_PLATFORM)) {
+ pr_err("tdx: no TDX private KeyIDs available\n");
+ return false;
+ }
+
+ if (!enable_virt_at_load) {
+ pr_err("tdx: tdx requires kvm.enable_virt_at_load=1\n");
+ return false;
+ }
+
+ return true;
}
static int __init __tdx_bringup(void)
@@ -103,7 +120,7 @@ static int __init __tdx_bringup(void)
void tdx_cleanup(void)
{
if (enable_tdx) {
- __do_tdx_cleanup();
+ __tdx_cleanup();
kvm_disable_virtualization();
}
}
@@ -115,15 +132,8 @@ int __init tdx_bringup(void)
if (!enable_tdx)
return 0;
- if (!kvm_can_support_tdx()) {
- pr_err("tdx: no TDX private KeyIDs available\n");
- goto success_disable_tdx;
- }
-
- if (!enable_virt_at_load) {
- pr_err("tdx: tdx requires kvm.enable_virt_at_load=1\n");
+ if (!kvm_can_support_tdx())
goto success_disable_tdx;
- }
/*
* Ideally KVM should probe whether TDX module has been loaded
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: tdxinit-fix.diff --]
[-- Type: text/x-patch; name="tdxinit-fix.diff", Size: 1679 bytes --]
diff --git a/arch/x86/kvm/vmx/tdx.c b/arch/x86/kvm/vmx/tdx.c
index 0666dfbe0bc0..d8c008437e79 100644
--- a/arch/x86/kvm/vmx/tdx.c
+++ b/arch/x86/kvm/vmx/tdx.c
@@ -38,10 +38,17 @@ static void __do_tdx_cleanup(void)
* 'multiple enabling' scenario.
*/
WARN_ON_ONCE(!tdx_cpuhp_state);
- cpuhp_remove_state_nocalls(tdx_cpuhp_state);
+ cpuhp_remove_state_nocalls_cpuslocked(tdx_cpuhp_state);
tdx_cpuhp_state = 0;
}
+static void __tdx_cleanup(void)
+{
+ cpus_read_lock();
+ __do_tdx_cleanup();
+ cpus_read_unlock();
+}
+
static int __init __do_tdx_bringup(void)
{
int r;
@@ -68,7 +75,17 @@ static int __init __do_tdx_bringup(void)
static bool __init kvm_can_support_tdx(void)
{
- return cpu_feature_enabled(X86_FEATURE_TDX_HOST_PLATFORM);
+ if (!cpu_feature_enabled(X86_FEATURE_TDX_HOST_PLATFORM)) {
+ pr_err("tdx: no TDX private KeyIDs available\n");
+ return false;
+ }
+
+ if (!enable_virt_at_load) {
+ pr_err("tdx: tdx requires kvm.enable_virt_at_load=1\n");
+ return false;
+ }
+
+ return true;
}
static int __init __tdx_bringup(void)
@@ -103,7 +120,7 @@ static int __init __tdx_bringup(void)
void tdx_cleanup(void)
{
if (enable_tdx) {
- __do_tdx_cleanup();
+ __tdx_cleanup();
kvm_disable_virtualization();
}
}
@@ -115,15 +132,8 @@ int __init tdx_bringup(void)
if (!enable_tdx)
return 0;
- if (!kvm_can_support_tdx()) {
- pr_err("tdx: no TDX private KeyIDs available\n");
- goto success_disable_tdx;
- }
-
- if (!enable_virt_at_load) {
- pr_err("tdx: tdx requires kvm.enable_virt_at_load=1\n");
+ if (!kvm_can_support_tdx())
goto success_disable_tdx;
- }
/*
* Ideally KVM should probe whether TDX module has been loaded
next prev parent reply other threads:[~2024-11-28 3:34 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-11-27 20:10 [PATCH v3 0/3] " Paolo Bonzini
2024-11-27 20:10 ` [PATCH 1/3] KVM: Export hardware virtualization enabling/disabling functions Paolo Bonzini
2024-11-27 20:10 ` [PATCH 2/3] KVM: VMX: Refactor VMX module init/exit functions Paolo Bonzini
2024-11-27 20:10 ` [PATCH 3/3] KVM: VMX: Initialize TDX during KVM module load Paolo Bonzini
2024-11-28 3:00 ` Chao Gao
2024-11-28 3:04 ` Huang, Kai
2024-11-28 3:34 ` Huang, Kai [this message]
-- strict thread matches above, loose matches on Subject: below --
2024-10-28 13:20 [PATCH 0/3] KVM: VMX: Initialize TDX when loading KVM module Kai Huang
2024-10-28 13:20 ` [PATCH 3/3] KVM: VMX: Initialize TDX during KVM module load Kai Huang
2024-10-30 15:19 ` Sean Christopherson
2024-10-31 11:17 ` Huang, Kai
2024-10-31 20:22 ` Sean Christopherson
2024-10-31 21:21 ` Huang, Kai
2024-10-31 21:29 ` Edgecombe, Rick P
2024-11-06 14:19 ` Edgecombe, Rick P
2024-11-06 10:49 ` Huang, Kai
2024-11-06 15:01 ` Sean Christopherson
2024-11-06 20:06 ` Huang, Kai
2024-11-07 22:04 ` Sean Christopherson
2024-11-07 23:25 ` Huang, Kai
2024-10-31 21:52 ` Dan Williams
2024-10-31 22:37 ` Huang, Kai
2024-10-31 22:56 ` Dan Williams
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=7109da789f52dac408f1fb9e0440ad418af6e3ab.camel@intel.com \
--to=kai.huang@intel.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=pbonzini@redhat.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
all inboxes | Powered by JetHome®