Re: [PATCH v1 1/3] x86/tdx: Add TDX Guest event notify interrupt support

On Sat, 2023-03-25 at 23:20 -0700, Kuppuswamy Sathyanarayanan wrote:

Host-guest event notification via configured interrupt vector is useful in cases where a guest makes an asynchronous request and needs a callback from the host to indicate the completion or to let the host notify the guest about events like device removal. One usage example is, callback requirement of GetQuote asynchronous hypercall.

In TDX guest, SetupEventNotifyInterrupt hypercall can be used by the guest to specify which interrupt vector to use as an event-notify vector to the VMM.

"to the VMM" -> "from the VMM"?

Details about the SetupEventNotifyInterrupt hypercall can be found in TDX Guest-Host Communication Interface (GHCI) Specification, sec 3.5 "VP.VMCALL<SetupEventNotifyInterrupt>".

It seems we shouldn't mention the exact section number.

As per design, VMM will post the event completion IRQ using the same CPU in which SetupEventNotifyInterrupt hypercall request is received.

"in which" -> "on which"

So allocate an IRQ vector from "x86_vector_domain", and set the CPU affinity of the IRQ vector to the current CPU.

IMHO "current CPU" is a little bit vague. Perhaps just "to the CPU on which SetupEventNotifyInterrupt hypercall is made".

Also, perhaps it's better to mention to use IRQF_NOBALANCING to prevent the IRQ from being migrated to another cpu.

Add tdx_register_event_irq_cb()/tdx_unregister_event_irq_cb() interfaces to allow drivers register/unregister event noficiation handlers.

Reviewed-by: Tony Luck tony.luck@intel.com Reviewed-by: Andi Kleen ak@linux.intel.com Reviewed-by: Mika Westerberg mika.westerberg@linux.intel.com Acked-by: Kirill A. Shutemov kirill.shutemov@linux.intel.com Acked-by: Wander Lairson Costa wander@redhat.com Signed-off-by: Kuppuswamy Sathyanarayanan sathyanarayanan.kuppuswamy@linux.intel.com

---

arch/x86/coco/tdx/tdx.c | 163 +++++++++++++++++++++++++++++++++++++ arch/x86/include/asm/tdx.h | 6 ++ 2 files changed, 169 insertions(+)

diff --git a/arch/x86/coco/tdx/tdx.c b/arch/x86/coco/tdx/tdx.c index 055300e08fb3..d03985952d45 100644 --- a/arch/x86/coco/tdx/tdx.c +++ b/arch/x86/coco/tdx/tdx.c @@ -7,12 +7,18 @@ #include <linux/cpufeature.h> #include <linux/export.h> #include <linux/io.h> +#include <linux/string.h> +#include <linux/uaccess.h>

Do you need above two headers?

Also, perhaps you should explicitly include <.../list.h> and <.../spinlock.h>.

+#include <linux/interrupt.h> +#include <linux/irq.h> +#include <linux/numa.h> #include <asm/coco.h> #include <asm/tdx.h> #include <asm/vmx.h> #include <asm/insn.h> #include <asm/insn-eval.h> #include <asm/pgtable.h> +#include <asm/irqdomain.h> /* TDX module Call Leaf IDs */ #define TDX_GET_INFO 1 @@ -27,6 +33,7 @@ /* TDX hypercall Leaf IDs */ #define TDVMCALL_MAP_GPA 0x10001 #define TDVMCALL_REPORT_FATAL_ERROR 0x10003 +#define TDVMCALL_SETUP_NOTIFY_INTR 0x10004 /* MMIO direction */ #define EPT_READ 0 @@ -51,6 +58,16 @@ #define TDREPORT_SUBTYPE_0 0 +struct event_irq_entry {

• tdx_event_irq_cb_t handler;
• void *data;
• struct list_head head;

+};

+static int tdx_event_irq;

__ro_after_init?

+static LIST_HEAD(event_irq_cb_list); +static DEFINE_SPINLOCK(event_irq_cb_lock);

/*

• Wrapper for standard use of __tdx_hypercall with no output aside from
• return code.

@@ -873,3 +890,149 @@ void __init tdx_early_init(void) pr_info("Guest detected\n"); }

+static irqreturn_t tdx_event_irq_handler(int irq, void *dev_id) +{

• struct event_irq_entry *entry;
• spin_lock(&event_irq_cb_lock);
• list_for_each_entry(entry, &event_irq_cb_list, head) {

• if (entry->handler)
  

• 	entry->handler(entry->data);
  

• }
• spin_unlock(&event_irq_cb_lock);
• return IRQ_HANDLED;

+}

+/* Reserve an IRQ from x86_vector_domain for TD event notification */ +static int __init tdx_event_irq_init(void) +{

• struct irq_alloc_info info;
• cpumask_t saved_cpumask;
• struct irq_cfg *cfg;
• int cpu, irq;
• if (!cpu_feature_enabled(X86_FEATURE_TDX_GUEST))

• return 0;
  

• init_irq_alloc_info(&info, NULL);
• /*

• * Event notification vector will be delivered to the CPU
  

• * in which TDVMCALL_SETUP_NOTIFY_INTR hypercall is requested.
  

• * So set the IRQ affinity to the current CPU.
  

• */
  

• cpu = get_cpu();
• cpumask_copy(&saved_cpumask, current->cpus_ptr);
• info.mask = cpumask_of(cpu);
• put_cpu();

The 'saved_cpumask' related code is ugly. If you move put_cpu() to the end of this function, I think you can remove all related code:

cpu = get_cpu();

/* * Set @info->mask to local cpu to make sure a valid vector is * pre-allocated when TDX event notification IRQ is allocated * from x86_vector_domain. */ init_irq_alloc_info(&info, cpumask_of(cpu));

// rest staff: request_irq(), hypercall ...

put_cpu();

• irq = irq_domain_alloc_irqs(x86_vector_domain, 1, NUMA_NO_NODE, &info);

Should you use cpu_to_node(cpu) instead of NUMA_NO_NODE?

• if (irq <= 0) {

• pr_err("Event notification IRQ allocation failed %d\n", irq);
  

• return -EIO;
  

• }
• irq_set_handler(irq, handle_edge_irq);
• cfg = irq_cfg(irq);
• if (!cfg) {

• pr_err("Event notification IRQ config not found\n");
  

• goto err_free_irqs;
  

• }

You are depending on irq_domain_alloc_irqs() to have already allocated a vector on the local cpu. Then if !cfg, your code of calling irq_domain_alloc_irqs() to allocate vector is broken.

So, perhaps you should just WARN() if vector hasn't been allocated to catch bug.

WARN(!(irq_cfg(irq)->vector));

• if (request_irq(irq, tdx_event_irq_handler, IRQF_NOBALANCING,

It's better to add a comment to explain why using IRQF_NOBALANCING.

• 	"tdx_event_irq", NULL)) {
  

• pr_err("Event notification IRQ request failed\n");
  

• goto err_free_irqs;
  

• }
• set_cpus_allowed_ptr(current, cpumask_of(cpu));
• /*

• * Register callback vector address with VMM. More details
  

• * about the ABI can be found in TDX Guest-Host-Communication
  

• * Interface (GHCI), sec titled
  

• * "TDG.VP.VMCALL<SetupEventNotifyInterrupt>".
  

• */
  

• if (_tdx_hypercall(TDVMCALL_SETUP_NOTIFY_INTR, cfg->vector, 0, 0, 0)) {

• pr_err("Event notification hypercall failed\n");
  

• goto err_restore_cpus;
  

• }
• set_cpus_allowed_ptr(current, &saved_cpumask);
• tdx_event_irq = irq;
• return 0;

+err_restore_cpus:

• set_cpus_allowed_ptr(current, &saved_cpumask);
• free_irq(irq, NULL);

+err_free_irqs:

• irq_domain_free_irqs(irq, 1);
• return -EIO;

+} +arch_initcall(tdx_event_irq_init)