From: Yang Weijiang <weijiang.yang@intel.com>
To: seanjc@google.com, pbonzini@redhat.com, dave.hansen@intel.com,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org
Cc: peterz@infradead.org, chao.gao@intel.com,
rick.p.edgecombe@intel.com, mlevitsk@redhat.com,
john.allen@amd.com, weijiang.yang@intel.com
Subject: [PATCH v7 06/26] x86/fpu/xstate: Create guest fpstate with guest specific config
Date: Fri, 24 Nov 2023 00:53:10 -0500 [thread overview]
Message-ID: <20231124055330.138870-7-weijiang.yang@intel.com> (raw)
In-Reply-To: <20231124055330.138870-1-weijiang.yang@intel.com>
Use fpu_guest_cfg to calculate guest fpstate settings, open code for
__fpstate_reset() to avoid using kernel FPU config.
Below configuration steps are currently enforced to get guest fpstate:
1) Kernel sets up guest FPU settings in fpu__init_system_xstate().
2) User space sets vCPU thread group xstate permits via arch_prctl().
3) User space creates guest fpstate via __fpu_alloc_init_guest_fpstate()
for vcpu thread.
4) User space enables guest dynamic xfeatures and re-allocate guest
fpstate.
By adding kernel dynamic xfeatures in above #1 and #2, guest xstate area
size is expanded to hold (fpu_kernel_cfg.default_features | kernel dynamic
xfeatures | user dynamic xfeatures), then host xsaves/xrstors can operate
for all guest xfeatures.
The user_* fields remain unchanged for compatibility with KVM uAPIs.
Signed-off-by: Yang Weijiang <weijiang.yang@intel.com>
---
arch/x86/kernel/fpu/core.c | 48 ++++++++++++++++++++++++++++--------
arch/x86/kernel/fpu/xstate.c | 2 +-
arch/x86/kernel/fpu/xstate.h | 1 +
3 files changed, 40 insertions(+), 11 deletions(-)
diff --git a/arch/x86/kernel/fpu/core.c b/arch/x86/kernel/fpu/core.c
index 516af626bf6a..985eaf8b55e0 100644
--- a/arch/x86/kernel/fpu/core.c
+++ b/arch/x86/kernel/fpu/core.c
@@ -194,8 +194,6 @@ void fpu_reset_from_exception_fixup(void)
}
#if IS_ENABLED(CONFIG_KVM)
-static void __fpstate_reset(struct fpstate *fpstate, u64 xfd);
-
static void fpu_init_guest_permissions(struct fpu_guest *gfpu)
{
struct fpu_state_perm *fpuperm;
@@ -216,25 +214,55 @@ static void fpu_init_guest_permissions(struct fpu_guest *gfpu)
gfpu->perm = perm & ~FPU_GUEST_PERM_LOCKED;
}
-bool fpu_alloc_guest_fpstate(struct fpu_guest *gfpu)
+static struct fpstate *__fpu_alloc_init_guest_fpstate(struct fpu_guest *gfpu)
{
+ bool compacted = cpu_feature_enabled(X86_FEATURE_XCOMPACTED);
+ unsigned int gfpstate_size, size;
struct fpstate *fpstate;
- unsigned int size;
- size = fpu_user_cfg.default_size + ALIGN(offsetof(struct fpstate, regs), 64);
+ /*
+ * fpu_guest_cfg.default_features includes all enabled xfeatures
+ * except the user dynamic xfeatures. If the user dynamic xfeatures
+ * are enabled, the guest fpstate will be re-allocated to hold all
+ * guest enabled xfeatures, so omit user dynamic xfeatures here.
+ */
+ gfpstate_size = xstate_calculate_size(fpu_guest_cfg.default_features,
+ compacted);
+
+ size = gfpstate_size + ALIGN(offsetof(struct fpstate, regs), 64);
+
fpstate = vzalloc(size);
if (!fpstate)
- return false;
+ return NULL;
+ /*
+ * Initialize sizes and feature masks, use fpu_user_cfg.*
+ * for user_* settings for compatibility of exiting uAPIs.
+ */
+ fpstate->size = gfpstate_size;
+ fpstate->xfeatures = fpu_guest_cfg.default_features;
+ fpstate->user_size = fpu_user_cfg.default_size;
+ fpstate->user_xfeatures = fpu_user_cfg.default_features;
+ fpstate->xfd = 0;
- /* Leave xfd to 0 (the reset value defined by spec) */
- __fpstate_reset(fpstate, 0);
fpstate_init_user(fpstate);
fpstate->is_valloc = true;
fpstate->is_guest = true;
gfpu->fpstate = fpstate;
- gfpu->xfeatures = fpu_user_cfg.default_features;
- gfpu->perm = fpu_user_cfg.default_features;
+ gfpu->xfeatures = fpu_guest_cfg.default_features;
+ gfpu->perm = fpu_guest_cfg.default_features;
+
+ return fpstate;
+}
+
+bool fpu_alloc_guest_fpstate(struct fpu_guest *gfpu)
+{
+ struct fpstate *fpstate;
+
+ fpstate = __fpu_alloc_init_guest_fpstate(gfpu);
+
+ if (!fpstate)
+ return false;
/*
* KVM sets the FP+SSE bits in the XSAVE header when copying FPU state
diff --git a/arch/x86/kernel/fpu/xstate.c b/arch/x86/kernel/fpu/xstate.c
index aa8f8595cd41..253944cb2298 100644
--- a/arch/x86/kernel/fpu/xstate.c
+++ b/arch/x86/kernel/fpu/xstate.c
@@ -559,7 +559,7 @@ static bool __init check_xstate_against_struct(int nr)
return true;
}
-static unsigned int xstate_calculate_size(u64 xfeatures, bool compacted)
+unsigned int xstate_calculate_size(u64 xfeatures, bool compacted)
{
unsigned int topmost = fls64(xfeatures) - 1;
unsigned int offset = xstate_offsets[topmost];
diff --git a/arch/x86/kernel/fpu/xstate.h b/arch/x86/kernel/fpu/xstate.h
index 3518fb26d06b..c032acb56306 100644
--- a/arch/x86/kernel/fpu/xstate.h
+++ b/arch/x86/kernel/fpu/xstate.h
@@ -55,6 +55,7 @@ extern void fpu__init_cpu_xstate(void);
extern void fpu__init_system_xstate(unsigned int legacy_size);
extern void *get_xsave_addr(struct xregs_state *xsave, int xfeature_nr);
+extern unsigned int xstate_calculate_size(u64 xfeatures, bool compacted);
static inline u64 xfeatures_mask_supervisor(void)
{
--
2.27.0
next prev parent reply other threads:[~2023-11-24 7:59 UTC|newest]
Thread overview: 104+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-11-24 5:53 [PATCH v7 00/26] Enable CET Virtualization Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 01/26] x86/fpu/xstate: Always preserve non-user xfeatures/flags in __state_perm Yang Weijiang
2023-11-30 17:24 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 02/26] x86/fpu/xstate: Refine CET user xstate bit enabling Yang Weijiang
2023-11-24 9:40 ` Peter Zijlstra
2023-11-27 2:55 ` Yang, Weijiang
2023-11-28 1:31 ` Edgecombe, Rick P
2023-11-28 8:50 ` Peter Zijlstra
2023-11-28 1:31 ` Edgecombe, Rick P
2023-11-28 7:52 ` Yang, Weijiang
2023-11-30 17:26 ` Maxim Levitsky
2023-12-01 6:51 ` Yang, Weijiang
2023-12-05 9:53 ` Maxim Levitsky
2023-12-06 1:03 ` Yang, Weijiang
2023-12-06 15:57 ` Maxim Levitsky
2023-12-08 14:57 ` Yang, Weijiang
2023-12-08 15:15 ` Maxim Levitsky
2023-12-13 9:30 ` Yang, Weijiang
2023-12-13 13:31 ` Maxim Levitsky
2023-12-13 17:01 ` Chang S. Bae
2023-12-14 3:12 ` Yang, Weijiang
2023-11-24 5:53 ` [PATCH v7 03/26] x86/fpu/xstate: Add CET supervisor mode state support Yang Weijiang
2023-11-24 9:45 ` Peter Zijlstra
2023-11-27 4:06 ` Yang, Weijiang
2023-11-28 1:34 ` Edgecombe, Rick P
2023-11-30 17:27 ` Maxim Levitsky
2023-12-01 7:01 ` Yang, Weijiang
2023-12-05 9:53 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 04/26] x86/fpu/xstate: Introduce XFEATURE_MASK_KERNEL_DYNAMIC xfeature set Yang Weijiang
2023-11-28 1:46 ` Edgecombe, Rick P
2023-11-28 8:00 ` Yang, Weijiang
2023-11-30 17:33 ` Maxim Levitsky
2023-12-01 7:49 ` Yang, Weijiang
2023-12-05 9:55 ` Maxim Levitsky
2023-12-06 3:00 ` Yang, Weijiang
2023-12-06 16:11 ` Maxim Levitsky
2023-12-08 15:57 ` Yang, Weijiang
2023-11-24 5:53 ` [PATCH v7 05/26] x86/fpu/xstate: Introduce fpu_guest_cfg for guest FPU configuration Yang Weijiang
2023-11-28 14:58 ` Edgecombe, Rick P
2023-11-29 14:12 ` Yang, Weijiang
2023-11-29 17:08 ` Edgecombe, Rick P
2023-11-30 13:28 ` Yang, Weijiang
2023-11-30 17:29 ` Maxim Levitsky
2023-11-30 18:02 ` Edgecombe, Rick P
2023-11-30 17:29 ` Maxim Levitsky
2023-11-24 5:53 ` Yang Weijiang [this message]
2023-11-28 15:19 ` [PATCH v7 06/26] x86/fpu/xstate: Create guest fpstate with guest specific config Edgecombe, Rick P
2023-11-29 14:16 ` Yang, Weijiang
2023-11-30 17:36 ` Maxim Levitsky
2023-12-01 8:36 ` Yang, Weijiang
2023-12-05 9:57 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 07/26] x86/fpu/xstate: Warn if kernel dynamic xfeatures detected in normal fpstate Yang Weijiang
2023-11-28 15:25 ` Edgecombe, Rick P
2023-11-29 14:18 ` Yang, Weijiang
2023-11-24 5:53 ` [PATCH v7 08/26] KVM: x86: Rework cpuid_get_supported_xcr0() to operate on vCPU data Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 09/26] KVM: x86: Rename kvm_{g,s}et_msr() to menifest emulation operations Yang Weijiang
2023-11-30 17:36 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 10/26] KVM: x86: Refine xsave-managed guest register/MSR reset handling Yang Weijiang
2023-11-30 17:36 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 11/26] KVM: x86: Add kvm_msr_{read,write}() helpers Yang Weijiang
2023-11-30 17:37 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 12/26] KVM: x86: Report XSS as to-be-saved if there are supported features Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 13/26] KVM: x86: Refresh CPUID on write to guest MSR_IA32_XSS Yang Weijiang
2023-11-30 17:37 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 14/26] KVM: x86: Initialize kvm_caps.supported_xss Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 15/26] KVM: x86: Load guest FPU state when access XSAVE-managed MSRs Yang Weijiang
2023-11-30 17:38 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 16/26] KVM: x86: Add fault checks for guest CR4.CET setting Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 17/26] KVM: x86: Report KVM supported CET MSRs as to-be-saved Yang Weijiang
2023-11-30 17:40 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 18/26] KVM: VMX: Introduce CET VMCS fields and control bits Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 19/26] KVM: x86: Use KVM-governed feature framework to track "SHSTK/IBT enabled" Yang Weijiang
2023-11-30 17:40 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 20/26] KVM: VMX: Emulate read and write to CET MSRs Yang Weijiang
2023-11-30 17:41 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 21/26] KVM: x86: Save and reload SSP to/from SMRAM Yang Weijiang
2023-11-30 17:42 ` Maxim Levitsky
2023-12-01 2:23 ` Chao Gao
2023-12-04 0:45 ` Yang, Weijiang
2023-12-05 10:02 ` Maxim Levitsky
2023-12-01 8:55 ` Yang, Weijiang
2023-11-24 5:53 ` [PATCH v7 22/26] KVM: VMX: Set up interception for CET MSRs Yang Weijiang
2023-11-30 17:44 ` Maxim Levitsky
2023-12-01 6:33 ` Chao Gao
2023-12-05 10:04 ` Maxim Levitsky
2023-12-01 9:45 ` Yang, Weijiang
2023-12-05 10:07 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 23/26] KVM: VMX: Set host constant supervisor states to VMCS fields Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 24/26] KVM: x86: Enable CET virtualization for VMX and advertise to userspace Yang Weijiang
2023-11-30 17:46 ` Maxim Levitsky
2023-12-01 16:15 ` Yang, Weijiang
2023-12-05 10:07 ` Maxim Levitsky
2023-11-24 5:53 ` [PATCH v7 25/26] KVM: nVMX: Introduce new VMX_BASIC bit for event error_code delivery to L1 Yang Weijiang
2023-11-24 5:53 ` [PATCH v7 26/26] KVM: nVMX: Enable CET support for nested guest Yang Weijiang
2023-11-30 17:53 ` Maxim Levitsky
2023-12-04 8:50 ` Yang, Weijiang
2023-12-05 10:12 ` Maxim Levitsky
2023-12-06 9:22 ` Yang, Weijiang
2023-12-06 17:24 ` Maxim Levitsky
2023-12-08 15:15 ` Yang, Weijiang
2023-12-08 15:22 ` Maxim Levitsky
2023-12-12 8:56 ` Yang, Weijiang
2023-12-12 11:09 ` Maxim Levitsky
2023-12-15 2:29 ` [PATCH v7 00/26] Enable CET Virtualization Yang, Weijiang
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=20231124055330.138870-7-weijiang.yang@intel.com \
--to=weijiang.yang@intel.com \
--cc=chao.gao@intel.com \
--cc=dave.hansen@intel.com \
--cc=john.allen@amd.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mlevitsk@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=rick.p.edgecombe@intel.com \
--cc=seanjc@google.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;
as well as URLs for NNTP newsgroup(s).