test: parallel vmexit

Index: qemu-kvm/kvm/user/test/lib/x86/smp.c
===================================================================
@@ -82,7 +82,7 @@ static void setup_smp_id(void *data)
     asm ("mov %0, %%gs:0" : : "r"(apic_id()) : "memory");
 }
 
-void on_cpu(int cpu, void (*function)(void *data), void *data)
+void on_cpu(int cpu, void (*function)(void *data), void *data, int wait)
 {
     spin_lock(&ipi_lock);
     if (cpu == smp_id())

@@ -93,9 +93,11 @@ void on_cpu(int cpu, void (*function)(vo
 	apic_icr_write(APIC_INT_ASSERT | APIC_DEST_PHYSICAL | APIC_DM_FIXED
                        | IPI_VECTOR,
                        cpu);
-	while (!ipi_done)
-	    ;
-	ipi_done = 0;
+	if (wait) {
+		while (!ipi_done)
+		    ;
+		ipi_done = 0;
+	}
     }
     spin_unlock(&ipi_lock);
 }

@@ -109,6 +111,6 @@ void smp_init(void)
 
     setup_smp_id(0);
     for (i = 1; i < cpu_count(); ++i)
-        on_cpu(i, setup_smp_id, 0);
+        on_cpu(i, setup_smp_id, 0, 1);
 
 }

Index: qemu-kvm/kvm/user/test/lib/x86/smp.h
===================================================================
@@ -9,7 +9,7 @@ void smp_init(void);
 
 int cpu_count(void);
 int smp_id(void);
-void on_cpu(int cpu, void (*function)(void *data), void *data);
+void on_cpu(int cpu, void (*function)(void *data), void *data, int wait);
 void spin_lock(struct spinlock *lock);
 void spin_unlock(struct spinlock *lock);
 

Index: qemu-kvm/kvm/user/test/x86/smptest.c
===================================================================
@@ -20,6 +20,6 @@ int main()
     ncpus = cpu_count();
     printf("found %d cpus\n", ncpus);
     for (i = 0; i < ncpus; ++i)
-	on_cpu(i, ipi_test, (void *)(long)i);
+	on_cpu(i, ipi_test, (void *)(long)i, 1);
     return 0;
 }

Index: qemu-kvm/kvm/user/test/x86/vmexit.c
===================================================================
@@ -63,14 +63,14 @@ static void nop(void *junk)
 
 static void ipi(void)
 {
-	on_cpu(1, nop, 0);
+	on_cpu(1, nop, 0, 1);
 }
 
 static void ipi_halt(void)
 {
 	unsigned long long t;
 
-	on_cpu(1, nop, 0);
+	on_cpu(1, nop, 0, 1);
 	t = rdtsc() + 2000;
 	while (rdtsc() < t)
 		;

Index: qemu-kvm-test/kvm/user/test/x86/vmexit.c
===================================================================
@@ -85,6 +85,9 @@ static struct test {
 	{ vmcall, "vmcall", },
 	{ mov_from_cr8, "mov_from_cr8" },
 	{ mov_to_cr8, "mov_to_cr8" },
+};
+
+static struct test smp_tests[] = {
 	{ ipi, "ipi", is_smp },
 	{ ipi_halt, "ipi+halt", is_smp },
 };

@@ -113,14 +116,25 @@ static void do_test(struct test *test)
 
 #define ARRAY_SIZE(_x) (sizeof(_x) / sizeof((_x)[0]))
 
-int main(void)
+static void do_tests(void *junk)
 {
 	int i;
 
-	smp_init();
-
 	for (i = 0; i < ARRAY_SIZE(tests); ++i)
 		do_test(&tests[i]);
+}
+
+int main(void)
+{
+	int i, n;
+
+	smp_init();
+
+	for (n = cpu_count(); n > 0; n--)
+		on_cpu(n-1, do_tests, 0, 0);
+
+	for (i = 0; i < ARRAY_SIZE(smp_tests); ++i)
+		do_test(&smp_tests[i]);
 
 	return 0;
 }

Index: qemu-kvm-parallel-vmexit/kvm/user/test/lib/x86/smp.c
===================================================================
@@ -82,24 +82,38 @@ static void setup_smp_id(void *data)
     asm ("mov %0, %%gs:0" : : "r"(apic_id()) : "memory");
 }
 
-void on_cpu(int cpu, void (*function)(void *data), void *data)
+static void __on_cpu(int cpu, void (*function)(void *data), void *data,
+                     int wait)
 {
     spin_lock(&ipi_lock);
     if (cpu == smp_id())
 	function(data);
     else {
+	ipi_done = 0;
 	ipi_function = function;
 	ipi_data = data;
 	apic_icr_write(APIC_INT_ASSERT | APIC_DEST_PHYSICAL | APIC_DM_FIXED
                        | IPI_VECTOR,
                        cpu);
-	while (!ipi_done)
-	    ;
-	ipi_done = 0;
+	if (wait) {
+		while (!ipi_done)
+		    ;
+	}
     }
     spin_unlock(&ipi_lock);
 }
 
+void on_cpu(int cpu, void (*function)(void *data), void *data)
+{
+    __on_cpu(cpu, function, data, 1);
+}
+
+void on_cpu_async(int cpu, void (*function)(void *data), void *data)
+{
+    __on_cpu(cpu, function, data, 0);
+}
+
+
 void smp_init(void)
 {
     int i;

Index: qemu-kvm-parallel-vmexit/kvm/user/test/lib/x86/smp.h
===================================================================
@@ -10,6 +10,7 @@ void smp_init(void);
 int cpu_count(void);
 int smp_id(void);
 void on_cpu(int cpu, void (*function)(void *data), void *data);
+void on_cpu_async(int cpu, void (*function)(void *data), void *data);
 void spin_lock(struct spinlock *lock);
 void spin_unlock(struct spinlock *lock);
 

Index: qemu-kvm-parallel-vmexit/kvm/user/test/x86/vmexit.c
===================================================================
@@ -80,22 +80,38 @@ static struct test {
 	void (*func)(void);
 	const char *name;
 	int (*valid)(void);
+	int parallel;
 } tests[] = {
-	{ cpuid, "cpuid", },
-	{ vmcall, "vmcall", },
-	{ mov_from_cr8, "mov_from_cr8" },
-	{ mov_to_cr8, "mov_to_cr8" },
-	{ ipi, "ipi", is_smp },
-	{ ipi_halt, "ipi+halt", is_smp },
+	{ cpuid, "cpuid", .parallel = 1,  },
+	{ vmcall, "vmcall", .parallel = 1, },
+	{ mov_from_cr8, "mov_from_cr8", .parallel = 1, },
+	{ mov_to_cr8, "mov_to_cr8" , .parallel = 1, },
+	{ ipi, "ipi", is_smp, .parallel = 0, },
+	{ ipi_halt, "ipi+halt", is_smp, .parallel = 0, },
 };
 
+unsigned iterations;
+volatile int nr_cpus_done;
+
+static void run_test(void *_func)
+{
+    int i;
+    void (*func)(void) = _func;
+
+    for (i = 0; i < iterations; ++i)
+        func();
+
+    nr_cpus_done++;
+}
+
 static void do_test(struct test *test)
 {
 	int i;
 	unsigned long long t1, t2;
-	unsigned iterations = 32;
         void (*func)(void) = test->func;
 
+        iterations = 32;
+
         if (test->valid && !test->valid()) {
 		printf("%s (skipped)\n", test->name);
 		return;

@@ -104,8 +120,17 @@ static void do_test(struct test *test)
 	do {
 		iterations *= 2;
 		t1 = rdtsc();
-		for (i = 0; i < iterations; ++i)
-			func();
+
+		if (!test->parallel) {
+			for (i = 0; i < iterations; ++i)
+				func();
+		} else {
+			nr_cpus_done = 0;
+			for (i = cpu_count(); i > 0; i--)
+				on_cpu_async(i-1, run_test, func);
+			while (nr_cpus_done < cpu_count())
+				;
+		}
 		t2 = rdtsc();
 	} while ((t2 - t1) < GOAL);
 	printf("%s %d\n", test->name, (int)((t2 - t1) / iterations));

Index: qemu-kvm-parallel-vmexit/kvm/user/test/x86/vmexit.c
===================================================================
@@ -17,6 +17,13 @@ static inline unsigned long long rdtsc()
 	return r;
 }
 
+static unsigned int inl(unsigned short port)
+{
+    unsigned int val;
+    asm volatile("inl %w1, %0" : "=a"(val) : "Nd"(port));
+    return val;
+}
+
 #define GOAL (1ull << 30)
 
 #ifdef __x86_64__

@@ -76,6 +83,11 @@ static void ipi_halt(void)
 		;
 }
 
+static void inl_pmtimer(void)
+{
+    inl(0xb008);
+}
+
 static struct test {
 	void (*func)(void);
 	const char *name;

@@ -86,6 +98,7 @@ static struct test {
 	{ vmcall, "vmcall", .parallel = 1, },
 	{ mov_from_cr8, "mov_from_cr8", .parallel = 1, },
 	{ mov_to_cr8, "mov_to_cr8" , .parallel = 1, },
+	{ inl_pmtimer, "inl_from_pmtimer", .parallel = 1, },
 	{ ipi, "ipi", is_smp, .parallel = 0, },
 	{ ipi_halt, "ipi+halt", is_smp, .parallel = 0, },
 };