X86/Hyper-V: Enable IPI enlightenments

Hyper-V supports hypercalls to implement IPI; use them.

Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
Cc: olaf@aepfle.de
Cc: sthemmin@microsoft.com
Cc: gregkh@linuxfoundation.org
Cc: jasowang@redhat.com
Cc: Michael.H.Kelley@microsoft.com
Cc: hpa@zytor.com
Cc: apw@canonical.com
Cc: devel@linuxdriverproject.org
Cc: vkuznets@redhat.com
Link: https://lkml.kernel.org/r/20180516215334.6547-2-kys@linuxonhyperv.com

4 files changed, 160 insertions, 0 deletions

diff --git a/arch/x86/hyperv/hv_apic.c b/arch/x86/hyperv/hv_apic.c
index ca20e31d311c..3e0de61f1a7c 100644
--- a/arch/x86/hyperv/hv_apic.c
+++ b/arch/x86/hyperv/hv_apic.c
@@ -33,6 +33,8 @@
 #ifdef CONFIG_X86_64
 #if IS_ENABLED(CONFIG_HYPERV)
 
+static struct apic orig_apic;
+
 static u64 hv_apic_icr_read(void)
 {
         u64 reg_val;
@@ -88,8 +90,123 @@ static void hv_apic_eoi_write(u32 reg, u32 val)
         wrmsr(HV_X64_MSR_EOI, val, 0);
 }
 
+/*
+ * IPI implementation on Hyper-V.
+ */
+static bool __send_ipi_mask(const struct cpumask *mask, int vector)
+{
+        int cur_cpu, vcpu;
+        struct ipi_arg_non_ex **arg;
+        struct ipi_arg_non_ex *ipi_arg;
+        int ret = 1;
+        unsigned long flags;
+
+        if (cpumask_empty(mask))
+                return true;
+
+        if (!hv_hypercall_pg)
+                return false;
+
+        if ((vector < HV_IPI_LOW_VECTOR) || (vector > HV_IPI_HIGH_VECTOR))
+                return false;
+
+        local_irq_save(flags);
+        arg = (struct ipi_arg_non_ex **)this_cpu_ptr(hyperv_pcpu_input_arg);
+
+        ipi_arg = *arg;
+        if (unlikely(!ipi_arg))
+                goto ipi_mask_done;
+
+        ipi_arg->vector = vector;
+        ipi_arg->reserved = 0;
+        ipi_arg->cpu_mask = 0;
+
+        for_each_cpu(cur_cpu, mask) {
+                vcpu = hv_cpu_number_to_vp_number(cur_cpu);
+                /*
+                 * This particular version of the IPI hypercall can
+                 * only target upto 64 CPUs.
+                 */
+                if (vcpu >= 64)
+                        goto ipi_mask_done;
+
+                __set_bit(vcpu, (unsigned long *)&ipi_arg->cpu_mask);
+        }
+
+        ret = hv_do_hypercall(HVCALL_SEND_IPI, ipi_arg, NULL);
+
+ipi_mask_done:
+        local_irq_restore(flags);
+        return ((ret == 0) ? true : false);
+}
+
+static bool __send_ipi_one(int cpu, int vector)
+{
+        struct cpumask mask = CPU_MASK_NONE;
+
+        cpumask_set_cpu(cpu, &mask);
+        return __send_ipi_mask(&mask, vector);
+}
+
+static void hv_send_ipi(int cpu, int vector)
+{
+        if (!__send_ipi_one(cpu, vector))
+                orig_apic.send_IPI(cpu, vector);
+}
+
+static void hv_send_ipi_mask(const struct cpumask *mask, int vector)
+{
+        if (!__send_ipi_mask(mask, vector))
+                orig_apic.send_IPI_mask(mask, vector);
+}
+
+static void hv_send_ipi_mask_allbutself(const struct cpumask *mask, int vector)
+{
+        unsigned int this_cpu = smp_processor_id();
+        struct cpumask new_mask;
+        const struct cpumask *local_mask;
+
+        cpumask_copy(&new_mask, mask);
+        cpumask_clear_cpu(this_cpu, &new_mask);
+        local_mask = &new_mask;
+        if (!__send_ipi_mask(local_mask, vector))
+                orig_apic.send_IPI_mask_allbutself(mask, vector);
+}
+
+static void hv_send_ipi_allbutself(int vector)
+{
+        hv_send_ipi_mask_allbutself(cpu_online_mask, vector);
+}
+
+static void hv_send_ipi_all(int vector)
+{
+        if (!__send_ipi_mask(cpu_online_mask, vector))
+                orig_apic.send_IPI_all(vector);
+}
+
+static void hv_send_ipi_self(int vector)
+{
+        if (!__send_ipi_one(smp_processor_id(), vector))
+                orig_apic.send_IPI_self(vector);
+}
+
 void __init hv_apic_init(void)
 {
+        if (ms_hyperv.hints & HV_X64_CLUSTER_IPI_RECOMMENDED) {
+                pr_info("Hyper-V: Using IPI hypercalls\n");
+                /*
+                 * Set the IPI entry points.
+                 */
+                orig_apic = *apic;
+
+                apic->send_IPI = hv_send_ipi;
+                apic->send_IPI_mask = hv_send_ipi_mask;
+                apic->send_IPI_mask_allbutself = hv_send_ipi_mask_allbutself;
+                apic->send_IPI_allbutself = hv_send_ipi_allbutself;
+                apic->send_IPI_all = hv_send_ipi_all;
+                apic->send_IPI_self = hv_send_ipi_self;
+        }
+
         if (ms_hyperv.hints & HV_X64_APIC_ACCESS_RECOMMENDED) {
                 pr_info("Hyper-V: Using MSR based APIC access\n");
                 apic_set_eoi_write(hv_apic_eoi_write);
diff --git a/arch/x86/hyperv/hv_init.c b/arch/x86/hyperv/hv_init.c
index 71e50fc2b7ef..6bc90d68ac8b 100644
--- a/arch/x86/hyperv/hv_init.c
+++ b/arch/x86/hyperv/hv_init.c
@@ -91,12 +91,19 @@ EXPORT_SYMBOL_GPL(hv_vp_index);
 struct hv_vp_assist_page **hv_vp_assist_page;
 EXPORT_SYMBOL_GPL(hv_vp_assist_page);
 
+void  __percpu **hyperv_pcpu_input_arg;
+EXPORT_SYMBOL_GPL(hyperv_pcpu_input_arg);
+
 u32 hv_max_vp_index;
 
 static int hv_cpu_init(unsigned int cpu)
 {
         u64 msr_vp_index;
         struct hv_vp_assist_page **hvp = &hv_vp_assist_page[smp_processor_id()];
+        void **input_arg;
+
+        input_arg = (void **)this_cpu_ptr(hyperv_pcpu_input_arg);
+        *input_arg = page_address(alloc_page(GFP_KERNEL));
 
         hv_get_vp_index(msr_vp_index);
 
@@ -217,6 +224,16 @@ static int hv_cpu_die(unsigned int cpu)
 {
         struct hv_reenlightenment_control re_ctrl;
         unsigned int new_cpu;
+        unsigned long flags;
+        void **input_arg;
+        void *input_pg = NULL;
+
+        local_irq_save(flags);
+        input_arg = (void **)this_cpu_ptr(hyperv_pcpu_input_arg);
+        input_pg = *input_arg;
+        *input_arg = NULL;
+        local_irq_restore(flags);
+        free_page((unsigned long)input_pg);
 
         if (hv_vp_assist_page && hv_vp_assist_page[cpu])
                 wrmsrl(HV_X64_MSR_VP_ASSIST_PAGE, 0);
@@ -260,6 +277,16 @@ void __init hyperv_init(void)
         if ((ms_hyperv.features & required_msrs) != required_msrs)
                 return;
 
+        /*
+         * Allocate the per-CPU state for the hypercall input arg.
+         * If this allocation fails, we will not be able to setup
+         * (per-CPU) hypercall input page and thus this failure is
+         * fatal on Hyper-V.
+         */
+        hyperv_pcpu_input_arg = alloc_percpu(void  *);
+
+        BUG_ON(hyperv_pcpu_input_arg == NULL);
+
         /* Allocate percpu VP index */
         hv_vp_index = kmalloc_array(num_possible_cpus(), sizeof(*hv_vp_index),
                                     GFP_KERNEL);
diff --git a/arch/x86/include/asm/hyperv-tlfs.h b/arch/x86/include/asm/hyperv-tlfs.h
index 416cb0e0c496..332e786d4deb 100644
--- a/arch/x86/include/asm/hyperv-tlfs.h
+++ b/arch/x86/include/asm/hyperv-tlfs.h
@@ -164,6 +164,11 @@
  */
 #define HV_X64_DEPRECATING_AEOI_RECOMMENDED     (1 << 9)
 
+/*
+ * Recommend using cluster IPI hypercalls.
+ */
+#define HV_X64_CLUSTER_IPI_RECOMMENDED         (1 << 10)
+
 /* Recommend using the newer ExProcessorMasks interface */
 #define HV_X64_EX_PROCESSOR_MASKS_RECOMMENDED   (1 << 11)
 
@@ -329,10 +334,14 @@ struct hv_tsc_emulation_status {
 #define HV_X64_MSR_HYPERCALL_PAGE_ADDRESS_MASK  \
                 (~((1ull << HV_X64_MSR_HYPERCALL_PAGE_ADDRESS_SHIFT) - 1))
 
+#define HV_IPI_LOW_VECTOR       0x10
+#define HV_IPI_HIGH_VECTOR      0xff
+
 /* Declare the various hypercall operations. */
 #define HVCALL_FLUSH_VIRTUAL_ADDRESS_SPACE      0x0002
 #define HVCALL_FLUSH_VIRTUAL_ADDRESS_LIST       0x0003
 #define HVCALL_NOTIFY_LONG_SPIN_WAIT            0x0008
+#define HVCALL_SEND_IPI                         0x000b
 #define HVCALL_FLUSH_VIRTUAL_ADDRESS_SPACE_EX  0x0013
 #define HVCALL_FLUSH_VIRTUAL_ADDRESS_LIST_EX   0x0014
 #define HVCALL_POST_MESSAGE                     0x005c
@@ -706,4 +715,10 @@ struct hv_enlightened_vmcs {
 #define HV_STIMER_AUTOENABLE            (1ULL << 3)
 #define HV_STIMER_SINT(config)          (__u8)(((config) >> 16) & 0x0F)
 
+struct ipi_arg_non_ex {
+        u32 vector;
+        u32 reserved;
+        u64 cpu_mask;
+};
+
 #endif
diff --git a/arch/x86/include/asm/mshyperv.h b/arch/x86/include/asm/mshyperv.h
index 162977b82e2e..1eff91599c2b 100644
--- a/arch/x86/include/asm/mshyperv.h
+++ b/arch/x86/include/asm/mshyperv.h
@@ -122,6 +122,7 @@ static inline void hv_disable_stimer0_percpu_irq(int irq) {}
 #if IS_ENABLED(CONFIG_HYPERV)
 extern struct clocksource *hyperv_cs;
 extern void *hv_hypercall_pg;
+extern void  __percpu  **hyperv_pcpu_input_arg;
 
 static inline u64 hv_do_hypercall(u64 control, void *input, void *output)
 {