From: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
To: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
x86@kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
David Hildenbrand <david@redhat.com>,
Andrea Arcangeli <aarcange@redhat.com>,
Josh Poimboeuf <jpoimboe@redhat.com>,
Juergen Gross <jgross@suse.com>, Deep Shah <sdeep@vmware.com>,
VMware Inc <pv-drivers@vmware.com>,
Vitaly Kuznetsov <vkuznets@redhat.com>,
Wanpeng Li <wanpengli@tencent.com>,
Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>
Cc: Peter H Anvin <hpa@zytor.com>,
Dave Hansen <dave.hansen@intel.com>,
Tony Luck <tony.luck@intel.com>,
Dan Williams <dan.j.williams@intel.com>,
Andi Kleen <ak@linux.intel.com>,
Kirill Shutemov <kirill.shutemov@linux.intel.com>,
Sean Christopherson <seanjc@google.com>,
Kuppuswamy Sathyanarayanan <knsathya@kernel.org>,
linux-kernel@vger.kernel.org
Subject: [PATCH v8 07/11] x86/tdx: Add HLT support for TDX guest
Date: Mon, 4 Oct 2021 19:52:01 -0700 [thread overview]
Message-ID: <20211005025205.1784480-8-sathyanarayanan.kuppuswamy@linux.intel.com> (raw)
In-Reply-To: <20211005025205.1784480-1-sathyanarayanan.kuppuswamy@linux.intel.com>
From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Per Guest-Host-Communication Interface (GHCI) for Intel Trust
Domain Extensions (Intel TDX) specification, sec 3.8,
TDVMCALL[Instruction.HLT] provides HLT operation. Use it to implement
halt() and safe_halt() paravirtualization calls.
The same TDX hypercall is used to handle #VE exception due to
EXIT_REASON_HLT.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
---
Changes since v7:
* Added section title to spec reference in commit log and comments.
* Added extra comments as per review suggestion.
Changes since v6:
* None
Changes since v5:
* Replaced sti with STI in commit log and comments.
* Added comments for _tdx_hypercall() usage in _tdx_halt().
* Added new helper function _tdx_halt() to contain common
code between tdx_halt() and tdx_safe_halt().
* Renamed tdg_->tdx_.
* Removed BUG_ON() and used WARN_ONCE() for HLT emulation
failure.
Changes since v4:
* Added exception for EXIT_REASON_HLT in __tdx_hypercall() to
enable interrupts using sti.
Changes since v3:
* None
arch/x86/kernel/tdcall.S | 30 ++++++++++++++++
arch/x86/kernel/tdx.c | 75 ++++++++++++++++++++++++++++++++++++++--
2 files changed, 103 insertions(+), 2 deletions(-)
diff --git a/arch/x86/kernel/tdcall.S b/arch/x86/kernel/tdcall.S
index 2e70133bebf2..1b9649ec2e29 100644
--- a/arch/x86/kernel/tdcall.S
+++ b/arch/x86/kernel/tdcall.S
@@ -40,6 +40,9 @@
*/
#define tdcall .byte 0x66,0x0f,0x01,0xcc
+/* HLT TDVMCALL sub-function ID */
+#define EXIT_REASON_HLT 12
+
/*
* __tdx_module_call() - Helper function used by TDX guests to request
* services from the TDX module (does not include VMM services).
@@ -240,6 +243,33 @@ SYM_FUNC_START(__tdx_hypercall)
movl $TDVMCALL_EXPOSE_REGS_MASK, %ecx
+ /*
+ * For the idle loop STI needs to be called directly before
+ * the TDCALL that enters idle (EXIT_REASON_HLT case). STI
+ * enables interrupts only one instruction later. If there
+ * are any instructions between the STI and the TDCALL for
+ * HLT then an interrupt could happen in that time, but the
+ * code would go back to sleep afterwards, which can cause
+ * longer delays.
+ *
+ * This leads to significant difference in network performance
+ * benchmarks. So add a special case for EXIT_REASON_HLT to
+ * trigger STI before TDCALL. But this change is not required
+ * for all HLT cases. So use R15 register value to identify the
+ * case which needs STI. So, if R11 is EXIT_REASON_HLT and R15
+ * is 1, then call STI before TDCALL instruction. Note that R15
+ * register is not required by TDCALL ABI when triggering the
+ * hypercall for EXIT_REASON_HLT case. So use it in software to
+ * select the STI case.
+ */
+ cmpl $EXIT_REASON_HLT, %r11d
+ jne skip_sti
+ cmpl $1, %r15d
+ jne skip_sti
+ /* Set R15 register to 0, it is unused in EXIT_REASON_HLT case */
+ xor %r15, %r15
+ sti
+skip_sti:
tdcall
/* Restore output pointer to R9 */
diff --git a/arch/x86/kernel/tdx.c b/arch/x86/kernel/tdx.c
index f7885c777a09..3d0416515506 100644
--- a/arch/x86/kernel/tdx.c
+++ b/arch/x86/kernel/tdx.c
@@ -5,6 +5,7 @@
#define pr_fmt(fmt) "tdx: " fmt
#include <asm/tdx.h>
+#include <asm/vmx.h>
/* TDX Module call Leaf IDs */
#define TDGETVEINFO 3
@@ -53,6 +54,62 @@ static inline u64 _tdx_hypercall(u64 fn, u64 r12, u64 r13, u64 r14,
return out->r10;
}
+static __cpuidle void _tdx_halt(const bool irq_disabled, const bool do_sti)
+{
+ u64 ret;
+
+ /*
+ * Emulate HLT operation via hypercall. More info about ABI
+ * can be found in TDX Guest-Host-Communication Interface
+ * (GHCI), sec 3.8 TDG.VP.VMCALL<Instruction.HLT>.
+ *
+ * The VMM uses the "IRQ disabled" param to understand IRQ
+ * enabled status (RFLAGS.IF) of TD guest and determine
+ * whether or not it should schedule the halted vCPU if an
+ * IRQ becomes pending. E.g. if IRQs are disabled the VMM
+ * can keep the vCPU in virtual HLT, even if an IRQ is
+ * pending, without hanging/breaking the guest.
+ *
+ * do_sti parameter is used by __tdx_hypercall() to decide
+ * whether to call STI instruction before executing TDCALL
+ * instruction.
+ */
+ ret = _tdx_hypercall(EXIT_REASON_HLT, irq_disabled, 0, 0, do_sti, NULL);
+
+ /*
+ * Use WARN_ONCE() to report the failure. Since tdx_*halt() calls
+ * are also used in pv_ops, #VE error handler cannot be used to
+ * report the failure.
+ */
+ WARN_ONCE(ret, "HLT instruction emulation failed\n");
+}
+
+static __cpuidle void tdx_halt(void)
+{
+ const bool irq_disabled = irqs_disabled();
+ const bool do_sti = false;
+
+ /*
+ * Non safe halt is mainly used in CPU off-lining
+ * and the guest will stay in halt state. So,
+ * STI instruction call is not required (set
+ * do_sti as false).
+ */
+ _tdx_halt(irq_disabled, do_sti);
+}
+
+static __cpuidle void tdx_safe_halt(void)
+{
+ /*
+ * Since STI instruction will be called in __tdx_hypercall()
+ * set irq_disabled as false.
+ */
+ const bool irq_disabled = false;
+ const bool do_sti = true;
+
+ _tdx_halt(irq_disabled, do_sti);
+}
+
unsigned long tdx_get_ve_info(struct ve_info *ve)
{
struct tdx_module_output out = {0};
@@ -79,8 +136,19 @@ unsigned long tdx_get_ve_info(struct ve_info *ve)
int tdx_handle_virtualization_exception(struct pt_regs *regs,
struct ve_info *ve)
{
- pr_warn("Unexpected #VE: %lld\n", ve->exit_reason);
- return -EFAULT;
+ switch (ve->exit_reason) {
+ case EXIT_REASON_HLT:
+ tdx_halt();
+ break;
+ default:
+ pr_warn("Unexpected #VE: %lld\n", ve->exit_reason);
+ return -EFAULT;
+ }
+
+ /* After successful #VE handling, move the IP */
+ regs->ip += ve->instr_len;
+
+ return 0;
}
void __init tdx_early_init(void)
@@ -92,5 +160,8 @@ void __init tdx_early_init(void)
setup_force_cpu_cap(X86_FEATURE_TDX_GUEST);
+ pv_ops.irq.safe_halt = tdx_safe_halt;
+ pv_ops.irq.halt = tdx_halt;
+
pr_info("Guest initialized\n");
}
--
2.25.1
next prev parent reply other threads:[~2021-10-05 2:52 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-05 2:51 [PATCH v8 00/11] Add TDX Guest Support (Initial support) Kuppuswamy Sathyanarayanan
2021-10-05 2:51 ` [PATCH v8 01/11] x86/paravirt: Move halt paravirt calls under CONFIG_PARAVIRT Kuppuswamy Sathyanarayanan
2021-10-05 20:13 ` Josh Poimboeuf
2021-10-05 2:51 ` [PATCH v8 02/11] x86/tdx: Introduce INTEL_TDX_GUEST config option Kuppuswamy Sathyanarayanan
2021-10-05 4:53 ` Randy Dunlap
2021-10-05 13:29 ` Sathyanarayanan Kuppuswamy Natarajan
2021-10-05 14:09 ` Dave Hansen
2021-10-05 14:31 ` Sean Christopherson
2021-10-05 14:43 ` Kuppuswamy, Sathyanarayanan
2021-10-05 14:13 ` Borislav Petkov
2021-10-05 14:48 ` Kuppuswamy, Sathyanarayanan
2021-10-05 17:29 ` Borislav Petkov
2021-10-05 20:21 ` Josh Poimboeuf
2021-10-05 20:38 ` Kuppuswamy, Sathyanarayanan
2021-10-05 20:17 ` Josh Poimboeuf
2021-10-05 20:33 ` Sean Christopherson
2021-10-05 20:42 ` Dave Hansen
2021-10-05 20:37 ` Kuppuswamy, Sathyanarayanan
2021-10-05 2:51 ` [PATCH v8 03/11] x86/cpufeatures: Add TDX Guest CPU feature Kuppuswamy Sathyanarayanan
2021-10-05 21:04 ` Josh Poimboeuf
2021-10-05 21:19 ` Borislav Petkov
2021-10-05 21:41 ` Kuppuswamy, Sathyanarayanan
2021-10-06 3:42 ` Josh Poimboeuf
2021-10-06 4:33 ` Kuppuswamy, Sathyanarayanan
2021-10-06 5:03 ` Josh Poimboeuf
2021-10-06 12:47 ` Borislav Petkov
2021-10-06 14:11 ` Josh Poimboeuf
2021-10-06 14:26 ` Borislav Petkov
2021-10-06 14:25 ` Josh Poimboeuf
2021-10-06 15:26 ` Borislav Petkov
2021-10-06 15:43 ` Kuppuswamy, Sathyanarayanan
2021-10-06 16:20 ` Borislav Petkov
2021-10-05 2:51 ` [PATCH v8 04/11] x86/tdx: Add Intel ARCH support to cc_platform_has() Kuppuswamy Sathyanarayanan
2021-10-05 4:47 ` Randy Dunlap
2021-10-05 12:29 ` Kuppuswamy, Sathyanarayanan
2021-10-05 21:16 ` Josh Poimboeuf
2021-10-05 21:42 ` Kuppuswamy, Sathyanarayanan
2021-10-06 18:02 ` Borislav Petkov
2021-10-06 18:14 ` Kuppuswamy, Sathyanarayanan
2021-10-05 2:51 ` [PATCH v8 05/11] x86/tdx: Add __tdx_module_call() and __tdx_hypercall() helper functions Kuppuswamy Sathyanarayanan
2021-10-06 5:53 ` Josh Poimboeuf
2021-10-06 16:52 ` Kuppuswamy, Sathyanarayanan
2021-10-07 9:33 ` Borislav Petkov
2021-10-07 16:55 ` Kuppuswamy, Sathyanarayanan
2021-10-05 2:52 ` [PATCH v8 06/11] x86/traps: Add #VE support for TDX guest Kuppuswamy Sathyanarayanan
2021-10-06 18:40 ` Josh Poimboeuf
2021-10-07 17:06 ` Borislav Petkov
2021-10-07 17:22 ` Kuppuswamy, Sathyanarayanan
2021-10-07 17:32 ` Borislav Petkov
2021-10-17 17:15 ` Dave Hansen
2021-10-18 10:53 ` Borislav Petkov
2021-10-18 14:05 ` Dave Hansen
2021-10-18 14:09 ` Borislav Petkov
2021-10-09 3:56 ` Lai Jiangshan
2021-10-11 15:06 ` Sean Christopherson
2021-10-11 16:49 ` Andi Kleen
2021-10-05 2:52 ` Kuppuswamy Sathyanarayanan [this message]
2021-10-06 19:17 ` [PATCH v8 07/11] x86/tdx: Add HLT " Josh Poimboeuf
2021-10-07 19:25 ` Kuppuswamy, Sathyanarayanan
2021-10-08 17:31 ` Borislav Petkov
2021-10-08 17:38 ` Kuppuswamy, Sathyanarayanan
2021-10-08 17:59 ` Borislav Petkov
2021-10-05 2:52 ` [PATCH v8 08/11] x86/tdx: Wire up KVM hypercalls Kuppuswamy Sathyanarayanan
2021-10-06 19:34 ` Josh Poimboeuf
2021-10-06 19:40 ` Borislav Petkov
2021-11-05 20:59 ` Sean Christopherson
2021-11-12 16:17 ` Sathyanarayanan Kuppuswamy
2021-10-05 2:52 ` [PATCH v8 09/11] x86/tdx: Add MSR support for TDX guest Kuppuswamy Sathyanarayanan
2021-10-05 23:22 ` Josh Poimboeuf
2021-10-06 0:48 ` Kuppuswamy, Sathyanarayanan
2021-10-06 19:49 ` Josh Poimboeuf
2021-10-08 2:16 ` Kuppuswamy, Sathyanarayanan
2021-10-05 2:52 ` [PATCH v8 10/11] x86/tdx: Don't write CSTAR MSR on Intel Kuppuswamy Sathyanarayanan
2021-10-05 2:52 ` [PATCH v8 11/11] x86/tdx: Handle CPUID via #VE Kuppuswamy Sathyanarayanan
2021-10-06 20:26 ` Josh Poimboeuf
2021-10-08 2:25 ` Kuppuswamy, Sathyanarayanan
2021-10-11 18:16 ` Josh Poimboeuf
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=20211005025205.1784480-8-sathyanarayanan.kuppuswamy@linux.intel.com \
--to=sathyanarayanan.kuppuswamy@linux.intel.com \
--cc=aarcange@redhat.com \
--cc=ak@linux.intel.com \
--cc=bp@alien8.de \
--cc=dan.j.williams@intel.com \
--cc=dave.hansen@intel.com \
--cc=david@redhat.com \
--cc=hpa@zytor.com \
--cc=jgross@suse.com \
--cc=jmattson@google.com \
--cc=joro@8bytes.org \
--cc=jpoimboe@redhat.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=knsathya@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=pbonzini@redhat.com \
--cc=pv-drivers@vmware.com \
--cc=sdeep@vmware.com \
--cc=seanjc@google.com \
--cc=tglx@linutronix.de \
--cc=tony.luck@intel.com \
--cc=vkuznets@redhat.com \
--cc=wanpengli@tencent.com \
--cc=x86@kernel.org \
/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).